Resensitized là gì
Page 2
demonstrated effective elimination of primary tumors and disseminated metastases of FBL-3 tumors in B6 mice by a single iv injection of immune lymphocytes. These lymphocytes were resensitized in vitro to the FBL-3 tumor and expanded in Interleukin-2. To make our approach more applicable to the clinical : situation in cancer patients, we investigated the efficacy of adoptive transfer of fetal antigen-sensitized spleen cells on preestablished metastases. The primary tumor was allowed to grow for several days. After it had metastasized to the lungs, it was amputated; then the spleen cell transfers were started. Adoptive transfer of fetal antigen-sensitized spleen cells significantly inhibited metastases in mice in which the primary tumors were amputated. However, adoptive transfer of fetal antigen-sensitized spleen cells did not inhibit tumor growth or metastases in mice in which primary tumors were left intact. Another observation made was that the metastatic index in amputated control animals was significantly higher than that in the unamputated control group. It could be anticipated that surgery alone or adoptive transfer of immune cells alone is not beneficial in such tumors as this one. Surgery and repeated inoculation of fetal immune cells seem to improve antimetastatic therapy to a significant level in this system. These findings on the role of oncofetal antigens in inhibiting tumor growth and metastases of a weakly immunogenic, metastatic tumor have direct relevance to human cancer, because the most commonly associated features of human cancers are their weak immunogenicity, their spontaneous metastatic behavior, and the need for therapy after the tumor has progressed beyond a certain stage.
(13) Cantor H, Boyse EA. Functional subclasses of T lymphocytes bearing different Ly antigens. I. The generation of functionally distinct T cell subclass is a differentiative process independent of antigen. J Exp Med 1975; 141:1376-1389. (14) Kisielow P, Hirst JA, Shiku H, et al. Ly antigens as markers for functionally distinct subpopulations of thymus-derived lympho cytes of the mouse. Nature 1975; 253:219-220. (15) JADINSKI J, Cantor H, TADAKUMA T, Peavy DL, Pierce CW. Sepa ration of helper T cells from suppressor T cells expressing different Ly components. I. Polyclonal activation: Suppressor and helper activities are inherent properties of distinct T cell subsets. J Exp Med 1976; 143:1382-1390. (16) HUBER B, DEVINSKY O, GERSHON RK, CANTOR H. Cell mediated immunity: Delayed type hypersensitivity and cytotoxic responses are mediated by different T cell subclasses. J Exp Med 1976; 143:1534-1539. (17) VADAS MA, MILLER JF, MCKENZIE IF, et al. Ly and la antigen phenotypes of T cells involved in delayed type hypersensitivity and in suppression. J Exp Med 1976; 144:10-19. (18) BEVERLEY PC, Woody J, DUNKLEY M, FELDMAN M, McKenzie I. Separation of suppressor and killer T cells by surface phenotypes. Nature 1976; 262:495-497. (19) SHEN FW, McDougaL JS, BARD J, CORT SP. Developmental and communicative interrelations of Ly123 and Lyl cell sets. J Exp Med 1980; 151:566-572. (20) GAUTAM SC, Fyfe D, DEODHAR SD. Augmentation of immune re sponse to syngeneic fibrosarcoma T241 with in vivo protection. Cancer Immunol Immunother 1979; 6:81-88. (21) Crile G JR, DEODHAR SD. Role of pre-operative irradiation in prolonging concomitant immunity and preventing metastases in mice. Cancer 1971; 27:629-634. (22) Baldwin RW. Relevant animal models for tumor immunotherapy. Cancer Immunol Immunother 1976; 1:197-198. (23) Martin DS, STOLFI R, FUGMAN RA. Animal models for tumor im munotherapy. Cancer Immunol Immunother 1977; 2:77-79. (24) Howell SB, Dean JH, Esser EC, Law LN. Cell interactions in adoptive immune rejection of a syngeneic tumor. Int J Cancer 1974; 14:662-674. (25) GAUTAM SC, Deodhar SD. Presence of suppressor cells in spleens of mice bearing a weakly immunogenic syngeneic tumor. Cancer Res 1979; 39:2945-2951. (26) Fujimoto S, GREENE M, SEHON AH. Regulation of the immune response to tumor antigens. I. Immuno-suppressor cells in tumor bearing hosts. J Immunol 1976; 116:791-799. (27) BRODER S, WALDMANN TA. The suppressor-cell network in cancer. N Engl J Med 1978; 299:1281-1284. (28) Fernandez-Cruz E, GILMAN SC, Feldman JD. Immunotherapy of a chemically induced sarcoma in rats: Characterization of the effector T cell subset and nature of suppression. J Immunol 1982; 128:1112 1117. (29) BURTON RC, PLATE JM. Helper factor(s) augment in vitro induction of tumor-specific immunity. Cell Immunol 1981; 58:225-237. (30) Shiku H, TAKAHASHI T, BEAN MA, OLD LJ, OETTGEN HF. Ly phenotype of cytotoxic T cells for syngeneic tumor. J Exp Med 1976; 144:1116-1120. (31) STUTMAN O, SHEN FW, Boyse EA. Ly phenotype of T cells cytotoxic for syngeneic mouse mammary tumors: Evidence for T cell inter actions. Proc Natl Acad Sci USA 1977; 74:5667-5671. (32) Sinclair NR, McFarlane DL, Low JM. Cell mediated cytotoxicity against syngeneic tumors. Cell Immunol 1981; 59:330-344. (33) Shimizu K, Shen FW. Role of different T cell sets in the rejection of syngeneic chemically induced tumors. J Immunol 1979; 122:1162 1165. (34) Nelson M, Nelson DS, McKenzie IF, BLANDEN RV. Thy and Ly markers on lymphocytes initiating tumor rejection. Cell Immunol 1981; 60:34-42. (35) BHAN AK, Perry LL, CANTOR H, McClusKEY RT, BENACERRAF B, GREENE MI. The role of T cell sets in the rejection of a methylcholanthrene-induced sarcoma (S1509a) in syngeneic mice. Am J Pathol 1981; 102:20-27. (36) GREENBERG PD, CHEEVER MA, FEFER A. Eradication of disseminated murine leukemia by chemoimmunotherapy with cyclophosphamide and adoptively transferred immune syngeneic Lyt-1+2- lymphocytes. J Exp Med 1981; 154:952-963.
(1) Abelev GI. Production of embryonal serum alpha-globulin by hep atomas. Review of experimental and clinical data. Cancer Res 1968; 28:1344-1350. (2) CINADER B. The future of tumor immunology. Med Clin N Am 1972; 56:801-836. (3) COGGIN JH JR, ANDERSON NG. Cancer differentiation and embryonic antigens: Some central problems. Adv Cancer Res 1974; 19:105 165. (4) Chism SE, BURTON RC, WARNER NL. Immunogenicity of oncofetal antigens: A review. Clin Immunol Immunopathol 1978; 11:346 373. (5) Jones PC, SIDELL N, Irie RF. Embryonic antigens and tumor cell cytolysis. Cancer Immunol Immunother 1980; 8:211-214. (6) Ting CC, RODRIGUES D, HERBERMAN RB. Expression of fetal antigens and tumor-specific antigens in SV40-transformed cells. II. Tumor transplantation studies. Int J Cancer 1973; 12:519–523. (7) Baldwin RW, Glaves D, Voss BM. Immunogenicity of embryonic Page 3
not increase either tumor incidence or multiplicity in any resistant strain. Results similar to these were found by Witschi and Lock (5), although the effect of BHT on our BALB/cBy subline was much greater than the 50% increase they described for BALB/c mice. We then sought to differentiate this genome-dependent effect on promotion from that on initiation by examining urethan dose-dependence in these same strains. When the urethan dose was increased by the animals' receiving multiple injections of the standard 1 mg/g body weight dose, the same strain-dependence on adenoma induction was shown as that noted when BHT was used as the promoter (table 4). The number of tumors was increased several times in BALB mice, A-strain mice were stimulated by about 50%, and no effect was observed in the three resistant strains. These results suggest that whatever mechanism regulates adenoma susceptibility, neither repeated dosing with a complete carcinogen nor treatment with a promoter can overcome complete genetic resistance. Tumor incidence and tumor number could be increased by either procedure in the
10 20 NUMBER OF TUMORS treatment in sensitive SWR/J mice and (SWR X BALB)Fprogeny. Ten SWR mice and 25 (SWR X BALB)F, mice were used.
Strain-Dependent Enhancement of Tumor Number by Multiple BHT or Multiple Urethan Treatments One approach for the elucidation of the mechanism by which the BALB gene regulates susceptibility to adenoma induction is to test whether it predominantly affects the initiation of the neoplastic process or whether it influences those factors which can be modulated by tumor-promoting agents. BHT causes lung toxicity when injected ip (23). BHT also can inhibit adenoma formation if administered simultaneously with urethan (5), or it can increase the number of adenomas formed when it is given as repeated ip injections following a single urethan dose (4, 5, 18). It is probable that metabolites of BHT, rather than BHT itself, regulate these effects (24, 25) and that different metabolites are involved in each process (25). The capacity of BHT to promote adenomas in one adenoma-sensitive strain (A), the intermediately adenoma-sensitive BALB strain, and the three adenoma-resistant strains was tested (table 3). BHT increased tumor number threefold to fourfold in BALB mice and by 50% in A-mice. Over the range of BHT concentrations tested, no increase in tumor multiplicity with increasing BHT concentrations was observed. BHT treatment did
Table 3.-Effects of multiple injections of BHT on the promotion of urethan-induced lung adenoma in inbred strains Treatment (mg/kg Mortality: No. of mice dead/No. of Incidence: No. mice with tumors/total Mean tumor Mouse strain body wt)" mice treated No. of surviving mice No.ASEM A/J Corn oil 0/12 12/12 22.3+1.47 BHT (400) 0/11 11/11 31.1+2.51 BALB/cByJ Corn oil 0/11 9/11 1.36+0.36 BHT (150) 1/22 21/21 5.0+0.726 BHT (300) 6/23 17/17 4.7120.45 BHT (400) 15/38 22/23 3.26+0.51% C57BL/6J Corn oil 0/5 3/5 0.60+0.24 BHT (400) 5/18 7/13 0.62+0.18 DBA/2J Corn oil 0/5 1/6 0.17 +0.17 BHT (400) 1/20 3/19 0.26+0.15 C3H/2IBG Corn oil 0/5 1/6 0.17+0.17 BHT (400) 10/20 0/10 0
Mice were given injections of 1 mg urethan/g body wt, followed by six weekly injections of either BHT or the corn oil vehicle. Page 4
Synergism With N-Nitrosobutylurea in Castrated Male WF Rats 1,2,3
ABSTRACT—Inbred male WF rats were castrated at 40 days of age and divided into 5 groups. Group I was given no further treatment. Groups III, IV, and V received pellet implants of 5.0 mg diethylstilbestrol (DES) concurrently with castration. At 50-55 days of age, groups II, IV, and V were given drinking water containing 5.0 mg N-nitrosobutylurea (NBU) per day for 30 days (subthreshold dose). At the termination of NBU treatment, group V further received daily sc injections of 2-bromoergocryptine (CB-154; 0.4 mg/100 g body wt) four times a week throughout the experiment. None of castrated rats or rats castrated and treated with NBU alone developed hepatic tumors (HT) and pituitary tumors (PT). Incidences of HT and PT in groups III, IV, and V were 4/9 (44%) and 7/9 (78%), 15/17 (88%) and 12/17 (71%), and 17/20 (85%) and 4/20 (20%), respectively. The treatment of DES alone resulted in the concurrent development of HT and PT in castrated male rats (group III), and further NBU treatment significantly increased the incidence of HT (group IV). CB-154 treatment did not change the incidence of HT, the number of HT per rat, and the liver weight, although it significantly reduced the incidence of PT, the pituitary weight, and the serum prolactin level in castrated male rats given DES and NBU (group V). These results indicate that DES itself had a direct carcinogenic effect on the liver; this effect was not mediated by prolactin, and NBU increased the effect of DES in this process. - JNCI 1983; 70:937-942.
A highly inbred strain of WF male rats maintained in our laboratory housed in temperature-controlled (24+1°C) and light-controlled (on at 7:00 a.m., off at 6:00 p.m.) room and given a standard commercial diet (Oriental MF; Oriental Yeast Co., Ltd., Tokyo, Japan) and tap water. Rats were castrated at 40 days of age and divided into 5 groups. Group I was given no further treatment. Groups III, IV, and V received subcutaneous implantations of pellets containing 5.0 mg DES and 45.0 mg cholesterol concurrently with castration treatment. Pellets were replaced every 2 months. At 50-55 days of age, groups II, IV, and V were given 5.0 mg NBU/day, dissolved in the drinking water (250 ppm solution), for 30 days (subthreshold dose). At the termination of NBU treatment, group V further received daily sc injections of CB-154 (0.4 mg/100 g body wt), dissolved in 0.9% NaCl (2.0 mg/ml solution), throughout the experiment. Castrated male rats treated differently were as follows: Group I, control; Group II, NBU alone; Group III, DES alone; Group IV, DES+NBU; and Group V, DES+NBU+CB-154. Blood was collected from either the tail or the axillary vein in each rat at intervals of 30 days for 12 months for prolactin radioimmunoassay. During the experiment moribund or dead rats were completely autopsied. One year after the initial treatment of NBU, all surviving rats were killed. HT, PT, liver, pituitary gland, mammary gland, epididy
ABBREVIATIONS USED: CB-154=2-bromoergocryptine; DES=diethylstilbestrol; H & E=hematoxylin and eosin; HT=hepatic tumor(s); MT=mammary tumor(s); NBU=N-nitrosobutylurea; PT=pituitary tumor(s).
There is epidemiologic evidence that DES and other synthetic estrogens are carcinogenic in humans (1, 2). A number of experiments with experimental animals also have shown that estrogens induce tumors mainly in tissues that are susceptible to the hormone (2-5). An association between hepatocellular adenoma and oral contraceptives (6, 7) was first pointed out by Baum et al. in 1973 (8). However, the experimental evidence is scanty (4, 9-17). In a series of our studies of rat mammary tumorigenesis treated by NBU or irradiation (12-14), we have observed the increased development of not only MT and PT but also HT in castrated male WF rats given the combined treatment of DES and NBU or irradiation (14). This increased incidence of HT was unexpected because the liver has been considered to be relatively resistant to the exposure of NBU or irradiation (12, 15), and it has been widely accepted that liver carcinogenesis chemically induced in rats is modified by hormones, especially male sex hormones (16). Therefore, HT in our system seems to be a good experimental model for liver cell adenoma in women taking oral contraceptives. Because DES treatment induces the development of PT and elevates the circulative level of prolactin in rats (5, 14), it is uncertain whether DES acts directly on the liver cell or whether its action is mediated via the pituitary gland. The present study was performed to clarify this problem with the use of the ergot drug CB-154, a potent prolactin suppressor (17), concurrently with DES treatment.
Received August 12, 1982; accepted December 29, 1982. 2 Supported by a Grant-in-Aid for Cancer Research and by a Grant-inAid for Scientific Research from the Ministry of Education, Science and Culture, Japan. Animals were maintained under the guidelines set forth by the Research Institute for Nuclear Medicine and Biology, Hiroshima University. * Department of Oral Anatomy (2d Division), School of Dentistry, Hiroshima University, Kasumi 1-2-3, Hiroshima 734, Japan. 6 Present address: % Dr. D. H. Nelson, Division of Endocrinology and Metabolism, Department of Medicine and Physiology, University of Utah, Medical Center, Salt Lake City, Utah 84132. 6 Address reprint requests to Dr. Sumi at Hiroshima University. Department of Pathology, Research Institute for Nuclear Medicine and Biology, Hiroshima University. 8 N-Nitrosobutylurea was generously provided by Dr. M. Nakadate, Department of Synthetic Chemistry, National Institute of Hygienic Sciences, Tokyo, Japan. Rat prolactin for radioimmunoassay was supplied by the National Institute of Arthritis, Diabetes, aned Digestive and Kidney Diseases. We express our gratitude to Professor H. Nagasawa, Meiji University, Kanagawa, Japan, for stimulating discussions. We also thank Mr. T. Nishioka and Miss M. Sasaki for their excellent technical assistance.
"Rats were castrated when 40 days of age. No. of rats that lived >6 mo after the initial NBU treatment. HT include both neoplastic nodule and hepatocellular carcinoma. dEither a gland weighing >30 mg or a macroscopically visible tumor was interpreted as PT. Rats were given 5.0 mg NBU/day in drinking water for 30 days when 50-55 days of age. 'A 5.0-mg DES pellet was grafted subcutaneously on the back and replaced every 2 mo. 8 Rats received daily sc injections of CB-154 (0.4 mg/100 g body wt) four times/wk throughout the experiment.
mis, adrenal gland, and kidney were removed, weighed, and fixed in 10% Formalin and/or Bouin's solution. Paraffin sections were routinely stained with H&E and examined histologically.
Incidences of HT and PT.-The incidences of HT and PT in 5 groups are summarized in table 1. The incidences of HT and PT in groups III, IV, and V were 4/9 (44%) and 7/9 (78%), 15/17 (88%) and 12/17 (71%), and 17/20 (85%) and 4/20 (20%), respectively. None of the castrated rats or rats castrated and treated with NBU alone developed HT and PT (groups I and II). The development of HT and PT occurred concurrently in rats treated with DES alone (group III). In addition, further treatment with NBU (group IV) increased the incidence of HT two times greater than that in group III, whereas it did not increase the incidence of PT. Meanwhile, the incidence of HT (group V) in rats treated with CB-154 and given DES and NBU was not altered, although the incidence of PT was significantly reduced from 71 to 20%. Number and size of HT.–Table 2 shows the mean number
CB-154 DES and cast. Latency (MO) TEXT-FIGURE 1.-Changes in serum prolactin (PRL) levels in DES-treated and castrated (cast.) male WF rats given no further treatment (group III, o), given NBU (group IV, 0), or given NBU and CB-154 (group V, A). Rats that developed PT in group V (@).
of HT and the distribution of HT with diameters larger than 5 mm in groups III, IV, and V. Most of the HT in these 3 groups were observed as multiple nodules of various
"Rats were castrated when 40 days of age. No. of rats that lived >6 mo after the initial NBU treatment. "Tumors include both neoplastic nodules and hepatocellular carcinomas. dA 5.0-mg DES pellet was grafted subcutaneously on the back and replaced every 2 mo. * Rats were given 5.0 mg NBU/day in drinking water for 30 days when 50-55 days of age. Rats received daily sc injections of CB-154 (0.4 mg/100 g body wt) four times/wk throughout the experiment.
TEXT-FIGURE 2.—Body weights and weights of pituitary glands and livers in DES-treated and castrated male WF rats given no further treatment, given NBU, or given NBU and CB-154 (groups III, IV, and V). Pituitary weight in castrated male rats given no further treatment or rats given NBU treatment alone (groups I and II). o, absolute; -, relative.
did not develop PT, the mean weight of the pituitary gland was not different from the mean weights in both rats given no further treatment (group I) and rats given NBU alone (group II). The mean body weights were not significantly different among groups III, IV, and V. Histologic observation of HT.-Histologic examination of hepatocellular lesions was diagnosed by the recent recommended classification (18), which is summarized in table 3. Most of the rats that had hepatocellular carcinoma also had neoplastic nodules and foci of cellular alteration (figs. 1-3). If a rat had these 3 lesions concurrently, we counted this rat
sizes on the surface of the liver. The number of HT increased in rats given DES and NBU (group III). There was no significant difference in both number of and distribution of HT of various sizes between rats given DES and NBU (group IV) and rats further treated with CB-154 (group V). Therefore, growth of HT was not affected by treatment with CB-154. Serum prolactin levels.—Changes of serum prolactin levels in groups III, IV, and V during the 12 months after initial NBU treatment are shown in text-figure 1. The levels of serum prolactin in rats treated with CB-154 (group V) were markedly lower throughout the experiment than were those in the other 2 groups (groups III and IV) and were mostly below about 200 ng/ml. A few samples at 10 and 12 months in group V showed the level above that; these samples were collected from the rats in whom PT was discovered at autopsy. Meanwhile, prolactin levels in groups III and IV were almost above 200 ng/ml and increased sharply about 6 months after the initial NBU treatment. These elevations during the late stage of the experiment appear to be associated with the development of PT. CB-154 treatment to rats given DES and NBU significantly affected the serum prolactin levels. Liver and pituitary weights.—The absolute or relative weights of livers and pituitary glands in each group are shown in text-figure 2. In rats given DES and NBU (group IV), the mean liver and pituitary weights were increased when compared with those in rats given DES alone (group III). The mean liver weights were not different between rats given DES and NBU (group IV) and rats further treated with CB-154 (group V), although the mean pituitary weight was reduced by about 74% in group V when compared with that in group IV. Furthermore, in 16 rats of group V that
as having hepatocellular carcinoma as described previously in (14). Inasmuch as neoplastic nodules have been considered to be neoplastic in nature (18), we interpreted hepatocellular carcinoma and neoplastic nodules as HT. In groups III and IV, neoplastic nodules and hepatocellular carcinoma were 3 (33%) and 1 (11%) and 4 (24%) and 11 (65%), respectively. The highly differentiated HT was observed in rats given DES alone (group III) more thar it was in rats given DES and NBU (group IV). The distribution of histologic classification in rats further treated with CB-154 (group V) did not differ significantly from that observed in rats given DES and NBU (group IV). No metastasis formation was detected in any HT case, and no HT was associated with cirrhosis. Furthermore, about 50% of the rats with HT showed a positive reaction for a-fetoprotein (14), whereas none of the rats had neoplastic lesions of bile duct origin. Histologic observation of PT.-The glands weighing 30 mg, as proposed by Clifton and Meyer (19), or macroscopically visible tumors, which are blood-rich, dark red, and elastic soft, were interpreted to be PT (fig. 4). Microscopically, they fell within the classification of hemorrhagic chromophobe adenoma; tumor cells were round and medium size with pale nuclei, and cytoplasms were weak eosinophilic or chromophobic. Mammary glands of PT-bearing rats were hyperplastic when the milk secretion was found (14). The development of MT, most of which were adenocarcinomas, was associated with the presence of PT.
Prolactin levels in groups III and IV increased sharply 6 months after the initial NBU treatment. These elevations of serum prolactin levels apparently were correlated with the development of PT whose incidence was 7/9 (78%) in group III and 12/17 (71%) in group IV. In group V serum prolactin levels were mostly below the 200 ng/ml level and were markedly lower than those in groups III and IV throughout the experiment, although a few samples showed the next higher level. These samples with the high serum prolactin level at 10 and 12 months were collected from rats that developed PT, which was recognized at autopsy (group V). CB-154 treatment of rats given DES and NBU significantly reduced not only the incidence of PT from 71 to 20% but also reduced the mean weight of the pituitary gland from 57.4 to 14.7 mg. Furthermore, in 16 rats that did not develop PT (group V), pituitary gland weight was reduced as compared with that in both rats given no further treatment (group I) and rats given NBU treatment (group II). Thus CB-154 was considered effective in the suppression of PT development and growth and in the suppression of prolactin secretion. On the contrary, CB-154 was ineffective in the reduction of the incidence of HT, the mean number of HT, and the mean weight of liver, and CB-154 did not alter the distribution of the histologic classification of hepatocellular lesions (group V). Moribund or dead rats during the late stage of the experiment had already developed HT when they were completely autopsied, and the serum prolactin levels from these rats were low (group V). The development of HT was considered not to be associated with the prolactin level. Therefore, prolactin does not seem to play a key role in the hepatic tumorigenesis in this study. We have had the belief that DES action on the liver cell is not mediated via the pituitary gland in the secretion of prolactin, but we do not contradict that other pituitary hormones may be able to affect the liver cell tumor formation. Recent reports have shown that estrogen, prolactin, and growth hormone bound specifically to the livers of several species, indicating the presence of receptors for each hormone on the liver (21-23). In general, implantation of a DES pellet induces the development of PT that possess mammosomatotropin (5, 26), and these hormones have been shown to promote the chemical carcinogen-induced hepatocarcinogenesis in rats (25, 26). Our previous study also showed that grafted mammotropic hormone-secreting PT, originally induced by DES, that have been transplanted in the syngeneic rat shortened the latent period of HT in castrated male rats given DES and NBU, particularly during the early stage of the experiment, but they did not increase the incidence of HT at the termination of the experiment (14). Furthermore, hypophysectomy or lesions in the median eminence area of the hypothalamus inhibited the hepatic carcinogenicity of several powerful carcinogens (27, 28). Thus growth hormone might be the main important pituitary hormone responsible for liver tumor induction by chemical carcinogens, although its effect might be considered not significant in this study. The findings of the present study suggest that DES may act directly on the liver cell as a weak carcinogen. This supposition is supported by our recent observation that the prolonged administration of antiestrogens inhibit the development of HT in castrated male rats treated with DES and
In the present study we investigated whether DES itself has a direct carcinogenic effect on the livers of castrated male rats given DES and NBU. About 44% of the castrated male rats given DES alone developed HT (group III). This result is evidence that HT were induced by DES; this supports the report of Reznik-Schüller (4) that 29% of the European hamsters treated with subcutaneous implants of DES pellets developed liver tumors. In addition, further treatment by NBU in these DES-treated rats increased the incidence of HT two times greater than it did in rats without NBU treatment (group III), indicating that the development of HT was significantly elicited by the subthreshold dose of NBU that, even in female rats, could not induce the tumorigenesis of the mammary gland as primary target tissue (12). The liver has not been considered to be the target tissue of NBU (12, 15). Therefore, a small number of chemical agents such as NBU in this study, which has appeared noncarcinogenic, might have contributed to the induction of the tumor formation when the hormone environment in the host is modified by exogenous hormones. The interaction of NBU and DES on the metabolic activation or the detoxification is not obvious in this study. However, NBU as well as irradiation (14) appears to have an initiating effect. Takizawa and Nishihara (20) pointed out the multicarcinogenicity of NBU in regard to its spectrum of target organ. NBU also might have promoted a initiating effect of DES by acting as a preconditioning factor and/or as an immunosuppressor. In general, chemical carcinogens have been known to suppress the host immunologic competence.
