- Email: [email protected]
Induction of liver and lung tumours in rats by the simultaneous administration of sodium nitrite and morpholine
Fd Cosmet. ToxicoL Vol. 11, pp. 819-825. Pergamon Press 1973. Printed in Great Britain
Induction of Liver and Lung Tumours in Rats by the Simultaneous Administration of Sodium Nitrite and Morpholine P. M. NEWBERNE and R. C. SHANK Department of Nutrition and Food Science, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, USA (Received 23 May 1973)
Abstract--Long-term feeding studies were conducted in Sprague-Dawley rats using various dietary concentrations (up to 1000ppm each) of nitrite and morpholine or 5 or 50 ppm N-nitrosomorpholine. The animals fed nitrite and morpholine at moderate to high concentrations developed hepatocellular carcinomas and angiosarcomas, histologically identical to those induced by preformed N-nitrosomorpholine.The data suggestthat in vivo nitrosation of morpholine does occur in the rat, presumably in the acidic gastric environment.
INTRODUCTION The carcinogenicity of preformed nitrosamines has been documented in many animals and species (Magee & Barnes, 1967) and continues to be of interest. However, there is increasing concern about nitrosation of dietary amines by sodium nitrite, a common food additive, and the possible relationship of this reaction to cancer in man (Druckrey, Preussmann, lvankovic & Schm~hl, 1967; Food and Cosmetics Toxicology, 1968; Lancet, 1968; Lijinsky & Epstein, 1970). Nitrosamines have been found in such varied items as tobacco smoke and mushrooms (Food and Cosmetics Toxicology, 1968) and, more significantly, in staple foods such as grains, pasteurized milk and cheese (Hedler & Marquardt, 1968) and nitrite-treated cheese, fish, smoked fish and meat (Ender & (~eh, 1968; Food and Cosmetics Toxicology, 1968). Furthermore, there is the disturbing suggestion that nitrosamines may be formed from nitrite and secondary amines under the acidic conditions of the human stomach (Druckrey, Steinhoff, Beuthner, Schneider & Kl~rner, 1963; Sander, 1967). Although Sander, Schweinsberg & Menz (1968) had demonstrated the in vivo formation of nitrosamines in the stomachs of rats simultaneously fed nitrite and secondary amines of fairly low pH, the interest in this phenomenon was raised considerably when Sander & Seif (1969) showed that patients given nitrite and diphenylamine produced nitrosodiphenylamine in the stomach. These considerations lend an urgency to this area of environmental carcinogenesis paralleled only by the recently recognized mycotoxin hazard (Ciegler, Kadis & Ajl, 1971). Whether or not biologically significant amounts of nitrosamines are formed in the human stomach from nitrite and the amines normally present in food products is a matter of conjecture. However, research to date strongly suggests the need to learn more about in vivo nitrosation and its relation to carcinogenesis. Only a few studies have dealt with the long-term administration of secondary amines and nitrite to intact animals. Druckrey et aL 819
820
P. M. NEWBERNE
and
R. C. SHANK
(1963) reported negative results when sodium nitrite and diethylamine were fed to rats, and Sander et al. (1968) observed that diethylamine and N-methylaniline were also without carcinogenic effect when fed along with nitrite to rats. However, Sander & Burkle (1969) have reported the occurrence of hepatic and oesophageal tumours in rats fed N-methylbenzylamine or morpholine along with sodium nitrite. More recently, Greenblatt, Mirvish & So (1971) have presented data which suggest that an increased incidence of lung adenomas in mice resulted from in vtvo nitrosation of secondary amines by dietary nitrite. The work reported here provides further evidence that dietary nitrite, as the sodium salt, and a secondary amine (morpholine) can induce liver and lung tumours identical to those induced by N-nitrosomorpholine, a preformed nitrosamine. EXPERIMENTAL Female Sprague-Dawley rats were fed from conception an agar-gel diet, the composition of which has been reported previously (Wogan & Newberne, 1967). The diet contained either N-nitrosomorpholine or nitrite and morpholine at various concentrations including those listed in Table 1. In addition, nitrite and morpholine were each fed separately at 1000 ppm but these are not listed because the results were negative. Following parturition, litters were adjusted to eight each and left with the mother until weaning; they then continued