- Email: [email protected]
The effect of age and exposure duration on cancer induction by a known carcinogen in rats, mice, and hamsters
TOXICOLOGY
AND
APPLIED
PHARMACOLOGY
68, 120- 130 ( 1983)
The Effect of Age and Exposure Duration on Cancer Induction Known Carcinogen in Rats, Mice, and Hamsters
by a
ROBERT T.DREw,"~ GARY A.BOORMAN,JOSEPH KHASEMAN, ERNEST E. MCCONNELL, WILLIAM M.BusEY,~ AND JOHN A. MOORE National
Institute of Environmental Health Sciences, P.O. Box 12233, Research Triangle Park, North Carolina 27709
Received
September
20, 1982; accepted
December
7, 1982
The Effect of Age and Exposure Duration on Cancer Induction by a Known Carcinogen in Rats, Mice, and Hamsters. DREW. R. T., BOORMAN, G. A., HASEMAN, J. K.. MCCONNELL, E. E., BUSEY, W. M., AND MOORE, J. A. (1983). Toxicol. Appl. Pharmacol. 68, 120-I 30. Female Golden Syrian hamsters, F-344 rats, Swiss CD-l mice, and B6C3Fl hybrid mice were exposed 6 hr/day, 5 days/week to carcinogenic levels of vinyl chloride (VC) for 6, 12. 18, or 24 months (rats and hamsters only). Other groups of rodents were held for 6 or 12 months and then exposed for 6 or 12 months. At the end of the study the incidence of VC-induced neoplasms was compared in each of the groups to assessthe effectsof duration of exposure and age at the start of exposure on carcinogenicity of VC. In rats, with early initial exposure, hemangiosarcomas, hepatocellular carcinomas. and mammary gland carcinomas occurred with increasing incidence with longer exposure duration. Rats held for 6 months before exposure developed V&elated neoplasms, while rats held 12 months before the start of exposure failed to show a significantly increased incidence of these neoplasms. In hamsters, hemangiosarcomas, mammary gland carcinomas, gastric adenocarcinomas, and skin carcinomas resulted from VC exposure. The highest incidence of malignant neoplasms occurred in hamsters exposed for the first 12 months, whereas exposure begun after 12 months of age did not cause neoplasms. In both strains of mice, VC exposure during the first 6 months of the experiment induced a high incidence of hemangiosarcomas and mammary gland carcinomas. Swiss mice also developed lung carcinomas after only 6 months of exposure. In all three rodent species an initial 12 month exposure to VC was adequate to detect its carcinogenic potential, but the shortened survival of VC exposed mice and hamsters precluded a meaningful comparison with longer periods of exposure. Exposures were most effective when started early in life.
The National Cancer Institute/National Toxicology Program has tested over 200 chemicals in rodent lifetime bioassays for potential toxicity and carcinogenicity (Ward et al., 1979). In the standard bioassay, animals are exposed to the chemical by an appropriate route (water, feed, gavage, etc.) for 24 months. Inhalation exposure is chosen for many eni Current address: Medical Department, Brookhaven National Laboratory, Upton, New York 11973. 2 Person to whom reprint requests should be addressed. 3 Current address: Experimental Pathology Laboratories, P.O. Box 474, Hemdon, Va. 22070. 0041-008X/83 Copynghl All
rights
0 of
1983 reproduction
$3.00 by
120
Academic in
vironmental chemicals since it most closely mimics the situation in man. However, inhalation exposure is significantly more expensive than other methods, and the number of chemicals that can be tested is limited by the number of available chambers. Most carcinogenesis studies call for lifetime exposures. However, there is a time at which the development of the neoplasm has progressed to an irreversible point. At this point, further exposures to the carcinogen are unnecessary and, if the chemical exerts other toxicity as well, further exposures may even
any
Press.
Inc.
form
reserved.
AGE AND EXPOSURE
DURATION
reduce the incidence of neoplasia because of shortened life span. If it could be shown that exposures of 6 or 12 months’ duration were sufficient to demonstrate carcinogenicity, more chemicals could be studied in a given period of time. The purpose of this study was twofold: to compare the neoplasm incidence resulting from exposures over various portions of an animal’s lifetime and to determine the optimal period of an animal’s life span over which to conduct the exposures. Specifically, would a 6- or 12-month exposure be as effective as the current 24-month exposure, and if so, when should the exposures be performed? Vinyl chloride (VC) was the model compound since its carcinogenic properties have been extensively studied (Maltoni, 1977; Maltoni and Lefemine, 1974; Maltoni et al., 198 1; Keplinger et al., 1975). Because of the large numbers of animals required for the study and since the main purpose was to compare carcinogenic incidences across groups, only female rodents were used. METHODS Animals. Eight hundred female Fischer-344 rats, 800 female Golden Syrian hamsters, 600 female hybrid B6C3Fl (female C57BL/6N X male C3H/HeN), and 600 female CD-I Swiss mice were purchased at 5 to 6 weeks of age (Charles River, Portage, Mich.). Each group was received at one time, ear tagped, and weighed weekly for 3 weeks. Animals weighing within 1 SD of the mean weight were included for long-term studies. Rats and hamsters were randomly assigned to 1 of 10 exposure groups or three control groups while mice were randomly separated into eight exposure groups and two control groups. The actual number of animals in each group and the duration of exposure of each group are shown in Fig. 1. Control groups and groups exposed during the first 6 months of their lifetimes contained larger numbers of animals allowing for sequential termination at various times after the exposure. Animals killed sequentially to follow the progression of lesions and their controls are not included in this report. Included in this report are results from 2089 rodents allowed to complete their life span and then necropsied when found moribund or dead. Animals were housed in stainless-steel hanging wire cages (16 X 11 X 7 in.). Each cage was subdivided into four compartments, each housing one rat. two hamsters,
121
VERSUS CANCER INDUCTION VINYL
CHLORIDE
EXPOSURE
REGIMEN
Months 0
6
-
Expose
12
N
----Hold 0 Rots
I6
N and
Hamsters
= 56 = 54
24
Rots,Hamsters Mice
only
FIG. 1. Protocol design. The solid lines indicate exposure periods and the dotted lines are holding periods.
or three mice. The exposure units were returned to housing racks at the end of each exposure period. Both the chambers and housing racks had automatic watering systems. An open formula pelleted rodent diet (NIH-31) (Zeigler Brothers, Gardiner, Pa.) was fed ad libitum to the animals except during the exposure period. Water was available throughout the study. Animals were checked twice daily to assesstheir clinical condition. During preexposure and postexposure holding periods, animals were held in similar cages. These animals and controls had food withheld during VC exposure periods. Vinyl chloride exposures. Exposures were in 1.3-m’ chambers similar to those described by Hinners et al. (1968). Each chamber held 36 exposure cages in six tiers, thereby providing space for 144 rats, 288 hamsters, or 432 mice per chamber. Two chambers were used to expose rats while one chamber was used for mice and one for hamsters. Animals were rotated into the chambers such that during the study no more than five groups were being exposed at any time (Fig. 1). The temperature in the chambers was 21 + 2’C with a relative humidity of 55 f 10%. The location of the animals in the chamber was rotated daily to minimize exposure bias. Chambers had an airflow of 300 liters/min and a slight negative pressure (0.5 in. HZO). Commercial cylinders of liquid VC were the source of VC vapor which was regulated to 20 psi and routed to the chambers. Chamber concentrations were monitored six times a day with a Miran IA IR analyzer. The chamber exhaust was scrubbed through a dual-activated charcoal system (Calgon Corp.), and the exhaust was monitored on the same schedule as above by a second Miran analyzer to assure complete removal of the VC. For personnel protection, a Miran analyzer was used to continuously monitor the room air in the chamber rooms. Concentrations. At the time this study was designed, the most recent published reports (Maltoni and Lefemine, 1974; Keplinger et al., 1975) indicated that VC concentrations as low as 50 ppm could produce tumors. Since Keplinger et al. (1975) had published an interim report,
122
DREW ET AL.