NBU (Sumi C: Unpublished data). DES appears to act in both the initiation and promotion processes of HT formation. As discussed previously (14), DES also might have played a role as a conditioning factor; its sustained stimuli on the male liver cells conceivably caused chronic cell injury, which rendered the cells more susceptible to the carcinogens such as NBU or to irradiation. Estrogens have been reported to cause hepatic dysfunction with depression of hepatic transport systems for organic anions (29), and the metabolic activation of DES and its possible relation to carcinogenic effect have been recently investigated (30, 31). The carcinogenic effect of DES might have been enhanced by the castration. We have recently found that none of the intact male or female rats developed HT when DES was administrated, whereas castrated female rats had a low incidence of HT by the same treatment (Sumi C: Unpublished data). The condition of being castrated might not only negate counteraction by the male sex hormones to DES but also might cause serious changes in the overall hormone balance of host rats. The increased development of HT in our system was primarily unexpected because, in general and as induced by various chemicals, rat liver carcinogenesis has been well known to be enhanced by the male sex hormones (16). However, our findings may be relevant to the recent observation of liver cell adenoma associated with the use of oral contraceptives (2, 6-8). Since the paper of Baum et al. (8), it has been known that the incidence of hepatocellular tumor development increased among women under the influence of oral contraceptives. Inasmuch as the incidence of primary HT in individuals untreated with sex hormones is rare in Europe and the United States, an association has been suggested between the use of oral contraceptives and the development of primary HT in women. However, the carcinogenicity of oral contraceptives on the liver was concluded to be negative by the Committee on Safety of Medicine (3) because long-term treatment of rats with contraceptives did not increase the incidence of hepatomas. Therefore, the high yield in rats of HT that was produced in our experimental system is considered to be a useful tool for the study of the mechanism of the HT associated with steroids and for the detection of the carcinogenicity of unknown chemical agents that act synergistically with these steroids on the liver. Further investigations are required.
istry patients. J Natl Cancer Inst 1977; 58:167-171. (8) Baum JK, Holtz F, BROOKSTEIN JJ, Klein EW. Possible association between benign hepatomas and oral contraceptives. Lancet 1973; 2:926-929. (9) Taper HS. The effect of estradiol-17-phenylpropionate and estradiol benzoate on N-nitrosomorpholine-induced liver carcinogenesis in ovariectomized female Sprague-Dawley rats. Cancer 1978; 42:462 467. (10) Yager JD JR, Yager R. Oral contraceptive steroids as promoters of hepatocarcinogenesis in female Sprague-Dawley rats. Cancer Res 1080; 40:3680-3685. (11) WANLESS IR, Medline A. Role of estrogens as promotors of hepatic neoplasia. Lab Invest 1982; 46:313-320. (12) YOKORO K, NAKANO M, ITO A, NAGAO K, Kodama Y, Hamada K. Role of prolactin in rat mammary carcinogenesis: Detection of carcinogenicity of low-dose carcinogens and of persisting dormant cancer cells. J Natl Cancer Inst 1977; 58:1777-1783. (13) Yokoro K, Sumi C, Ito A, Hamada K, Kanda K, KOBAYASHI T. Mammary carcinogenic effect of low-dose fission radiation in Wistar/Furth rats and its dependency on prolactin. JNCI 1980; 64:1459-1466. (14) Sumi C, Yokoro K, Kapitani T, ITO A. Synergism of diethylstilbestrol and other carcinogens in concurrent development of hepatic, mammary and pituitary tumors in castrated male rats. JNCI 1980; 65:169-175. (15) ODASHIMA S. Leukemogenesis of N-nitrosobutylurea in the rats. I. Effect of various concentrations in the drinking water to female Donryu rats. Gan 1970; 61:245-253. (16) FIRMINGER HI, REUBER MD. Influence of adenocortical, androgenic, and anabolic hormones on the development of carcinoma and cirrhosis of the liver in AXC rats fed N-2-fluorenyldiacetamide. J Natl Cancer Inst 1961; 27:559-595. (17) Brooks CL, Welsch CW. Reduction of serum prolactin in rats by 2 ergot alkaloids and 2 ergoline derivatives: A comparison. Proc Soc Exp Biol Med 1974; 146:863-867. (18) SQUIRE RA, LEVITT MH. Report of a workshop on classification of
REFERENCES (1) HERBST AL, ULFELDER H, POSKANZEW DC. Adenocarcinoma of the vagina. Association of maternal stilbestrol therapy with tumor appearance in young women. N Engl J Med 1971; 284:878-881. (2) LINGEMAN CH. Hormones and hormonomimetic compounds in the etiology of cancer. Carcinogenic hormones. Recent Results Cancer Res 1979; 66:1-48. (3) Committee on Safety of Medicines. Carcinogenicity testing of oral Page 5
Figure 1.—Gross appearance of HT in castrated male WF rats given DES alone, diagnosed histologically as hepatocellular carcinoma. Large arrow, HT: double-headed arrow, right kidney; small arrow, spleen. FIGURE 2.-Histologic appearance of foci of cellular alteration in castrated male WF rats given DES and NBU. Foci are compressing the surrounding liver parenchyma; cells in foci do not contain lipofusion-like pigmentation, whereas adjacent liver cells do and various degenerative changes occur. H&E. Bar=250 um. FIGURE 3.— Histologic appearance of well-differentiated hepatocellular carcinoma in castrated male WF rats given DES and NBU. Note papillary proliferation of tumor cells in dilated sinusoid. H&E. Bar=25 um. FIGURE 4.—Gross appearance of PT in castrated male WF rats given DES and NBU; gland weighs 187 mg. Huge tumor mass, dark red and elastic soft, Page 6
F19 cells (table 3), it is tempting to speculate that FBP and RCA-I may be binding to viral encoded or induced surface components. Although a strict correlation does not exist, data from a number of sources suggest that the tumorigenic potential in the Ad2-transformed rat cell lines may relate to the fraction of the viral genome represented in the transformed cell line (reviewed in (17)]. Therefore, the increased viral information may, upon interaction with various host cellular components, provide for both increased exposure of galactose and fucose residues at the cell surface and increased variability of surface structural features within the cell population affected (see text-figs. 3 and 4, F4 and T2C4). This increased heterogeneity may provide a selective advantage for the induction of tumors in animal hosts, because conceivably, variations in populations of surface carbohydrate molecules in the malignant cells could isolate and protect the cells from outside influences and regulating forces such as the host immune system. Another possibility may be that the greater tumorigenic potential of the F4 and T2C4 lines merely reflects the presence within these heterogeneous populations (and absence from the less heterogeneous F17 and F19 populations) of highly malignant cell variants characterized by specific surface changes that allow their escape from host surveillance and also provide for their uncontrolled proliferation, invasive properties, and/or ability to metastasize. The demonstration of lectin-binding heterogeneity within the Ad2-transformed cells lines should allow for selection of cells from each line with specifically defined surface features so that these hypotheses may be tested.
(3) Harwood LM, GALLIMORE PH. A study of the oncogenicity of adenovirus type 2 transformed rat embryo cells. Int J Cancer 1975: 16:498-508. (4) Wold WS, Green M. Adenovirus type 2 early polypeptides immu noprecipitated by antisera to five lines of adenovirus-transformed rat cells. J Virol 1979; 30:297–310. (5) Binger MH, FLINT SJ, REKOSH DM. Expression of the gene encoding the adenovirus DNA terminal protein precursor in productively infected and transformed cells. J Virol 1982; 42:488–501. (6) GALLIMORE PH, SHARP PA, SAMBROOK J. Viral DNA in transformed cells. II. A study of the sequences of adenovirus 2 DNA in nine lines of transformed rat cells using specific fragments of the viral genome. J Mol Biol 1974; 89:49–79. (7) SAMBROOK J, Botchan M, GALLIMORE P, et al. Viral DNA sequences in cells transformed by simian virus 40, adenovirus type 2, and adenovirus type 5. Cold Spring Harbor Symp Quant Biol 1974: 39:615-632. (8) Flint SJ, GALLIMORE PH, SHARP PA. Comparison of viral RNA sequences in adenovirus 2-transformed and lytically infected cells. J Mol Biol 1975; 96:47-68. (9) Persson H, Signas C, PHILIPSON L. Purification and characterization of an early glycoprotein from adenovirus type 2-infected cells. J Virol 1979; 29:938-948. (10) Chen LB, GALLIMORE PH, McDougall JK. Correlation between tumor induction and the large external transformation sensitive protein on the cell surface. Proc Natl Acad Sci USA 1976; 73:3570 3574. (11) Boldt DH, MacDERMOTT RP, JOROLAN EP. Interaction of plant lectins with purified human lymphocyte population: Binding characteristics and kinetics of proliferation. J Immunol 1975; 114:1532 1536. (12) Brown JC, HUNT RC. Lectins. Int Rev Cytol 1978; 52:277-349. (13) SCATCHARD G. The attractions of proteins for small molecules and ions. Ann NY Acad Sci 1949; 51:660-672. (14) RODBARD D. Mathematics of hormone-receptor interaction. I. Basic principles. In: O'Malley BW, Means AR, eds. Receptors for repro ductive hormones. New York: Plenum Press, 1973:278-297. (15) HORAN PK, WHEELESS LL. Quantitative single cell analysis and sorting. Science 1977; 198:149-157. (16) Jacobs S, CUATRECASAS P. The mobile receptor hypothesis and "cooperativity” of hormone binding. Application to insulin. Biochim Biophys Acta 1976; 433:482-495. (17) McDougaLL JK, CHEN JB, DUNN AR, GALLIMORE PH. Adenovirus integration and oncogenicity from integration and excision of DNA molecules. In: Colloquium de Gesellschaft für Biologische Chemie. Vol 28. New York: Springer-Verlag, 1978:59–70. Page 7
(3) Clamon GH, SPorn MB, Smith JM, SAFFIOTTI U. a- and B-retinyl acetate reverse metaplasias of vitamin A deficiency in hamster trachea in organ culture. Nature 1974; 250:64-66. (4) STRICKLAND S, Mahdavi V. The induction of differentiation in tera tocarcinoma stem cells by retinoic acid. Cell 1978; 15:393-403. (5) JETTEN AM, JETTEN ME, SHERMAN MI. Stimulation of differentiation of severval murine embryonal carcinoma cell lines by retinoic acid. Expt Cell Res 1979; 124:381-391. (6) Tchao R. Keratinization and the effect of vitamin A in aggregates of a squamous carcinoma cell line NBT II. In Vitro 1980; 16:407– 414. (7) SPORN MB, NEWTON DL. Chemoprevention of cancer with retinoids. Fed Proc 1979; 38:2528-2534. (8) Peto R, DOLL R, BUCKLEY JD, SPORN MB. Can dietary beta-carotene materially reduce human cancer rates. Nature 1981; 290:201-208. (9) JETTEN AM, JETTEN MR, SHAPIRO SS, Poon JP. Characterization of the action of retinoids on mouse fibroblast cell lines. Expı Cell Res 1979; 119:289-299. (10) Haddox MK, Russell DH. Cell cycle-specific locus of vitamin A inhibition of growth. Cancer Res 1979; 39:2476-2480. (11) SHAPIRO SS, Poon JP. Retinoic acid-induced alterations of growth and morphology in an established epithelial line. Expt Cell Res 1979; 119:349-357. (12) Lotan R. Different susceptibilities of human melanoma and breast carcinoma cell lines to retinoic acid-induced growth inhibition. Cancer Res 1979; 39:1014-1019. (13) LACROIX A, LIPPMAN ME. Binding of retinoids to human breast cancer cell lines and their effects on cell growth. J Clin Invest 1980; 65:586-591. (14) Dicker P, ROZENGURT E. Retinoids enhance mitogenesis by tumor promoter and polypeptide growth factors. Biochem Biophys Res Commun 1979; 91:1203-1210. (15) Adamo S, De Luca LM, AKALOVSKY I, BHAT PV. Retinoid-induced adhesion in cultured, transformed mouse fibroblasts. JNCI 1979; 62:1473-1478. (16) Shin S, FrEEDMAN VH, Risser R, POLLACK R. Tumorigenicity of virus-transformed cells in nude mice is correlated specifically with anchorage independent growth in vitro. Proc Natl Acad Sci USA 1975; 72:4435-4439. (17) BENNETT DC, Peachey LA, DURBIN H, RUDLAND PS. A possible mammary stem cell line. Cell 1978; 15:283-298. (18) Rudland PS, BENNETT DC, WARBURTON MJ. Hormonal control of growth and differentiation of cultured mammary gland epithelial cells. Cold Spring Harbor Conf Cell Proliferation 1979; 6:677-699. (19) Rudland PS, BENNETT DC, Ritter MA, Newman RA, WARBURTON MJ. Differentiation of a rat mammary stem cell line in culture. In: 1980:341-365. (20) Young S, Hallowes RC. Tumors of the mammary gland. In: Tu
induced tumor epithelium from rat mammary glands. J Cell Biol 1977; 73:561-577. (28) JIMENEZ DE Asua L, CARR B, CLINGAN D, RUDLAND PS. Specific glucocorticoid inhibition of growth promoting effects of prostaglan din F2a on 3T3 cells. Nature 1977; 265:450-452. (29) Riddle PN. Analysis of cinemicrographs. In: Time-lapse cinemicros copy. London: Academic Press, 1979:82-90. (30) Davies AJ. Immunodeficient animals. In: Sparrow S, ed. Immuno deficient animals for cancer research. London: Macmillan, 1980:1 4. (31) Williams JC, GUSTERSON B, HUMPHREYS J, et al. N-methyl-N-nitro sourea-induced rat mammary tumors. Hormone responsiveness but lack of spontaneous metastasis. JNCI 1981; 66:147-155. (32) FRIEND C, Scher W, HOLLAND JG, Sato T. Hemoglobin synthesis in murine virus-induced leukemic cells in vitro: Stimulation of erythroid differentiation by dimethyl sulfoxide. Proc Natl Acad Sci USA 1971; 68:378-382. (33) REUBEN RC, WIFE RL, BRESLOW R, RIFKIND RA, MARKS PA. A new group of potent inducers of differentiation in murine erythroleu kemic cells. Proc Natl Acad Sci USA 1976; 73:862-866. (34) SANTORO MG, BENEDETTO A, JAFFE BM. Prostaglandin A, induces differentiation in Friend erythroleukemic cells. Prostaglandins 1978; 17:719-727. (35) STRICKLAND S, SAWEY MJ. Studies on the effect of retinoids on the differentiation of teratocarcinoma stem cells in vitro and in vivo. Dev Biol 1980; 78:76-85. (36) Speers WC. Conversion of malignant murine embryonal carcinomas to benign teratomas by chemical induction of differentiation in vivo. Cancer Res 1982; 42:1843-1849. (37) BREITMAN TR, SELONICK SE, COLLINS SJ. Induction of differentiation of the human promyelocytic leukemia cell line (HL-60) by retinoic acid. Proc Natl Acad Sci USA 1980; 77:2936-2940. (38) MÜLLER-SALAMIN L, MATTER A, LASNITZKI I. Interactions of retinoic acid and 3-methylcholanthrene on the fine structure of mouse prostate epithelium in vitro. JNCI 1979; 63:485-495. (39) Dexter DL, BARBOSA JA, CALABRESI P. N,N-Dimethylformamide induced alteration of cell culture characteristics and loss of tumorigenicity in cultured human colon carcinoma cells. Cancer Res 1979; 39:1020-1025. (40) Dexter DL. N, N-Dimethylformamide-induced morphological differ entiation and reduction of tumorigenicity in cultured rhabdomyo sarcoma cells. Cancer Res 1977; 37:3136-3140. (41) MEHTA RG, CERNY WL, Moon RC. Distribution of retinoic acid binding proteins in normal and neoplastic mammary tissues. Can cer Res 1980; 40:47-49. (42) JETTEN AM, JETTEN ME. Possible role of retinoic acid binding protein in retinoid stimulation of embryonal carcinoma cell differentiation. Nature 1979; 278:180-182. (43) Matusik RJ, ROSEN JM. Prolactin induction of casein mRNA in organ cultures. J Biol Chem 1978; 253:2343-2347. (44) Sopowit SC, ROSEN JM. Hormonal induction of casein gene expres sion is limited to a small subpopulation of 7,12-dimethylbenz(alanthracene induced mammary cells. Cancer Res 1982; 42:1355 1360. (45) Welsch CW, GOODRICH-Smith M, BROWN CK, CROWE N. Enhance ment by retinyl acetate of hormone-induced mammary tumorigen esis in female GR/A mice. JNCI 1981; 67:935-938. (46) MEHTA RG, Moon RC. Inhibition of DNA synthesis by retinyl acetate during chemically induced mammary carcinogenesis. Can cer Res 1980; 40:1109-1111. (47) McCormicK DL, BURNS FJ, ALBERT RE. Inhibition of rat mammary rcinogenesis by short dietary exposure to retinyl acetate. Cancer Res 1980; 40:1140-1143. (48) THOMPSON HJ, Beccı PJ, BROWN CC, Moon RC. Effect of the duration of retinyl acetate feeding on inhibition of l-methyl-1nitrosourea-induced mammary carcinogenesis in the rai. Cancer Res 1979; 39:3977-3980. (49) Gusterson B, WARBURTON MJ, MITCHELL D, ELLISON M, NEVILLE AM, RUDLAND PS. Distribution of myoepithelial cells and basement membrane proteins in the normal, benign and malignant breast. Page 8
287:160-161. (51) LOTAN R. Effects of vitamin A and its analogs (retinoids) on normal and neoplastic cells. Biochim Biophys Acta 1980; 605:33-91. (52) LEVINE L, Ohuchi L. Retinoids as well as tumor promoters enhance deacylation of cellular lipids and prostaglandin production in MDCK cells. Nature 1978; 276:274-275. (53) REED ND, MANNING JK, Baker PJ, ULRICH JT. Analysis of thymus independent immune responses using nude mice. In: Rygard J, Polvsen CO, eds. Proceedings of the first international workshop on nude mice. Stuttgart: Fischer Verlag 1974:95–103. (54) Huggins C, Grand LC, BRILLANTES FP. Mammary cancer induced by a single feeding of polynuclear hydrocarbons and its suppression. Nature 1961; 189:204-207. (55) Dao TL. Mammary cancer induction by 7,12-dimethylbenz[a]an
(57) Moon RC. Relationship between previous reproductive history and
Figure 1.-Morphology of cells treated with RA. A) Confluent culture of cuboidal Rama 25 cells in BM for 2 days. B) Confluent culture of criss-cross elongated Rama 521 cells treated with 3.3 uM RA and BM for 2 days. C) Confluent culture of Rama 259 cells treated for 2 days with BM and RA. D) Confluent culture of Rama 25 cells treated for 2 days with BM and RA showing droplet cells and a dome (D). A-D) Living cultures were photographed with phase-contrast optics. White bar=100 um. X 190
n-propylamine Derivatives 1,2,3
W. Lijinsky,4 M. D. Reuber, 4 J. E. Saavedra, 4 and G. M. Singer 4
compound has been identified as a metabolite of several Nnitrosomethyl-n-alkylamines (5) that gave rise to tumors of the bladder in rats when given by gavage (1), and the concentration of the gavage solution (63 mg/ml) was equimolar with that of the nitrosomethylalkylamine treatments. N-Nitrosomethyl(3-carboxypropyl)amine, also a urinary metabolite of several N-nitrosomethyl-n-alkylamines having an even number of carbons in the chain, was administered as a 2-mM solution because nitrosamino acids are usually weak carcinogens. The concentrations of compound in the various treatment groups are given in table 1.
ABSTRACT — The carcinogenicity of N-nitrosomethyl-n-propylamine and five of its derivatives, including N-nitrosomethyl-n-butylamine, was compared by oral administration of the compounds to inbred F344 rats. N-Nitromethyl-n-propylamine and N-nitrosomethyl-n-butylamine given in drinking water induced tumors of the upper gastrointestinal tract, mainly carcinomas of the esophagus, and appeared to be of comparable potency. N-Nitrosomethyl(2-hydroxypropyl)amine also mainly induced esophageal carcinomas (100% incidence) and lung tumors, whereas N-nitrosomethyl(2,3-dihydroxypropyl)amine mainly induced nasal cavity tumors and gave rise to a high incidence of esophageal tumors; however, it appeared to be less potent than the monohydroxy compound. N-Nitrosomethyl(2-oxopropyl)amine, the ketone corresponding to N-nitrosomethyl(2-hydroxypropyl)amine, was a more potent carcinogen than the latter at comparable doses in drinking water and gave rise to a high incidence of esophageal tumors and tumors of the trachea; female rats had a high incidence (15/20) of angiosarcomas of the liver, but only 2 male rats died with this tumor. When N-nitrosomethyl(2-oxopropyl)amine was administered at a lower dose in drinking water or at the same dose given by gavage, the incidence of esophageal tumors was lower and there were fewer carcinomas. After administration of large doses in drinking water to male and female rats, N-nitrosomethyl(3-carboxypropyl)amine, a urinary metabolite of several N-nitrosomethyl-n-alkylamines that induce tumors of the urinary bladder in rats, gave rise to a high incidence of transitional cell carcinomas of the bladder. The time to death of animals with these tumors was long, and there were few other tumors.—JNCI 1983; 70:959-963.
Chemicals.-N-Nitrosomethyl-n-propylamine and N-nitrosomethyl-n-butylamine were prepared by nitrosation of the corresponding amines (Aldrich Chemical Co., Milwaukee, Wis.) and were distilled under reduced pressure. Their mass spectra and NMR spectra showed them to be of high purity. N-nitrosomethyl(2-hydroxypropyl)amine, N-nitrosomethyl(2-oxopropyl)amine, and N-nitrosomethyl(3-carboxypropyl)amine were prepared as described by Singer et al. (5). N-Nitrosomethyl(2,3-dihydroxypropyl)amine. First N-nitrosomethyl(2,3-dihydroxypropylamine was prepared by the addition of 26 ml (0.5 mol) of glycidol in 50 ml methanol to 60 g (1.9 mol) of methylamine in 500 ml methanol at 0°C, after which the solution was stirred at room temperature for 4 hours. The methanol and excess methylamine were removed on a rotary evaporator, and the residue was vacuum distilled to give 25 g (46%) of the amine (boiling point, 856°C at 0.1 mm Hg; infrared spectrophotometry (film), 3,360, 2,940, 1,460, 1,080, 1,040, and 860 cm NMR (CDC13), 8 2.42 (s, 3H), 2.66 (d, 2H), and 3.4–3.91 (m, 5H);MS, m/z (%), 106 (M* + 1, 17), 105 (M", 2), 88 (4), 87 (22), 86 (3), 75 (6), 74 (100), 70 (9), 59 (12), 58 (21), and 56 (43)]. For the preparation of the nitroso derivative, a solution of 19.5 g (0.186 mol) of N-methyl(2,3-dihydroxypropyl)amine
A number of N-nitrosomethyl-n-alkylamines containing alkyl chains of different lengths were administered to rats by gavage, and their carcinogenic effectiveness was compared (1). Similarly, the effects on carcinogenicity of various substituents in the ethyl group of N-nitrosomethylethylamine have been investigated (2). In the latter study, the considerable influence of the substituents on the potency of the resulting nitrosamine led us to study derivatives of Nnitrosomethyl-n-propylamine. Again, the objective was similar-to gain insight into the relative importance of oxidation of the methyl group and of the substituted n-propyl group in carcinogenesis. This information can then be compared with the results of in vitro studies of metabolism by rat liver microsomes (3, 4). The derivatives of N-nitrosomethyl-n-propylamine (textfig. 1) used in this study include those with one or more hydroxyl groups, a methyl group, a carbonyl group, or a carboxyl group in the propyl chain. It was originally planned to administer all of the compounds in drinking water at equimolar concentrations to make comparisons of potency easier. However, because of their great toxicity, Nnitrosomethyl-n-propylamine and N-nitrosomethyl-n-butylamine were given at low concentrations. One compound, N-nitrosomethyl(2-oxopropyl)amine, was administered to rats by gavage as well as in the drinking water. This
ABBREVIATIONS USED: MS=mass spectrometry; m/z=mass/charge; NMR=nuclear magnetic resonance.
Received September 7, 1982; accepted December 7, 1982. Supported by Public Health Service contract NOICO-75380 from the National Cancer Institute. The contents of this publication do not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the U.S. Government.
methanol to give 14 g (56%) N-nitrosomethyl(2,3-dihydroxypropyl)amine (infrared spectrophotometry (film), 3,380, 2,920, 1,440, 1,420, 1,330, and 1,030 cm-?; NMR (acetonedo), 8 3.12 (s, 2.45H) syn CH3, 3.88 (s, 0.55H) anti-CH3, 3.63 (d, 2H), and 3.44-4.12 (complex, 5H); UV spectrophotometry, Imax (€), 338 (97); MS, m/2 (%), 135 (1.8), 134 (M*, 0.5), 116 (2), 105 (1.4), 104 (0.9), 103 (3), 87 (5), 86 (3), 75 (3), 74 (15), 73 (46), 72 (9), 61 (100), 60 (5), 59 (8), 58 (17). 57 (10), and 56 (11)]. Analysis: C4H10N203; calculated: C, 35.82; H, 7.51; N, 20.89 found: C, 35.70; H, 7.63; N, 21.04
in 180 ml water at 5°C was acidified with 24 ml concentrated hydrochloric acid. To the resulting solution was added 33 g (0.48 mol) sodium nitrite in small lots over 15 minutes. The ice bath was removed, and the reaction mixture was stirred at 25°C for 2 hours. The product was extracted into ethyl acetate, which was then evaporated leaving a yellow oil. The crude material was treated with methanol-HCl until evolution of methyl nitrite ceased. The material was reextracted into ethyl acetate, washed with 5% sodium bicarbonate solution, and dried over anhydrous sodium sulfate. The solvent was removed on a rotary evaporator, and the product was chromatographed through dry-packed silica gel and eluted with 10:1 chloroform:
Animal treatments.—A group of 20 F344 rats 7-8 weeks old from the colony of the NCI-Frederick Cancer Research Facility, bred and maintained in a barrier facility, was used for each treatment. These animals display nearly uniform weights and growth characteristics, adult males weighing approximately 400 g and females weighing approximately 200 g. The animals were housed 4 to a plastic cage with a wire-mesh bottom and fed Rockland rat diet in pellets ad libitum. For those groups receiving the nitrosamines in drinking water, a solution of the nitrosamine was given a the rate of 20 ml per rat per day, 5 days a week; all of the solution was consumed with little spillage. So that water deficiency could be avoided, on the other 2 days of each week tap water was given ad libitum, as it was continuously to those groups in which treatment was by gavage. All six compounds were given to female rats, but groups of males used for N-nitrosomethyl(3-carboxypropyl)amine treatment and for three types of treatment with N-nitroso
Oropharynx 7 (5) Tongue 3 (2) Epiglottis 3 (2) Oropharynx 8 (5) Tongue 2 (2) Leukemia 2 Trachea 6 (0) Bronchus carcinoma 1 Trachea 11 (10) Kidney pelvis carcinoma 1 Pancreas acinar cell adenoma 1 Tongue 4 (2) Trachea 2 (1) Leukemia 9 Lung adenocarcinoma 8 Lung squamous cell carcinoma 7 Trachea 1 (0) Tongue 1 (1) Epiglottis 1 (1) Kidney interstitial cell sarcoma 1 Tongue 3 (2) Lung 2 (1) Pituitary gland 3 (2) Uterus carcinoma 1 Bladder polyp 1
methyl(2-oxopropyl)amine. For economic reasons both sexes were not used for all treatments. The treatments lasted a prescribed number of weeks, after which time the animals were observed until natural death. Each animal was necropsied. All lesions and major organs were fixed in Formalin for microscopic examination. No special group of untreated controls was kept, since groups of untreated F344 rats are continuously maintained in the facility. Contemporary control animals did not begin to die until the 80th week; there was 90% survival at 100 weeks.
The survival times among treated animals are shown in table 1. In all groups except those given N-nitrosomethyl(3carboxypropyl)amine, most of the rats were dead by week 50, and the cause of death was almost invariably tumors induced by the treatments; tumors of the esophagus, liver, and nasal cavity were the most common tumors. The number of animals with each type of tumor is also given in table 1. The bladder tumors induced by N-nitrosomethyl(3-carboxypropyl)amine took considerably longer to kill the animals. The mortality rates of the rats treated with N-nitrosomethyl-n-propylamine and its higher homologue N-nitrosomethyl-n-butylamine, which were administered at approximately equimolar doses totaling 0.3 mmol, were similar. This result suggests that these two compounds are of similar carcinogenic potency, since, in both cases, the rats died of tumors of the upper gastrointestinal tract, mainly carcinomas of the esophagus. By the same criterion the oxygenated N-nitrosomethyl-n
propylamines were considerably less potent than was the parent compound. Again, the common cause of death in these groups was neoplasms of the esophagus, but a smaller dose of N-nitrosomethyl-n-propylamine led to death of the animals almost as rapidly as did the derivatives. N-Nitrosomethyl(2-hydroxypropyl)amine and N-nitrosomethyl(2,3dihydroxypropyl)amine were much less potent by this criterion than was N-nitrosomethyl-n-propylamine, whereas Nnitrosomethyl(2-oxopropyl)amine was only somewhat less potent. The mortality rates of male and female rats given 100 mg N-nitrosomethyl(2-oxopropyl)amine per liter were similar, but 15 females had angiosarcomas of the liver compared with 2 of the males. Tumors of the nasal cavity (squamous cell carcinomas and olfactory adenocarcinomas) and of the trachea were common in the rats treated with N-nitrosomethyl(2-oxopropyl)amine and N-nitrosomethyl(2,3-dihydroxypropyl)amine, but they were rare or absent in other groups. Animals given N-nitrosomethyl(2-hydroxypropyl)amine had a high incicence of lung tumors (squamous cell carcinomas and adenocarcinomas) not seen in the other groups. In contrast, the rats treated with N-nitrosomethyl(3-carboxypropyl)amine, although they survived much better than any of the other groups, had almost exclusively transitional cell carcinomas of the urinary bladder, the only exceptions being 2 females with liver tumors as well.