on their respective diets. Some of the F~ generation were bred to provide Fz generations for long-term carcinogenic studies. The diets were freshly prepared each week and were stored in covered containers under refrigeration. Analytical-grade sodium nitrite (J. T. Baker Chemical Co., Phillipsburg, N.J.) and morpholine (purified, Fisher Scientific Co., Fairlawn, N.J.) were mixed into the diet to provide the various desired concentrations. N-Nitrosomorpholine (Schuchardt GmbH, Munich, Federal Republic of Germany) was fed as a positive control at a dietary level of 5 or 50 ppm (dry weight). All animals were observed at least twice daily and were weighed weekly, and complete autopsies were performed at death or when the rats were moribund. Routine histopathology studies were conducted on all animals. RESULTS Significant differences in tumour incidence were not observed between sexes or between generations at the highest concentrations of nitrite and morpholine; for this reason, results from both sexes and both generations were combined for each of the groups fed the high levels of nitrite and morpholine and for those fed the two levels of N-nitrosomorpholine, and these combined data are listed as single treatment groups in Table 1. Of the 159 rats fed a diet containing 1000 ppm sodium nitrite and I000 ppm morpholine, almost 100~o (156/159) developed hepatocellular carcinoma and about 2 5 ~ of these had coexisting hepatic angiosarcomas, which appeared to derive from sinusoidal endothelium or from Kupffer cells. More than 68 ~o of the liver-cell tumours metastasized to the lung, where primary angiosarcomas were also present in about 25 ~o of the cases. When the concentration of morpholine was decreased to 50 ppm, there was a concomitant drop in the incidence of all tumours and there were fewer metastases of liver tumours to the lung. When morpholine was decreased to 5 ppm, only 3/160 animals developed liver-cell carcinoma even though the nitrite intake remained high (1000 ppm). None of these animals had detectable metastatic lesions, nor were angiosarcomas observed in any of them.
TUMOUR INDUCTION BY NITRITE AND MORPHOLINE
E
¢.0
~o
.='~
o~
E
o
~0
Z
b
@
:II
E
-6
o
o ~ ~_~oo
@
.-=
@
z Z i
Z t
821
822
P. M. NEWBERNE a n d R. C. SHANK
FIG. 1. Liver tumours with peritoneal visceral extension. The closed arrow indicates a nodule subsequently confirmed as a hepatocellular carcinoma. The open arrow indicates a tumour which histological examination revealed as a haematoma with a surrounding angiosarcomatous component. FIG. 2. Multicentric liver turnout typical of those induced by monocrotaline, aflatoxin and acetylaminofluorene and now observed both with N-nitrosomorpholine and with nitrite and morpholine. This type of tumour was more common with the lower level of N-nitrosomorpholine or when either nitrite or morpholine was decreased from the 1000 ppm concentration. FIG. 3. Section taken from the nodule indicated by the closed arrow in Fig. 1. The histological arrangement is typical of the trabecular-type liver-cell carcinoma. Haematoxylin and eosin × 190. FI6. 4. Section taken from an anaplastic nodule in the mesenteric extension of Fig. 1. The rapidly growing area is characteristic of a large proportion of nitrite-morpholine induced liver-cell tumours. Haematoxylin and eosin × 410.
TUMOUR INDUCTION BY NITRITE AND MORPHOLINE
823
FIG. 8. Gross appearance of metastatic lesions of liver tumours to lung ( m o w e d ) . The large number of metastatic lesions, numbering up to 27/lobe, indicated a fast-growing, early metastasizing liver tumour. FIG. 9. A characteristic liver-cell tumour, which had metastasized to the lung. The cells are typical of the hepatocellular carcinoma resident in the rat from which this lung section was taken. Haematoxylin and eosin x 190. FIG. 10. Left portion ( m o w e d ) indicates a section of an angiosarcoma in the lung of a rat with a coexisting liver-cell carcinoma. The lesion on the right (arrowed) indicates a carcinoma metastatic from the liver. Haematoxylin and eosin x 32. FIG. 11. High-power view of the section taken from the lesion on the left of Fig. 10. The cells are pleomorphic, bizarre and hyperchromatic, typical of angiosarcomas observed in these studies. Haematoxylin and eosin x 380. FIG. 12. High-power view of the section taken from lesion on the right of Fig. 10, showing a typical anaplastic hepatoceilular carcinoma with numerous mitotic figures. Haematoxylin and eosin × 425.