their current results were sought. They exposed COBS rats, CD- 1 Swiss mice, and Golden Syrian hamsters to 0, 50, 200, or 2500 ppm VC for 7 hr/day. Their results (persona1 communication) indicated that exposure to 50 ppm did not shorten the life span of rats but that 200 ppm reduced the mean life from 30 to 21 months. Since many groups in the current study would receive only 6 months’ exposure, 100 ppm was selected for rats. The data also indicated that exposure to 50 ppm reduced the lifespan of CD-I mice; however, no evidence was available at that time that lower concentrations would produce tumors. Therefore, 50 ppm was chosen for mice. The studies of Keplinger et al. (1975) also showed that hamsters appeared more resistant to VC toxicity with only a slight reduction in life span after 250 ppm; therefore, a level of 200 ppm was selected for this study. Pathology. Each animal, when found moribund or dead, was subjected to a complete necropsy. In addition to macroscopic lesions noted during the postmortem examination, the following tissues were routinely collected: brain, pituitary gland, salivary gland, thyroid, nasal turbinates (three levels for hamsters and rats, two levels for mice), heart, liver, kidneys, spleen, adrenal gland, stomach, small intestine, large intestine, pancreas, sternum, and lungs. Tissue sections were prepared, stained with hematoxylin and eosin, and examined microscopically. Statistical analysis. Differences in survival between the various groups were assessedby life table analyses (Cox, 1972). For the statistical analysis of neoplastic lesions, two different methods of adjusting for intercurrent mortality were employed. Each used the classical methods for combining contingency tables developed by Mantel and Haenszel (1959). For lesions judged to be “fatal” (i.e., that either directly or indirectly caused the death of the animal), life table methodology (Cox, 1972) was employed to compare incidence rates. By this approach, the proportions of tumor-bearing animals in the VC and control groups were compared at each point in time at which an animal died with a neoplasm of interest. The denominators of these proportions were the total number of animals at risk in each group. These results were combined by Mantel-Haenszel methods to obtain an overall p value. Examples of fatal neoplasms are hemangiosarcoma, mammary gland carcinoma, and hepatocellular carcinoma. For lesions judged to be “nonfatal” (i.e., that were merely observed at autopsy in animals dying of an unrelated cause), the incidental tumor test proposed by Peto et al. (1980) was employed. With this approach, neoplasm rates in exposed and control animals were compared in successive time intervals. The denominators of these rates were the number of animals on which necropsies were performed during the time interval. The individual time interval comparisons were then combined by the previously described methods to obtain a single overall result. Examples of nonfatal neoplasms are pituitary adenoma and adrenal cortical adenoma. A detailed discussion of
the statistical analysis of fatal and nonfatal neoplasms is given by Peto et al. (1980).
RESULTS Chamber concentrations. During the 2-year study no problems were encountered with vinyl chloride generation. The mean exposure levels were within 1% of the target values. Mortality. The mean survival times for all groups of animals are shown in Table 1. In rats, only those exposed for 1 year or more, beginning at 10 weeks of age exhibited any life span shortening (Fig. 2). This effect was presumably due to cancer since no other overt VC-induced pathologic changes were noted. In all groups of hamsters where the exposure began early in life, the life span was significantly reduced (Fig. 3). These deaths in hamsters resulted from vinyl chloride toxicity with the animals dying primarily from hepatic necrosis. None of the hamsters scheduled for exposure for 24 months survived past 18 months; therefore, in the remaining tables those groups have been combined. It should be noted that the mean life span of the control hamsters was only 463 days. Hamsters scheduled for exposure from 12 to 24 months had significantly shortened life spans. However, since three of those animals survived more than 18 months, these data have not been combined with the data from 12 to 18 months. All VC-exposed B6C3Fl mice exhibited significant life span shortening (Fig. 4). This decrease was due to the development of neoplasms since liver necrosis was not a common finding. None of the scheduled 0 to 18-monthexposed mice survived more than 1 year; therefore, these data have been combined with the data from 0 to 12 months for analysis of neoplasm incidence. These mice survived about 315 days after the start of a 6-month exposure regardless of whether the exposure began early in life, after a 6-month holding period, or after holding the mice for 1 year. Exposure to 50 ppm VC early in life significantly shortened the life span of the Swiss
AGE AND EXPOSURE
DURATION
123
VERSUS CANCER INDUCTION
TABLE 1 MEANSURVIVALOFFEMALEANIMALSTHROUGHOUTTHESTUDY" Time (days) Exposure period (months)
Fischer-344 rats
Golden Syrian hamsters
0
703 f 15
463 ?Y11
780 + 21
474 f 14
O-6 O-12 O-18 O-24
682 634 575 622
k + f f
14 20’ 156 II*
390 355 342 347
* + + *
126 lob gb 9b
316+ 8”’ 301 f 5b.d 304k 5b -
340 + lo”,’ 347 f 9b.d 321 5~ 7’ -
O-6 6-12 12-18 18-24
682 703 688 708
+ rt + +
14 24 25 21
390 468 456 499
AZ12b + 16 zk 17 + 14
316+ 8’,’ 480 f 13”’ 695 + 12b,c -
340 Ik lob,’ 472 f 15 521 -+ 15 -
O-12 6-18 12-24
634 f 20’ 659 f 17 717 f 17
355 f lob 455 f 13 424 + l5b
301 f 5”d 479 + 9b.d 632 f 12”d
347 f 9b,d 443 iz 16 472 f 20
B6C3Fl mice
Swiss CD-l mice
’ Average lifetime in days from the day the first animals were exposed (2 f SE). b Different from controls (p < 0.01). c Different from all other groups exposed for 6 months (p < 0.01). d Different from all other groups exposed for 1 year (p < 0.01).
mice (Fig. 5). In spite of the reduced lifespan of the 0 to 1&month exposed mice, a number of mice survived longer than 12 months, and the data were not combined with that from the 0 to lZmonth-exposed animals. When exposures to 50 ppm VC were initiated at 8
months of age or later, there was no effect on lifespan with this strain of mice. Other signs of VC exposure were not evident. Tumor incidence in rats. Hemangiosarcomas, mammary neoplasms, and hepatocellular carcinomas were strongly associated with
100
20
4x)
460
540
600
660
720 Time (days)
180
840
900
FIG. 2. Percentage survival for female Fischer rats exposed to vinyl chloride for 0 (control, a) O-6 (b), O-12 (c), O-l 8 (d), and O-24 (e) months. Decreasing survival is seen with increasing exposure duration. Survival of rats held for any time interval before exposure was similar to controls (Table 1). For brevity, data from these studies are not shown.
DREW
124
ET
AL.
Time (days)
FIG. 3. Percentage a), O-6 (b), O-12 (c). duration. Survival of I). For brevity, these
survival for female Golden Syrian hamsters exposed to vinyl chloride for 0 (control, O-l 8 (d), and O-24 (e) months. Decreasing survival is seen with increasing exposure hamsters held for any time interval before exposure was similar to controls (Table data are not shown.
exposure to VC in rats (Table 2). The incidence of hemangiosarcomas was a function of the duration of exposure. Exposures for 6 months induced a low incidence of hemangiosarcomas only if begun early in life. Exposures of a year or more produced the neoplasm, except when started after 1 year, but the incidence was higher if exposures began early in life. Mammary gland adenocarcinomas were also evident after a year or more
exposure to VC; however, in this case the exposures had to begin prior to 8 months of age for this neoplasm to develop. The incidence of fibroadenomas of the mammary gland was increased by exposure to VC in most groups, but the increase was not always related to the duration of the exposure. The incidence of these neoplasms was higher when young rats were exposed. Hepatocellular carcinomas were induced in rats in a dose-related manner when
Time (days)
FIG. 4. Percentage survival for female B6C3Fl mice exposed to vinyl chloride for 0 (control, a), 6-12 (b), 6-18 (c), 12-18 (d), and 12-24 (e) months. Mice first treated after 6 months showed greater life shortening than those after 12 months. The survival of B6C3Fl mice treated from the beginning was essentially independent of exposure duration with similar mean survival times (Table I). For brevity these data are not plotted on this figure.