It appears that the carcinogenic potency of N-nitrosomethyl-n-alkylamines increases considerably with increasing chain length from -ethyl to -n-propyl, but an additional Page 9
carbon atom in the chain, as in N-nitrosomethyl-n-butylamine, did not further increase potency. However, as measured by the rate of mortality from tumors, all of the other substituents in the propyl group decreased the carcinogenic potency. In the case of the 2-hydroxypropyl derivative, the carcinogenic potency was considerably reduced. Rats treated with six times the dose of N-nitrosomethyl(2-hydroxypropyl)amine died at the same rate from esophageal tumors as did the group treated with N-nitrosomethyl-n-propylamine. The corresponding ketone N-nitrosomethyl(2-oxopropyl)amine was somewhat more potent than the alcohol, but it was less potent than the unsubstituted propyl compound. Animals treated with N-nitrosomethyl(2-oxopropyl)amine had a lower mortality rate from esophageal tumors than those receiving half that dose of N-nitrosomethyl-n-propylamine. The mortality rate of rats treated with six times the dose of the ketone was greater than that of rats treated with the unsubtituted propyl compound. N-Nitrosomethyl(2-oxopropyl)amine is partially reduced in vivo to the alcohol Nnitrosomethyl(2-hydroxypropyl)amine, which appears in the urine (5). The rats receiving the 2-oxopropyl compound by gavage suffered considerable toxicity; 5 rats died or were moribund after 8 weeks of treatment, which was then stopped. Although these animals had then received a dose comparable with the dose (100 mg/liter) given in drinking water, the numbers of animals with malignant tumors of the upper gastrointestinal tract and nasal cavity were smaller. The rats in this group developed severe pericentral necrosis of the liver with severe acute hemorrhage, followed by hemorrhage in many other organs of the body, such as the thymus, lymph nodes, and meninges, as well as old and recent hemorrhage into the gastrointestinal tract. Most of these rats did not survive long enough to develop many neoplasms, but several had papillomas of the esophagus or early monocytic leukemia, the latter being common in old untreated F344 rats. It can be assumed that the periodic treatment with larger doses of this nitrosoamine is less effective than are continuous smaller doses given in drinking water. There were no bladder tumors in the rats treated with Nnitrosomethyl(2-oxopropyl)amine, which suggests that despite the presence of this metabolite in the urine of rats treated with even-numbered long-chain N-nitrosomethyl-nalkylamines, this metabolite is not responsible for the bladder tumors induced by these compounds (1). Since the longchain N-nitrosomethyl-n-alkylamines above octyl- did not induce tumors of the upper gastrointestinal tract, the extent of formation of N-nitrosomethyl(2-oxopropyl)amine from them in vivo must be insufficient to induce those tumors. There is evidence that nitrosamines can induce tumors of the upper gastrointestinal tract systemically and not necessarily through direct contact (6). N-Nitrosomethyl(2-oxopropyl)amine at 100 mg/liter in drinking water induced a high incidence (15/20) of angiosarcomas of the liver in female rats; only 2 male rats given this treatment developed liver tumors. This seems to be a rare case of one sex of rats being susceptible to a particular nitrosamine-induced tumor, while the other sex is not. However, the other treatments with this compound (25 mg/liter in water and 63 mg/ml in oil) were given only to males,
which did not develop liver tumors, and it is not known whether liver tumors would have developed in female rats given those treatments. Possibly, the higher dose per unit body weight received by the females induced greater liver toxicity and consequently more liver tumors. In contrast, this compound given by sc injection into Syrian hamsters was reported to give rise to tumors of the nasal cavity, pancreas, liver, and kidneys (7). N-Nitrosomethyl(3-carboxypropyl)amine is a common urinary metabolite of N-nitrosomethyl-n-alkylamines containing even numbers of carbons in the chain (5). These nitrosamines, other than butyl- and hexyl-, have regularly induced transitional cell carcinomas of the urinary bladder and, in most cases, these were the only tumors induced in male rats (1). Our finding that N-nitrosomethyl(3-carboxypropyl)amine is a bladder carcinogen in rats seems to support the suggestion of Okada et al. (8) that this nitrosamino acid is the proximate carcinogenic metabolite of those Nnitrosomethyl-n-alkylamines. This could be true, but the doses of the nitrosamino acid received by each rat in these experiments were considerably higher than those of the Vnitrosomethyl-n-alkylamines that gave rise to bladder tumors. Even if the conversion of the latter to the nitrosamino aci were 100% (and it was not, since other metabolites were found in the urine (5)], the N-nitrosomethyl(3-carboxypropyl)amine would seem to be insufficiently potent to be considered the proximate carcinogen. The pharmacology of these compounds has not, however, been studied in depth, and the explanation for the discrepancy might lie in the pharmacodynamics of the compounds. It is equally possible that N-nitrosomethyl(3-carboxypropyl)amine, which is itself metabolized in rats (5), might have in common with the Nnitrosomethyl-n-alkylamines a metabolite that is the proximate carcinogen. The N-nitrosomethyl-n-alkylamines are usually mutagenic to bacteria (9), whereas N-nitrosomethyl(3-carboxypropyl)amine is nonmutagenic to Salmonella tiphimurium (Andrews AW, Lijinsky W: Unpublished observations). It is not clear why the rat esophagus is so common a target for the derivatives of N-nitrosomethyl-n-propylamine examined. It is clear, however, that substitution in the propyl group decreased carcinogenic effectiveness, which is not supportive of the idea of Krüger that ß-oxidation of the propyl group was an important step in activation of propylnitrosamines (10). The present results are consistent with those of Blattmann and Preussmann (11) and others who found that oxidation of nitrosamines can take place at many sites on the molecule other than the a-carbon atoms. These alternative oxidations can possibly lead to reduction in effectiveness due, for example, to easier excretion. This could be the reason for the lower effectiveness of N-nitrosomethyl(2-hydroxypropyl)amine and the still lower effectiveness of N-nitrosomethyl(2,3-dihydroxypropyl)amine. These compounds could all act by methylation of DNA in the rat esophagus, as suggested by Hodgson et al. (12) Such methylation would require formation of a methylating agent from the nitrosamines by oxidation of the propyl group at the a-position, forming an aldehyde, which in all of these cases is a common product of intermediary metabolism, pyruvaldehyde, lactaldehyde, or glyceraldehyde.
However, oxidation of the methyl group of these substituted propylnitrosamines would release activated forms of these aldehydes, and the effect of these forms is unknown. Formaldehyde is commonly formed by incubation of N-nitrosomethyl-n-alkylamines of this type with rat liver microsomes, but the aldehyde derived from oxidation of the other side of the molecule is formed in much smaller amounts, and sometimes it is not detected at all (Farrelly JG: Unpublished observations). The precise mechanism by which N-nitrosomethyl-n-alkylamines induce esophageal tumors in rats is unknown. Unfortunately, studies of metabolic activation in the esophagus and bladder of the rat are made difficult by the sparseness of these tissues.
nitrosomethyl-n-butylamine and nitrosomethylphenylethylamine. Cancer Res 1982; 42:2105-2109. (5) SINGER GM, LIJINSKY W, BUETTNER L, McClusKY GA. Relationship of rat urinary metabolites of N-nitrosomethyl-n-alkylamine to blad der carcinogenesis. Cancer Res 1981; 41:4942-4946. (6) DRUCKREY H, PREUSSMANN R, IVANKOVIC S, SCHMÄHL D. Organotrope carcinogene Wirkungen bei 65 Verschiedenen N-Nitroso-Verbin dungen an BD-Ratten. Z Krebsforsch 1967; 69:103-201. (7) Pour P, GIngelL R, LANGENBACH R, et al. Carcinogenicity of N nitrosomethyl(2-oxopropyl)amine in Syrian hamsters. Cancer Res 1980; 40:3585-3590. (8) OKADA M, Suzuki E, Mochizuki M. Possible important role of urinary N-methyl-N(3-carboxypropyl) nitrosamine in the induction of bladder tumors in rats by N-methyl-N-dodecylnitrosamine. Gan 1976; 67:771-772. (9) ANDREWS AW, LIJINSKY W. Mutagenicity of 45 nitrosamines in Sal
(1) LIJINSKY W, SAAVEDRA J, REUBER MD. Induction of carcinogenesis
ABSTRACT—Friend erythroleukemia cells (FL cells) derived from DBA/2 mice may be induced to differentiate in vitro by addition of dimethyl sulfoxide (DMSO) to the culture medium. Transglutaminase (glutaminyl-peptide y-glutamyltransferase, EC 2.3.2.13) (TGase) activity was detected in the lysates of uninduced FL cells and was markedly increased in DMSO-treated cells. DMSO induced TGase activity of differentiating FL cells in a dose-dependent manner over the concentration range 0-280 mm. The increase in TGase activity was observed after 1 day's cultivation of the cells with 280 MM DMSO and continued to 4 days. Another differentiation inducing agent, butyric acid (1.4 MM), was as effective as DMSO in enhancing TGase activity in FL cells. Treatment of controls, a lymphoma cell line and mouse erythrocytes, with DMSO or butyric acid had no effect. These results suggest that the increase in TGase activity in DMSO-or butyric acid-treated FL cells may be related to cell differentiation.- JNCI 1983; 70:965-969.
When grown in tissue culture, FL cells may be induced to differentiate along the erythroid pathway by addition of DMSO (1) to the growth medium or by various other agents (2). Treatment of FL cells with inducing agents leads to a gradual arrest of cell division (3-5), which is accompanied by chromatin condensation (6) and a decrease in cell volume (7). Several biochemical markers have been identified after induction, including the accumulation of globin mRNA (813) and globin chain synthesis (14, 15), increases in heme synthesis (6, 16), synthesis of carbonic anhydrase (17), appearance of erythrocyte-specific membrane antigens (18, 19), and membrane proteins such as spectrin (20, 21), and
YAC-1, a lymphoma cell line, was used as described elsewhere (27). Cells were cultured in RPMI-1640 medium containing 10% fetal bovine serum (Reheis Chemical Co., Kankakee, III.), 100 U penicillin/ml and 100 ug streptomycin/ml (Flow Laboratories, Rockville, Md.) at 37°C in a 95% air-5% CO2 atmosphere. Cells were plated at 1X10%/ ml. Recultured cells or cells treated with DMSO or butyric acid (Sigma Chemical Co., St. Louis, Mo.) were obtained from previous culture of exponentially growing cells. TGase activity assay. — TGase was assayed essentially as described (28) by incorporating [H]putrescine (New England Nuclear Corp., Boston, Mass.) into N,N-dimethylated casein (Calbiochem-Behring Corp., La Jolla, Cal.) in the presence of Ca2+. The cell suspensions were washed once with 50 mM Tris-HCl buffer (pH 7.5) containing 0.15 M NaCl. The pellet was resuspended in the same buffer at a final concentration of 5.0x10' cells/ml and then frozen and thawed twice. The assay mixture (at 37°) consisted of 10 ul CaCl2 (50 mm), 20 ul N,N'-dimethylated casein (2%), 10 ul [H]putrescine (12 mM, sp act 2 uCi/120 nmol), and 40 ul cell lysate. After various times (0-90 min), 10-ul aliquots were placed on 1-cm? Whatman 3MM filter paper and treated with 10% trichloroacetic acid. The filter papers were washed with 5% trichloroacetic acid followed by an ethanolacetone (1:1) mixture and then by acetone, and the radioactivity of the insoluble material was measured in a Packard liquid scintillation counter. We calculated the specific activity of TGase by subtracting the zero-time control from the experimental value. One unit of enzyme activity is defined as nanomoles of putrescine incorporated per 30 minutes with 0.4 mg N,N'-dimethylated casein as acceptor substrate. Protein determination. — The procedure of the Bio-Rad protein assay was used. The principle of this dye-binding method was first described by Bradford (29). Hemoglobin determination.—Cells were collected and washed with a phosphate-buffered saline solution (pH 7.5) and lysed by freezing and thawing 5x10' cells/ml. Following centrifugation at 20,000Xg for 30 minutes, the supernatant was collected and assayed for hemoglobin content as described (30). The 20 ul supernatant was mixed with 0.2 ml 1% benzidine solution in 90% acetic acid and 0.2 ml freshly prepared 1% hydrogen peroxide. After incubation at room temperature for 20 minutes, 2 ml 10% acetic acid was added
TGase (glutaminyl-peptide Y-glutamyltransferase, EC 2.3.2.13) catalyzes an acyl transfer reaction between the ycarboxamide groups of protein-bound glutamine and primary amino groups. When the amine is the e-amino residue of peptide-bound lysine, intermolecular or intramolecular €-(y-glutamyl) lysine isopeptide cross-links are formed. TGase enzymes exist widely in nature and have been shown to perform several biologically important functions, i.e., the stabilization of the fibrin structure during blood coagulation (23) and the production of the vaginal plug by postejaculatory clotting of rodent seminal plasma (24). Work on human erythrocytes (25, 26) has indicated that TGase plays a role in Ca2+-mediated cross-linking of erythrocyte membrane proteins. The studies reported here demonstrate a marked increase in TGase activity during erythroid differentiation of FL cells. This phenomenon may be related to stabilization of cell membranes by cross-linking membrane proteins.
ABBREVIATIONS USED: DMSO=dimethyl sulfoxide; FL=Friend erythroleukemia; TGase=transglutaminase (glutaminyl-peptide y-glutamyltransferase, EC 2.3.2.13).
Cell lines. —GM 979, a Friend virus-infected leukemia cell line derived from DBA/2 ice, was obtained from G. H. Lyman, University of South Florida College of Medicine.
Received September 10, 1982; accepted December 28, 1982. Department of Medical Microbiology and Immunology, University of South Florida College of Medicine, Tampa, Fla. 33612.
TGase activity was present in cell lysates of cultured FL cell line GM 979 as determined by assaying the incorporation of [H]putrescine into N,N’-dimethylated casein in the presence of Ca2+. The enzyme activity in GM 979 cells at exponential growth was in the range of 4 to 6 U/mg protein. The TGase increased eightfold to fifteenfold in cells cultured for 4 days in medium containing 280 mM DMSO. Enzyme activity was not detected in intact cell preparations but was detected only in cell lysates. Without addition of Ca2+, a time dependent, low level of [H]putrescine incorporation could be detected. However, upon addition of 1 mM EDTA in the assay mixture, the incorporation of [H]putrescine into N,N'-dimethylated casein was no longer detectable. The ability of several amines to inhibit DMSO-induced TGase activity was examined in GM 979 cells. The synthetic amine, cystamine, has been found capable of inactivating TGase (31). 1,5-Pentanediamine, histamine, methylamine, and ethylamine, as well as unlabeled, cold putrescine, are known to serve as substrates and compete with [H]putrescine for y-glutamyl acceptor binding sites. As shown in textfigure 1, all of the amines tested inhibited [H]putrescine incorporation. Cystamine was the most potent inhibitor, inhibiting 50% of the total enzyme activity at a concentration of 0.3 mm; 100% inhibition was obtained with a concentration of 5.8 mM. Histamine was as effective as putrescine in inhibiting [H]putrescine incorporation, followed by 1,5-pentanediamine, methylamine, and ethyla
DMSO ( mm ) TEXT-FIGURE 2.– TGase activity in FL cells treated with DMSO at differ ent concentrations. FL cells were cultured with DMSO at different concentrations for 3 days (O), or 4 days followed by measurement of TGase activity in cell lysates.
mine, which were less active than histamine in competing with putrescine for acceptor binding sites on N.V"-dimethylated casein. DMSO induced TGase activity in a dose-dependent manThe enzyme activities detected in FL cells treated with DMSO for 3 or 4 days showed a linear relationship with DMSO concentration over the range of 0-280 mM (text-fig. 2). Four-day treatment resulted in consistently higher enzyme activity than that achieved by 3-day treatment. DMSO concentration higher than 280 mM was not tested because of its toxic effects. Birckbichler et al. (32) reported that enhanced TGase activity correlates with the nonproliferating state of cultured cells. To distinguish whether the increase in TGase activity in DMSO-treated FL cells was associated with cell differentiation or cell proliferation, we established growth curves of GM 979 in the presence or absence of DMSO and timedependent TGase activities in these cultured cells. As shown in text-figure 3, the increase in TGase activity was observed after 1 day's cultivation of GM 979 in 280 mM DMSO, and it continued to 4 days. During the period of exponential growth, days 1-3 after culturing, enzymatic activity of untreated GM 979 was low and relatively constant. On the 4th day, when cells ceased growing, TGase activity increased 2.3-fold as compared to 1 day before. This observation is consistent with the finding of Birckbichler et al. (32) that TGase activity is increased in nonproliferating cells. However, a 6.5-fold higher activity was found in DMSO-treated cells when compared with untreated, 4-day cultured cells. This result suggests that DMSO-induced TGase activity is more likely associated with cell differentiation than with cell proliferation. This hypothesis was further supported by the finding that another differentiation inducing agent, i.e., butyric acid, was as effective as DMSO in enhancing TGase activity in FL cells. Also, 1.4 mM butyric acid induced GM 979 cell differentiation. Cultivation of GM 979 cells for 3
CONCENTRATION (mm) TEXT-FIGURE 1.-Inhibition of TGase in FL cell lysates. Log dose-response curves for inhibition of DMSO-induced TGase in FL cells by cystamine Page 10
The role of TGase in erythroid differentiation requires further investigation. The continuous increase in TGase activity during exponential growth of DMSO-treated FL cells showed no direct correlation to the nonproliferating state of the cells. However, DMSO- or butyric acid-treated FL cells, which are committed to terminal differentiation, eventually lost their proliferative capacity. It is possible that the higher TGase activity in differentiating FL cells is related to the stabilization of cell membranes by crosslinking membrane proteins which, in turn, leads to an irreversible arrest of cell division. Alternatively, the enzyme's role may be to post modify membrane proteins or other proteins with small metabolites such as polyamines and provide signals leading to the terminal nonproliferating state of the cells. It has been reported that small amounts of putrescine and spermidine are bound covalently to proteins through the catalytic reaction of TGase in human blood lymphocytes after treatment with mitogens (33). The presence of the erythrocyte membrane-associated protein spectrin in FL cells has been reported. Spectrin increased tenfold to twentyfold after exposure of the cells to DMSO. The intracellular concentration of spectrin reached a peak on the 3d day of DMSO treatment, after which levels fell to those found in mouse erythrocytes (21). The synthesis of transmembrane glycoprotein band 3 was also reported in FL cells treated with DMSO (22). A greater than tenfold increase was observed, and maximal synthesis was attained 3-4 days after the beginning of induction. Results from experiments with Ca't-loaded human erythrocytes (26) demonstrated that the Ca2+ can switch on the latent TGase in erythrocytes, which causes the covalent fusion of membrane proteins by y-glutamyl-e-lysine bridges. The formation of membrane protein polymers was accompanied by the disappearance of band 4.1 and a reduction in the amounts of proteins in the spectrin and band 3 regions (26). The involvement of spectrin and band 3 in TGase-catalyzed reactions in differentiating FL cells is a viable possibility. Further analyses of these complex metabolic changes appear warranted. Nevertheless, it appears from the results of this study, as well as those based on analysis of different biochemical markers, that differentiation of leukemia cells in vitro by agents such as DMSO is reflected by major changes of enzyme and membrane protein activities.
vitro of murine (Friend) virus-induced leukemic cells. Natl Cancer Inst Monogr 1966; 22:504-514. (7) Loritz F, BERNSTEIN A, Miller RG. Early and late volume change during erythoid differentiation of cultured Friend leukemic cells. J Cell Physiol 1977; 90:423-438. (8) Aviv H, Volch Z, Bastos R, LEVY S. Biosynthesis and stability of globin mRNA in cultured erythroleukemia Friend cells. Cell 1976: 8:495-503. (9) CONKIE D, AFFARA N, Harrison PR, Paul J, JONES K. In situ localization of globin messenger RNA formation. II. After treatment of Friend virus-transformed mouse cells with dimethyl sulf. oxide. J Cell Biol 1974; 63:414-419. (10) Curtis PJ, Mantei N, VAN DEN Berg J, WEISSMANN C. Presence of a putative 15$ precursor to B-globin mRNA but not to a-globin mRNA in Friend cells. Proc Natl Acad Sci USA 1977; 74:3184 3188. (11) Nudel U, SALMON J, FIBACH E, et al. Accumulation of a- and B globin messenger RNAs in mouse erythroleukemia cells. Cell 1977; 12:463-469. (12) Ostertag W, Crozier T, Kluge N, Melderis H, DUBE S. Action of 5-bromodeoxyuridine on the induction of haemoglobin synthesis in mouse leukemia cells resistant to 5 BUDR. Nature (New Biol] 1973; 243:203-205. (13) Ross J, Ikawa Y, LEDER P. Globin messenger-RNA induction during erythroid differentiation of cultured leukemia cells. Proc Natl Acad Sci USA 1972; 69:3620-3623. (14) BOYER SH, Wuu KD, Noyes AN, et al. A. Hemoglobin biosynthesis in murine virus-induced leukemic cells in vitro: Structure and amounts of globin chains produced. Blood 1972; 40:823-835. (15) Ostertag W, MELDERIS H, STEINHEIDER G, KLUGE N, DUBE S. Synthesis of mouse hemoglobin and globin mRNA in leukemia cell cultures. Nature (New Biol] 1972; 239:231-234. (16) Ebert PS, Ikawa Y. Induction of S-aminolevulinic acid synthetase during erythroid differentiation of cultured leukemia cells. Proc Soc Exp Biol Med 1974; 146:601-604. (17) KABAT D, SHERTON CC, Evans LH, BIGLEY R, KOLER RD. Synthesis of erythrocyte-specific proteins in cultured Friend leukemia cells. Cell 1975; 5:331-338. (18) FURUSAWA M, IKAWA Y, SUGANO H. Development of erythrocyte membrane-specific antigens in clonal cultured cells of Friend virus induced tumor. Proc Jpn Acad 1971; 47:220-225. (19) MacDonalD ME, LETARTE M, BERNSTEIN A. Erythrocyte membrane antigen expression during Friend cell differentiation: Analysis of two non-inducible variants. J Cell Physiol 1978; 96:291-301. (20) ARNDT-Jovin DT, Ostertag W, Eisen H, Klink F, Jovin TM. Studies of cellular differentiation by automated cell separation. Two model systems: Friend virus transformed cells and Hydra attenuata. J His tochem Cytochem 1976; 24:332-347. (21) Eisen H, Bach R, EMERY R. Induction of spectrin in erythroleukemic cells transformed by Friend virus. Proc Natl Acad Sci USA 1977; 74:3898-3902. (22) Sabban EL, Sabatini DD, MARCHESI VT, ADESNIK M. Biosynthesis of erythrocyte membrane protein band 3 in DMSO-induced Friend erythroleukemia cells. J Cell Physiol 1980; 104:261-268. (23) LORAND L. Fibrinoligase: The fibrin-stabilizing factor system of blood plasma. Ann NY Acad Sci 1972; 202:6-58. (24) Williams-Ashman HG, Notides AC, Pabalan SS, LORAND L. Trans amidase reactions involved in the enzymatic coagulation of semen: Isolation of y-glutamyl-e-lysine dipeptide from clotted secretion protein of guinea pig seminal vesicle. Proc Natl Acad Sci USA 1972; 69:2322-2325. (25) Anderson DR, Davis JL, CARRAWAY KL. Calcium-promoted changes of the human erythrocyte membrane. Involvement of specirin, transglutaminase, and a membrane-bound protease. J Biol Chem 1977; 252:6617-6623. (26) SIEfRING GE Jr, APOSTOL AB, Velasco PT, Lorand L. Enzymatic
(1) Friend C, Scher W, HOLLAND JG, Sato T. Hemoglobin synthesis in murine virus induced leukemic cells in vitro: Stimulation of erythroid differentiation by dimethyl sulfoxide. Proc Natl Acad Sci USA 1971; 68:378-382. (2) Marks PA, RIFKIND RA. Erytholeukemic differentiation. Annu Rev Biochem 1978; 47:419-448. (3) FIBACH E, REUBEN RC, RIFKIND RA, Marks PA. Effect of hexa methylene bisacetamide on the commitment of differentiation of murine erythroleukemia cells. Cancer Res 1977; 37:440-444. (4) GUSELLA JF, Housman D. Induction of erythroid differentiation in vitro by purines and purine analogues. Cell 1976; 8:263-269. (5) Preisler HD, LUTTON JD, Giladi M, Goinstein K, ZANJANI ED. Loss of clonogenicity in agar by differentiating erythroleukemia Page 11
and the cat as in man. When only rates for all leukemias in entire dogs and cats were compared, there was an approximate range of 20% more male expression in the dog to a 30% excess expression in the cat (table 5). Human subpopulations in the United States also show variable percent excesses of male over female age-adjusted leukemia rates, some in the same ranges as those in the cat and the dog, but most falling between 40 and 60% or approximately two times the dog to cat excess percentages (13). It has been hypothesized that the major reason for excess human male cancer at most sites was a more efficient immunologic system in human females (14). Female dogs and cats also appear better able to survive in our environment than their male counterparts, for in both species, as in humans, the survival gap consistently widens between male and female with increasing age (10). Moreover, in an unpublished analysis of mortality data for dogs and cats collected in the two population surveys used in this study, females had lower age-adjusted, disease-only mortality rates among both neutered and entire groups than their male counterparts. In contrast to humans, dogs, and cats, other species for which there have been sufficient data available tend to have excess expression of leukemia in the female. These species are the mouse (15), rat (16), guinea pig (17), horse (7), and cow (7). Thus the leukemia patterns of humans, dogs, and cats differ in sex expression preference from the other named species. The reasons for this difference may be important in understanding the etiology and/or pathogenesis of leukemia and why sex hormones can affect risk. The two most obvious differences are physical environment and diet. The dog and the cat tend to share closely the same physical environment of man, and some aspects of this environment may relate to male leukemia excess. It is also possible, however, that diet may be more important, in that humans, dogs, and cats eat meat and animal products as a major dietary source, whereas the other species do not. Perhaps animal-source steroid or protein metabolism is in some way involved. In addition to male-female differences, effects of sex hormone removal via neutering were also present (text-figs. 4, 5). The effects of neutering appeared to be related partially to species, partially to sex, and partially to the age when neutering occurred. Species and sex differences were found in the female only. In the males of both species, entire male age-adjusted rates were approximately 60% more than neutered rates, whereas for the neutered female compared to the entire female, there were marked differences between dogs and cats (tables 4, 5). There was little or no change due to neutering in the female dog, where neutered expression was almost the same as the entire expression. However, in the female cat, neutering reduced expression by approximately 40-50% of the entire female age-adjusted rate. The reason for these female differences was not clear and should be pursued further, especially since males of both species showed the same magnitude of reduced risk from neutering. Two areas that might be fruitful for study are the differences in estrous cycles between species and the effect of pregnancy, which would generally occur more frequently in cats. Neutering in three strains of mice (15) resulted in higher leukemia expression in those neutered, which was exactly opposite to the results shown in this paper for the dog and
the cat. Neutered male horses and cows (7) also had excess leukemia expression as was found in the mouse. Thus there appears to be a consistency of opposite neutering as well as opposite male-female effects in the dog and the cat and perhaps man also, when compared to the effects seen in rodents and herbivores. In addition, age at neutering appears to be important in the neutered risk effect for both males and females (table 6). It is obvious that the major part of the reduced risk effect of neutering in the feline female and male and the canine male was due to neutering before maturity (table 6). Interestingly, there was a definite pattern of increasing and decreasing risk to the ratios obtained when prior age-adjusted, ageneutered leukemia incidence rates were divided by subsequent age-adjusted, age-neutered incidence rates (table 7). The same pattern was present in the female regardless of species. Thus even though neutering did not affect the leukemia expression of the female dog, age at neutering showed a similar pattern of risk effect as seen in the female cat, where neutering reduced leukemia expression by approximately 40-50% (table 5). With a slight reduction in the 4- to 8-year incidence rate of the neutered male cat, a similar ratio pattern could also be seen in males of both species, where the effect of neutering was the same (table 7). Further work is necessary to gain an understanding of the pattern of increasing and decreasing risk effect of age at neutering. In summary, age and sex preferences of leukemia expression in the cat and the dog resemble those in humans. The sex preference contrasts with what has been previously found in some rodent and herbivore species. In addition, neutering reduced risk markedly in the female cat but less so in the male cat and dog and did not appear to affect risk in the female dog. However, the choice of the age of neutering and the sex had a profound interacting effect on whether neutering decreased, increased, or did not change expression in both the cat and the dog.
(1) PRIESTER WA, MANTEL N. Occurrence of tumors in domestic animals. Data from 12 United States and Canadian colleges of veterinary medicine. J Natl Cancer Inst 1971; 47:1333-1344. (2) Priester WA. Skin tumors in domestic animals. Data from 12 United States and Canadian colleges of veterinary medicine. J Natl Cancer Page 12
rural areas where hygienic conditions are poor. Also, many patients reported multiple female partners. As already noted, carcinoma of the penis is prevalent in areas where there is also a high incidence of carcinoma of the cervix. Preliminary studies by Dr. Jenson suggest an association with the papillomavirus, which is not surprising in view of the epidemiologic data.
Dr. Helena de Restrepo reviewed information on lung cancer noting that in Latin America the incident rates are lower than in developed countries. As in other areas, however, lung cancer is associated with the use of cigarettes and alcohol. No differences were found, however, between the incidence of lung cancer and the different types of tobacco that are used. In Colombia, carcinomas of the larynx are more common than carcinomas of the lung. Dr. Rei T. Ravenholt indicated that cigarette smoking has caused the major health problems in the late 20th century.
Lymphomas Dr. Gregory T. O'Conor discussed lymphomas and the need for uniformity in classification and in diagnostic criteria. He emphasized that the lymphomas and lymphoid leukemias represent a group of distinct biologic entities and should not be regarded as a single disease with morphologic variations. He noted that much of the available incidence data such as those reported in previous editions of "Cancer Incidence in Five Continents” were based on inadequate or very limited classification systems. Dr. O’Conor highlighted opportunities for epidemiologic studies of lymphoid neoplasms, since it is already evident that there are large differences in the geographic distribution of many types of lymphomas and leukemias and there are often special characteristics that may provide clues to etiology. Important examples are Burkitt's lymphoma in Africa and T-cell lymphoma in Japan. Some observations require further epidemiologic investigation. With particular reference to Latin America, they include the geographic and ethnic variations in incidence and age distribution of different types of acute lymphoblastic leukemia, the changing time trends for specific lymphomas and leukemia subtypes, the deficiency of nodular lymphomas and the excess of extranodal lymphomas in developing countries, the high frequency of nasal lymphoma in Peru, and the occurrence of adult T-cell lymphoma in the Caribbean area and its relationship to human T-cell leukemia virus. There is every indication that a better definition of lymphomas and leukemias, based on modern classification systems utilizing functional parameters as well as morphology, offers an opportunity to gain new insights into the natural history and potential control of this important group of neoplasms. Dr. Daniel Knowles described the usefulness of cell markers for the diagnosis of lymphoproliferative cancers and emphasized those that can be measured in any pathology laboratory. He also described the antigenic changes that occur during lymphocyte differentiation. Cell marker analysis can distinguish benign, reactive lymphoid proliferations (which are immunologically polyclonal) from neoplastic lymphoid proliferations (which are immunologically monoclonal). In many cases, these studies can also determine the cell origin of the lymphoid cancers. Clinicopathologic correlations with therapeutic and prognostic implications, not always possible by morphologic criteria alone, have been a beneficial result from these studies. The panel of cell markers which are most helpful consists of the immunofluorescent demonstration of la antigens and surface immunoglobulin for B-cells and sheep erythrocyte rosette formation for T-cells. This panel can divide lymphoid proliferations into 4 categories-polyclonal, monoclonal Bcell, T-cell, and non-B-, non-T-cell.