F.C.T.
11/5----!1
824
P. M. NEWBERNE
and R. C. SHANK
Decreasing the nitrite concentration to 50 ppm and maintaining the morpholine level at I000 ppm resulted in liver-cell carcinomas in 4/120 animals. Studies on both the group given the high level of N-nitrosomorpholine (50 ppm) and that given the high level of nitrite and morpholine (1000 ppm of each) are completed, so the statistics on these groups are firm. Results of the other concentrations may be modified slightly when all the animals have been accounted for, but the trends are clear and the high incidence and rapid induction of tumours with nitrite and morpholine imply a more complex system than is indicated by kinetic studies. The tumours induced by N-nitrosomorpholine and by nitrite and morpholine were morphologically identical. Grossly, most livers presented massive, multicentric nodular tumours with a remarkable vascular component, often with extensive haemorrhage or large haematomas (Fig. !). There was extensive destruction of the liver by greyish-white nodules and dark blood-filled cavities; rapid extension throughout the abdominal cavity was usual. In all groups, however, some rats had smaller liver tumours (Fig. 2) resembling those induced by other hepatocarcinogens, including aflatoxin, acetylaminofluorene and monocrotaline. The rapidly growing soft tumorous nodules, away from frankly necrotic or haemorrhagic areas, were histologically either trabecular (Fig. 3) or anaplastic (Fig. 4), with the latter predominating. Highly vascular areas consisting of blood-filled spaces (Fig. 5) lined either by neoplastic liver cells or by sarcomatous cells derived from the endothelium (Fig. 6) were present in many of the tumours. In some cases, rupture of blood-filled spaces into the peritoneal cavity resulted in death from haemorrhage. In many tumour-bearing rats, somewhat firm, solid angiosarcomas (Fig. 7) coexisted with liver cell carcinomas and both types of tumours metastasized to the lung (Figs 8-12). Since primary angiosarcomas also arose in the lungs of some animals, it was not always possible to determine whether the sarcomas were primary to the lung or were metastatic from the liver. Generally, the enormous size of the liver and the rapidity and marked degree of peritoneal extension and lung metastasis exceeded that observed with any of several standard liver carcinogens studied in our laboratories. The rapid extension and metastasis may have been related to the vascularity of the tumours, but this is speculative.
DISCUSSION These data support the concept that in vivo nitrosation does occur, presumably under the acidic conditions in the rat's stomach. It is quite likely, however, that nitrosation occurs in some other biological compartment, or that some other modification takes place which results in chemical conversion favourable to carcinogenesis. The original concepts of Sander (1967) and Lijinsky & Epstein (1970) are extremely provocative, but we must bear in mind that nitrates and nitrites, as well as amines, have been abundant in our environment for many years without an appreciable increase in chemically related cancer of the types produced in experimental animals. We must therefore proceed with caution in attempting to extrapolate to man the results obtained in animals.
Acknowledgements~This work was supported in part
by Contract No. FDA 71-81 USPHS, Food and Drug Administration, Department of HEW. The opinions expressed in this paper are those of the authors and do not necessarily represent those of the FDA.