AGE AND EXPOSURE
DURATION
Time (days)
FIG. 5. Percentage survival for female CD- 1 Swiss mice to vinyl chloride for 0 (control, a), O-6 (b), O12 (c), and O-18 (d) months. All exposures were associated with a marked diminution of lifespan. In animals held for a period of time before exposure, life spans were generally not shortened (Table I). For brevity, these data are not shown. exposed
exposures began early in life. The incidence of this tumor was significantly elevated after 12 months’ exposure only if the exposures started early in life. Six months’ exposure only induced the tumor if exposures were begun at 8 months of age. This case was the only situation where aging the rats for 6 months
125
VERSUS CANCER INDUCTION
prior to exposure caused a higher incidence of cancer. Other neoplasms occasionally seen in both control and exposed rats included pituitary adenomas, follicular cell and C-cell tumors of the thyroid, pheochromocytomas and cortical and medullary tumors of the adrenal gland, tumors of the urinary bladder, and lymphoretitular tumors. The incidence of these tumors was not related to exposure. Tumor incidence in hamsters. In hamsters, hemangiosarcomas, mammary gland carcinomas, stomach adenomas, and skin carcinomas were produced by exposure to 200 ppm VC (Table 3). The highest incidence of hemangiosarcomas and stomach adenomas was seen in animals exposed early in life for only 6 months. Further exposures did not increase the incidence of these tumors in spite of a minimal further decrease in life span. To produce the highest stomach adenoma incidence, 6 months’ exposure was sufficient, starting at either 2 or 8 months of age. If exposures were started at 2 months of age, the incidence of mammary gland carcinoma was related to concentration X time through the first year of
TABLE 2 FREQUENCY(%)•
FTUMORSWITHINCREASEDINC~DENCEINFEMALEFISCHER-~~~ ADMINISTERED
Hemangiosarcoma
EXpJSWe period (months) 0
Liver 0.9 (l/112)
VINYL Mammary
All sites 1.8 (2/112)
Fibroadenoma 21.4 (24/112)
RATS
CHLORIDE gland Neoplastic nodule3 in liver
Adenocarcinoma 4.5 (5/l
12)
3.6 (4/l 12)
Hepatocellular carcinoma 0.9(1/112)
O-6 o-12 O-18 O-24
5.3 20.0 23.6 34.7
(4/76)# (1 I/55)” (13/55)” (19/55)b
5.3 21.4 27.3 43.6
(4i76) (12/56)” (15/55)” (24/55)b
36.8 50.0 43.6 47.3
(28/76)” (28/56)’ (24/55)” (26/55)”
7.9 19.6 16.4 9. I
(b/76) (I l/56)’ (9/55)” (5/55)”
20.0 35.7 13.0 10.9
(15175)” (20156)” (7/54)0 (6/55)”
4.0 7. I 14.8 16.4
(3/75) (4/56) (8/54)6 (9/55)O
O-6 b-12 12-18 18-24
5.3 3.8 0.0 0.0
(4176)’ (2/52) (O/5 I) (O/53)
5.3 3.8 0.0 0.0
(4/76) (2/53) (O/53) (O/53)
36.8 43.4 32.1 37.7
(28176)” (23/53)’ (17/53) (20/53)’
7.9 3.8 5.7 3.8
(b/76) (2/53) (3/53) (2/53)
20.0 19.2 3.9 7.5
( I5/75)” (10/52)” (2/5 I) (4/53)
4.0 11.5 0.0 1.9
(3/75) (b/52)’ (O/5 I) (I/53)
o-12 6-18 12-24 a Different b Different ‘Different
20.0 (I I/55)” 9.3 (5/54)’ 4.1 (2/49) from controls, from all other from controls,
21.4 (12/56)” 9.1 (5/55)5 4.0 (2/50)
50.0 (28/56)” 29.1 (16/55)” 30.0 ( 1S/50)
p < 0.01 (life table analysis). groups, p < 0.01 (life table analysis). p < 0.05 (life table analysis).
19.6 (I I/56)” 7.3 (4155) 0.0 (O/50)
35.7 (20/56)” 7.4 (4154) 8.2 (4/49)
7.1 (4156)” 1.9 (l/54) 0.0 (O/49)
126
DREW
ET
TABLE
AL. 3
FREQUENCY(%) OF TUMORS SHOWING INCREASEDINCIDENCEIN FEMALE GOLDEN SYRIAN HAMSTERS ADMINISTERED VIWL CHLORIDE Exposure period (months)
Hemangiosarcoma (all sites)”
0
0.0 (O/143)
O-6 o-12 O-18 O-6 6-12 12-18
18-24 o-12 6-18 12-24
Mammary gland carcinoma
Stomach adenoma
Skin carcinoma
0.0 (01143)
3.6 (5/138)
0.0 (O/133)
14.8 (13/88)” 7.1 (4/52)b 1.9 (2/103)
32.2 (28/87)b 59.6 (31/52)b 46.1 (47/102)b
26.1 (23/88)b 6.0 (3/50)’ 19.8 (20/101)b
2.5 (2/80) 18.8 (9/48)b 3.3 (3/90)
14.8 5.7 0.0 0.0
32.2 3.8 0.0 1.9
26.1 28.3 12.2 0.0
(13/88)b (3/53)’ (O/50) (O/52)
7.7 (4/52)b 2.3 (l/44) 0.0 (O/43)
(28/87)b (2/52)’ (O/50) (l/52)
59.6 (3 1/52)b
13.6 (6/44)b 0.0 (O/42)
(23/88)b (15/53)b (6149)’ (O/52)
6.0 (3/50)’ 22.7 (10/44)’ 7.3 (3/41)
2.5 0.0 0.0 0.0
(2/80) (O/49) (O/46) (O/50)
2.5 (2/80) 0.0 (O/38) 0.0 (O/30)
’ These tumors occurred primarily in the skin, spleen, and liver. b Different from controls, p < 0.01 (life table analysis). ’ Different from controls, p < 0.05 (life table analysis).
exposure. Further exposures did not increase mammary gland carcinomas in B6C3Fl mice the incidence. If exposures were deferred until if started early in life or at 8 months of age. the hamsters were 8 months old, the incidence Exposures beginning after 1 year of age proof carcinomas of the mammary gland was duced mammary gland carcinomas in this markedly reduced. A 1-year exposure starting strain but at lower incidences. at 8 months of age did induce mammary carIn Swiss mice 6 months’ exposure to 50 cinomas but not nearly to the degree devel- ppm VC produced significant incidences of mammary gland carcioped by young hamsters exposed for 1 year. hemangiosarcomas, The incidence of skin carcinomas was only nomas, and lung carcinomas. Exposure besignificant in the group exposed for 1 year. yond that time had little effect on neoplasm Exposed hamsters also occasionally devel- incidence. If exposures were deferred for 6 or oped follicular adenomas of the thyroid and 12 months, the incidence was decreased for lung adenomas, primarily in the groups where each tumor type. exposure began early in their lifetime. HowLymphosarcomas were seen in both strains ever, the incidence was low and was not re- of control mice with the incidence being about lated to the exposure. 32% and 17% in B6C3Fl mice and Swiss mice, Tumor incidence in mice. Exposure of mice respectively. These tumors did not increase in to 50 ppm VC induced hemangiosarcomas an exposure-related way. Very few other tuand mammary gland carcinomas in both mors were seen in the Swiss mice. In B6C3Fl strains of mice and lung carcinomas in Swiss mice, additional tumors included phenmice only (Table 4). In B6C3Fl mice, ex- chromocytomas, adenomas, and carcinomas posure of 6 months or more induced 60 to of the adrenal gland, lung adenomas, and both 70% hemangiosarcomas, regardless of the age adenomas and carcinomas of the liver. None of the animals at the start of exposure. Six of these tumors appeared to be related to the months’ exposure was sufficient to induce exposure to VC.
AGE
AND
EXPOSURE
DURATION
VERSUS
TABLE FREQUENCY
Exposure period (months)
(%) OF TUMORS
Hemangiosarcoma
SHOWING INCREASED ADMINISTERED
CANCER
127
INDUCTION
4
INCIDENCE IN B6C3Fl VINYL CHLORIDE
(ah sites)’
Mammary
AND CD-I
SWISS FEMALE
gland carcinoma Lung
B6C3Fl
Swiss
MICE
B6C3Fl
SWiSS
carcinoma SW&
12.7 (9/71)
0
5.8 (4/69)
1.4 (l/71)
4.3 (3169)
2.8 (2/71)
O-6 o-12 O-18
68.7 (46/67)b 76.7 (69190)” -
43.3 (29/67)b 63.8 (30147)’ 44.4 (20/45)b
43.2 (29/67)b 41.1 (37/90)b -
49.3 (33/67)b 46.8 (22/47)b 48.9 (22/45)b
27.7 (18/65)b 31.9 (15/47)b 24.4 (1 1/45)b
O-6 6-12 12-18
68.7 (46/67)b 64.3 (27/42)b 58.8 (30/51)b
43.3 (29,‘67)b 22.4 (1 1/49)b 9.4 (5/53)
43.2 (29/67)b 31.0 (13/42)” 7.8 (4/51)’
49.3 (33,‘67)b 26.5 (13149)” 3.8 (2/53)
27.7 (18/65)b 26.5 (13/49)’ 13.2 (7153)
o-12 6-18 12-24
76.7 (69/90)b 62.5 (30/48)b 60.4 (29/48)b
63.8 (30/47)b 37.0 (1 7/46)b 6.0 (3/50)
41.1 (37/9O)b 18.8 (9/48)b 8.3 (4/48)b
46.8 (22/47)b 17.8 (8/45)b 0.0 (O/50)
31.9 (15/47)” 19.6 (9/46)’ 6.0 (3150)
a These tumors occurred b Different from controls, ’ Different from controls,
primarily in the peritoneal (both p < 0.01 (life table analysis). p < 0.05 (life table analysis).