The session on cancer and nutrition was chaired by Dr. Maurice E. Shils who indicated that not only is the relation between diet and cancer complex, but also the long-term influence of diet makes research difficult. In reference to intervention programs, he concluded that there is no real evidence as to what dietary changes should be made in South America to reduce cancer. Diets can affect the incidence of cancer in several ways. These include ingestion of carcinogens, conversion of certain substrates to carcinogens in the body, interference with the inactivation or transport of carcinogens, affecting promotion of cells already stimulated, and overnutrition. At least four substances (ß-carotene, vitamin C, selenium, and a-tocopherol) are known to ameliorate the effects of carcinogens. Despite their diverse structure, these substances have one common attribute: They are all reducing agents. Dr. Evan A. Stein discussed the relation between serum cholesterol and cancer. Although serum cholesterol levels are related to the overall fat intake, there is an apparent inverse relationship between serum levels and mortality from cancer. Found in most epidemiologic studies, this relationship seems paradoxical in view of the association between high-fat diets and cancer. At present, the cholesterol-cancer relationship is still controversial, since there is little evidence to support more than an association. Although most of the epidemiologic data that relate lipids to cancer have come from North America and Western Europe, little work has been done in South America, where coronary artery disease and serum lipid levels are lower. Dr. Alfredo Lopez presented preliminary data from two case-control studies which showed that decreased vitamin A and ß-carotene serum levels correlate with an increased incidence of lung cancer in Louisiana. Vitamins E and C were also reduced in many patients. These data supported a theme prevalent at the Conference that nutritional deficiencies may not cause cancer directly, but they may increase the susceptibility to naturally occurring carcinogens. According to Dr. Lopez,' borderline deficiencies of several vitamins may be as deleterious as a severe deficiency of only one. Since diets in the United Stated are different from those in South America, it seems that collaborative simultaneous studies on various populations could yield important information. At the conclusion, Dr. Jorge Litvak, on behalf of The Pan American Health Organization, thanked all participants for their contribution. He emphasized that the Organization, as part of its plans to expand research on cancer epidemiology in Latin America, will assist investigators in both North and South America by arranging for contacts, preparing research proposals, encouraging investigators to submit applications to the Organization's research grant program, and assisting investigators to obtain funds from other agencies. Page 13
In “The Causes of Cancer: Quantitative Estimates of Avoidable Risks of Cancer in The United States Today,” by Richard Doll and Richard Peto, page 1197, entry in table 1 should read: “... each contributing less than 4% of deaths” (not“. . . each contributing less than 3% of deaths”).
sium to be held on September 26-30, 1983. The main topics of the conference will be: cyclic AMP, cyclic GMP, and calcium as signals; control of intermediary metabolism by insulin; neurochemistry; steroids and gene expression; control of cell multiplication by growth factors; signals in development and differentiation. Detailed information on the program, registration, etc., are available from Dr. J. E. Dumont, Institute of Interdisciplinary Research, Faculty of Medicine, Free University of Brussels, Campus Hôpital Erasme, 808 route de Lennik, B-1070 Brussels, Belgium.
Fourth International Congress of Dermatologic Surgery
The Journal will publish notices of major meetings dealing with cancer. Submission of notices 6 months in advance is requested to permit timely publication.
New Educational Institution—The European School of Oncology
This conference, organized by the International Society for Dermatologic Surgery and the Department of Medical and Surgical Dermatology of the Granada University, will be held in Granada, Spain, on October 25, 1983. The ISDS brings dermatosurgeons together so they may exchange innovative techniques and work for standardization and improvement of training programs in dermatologic surgery. For registration forms and additional information please contact Dr. V. Delgado Florencio, OTECSA, Recogidas, 21-1o, Granada, Spain.
Third International Symposium on Medical Virology
The European School of Oncology, promoted by the European Society of Surgical Oncology, will present postgraduate training in the field of clinical oncology. The following is a program of courses being offered for 1983: Breast Cancer; Dr. Veronesi (Milan) Chairman; course date May 30 June 8, 1983. Principles of Medical Oncology; Dr. Pinedo (Amsterdam) Chairman; course date June 12-19, 1983. Malignant Melanoma; Dr. Levene (London) Chairman; course date October 24-30, 1983. Genito-urinary Tract Tumors; Dr. Denis (Brussels) Chairman; course date November 7-13, 1983. Colo-rectal Cancer; Dr. Goligher (Leeds) Chairman; course date No vember 21-27, 1983. All the courses will be held at the Pomerio Castle, 40 minutes from Milan. English will be the official language. Additional information and requests for participation should be addressed to the European School of Oncology, Via Venezian, 1, 20133 Milan, Italy (telephone: (02)294662).
The Medical Microbiology Division, Department of Pathology, University of California, Irvine, Medical Center will sponsor this symposium to be held at the Disneyland Hotel, Anaheim, California, on October 19-21, 1983. This is an approved program of Continuing Medical Education. Physicians attending this course may report up to 18-34 hours for formal (Category I) credit toward the California Medical Association Certificate in Continuing Medical Education and the American Medical Association Physician Recognition Award. For further information contact Dr. Luis M. de la Maza, Department of Pathology, University of California, Irvine, Medical Center, Orange, California 92668 (telephone: 714-634-6868).
Second International Conference on Safety Evaluation and Regulation of Chemicals
Design and Analysis of Scientific Experiments Course
The Massachusetts Institute of Technology will offer this one-week course to be held on July 11-16, 1983. Applications will be made to the physical, chemical, biological, medical, engineering, and industrial sciences, and to experimentation in psychology and economics. The course will be taught by Professors Harold Freeman and Paul Berger. Further particulars may be obtained by writing to the Director of Summer Session, Room E19-356, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139.
This course, sponsored by the American College of Toxicology, organized by Boston University School of Medicine, Bio-Research Institute, Cambridge, and Mallory Institute of Pathology, Boston, City Hospital, will be held at the Hotel Sonesta, Cambridge, Massachusetts, on October 2428, 1983. The first two days of this conference will focus on the safety evaluation and regulation of chemicals as viewed by representatives of state and federal governments and industry; procedures for compliance with regulations and the testing required will also be included. The last three days (International Conference) will deal with regulations of drugs, foods, cosmetics, and toxic substances throughout the world from the point of view of government and international agencies, and with method development for safety evaluation. The Academy of Toxicological Sciences has approved accreditation for AMA Category I Continuing Medical Education credit. For additional information please contact Dr. Freddy Homburger, Bio-Research Institute, 9 Commercial Avenue, Cambridge, Massachusetts 02141 (telephone: 617864-8735).
Eighth European Symposium on Hormones and Cell Regulation
Sainte-Odile (near Strasbourg), France, will be the site of this sympo
Fifth World Congress of Cryosurgery
The Philippine International Convention Center, Manila, Philippines, will be the site of this conference to be held on November 17-19, 1983. The program will include cryobiology, cryoimmunology, cryotherapy with multi-disciplinary approach, and cryopreservation in addition to scientific sessions. Registration rates will increase after July 31, 1983. Abstracts should be sent to Dr. Sajio Sumida, The National Fukuoka Central Hospital 2-2, Jonai, Chuoku, Fukuoka 810, Japan (telephone: 092-714-0151) for receipt no later than May 31, 1983. For additional registration information please contact Dr. Samuel M. Tanchoco, Suite 31, Lagaspi Towers, 300 Roxas Blvd. corner Vito Cruz, Manila, Philippines (telephone: 59-40-80/59-5737).
Inn, Newport, Rhode Island, on August 15-18, 1983. The symposium will focus on the use of digital imaging, NMR, computed tomography, ultrasound, nuclear medicine, angiography, and diagnostic imaging in diagnosing adult and pediatric diseases and problems of the gastrointestinal, reproductive, genitourinary, skeletal, thoracic, and neurologic systems. Invasive and noninvasive techniques and the future of the imaging modalities will be discussed in lectures and small group workshops. Accreditation has been approved for 21-1/2 hours, Category I, AMA Physician's Recognition Award. For additional information contact Educational Resources Associates, Inc., P. O. Box 369, Brookline, Massachusetts 02146 (telephone: 617-738-8859/8861).
Safety Management Techniques Course
Fourth International Conference on the Adjuvant Therapy of Cancer
The National Safety Council will sponsor this course to be held in Dayton, Ohio, on September 12-15, 1983. For additional information contact Mr. James C. Schaffner, Director, Dayton/Miami Valley Safety Council, 1980 Winters Bank Tower, Dayton, Ohio 45423 (telephone: 513-226-1444).
The University of Arizona Cancer Center will sponsor this conference to be held at the Tucson Convention Center, Tucson, Arizona, on March 21-24, 1984. Drs. Stephen E. Jones and Sydney E. Salmon will serve as conference co-chairmen. The deadline for submission of abstracts (prepared in the format of AACR/ASCO) is November 1, 1983. For abstract forms or further information please contact Ms. Mary Humphrey, Conference Coordinator, University of Arizona Cancer Center, Tucson, Arizona 85724 (telephone: 602-626-6044).
Second International Congress of the T. and L. de Beaumont Bonelli Foundation for Cancer Research
“The Role of Viruses in Human Cancer" will be the subject of this meeting to be held in Naples, Italy, on September 22-24, 1983. The purpose of the conference is to provide current results ard perspectives in main areas in which viruses have been potentially associated with human malignancies. Topics will include: hepatitis B virus, herpesviruses, papovaviruses and retroviruses epidemiology, etiology, molecular biology, diagnosis, and prevention. For additional information contact Dr. Gaetano Giraldo, V. Le Elena 17/B, 80122 Naples, Italy (telephone: (39). (81)-685.013 - 684.047).
Second International Conference on Malignant Lymphoma
American College of Physicians Medical Oncology Review Course
Lugano, Switzerland, will be the site of this conference to be held at the Palazzo dei Congressi on June 13-16, 1984. This conference will deal with all important approaches in the diagnosis and treatment of Hodgkin's and non-Hodgkin's lymphoma in children and adults. Most lectures will be given by invited speakers. Selected controversial issues will be discussed by panels of leading experts. The program will include free presentations and a poster session. The deadline for receipt of abstracts is January 15, 1984. For abstract forms and further information please contact Dr. F. Cavalli, Head, Division of Oncology, Ospedale San Giovanni, 6500 Bellinzona, Switzerland.
Honolulu, Hawaii, will be the site of this course to be held on October 17-22, 1983. Dr. Thomas C. Hall will be the course director. For further information contact Ms. Maxine Topping, Postgraduate Division, American College of Physicians, 4200 Pine Street, Philadelphia, Pennsylvania 19104 (telephone: 215-243-1200 or 800-523-1546).
Fourth Annual Symposium on Diagnostic Imaging
Second International Conference on Human Tumor Markers
The Section on Radiation Medicine of Brown University School of Medicine and the Department of Diagnostic Radiology of Rhode Island Hospital will cosponsor this symposium to be held at the Sheraton-Islander
The International Society for Preventive Oncology (ISPO), the Austrian Cancer Society-Austrian Cancer League, and the Association of Clinical Scientists-USA will sponsor this conference to be held in Vienna, Austria, on February 20-22, 1984. The scientific program is designed for an update on the biology and clinical application of tumor markers. Panel discussions, poster presentations, and round table luncheons are devoted to the presentation of established and new tumor markers with emphasis on their sensitivity and specificity at various stages of tumor growth. Page 14
The cover illustration was suggested by a special report on diet, nutrition, and cancer that appears in this issue. (See paper by Palmer and Bakshi on pages 1151-1170.) in evaluating the role of diet in carcinogenesis, a committee of the National Academy of Sciences emphasizes the importance of including fruits, vegetables, and whole grain cereal products in the daily diet. (Illustration by Ms. Betty Hebb, Medical Arts and Photography Branch, National Institutes of Health]
Text: Send one original and two copies. Text-figures (graphs, charts, and line drawings): Send one original and two copies. Figures (photomicrographs and photographs): Send three original glossy prints, unmounted. Tables: Send tabular material in typewritten form; do not submit tables as glossy prints to be photographed. The entire manuscript including footnotes, legends, and references must be typed double spaced. Single-spaced portions of a manuscript will be returned to the author for retyping. The mailing address is: Journal of the National Cancer Institute, Westwood Building, Room 850, 5333 Westbard Avenue, Bethesda, Maryland 20816.
Manuscripts should be accompanied by a covering letter from the author responsible for correspondence about the manuscript. The covering letter should contain a statement that the manuscript has been seen and approved by all authors and that all persons listed as authors have made significant (not merely administrative, technical, or supervisory) contributions to the content of the paper. The letter should also state that the manuscript or substantial portions thereof have not been previously published nor are they under consideration for publication in another journal. Data too extensive for inclusion in a manuscript, but which may be of interest to other investigators, may be deposited with the National Auxiliary Publications Service (NAPS), c/o Microfiche Publications, P.O. Box 3513, Grand Central Station, New York, New York 10017. Authors using the NAPS facilities should indicate in a footnote how their additional data may be obtained.
Approximately one-half of all manuscripts submitted to the Journal are accepted for publication. Publication of manuscripts is contingent on transfer of copyright to the Journal. Forms for this purpose will be sent to authors when a manuscript has been approved for publication. All material published in the Journal is in the public domain; therefore, permission to reproduce any portions is not necessary. As a matter of courtesy, communication should be made to the author(s) concerned.
TITLE: Be brief (no more than 14 words, if possible), but for indexing purposes include all key words. The annual subject index is prepared directly from the titles. Titles that form complete sentences are not acceptable. Do not use abbreviations or chemical formulas in titles.
TABLES: Submit tabular material typed double spaced on a separate page. Computer print-outs of tables are acceptable only if they are double spaced and with satisfactory title, column headings, legends, and other textual material. Provide headings for all columns of a table and indicate units of measurement where applicable.
BYLINE: Supply the complete first name, middle initial, and last name of each author. Avoid use of nicknames. Academic degree is optional; however, if a degree is provided for one author, it should be provided for all authors on a manus nuscript. Include the name of the institution, address, and zip code for each author. Indicate address for reprint requests in a footnote on the byline.
TEXT-FIGURES: Refer to graphs, charts, and line drawings as textfigures. Make the abbreviations on text-figures identical to the abbreviations used in the text and legends. Symbols and lettering should be large enough to be legible if reduced. Glossy prints or original drawings in black ink are acceptable; orient vertically instead of horizontally.
ACKNOWLEDGMENTS: Financial aid in support of research should be recognized in a footnote. Aid in technical fields, animal care, photography, stenography, and other areas may be mentioned in a separate footnote. Indicate the title (Dr., Mr., Mrs., Miss, Ms.) and affiliation of each person acknowledged.
FIGURES: Refer to photomicrographs and photographs as figures. Figures and their legends should fit a 7.5. by 9-inch dimension. Calculate that nine typewritten lines of legend will require 1 inch of vertical space. Type figure legends on a separate sheet and identify in the legend any lettering used on the figure. Trim photographs to show only important areas. Do not number face of photographs; use the back or overlay to indicate the top and number of each. Printed indicators may be included on photographs; do not use hand-drawn indicators. Insert internal scale markers on all photomicrographs. Color photographs must be grouped as a single plate. For color reproduction, submit both color transparencies and high-quality black and white reproductions. The need for color must be documented before it will be considered.
ABSTRACT: Provide an abstract, 150 words or less, on a separate page. The abstract should describe only results of the work being reported. Speculation, historical background, or reference to published works will be deleted from abstracts. Any inbred strains, noninbred stocks, cell lines, or other living materials used in the work should be noted in the abstract.
TEXT: Avoid the use of laboratory jargon and abbreviations not in common usage. When possible, refrain from the use of abbreviations that form words (exceptions are PAGE, ALL). Whenever possible, write mathematical equations and chemical formulas on a single line; submit complicated chemical structures as artwork. Avoid repetition of data in the text when it is presented in tables or text-figures.
REFERENCES: Use any recent issue of the Journal as a style guide for reference lists. Number references in order of citation in both text and reference list. For periodicals, provide author(s), title of paper, journal, year, volume, and first and last page numbers. For books, give the author(s), title, name(s) of editor(s), location and name of publisher, year, pertinent chapter or section title, and page numbers. Follow Index Medicus for punctuation and abbreviations. Citations of papers in preparation or submitted for publication, unpublished observations, and personal communications should be noted parenthetically in the text. Do not include them in the reference list. The accuracy of all references is the responsibility of the authors and not of the editorial staff.
Use the following style sources: Webster's Third New International Dictionary and Dorland's Illustrated Medical Dictionary (25th ed) for spelling; Council of Biology Editors Style Manual (4th ed) for scientific abbreviations and symbols: Government Printing Office Style Manual for other abbreviations and symbols, punctuation, and hyphenation; Biochemical Nomenclature and Related Documents (IUB, 1978) for abbreviations of chemical substances; Enzyme Nomenclature, recommendations (1978) of the IUB for recommended names of enzymes and information on nomenclature of isozymes; The Merck Index (9th ed) for chemical compounds; “Standard Nomenclature for Inbred Strains of Mice: Seventh Listing" by Joan Staats (Cancer Res 1980;40: 2083-2128) for inbred mouse strains; “Standardized Nomenclature for Inbred Strains of Rats: Fourth Listing" by Joan Staats and Michael Festing (Transplantation 1973;16:221-245) for designations of inbred rat strains; Animals for Research (9th ed) prepared by the Institute of Laboratory Animal Resources, National Academy of Sciences, for designations of laboratory animal stocks and strains.
LETTERS TO THI EDITOR should pertain to recent articles in the Journal and should not contain new data unless relevant to the discussion of a particular paper. The authors addressed will be invited to reply to comments before correspondence is published. Letters will be edited by the Staff, subject to contributor approval.
FURTHER DETAILS on the preparation of manuscripts are set forth in the document “Uniform Requirements for Manuscripts Submitted to Biomedical Journals” prepared by the International Steering Committee of medical editors (see Am Rev Respir Dis 1979; 119:3-10/Ann Intern Med 1979;90:95-99/Br Med J 1979;1:532-535/ or Lancet 1979;1:428-430).
UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON, D.C.: 1983 Page 15
The possibility that murine lung adenomas may arise from two different cell types is an intriguing one. It assumes greater importance in connection with our continuing research, which deals with adenocarcinoma development and the Druckery effect. As Kauffman et al.' noted, papillary tumors may have malignant potential not possessed by solid tumors. Indeed, we have observed that large murine lung tumors induced by urethan appear invariably to be papillary in nature, whereas either set of morphologic characteristics may be dominant in small adenomas. ELLEN J. O'FLAHERTY, Ph.D. University of Cincinnati MICHAEL L. DOURSON, Ph.D. Environmental Protection Agency Cincinnati, Ohio 45268 Page 16
the possibility existed that IgM antibodies could be recognized as IgG by this reagent. However, a few serum samples from carcinoma patients had the highest levels of IgG antibodies to FA. Furthermore, certain serum samples in all groups were positive for IgM antibody and not for IgG antibody (table 3). To further substantiate these observations, we analyzed selected serum samples for IgG antibodies using a y-chain-specific enzyme-conjugated anti-human IgG reagent. Results presented in text-figure 2 reveal that the serum samples (#3 and #2) that were positive for IgG antibodies by heavy- and light-chain enzyme anti-IgG conjugate were also positive by y-chain enzyme anti-IgG conjugate. Both of these serum samples contained detectable levels of anti-FA of the IgM class. Furthermore, samples #1 and #4, which were negative by both IgG reagents, were positive for IgM antibodies. These results confirmed that certain serum samples did contain anti-FA antibodies of the IgG class and that in certain serum samples there was no association between the two classes of antibodies.
Distribution of FA Activity in Various Tumor and Normal Cells
attempted to develop ELISA to detect anti-FA antibodies in human sera and FA activity in various tumor and normal cells. Text-figure 1 shows that anti-FA antibodies of both the IgM and IgG classes could be detected in sera from melanoma patients by the indirect ELISA method. The incidence of anti-FA antibodies in sera from cancer patients with various histologic types of cancers and normal volunteers is listed in table 3. There were no obvious differences in incidence between the sera from normal volunteers and cancer patients. The incidence of IgM antibodies was significantly higher than the incidence of IgG antibodies in sera from melanoma and carcinoma patients and in normal volunteers. These differences were statistically significant by the McNemar's test (P<0.01). Analysis of the data by the odds ratio method revealed that there was no statistically significant association between the presence of IgM and IgG antibodies in serum samples from melanoma and normal volunteers (odds ratio of 0.71 for melanoma and 1.59 for normal volunteers). For the sarcoma group there was a positive association between the presence of IgM and IgG antibodies (odds ratio = 6.75), whereas for the carcinoma group there was a negative association (odds ratio - 0). Since the enzyme-conjugated anti-human IgG used in this series of experiments was against heavy and light chains,
A number of solid tumor and normal tissues obtained at biopsy or autopsy were analyzed for the presence of FA activity by competitive inhibition in ELISA. The tissues were finely minced and passed through a 60-mesh stainless steel screen. The cells were washed three times with 20-25 volumes of PBS/T and then used in the assay as described in “Materials and Methods." The FA activity was widely distributed in tumor tissues of various histologic types (text-fig. 3). The degree of inhibition in ELISA ranged from 2 to 100% for tumor tissues, 2 to 70% for apparently normal tissues of cancer patients, 0 to 25% for normal tissues of noncancer patients, and 8 to 100% for human fetal tissues. If one considers a 20% or greater inhibition (approximately, mean + 2 SD inhibition caused by normal tissues of noncancer patients) as a criterion for positive activity for the presence of immunologically similar antigen, 89% (23/26) melanoma, 73% (11/15) sarcoma, and 68% (13/19) carcinoma tumor tissues were positive for the FA activity. Among apparently normal tissues (which included tissues from skin, muscle, brain, and liver) from a melanoma, a sarcoma, and a carcinoma patient, 58% (7/12) exhibited FA activity. The presence of FA activity was not associated with one particular histologic type of tissue;
TABLE 3.- Incidence of anti-FA antibodies of IgG and IgM classes in sera from cancer patients and normal volunteers by ELISA“ with use of aqueous phase of CM-extracted spent culture medium of M14 cell line as the target antigen
Melanoma patient Sarcoma patient Carcinoma patient Normal volunteer
1:86+92 1:137+161 1:268+605 1:119-96
1:182-170 1:144+137 1:103271 1:117–102
Heat Stability of FA Heating of the FA fraction (0.1 mg protein/ml) at 100°C for 5 minutes did not affect the antigenic activity. The extent of inhibition in competitive ELISA by unheated and heated FA fractions in three replicates was 83£2.1% and 7723.2%, respectively.
o 20 80 320 1280 5120 1/ Serum Dilution TEXT-FIGURE 2.-Anti-FA level of IgM and IgG classes in selected serum samples from melanoma (#1 and #2) and carcinoma (#3 and #4) patients as assessed by indirect ELISA. Assay conditions are described in “Materials and Methods.” Three different alkaline phosphatase-conjugated second antibodies were used for each serum. Z, enzyme-conjugated u-chain-specific goat anti-human IgM; o, enzyme-conjugated heavy- and light-chain-specific goat anti-human IgG; &, enzyme-conjugated y-chain-specific goat anti-human IgG.
Susceptibility of FA to Various Enzymes Analysis of the FA-containing fraction for protein and carbohydrate revealed the presence of both components. Since the carbohydrate detected in the FA fraction could be associated with glycoproteins or other complex molecules, we treated the FA-containing fraction with various immobilized enzymes and lectins and then tested for its ability to compete for the anti-FA antibody in ELISA. Results presented in table 4 show that immobilized proteolytic enzymes (protease, trypsin, papain, and a-chymotrypsin), lipases (phospholipase A2, phospholipase C, and lipase), and neuraminidase did not affect the antigenic activity of FA. The extent of inhibition in competitive ELISA by the FA fraction treated with these enzymes ranged from 74.5 to 85.8% as compared to 87.3% by the untreated fraction. However, treatment with B-galactosidase and hyaluronidase reduced the inhibition significantly (12.4 and 8.0%, respectively, vs. 87.3% by the untreated fraction). These results suggested that the antigenic activity resided in the carbohydrate portion of the FA molecule.
rather, all four types of apparently normal tissues of the melanoma patient and some tissues of the carcinoma and sarcoma patients were positive for the FA activity. Among normal tissues (group V) that included five skin, five muscle, five brain, and five liver samples from 5 individuals who had no known history of cancer and who died from causes other than cancer, 5% (1/20) were positive for the FA activity. This FA-positive tissue was skin (group V in text-fig. 3), and the extent of inhibition in competitive ELISA was only 25% as opposed to 44-59% (mean) inhibition by malignant tissues and 30% by apparently normal tissues from cancer patients. Eighty-seven percent (13/15) human fetal tissues caused significant inhibition (52.9=27.6%) in ELISA (group VI in text-fig. 3). The fetal tissues included five brain, five skin, and five muscle samples. All five brain and all five skin fetal tissues caused greater than 20% inhibition (75.6–23.8% and 54.2£17.1%, respectively), whereas 60% (3/5) of fetal muscle tissues were positive for the FA activity with a mean I standard deviation of 26£15.9%. The FA activity was expressed by 83-91% of cultured tumor and human normal cell lines. Cells were grown as monolayers and harvested by being scraped. Harvested cells were washed three times, and 250 ul of packed cells were used for competitive inhibition in ELISA. Of 22 melanoma cell lines, 20 (91%) caused greater than 20% inhibition in ELISA, with a mean I standard deviation of 58.4–28.2%. Among sarcoma and carcinoma cell lines, 83% (5/6, 44.2=22.7% inhibition) and 88% (7/8, 47.5223.8% inhibition) were positive for FA, respectively. Ninety percent (9/ 10) of cultured normal fibroblast (muscle and skin) lines were positive for FA, with 56.3227.4% inhibition in ELISA. The normal fibroblast lines were long-term cultures.
TEXT-FIGURE 3.—Distribution of FA activity in various tumor and normal tissues as assessed by their ability to inhibit binding between allogeneic antibody (serum B) and FA fraction in ELISA (see "Materials and Methods”). The horizontal dotted line at 20% (approximately, mean + 2 SD of normal tissues) inhibition is drawn to separate FA-positive and FA-negative tissues. I, melanomas; II, sarcomas; III, carcinomas; IV, apparently normal tissues of cancer patients; V, normal tissues from individuals who had no history of cancer; VI, human fetal tissues. The normal tissues either from cancer patients or noncancer patients were from the skin, muscle, brain, and liver. The fetal tissue group was comprised of tissues from the brain, skin, and muscle-three each. Thin vertical line with small horizontal line in each tissue group represents mean I SD, respectively.
able to remove significant levels of FA activity as assessed by reduction in inhibition in competitive ELISA (table 5). Since ß-D-galactose is one of the sugars that binds with ricin, peanut, and soybean lectins and since Bandeirea simplifolia lectin, which is specific a-D-galactose, did not remove the FA activity, these results (table 5) suggested that the FA activity was removed by lectins that bind to the B-isomer of galactose. Thus B-D-galactose or its derivative appeared to be one of the key constituents of the determinant recognized by allogeneic antibody.
Figure 2 illustrates the analysis of the FA-containing fraction obtained from ricin-agarose affinity chromatography by gradient PAGE under nondissociating conditions. Under the experimental conditions with the use of basic buffer in the electrode chambers, most of the proteinaceous material remained in the high-molecular-weight region of the gel (fig. 2, gel A). To localize the antigenic activity, we washed corresponding unstained gels three times with PBS/ T and sliced them into 2-mm sections. Each slice was reacted with the allogeneic anti-FA serum B that was quantitatively adsorbed with lymphoblastoid cells of the M14 donor. After being washed four times with PBS/T, the slices were incubated with alkaline phosphatase-conjugated goat anti-human IgG for 30 minutes and then were washed again four times with PBS/T. The washed slices were transferred to new tubes containing 1.0 ml p-nitrophenyl phosphate (1 mg/ml) and incubated at room temperature. Aliquots (200 ul) were withdrawn at different intervals to determine absorbance at 405 nm. After an appropriate incubation time, absorbance of each sample was corrected for background noise by subtraction of the readings from corresponding slices of control gels. The results were plotted as A OD versus gel length. As seen in text-figure 5, most of the antigenic activity was associated with slices obtained from the upper portion representing the more than 400K-M, region. Such a slow rate of migration of proteinaceous material and antigenic activity into the gel under nondissociating conditions could be due either to large aggregates of molecules or to lack of predominance of anionic charges on the molecules. If the latter was true, then the isolation procedures, i.e., CM extraction and ricin-agarose affinity, would have isolated acidic or neutral proteins selectively, which was very unlikely. When the same material (ricin-agarose affinity-purified FA) was analyzed by SDS-PAGE (after treatment of the sample with 1% SDS and preequilibration of gels with 0.1%
Ricin Affinity Chromatography
The partially purified fraction of M14 spent culture medium obtained from the aqueous phase after CM extraction was subjected to ricin-agarose affinity chromatography. The material retained by the column was eluted with 15% D-galactose containing Tris-HCl buffer (pH 8.0). Each fraction was analyzed for FA activity in ELISA for competitive inhibition. As shown in text-figure 4, the desorption buffer eluted the FA activity from the column. The fractions
"FA activity was assessed by competitive inhibition in ELISA. * Each OD value represents an average of three replicates. *Percent reduction=(1-(% inhibition by lectin-treated FA/% inhibition by untreated FA)]x100. The untreated FA sample was mixed with agarose (Sepharose 4B) beads, incubated at room temperature for 30 min, and centrifuged at 7,000xg for 10 min. The agarose beads contained no lectins.