TUMOUR INDUCTION BY NITRITE AND MORPHOLINE
825
REFERENCES Ciegler, A., Kadis, S. & Ajl, S. (1971). Microbial Toxins. A Comprehensive Treatise. Vol. VI. Fungal Toxins. Academic Press, New York. Druckrey, H., Preussmann, R., Ivankovic, S. u. Schm~ihl, D. (1967). Organotrope carcinogene Wirkungen bei 65 verschiedenen N-Nitroso-Verbindungen an BD-Ratten. Z. Krebsforsch. 69, 103. Druckrey, H., Steinhoff, D., Bcuthner, H., Schneider, H. u. Klarner, P. (1963). Priifung von Nitrit auf chronisch toxische Wirkung an Ratten. Arzneimittel-Forsch. 13, 320. Ender, F. & (?.eh, L. (1968). Occurrence of nitrosamines in foodstuffs for human and animal consumption. Fd Cosmet. Toxicol. 6, 569. Food and Cosmetics Toxicology (1968). Nitrosamines: A jig-saw puzzle with missing pieces, ibm 6, 647. Greenblatt, M., Mirvish, S. & So, B. T. (1971). Nitrosamine studies: Induction of lung adenomas by concurrent administration of sodium nitrite and secondary amines in Swiss mice. J. natn. Cancer Inst. 46, 1029. Hedler, Liselotte & Marquardt, P. (1968). Occurrence of diethylnitrosamine in some samples of food. Fd Cosmet. Toxicol. 6, 341. Lancet (1968). Nitrites, nitrosamines, and cancer, ibid i, 1071. Lijinsky, W. & Epstein, S. S. (1970). Nitrosamines as environmental carcinogens. Nature, Lond. 225, 21. Magee, P. N. & Barnes, J. M. (1967). Carcinogenic nitroso compounds. Adv. Cancer Res. 10, 163. Sander, J. (1967). Kann Nitrit in der menschlichen Nahrung Ursache einer Krebsentstehung durch Nitrosaminbildung sein? Arch. Hyg. Bakt. 151, 22. Sander, J. u. Btirkle, G. (1969). Induktion maligner Tumoren bei Ratten durch gleichzeitige Verfiitterung yon Nitrit und sekundaren Aminen. Z. Krebsforsch. 73, 54. Sander, J., Schweinsberg, F. u. Menz, H.-P. (1968). Untersuchungen fiber die Entstehung cancerogener Nitrosamine im Magen. Hoppe-Seyler's Z. physiol. Chem. 349, 1691. Sander, J. u. Self, F. (1969). Bakterielle Reduktion yon Nitrat im Magen des Menschen als Ursache einer Nitrosamin-Bildung. Arzneimittel-Forsch. 19, 1091. Wogan, G. N. & Newberne, P. M. (1967). Dose-response characteristics of aflatoxin BI carcinogenesis in the rat. Cancer Res. 27, 2370.
Induction de tumeurs h~patiques et pulmonaires chez le rat par l'administration simultan~e de nitrite de sodium et de morpholine R~um6--Dans des essais nutritionnels de Iongue dur6e, on a fait prendre b, des rats SpragueDawley soit diff6rentes doses (jusqu'h 1000 ppm du r6girne) de nitrite et de morpholine, soit 5 ou 50 ppm de N-nitrosomorpholine. Les animaux qui recevaient le nitrite et la morpholine ~t des doses mod6r~s ou 61ev6es ont contract6 des carcinomes et des angiosarcomes h6patocellulaires histologiquement identiques b, ceux provoqu~s par la N-nitrosomorpholine pr6form6e. Les donn6es sugg~rent que la nitrosation in vioo de la morpholine chez le rat se produit probablement dans le milieu acide de l'estomac.
Induktion von Leber- und Lungentumoren bei Ratten durch die gleichzeitige Anwendung von Natrinmnitrit und Morpholin Zusammenfassung--Langzeitverfiitterungsversuche wurden mit Sprague-Dawley-gatten unter Anwendung verschiedener Konzentrationen (bis 1000 ppm jeder Verbindung) von Nitrit und Morpholin oder 5 oder 50 ppm N-Nitrosomorpholin durchgefiihrt. Die Tiere, welche Nitrit und Morpholin in m~sigen bis hohen Konzentrationen im Futter erhielten, entwickelten hepatocellul~e Carcinome und Angiosarcome, histologisch identisch mit den vou ausserhalb des K6rpers gebildetem N-Nitrosomorpholin induzierten. Die Daten lassen annehmen, dass tn-vivo-Nitrosation yon Morpholin in der Ratte vorkommt, vermutlich innerhalb des sauren Mageninhalts.