DISCUSSION The purpose of this study was to evaluate the effect of duration of exposure and age at first exposure on tumor incidence in three rodent species with a known and well-characterized carcinogen. Vinyl chloride has been extensively studied in rats, mice, and hamsters (Maltoni, 1977; Maltoni et al., 198 1). In rats, VC produced hemangiosarcomas, hepatocellular carcinomas, mammary carcinomas, and at higher doses, zymbal gland carcinomas and olfactory neuroblastomas of the nasal cavity. Hamsters and mice exhibit lung neoplasms, mammary carcinomas, and hemangiosarcomas after VC exposure. In addition, hamsters develop neoplasms in the nonglandular portion of the stomach. The VC exposures in this study produced four major neoplasms in rats, four in hamsters, two in B6C3Fl mice, and three in Swiss mice. The pattern of tumors in this study (Tables 2 to 4) was consistent with the observations reported above. Since nasal neoplasms
strains),
subcutis
(B6C3Fl),
and skin (Swiss).
were recently reported to be induced by formaldehyde (Swenberg et al., 1980), the nasal cavities of all animals in this study were examined. The only VC-related neoplasms at this site were squamous cell carcinomas in rats that appeared to arise in the nasal lacrimal duct. In rats with early initial exposures, the incidence of hemangiosarcomas and hepatocellular carcinomas increased with duration of exposure. However, the incidence of neoplastic nodules of the liver peaked after 1 year. While there is some controversy about the interpretation of the dose relationship of neoplastic nodules, if the incidence of neoplastic nodules is combined with hepatocellular carcinomas, the maximum liver neoplasm response was seen after a year of exposure. The maximum incidence for benign and malignant (combined) mammary gland tumors was seen after a year of exposure in young animals. Increasing the duration of exposure from 12 to 18 months did not, in general, increase the incidence of these neoplasms. Thus, it ap-
128
DREW
pears that an exposure of one year is sufficient to induce all the major types of neoplasms seen in rats. In mice and hamsters, 6 months’ exposure, beginning at 2 months of age, resulted in the significant tumor incidence in eight of the nine major tumors observed. Further exposure resulted in higher incidence only in hemangiosarcomas in Swiss mice and mammary gland carcinomas in hamsters. The incidence of hemangiosarcomas and stomach adenomas in hamsters was actually reduced with further exposure. However, this finding may be related to the fact that the life spans in chronically exposed hamsters were reduced, possibly to the point where they died before a potential neoplasm could be expressed. The incidence of the other neoplasms in mice was similar after either 6 or 12 months’ exposure. Thus, if begun early in life, 6 months exposure to VC is sufficient to demonstrate carcinogenicity in both mice and hamsters. Unfortunately, the high mortality in both species precludes selection of an optimum exposure duration in these species. Hehir et al. ( 198 1) exposed mice and rats to VC, assessing the effects of single I-hr exposures at concentrations of up to 50,000 ppm. In another experiment they kept concentration X time (CT) constant at 5000 ppm-hr and exposed animals for 1 hr/day to either 50 ppm for 100 days or 500 ppm for 10 days. No rats developed VC-related tumors in their studies. A single I-hr exposure to 50,000 ppm increased the lung adenoma incidence in female mice and induced 2/76 lung carcinomas. Only when the concentration was 500 ppm did exposure to 5000 ppm-hr increase the number of lung adenomas. They concluded that in mice a CT between 50,000 and 500,000 ppm-hr is a definite carcinogenic dose. In our mice, the minimum exposure was about 36,000 ppm-hr, thus our results suggest that 36,000 ppm-hr is a carcinogenic dose. However, our tumor incidence was much higher, presumably because the exposures were spread out over 6 months.
ET
AL.
One goal of these studies was to assessthe optimum age at which to begin inhalation carcinogenesis studies. Groth et al. ( 198 1) observed an increase in hemangiosarcoma production with age. They exposed male and female adult Sprague-Dawley rats of four different ages (6, 18, 33, and 53 weeks) to 940 ppm VC for 7 hr/day, 5 days/week for 24 weeks. Exposures were scheduled for 1 year, but because of pneumonia the studies were terminated 43 weeks after exposures began. This procedure resulted in a maximum age at termination of 49, 6 1, 76, and 96 weeks. However, the mean number of exposure weeks for all groups of animals was only about 24 weeks. In their female rats, the incidence of hemangiosarcomas was 5, 15,47, and 20% in animals of 6, 18, 33, or 53 weeks of age, respectively, at the beginning of the exposures. In males the incidence was 0, 0, 7, and 24%, respectively. They concluded that older rats were more susceptible to VC-induced hemangiosarcomas, and that females were more sensitive than males. Furthermore, they suggest that bioassay studies be modified by using older animals and exposing for a shorter duration. The data reported here do not support the conclusions of Groth et al. ( 198 1). We concur that females are more sensitive to VC; however, we strongly disagree with the conclusion that older animals are more susceptible to VC induced neoplasms. Groth et al. (198 1) only reported the incidence of hemangiosarcomas, a tumor that develops late in an animal’s life, regardless of when the exposure occurred. Since animals which were exposed at a younger age were also killed at a younger age, it is probable that the animals did not live long enough to develop readily observable hemangiosarcomas. In contrast, all the animals in our study were allowed to live out their natural life span, being killed only when found moribund. Other differences between the study of Groth ei al. (198 1) and ours include the fact that they used a different strain of rat (Sprague-Dawley), a higher concentration of
AGE
AND
EXPOSURE
DURATION
VERSUS
VC, and four different lots of animals. The most plausible explanation for the markedly different results in the two studies is the fact that in the study of Groth et al. (198 1) the younger rats were not allowed to reach the age where the hemangiosarcomas would have been readily observable. The data reported here indicate that inhalation carcinogenesis studies should begin early in life. Maltoni et al. (1982) also demonstrated that animals exposed early in life are more susceptible to the carcinogenic effects of VC. They showed that l-day-old rats exposed to VC had a much higher incidence of hemangiosarcomas and hepatomas than rats exposed to a similar dosage regimen at 11 weeks of age. When Sinha and Dao (1980) gave rats a single iv injection of 7,12-dimethylbenz[a]anthracene (DMBA), 80 to 90% of rats exposed at 60 to 70 days of age, and 40% of rats exposed at 90 days of age developed tumors. However, animals given DMBA at 120 days of age were refractory to tumor induction. In summary, the generally accepted notion that young animals are more sensitive
CANCER
129
INDUCTION
to the carcinogenic effects of most chemicals, including VC, is still valid. It is interesting to note that the time required for development of hemangiosarcoma generally remained the same, regardless of the rodent age at initiation of exposure (Table 5). B6C3Fl mice exposed to VC for 6 months after a holding period of 0, 6, or 12 months died with hemangiosarcoma an average of 343 days after exposure began. Swiss mice and rats showed a similar trend. In hamsters, the low incidence of hemangiosarcomas and the fact that deaths occurred earlier precluded making such an assessment. In Swiss mice where the toxicity of VC reduced the life span so that it was shorter than the time to tumor, a marked reduction in tumor incidence occurred. This observation confirms that doses that result in marked dimunition of life span should be avoided in screening for potential carcinogens. These data clearly show that limited exposures to VC will demonstrate carcinogenicity. For this carcinogen, a 1-year exposure appears as effective as a 2-year exposure for
TABLE
5
THETIMEREOUIREDTOINDUCEDEATHFROMHEMANGIOSARCOMAS B6C3Fl Exposure period (months)
Incidence
female
mice
CD-l
Induction time”
Incidence
Swiss female
mice
Female
Induction time’
Incidence
rats
Induction time’
0
4169
O-6 o-12 O-18 O-24
46167 32144 37146
343 t313 2 313 + -
8 5 4
29167 30147 20145
369 f 12 350 * 9 350 k 10
-
4176 12156 15155 24155
O-6 6-12 12-18 18-24
46167 27142 3015 1 -
343 -+ 8 344 * 10 343 f 12
29167 1 l/49 5153
-
369 f 12 340 k 15 226 + 39 -
4176 2153 o/53 o/53
716 -+ 60 613 + 34 -
o-12 6-18 12-24
32144 30148 29/48
313 * 319 + 304*
30147 1 J/46 3150
350 k 9 323 k 16 124 + 35
12156 5155 2150
671 -t 48 537 k 29 390 + 87
’ Survival
-
in days from
l/J1
Fischer-344
-
-
5 6 5
the date each animal
received
2/l 12
the first exposure
to vinyl
chloride
(i f SE).