Recently, Irie et al. (8) produced two different anti-OFAI antibodies of monoclonal origin from Epstein-Barr virustransformed human B-lymphoid cells. These antibodies, designated as anti-OFA-I-1 and anti-OFA-1-2, recognized membrane-associated glycolipids that had different characteristics. Since initial serologic studies of the FA present in the M14 spent culture medium (5) showed immunologic similarities with OFA-I (6), FA fractions of M14 spent medium before and after ricin-affinity purification were tested in immune adherence (6,8) for their ability to react with anti-OFA-I-1 and anti-OFA-I-2. Neither of the two FA fractions was able to block reactivity between anti-OFA-I-2 and M14 cells in immune adherence. However, both fractions of FA blocked reactivity between anti-OFA-I-1 and M14 cells. The protein concentrations of FA fractions before and after ricin-affinity purification were 0.2 and 0.017 mg/ ml, respectively. The inhibition by ricin-affinity-purified FA fraction at 0.017 mg protein/ml was minimal. When tested at 1:10 dilution, no blocking was observed. Therefore, in subsequent experiments, the ricin-affinity-purified fraction was concentrated tenfold. For immunodepletion studies, an excess of anti-OFA-I-1 monoclonal antibodies (IgM class) was reacted with agarosebound u-chain-specific rabbit anti-human IgM antibodies as described elsewhere (25). After being extensively washed with Veronal-buffered saline, the IgM-containing immunobeads were used for four successive immunodepletions of ricin-affinity-purified FA fractions (0.17 mg protein/ml). The mixture of first immunodepletion was comprised of 2 ml FA fraction and 50 ug IgM (anti-OFA-I-1) bound to 25 mg rabbit anti-human IgM agarose beads. The mixture was incubated at 37°C for 30 minutes and centrifuged at 1,000Xg for 10 minutes. After an aliquot of 400 ul was saved, the supernatant was treated three times in succession with fresh batches of 25 mg anti-OFA-I-1-containing immunobeads. The aliquots of the four immunodepleted supernatants were tested in immune adherence and competitive inhibition ELISA. In addition, 2 ml of the FA fraction was
SDS), at least five bands stainable by Coomassie blue were observed (fig. 3, gel B). These bands were near the 22K-, 40K-, 60K-, 70K-, and 280K-M, regions of the gel. Analysis of the unstained gel for localization of the FA activity by the modified ELISA revealed that the antigenic activity was in the region of 60K-70K M (text-fig. 6). This region contained two bands. These results suggested that FA activity present in the ricin-affinity-purified fraction of M14 spent medium may have been in the aggregated form of antigenically active molecules of 60K and/or 70K daltons.
MW MARKERS 94K † 43K9 20K1 14,4K TEXT-FIGURE 6.— Association of FA activity with various regions of SDS PAGE (fig. 3). FA activity in each slice of the gel was analyzed by modified ELISA as described in the text. MW=molecular weight.
* Anti-OFA-I-1 (monoclonal) was provided by Dr. R. F. Irie. "FA fraction before ricin-affinity purification was used as target as described in "Materials and Methods." Each OD value represents an average of three replicates. d Starting material denotes ricin-affinity-purified FA fraction that was used for immunodepletion by anti-OFA-I-1 immunobeads and control immunobeads.
treated in a similar manner, with 25 mg rabbit anti-human IgG that was not coated with anti-OFA-I-1. The supernatants from these treatments were used as controls. With the use of anti-OFA-I-1 monoclonal antibody and M14 cells in immune adherence, the blocking ability of the FA fraction was almost completely abolished after the first immunodepletion with the immunobead-bound anti-OFAI-1, whereas no significant reduction in the blocking occurred after four depletions of the FA fraction with control immunobeads. The degree of inhibition in competitive ELISA by immunoundepleted and immunodepleted samples (FA fraction) was almost the same (table 6). These results suggested that anti-OFA-I-1 antibodies were unable to react with the FA detected by the ELISA with the use of allogeneic antibody. To confirm this observation, we reacted anti-OFA-I-1 with the FA fraction (before ricin-affinity purification) in an indirect ELISA with alkaline phosphatase-conjugated goat anti-human IgM as the second antibody. Although this monoclonal antibody showed reactivity with the FA fraction by immune adherence, no detectable reactivity was observed in ELISA with the use of the FA fraction as the target antigen, even at the lowest (1:10) dilution of the antibody. This difference might have occurred because the OFA-I-1 antigen recognized by the monoclonal antibody, which might have been present in the FA fraction, did not adhere to the Immulon plates under the experimental conditions described in “Materials and Methods." Thus the FA recognized in this study is different from the OFA-I-1 and the OFA-I-2.
The inability of proteolytic enzymes to inactivate FA could be due either to inaccessibility of the protein core that is surrounded by polysaccharide or to the fact that the protein part does not form the antigenic determinant of the FA. The FA fraction did contain easily detectable carbohydrates. Since proteolytic or lipolytic enzymes themselves did not affect the FA activity, its destruction by B-galactosidase and hyaluronidase was not due to contamination of these glycosidases by proteolytic and lipolytic enzymes. The reason for inactivation of FA by hyaluronidase is not clear. However, it has been reported that crude testicular hyaluronidase can cleave terminal B-galactose residues from glycoproteins and from proteoglycans (29). Such activity could be responsible for inactivation of the FA. Therefore, we concluded that the antigenic activity is associated with the carbohydrate moiety of the FA molecule. Analysis of ricin-affinity-purified FA by gradient PAGE under native conditions clearly revealed that the proteinaceous material was retained in the high-molecular-weight region. Also, the FA activity was not sharply delineated. Analysis under dissociating conditions (SDS-gradient PAGE) of the FA after treatment with the SDS showed at least five proteinaceous bands, two of which were in the 60to 70K-M, region that contained FA activity as well. Since under dissociating conditions high-molecular-weight bands had disappeared and new ones appeared, it seemed that at least some of these bands originated from the high-molecular-weight region. However, it is also possible that some of the bands in the low-molecular-weight region originated from the material that was excluded in native gels due to its acidic nature. Nevertheless, the fact remains that apparently high-molecular-weight component(s) observed by native PAGE that contained FA activity disappeared after SDS treatment, and FA activity migrated to the 60- to 70Kdalton region. Because the FA-containing fraction was preheated in the presence of 1% SDS at 100°C for PAGE analysis under dissociating conditions, the possibility that low-molecular-weight components, e.g., glycolipids, complexed with a protein or albuminoid material is highly unlikely. Other antigens associated with M 14 were either separated into fractions that did not contain FA or inactivated by the extraction procedures used. This was evident by the lack of reactivity of some of the melanoma sera with various fractions of spent medium, although these sera did react positively with the concentrated spent medium. The extracted FA was devoid of HLA activities that are associated with the M14 cells (5). This was demonstrated by the lack of reactivity between the FA fraction and certain sera from melanoma patients who had received tumor cell vaccine and BCG as immunotherapy. These serum samples contained antibodies to HLA as evidenced by their reactivity both to M14 cells and to the autologous ML 14 cells. Absorption of these sera with ML 14 cells completely removed the antibody activity against M14 cells (5). Furthermore, the incidence of reactivity and the mean antibody titers of ML14 absorbed and unabsorbed sera that retained reactivity against CDM-grown M14 cells after absorption with ML14 cells were almost the same when the FA fraction was used as the target antigen.
It has become obvious that the antigenic composition of human malignant melanoma is highly heterogeneous. On the basis of cross-reactivities in serologic assays and absorption studies, at least 3 groups of antigens associated with human melanomas have been recognized: a) FA, b) groupspecific antigens, and c) individually specific antigens (26). Our earlier report suggested that many of these antigens could be detected in spent culture medium of melanoma cells and could be separated by biochemical means (5, 27). Despite extensive studies of FA associated with human cancer, there is little information regarding their physical and chemical properties. Results presented in this report suggest that FA isolated by CM extraction of the concentrated and processed spent culture medium of the M14 cell line may be a glycoprotein of 60-70K daltons. This conclusion is drawn on the basis of the observations that a) FA activity is heat-stable; b) it can be destroyed by certain glycosidases, e.g., B-galactosidase and hyaluronidase, but not by neuraminidase; c) it binds to lectins that have specificity for B-galactose; and d) after SDS-PAGE analysis antigenic activity can be demonstrated in the region of 6070K daltons. This region of the gel contained two bands stainable by Coomassie blue. Because of the closeness of the two bands, we were unable to discern whether or not both bands were antigenic. Attempts to localize the FA activity by the Electro-Blot method with the use of nitrocellulose membrane (28) were unsuccessful. This failure, we believe, was due to antibodies of low affinity in the allogeneic serum. Page 17
(6) Irie RF, Irie K, Morton DL. A membrane antigen common to human cancer and fetal brain tissue. Cancer Res 1976; 36:3510 3517. (7) IRIE RF, GIULIANO AE, MORTON DL. Oncofetal antigen: A tumor associated fetal antigen immunogenic in man. JNCI 1979; 63:367 373. (8) Irie RF, Sze LL, Saxton RE. Human antibody to OFA-I, a tumor antigen, produced in vitro by Epstein-Barr virus-transformed human B-lymphoid cell lines. Proc Natl Acad Sci USA 1982; 79:5666 5670. (9) LAINE RA, HAKOMORI S. Incorporation of exogenous glycosphingo lipids in plasma membranes of cultured hamster cells and concurrent change of growth behavior. Biochem Biophys Res Commun 1973; 54:1039-1054. (10) Irie RF, Irie K, MORTON DL. Natural antibody in human serum to a neoantigen in human cultured cells grown in fetal bovine serum. JNCI 1974; 52:1051-1058. (17) GUPTA RK, Irie RF, MORTON DL. Antigens on human tumor cells assayed by complement fixation with allogeneic sera. Cancer Res 1978; 38:2573-2580. (12) CHEE DO, Boddie AW, Roth JA, HOLMES EC, MORTON DL. Pro duction of melanoma-associated antigen(s) by a defined malignant melanoma cell strain grown in chemically defined medium. Cancer Res 1976; 36:1503-1509. (13) RENK CM, GUPTA RK, Morton DL. Inhibition of normal allogeneic lymphocyte mitogenesis by a factor released from human tumor cells in culture. Cancer Immunol Immunother 1981; 11:7-16. (14) PELLEGRINO MA, FERRONE S, REISFELD RA, IRIE RF, GOLUB SH. Expression of histocompatibility (HLA) antigens on tumor cells and normal cells from patients with melanoma. Cancer 1977; 40:36-41. (15) GUPTA RK, Morton DL. Suggestive evidence for in vivo binding of specific antitumor antibodies of human melanoma. Cancer Res 1975; 35:58-62. (16) ENGVAL E, PERLMANN P. Enzyme-linked immunosorbent assay, ELISA. III. Quantitation of specific antibodies by enzyme-labeled anti-immunoglobulin in antigen-coated tubes. J Immunol 1972; 109:129-135. (17) CRAVEN GR, STEERS E JR, ANFINSEN CB. Purification, composition, and molecular weight of the beta-galactosidase of Escherichia coli K12. J Biol Chem 1965; 240:2468-2477. (18) Yang C, Srivastav PN. Purification and properties of hyaluronidase from bull sperm. J Biol Chem 1975; 250:79-83. (19) BROCKMAN HL, Law JH, Kezny FJ. Catalysis by adsorbed enzymes: The hydrolysis of triproprionin by pancreatic lipase adsorbed to
(22) Lowry OH, ROSEBROUCH NJ, Farr AL, RANDALL RJ. Protein mea surement with the Folin phenol reagent. J Biol Chem 1951; 193:265-275. (23) Spiro RG. Analysis of sugars found in glycoproteins. Complex car bohydrates. Methods Enzymol 1966; 8:7-9. (24) Fleiss JL. Statistical methods for rates and proportions. 2d ed. New York: Wiley, 1981:61, 114. (25) GUPTA RK, SILVER HK, REISFELD RA, Morton DL. Isolation and immunochemical characterization of antibodies from the sera of cancer patients which are reactive against human melanoma cell membranes by affinity chromatography. Cancer Res 1979; 39:1683-1695. (26) Old LJ. Cancer immunology: The search for specificity-G.H.A. Memorial lecture. Cancer Res 1981; 41:361-375. (27) Gupta RK, Silver HK, MORTON DL. Production and characteriza tion of xenogeneic antisera to tumor-associated antigen(s). J Surg Oncol 1980; 13:75-89. (28) TowBin H, STAEHELIN T, GORDON J. Electrophoretic transfer of Page 18
TABLE 2.-Effector-to-target cell ratios and monocyte- and macrophage-mediated tumor cytolysis Mean % [H]dThd release = SE from HeLa cells by: Effector:target cell Blood Peritoneal monocytes macrophages 15:1 37.2+1.7 41.425.4 10:1 14.9+4.0 43.9£3.9 5:1 9.8+1.7 22.7+1.6
TABLE 3.—Monocyte- and macrophage-mediated cell lysis Mean % [H]dThd release SE by: Target cell (No. of women studied) Peritoneal macBlood monocytes rophages Hela (33) 19.7+2.5 42.5+3.3 HL60 (4) 4.4+3.0 38.3+10.2 K562 (7) 34.2+11.4 49.6+13.2 HPAF (4) 9.4+2.9 29.0+8.3 Normal fibroblasts (3) 0.0+1.5 5.8+1.4
cytes (20)]. Preliminary experiments showed that isolated platelets, even in numbers greatly exceeding those contaminating the monocyte preparations, did not produce Oz or H2O2 with or without 200 nM PMA. Other procedures. —Nonspecific esterase (a-naphthyl butyrate esterase) stains were done with reagents from Technicon Instruments Corp., Tarrytown, N.Y., (21), and peroxidase stains were done by the method of Kaplow (22). Sheep erythrocytes were opsonized as previously described (23).
Characterization of monocytes and peritoneal macrophages. — The numbers of blood monocytes from the 3 groups of women did not differ (table 1). As reported previously (15), endometriosis patients (group III) had more peritoneal cells than did infertile women with inactive pelvic inflammatory disease (group II) or normal women (group I). The peritoneal fluids from all 3 groups had comparable percentages of leukocytes. The peritoneal macrophages were larger than the monocytes (9-12 vs. 15-25 um in diameter), and they spread more rapidly on tissue culture plates. The monocytes and macrophages stained equally well for nonspecific esterase and peroxidase, and they both phagocytized latex spheres and anti-sheep erythrocyte IgG-coated sheep erythrocytes.
Tumor cell killing by monocytes and peritoneal macrophages. Lysis of HeLa tumor cells by blood monocytes and peritoneal macrophages depended on the effector-to-target cell ratio, with more specific release of [H]dThd occurring at the higher ratios. Table 2 demonstrates results of an experiment representative of three comparable experiments in which cells from these separate patients were used. When examined as matched pairs (i.e., blood monocytes and peritoneal macrophages) from individual subjects, peritoneal macrophages always lysed more HeLa cells than did the monocytes, and the macrophage-mediated lysis was statistically greater than the monocyte-mediated lysis (P<0.01 for group I, P<0.05 for II, and P<0.01 for III; text-fig. 1). The tumoricidal effects of the monocytes or macrophages in one group did not differ from those of the other groups (P>0.1). This same pattern of monocyte- and macrophagemediated tumor cytolysis was seen with other tumor target cells (e.g., with K562, HL60, and HPAF; table 3). Early passage, normal skin fibroblasts were essentially not lysed by the monocytes or macrophages. Supernatants or lysates of cytolytic macrophages or monocytes (75, 50, or 25% vol/ vol) coincubated with the [H]dThd-labeled HeLa cells for 60 hours did not lyse those tumor cells. Lack of enhancement of human mononuclear phagocyte-mediated tumor cytolysis by LPS.—Mouse peritoneal macrophages are very sensitive to the activating effects of LPS (5, 6, 24). We
n6 I II MI TEXT-FIGURE 1.—Lysis of HeLa tumor cells by blood monocytes and peritoneal macrophages. HeLa lysis (percent [H]dThd release) was determined for matched pairs (monocytes and macrophages) from each patient. indicate means from each group. Cells in I are from normal, fertile women; in II, from infertile women with past salpingo-oophoritis; and in III, from infertile women with endometriosis. The effector-totarget cell ratio was 10:1 for both monocytes and macrophages. Monocytes are different from macrophages in group I (P<0.01 by MannWhitney U test), group II (P<0.05), and group III (P<0.01).
TEXT-FIGURE 2.-Lysis of Hela tumor cells by blood monocytes and peritoneal macrophages in the presence of LPS. Lysis assays were done in the presence of different amounts of E. coli K235 LPS and of 10% heat-inactivated autologous serum. Effector-to-target cell ratio was 10:1 for both monocytes and macrophages. indicate means. Tumor cytolysis by monocytes or macrophages in the presence of all doses of LPS did not differ from the respective control groups (P>0.1; Mann-Whitney U test). At all doses of LPS, macrophages lysed more than did monocytes (P<0.002 in the 0, 10', and 102 LPS groups and P<0.02 in the 108 and 10* LPS groups; Mann-Whitney U test).
hypothesized that human mononuclear phagocyte tumor cell killing would likewise be enhanced and that peritoneal macrophages [being a more differentiated form of a blood monocyte (25)] would be more sensitive to the effects of LPS. Text-figure 2 shows that E. coli K235 LPS did not augment the tumoricidal effect of monocytes or macrophages (P>0.1). Another LPS preparation (Westphal-extracted E. coli 0128:B12) also did not enhance the killing. Lack of involvement of reactive oxygen species in the killing process. --Mouse peritoneal macrophages stimulated by PMA kill tumor cells by secreting H2O2 (26). Likewise, mouse peritoneal macrophage-mediated destruction of antibody-coated microbes or antibody-coated tumor cells is caused by H2O2 (27, 28). Also, PMA-treated human monocytes lyse erythrocytes by an H2O2 and/or 02-dependent reaction (29, 30). In our experiments here, even though the peritoneal macrophages were more cytolytic than the monocytes, the blood monocytes and peritoneal macrophages had comparable basal and PMA-stimulated H2O2 and O2 (text-fig. 3). To help determine if spontaneous human mononuclear phagocyte-mediated tumor cytolysis is mediated by a reactive oxygen species, we did experiments using known inhibitors of reactive oxygen species in doses shown in other systems to inhibit or quench these potential mediators (2630) (table 4). Catalase (which reduces H2O2 to O2 and H2O), SD (which converts Oz to H2O2), mannitol (which inhibits hydroxyl radical activity), and l-histidine (which inhibits singlet oxygen activity) did not reduce the monocyte- or macrophage-mediated HeLa cell killing. Data, not included in table 4, also showed that lower amounts of these potential inhibitors did not reduce the killing.
OPMA 200 nMPMA O PMA 200 nMPMA O2 Production H2O2 Production TEXT-FIGURE 3.—O2 and H2O2 production by blood monocytes and peri toneal macrophages without and with 200 nM PMA. As determined by the Student's t-test, monocytes and macrophages without PMA did not produce different amounts of Oz (P>0.3) or H2O2 (P>0.8), and monocytes and macrophages with 200 nM PMA did not produce different amounts of Oz (P>0.2) or H2O2 (P>0.6). As compared to no PMA, 200 NM PMA enhanced monocyte Oz (P<0.0001) and H202 (P<0.05) production and macrophage O2 (P<0.001) and H2O2 (P<0.02) production.
TABLE 4.-Lack of inhibition of human mononuclear phagocytemediated tumor cytolysis by reactive oxygen species scavengers and quenchers Mean % [ H]dThd release I SE
Tissue macrophages arise from circulating blood monocytes (31, 32). In the tissue, the monocytes differentiate into qualitatively distinct cells. This study demonstrates morphologic and functional differences between blood monocytes and peritoneal macrophages isolated from the same normal or infertile woman. The resident human peritoneal macrophages from normal and infertile women are larger and manifest more tumor cytolysis than do blood monocytes from these same women. Thus their enhanced tumoricidal activity represents a functional change associated with the in vivo differentiation of the blood monocytes to peritoneal macrophages. Attempts to identify macrophage-activating factors in peritoneal fluid have thus far been unsuccessful (Weinberg JB: In preparation). Other studies analyzing the antitumor effects of human tissue mononuclear phagocytes have also demonstrated that macrophages in peritoneal fluid (10-12, 33), alveoli (34), and pleural fluid (12) are cytostatic and cytolytic for various tumor cells in vitro. Many of the subjects studied had diseases including cancer (12, 33), and differences in the collection and purification methods make it difficult to directly compare their results to ours. However, our study does show that peritoneal macrophages freshly isolated from the physiologic peritoneal fluid of normal women, like those of subjects with renal failure (10, 11), cancer (12, 33), and indwelling peritoneal catheters (12), are spontaneously cytolytic for tumor cells in vitro.
1 LPS is extremely potent in inducing mouse macrophages to kill tumor cells in vitro (4-6). Human monocytes are also 1 very sensitive to LPS, as evidenced by the induction of tissue i factor (35) or interleukin 1 elaboration (36, 37) by nano: gram quantities of LPS. However, our studies show that freshly isolated blood monocytes and peritoneal macro: phages do not have enhanced tumor cell killing in assays I done in the presence of 102-104 ng/ml. Others showed that monocytes, cultured for 4 or 5 days and then treated with LPS, do have enhanced tumor cytolytic activity (38, 39). The precise mechanism(s) by which mononuclear phagocytes kill tumor cells in vitro is not known. The tumor cell killing is nonphagocytic, requires effector-target cell contact, and is not reproduced by cell-free supernatants from or lysates of the tumoricidal phagocytes. Reactive oxygen species are known to act as effectors of cell lysis in various systems, including human monocyte-mediated lysis of antibody-coated erythrocytes or tumor cells (26-30, 40). Mouse macrophages (activated by in vivo infection) and human alveolar macrophages have normal cytocidal activity in the presence of catalase or SD (23, 34, 41). We also show here that the spontaneous, antibody-independent tumoricidal effects of human monocytes and macrophages are not inhibited by catalase, SD, mannitol, or L-histidine. Although it is possible that these quenchers might be excluded from areas of macrophage-tumor cell membrane apposition (42), these results suggest that the reactive oxygen species do not operate in this process. Sorrel et al. (41) showed normal tumor cytotoxicity by mouse macrophages in “anaerobic” incubations; our multiple attempts to determine the necessity for oxygen in the killing process over a 60-hour culture revealed no inhibition of mononuclear phagocyte tumor cell killing by anaerobic (or low oxygen) conditions (data not shown). However, we, as others, were never confident that ultimate and permanent exclusion of oxygen from the microenvironment of the phagocyte and tumor cell had been achieved (41-43). Whether the tumor cells might evoke any or differing H2O2 or Oī responses from the monocytes or macrophages is not known. However, the basal and PMAstimulated H202 and O2 production by blood monocytes and peritoneal macrophages was essentially the same, even though the macrophages had more lytic capacity for tumor cells. These observations provide further support for the theory that spontaneous tumor cell killing by human mononuclear phagocytes is not mediated by reactive oxygen species.
icity: Development of macrophage cytotoxic activity requires completion of a sequence of short-lived intermediary reactions. J Im munol 1978; 121:2035-2042. (5) WEINBERG JB, CHAPMAN HA, Hibbs JB JR. Characterization of the effects of endotoxin on macrophage tumor cell killing. J Immunol 1978; 121:72-80. (6) Pace JL, Russell SW. Activation of mouse macrophages for tumor cell killing. I. Quantitative analysis of interactions between lym phokine and lipopolysaccharide. J Immunol 1981; 126:1863–1867. (7) Rinehart JJ, VESSELLA R, LANGE P, Kaplan ME. Characterization and comparison of human monocyte- and macrophage-induced tumor cell cytotoxicity. J Lab Clin Med 1979; 93:361-369. (8) CAMERON DJ, CHURCHILL WH. Cytotoxicity of human macrophages for tumor cells. Enhancement by human lymphocyte mediators. J Clin Invest 1979; 63:977-984. (9) HAMMERSTROM J. Human monocyte-mediated cytotoxicity to K562 cells: Activation by lymphokines. Scand J Immunol 1979; 10:575 584. (10) GOUGEROT MA, DIMITRIU A, DY M, LEGRAND L, HAMBURGER J. Demonstration of a human lymphokine which renders macrophages cytotoxic. Ann Immunol (Paris) 1977; 128C:624-632. (11) HAMMERSTROM J. Structure and function of human effusion macro phages from patients with malignant and benign disease. 2. In vitro cytostatic and cytolytic effect on human tumor cell lines. Acta Pathol Microbiol Scand [C] 1980; 88:201-209. (12) MANTOVANI A, SHAVIT ZB, Peri G, et al. Natural cytotoxicity on
(1) Evans R, ALEXANDER P. Mechanisms of extracellular killing of nu cleated mammalian cells by macrophages. In: Nelson DS, ed. Immunobiology of the macrophage. New York: Academic Press, 1975:100. (2) WEINBERG JB, Hibbs JB JR. The role of macrophages in cancer resistance and therapy. In: Waters H, ed. The handbook of cancer immunology. New York: Garland STPM Press, 1978:51-90. (3) BLANDEN RV, HAPEL AJ, DOHERTY PC, ZINKERNAGEL RM. Lympho cyte-macrophage interactions and macrophage activation in the expression of antimicrobial immunity in vivo. In: Nelson DS, ed. Immunobiology of the macrophage. New York: Academic Press, 1976:367-400. (4) Ruco LP, Meltzer MS. Macrophage activation for tumor cytotox
vation of macrophages in vivo and in vitro. Correlation between hydrogen peroxide release and killing of Trypanosoma cruzi. J Exp Med 1979; 149:1056–1068. (28) NATHAN C, COHN ZA. Role of oxygen-dependent mechanisms in antibody-induced lysis of tumor cells by activated macrophages. J Exp Med 1980; 152:198–208. (29) Weiss SJ, LOBUGLIO AF, KESSLER HB. Oxidative mechanisms of monocyte-mediated cytotoxicity. Proc Natl Acad Sci USA 1980; 77:584-587. (30) Klassen DK, SAGONE AL JR. Evidence for both oxygen and non oxygen dependent mechanisms of antibody sensitized target cell lysis by human monocytes. Blood 1980; 56:985-992. (31) VAN FURTH R, COHN ZA. The origin and kinetics of mononuclear phagocytes. J Exp Med 1968; 128:415-425. (32) VOLKMAN A. Monocyte kinetics and their changes in infection. In: Nelson DS, ed. Immunobiology of the macrophage. New York: Academic Press, 1975:291. (33) MANTOVANI A, POLENTARUTTI N, Peri G, et al. Cytotoxicity on tumor
Temperance Society Members 1.2
ABSTRACT—Cancer risk was studied in 781 male Seventh-Day Adventists (SDA) and 808 male members of other temperance societies. Standardized morbidity ratios for all cancers were 0.69 among SDA and 1.05 among other temperants. Significantly decreased risks of cancers were noted among SDA for cancer of the colon (observed/ expected (0/E): 0.13), cancer of the respiratory system (O/E: 0.17), cancer of the lung (0/E: 0.15), and cancer of the bla der including papilloma (O/E: 0.13). No significant deviations from expectations were noted among members of other temperance societies. Thus risks of tobacco-associated cancers were markedly decreased among SDA. The risk of alcohol-associated cancers (cancers of the oral cavity, pharynx, esophagus, and larynx) taken together was also decreased (0/E: 0.7), although not significantly so. When the results were compared with those of a previous study of Danish brewery workers who had a high average daily beer intake, the present investigation provides further support that the alleged association between beer consumption and the occurrence of rectal cancer is of a noncausal nature. The explanation for the decreased risk of colon cancer should probably be sought in the dietary practices of SDA.—JNCI 1983; 70:1011-1014.
was also studied. Approximately half of the group investigated were SDA. This study provided us with an opportunity to examine cancer patterns in SDA outside the United States, where most previous studies of SDA have been performed. The existence of a nationwide cancer registry in Denmark makes it possible to study cancer morbidity, thus avoiding the possible bias involved in mortality studies. In addition to a description of the total cancer pattern among SDA, the risk of cancer is compared with cancer risks in a group of male brewery workers. These workers who have an alcohol intake, notably in the form of beer, well above the average were previously investigated in exactly the same fashion as were the temperance society members by means of a cohort study (5).