Induction of Liver and Lung Tumours in Rats by the Simultaneous Administration of Sodium Nitrite and Morpholine P. M. NEWBERNE and R. C. SHANK Department of Nutrition and Food Science, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, USA (Received 23 May 1973)
Abstract--Long-term feeding studies were conducted in Sprague-Dawley rats using various dietary concentrations (up to 1000ppm each) of nitrite and morpholine or 5 or 50 ppm N-nitrosomorpholine. The animals fed nitrite and morpholine at moderate to high concentrations developed hepatocellular carcinomas and angiosarcomas, histologically identical to those induced by preformed N-nitrosomorpholine.The data suggestthat in vivo nitrosation of morpholine does occur in the rat, presumably in the acidic gastric environment.
INTRODUCTION The carcinogenicity of preformed nitrosamines has been documented in many animals and species (Magee & Barnes, 1967) and continues to be of interest. However, there is increasing concern about nitrosation of dietary amines by sodium nitrite, a common food additive, and the possible relationship of this reaction to cancer in man (Druckrey, Preussmann, lvankovic & Schm~hl, 1967; Food and Cosmetics Toxicology, 1968; Lancet, 1968; Lijinsky & Epstein, 1970). Nitrosamines have been found in such varied items as tobacco smoke and mushrooms (Food and Cosmetics Toxicology, 1968) and, more significantly, in staple foods such as grains, pasteurized milk and cheese (Hedler & Marquardt, 1968) and nitrite-treated cheese, fish, smoked fish and meat (Ender & (~eh, 1968; Food and Cosmetics Toxicology, 1968). Furthermore, there is the disturbing suggestion that nitrosamines may be formed from nitrite and secondary amines under the acidic conditions of the human stomach (Druckrey, Steinhoff, Beuthner, Schneider & Kl~rner, 1963; Sander, 1967). Although Sander, Schweinsberg & Menz (1968) had demonstrated the in vivo formation of nitrosamines in the stomachs of rats simultaneously fed nitrite and secondary amines of fairly low pH, the interest in this phenomenon was raised considerably when Sander & Seif (1969) showed that patients given nitrite and diphenylamine produced nitrosodiphenylamine in the stomach. These considerations lend an urgency to this area of environmental carcinogenesis paralleled only by the recently recognized mycotoxin hazard (Ciegler, Kadis & Ajl, 1971). Whether or not biologically significant amounts of nitrosamines are formed in the human stomach from nitrite and the amines normally present in food products is a matter of conjecture. However, research to date strongly suggests the need to learn more about in vivo nitrosation and its relation to carcinogenesis. Only a few studies have dealt with the long-term administration of secondary amines and nitrite to intact animals. Druckrey et aL 819
820
P. M. NEWBERNE
and
R. C. SHANK
(1963) reported negative results when sodium nitrite and diethylamine were fed to rats, and Sander et al. (1968) observed that diethylamine and N-methylaniline were also without carcinogenic effect when fed along with nitrite to rats. However, Sander & Burkle (1969) have reported the occurrence of hepatic and oesophageal tumours in rats fed N-methylbenzylamine or morpholine along with sodium nitrite. More recently, Greenblatt, Mirvish & So (1971) have presented data which suggest that an increased incidence of lung adenomas in mice resulted from in vtvo nitrosation of secondary amines by dietary nitrite. The work reported here provides further evidence that dietary nitrite, as the sodium salt, and a secondary amine (morpholine) can induce liver and lung tumours identical to those induced by N-nitrosomorpholine, a preformed nitrosamine. EXPERIMENTAL Female Sprague-Dawley rats were fed from conception an agar-gel diet, the composition of which has been reported previously (Wogan & Newberne, 1967). The diet contained either N-nitrosomorpholine or nitrite and morpholine at various concentrations including those listed in Table 1. In addition, nitrite and morpholine were each fed separately at 1000 ppm but these are not listed because the results were negative. Following parturition, litters were adjusted to eight each and left with the mother until weaning; they then continued on their respective diets. Some of the F~ generation were bred to provide Fz generations for long-term carcinogenic