716 671 643 666
+ + 4 If;
60 48 16 11
130
DREW
producing carcinogen-related tumors, suggesting a potential for reducing both cost and the need for chamber space. However, it cannot be assumed that all inhaled carcinogens will behave in a similar manner. Whether or not there is a correlation between carcinogen potency and the ability of a l-year exposure interval to produce neoplasms remains to be determined. Additional studies are needed to learn if exposures for 1 year are adequate to predict the lifetime effect for potent carcinogens in general, and the extent to which this exposure duration can assess the activity of weak carcinogens. Whether the results reported for VC-induced neoplasms also apply to tumor induction by other routes also needs to be determined. In summary, the data clearly demonstrate that: (1) regardless of VC exposure duration, a higher incidence of neoplasms is observed when exposures are started early in life [in direct contrast to the conclusions of Groth et al. (198 l)]; (2) in spite of high mortality, both mice and hamsters will develop VC-induced neoplasms after only 6 months of exposure; and (3) in rats, a year of exposure was sufficient to demonstrate increased incidences of all major types of VC-induced neoplasms. If these principles can be demonstrated for other carcinogens by various exposure routes, it may be possible to significantly shorten the duration of their exposure. However, if exposure durations are shortened, the animals should be observed throughout their natural lifetime. ACKNOWLEDGMENTS The exposures for this study were carried out by the Becton Dickinson Research Center under contract to the NIEHS. The authors are grateful to Dr. Finis L. Cavender and Dr. Connie J. Stone and their staff.
REFERENCES
ET
AL.
duction of hemangiosarcoma. Environ.
Health
Per-
spect. 41, 53-57.
HEHIR, R. M., MCNAMARA, B. P., MCLAUGHLIN, J., WILLIGAN, D. A., BIERBOWER, G., AND HARDISTY, J. F. (1981). Cancer induction following single and multiple exposures to a constant amount of vinyl chloride monomer. Environ. Health Perspect. 41, 63-72. HINNERS, R. G., BURKART, J. K., AND PUNTE, C. L. (1968). Animal inhalation exposure chambers. Arch. Environ.
Health
16, 194-206.
KEPLINGER, M. L., GOODE, J. W., GORDON, D. E., AND CALANDARA, J. C. (1975). Interim results of exposure of rats, hamsters, and mice to vinyl chloride. Ann. N. Y. Acad. Sci. 246, 2 19-224.
MALTONI, C. (1977) Vinyl chloride carcinogenicity: An experimental model for cam&genesis studies. In Origins of Human Cancer (H. W. Hiatt, J. D. Watson, and J. A. Winsten. eds.), p. 119-146. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. MALTONI, C., AND LEFEMINE, G. (1974). Carcinogenicity bioassaysof vinyl chloride. 1. Research plans and early results. Environ. Rex 41, 387-405. MALTONI, C., LEFEMINE, G., CILIBERTI, A., COTTI, G., AND CARETTI, D. (198 1). Carcinogenicity bioassays of vinyl chloride monomers. A model of risk assessment on an experimental basis. Environ. Health Perspect. 41, 3-30.
MALTONI, C., LEFEMINE, G., CILIBERTI, A., Co~-rr, G., AND CARE~I, D. (1982). Vinyl chloride: A model carcinogen for risk assessment. In Genotoxic E&cts of Airborne Agents (R. R. Tice, D. L. Costa, and K. M. Schaich, eds.), p. 329-344. Plenum, New York. MANTEL, N., AND HAENSZEL, W. (1959). Statistical aspects of the analysis of data from retrospective studies of disease. J. Nat. Cancer Inst. 22, 719-758. PETO, R.. PIKE, M., DAY, N., GRAY, R., LEE, P., PARISH, S.. PETO, J., RICHARDS, S., AND WAHRENDORF, J. (1980). Guidelines for simple, sensitive signihcance tests for carcinogenic effects in long term animal experiments. In IARC Monographs on the Long Term and Short Term Screening Assays for Carcinogenesis: Critical Appraisal. WHO, Geneva, Suppl. 2. pp. 321.
A
3 I l-
SINHA, D. K., AND DAO, T. L. (1980). Induction of mammary tumors in aging rats by 7,12-dimethylbenz(a)anthracene: Role of DNA synthesis during carcinogenesis. J. Nat. Cancer Inst. 64, 519-521. SWENBERG, J. A., KERNS, W. D., MITCHELL, R. I., GRALLA, E. J., AND PAVKOV, K. L. (1980). Induction of squamous cell carcinoma of the rat nasal cavity by inhalation exposure to formaldehyde vapor. Cancer Rex 40, 3398-3402.
Cox, D. R. (1972). Regression models and life tables. J. R. Stat. Sot. 34, 187-220. GROTH, D. H., COATE, W. B., ULLAND, B. M., AND HORNUNG, R. W. (1981). Effects of aging on the in-
WARD, J. M., GOODMAN, D. G., SQUIRES,R. A., CHU, K. C., AND LINHART, M. S. (1979). Neoplastic and nonneoplastic lesions in aging (C57BL/6NX C3H/Hen) F, (B6C3Fl) mice. J. Nat. Cancer Insf. 63, 849-854.
AND
APPLIED
PHARMACOLOGY
68, 120- 130 ( 1983)
The Effect of Age and Exposure Duration on Cancer Induction Known Carcinogen in Rats, Mice, and Hamsters
by a
ROBERT T.DREw,"~ GARY A.BOORMAN,JOSEPH KHASEMAN, ERNEST E. MCCONNELL, WILLIAM M.BusEY,~ AND JOHN A. MOORE National
Institute of Environmental Health Sciences, P.O. Box 12233, Research Triangle Park, North Carolina 27709
Received
September
20, 1982; accepted
December
7, 1982
The Effect of Age and Exposure Duration on Cancer Induction by a Known Carcinogen in Rats, Mice, and Hamsters. DREW. R. T., BOORMAN, G. A., HASEMAN, J. K.. MCCONNELL, E. E., BUSEY, W. M., AND MOORE, J. A. (1983). Toxicol. Appl. Pharmacol. 68, 120-I 30. Female Golden Syrian hamsters, F-344 rats, Swiss CD-l mice, and B6C3Fl hybrid mice were exposed 6 hr/day, 5 days/week to carcinogenic levels of vinyl chloride (VC) for 6, 12. 18, or 24 months (rats and hamsters only). Other groups of rodents were held for 6 or 12 months and then exposed for 6 or 12 months. At the end of the study the incidence of VC-induced neoplasms was compared in each of the groups to assessthe effectsof duration of exposure and age at the start of exposure on carcinogenicity of VC. In rats, with early initial exposure, hemangiosarcomas, hepatocellular carcinomas. and mammary gland carcinomas occurred with increasing incidence with longer exposure duration. Rats held for 6 months before exposure developed V&elated neoplasms, while rats held 12 months before the start of exposure failed to show a significantly increased incidence of these neoplasms. In hamsters, hemangiosarcomas, mammary gland carcinomas, gastric adenocarcinomas, and skin carcinomas resulted from VC exposure. The highest incidence of malignant neoplasms occurred in hamsters exposed for the first 12 months, whereas exposure begun after 12 months of age did not cause neoplasms. In both strains of mice, VC exposure during the first 6 months of the experiment induced a high incidence of hemangiosarcomas and mammary gland carcinomas. Swiss mice also developed lung carcinomas after only 6 months of exposure. In all three rodent species an initial 12 month exposure to VC was adequate to detect its carcinogenic potential, but the shortened survival of VC exposed mice and hamsters precluded a meaningful comparison with longer periods of exposure. Exposures were most effective when started early in life.
The National Cancer Institute/National Toxicology Program has tested over 200 chemicals in rodent lifetime bioassays for potential toxicity and carcinogenicity (Ward et al., 1979). In the standard bioassay, animals are exposed to the chemical by an appropriate route (water, feed, gavage, etc.) for 24 months. Inhalation exposure is chosen for many eni Current address: Medical Department, Brookhaven National Laboratory, Upton, New York 11973. 2 Person to whom reprint requests should be addressed. 3 Current address: Experimental Pathology Laboratories, P.O. Box 474, Hemdon, Va. 22070. 0041-008X/83 Copynghl All
rights
0 of
1983 reproduction
$3.00 by
120
Academic in
vironmental chemicals since it most closely mimics the situation in man. However, inhalation exposure is significantly more expensive than other methods, and the number of chemicals that can be tested is limited by the number of available chambers. Most carcinogenesis studies call for lifetime exposures. However, there is a time at which the development of the neoplasm has progressed to an irreversible point. At this point, further exposures to the carcinogen are unnecessary and, if the chemical exerts other toxicity as well, further exposures may even
any
Press.
Inc.
form
reserved.