Temperance society membership.—The National League of Temperance Societies (Afholdsselskabernes Landsforbund) consists of a number of independent societies with a decentralized organizational pattern and no central record system. The groups who are members of the national association are quite different. The association accommodates temperance societies that have memberships consisting of a mixture of lifelong nondrinkers and reformed alcoholics in unknown proportions, although as much as 25-50% may fall into the reformed alcoholic category. At the other end of the spectrum, some religious groups such as the SDA hold group memberships; some of these originate from SDA families and they are likely to have been lifetime alcohol abstainers. Each temperance society member signs a declaration once a year that he will not drink. With the assistance of the National League of Temperance Societies, we obtained lists of current and all previous
Groups at high and low risk of cancer may provide clues to factors that influence the risk of the disease. Investigations of alcoholics and other groups having a high daily intake of alcohol, as well as case-control studies, indicated many years ago that alcohol is a risk factor for certain upper aerodigestive tract cancers. Similarly, the role of tobacco in a number of cancers is well established. In consonance with such well-established associations of cancer with alcohol and tobacco, SDA, who by church prescription abstain from smoking and the use of alcoholic beverages, have a much lower risk of developing cancers of the lung, upper aerodigestive tract, and bladder in the United States than the average white U.S. population (1, 2). Similarly, findings of decreased risks emerge from studies of U.S. Mormons (3, 4). In addition to the low risk for the smoking- and alcoholrelated cancers, deficits have been noted for a number of other cancers among SDA; these deficits result in a lower overall cancer rate that is approximately two-thirds of the average U.S. rates (1, 2). The low risk of colorectal cancer among SDA in particular has been associated with the lactoovo-vegetarian diet followed by a large proportion of SDA. The finding of similarly low rates among Mormons (3, 4) points to the possibility that more general health habits are of importance in these cancers. Fatality rates are influenced by diagnoses in early stages of the disease, which is likely to have an impact on the mortality data used by most investigators as indicators of cancer morbidity. As part of the ongoing investigations of the association between alcohol and cancer, and in particular of the evaluation of the alleged association between beer consumption and rectal cancer by the study of Danish brewery workers, a group of male Copenhagen temperance society members
ABBREVIATIONS USED: CI=confidence interval; ICD=International Classification of Diseases; O/E=observed/expected; SDA=Seventh-Day Adventist(s).
Received August 3, 1982; accepted January 7, 1983. ? The study was in part carried out within the Program on Alcohol and Cancer of the International Agency for Research on Cancer, Lyon, France, with the support of the National Institute of Alcohol Abuse and Alcoholism, Alcohol, Drug Abuse, and Mental Health Administration (contract HSM 72-72-116 and ADM 281-77-0026). 3 Danish Cancer Registry, Strandboulevarden 49, DK-2100 Copen Page 19
male members of the Copenhagen fractions of the Danish Temperance Union (i.e., SDA), Blue Cross, Good Templars, the Baptist Temperance Mission, and the White Cross. Since we assembled the cohort to parallel a previously investigated cohort of 14,313 members of the Danish brewery workers (5), only male members between 1939 and 1963 were included, which left 1,882 persons in the study group (table 1). Follow-up. We obtained as much identifying information as possible from each temperance society. However, records were of varying quality, but by means of follow-up in municipality registries, the National Central Person Registry, and the National Central Death Registry all but 78 persons (4.1%) could be identified and followed until death, emigration, or December 31, 1977. Those persons not identified were excluded from the investigation and so were 43 persons who left the country or died before the start of the Danish Cancer Registry on January 1, 1943. The remaining total of 1,752 persons was computer linked with the Danish Cancer Registry on the basis of the national identification number (Central Person Number) whenever
available in the cohort. For the remainder of the temperance cohort, a computerized search of the Cancer Registry file located all male patients born on the same date as the cohort member. Linkage of records in the two files pertaining to the same individual was then undertaken visually on the basis of concordance of full name, sex, date and place of birth, and address. These linkage procedures for follow-up in the Cancer Registry have previously been shown to be valid (6). Linkage was undertaken by the staff of the Registry who had no knowledge of a person's adherence to the SDA or to another temperance society. Because the files of the Blue Cross temperance society turned out to be incomplete, this group of 163 individuals was excluded from further study, which left a total of 1,589 persons included in the analyses (see table 2). Observed and expected numbers of cancer.—The 4-digit modified version of the Seventh Revision of the ICD has been used for the coding of cancer cases in the Danish Cancer Registry during the period 1943-77. The registry may be regarded as virtually complete. After the coded contents of the Danish Cancer Registry and the League Temperance Society records were linked, the observed number of cases was computed. Person-years at risk were computed for sex-specific, 5-year age groups and 5-year calendar time periods for each person. Taken into account were the date of entry into the cohort, age, and time of leaving the cohort either at death, at emigration, or on December 31, 1977. We derived expected numbers of cancer cases by multiplying sex-, age-, and calendar time-specific incidence rates for Copenhagen with the corresponding person-years, thus indirectly standardizing for age, sex, and time trends by summarizing over the various cells. Diagnostic categories compared were identical for observed and expected numbers at the 4-digit level of the ICD.
Table 2.–Observed and expected numbers of cancer cases during 1943-1977 among male Copenhagen temperants ICD (7th rev.) Danish Temperance Union members Members of other temperance societies Site classification Observed Expected O/E 95% CI Observed Expected O/E 95% CI 140-145, 147-148 Buccal cavity and pharynx 6 9 4.5 2.0 150-159 Digestive organs, peritoneum 26 33.6 0.8 0.5-1.1 63 58.7 1.1 0.8-1.4 150 Esophagus 0 2.1 0.0-1.8 6 3.7 0.6-3.5 151 Stomach 10 1.1 0.5-2.0 20 16.0 1.3 0.8-1.9 153 Colon (including rectosig 1 9 13.3 0.7 0.3-1.3 moid) 154 Rectum (excluding anus) 11 7.3 0.8-2.7 14 12.8 1.1 0.6-1.8 157 Pancreas 3 3.5 0.9 0.2-2.5 4 6.1 0.7 0.2-1.9 146, 160-164 Respiratory system 4 23.7 0.1-0.4 39 40.9 1.0 0.7-1.3 161 Larynx 0 1.7 0.0-2.2 4 2.7 1.5 0.4-3.8 162.1 Lung, primary 3 20.4 0.0-0.4 35 35.3 1.0 0.7-1.4 177-179 Male genital organs 10 11.1 0.9 0.4-1.7 22 18.4 1.2 0.8-1.8 180-181 Urinary system 5 11.0 0.5 0.2-1.1 20 18.6 1.1 0.7-1.7 180 Kidney 4 3.3 1.2 0.3–3.1 8 5.6 1.4 0.6-2.8 181 Bladder (including papil 1 7.5 0.1 0.0-0.7 11 12.6 0.9 0.4-1.6 loma) 200-205 Lymphatic, hematopoietic, 7 4.8 1.5 0.6-3.0 7 7.6 0.9 0.4-1.9 and other and unspecified 6 5.4 1.1 0.4-2.4 6 9.1 0.2-1.4 140-205 All malignant neoplasms 64 92.5 0.7 0.5-0.9 166 157.4 1.1 0.9-1.2 No. of persons in group 781 808 Person-years Page 20
je therein reme d brand d. I
Total person-years of observation, 290,186. “The Seventh Revision of the International Classification of Diseases applies to deaths occurring from 1965 to 1968 and the Eighth Revision, sastre - from 1969 to 1977. Two-sided P-value. results being obtained for “diesel fume exposure." Analysis er, the slightly smaller than the corresponding values in table 2.) Although the mortality rate for all causes of death combined restricted to individuals who retired from 1950 onward also showed a significant trend, the increases in relative risks yielded relative risks of 1.0 (based on 231 deaths), 1.23 bears were much smaller than those for lung cancer, being respec- (based on 393 deaths), and 1.40 (based on 273 deaths) br the tively 1.00, 1.04, and 1.09 for nonexposed, possibly exposed, respectively for individuals nonexposed, possibly exposed, and probably exposed to diesel fumes. These figures sugactiver and probably exposed groups. Coal exposure.-When trend analyses were conducted for gested that the association was with diesel fumes rather than coal exposure in a similar way to the trend analyses for with coal, but this conclusion must, of course, remain tendiesel exposure, only lung cancer showed consistent eleva- tative. tions in risk with increasing exposure, with a corresponding A further complication arose because individuals involved statistically significant trend. The data are shown in table in the maintenance of steam locomotives may have been 4. The corresponding relative risks for all causes of death exposed to high levels of asbestos (Terracini B: Personal combined were 1.00, 1.04, and 1.08, for nonexposed, possibly communication), and asbestos-related deaths have been reexposed, and probably exposed groups, respectively, again ported in such individuals (20); thus part of the observed the increase being considerably less than the corresponding associations with lung cancer could be due to asbestos increase for lung cancer. exposure. However, when those occupational groups that Joint exposure of diesel fumes, coal, and asbestos.—There is included individuals involved in locomotive maintenance considerable overlap between occupations involving proba- were excluded from analysis, the relative risks for individuals ble exposure to diesel fumes and probable exposure to coal nonexposed, possibly exposed, and probably exposed to dust as indicated by the similarity of the results presented in tables 3 and 4. This was a consequence of the transition TABLE 3.-Cancers of the trachea, bronchus, and lung in workers from steam power to diesel power over the years. Most exposed to varying levels of diesel fumes members of the cohort would have had been in employment Level of exposure (P-value) during the years when that transition occurred, which would nake it difficult for one to distinguish between the diesel Parameter Non- Possibly Probably ume and coal associations with lung cancer. Analyses re exposed exposed exposed tricted to individuals who retired before 1950, i.e., who Observed No. of 239 407 279 vould probably have had essentially coal dust rather than deaths liesel fume exposure, yielded relative risks of 1.0 (based on Person-years of 85,200 121,850 74,571 observation ; deaths), 0.70 (based on 13 deaths), and 0.44 (based on 7 Page 21
confounding because of misclassification of tobacco was present to the extent that study subjects did not perfectly recall and report tobacco histories. With a sample of respondents, we conducted a brief telephone re-interview. There was 97% concordance between the history of nonfilter smoking (yes or no) reported in the re-interview and that reported in the original home interview. We do not know whether the more complicated tobacco questions would have shown high accordance, and we do not know if the answers were accurate as well as replicable. Greenland (29) discussed the problem of misclassification of covariates leading to residual confounding, and Morrison et al. (12) and Morrison (paper presented at the annual meeting of the Society for Epidemiologic Research) discussed the particular implications for coffee, tobacco, and bladder cancer. Whether residual confounding accounts for the entire observed excess RR of 40% cannot be determined. Confounding by correlates of coffee drinking other than tobacco exposure may also have contributed to the persistent but inconsistent relation between bladder cancer and coffee. As noted, control for a variety of factors did not materially alter the estimates. Overall, people who never drank coffee are a small minority of American adults, and the distinguishing characteristics of this minority are not well understood. They may differ from coffee drinkers on a variety of healthrelated variables, but it is not clear what other correlates of coffee drinking might be related to bladder cancer. Further studies of typical U.S. populations are not likely to provide more precise estimates than those from this unusually large study, and they are not likely to avoid the bias created by residual confounding by tobacco. Additional studies of populations with a low prevalence of coffee drinking, e.g., Mormons or Seventh-Day Adventists, may illuminate the comparison of nondrinkers and drinkers. (We are pursuing this possibility by continuing the present study in Utah.) Studies of populations who drink more coffee than the U.S. population, e.g., Scandinavians, may also reveal whether very high doses of coffee affect bladder cancer risk. In addition, the completion of laboratory tests in progress will further our understanding of the effects of coffee. Meanwhile, the available evidence from our study, other epidemiologic studies, and laboratory experiments suggests that coffee plays either a small biologic role in human bladder cancer or no role at all.
(6) WYNDER EL, ONDERDONK J, MANTEL N. An epidemiological investi gation of cancer of the bladder. Cancer 1963; 13:1388–1406. (7) Miller CT, NEUTEL CI, NAIR RC, Marrett LD, Last JM, Collins WE. Relative importance of risk factors in bladder carcinogenesis. J Chronic Dis 1978; 31:51-56. (8) Kessler II, Clark JP. Saccharin, cyclamate, and human bladder ! cancer. JAMA 1978; 240:349–355. (9) Mettlin C, GRAHAM S. Dietary risk factors in human bladder cancer. Am J Epidemiol 1979; 110:255-263. (10) Howe GR, BURCH JD, MILLER AB, et al. Tobacco use, occupation, coffee, various nutrients, and bladder cancer. JNCI 1980; 64:701 713. (11) CARTWRIGHT RA, ADIB R, GLASHAN R, GRAY BK. The epidemiology of bladder cancer in West Yorkshire. A preliminary report on non occupational aetiologies. Carcinogenesis 1981; 2:343-347. (12) MORRISON AS, BURING JE, VERHOEK WG, et al. Coffee drinking and cancer of the lower urinary tract. JNCI 1982; 68:91-94. (13) Nagao M, TAKAHASHI Y, YAMANAKA H, SUGIMURA T. Mutagens in coffee and tea. Mutat Res 1979; 68:101-106. (14) MAHER VM, QUELLETTE LM, MITTLESTAT M, McCORMICK JJ. Syn ergistic effect of caffeine on the cytoxicity of ultraviolet irradiation and of hydrocarbon epoxides in strains of Xeroderma pigmentosum. Nature 1975; 258:760-763. (15) CHANG CC, Phillips C, TROSKO JE, Hart RW. Mutagenetic and epigenetic influence of caffeine on the frequencies of UV-induced ouabain-resistant Chinese hamster cells. Mutat Res 1977; 45:125 136. (16) Hava P, HEJLOVA A, Soskova L. Antimutagenic effects of caffeine during nitrosoguanidine-induced mutagenesis of Salmonella typhi murium cells and phages. Folia Microbiol (Praha) 1978; 23:45-54. (17) MENNIGMANN HD, Pons FW. Mutation induction by thymidine dep rivation in Escherichia coli B/r. I. Influence of caffeine. Mutat Res 1979; 60:13-23. (18) Hoover RN, STRASSER PH, Child MA, et al. Progress report to the Food and Drug Administration from the National Cancer Institute concerning the National Bladder Cancer Study. Bethesda, Md.: National Cancer Institute, 1979. (19) WAKSBERG J. Sampling methods for random digit dialing. J Am Star Assoc 1978; 73:40–46. (20) Gart JJ. Point and interval estimation of the common odds ratio in the combination of 2x2 tables with fixed marginals. Biometrika 1970; 57:471-475. (21) BRESLOW NE, Day NE. Statistical methods in cancer research. I. The analysis of case-control studies. Lyon, France: IARC, 1980. (22) Graham DM. Caffeine—its identity, dietary sources, intake and biological effects. Nutr Rev 1978; 36:97–102. (23) NOMURA T. Diminution of tumourigenesis initiated by 4-nitroquino line-1-oxide by post treatment with caffeine in mice. Nature 1976; 260:547-549. (24) Hicks RM, SEVERS N, CHOWANIEC J. Further investigations with a 2 stage model designed to detect low potency bladder carcinogens and cocarcinogens. In: Toxicology research projects directory. Vol 4, issue 10. Springfield, Va.: Natl Technical Information Service, 1979. (25) JOHANSSON SL. Carcinogenicity of analgesics. Long-term treatment of Sprague-Dawley rats with phenacetin, phenazone, caffeine and paracetomol. Int J Cancer 1981; 27:521-530. (26) WURZNER HP, LINDSTROM E, Vuataz L, LAGINBUHL H. A 2-year feeding study of instant coffee in rats. II. Incidence and types of neoplasms. Food Cosmet Toxicol 1977; 15:289-296. (27) Palm PE, ARNOLD EP, RACHWALL PC, LEYCZEK JC, TEAGUE KW, KENSLER CJ. Evaluation of the teratogenic potential of freshbrewed coffee and caffeine in the rat. Toxicol Appl Pharmacol 1978; 44:1-16. (28) Parsons WD, Neims AH. Effect of smoking on caffeine clearance.
(1) COLE P. Coffee-drinking and cancer of the lower urinary tract. Lancet Page 22
Differences in Breast Cancer Risk Factors According to the Estrogen Receptor Level of the Tumor 1,2,3
Nancy G. Hildreth, 4,5 Jennifer L. Kelsey, 6 Arnold J. Eisenfeld,' Virginia A. LiVolsi, Theodore R. Holford, and Diana B. Fischer 6, 9
ABSTRACT-One hundred and forty-eight postmenopausal breast cancer cases and 585 postmenopausal controls were included in an investigation of whether various risk factors for breast cancer are associated with the level of estrogen (E) receptor (ER) protein in the tumor. In an intracase analysis, the tumor ER level was positively associated with nulliparity, late age at first live birth, a history of benign breast disease, and having breast-fed at least 1 child and was negatively associated with previous use of E replacement therapy. A casecontrol analysis suggested that the first three variables, established risk factors for breast cancer, are associated with an increased risk for malignant breast tumors that are ER-positive but not for those that are ER-negative. This analysis did not provide a clear interpretation of the findings in the intracase comparisons with regard to prior breastfeeding and the use of E replacement therapy.—JNCI 1983; 70:10271031.
trols aged 45-74 years was drawn from these services to obtain a 3:1 ratio of controls to breast cancer cases in the larger case-control study (4). A variety of diagnostic groups was represented, the largest being diseases of the gallbladder, which accounted for 12.1% of the diseases diagnosed in the entire control group. No other single diagnostic group accounted for more than 5% of the controls. Excluding controls with gallbladder disease altered only slightly the distribution of characteristics of the entire control group, so no diagnostic groups were selectively excluded. Women were excluded as either cases or controls if they had a history of cancer of the breast or a genital organ, resided outside of Connecticut, or were unable to speak English. One control who had never had a menstrual period was also excluded. Only postmenopausal women were included in the present study since the initial age restriction of 45-74 years imposed by the larger study (4) resulted in too few premenopausal and perimenopausal women to permit a separate, meaningful statistical analysis. A woman was considered to be postmenopausal if she met one of the following criteria: 1) She had had a hysterectomy with a bilateral oophorectomy at any time prior to the interview, or 2) she had not had a menstrual period in the 2 years prior to the interview, unless the woman was less than 55 years age and had cessation of menses resulting from a hysterectomy with preservation of at least one ovary, in which case she was not considered to be postmenopausal.
Epidemiologic and experimental research strongly suggests that estrogens are centrally involved in the pathogenesis of breast cancer in both experimental animals and humans (1). For E to influence the physiologic activity or growth rate of a population of cells, cytoplasmic ER must be present (2). Since malignant breast tumors are known to exhibit marked variability in the concentration of this protein (3), the present epidemiologic investigation was performed to determine whether various risk factors for breast cancer were associated with the ER level of the tumor.
ABBREVIATIONS CL=confidence limit(s); CP=cytosol protein; E=estrogen(s); ER=E receptor(s); ER=ER-negative; ER*=ER-positive; OR=odds ratio(s).
Study design and subjects. This study used a case-control design and was linked to a larger case-control study of the epidemiology of cancers of the breast, ovary, and endometrium, details of which have been reported previously (4). Cases in this report were postmenopausal women of ages 45-74 years who were admitted to two hospitals in New Haven, Conn., between July 1977 and June 1980, who had a newly diagnosed and biopsy-confirmed breast cancer, and who had an ER assay performed on a tumor specimen. All breast tumors were assayed for cytoplasmic ER in a similar manner in one laboratory according to a procedure described in detail elsewhere (5). ER concentration is expressed in terms of femtomoles per milligram CP. In the analyses that dichotomize ER levels into positive and negative, tumors containing less than 30 fmol ER/mg CP are considered to be ER and those with 30 fmol/mg CP or more are classified as ER". This dichotomous classification will hereafter be referred to as the ER status of the tumor. Controls consisted of postmenopausal women of ages 4574 years admitted between July 1977 and March 1979 to the general surgery, orthopedic, and otolaryngology inpatient services of the same two hospitals. With the use of admitting records, a systematic sample of all potential con
Received September 21, 1982; accepted February 4, 1983. ? Supported in part by Public Health Service (PHS) grant 5R01CA20556 and CA-16359 (Yale Cancer Center Grant) from the National Cancer Institute (NCI) and by PHS training grant 1-T32-CA-09279-03 to Dr. Hildreth from the NCI. * Research procedures were in accord with the ethical standards of the Human Investigation Committee, Yale University School of Medicine. * Department of Preventive, Family, and Rehabilitation Medicine, The University of Rochester School of Medicine and Dentistry, Box 644, 601 Elmwood Ave., Rochester, N.Y. 14642. 5 This paper is based on dissertation research by Dr. Hildreth in partial fulfillment of the requirements for the Ph.D. degree at Yale University. © Department of Epidemiology and Public Health, Yale University School of Medicine, New Haven, Conn. 06510. Department of Obstetrics and Gynecology, Yale University School of Medicine. Department of Pathology, Yale University School of Medicine. 'We thank the physicians, staff, and patients of Yale-New !Haven Hospital and the Hospital of St. Raphael for their cooperation in this study, Ms. Theresa O'Connor for programming assistance, and Ms. Irene Visintin and Ms. Paula Preston for technical assistance.
The proportions of eligible women interviewed were 74% of cases and 72% of controls. However, 28% of the interviewed cases did not have an ER assay and, therefore, could not be included in the present study. This exclusion of cases did not introduce any obvious bias, because cases with an ER assay were not different from those without an assay in respect to any of the variables considered, including reproductive and menstrual history variables, prior benign breast disease, use of oral contraceptives and E-replacement therapy, body build indicators, and clinical stage of disease. In total, 148 breast cancer cases and 585 controls were available for statistical analysis. A standardized structured questionnaire administered by trained interviewers was used to collect information on exposure to risk factors and other relevant variables. So that recall could be facilitated, women were shown actual samples of E replacement therapy compounds and pictures of oral contraceptive packets. If a woman was uncertain as to the type of E used or the nature of previous surgery on her reproductive system, her physician was contacted. Analysis of data.-The data analysis in the present study can be divided into two parts. The first is an analysis of the breast cancer cases without controls for the purpose of determining whether the ER status (ER* vs. ER) as well as the actual ER concentration of malignant breast tumors varied according to any of the known or suspected risk factors for breast cancer. In the second part, variables identified as important in the intracase comparisons were considered in a case-control analysis to estimate the degree of association between various risk factors and breast cancer according to the ER status of the tumor. The measure of association used in both the intracase and case-control comparisons is the OR, which is an approximation to the relative risk (6). In the intracase comparisons, an OR of 3 implies that among women with breast cancer, those positive for the risk factor are three times more likely than those negative for the risk factor to have an ER* tumor (as opposed to an ER tumor). In the case-control comparisons where cases are distinguished according to the ER status of the tumor, an OR of 3 for ER* tumors would mean that women positive for the risk factor have an incidence rate of ER breast cancer estimated to be three times the rate of ER+ cancer in women negative for the risk factor, and an OR of 3 for ER tumors would mean that women positive for the risk factor have an incidence rate of ER breast cancer estimated to be three times the rate of ER cancer in women negative for the risk factor.
the mean ages being 62.7 and 60.5 years, respectively (P=0.07). No associations were noted with any of the other demographic variables considered, including marital status, race, education, or religion. In addition, the ER status of the tumor was independent of tumor histology and clinical stage of disease. With regard to reproductive and menstrual history variables, women with an ER* tumor were more likely to have never had a live birth than were women with an ER tumor, and among those who had a live birth, the percentage of ER* tumors increased linearly with the age at which a woman delivered her first live birth (table 2). In contrast, neither the number of live births nor the number of nonlive births was found to be associated with the ER status of the tumor. If one considers only women who had a live birth, ER* cases were more likely than were the ER ca: to have breast-fed at least 1 child, although the association was only of borderline statistical significance (table 2). The association with actual months of breast-feeding could not be evaluated adequately since too few women had breastfed for a long period of time. Very similar distributions were observed when the ER and ER cases were compared in terms of menstrual history variables, including age at menarche, age at menopause, history of bilateral oophorectomy, and prior hysterectomy. When exogenous E were considered, the ER status of the tumor was negatively associated with the number of milligram-months a woman was exposed to E replacement therapy (table 2). Although not shown, a similar association was seen when actual months of use of E replacement therapy and when only conjugated E were considered. Since none of the cases had used these compounds within 1 month of the time of surgery and since only 6 cases had used them within 2 months, the association with use of E replacement therapy is unlikely to be a result of occupation of receptor sites by exogenous E, a situation that would render them unavailable for detection. In any event, exclusion of the 6 cases with recent hormone use did not alter this finding. The association of ER status with use of oral contraceptives could not be assessed adequately since only 4 of the cases had ever used these compounds. Finally, as shown in table 2, a positive history of benign breast disease was more often reported by women with ER* tumors than by those with ER tumors, and the association had marginal statistical significance. The ER and ER cases were very similar with regard to other risk factors for breast cancer, including family history of breast cancer and
Analysis of Cases: Association of Case Characteristics With Tumor ER Status
The distribution of the 148 breast cancer cases according to the level of ER in the tumor is given in table 1. The value range was 0-729 fmol/mg CP, with a mean of 94 fmol/mg CP and a median of 50 fmol/mg CP. The tumors were classified as ER+ in 70.3% of the cases, and the remainder (29.7%) were ER On the average, women with ER+ tumors tended to be approximately 2 years older than those with ER tumors,
TABLE 2.—Distribution of ER* cases and ER cases by selected variables: OR, 95% CL, and tests of statistical significance
"OR are calculated as odds of an ER* tumor relative to the odds of an ER tumor (see “Methods”). Fisher's exact probability test. *Parous women only are included. dTest for linear trend. Milligram-months of use is defined as the No. of months of use of each compound multiplied by the milligrams of E in that compound summed over all compounds a woman had used.
body build indicators such as weight, height, and Quetelet's index (weight/height?). To examine the association of a variable with the ER status of the tumor while controlling for the effects of other variables, we performed a linear logistic regression analysis (7). Included in the model were nulliparity, age at first live í birth, ever having breast-fed, milligram-months of use of E : replacement therapy, history of benign breast disease, and age, variables that were significantly or marginally significantly associated with the ER status of the tumor in the bivariate analyses. With the exception of age, all of these variables were significantly (P<0.05) associated with the ER status of the tumor in the presence of the other variables. Adjusted OR, presented in table 3, provide an estimate of the magnitude of these associations.
Analysis of Cases: Association of Case Characteristics With Tumor ER Concentration The foregoing analyses were repeated preserving the actual ER concentration of the tumor. Since the ER concentration is log normally distributed, all statistical analyses were based on the transformed data (i.e., log10 ER concentration). Consistent with the previous results, a higher mean log10 ER concentration was found among women who were nulliparous, who had a late age at first live birth, who had ever breast-fed at least 1 child, and who had a history of benign breast disease, whereas an inverse association was still seen between the logarithm of the ER concentration and milligram-months of use of E replacement therapy.
TABLE 4.-Multiple regression coefficients and tests of statistical significance for the association between selected variables and the ER concentration (log10) of the tumora
TABLE 3.-Adjusted OR, 95% CL, and tests of statistical significance for the associations between selected variables and the ER status of the tumora
TABLE 5.—Unadjusted and adjusted" OR, and 95% CL for the associations between selected variables and breast cancer according to the ER status of the tumor
"Each adjusted OR is adjusted by means of linear logistic regression for all other variables in the table as well as for six other variables found to be associated with breast cancer in the main study (4), i.e., age at menopause, age at menarche, family history of breast cancer, being born in Europe, prior bilateral oophorectomy, and weight. Controlling for age did not alter these findings. 642 ER cases, 100 ER* cases, and 535 controls.
Each of these five variables was found to be independently associated (P<0.05) with the ER concentration of the tumor when they were included along with age in a multiple regression model (table 4). However, the multiple correlation coefficient associated with the model indicated that these patient characteristics explain only 17.6% of the total variation in the ER concentration among the cases. No additional associations were found with any of the other variables originally considered.
The ER* and ER cases were next compared to the control group with respect to each of the five variables associated with both the ER status and ER concentration of the tumor. Table 5 shows that an increased risk of ER* breast cancer was associated with nulliparity, a late age at first live birth, a history of having breast-fed at least 1 child, and a history of benign breast disease. However, for the first three variables, these associations were not quite statistically significant, the lower 95% CL being 0.9, 0.9, and 0.8, respectively. In contrast, these four factors were not associated with an increased risk for an ER breast tumor and, in fact, the OR were less than 1.0 in all cases. Use of E replacement therapy did not alter the risk of an ER* tumor but was associated with a somewhat increased risk for an ER tumor.