studies. The diets were freshly prepared each week and were stored in covered containers under refrigeration. Analytical-grade sodium nitrite (J. T. Baker Chemical Co., Phillipsburg, N.J.) and morpholine (purified, Fisher Scientific Co., Fairlawn, N.J.) were mixed into the diet to provide the various desired concentrations. N-Nitrosomorpholine (Schuchardt GmbH, Munich, Federal Republic of Germany) was fed as a positive control at a dietary level of 5 or 50 ppm (dry weight). All animals were observed at least twice daily and were weighed weekly, and complete autopsies were performed at death or when the rats were moribund. Routine histopathology studies were conducted on all animals. RESULTS Significant differences in tumour incidence were not observed between sexes or between generations at the highest concentrations of nitrite and morpholine; for this reason, results from both sexes and both generations were combined for each of the groups fed the high levels of nitrite and morpholine and for those fed the two levels of N-nitrosomorpholine, and these combined data are listed as single treatment groups in Table 1. Of the 159 rats fed a diet containing 1000 ppm sodium nitrite and I000 ppm morpholine, almost 100~o (156/159) developed hepatocellular carcinoma and about 2 5 ~ of these had coexisting hepatic angiosarcomas, which appeared to derive from sinusoidal endothelium or from Kupffer cells. More than 68 ~o of the liver-cell tumours metastasized to the lung, where primary angiosarcomas were also present in about 25 ~o of the cases. When the concentration of morpholine was decreased to 50 ppm, there was a concomitant drop in the incidence of all tumours and there were fewer metastases of liver tumours to the lung. When morpholine was decreased to 5 ppm, only 3/160 animals developed liver-cell carcinoma even though the nitrite intake remained high (1000 ppm). None of these animals had detectable metastatic lesions, nor were angiosarcomas observed in any of them.
TUMOUR INDUCTION BY NITRITE AND MORPHOLINE
E
¢.0
~o
.='~
o~
E
o
~0
Z
b
@
:II
E
-6
o
o ~ ~_~oo
@
.-=
@
z Z i
Z t
821
822
P. M. NEWBERNE a n d R. C. SHANK
FIG. 1. Liver tumours with peritoneal visceral extension. The closed arrow indicates a nodule subsequently confirmed as a hepatocellular carcinoma. The open arrow indicates a tumour which histological examination revealed as a haematoma with a surrounding angiosarcomatous component. FIG. 2. Multicentric liver turnout typical of those induced by monocrotaline, aflatoxin and acetylaminofluorene and now observed both with N-nitrosomorpholine and with nitrite and morpholine. This type of tumour was more common with the lower level of N-nitrosomorpholine or when either nitrite or morpholine was decreased from the 1000 ppm concentration. FIG. 3. Section taken from the nodule indicated by the closed arrow in Fig. 1. The histological arrangement is typical of the trabecular-type liver-cell carcinoma. Haematoxylin and eosin × 190. FI6. 4. Section taken from an anaplastic nodule in the mesenteric extension of Fig. 1. The rapidly growing area is characteristic of a large proportion of nitrite-morpholine induced liver-cell tumours. Haematoxylin and eosin × 410.
TUMOUR INDUCTION BY NITRITE AND MORPHOLINE
823
FIG. 8. Gross appearance of metastatic lesions of liver tumours to lung ( m o w e d ) . The large number of metastatic lesions, numbering up to 27/lobe, indicated a fast-growing, early metastasizing liver tumour. FIG. 9. A characteristic liver-cell tumour, which had metastasized to the lung. The cells are typical of the hepatocellular carcinoma resident in the rat from which this lung section was taken. Haematoxylin and eosin x 190. FIG. 10. Left portion ( m o w e d ) indicates a section of an angiosarcoma in the lung of a rat with a coexisting liver-cell carcinoma. The lesion on the right (arrowed) indicates a carcinoma metastatic from the liver. Haematoxylin and eosin x 32. FIG. 11. High-power view of the section taken from the lesion on the left of Fig. 10. The cells are pleomorphic, bizarre and hyperchromatic, typical of angiosarcomas observed in these studies. Haematoxylin and eosin x 380. FIG. 12. High-power view of the section taken from lesion on the right of Fig. 10, showing a typical anaplastic hepatoceilular carcinoma with numerous mitotic figures. Haematoxylin and eosin × 425.