AGE AND EXPOSURE
DURATION
reduce the incidence of neoplasia because of shortened life span. If it could be shown that exposures of 6 or 12 months’ duration were sufficient to demonstrate carcinogenicity, more chemicals could be studied in a given period of time. The purpose of this study was twofold: to compare the neoplasm incidence resulting from exposures over various portions of an animal’s lifetime and to determine the optimal period of an animal’s life span over which to conduct the exposures. Specifically, would a 6- or 12-month exposure be as effective as the current 24-month exposure, and if so, when should the exposures be performed? Vinyl chloride (VC) was the model compound since its carcinogenic properties have been extensively studied (Maltoni, 1977; Maltoni and Lefemine, 1974; Maltoni et al., 198 1; Keplinger et al., 1975). Because of the large numbers of animals required for the study and since the main purpose was to compare carcinogenic incidences across groups, only female rodents were used. METHODS Animals. Eight hundred female Fischer-344 rats, 800 female Golden Syrian hamsters, 600 female hybrid B6C3Fl (female C57BL/6N X male C3H/HeN), and 600 female CD-I Swiss mice were purchased at 5 to 6 weeks of age (Charles River, Portage, Mich.). Each group was received at one time, ear tagped, and weighed weekly for 3 weeks. Animals weighing within 1 SD of the mean weight were included for long-term studies. Rats and hamsters were randomly assigned to 1 of 10 exposure groups or three control groups while mice were randomly separated into eight exposure groups and two control groups. The actual number of animals in each group and the duration of exposure of each group are shown in Fig. 1. Control groups and groups exposed during the first 6 months of their lifetimes contained larger numbers of animals allowing for sequential termination at various times after the exposure. Animals killed sequentially to follow the progression of lesions and their controls are not included in this report. Included in this report are results from 2089 rodents allowed to complete their life span and then necropsied when found moribund or dead. Animals were housed in stainless-steel hanging wire cages (16 X 11 X 7 in.). Each cage was subdivided into four compartments, each housing one rat. two hamsters,
121
VERSUS CANCER INDUCTION VINYL
CHLORIDE
EXPOSURE
REGIMEN
Months 0
6
-
Expose
12
N
----Hold 0 Rots
I6
N and
Hamsters
= 56 = 54
24
Rots,Hamsters Mice
only
FIG. 1. Protocol design. The solid lines indicate exposure periods and the dotted lines are holding periods.
or three mice. The exposure units were returned to housing racks at the end of each exposure period. Both the chambers and housing racks had automatic watering systems. An open formula pelleted rodent diet (NIH-31) (Zeigler Brothers, Gardiner, Pa.) was fed ad libitum to the animals except during the exposure period. Water was available throughout the study. Animals were checked twice daily to assesstheir clinical condition. During preexposure and postexposure holding periods, animals were held in similar cages. These animals and controls had food withheld during VC exposure periods. Vinyl chloride exposures. Exposures were in 1.3-m’ chambers similar to those described by Hinners et al. (1968). Each chamber held 36 exposure cages in six tiers, thereby providing space for 144 rats, 288 hamsters, or 432 mice per chamber. Two chambers were used to expose rats while one chamber was used for mice and one for hamsters. Animals were rotated into the chambers such that during the study no more than five groups were being exposed at any time (Fig. 1). The temperature in the chambers was 21 + 2’C with a relative humidity of 55 f 10%. The location of the animals in the chamber was rotated daily to minimize exposure bias. Chambers had an airflow of 300 liters/min and a slight negative pressure (0.5 in. HZO). Commercial cylinders of liquid VC were the source of VC vapor which was regulated to 20 psi and routed to the chambers. Chamber concentrations were monitored six times a day with a Miran IA IR analyzer. The chamber exhaust was scrubbed through a dual-activated charcoal system (Calgon Corp.), and the exhaust was monitored on the same schedule as above by a second Miran analyzer to assure complete removal of the VC. For personnel protection, a Miran analyzer was used to continuously monitor the room air in the chamber rooms. Concentrations. At the time this study was designed, the most recent published reports (Maltoni and Lefemine, 1974; Keplinger et al., 1975) indicated that VC concentrations as low as 50 ppm could produce tumors. Since Keplinger et al. (1975) had published an interim report,
122
DREW ET AL.
their current results were sought. They exposed COBS rats, CD- 1 Swiss mice, and Golden Syrian hamsters to 0, 50, 200, or 2500 ppm VC for 7 hr/day. Their results (persona1 communication) indicated that exposure to 50 ppm did not shorten the life span of rats but that 200 ppm reduced the mean life from 30 to 21 months. Since many groups in the current study would receive only 6 months’ exposure, 100 ppm was selected for rats. The data also indicated that exposure to 50 ppm reduced the lifespan of CD-I mice; however, no evidence was available at that time that lower concentrations would produce tumors. Therefore, 50 ppm was chosen for mice. The studies of Keplinger et al. (1975) also showed that hamsters appeared more resistant to VC toxicity with only a slight reduction in life span after 250 ppm; therefore, a level of 200 ppm was selected for this study. Pathology. Each animal, when found moribund or dead, was subjected to a complete necropsy. In addition to macroscopic lesions noted during the postmortem examination, the following tissues were routinely collected: brain, pituitary gland, salivary gland, thyroid, nasal turbinates (three levels for hamsters and rats, two levels for mice), heart, liver, kidneys, spleen, adrenal gland, stomach, small intestine, large intestine, pancreas, sternum, and lungs. Tissue sections were prepared, stained with hematoxylin and eosin, and examined microscopically. Statistical analysis. Differences in survival between the various groups were assessedby life table analyses (Cox, 1972). For the statistical analysis of neoplastic lesions, two different methods of adjusting for intercurrent mortality were employed. Each used the classical methods for combining contingency tables developed by Mantel and Haenszel (1959). For lesions judged to be “fatal” (i.e., that either directly or indirectly caused the death of the animal), life table methodology (Cox, 1972) was employed to compare incidence rates. By this approach, the proportions of tumor-bearing animals in the VC and control groups were compared at each point in time at which an animal died with a neoplasm of interest. The denominators of these proportions were the total number of animals at risk in each group. These results were combined by Mantel-Haenszel methods to obtain an overall p value. Examples of fatal neoplasms are hemangiosarcoma, mammary gland carcinoma, and hepatocellular carcinoma. For lesions judged to be “nonfatal” (i.e., that were merely observed at autopsy in animals dying of an unrelated cause), the incidental tumor test proposed by Peto et al. (1980) was employed. With this approach, neoplasm rates in exposed and control animals were compared in successive time intervals. The denominators of these rates were the number of animals on which necropsies were performed during the time interval. The individual time interval comparisons were then combined by the previously described methods to obtain a single overall result. Examples of nonfatal neoplasms are pituitary adenoma and adrenal cortical adenoma. A detailed discussion of
the statistical analysis of fatal and nonfatal neoplasms is given by Peto et al. (1980).
RESULTS Chamber concentrations. During the 2-year study no problems were encountered with vinyl chloride generation. The mean exposure levels were within 1% of the target values. Mortality. The mean survival times for all groups of animals are shown in Table 1. In rats, only those exposed for 1 year or more, beginning at 10 weeks of age exhibited any life span shortening (Fig. 2). This effect was presumably due to cancer since no other overt VC-induced pathologic changes were noted. In all groups of hamsters where the exposure began early in life, the life span was significantly reduced (Fig. 3). These deaths in hamsters resulted from vinyl chloride toxicity with the animals dying primarily from hepatic necrosis. None of the hamsters scheduled for exposure for 24 months survived past 18 months; therefore, in the remaining tables those groups have been combined. It should be noted that the mean life span of the control hamsters was only 463 days. Hamsters scheduled for exposure from 12 to 24 months had significantly shortened life spans. However, since three of those animals survived more than 18 months, these data have not been combined with the data from 12 to 18 months. All VC-exposed B6C3Fl mice exhibited significant life span shortening (Fig. 4). This decrease was due to the development of neoplasms since liver necrosis was not a common finding. None of the scheduled 0 to 18-monthexposed mice survived more than 1 year; therefore, these data have been combined with the data from 0 to 12 months for analysis of neoplasm incidence. These mice survived about 315 days after the start of a 6-month exposure regardless of whether the exposure began early in life, after a 6-month holding period, or after holding the mice for 1 year. Exposure to 50 ppm VC early in life significantly shortened the life span of the Swiss
AGE AND EXPOSURE
DURATION
123
VERSUS CANCER INDUCTION
TABLE 1 MEANSURVIVALOFFEMALEANIMALSTHROUGHOUTTHESTUDY" Time (days) Exposure period (months)
Fischer-344 rats
Golden Syrian hamsters
0
703 f 15
463 ?Y11
780 + 21
474 f 14
O-6 O-12 O-18 O-24
682 634 575 622
k + f f
14 20’ 156 II*
390 355 342 347
* + + *
126 lob gb 9b
316+ 8”’ 301 f 5b.d 304k 5b -
340 + lo”,’ 347 f 9b.d 321 5~ 7’ -
O-6 6-12 12-18 18-24
682 703 688 708
+ rt + +
14 24 25 21
390 468 456 499
AZ12b + 16 zk 17 + 14
316+ 8’,’ 480 f 13”’ 695 + 12b,c -
340 Ik lob,’ 472 f 15 521 -+ 15 -
O-12 6-18 12-24
634 f 20’ 659 f 17 717 f 17
355 f lob 455 f 13 424 + l5b
301 f 5”d 479 + 9b.d 632 f 12”d
347 f 9b,d 443 iz 16 472 f 20
B6C3Fl mice
Swiss CD-l mice
’ Average lifetime in days from the day the first animals were exposed (2 f SE). b Different from controls (p < 0.01). c Different from all other groups exposed for 6 months (p < 0.01). d Different from all other groups exposed for 1 year (p < 0.01).