The positive association between breast-feeding and tumor ER level also requires confirmation. The inverse association between tumor ER level and use of E replacement therapy in this study could not be attributed to occupation of receptor sites by exogenous estrogens. An inverse association was also reported by Wallace et al. (11); although length of use was not considered in that study, ER cases were found to be four times as likely as ER* cases to have had prior exposure to exogenous estrogens (11). The only other study to consider length of use of E replacement therapy (9) found no association with tumor ER level, but this discrepancy may be attributable to differences between the two study populations, since Elwood and Godolphin (9) included both premenopausal and postmenopausal women in their analysis, whereas the present study was restricted to postmenopausal women. In the present study, the ER* and ER cases were compared to a control group so that the findings of the intracase analysis could be interpreted as OR for either ER* or ER breast cancer. Based on this analysis no conclusive interpretations can be made because most of the associations were relatively weak. It is difficult to say if the case-control analysis was affected by the controls not being entirely representative. However, this is a possibility since in the larger case-control study (4), where ER status was not considered, the OR for certain risk factors such as nulliparity, age at first live birth, and history of benign breast disease were toward the lower end of the range generally reported in the literature. Therefore, use of this control group, coupled with the relatively small sample sizes that resulted after division of the cases according to the ER status, may explain why some of the OR for the ER+ tumors were only of marginal statistical significance. Unfortunately, trends in these data cannot be compared and contrasted with those in other studies because no other investigators have used a control group when they analyzed ER data. Nonetheless, considering that nulliparity, age at first live birth, and previous benign breast disease are established risk factors for breast cancer, we believe our findings regarding them should at least be commented upon. Nulliparity, a late age at first live birth, and previous benign breast disease were clearly not associated with an increased risk of breast cancer among women with ER tumors. In contrast, the risk of an ER* tumor was elevated for each of these three variables, and, in each instance, the
In the present study, the ER level in malignant breast tumors, when considered as either a dichotomously or a continuously distributed variable, was found to be associated independently with five patient characteristics. Three of these variables, nulliparity, a late age at first live birth, and a history of benign breast disease, are established risk factors for breast cancer (8). In two previous studies (9, 10) no association was found between tumor ER status and nulliparity. However, Elwood and Godolphin (9) noted that women who had their first live birth at the age of 20 years or later were twice as likely to have an ER* tumor as were women whose first live birth occurred before the age of 20 years, a finding consistent with the present study. No other investigators have reported examining the association between tumor ER level and previous benign breast disease.
this study is that in the examination of the epidemiology of breast cancer, consideration of various subgroups of tumors on the basis of relevant biochemical or pathologic parameters (23) may increase our knowledge of the factors associated with the occurrence of breast cancer and ultimately may result in a clearer understanding of the disease process.
magnitude of the increased risk falls into the range generally reported in the literature (8). If these findings were replicated by others, it would suggest that ER* and ER breast tumors may in some ways be etiologically dissimilar. This observation would not be inconsistent with the large body of literature which indicates that these three factors increase the risk for breast cancer as a whole (8), since the majority of breast tumors both in this study and in other studies (3) are ER*. Although it is presently unknown why nulliparity, a late age at first live birth, and previous benign breast disease increase a woman's risk for breast cancer, current theories suggest that endogenous E are centrally involved and that the mammary tissue of women possessing any of these risk factors may receive greater E stimulation than the mammary tissue of women without these characteristics (12-16). In some cases, this excess E stimulation could result in the initiation or promotion of a neoplastic transformation in a population of cells. Should this be the case, then continued E stimulation of the preclinical neoplastic cells would be expected to promote cell division only in the cells that contain ER, because those that lack this cytoplasmic protein would be unresponsive to E. This preferential growth stimulus of the ER+ cells could then result in a relatively greater number of ER* cells than ER- cells in the clinically detectable tumor. Therefore, these three risk factors, by means of persistent or excessive Estimulation, could increase the risk for an ER* tumor. No biologic explanation can be proposed to explain how exogenous E (E replacement therapy) might be associated with an ER tumor as was suggested by the present data. Furthermore, the role of E replacement therapy in the etiology of breast cancer in general has not been clearly established (17-22). This inverse association between tumor ER level and use of E replacement therapy is in fact somewhat paradoxical, since the findings regarding nulliparity, a late age at first live birth, and prior benign breast disease suggest that endogenous E may increase the risk for an ER+ tumor. However, differences in the chemical composition of exogenous compared to endogenous E and differences in the ages of exposure to exogenous versus endogenous E may at least partly account for these findings. Since breast-feeding is generally thought not to affect the risk for breast cancer and since duration of breast-feeding could not be considered in the present study, the finding of a positive association between breast-feeding and tumor ER level should be confirmed in other studies before much attention is paid to possible mechanisms. The present study suggests that ER+ and ER- breast cancers may share some risk factors but not others. This statement, however, is made cautiously since some of the associations reported here are not entirely consistent with other studies, and, given the number of associations considered, some of these may represent chance occurrences. Fur thermore, a general lack of knowledge of ER levels in normal ; breast tissue as well as of the stability of ER levels in malignant tumors over time within the same women and the lack of data on serum levels of endogenous E make it difficult to interpret these findings with a high degree of certainty. Nonetheless, while the specific findings of this study remain to be verified, a more general implication of
REFERENCES (1) KIRSCHNER MA. The role of hormones in the etiology of human breast cancer. Cancer 1977; 39:2716-2726. (2) Gorski J, GANNON F. Current models of steroid hormone action. A critique. Annu Rev Physiol 1976; 38:425-449. (3) McGuire WL, CARBONE PP, Sears ME, Escher GC. Estrogen recep tors in human breast cancer: An overview. In: McGuire WL, Carbone PP, Vollmer EP, eds. Estrogen receptors in human breast cancer. New York: Raven Press, 1975:1-15. (4) KELSEY JL, Fischer DB, HOLFORD TR, et al. Exogenous estrogens and other factors in the epidemiology of breast cancer. JNCI 1981; 67:327-333. (5) EISENFELD AJ, Schwartz PE, MORRIS JM. Estrogen and progesterone receptors in vaginal and uterine adenocarcinomas following estro gen use. Gynecol Oncol 1980; 10:63-68. (6) Fleiss JL. Statistical methods for rates and proportions. New York: Wiley, 1973. (7) Cox DR. The analysis of binary data. London: Methuen, 1970. (8) KE JL. A review of the epidemiology of human breast cancer. Epidemiol Rev 1979; 1:74-109. (9) Elwood JM, GODOLPHIN W. Oestrogen receptors in breast tumours: Associations with age, menopausal status and epidemiological and clinical features in 735 patients. Br J Cancer 1980; 42:635-644. (10) LESSER ML, Rosen PP, SENIE RT, Duthie K, MENENDEZ-Botet C, Schwartz MK. Estrogen and progesterone receptors in breast carcinoma: Correlations with epidemiology and pathology. Cancer 1981; 48:299-309. (11) WallACE RB, SHERMAN BM, BEAN JA. Relationship between artificial menopause, previous estrogen consumption, and estrogen receptor content of breast neoplasms-preliminary communication. Oncol ogy 1980; 37:111-113. (12) SHERMAN BM, KORENMAN SG. Inadequate corpus luteum function: A pathophysiological interpretation of human breast cancer epide miology. Cancer 1974; 33:1306-1312. (13) KORENMAN SG. The endocrinology of breast cancer. Cancer 1980; 46:874-878. (14) Drife JO. Breast cancer, pregnancy, and the pill. Br Med J 1981; 283:778-779. (15) GRATTAROLA R. Anovulation and increased androgenic activity as breast cancer risk in women with fibrocystic disease of the breast. Cancer Res 1978; 38:3051-3054. (16) SITRUK-WARE LR, STERKERS N, Mowszowicz I, Mauvais-Jarvis P. Inadequate corpus luteal function in women with benign breast Page 23
3, 4 Olga G. Joly, D.D.S., Sc.D., Jay H. Lubin, Ph.D., and Magaly Caraballoso, M.D. 6,7
ABSTRACT—A retrospective epidemiologic study of 826 cytologically and/or histologically confirmed lung cancer cases (219 females and 607 males), 979 hospital controls, and 539 neighborhood controls was undertaken in Havana, Cuba, to investigate whether the high lung cancer mortality rates in this country could be explained by the cigarette and cigar consumption habits, including the smoking of darktobacco cigarettes. Relative risk(s) (RR) of lung cancer among cigarette smokers were 7.3 in women and 14.1 in men and increased consistently with various measures of exposure to smoke. The findings suggested that duration of smoking, daily number of cigarettes consumed, and inhalation practices have independent effects. Most Cubans smoked dark tobacco. RR were higher for dark-tobacco users than for light-tobacco users (RR 8.6 vs. 4.6 for women and 14.3 vs. 11.3 for men), but the differences were reduced after adjustment for amount smoked. Cigarette smoking was associated with all histologic types of lung cancer, although the risk for adenocarcinoma was lower than that for the other types. Men who smoked exclusively cigars had a fourfold risk of lung cancer. Mixed smokers (i.e., cigar and cigarette smokers) had a greater RR than cigarette-only smokers (15.0 vs. 14.1), which was perhaps related to the unusually deep and frequent inhalation of cigar smoke. The data support the hypothesis that smoking patterns account for the higher lung cancer mortality in Cuba than in other Latin American countries.—JNCI 1983; 70:1033-1039.
from patients admitted to all the city's 12 main general hospitals with a tentative diagnosis of lung cancer; females were ascertained from 1978 to 1980 and males from 1979 to 1980. In each of the participating hospitals, a staff member was assigned to ensure prompt identification and reporting of all suspected lung cancer patients to the Central Study Office located at the Institute of Oncology and Radiobiology. Only those patients who had their final diagnosis confirmed by cytology and/or histology were included in the study. To ensure consistency of criteria, an ad hoc Pathology Review Committee performed all typing according to the World Health Organization's Classification of Lung Cancer. A hospital control selected from patients with a current admission for a non-smoking-related disease was matched to each case by sex, age (£5 yr), hospital of admission, and admission date (+15 days). When several controls were available, the closest in admission date was chosen. Not accepted as controls were patients with diseases such as cancers of the upper digestive and respiratory tracts, pancreas, kidney, and urinary bladder; chronic bronchial disease; duodenal and gastric ulcers; and cardiovascular diseases known to be related to smoking. The main diagnostic categories among female and male hospital controls, respectively, were in percentages: digestive diseases (15.1 and 12.8), malignant and benign neoplasms (21.9 and 23.7), eye diseases (6.4 and 15.6), respiratory diseases including tuberculosis (14.5 and 9.3), traumatisms (7.3 and 15.6), diabetes and metabolic disorders (9.6 and 2.6), and skin diseases (4.1 and 7.3); in smaller proportions were the categories arthritis,
ABBREVIATION USED: RR=relative risk(s).
A report on the health conditions in the Americas in the early 1970's revealed that Cuba had the highest lung cancer mortality rates for women in the American region; ageadjusted rates per 100,000 were 8.7 in Cuba as compared to 5.8 in the United States, 3.5 in Argentina, and 2.3 in Uruguay. Corresponding rates for men were the fourth highest in the American region; they were 24.3 in Cuba as compared to 26.9 in the United States, 27.8 in Argentina, and 25.7 in Uruguay (1). Of possible significance is that a high proportion of cigarettes smoked in Latin America are made of a dark, locally grown tobacco that has a higher concentration of phenols and nicotine and is more alkaline than the light Virginia-grown type of tobacco (2). Skin painting experiments on mice suggested that refined tar from Colombian dark tobacco has about twice the carcinogenic potential as refined tar from U. S. tobacco (3). A study of mortality patterns in 11 Latin American cities showed that La Plata, Argentina, had the highest lung cancer male-adjusted mortality rate (59.0/100,000); the explanation offered was the known high consumption of dark tobacco among the population of this city (4). This paper reports results of a case-control study undertaken to evaluate the role of smoking habits in the causation of lung cancer among Cuban women and men.
Received September 21, 1982; accepted January 25, 1983. 2 Sponsored jointly by the International Agency for Research on Cancer (by subcontract to National Cancer Institute Prime Contract No. NOICP55666) and the Ministry of Public Health, Cuba. 3 Cancer Training Branch, Division of Resources, Centers, and Community Activities, National Cancer Institute, National Institutes of Health, Public Health Service, U.S. Department of Health and Human Services, Bethesda, Md. 20205. * Address reprint requests to Dr. Olga G. Joly, Blair Building, Room 722, National Institutes of Health, Bethesda, Md. 20205. “Environmental Epidemiology Branch, Division of Cancer Cause and Prevention, National Cancer Institute. National Cancer Registry, Institute of Oncology and Radiobiology, Havana, Cuba. ? We acknowledge the valuable cooperation of the staffs of the International Agency for Research on Cancer and the Ministry of Public Health, Cuba. We are grateful to Dr. D. J. Joly (formerly, Pan American Health Organization) and Dr. W. Blot (National Cancer Institute) for their critical comments and suggestions. We also thank Mrs. Elaine Carlson (National Cancer Institute) for her assistance in the preparation of the manuscript.
A case-control study of lung cancer was conducted in Havana, Cuba, a city of approximately 2 million inhabitants (one-fifth of the country's population). Cases were drawn
asthma, anemia, renal diseases, and other miscellaneous conditions. For each case residing in Havana, an additional neighborhood control was selected and was matched with regard to place of residence (on the same block or as close as possible), sex, and age (5 yr) in an effort to adjust for socioeconomic and general environmental factors. Selection of the neighborhood control was achieved through the cooperation of neighborhood community affairs committees who made available listings of residents along with information on sex and date of birth. Initially, two hospital controls were matched to each non-Havana resident case, but this was reduced to one hospital control after the 2d year of the study. As potential study subjects were identified, interviews were conducted with the use of a standardized structured questionnaire. Data were collected on demographic characteristics; educational level; lifelong smoking, residential, and occupational histories; alcohol consumption; and exposure to cooking fuels. The population proved extremely cooperative, with all persons contacted agreeing to participate in the study. Although the neighborhood controls were interviewed at home and the cases and hospital controls were interviewed in the hospital, minimal interviewer bias was anticipated since most hospital interviews were not conducted at bedside but in adjacent private surroundings. While complete interviewer blindness would have been impossible to achieve, the interviewers were made aware that the questionnaire was administered only to gather general health and disease data from both hospitalized and nonhospitalized populations. Data on a total of 826 confirmed lung cancer cases (219 females and 607 males), 979 hospital controls, and 539 neighborhood controls were analyzed with the use of procedures for matched (5) and unmatched (6, 7) studies. Hospital and neighborhood controls for both female and male cases were quite similar regarding sociodemographic characteristics, such as years of education, marital status, and age at interview (table 1). Furthermore, the distribution of the smoking-related variables in both control groups and the RR of lung cancer resulting from the comparison of cases with either control group with the use of matched or unmatched procedures were also similar. It was decided thenceforth to present only the unmatched analyses in which cases were compared with both types of controls combined into a single group. Measures of exposure to cigarette smoking (i.e., duration of smoking, number of cigarettes smoked daily, and frequency and depth of inhalation) were based on lifetime information on all cigarette brands used; 92% of the respondents smoked one or two brands, and 7% smoked a total of three brands. Identification of dark- versus light-tobacco smokers was made from reported brands. Measures of exposure to tar were based on information from only the four most popular cigarette brands currently smoked, which were analyzed by the Oak Ridge National Laboratory, Oak Ridge, Tenn., for yields of tar, nicotine, CO, and CO2. Therefore, a substantially smaller number of subjects were available for the investigation of smoke content since only persons consuming exclusively the four brands tested could be used.
The ages of the lung cancer patients ranged from 23 to 89 years, with approximately one-half of the females and twothirds of the males 60 years old or over at the time of diagnosis. Education level was similar in all groups, but slightly lower mean years of schooling were found among the cases. Approximately 60% of the women and 75% of the men were married (table 1).
In this study 167 of the 219 female cases (76.3%) and 595 of the 607 male cases (98.0%) ever smoked regularly, as compared to 31.0 and 80.3% of female and male controls, respectively. The corresponding proportions for female hospital and neighborhood controls were 30.5 and 31.8%, whereas for males they were 80.5 and 80.1%. A total of 12.2 and 8.1% of female and 38.9 and 39.9% of male hospital and neighborhood controls, respectively, smoked cigarettes for 50 years or more. Among case smokers, 1 female and 43 males smoked cigars only. In the following analyses cigaronly and pipe-only smokers were excluded.
The overall RR of lung cancer in cigarette smokers were 7.3 and 14.1 for females and males, respectively. Most e respondents who smoked consumed the local dark-tobacco cigarettes exclusively, as reported by 80 and 96% of female and male cases, respectively, and by 68 and 95% of their corresponding controls. There were increased risks of lung cancer in both sexes associated with smoking both light and dark tobaccos, but the excess was greater for dark tobacco (RR=8.6 for dark tobacco vs. 4.6 for light tobacco among sense females; RR=14.3 for dark tobacco vs. 11.3 for light tobacco among males) (table 2). With either kind of tobacco, the longer the duration of smoking or the greater the total number of cigarettes consumed the higher the risk, all trends being highly significant (P<0.001) (table 3). When gradients of increasing risk for duration of smoking and for number of cigarettes smoked daily were estimated with the use of a logistic model that included linear and quadratic effects, the differences in trend between dark- and light-tobacco users were not statistically significant. Analysis of all cigarette smokers regardless of the kind of tobacco they favored indicated that cases of both sexes not only had a much higher proportion of smokers than controls, but also smoked for many more years, consumed more cigarettes daily, and were exposed to higher amounts of tar
No. of years of smoking <20 20-29 30-39 40-49 250 Mean No. of cigarettes/ day 1-9 10-19 20-29 230 Lifetime tar exposure I II III
"All RR relative to those of nonsmokers; all tests of linear trend, P<0.001. Levels of tar in milligrams for females were: 1=<11,000 and II=211,000. For males they were: I=<15,640, II=15,640-24,079, and III=224,080.
Nonsmokers Duration of dark-to- bacco smoking, yr 1-19 20-39 240 Duration of light-to bacco smoking, yra 1-19 220 Duration of mixed-to bacco smoking, yr 1-39 240 No. daily, dark-tobacco 1-9 10-19 20-29 230 No. daily, light to bacco 1-19 220 No. daily, mixed to bacco 1-19 220
(table 4). Highly significant (P<0.001) linear trends in RR for these three measures of cigarette smoke exposure were noticed in both sexes; increases of 6.2-fold, 2.9-fold, and 4.5fold for duration of smoking, daily cigarette consumption, and lifetime tar exposure, respectively, were found between the lowest and the highest levels of exposure in females; the corresponding increases in males were 4.3-fold, 4.0-fold, and 1.6-fold. When RR for duration of smoking, daily cigarette consumption, and lifelong tar exposure were each computed with separate adjustment for the other two, there remained a consistent pattern of increasing risk with duration of smoking and mean number of daily cigarettes (table 5). The increasing risk with duration of smoking was greater for females (over fourfold) than for males (twofold), although the increases with the daily number of cigarettes were greater for males. While RR increased with tar intake, the risk patterns were less convincing than for duration of smoking and amount smoked. These results suggest that duration of smoking and number of cigarettes acted independently, whereas risk with tar intake was less clear, due perhaps to the high correlation of tar with duration of smoking. Because of the small number of brands tested and the high correlation between the smoke components, the effects of tar could not be isolated from other constituents of tobacco smoke, such as nicotine, CO, or CO2. The earlier in life cases began to smoke, and the more frequently and the more deeply they inhaled, the higher were their risks of developing lung cancer, even after control for duration of cigarette use; linear trends were significant and did not differ much between females and males (table 6). Cases of both sexes who reported cessation of smoking 14 years prior to interview had an increased risk of lung cancer that declined after the 5th year of smoking cessation.
Histologic Type of Lung Cancer The four main histologic types of lung cancer were almost equally represented among females; in males the squamous type was predominant (36%), followed by the poorly differentiated type (27%), and the undifferentiated type (23%), ! whereas adenocarcinoma was by far the less frequent (1470). The risk of lung cancer increased significantly with longer duration of cigarette smoking for all histologic types (P<0.001); the greatest risks were associated with the squamous type in females and the undifferentiated type in males (table 7). In spite of the distributional differences in duration of cigarette use and in histology, the risks in smokers of 50 years or more relative to those in nonsmokers were not much different between the sexes. The RR for 1-29 years of exposure were larger in males for all but the adenocarcinoma type. Thus females had a steeper gradient of RR with years of cigarette consumption than did males for the squamous. undifferentiated, and poorly differentiated types, whereas for adenocarcinoma the gradients with duration were similar in both sexes. Although squamous, undifferentiated, and poorly differentiated types of lung cancer did not differ significantly among themselves insofar as duration of smoking, females with all these three types were exposed significantly longer to cigarette smoke than those with adenocarcinomas (table 8). Adenocarcinoma was the most common histologic type among nonsmokers (48 and 42% in females and males, respectively) whereas it was infrequently seen among smok
TABLE 7.-RR of lung cancer according to duration of smoking and histologic type of lung cancer
Age started, yr 25 15-24 <15 Frequency of inhala tion Rarely or never Sometimes Always Depth of inhalation Never Slightly Medium Deeply Years since last smoked Current smokers 1-4 25
"Risks relative to those in nonsmokers; all trends were significant, P<0.001. b There were 54 females and 219 males with the squamous type. “There were 58 females and 138 males with the undifferentiated type. Inasmuch as there were no nonsmoking cases among the 138 males, one case was added in calculation of RR. dThere were 48 females and 165 males with the poorly differentiated type. € There were 58 females and 82 males with adenocarcinoma. Page 24
Cigar Smoking in Males A separate analysis was done on the 216 male cases and 389 controls who smoked either cigars only or cigars plus cigarettes (mixed smokers). (The 5 females, 1 case and 4 controls, who smoked cigars were excluded from analysis.) The risks of lung cancer in both groups of smokers were significantly greater than those in nonsmokers; those males who smoked only cigars had an overall risk of 4.4 and those
males who were mixed smokers had an overall risk of 15.0. Risks increased significantly with duration of cigar smoking in those who smoked only cigars as well as in those who were mixed smokers, but the increases with the number of cigars smoked daily in either group of smokers were not significant (table 9). Male cases who smoked cigars exclusively began their habit earlier in life than the controls; however, after control for duration it was found that age at which smoking was started or years since cessation of smoking did not influence the risk of lung cancer, although the power to detect effects independent of duration was small. In contrast, the RR did increase significantly with frequency and depth of smoke inhalation even after adjustment for duration of smoking (table 10).
The similarity in smoking prevalence and smoking characteristics between both groups of controls was likely due to the high prevalence of smoking in Cuba and the acceptance of hospital controls whose current admission was for a nontobacco-related disease. In fact, in a 1979 population-based survey in Havana, the frequency of smoking was found to be 33.0% in women and 78.0% in men over 30 years of age (Marcias Castro I: Personal communication). These smoking prevalences are considerably higher than those reported in the early 1970's for eight other Latin American cities, where the overall frequencies of smoking were 18.5 and 47.8%, respectively, in the age group of 25-74 years, ranging from 6.1% (Lima, Peru) to 28.1% (Santiago, Chile) for females and from 35.4% (Lima) to 55.4% (La Plata) for males (2). Furthermore, not only a greater proportion of Cubans than other Latin Americans smoked, but also those who did smoke consumed almost twice as many daily cigarettes, started smoking earlier in life, smoked for more years, and preferred the dark type of tobacco and nonfilter ciga
rettes. These differences in smoking patterns and the high risk of lung cancer related to cigarette smoking found in our study indicate that smoking habits may account for the higher lung cancer rate in Cuba than in other areas of Latin America. On the basis of calculations of attributable risk, we estimate that a maximum of 66% of female and 91% of male lung cancer cases in Cuba are due to smoking. Analyses of other variables, such as place of birth, residence, alcohol consumption, and environmental exposure measured by lifelong presence and type of nearby factories, occupation, and use of cooking fuels, did not discriminate between cases and controls. A separate analysis by residence (Havana vs. not Havana) failed to show a differential risk between lifetime rural or urban residency. It has been suggested (3, 4) that smoking the dark type of tobacco product is more deleterious than smoking the light variety. We did find a higher level of risk linked to smoking dark-tobacco cigarettes, but the excess was reduced after accounting for duration and number of cigarettes smoked daily. However, the power to detect a difference between the tobacco types was small due to the very low proportion of light-tobacco cigarette smokers. It has been assumed that, because of their alkalinity, dark-tobacco cigarettes could not be inhaled as deeply and as frequently as light-tobacco ones, but we found that the tobacco type had no influence on the inhalation practices. We believe that the differential risks by type of tobacco require further exploration, preferably in communities where the discrepancy between proportions of light- and dark-tobacco users is not as extreme as in Cuba. The RR of lung cancer associated with smoking were lower in women than in men, partly related to lower cigarette consumption by women. The RR tended to be slightly lower in women for all smoking categories, such as number of years smoked, mean number of daily cigarettes consumed, lifetime tar exposure, and age starting to smoke, in agreement with prior reports (8-11). However, differential risk disappears in frequently and deeply inhaling females for whom the risk is even slightly higher than that in males. This same finding was noted for squamous cell cancers, the histologic type with the strongest relation to smoking. Hence it appears that no biologic protection is afforded to smoking women, and the lower RR result from somewhat lower average smoke exposure within each of the broad categories of age smoking started, duration of smoking, and daily cigarette consumption. Although differences in the histologic classification criteria complicate the comparison with data from other studies, most studies have reported a preponderance of adenocarcinoma (12-16), squamous type (17), undifferentiated type (18), or similar proportions of undifferentiated type and adenocarcinoma (19, 20) in females. In our study, no histologic type was predominant among females. Among males, however, there was agreement in that the squamous type of cancer appeared to predominate (15, 21, 22). Similar to the results in other reports, the gradient in RR was steepest for squamous cell carcinomas, but increased risks were found for all types of lung cancer including adenocarcinoma. Because male cigarette smokers had a threefold higher risk of lung cancer than cigar-only smokers (14.1 vs. 4.4), it
was expected that mixed smokers (i.e., smokers of cigarettes plus cigars) would experience a risk intermediate between the risks of the first 2 groups. However, we found instead that cases who were mixed smokers had a slightly greater risk than those smoking cigarettes only (15.0 vs. 14.1); to account for this unexpected finding, we examined the inhalation practices among cases and controls. Results indicated that 31.7 and 41.5% of all cases who smoked cigars only were frequent and deep inhalers of cigar smoke, whereas among those who were mixed smokers the percentages were higher, 54.0 and 60.2%, respectively (table 11). Even greater differences in inhalation practices were observed between both groups of smokers in the controls. Conversely, the inhalation practices of cigarette smoke did not differ significantly between smokers of cigarettes exclusively and mixed smokers, except for inhalation frequency among the controls. It thus appears that the manner in which mixed smokers inhaled their cigarette smoke determined the manner in which they also inhaled their cigar smoke, which in turn may explain the high risk among the mixed smokers.
(1) Pan American Health Organization. Health conditions in the Amernization, 1977 (sci publication No. 337). sty (3) MUÑOZ N, CORREA P, Bock FG. Comparative carcinogenic effect of
two types of tobacco. Cancer 1968; 21:376-389. (4) PUFFER RR, GRIFFITH GW. Patterns of urban mortality: Report of the inter-American investigation of mortality. Washington, D.C.: Pan American Health Organization, 1967 (sci publication No. 151). (5) BRESLOW NE, Day NE, HALVERSEN KT, PRENTICE RL, Sabai C. Estimation of multiple relative risk functions in matched case control studies. Am J Epidemiol 1978; 108:299-307. (6) GART JJ. Point and interval estimation of the common odds ratio in the combination of 2x2 tables with fixed marginals. Biometrika 1970; 57:471-475. (7) PRENTICE RL, Pyke R. Logistic disease incidence models and case control studies. Biometrika 1970; 66:408-412. (8) Health Services and Mental Health Administration, U.S. Public Health Service. The health consequences of smoking. A report of the Surgeon General. Washington, D.C.: U.S. Govt Print Off, 1971 (DHEW publication No. (HSM)71-7513). (9) Office of the Assistant Secretary for Health, Office on Smoking and Health, U.S. Public Health Service. Smoking and health. A report of the Surgeon General. Washington, D.C.: U.S. Govt Print Off, 1979 (DHEW publication No. (PHS)79-50066). (10) The health consequences of smoking for women. A report of the Surgeon General. Washington, D.C.: U.S. Govt Print Off, 1980. (17) The health consequences of smoking. A report of the Surgeon General. Washington, D.C.: U.S. Govt Print Off, 1982. (12) Beamis JF JR, STEIN A, ANDREWS JL JR. Changing epidemiology of lung cancer. Increasing incidence in women. Med Clin North Am
Clinical Study of Recombinant DNA-Produced Leukocyte Interferon (Clone A) in an Intermittent Schedule in Cancer Patients 1.2
Jorge R. Quesada, M.D.,and Jordan U. Gutterman, M.D. 3, 4, 5
ABSTRACT — The effects of recombinant DNA-produced leukocyte interferon (IFLrA) were studied in 37 patients with metastatic cancer who received sequentially escalating doses of 9-86 million units (MU) of IFLrA by im injection twice weekly. The IFLrA was absorbed rapidly and reached a peak serum concentration 6-8 hours after injection. Serum concentration of IFLrA increased proportionately with the dose. The most common side effects included fever, chills, asthenia, anorexia, and weight loss, and leukopenia, granulocytopenia, and lymphopenia occurred frequently. Elevation of serum glutamic-oxaloacetic transaminase was frequent above doses of 50 MU. All side effects were reversible by discontinuation of the drug. Antibodies to IFLIA were detected in 3 patients while on treatment. The presence of antibodies coincided with drastic reduction in serum IFLrA concentration and, in 1 patient, with relapse of disease. Objective tumor responses were documented in patients with lymphomas but not in other groups of patients.—JNCI 1983; 70:1041-1046.
other patient with renal cell carcinoma had shown mixed effects to partially purified leukocyte interferon. A minimum of 4 weeks was required between the last anticancer treatment and the initiation of IFLrA treatment. Eligibility criteria included histologic confirmation of metastatic cancer, a performance status of 60 or more (Karnofsky scale), an expected survival of 12 or more weeks, and preserved renal, liver, and bone marrow functions. All patients signed informed consent forms. Pregnant patients and patients with intercurrent illnesses or clinically active viral diseases were excluded from the study. Interferon preparation. —IFLrA was produced and purified by Hoffmann-La Roche Inc. and Genentech (San Francisco, Calif.) as previously described (1,5) and was provided in ampuls of 3, 18, and 50 MU/ml. The purified protein was homogeneous by sodium dodecyl sulfate polyacrylamide gel electrophoresis with a specific activity of 2-4X108 U/mg protein. The final freeze-dried preparation contained human albumin and was reconstituted immediately before use. Study design.—All patients had thorough clinical examinations and careful staging of their disease before initiation of IFLrA treatment. Complete blood cell counts, differential count, platelet and reticulocyte counts, 12-channel blood chemistry, coagulation parameters, urinalysis, electrocardiogram, and chest X-ray were performed within 1 week of the first injection. Patients were treated in groups at sequentially escalating doses of 9, 18, 36, 50, 68, and 86 MU for a total of 12 weeks. IFLrA was injected im at a fixed dose for each
Recent production of leukocyte interferon in Escherichia coli by recombinant DNA techniques (1, 2) has led to the further purification of one of the 8-12 molecular species of leukocyte interferon (3, 4). This clone (clone A) has been purified to homogeneity with monoclonal antibodies, resulting in a protein with 2-4X10% antiviral U/mg protein (1,3, 5). IFLrA has been shown to have antiproliferative activity and to activate natural killer cells in vitro (6, 7). A preliminary clinical study demonstrated that IFLrA is well absorbed after im injection, is well tolerated in doses up to 198X106 U, and elicits biologic effects in vivo, including antitumor effects in patients with metastatic cancer (8). Results of a subsequent clinical study of IFLrA in an intermittent schedule are presented here with emphasis on tolerance, pharmacology, and antitumor effects.