F.C.T.
11/5----!1
824
P. M. NEWBERNE
and R. C. SHANK
Decreasing the nitrite concentration to 50 ppm and maintaining the morpholine level at I000 ppm resulted in liver-cell carcinomas in 4/120 animals. Studies on both the group given the high level of N-nitrosomorpholine (50 ppm) and that given the high level of nitrite and morpholine (1000 ppm of each) are completed, so the statistics on these groups are firm. Results of the other concentrations may be modified slightly when all the animals have been accounted for, but the trends are clear and the high incidence and rapid induction of tumours with nitrite and morpholine imply a more complex system than is indicated by kinetic studies. The tumours induced by N-nitrosomorpholine and by nitrite and morpholine were morphologically identical. Grossly, most livers presented massive, multicentric nodular tumours with a remarkable vascular component, often with extensive haemorrhage or large haematomas (Fig. !). There was extensive destruction of the liver by greyish-white nodules and dark blood-filled cavities; rapid extension throughout the abdominal cavity was usual. In all groups, however, some rats had smaller liver tumours (Fig. 2) resembling those induced by other hepatocarcinogens, including aflatoxin, acetylaminofluorene and monocrotaline. The rapidly growing soft tumorous nodules, away from frankly necrotic or haemorrhagic areas, were histologically either trabecular (Fig. 3) or anaplastic (Fig. 4), with the latter predominating. Highly vascular areas consisting of blood-filled spaces (Fig. 5) lined either by neoplastic liver cells or by sarcomatous cells derived from the endothelium (Fig. 6) were present in many of the tumours. In some cases, rupture of blood-filled spaces into the peritoneal cavity resulted in death from haemorrhage. In many tumour-bearing rats, somewhat firm, solid angiosarcomas (Fig. 7) coexisted with liver cell carcinomas and both types of tumours metastasized to the lung (Figs 8-12). Since primary angiosarcomas also arose in the lungs of some animals, it was not always possible to determine whether the sarcomas were primary to the lung or were metastatic from the liver. Generally, the enormous size of the liver and the rapidity and marked degree of peritoneal extension and lung metastasis exceeded that observed with any of several standard liver carcinogens studied in our laboratories. The rapid extension and metastasis may have been related to the vascularity of the tumours, but this is speculative.
DISCUSSION These data support the concept that in vivo nitrosation does occur, presumably under the acidic conditions in the rat's stomach. It is quite likely, however, that nitrosation occurs in some other biological compartment, or that some other modification takes place which results in chemical conversion favourable to carcinogenesis. The original concepts of Sander (1967) and Lijinsky & Epstein (1970) are extremely provocative, but we must bear in mind that nitrates and nitrites, as well as amines, have been abundant in our environment for many years without an appreciable increase in chemically related cancer of the types produced in experimental animals. We must therefore proceed with caution in attempting to extrapolate to man the results obtained in animals.
Acknowledgements~This work was supported in part
by Contract No. FDA 71-81 USPHS, Food and Drug Administration, Department of HEW. The opinions expressed in this paper are those of the authors and do not necessarily represent those of the FDA.