mice (Fig. 5). In spite of the reduced lifespan of the 0 to 1&month exposed mice, a number of mice survived longer than 12 months, and the data were not combined with that from the 0 to lZmonth-exposed animals. When exposures to 50 ppm VC were initiated at 8
months of age or later, there was no effect on lifespan with this strain of mice. Other signs of VC exposure were not evident. Tumor incidence in rats. Hemangiosarcomas, mammary neoplasms, and hepatocellular carcinomas were strongly associated with
100
20
4x)
460
540
600
660
720 Time (days)
180
840
900
FIG. 2. Percentage survival for female Fischer rats exposed to vinyl chloride for 0 (control, a) O-6 (b), O-12 (c), O-l 8 (d), and O-24 (e) months. Decreasing survival is seen with increasing exposure duration. Survival of rats held for any time interval before exposure was similar to controls (Table 1). For brevity, data from these studies are not shown.
DREW
124
ET
AL.
Time (days)
FIG. 3. Percentage a), O-6 (b), O-12 (c). duration. Survival of I). For brevity, these
survival for female Golden Syrian hamsters exposed to vinyl chloride for 0 (control, O-l 8 (d), and O-24 (e) months. Decreasing survival is seen with increasing exposure hamsters held for any time interval before exposure was similar to controls (Table data are not shown.
exposure to VC in rats (Table 2). The incidence of hemangiosarcomas was a function of the duration of exposure. Exposures for 6 months induced a low incidence of hemangiosarcomas only if begun early in life. Exposures of a year or more produced the neoplasm, except when started after 1 year, but the incidence was higher if exposures began early in life. Mammary gland adenocarcinomas were also evident after a year or more
exposure to VC; however, in this case the exposures had to begin prior to 8 months of age for this neoplasm to develop. The incidence of fibroadenomas of the mammary gland was increased by exposure to VC in most groups, but the increase was not always related to the duration of the exposure. The incidence of these neoplasms was higher when young rats were exposed. Hepatocellular carcinomas were induced in rats in a dose-related manner when
Time (days)
FIG. 4. Percentage survival for female B6C3Fl mice exposed to vinyl chloride for 0 (control, a), 6-12 (b), 6-18 (c), 12-18 (d), and 12-24 (e) months. Mice first treated after 6 months showed greater life shortening than those after 12 months. The survival of B6C3Fl mice treated from the beginning was essentially independent of exposure duration with similar mean survival times (Table I). For brevity these data are not plotted on this figure.
AGE AND EXPOSURE
DURATION
Time (days)
FIG. 5. Percentage survival for female CD- 1 Swiss mice to vinyl chloride for 0 (control, a), O-6 (b), O12 (c), and O-18 (d) months. All exposures were associated with a marked diminution of lifespan. In animals held for a period of time before exposure, life spans were generally not shortened (Table I). For brevity, these data are not shown. exposed
exposures began early in life. The incidence of this tumor was significantly elevated after 12 months’ exposure only if the exposures started early in life. Six months’ exposure only induced the tumor if exposures were begun at 8 months of age. This case was the only situation where aging the rats for 6 months
125
VERSUS CANCER INDUCTION
prior to exposure caused a higher incidence of cancer. Other neoplasms occasionally seen in both control and exposed rats included pituitary adenomas, follicular cell and C-cell tumors of the thyroid, pheochromocytomas and cortical and medullary tumors of the adrenal gland, tumors of the urinary bladder, and lymphoretitular tumors. The incidence of these tumors was not related to exposure. Tumor incidence in hamsters. In hamsters, hemangiosarcomas, mammary gland carcinomas, stomach adenomas, and skin carcinomas were produced by exposure to 200 ppm VC (Table 3). The highest incidence of hemangiosarcomas and stomach adenomas was seen in animals exposed early in life for only 6 months. Further exposures did not increase the incidence of these tumors in spite of a minimal further decrease in life span. To produce the highest stomach adenoma incidence, 6 months’ exposure was sufficient, starting at either 2 or 8 months of age. If exposures were started at 2 months of age, the incidence of mammary gland carcinoma was related to concentration X time through the first year of
TABLE 2 FREQUENCY(%)•
FTUMORSWITHINCREASEDINC~DENCEINFEMALEFISCHER-~~~ ADMINISTERED
Hemangiosarcoma
EXpJSWe period (months) 0
Liver 0.9 (l/112)
VINYL Mammary
All sites 1.8 (2/112)
Fibroadenoma 21.4 (24/112)
RATS
CHLORIDE gland Neoplastic nodule3 in liver
Adenocarcinoma 4.5 (5/l
12)
3.6 (4/l 12)
Hepatocellular carcinoma 0.9(1/112)
O-6 o-12 O-18 O-24
5.3 20.0 23.6 34.7
(4/76)# (1 I/55)” (13/55)” (19/55)b
5.3 21.4 27.3 43.6
(4i76) (12/56)” (15/55)” (24/55)b
36.8 50.0 43.6 47.3
(28/76)” (28/56)’ (24/55)” (26/55)”
7.9 19.6 16.4 9. I
(b/76) (I l/56)’ (9/55)” (5/55)”
20.0 35.7 13.0 10.9
(15175)” (20156)” (7/54)0 (6/55)”
4.0 7. I 14.8 16.4
(3/75) (4/56) (8/54)6 (9/55)O
O-6 b-12 12-18 18-24
5.3 3.8 0.0 0.0
(4176)’ (2/52) (O/5 I) (O/53)
5.3 3.8 0.0 0.0
(4/76) (2/53) (O/53) (O/53)
36.8 43.4 32.1 37.7
(28176)” (23/53)’ (17/53) (20/53)’
7.9 3.8 5.7 3.8
(b/76) (2/53) (3/53) (2/53)
20.0 19.2 3.9 7.5
( I5/75)” (10/52)” (2/5 I) (4/53)
4.0 11.5 0.0 1.9
(3/75) (b/52)’ (O/5 I) (I/53)
o-12 6-18 12-24 a Different b Different ‘Different
20.0 (I I/55)” 9.3 (5/54)’ 4.1 (2/49) from controls, from all other from controls,
21.4 (12/56)” 9.1 (5/55)5 4.0 (2/50)
50.0 (28/56)” 29.1 (16/55)” 30.0 ( 1S/50)
p < 0.01 (life table analysis). groups, p < 0.01 (life table analysis). p < 0.05 (life table analysis).
19.6 (I I/56)” 7.3 (4155) 0.0 (O/50)
35.7 (20/56)” 7.4 (4154) 8.2 (4/49)
7.1 (4156)” 1.9 (l/54) 0.0 (O/49)
126
DREW
ET
TABLE
AL. 3
FREQUENCY(%) OF TUMORS SHOWING INCREASEDINCIDENCEIN FEMALE GOLDEN SYRIAN HAMSTERS ADMINISTERED VIWL CHLORIDE Exposure period (months)
Hemangiosarcoma (all sites)”
0
0.0 (O/143)
O-6 o-12 O-18 O-6 6-12 12-18
18-24 o-12 6-18 12-24
Mammary gland carcinoma
Stomach adenoma
Skin carcinoma
0.0 (01143)
3.6 (5/138)
0.0 (O/133)
14.8 (13/88)” 7.1 (4/52)b 1.9 (2/103)
32.2 (28/87)b 59.6 (31/52)b 46.1 (47/102)b
26.1 (23/88)b 6.0 (3/50)’ 19.8 (20/101)b
2.5 (2/80) 18.8 (9/48)b 3.3 (3/90)
14.8 5.7 0.0 0.0
32.2 3.8 0.0 1.9
26.1 28.3 12.2 0.0
(13/88)b (3/53)’ (O/50) (O/52)
7.7 (4/52)b 2.3 (l/44) 0.0 (O/43)
(28/87)b (2/52)’ (O/50) (l/52)
59.6 (3 1/52)b
13.6 (6/44)b 0.0 (O/42)
(23/88)b (15/53)b (6149)’ (O/52)
6.0 (3/50)’ 22.7 (10/44)’ 7.3 (3/41)
2.5 0.0 0.0 0.0
(2/80) (O/49) (O/46) (O/50)
2.5 (2/80) 0.0 (O/38) 0.0 (O/30)
’ These tumors occurred primarily in the skin, spleen, and liver. b Different from controls, p < 0.01 (life table analysis). ’ Different from controls, p < 0.05 (life table analysis).