ABBREVIATIONS USED: APhralkaline phosphatase; IFLrA=recombinant DNA-produced leukocyte interferon, clone A; LDH=lactate dehydrogenase; MU=million units; NPDLL=nodular poorly differentiated lymphocytic lymphoma; SGOT=serum glutamic-oxaloacetic transaminase.
Patient population.—Twenty-one female and 16 male patients with metastatic cancer were entered in the study. Of these 37 patients, 11 had lymphoma, including 3 with welldifferentiated lymphocytic lymphoma, 3 with NPDLL, 2 with large cell lymphoma, and 1 with cutaneous T-cell lymphoma. Also included in this group of patients with lymphoma were 2 patients with Waldenström's macroglobulinemia. The study also included 9 patients with breast cancer, 9 with colon cancer, 4 with renal cell carcinoma, and 1 patient each with fibrohistiocytoma, Kaposi's sarcoma, osteosarcoma, and ovarian cancer. Patients' ages ranged from 17 to 78 years with a median age of 54 years. Thirty-one patients had received prior chemotherapy or hormonal therapy. All patients had failed prior treatment or had relapsed at the time they were entered into the study. Two patients had received previous treatment with interferon. One patient with lymphoma experienced objective tumor regression in a preliminary study with IFLIA (8); the
Received October 1, 1982; accepted February 2, 1983. ? Supported by Public Health Service grant CA-05831 from the National Cancer Institute, by a grant from the James E. Lyon Medical Research Foundation, and by a grant from Hoffmann-La Roche Inc. Department of Clinical Immunology and Biological Therapy, The University of Texas, M. D. Anderson Hospital and Tumor Institute, Texas Medical Center, 6723 Bertner Ave., Houston, Tex. 77030. * Senior Clayton Foundation investigator. Research was conducted in part by the Clayton Foundation for Research. We thank Juanita Gomez R.N., Jean Hermann R.N., Donna Jones R.N., Andrea Sacolick R.N., Paula Trahan R.N., Mrs. Judy Rotbein, Mrs. Carol Hunter, Mrs. Jannie Perkins, Mrs. Marion Stringer, Mrs. Kathy Marshall, and Mrs. Rosaleen Fisk for assistance in conducting the study and preparing the manuscript. We also acknowledge the participation of Dr. H. E. Spiegel, Department of Clinical Laboratory Research, HoffmannLa Roche Inc., Nutley, N.J., and Dr. M. Kramer, Department of Immunotherapy, Hoffmann-La Roche Inc., who conducted pharmacologic studies and itibody tests, and Dr. Seymour Fein, study coordinator, Hoffmann-La Roche Inc. Page 25
optimally stimulated by intermittent exposure. Further studies in this area will be required. IFLrA seems to have a preferential activity in patients with lymphoma, as seen with the partial remissions obtained in this study. Activity in other groups of patients seems marginal. Some studies have indicated selectivity of leukocyte interferon for cancers of lymphoid origin (22-24). This selectivity could be enhanced when a single molecular species is used. It is not possible at this time to determine whether the lack of response of solid tumors was due to the schedule selected for this study or to the resistance of the malignant cells to IFLrA. Of interest also is the fact that patients with lymphoma experienced more profound myelosuppression compared to the myelosuppression experienced by patients with solid tumors. A relationship between myelosuppression and tumor response has been observed with partially purified leukocyte interferon (15, 25, 27). Previous studies with partially purified polyclonal leukocyte interferon indicated activity in patients with breast (15, 25) and colon carcinomas (26) when exposed to daily administration of interferon. More recently, we have also reported antitumor activity in renal carcinoma (27) and chronic myelocytic leukemia (28). Conceivably, a combination of different molecular species of leukocyte interferon or a combination of different types of interferon may be required for a broader range of activity. In addition, partially purified polyclonal interferon preparations may contain antitumor substances other than interferon. There is a role for IFLrA in the treatment of cancer patients, but its precise indications, optimal doses, and schedules remain to be determined. This study contributes in defining some ancillary principles for further clinical application of interferons.
(7) Evinger M, Maeda S, PESTKA S. Recombinant human leukocyte interferon produced in bacteria has antiproliferative activity. J Biol Chem 1981; 256:2113-2114. (8) GUTTERMAN JU, Fein S, QUESADA J, et al. Recombinant leukocyte A interferon: Pharmacokinetics, single-dose tolerance and biologic effects in cancer patients. Ann Intern Med 1982; 96:549-556. (9) RUBENSTEIN S, FAMILLETTI PC, Pestka S. Convenient assay for inter ferons. J Virol 1981; 37:755-758. (10) STAEHELIN T, DURRER B, SCHMIDT J, et al. Monoclonal antibodies to human leukocyte interferons: Their use in assay and purification. 1 Tex Rep Biol Med 1982; 41:43-58. (11) VOLLER A, BIDWELL DE, BARTLETT A. The enzyme linked immuno sorbent assay (ELISA). Guernsey, England: Dynatech Europe, 1979. (12) Trown PW, KRAMER MW, DENNIS RA, et al. Antibodies to human leukocyte interferon in cancer patients. Lancet 1983; 1:81-84. (13) MOGENSEN KE, Daubas PH, GRESSER I, SERENI D, VARET B. Patient with circulating antibodies to interferon. Lancet 1981; 1:1227 1228. (14) VALLBRACHT A, TREUNER J, FLEMING B, JOESTERS KE, NICHTHAMMER D. Interferon neutralizing antibodies in a patient treated with human fibroblast interferon. Nature 1981; 289:496-497. (15) GUTTERMAN JU, BLUMENSCHEIN GR, ALEXANIAN R, et al. Leukocyte interferon-induced tumor regression in human metastatic breast i cancer, multiple myeloma, and malignant lymphoma. Ann Intern Med 1980; 93:399-406. (16) VERMA DS, SPITZER G, GUTTERMAN JU, ZANDER AR, McCREDIE KB, DickE KA. Human leukocyte interferon preparation blocks gran 1 ulopoietic differentiation. Blood 1979; 54:1423-1427. (17) GRESSER I, Tovey MG, MAURY C, CHOUROULINKOV I. Lethality of interferon preparations for newborn mice. Nature 1975; 258:76-78. (18) GRESSER I, Maury C, Tovey M. Progressive glomerulonephritis in mice treated with interferon preparations at birth. Nature 1976; 263:420-422. (19) SHERWIN S, KNOST J, Fein S, et al. A multiple dose phase I trial of recombinant leukocyte A interferon in cancer patients. JAMA 1982; 19:2461-2466. (20) LEAVITT RD, DUFFEY PL, WIERNIK PH. A phase I trial of twice daily recombinant human leukocyte A interferon (IFL-rA) in cancer patients. Proc Am Soc Clin Oncol 1982; 1:41 (C-162). (21) TAYLOR-PAPADIMITRIOU J. Effects of interferons on cell growth and function. In: Gresser I, ed. Interferon 1980. New York: Academic Press, 1980:13-42. (22) RUBIN BY, GUPTA SL. Differential efficacies of human type I and type II interferons as antiviral and antiproliferative agents. Proc Nazi Acad Sci USA, 1980; 77:5928-5932. (23) EINHORN S, STRANDER H. Is interferon tissue specific?: Effect of human leukocyte and fibroblast interferons on the growth of lymphoblas toid and osteosarcoma cell lines. J Gen Virol 1977; 35:573-577. (24) CHADHA KC, SRIVASTAVA BI. Comparison of the antiproliferative effects of human fibroblast and leukocyte interferons on various leukemic cell lines. J Clin Hematol Oncol 1981; 11:55-60. (25) BORDEN EC, HOLLAND JF, Dao TL, GUTTERMAN JU, Wiener L CHANG YC. Leukocyte-derived interferon (alpha) in human breast carcinoma. Ann Intern Med 1982; 97:1-6. (26) GUTTERMAN JU, QUESADA JR, Fein S. Clinical investigation of par tially pure and recombinant DNA derived leukocyte interferon in human cancer. Tex Rep Biol Med 1982; 41:626-633. (27) Quesada JR, SWANSON D, TRINDADE A, GUTTERMAN J. Renal cell carcinoma: Antitumor effects of leukocyte interferon. Cancer Res 1983; 43:940-947. (28) Talpaz M, McCredie K, GUTTERMAN J. Clinical investigation of Page 26
Hans F. Stich, * Hiroshi Ohshima, Brigitte Pignatelli, Jocelyne Michelon, and Helmut Bartsch
ABSTRACT—Extracts of betel nut (Areca catechu) were tested for their capacity to inhibit the endogenous formation of nitrosamines by i measurement of the amount of urinary N-nitroso-L-proline (NPRO) following ingestion of sodium nitrate (300 mg) and L-proline (300 mg) by 2 volunteers. A water extract of the dried nuts, an ether extract containing mainly (+)-catechin and (-)-epicatechin, and a caffeinei precipitated n-butyl alcohol extract containing primarily proanthocyanidins (tannins) strongly reduced the endogenous formation of NPRO. An average of 14.7 and 10.9 ug NPRO (8 expts per individual) was excreted in the urine of the 2 volunteers over a 24-hour period following the intake of sodium nitrate and L-proline. The water extract and the proanthocyanidin (tannin)-containing extract, both of which contain the dose equivalent of one-quarter of a nut, reduced the excreted NPRO to background levels, which varied from 0.5 to 3.6 ug and from 0.6 to 2.1 ug (6 expts) in 24-hour urine samples from the 2 volunteers. These results may exemplify the way in which naturally occurring phenolics, which are ingested daily in relatively large quantities, could affect the endogenous formation of carcinogenic nitrosamines.—JNCI 1983; 70:1047-1050.
30 minutes at room temperature. The suspension was filtered through coarse cellulose filter paper, and the residual nhexane was evaporated. The dry, defatted powder was extracted in distilled water (0.1% sodium metabisulfite) and then heated in a boiling water bath for 10 minutes. This extraction procedure was repeated five times. Filtrates were pooled and spun (4,000Xg) in large centrifuge bottles for 10 minutes and freeze-dried. This extract is referred to as “aqueous betel nut extract” (extract A). Ether extract: A portion of the freeze-dried aqueous betel nut extract was dissolved in distilled water and extracted six times with ether, 100 ml each time. The combined ether extracts were rotary evaporated to dryness and further dried under high vacuum. The residue was taken up in 20 ml water and freeze-dried. Catechins were determined by highpressure liquid chromatography with the use of commercial catechin standard. This extract is referred to as “catechin extract” or “fraction" (extract C). It contained (+)-catechin, (-)-epicatechin, and monomeric (+)-leucocyanidin. n-Butyl alcohol extraction and caffeine precipitation of proanthocyanidins: The aqueous phase from the previous extraction was rotary evaporated to remove residual ether, followed by extraction with n-butyl alcohol (six times, 1 vol each time). The n-butyl alcohol extract was rotary evaporated to dryness and dried further under high vacuum, and the residue was taken up in water. To this solution, 1.5% caffeine was added to precipitate polymeric phenolics, and the suspension was spun at 4,000xg for 5 minutes. The supernatant was collected, and the pellet was resuspended in 200 ml water. Both supernatant and pellet fractions were extracted three times with 200 ml methylene chloride to remove caffeine. The methylene chloride fraction was discarded, and the water fractions were freeze-dried. The resuspended pellet contained primarily polymeric proanthocyanidins (tannins) (extract B). The catechin and tannin fractions were subjected to two
Numerous chemical studies have shown the capacity of simple phenolics, phenolic acids, cinnamic acid and related compounds, or polymer phenolics to inhibit or enhance nitrosation of nitrosable compounds (1-4). These in vitro results are difficult to extrapolate to nitrosation reactions occurring within complex mixtures or an entire organism. Information on this issue can be obtained virtually only by direct examination of humans. Measurements of urinary NPRO following ingestion of known quantities of nitrate and L-proline (5, 6) offered the opportunity for study of various dietary factors which, by modifying endogenous nitrosation, could conceivably affect cancer incidences. As an example, we examined the effect of betel nut chewing. On a worldwide basis, over 250 million individuals use betel nut Areca catechu), tobacco (Nicotiana tabacum), and betel leaves (Piper betle) in a great variety of chewing mixtures. Considerable epidemiologic evidence points to a link between these chewing habits and the development of oral, pharyngeal, and esophageal carcinomas (7), whereas the incidence of stomach cancers does not seem to be increased among the betel quid- or tobacco-chewing population groups (8). In many rural areas of southern and northeast India, more than 90% of the population chews betel quid. The objective of this study was to reveal the effect of phenolic-containing betel nut extracts on the endogenous formation of NPRO under conditions that simulate the chewing of betel quids.
ABBREVIATION USED: NPRO=N-nitroso-L-proline.
'Received October 15, 1982; accepted January 27, 1983. 2 The TEA 502 detector was provided on loan by the Division of Cancer Cause and Prevention, National Cancer Institute, under Public Health Service contract NOICP-55717. This study was supported in part by the National Cancer Institute of Canada and by the Natural Sciences and Engineering Research Council of Canada. 3 Research procedures were in accord with the ethical standards of the Bioethics Committee, University of British Columbia.
Extracts of areca nuts. —Water extract: Whole, dry areca nuts were crushed to a fine powder in a blender and extracted twice with 500 ml n-hexane by being shaken for
dimensional chromatography on cellulose plates (cellulose MN 300 G) with the use of n-butyl alcohol:acetic acid:water (4:1:2.2) in the first direction and 2% aqueous acetic acid in the second direction. The identified components have been mentioned in the preceding paragraphs and in greater detail by Stich et al. (9). Analysis of NPRO.— NPRO levels in the urine were determined according to the procedure reported by Ohshima and Bartsch (5). A 15-ml sample of urine, to which 300 ng Nnitrosopipecolic acid had been added as an internal standard, was extracted three times with 25 ml ethyl acetate in the presence of 5 g NaCl and 1.5 ml of a 20% ammonium sulfamate solution in 3.6 N H2SO4. The combined ethyl acetate extracts were dried over anhydrous Na2SO4 and concentrated to dryness by a rotary evaporator. The residue was dissolved in 3 ml diethyl ether and derivatized with diazomethane for 5 minutes. The derivatized ethereal extract was concentrated to 0.5 ml under a stream of nitrogen; a 10-ul aliquot was used to determine NPRO methyl ester in a gas liquid chromatograph equipped with a thermal energy analyzer (TEA 502; Thermo Electron Corp., Waltham, Mass.). Chromatography conditions were as follows: a glass column (2 m X 3 mm inside diameter) containing 10% Carbowax 20 M on Chromosorb W-HP (80-100 mesh); temperature of column oven, 190°C; temperature of injection port, 220°C; carrier gas, argon (20 ml/min). An authentic sample of the methyl ester of NPRO was used as the reference compound for quantitation. Experimental design.—We studied 2 healthy volunteers, 1 male and 1 female, who had no medical complaints and who had not taken any medication before or during the experimental period. The male was 54 years old, and the female was 48 years old. On the day of the experiment, the volunteers consumed no breakfast, no other foods, and no beverages before noon. Around 10:00 a.m., they drank a glass of water. They ate their regular lunch and supper at 12:30 and 6:30 p.m., respectively. The volunteers followed a rigid regimen of intake of sodium nitrate, L-proline, and betel nut extracts. At 9:00 a.m., the volunteers drank 410 mg sodium nitrate (300 mg NO3 ) dissolved in 20 ml distilled water. At 9:30 a.m., they ingested 300 mg L-proline (E. Merck A.G., Darmstadt, Federal Republic of Germany) dissolved in 20 ml distilled water. Immediately thereafter, they took the first betel nut extract dissolved in 20 ml distilled water, and then 15 and 30 minutes later they took, respectively, the second and third betel nut extracts (each dissolved in 20 ml distilled water). Urine samples were collected over a 24-hour period.
proline greatly exceeded background levels found in the urine of the 2 volunteers on the days on which no sodium nitrate and L-proline were taken (table 1). The graphs in text-figure 1 show NPRO excretion patterns in two experiments that were performed 6 weeks apart, i.e., at the onset and at the end of the experimental series. The urinary NPRO levels of the 2 volunteers reached a maximum within the first 8 hours after sodium nitrate and L-proline ingestion. This pattern is comparable to that previously described for another volunteer (5, 6). The effect of betel nut chewing on endogenous nitrosation
In the absence of adverse biological effects of NPRO, urinary excretion of NPRO following ingestion of sodium nitrate and L-proline was monitored as an index for endogenous nitrosation (5, 6). Of the 2 volunteers, 1 (subject #1) produced relatively high levels of NPRO, whereas the other (subject #2) formed a lower amount, although both were on a virtually identical diet for about 1 month before the experiment (table 1, text-fig. 1). The excretion of NPRO over 24 hours following ingestion of sodium nitrate and L
TEXT-FIGURE 1.–Urinary excretion of NPRO by 2 volunteers (subject #1,
was simulated by the ingestion of sodium nitrate, L-proline, and betel nut extracts. The following betel nut extracts were used: an aqueous extract of the entire betel nut (extract A), a fraction containing predominantly polymeric proanthocyanidins (tannins) (extract B), and a fraction consisting of (+)-catechin, (-)-epicatechin, monomeric (+)-leucocyanidin, and trace quantities of two other unidentified compounds (extract C). One betel nut yielded approximately 840 mg of a freeze-dried water extract (extract A), 84 mg tannins (extract B), and 12.8 mg catechin-epicatechin (extract C). The inhibitory effect of the betel nut components on the endogenous formation of NPRO was examined by the ingestion of sodium nitrate, L-proline, and the various betel nut extracts. The betel nut extracts were taken three times at 15-minute intervals following the consumption of L-proline.
The amounts of 210 mg for extract A, 21 mg for extract B, and 3.2 mg for extract C correspond to the amount that would be obtained if one chewed one-quarter of a betel nut. In the northeastern regions of India, one-quarter of a betel nut is eaten at approximately 15-minute intervals (9). Thus the experimental protocol simulates the actual conditions found among betel nut chewers. A relatively strong inhibitory effect on NPRO formation was observed with all three examined betel nut extracts (table 1). Even the lowest tested doses of extracts A and B, which would correspond to the dose level from a small fraction of a betel nut, greatly reduced the formation of NPRO. The total amounts of NPRO excreted over 24 hours following ingestion of sodium nitrate, L-proline, and extract A or B were at or close to the NPRO background levels. The urinary excretion pattern of NPRO over 24-hour periods is given in text-figure 2. Each graph in this text-figure represents one experiment. The results also indicate that the three examined betel nut extracts exerted a relatively strong inhibitory effect on the endogenous formation of NPRO.
Considering that more than 1 g of phenolics is ingested daily by individuals on a Western diet (10), one should not be surprised to see a great interest in the examination of the genotoxic (10-12) and carcinogenic (13, 14) properties of phenolics. Somewhat less attention has been given to this group of compounds as possible anticarcinogenic (15), antigenotoxic (16, 17), and antinitrosation (1-4) agents. The inhibitory effect of phenolics on nitrosation seems to be primarily due to their trapping effect of nitrous acid (18), whereas their stimulating action is probably due to the formation of highly reactive derivatives of C-nitroso phenolics (3). The ingestion of betel nut extracts resulted in a relatively strong inhibitory effect on the endogenous formation of NPRO. Water extracts of betel nut (extract A) and proanthocyanidin (tannin)-containing fractions (extract B), consumed at dose levels corresponding to those that would result if one chewed one-quarter of a betel nut or less every 15 minutes, virtually abolished the increase in urinary NPRO levels following the intake of sodium nitrate and Lproline. These results are in good agreement with the observation that phenolics or phenolic-containing beverages reduce the formation of mutagenic compounds when added during nitrosation of methylurea (17) and with the inhibitory effect of various phenolics on in vitro nitrosation (1-4). In this study no enhanced formation of NPRO was observed in 2 volunteers following the ingestion of betel nut extracts. This result does not exclude the possibility that, at a different ratio between nitrate, L-proline, and phenolics, a stimulating effect similar to that described for in vitro test systems (4) may actually occur. The observed inhibitory effect of the betel nut extracts could conceivably exemplify a pattern that is valid for other phenolic-containing food items. Such foods are ingested with virtually every meal containing plant products and with a wide array of plant-derived alcoholic or nonalcoholic beverages. The possibility cannot be simply excluded that the concurrent ingestion of phenolics and nitrosable compounds may, in part, be responsible
HOURS TEXT-FIGURE 2.—Urinary excretion of NPRO by 2 volunteers (subject #1,
for the extremely low levels or even lack of nitrosamines within the gastric juice (19).
mutagens and carcinogens in the etiology of oral and esophageal carcinomas of betel nut and tobacco chewers. In: Stich HF, ed Carcinogens and mutagens in the environment. Vol III: Natural occurring compounds: Epidemiology and distribution. Boca Raton, Fla.: CRC Press, 1983. In press. (10) BROWN JP. A review of the genetic effects of naturally occurring flavonoids, anthraquinones and related compounds. Mutat Res 1980; 75:243-277. (11) Nagao M, Morita N, Yahagi T, et al. Mutagenicities of 61 flavonoids and 11 related compounds. Environ Mutagen 1981; 3:401-419. (12) Stich HF, Rosin MP, Wu CH, Powrie WD. The action of transition metals on the genotoxicity of simple phenols, phenolic acids and cinnamic acids. Cancer Lett 1981; 14:251-260. (13) PAMUKCU AM, YALCINER S, HATCHER J, BRYAN GT. Quercetin, a rat intestinal and bladder carcinogen present in bracken fern (Puendum aquilinum). Cancer Res 1980; 40:3468-3472. (14) Hirono I, UENO I, Hosaka S, et al. Carcinogenicity examination of quercetin and rutin in ACI rats. Cancer Lett 1981; 13:15-21. (15) WATTENBERG LW, Coccia JB, Lam LK. Inhibitory effects of phenolic compounds on benzo(a)pyrene-induced neoplasia. Cancer Res 1980; 40:2820-2823. (16) Wood AW, Huang M-T, Chang RL, et al. Inhibition of the muta Madelyn Kleiner Feder“ and Fred Gilbert
ABSTRACT — Tumor samples, obtained from a single patient at two points in his illness, have enabled us to study clonal evolution in a neuroblastoma. Cells from the primary tumor demonstrated considerable heterogeneity in terms of chromosome number; cells from 4 subsequent metastases were all nearly diploid; and cells from a tumor produced in a mouse by the injection of cells from the primary tumor were hypotriploid in modal number. All of the tumor samples contained the same marker chromosome rearrangements, indicating their origin from a common precursor. Each of the cell lines analyzed (including those from the patient's metastases, those from the tumor in a mouse, and those from the primary tumor after 11 months in continuous culture) also contained different and distinguishing chromosome abnormalities. The differences in karyotype among these tumor samples and cell lines presumably reflect the different selection pressures at work in each instance.—JNCI 1983; 70:1051-1056.
Studies of cytogenetic and biochemical parameters point to a unicellular or clonal origin for most cancers in humans. In both hematopoietic and solid tumors, for example, it has usually been found (1, 2) that all cells from the same neoplasm either have the same abnormal karyotype or share common marker chromosomes, indicating their origin from a single precursor cell. It has also been shown (3) that, in the lymphoproliferative cancers that can be characterized by cell surface-associated or secreted immunoglobulin markers (i.e., multiple myeloma, certain leukemias, and lymphomas), all of the tumor cells from an individual patient synthesize the same molecular species of immunoglobulin. In addition, females heterozygous at the X-linked locus for glucose-6-phosphate dehydrogenase generally express only one of the two isozymes in tumor tissue, though both types can be detected in normal tissue (3, 4). As tumor cells continue to divide over time, both in vivo and in vitro, additional cytogenetic changes can occur. The changes most commonly identified in vivo are duplications of whole chromosomes or chromosome segments and nonrandom involvement of particular chromosomes in structural rearrangements (5). The altered genotype, which is the consequence of these changes in karyotype, is likely to contribute to the generation of more aggressive sublines within a tumor, a process termed “clonal evolution.” Previous studies of the progression of cytogenetic changes in individual cancers have concentrated on the hematopoietic cancers in humans and on certain solid tumors in animals (1, 5, 6). We now present a description of clonal evolution in a human neuroblastoma, comparing the chromosome constitution of the untreated primary tumor with a recurrence at the primary tumor site and 3 metastases obtained from the same patient 11 months later, at the time of death.
the right adrenal gland. At the time the tumor was excised, the patient's bone marrow was judged to be free of tumor, and there was no radiologic evidence of metastatic spread of disease. Postoperatively, treatment was begun with cyclophosphamide, vincristine, and dacarbazine. Seven months later, a follow-up bone marrow was found to be positive for neuroblastoma by visual inspection. Chemotherapy was changed to cisplatin and teniposide, followed by melphalan. A course of total-body irradiation failed to halt the deterioration in the patient's condition, and he died 11 months from the time of diagnosis. At the initial surgery, a tumor sample was obtained aseptically and put into tissue culture. The cell line that grew was designated NB-69 and has been maintained in RPMI-1640 medium supplemented with 10 or 20% fetal calf serum. At the time of death, tissue samples were obtained from metastases to lung, liver, and sternum and from an abdominal mass at the site of the original tumor in the right suprarenal area. The cell lines that subsequently grew in culture were designated NB-69(2): They were from the lung, liver, sternum, and primary recurrence. Cells from NB-69 were injected into athymic (nu/nu) mice (on a BALB/c background) obtained from Harlan SpragueDawley, Inc., Madison, Wis.; 10' cells were injected into each of two sites per mouse in 3 mice. Seventy days after inoculation, 1 of the tumors that developed was excised and explanted back into tissue culture; the resulting cell line was designated NB-69TC. Chromosome preparations were made from each of the minced tumor tissue samples immediately and/or after 72 hours in culture. Additional preparations were made from each of the tumor cell lines at intervals over the succeeding months in culture. The methods of slide preparation and chromosome banding were as previously described (7). From a 3-day culture of the primary tumor, 200 metaphases were counted for the determination of the extent of heterogeneity within the tumor. For all subsequent tumor samples and for the preparations from cultured cell lines,
Received November 1, 1982; accepted February 18, 1983. ? Supported by a grant from the American Cancer Society and by Public Health Service (PHS) grant CA-14489 from the National Cancer Society to the Children's Cancer Research Center of Children's Hospital and by PHS grant GM-20138 from the National Institute of General Medical Sciences to the University of Pennsylvania Genetics Center. Animals were maintained under the guidelines of the University of Pennsylvania School of Medicine. * Department of Human Genetics, University of Pennsylvania School of Medicine, Philadelphia, Pa. 19104. 5 Department of Pediatrics, University of Pennsylvania School of Medicine, Philadelphia, Pa. 19104. 6 Address reprint requests to Dr. Gilbert at his present address: Department of Pediatrics, Annenberg 17-76, Mt. Sinai School of Medicine, New York, N.Y. 10029.
A 16-month-old male child presented with an abdominal mass that at surgery proved to be a neuroblastoma involving |