TUMOUR INDUCTION BY NITRITE AND MORPHOLINE
825
REFERENCES Ciegler, A., Kadis, S. & Ajl, S. (1971). Microbial Toxins. A Comprehensive Treatise. Vol. VI. Fungal Toxins. Academic Press, New York. Druckrey, H., Preussmann, R., Ivankovic, S. u. Schm~ihl, D. (1967). Organotrope carcinogene Wirkungen bei 65 verschiedenen N-Nitroso-Verbindungen an BD-Ratten. Z. Krebsforsch. 69, 103. Druckrey, H., Steinhoff, D., Bcuthner, H., Schneider, H. u. Klarner, P. (1963). Priifung von Nitrit auf chronisch toxische Wirkung an Ratten. Arzneimittel-Forsch. 13, 320. Ender, F. & (?.eh, L. (1968). Occurrence of nitrosamines in foodstuffs for human and animal consumption. Fd Cosmet. Toxicol. 6, 569. Food and Cosmetics Toxicology (1968). Nitrosamines: A jig-saw puzzle with missing pieces, ibm 6, 647. Greenblatt, M., Mirvish, S. & So, B. T. (1971). Nitrosamine studies: Induction of lung adenomas by concurrent administration of sodium nitrite and secondary amines in Swiss mice. J. natn. Cancer Inst. 46, 1029. Hedler, Liselotte & Marquardt, P. (1968). Occurrence of diethylnitrosamine in some samples of food. Fd Cosmet. Toxicol. 6, 341. Lancet (1968). Nitrites, nitrosamines, and cancer, ibid i, 1071. Lijinsky, W. & Epstein, S. S. (1970). Nitrosamines as environmental carcinogens. Nature, Lond. 225, 21. Magee, P. N. & Barnes, J. M. (1967). Carcinogenic nitroso compounds. Adv. Cancer Res. 10, 163. Sander, J. (1967). Kann Nitrit in der menschlichen Nahrung Ursache einer Krebsentstehung durch Nitrosaminbildung sein? Arch. Hyg. Bakt. 151, 22. Sander, J. u. Btirkle, G. (1969). Induktion maligner Tumoren bei Ratten durch gleichzeitige Verfiitterung yon Nitrit und sekundaren Aminen. Z. Krebsforsch. 73, 54. Sander, J., Schweinsberg, F. u. Menz, H.-P. (1968). Untersuchungen fiber die Entstehung cancerogener Nitrosamine im Magen. Hoppe-Seyler's Z. physiol. Chem. 349, 1691. Sander, J. u. Self, F. (1969). Bakterielle Reduktion yon Nitrat im Magen des Menschen als Ursache einer Nitrosamin-Bildung. Arzneimittel-Forsch. 19, 1091. Wogan, G. N. & Newberne, P. M. (1967). Dose-response characteristics of aflatoxin BI carcinogenesis in the rat. Cancer Res. 27, 2370.
Induction de tumeurs h~patiques et pulmonaires chez le rat par l'administration simultan~e de nitrite de sodium et de morpholine R~um6--Dans des essais nutritionnels de Iongue dur6e, on a fait prendre b, des rats SpragueDawley soit diff6rentes doses (jusqu'h 1000 ppm du r6girne) de nitrite et de morpholine, soit 5 ou 50 ppm de N-nitrosomorpholine. Les animaux qui recevaient le nitrite et la morpholine ~t des doses mod6r~s ou 61ev6es ont contract6 des carcinomes et des angiosarcomes h6patocellulaires histologiquement identiques b, ceux provoqu~s par la N-nitrosomorpholine pr6form6e. Les donn6es sugg~rent que la nitrosation in vioo de la morpholine chez le rat se produit probablement dans le milieu acide de l'estomac.
Induktion von Leber- und Lungentumoren bei Ratten durch die gleichzeitige Anwendung von Natrinmnitrit und Morpholin Zusammenfassung--Langzeitverfiitterungsversuche wurden mit Sprague-Dawley-gatten unter Anwendung verschiedener Konzentrationen (bis 1000 ppm jeder Verbindung) von Nitrit und Morpholin oder 5 oder 50 ppm N-Nitrosomorpholin durchgefiihrt. Die Tiere, welche Nitrit und Morpholin in m~sigen bis hohen Konzentrationen im Futter erhielten, entwickelten hepatocellul~e Carcinome und Angiosarcome, histologisch identisch mit den vou ausserhalb des K6rpers gebildetem N-Nitrosomorpholin induzierten. Die Daten lassen annehmen, dass tn-vivo-Nitrosation yon Morpholin in der Ratte vorkommt, vermutlich innerhalb des sauren Mageninhalts.