exposure. Further exposures did not increase mammary gland carcinomas in B6C3Fl mice the incidence. If exposures were deferred until if started early in life or at 8 months of age. the hamsters were 8 months old, the incidence Exposures beginning after 1 year of age proof carcinomas of the mammary gland was duced mammary gland carcinomas in this markedly reduced. A 1-year exposure starting strain but at lower incidences. at 8 months of age did induce mammary carIn Swiss mice 6 months’ exposure to 50 cinomas but not nearly to the degree devel- ppm VC produced significant incidences of mammary gland carcioped by young hamsters exposed for 1 year. hemangiosarcomas, The incidence of skin carcinomas was only nomas, and lung carcinomas. Exposure besignificant in the group exposed for 1 year. yond that time had little effect on neoplasm Exposed hamsters also occasionally devel- incidence. If exposures were deferred for 6 or oped follicular adenomas of the thyroid and 12 months, the incidence was decreased for lung adenomas, primarily in the groups where each tumor type. exposure began early in their lifetime. HowLymphosarcomas were seen in both strains ever, the incidence was low and was not re- of control mice with the incidence being about lated to the exposure. 32% and 17% in B6C3Fl mice and Swiss mice, Tumor incidence in mice. Exposure of mice respectively. These tumors did not increase in to 50 ppm VC induced hemangiosarcomas an exposure-related way. Very few other tuand mammary gland carcinomas in both mors were seen in the Swiss mice. In B6C3Fl strains of mice and lung carcinomas in Swiss mice, additional tumors included phenmice only (Table 4). In B6C3Fl mice, ex- chromocytomas, adenomas, and carcinomas posure of 6 months or more induced 60 to of the adrenal gland, lung adenomas, and both 70% hemangiosarcomas, regardless of the age adenomas and carcinomas of the liver. None of the animals at the start of exposure. Six of these tumors appeared to be related to the months’ exposure was sufficient to induce exposure to VC.
AGE
AND
EXPOSURE
DURATION
VERSUS
TABLE FREQUENCY
Exposure period (months)
(%) OF TUMORS
Hemangiosarcoma
SHOWING INCREASED ADMINISTERED
CANCER
127
INDUCTION
4
INCIDENCE IN B6C3Fl VINYL CHLORIDE
(ah sites)’
Mammary
AND CD-I
SWISS FEMALE
gland carcinoma Lung
B6C3Fl
Swiss
MICE
B6C3Fl
SWiSS
carcinoma SW&
12.7 (9/71)
0
5.8 (4/69)
1.4 (l/71)
4.3 (3169)
2.8 (2/71)
O-6 o-12 O-18
68.7 (46/67)b 76.7 (69190)” -
43.3 (29/67)b 63.8 (30147)’ 44.4 (20/45)b
43.2 (29/67)b 41.1 (37/90)b -
49.3 (33/67)b 46.8 (22/47)b 48.9 (22/45)b
27.7 (18/65)b 31.9 (15/47)b 24.4 (1 1/45)b
O-6 6-12 12-18
68.7 (46/67)b 64.3 (27/42)b 58.8 (30/51)b
43.3 (29,‘67)b 22.4 (1 1/49)b 9.4 (5/53)
43.2 (29/67)b 31.0 (13/42)” 7.8 (4/51)’
49.3 (33,‘67)b 26.5 (13149)” 3.8 (2/53)
27.7 (18/65)b 26.5 (13/49)’ 13.2 (7153)
o-12 6-18 12-24
76.7 (69/90)b 62.5 (30/48)b 60.4 (29/48)b
63.8 (30/47)b 37.0 (1 7/46)b 6.0 (3/50)
41.1 (37/9O)b 18.8 (9/48)b 8.3 (4/48)b
46.8 (22/47)b 17.8 (8/45)b 0.0 (O/50)
31.9 (15/47)” 19.6 (9/46)’ 6.0 (3150)
a These tumors occurred b Different from controls, ’ Different from controls,
primarily in the peritoneal (both p < 0.01 (life table analysis). p < 0.05 (life table analysis).
DISCUSSION The purpose of this study was to evaluate the effect of duration of exposure and age at first exposure on tumor incidence in three rodent species with a known and well-characterized carcinogen. Vinyl chloride has been extensively studied in rats, mice, and hamsters (Maltoni, 1977; Maltoni et al., 198 1). In rats, VC produced hemangiosarcomas, hepatocellular carcinomas, mammary carcinomas, and at higher doses, zymbal gland carcinomas and olfactory neuroblastomas of the nasal cavity. Hamsters and mice exhibit lung neoplasms, mammary carcinomas, and hemangiosarcomas after VC exposure. In addition, hamsters develop neoplasms in the nonglandular portion of the stomach. The VC exposures in this study produced four major neoplasms in rats, four in hamsters, two in B6C3Fl mice, and three in Swiss mice. The pattern of tumors in this study (Tables 2 to 4) was consistent with the observations reported above. Since nasal neoplasms
strains),
subcutis
(B6C3Fl),
and skin (Swiss).
were recently reported to be induced by formaldehyde (Swenberg et al., 1980), the nasal cavities of all animals in this study were examined. The only VC-related neoplasms at this site were squamous cell carcinomas in rats that appeared to arise in the nasal lacrimal duct. In rats with early initial exposures, the incidence of hemangiosarcomas and hepatocellular carcinomas increased with duration of exposure. However, the incidence of neoplastic nodules of the liver peaked after 1 year. While there is some controversy about the interpretation of the dose relationship of neoplastic nodules, if the incidence of neoplastic nodules is combined with hepatocellular carcinomas, the maximum liver neoplasm response was seen after a year of exposure. The maximum incidence for benign and malignant (combined) mammary gland tumors was seen after a year of exposure in young animals. Increasing the duration of exposure from 12 to 18 months did not, in general, increase the incidence of these neoplasms. Thus, it ap-
128
DREW
pears that an exposure of one year is sufficient to induce all the major types of neoplasms seen in rats. In mice and hamsters, 6 months’ exposure, beginning at 2 months of age, resulted in the significant tumor incidence in eight of the nine major tumors observed. Further exposure resulted in higher incidence only in hemangiosarcomas in Swiss mice and mammary gland carcinomas in hamsters. The incidence of hemangiosarcomas and stomach adenomas in hamsters was actually reduced with further exposure. However, this finding may be related to the fact that the life spans in chronically exposed hamsters were reduced, possibly to the point where they died before a potential neoplasm could be expressed. The incidence of the other neoplasms in mice was similar after either 6 or 12 months’ exposure. Thus, if begun early in life, 6 months exposure to VC is sufficient to demonstrate carcinogenicity in both mice and hamsters. Unfortunately, the high mortality in both species precludes selection of an optimum exposure duration in these species. Hehir et al. ( 198 1) exposed mice and rats to VC, assessing the effects of single I-hr exposures at concentrations of up to 50,000 ppm. In another experiment they kept concentration X time (CT) constant at 5000 ppm-hr and exposed animals for 1 hr/day to either 50 ppm for 100 days or 500 ppm for 10 days. No rats developed VC-related tumors in their studies. A single I-hr exposure to 50,000 ppm increased the lung adenoma incidence in female mice and induced 2/76 lung carcinomas. Only when the concentration was 500 ppm did exposure to 5000 ppm-hr increase the number of lung adenomas. They concluded that in mice a CT between 50,000 and 500,000 ppm-hr is a definite carcinogenic dose. In our mice, the minimum exposure was about 36,000 ppm-hr, thus our results suggest that 36,000 ppm-hr is a carcinogenic dose. However, our tumor incidence was much higher, presumably because the exposures were spread out over 6 months.
ET
AL.
One goal of these studies was to assessthe optimum age at which to begin inhalation carcinogenesis studies. Groth et al. ( 198 1) observed an increase in hemangiosarcoma production with age. They exposed male and female adult Sprague-Dawley rats of four different ages (6, 18, 33, and 53 weeks) to 940 ppm VC for 7 hr/day, 5 days/week for 24 weeks. Exposures were scheduled for 1 year, but because of pneumonia the studies were terminated 43 weeks after exposures began. This procedure resulted in a maximum age at termination of 49, 6 1, 76, and 96 weeks. However, the mean number of exposure weeks for all groups of animals was only about 24 weeks. In their female rats, the incidence of hemangiosarcomas was 5, 15,47, and 20% in animals of 6, 18, 33, or 53 weeks of age, respectively, at the beginning of the exposures. In males the incidence was 0, 0, 7, and 24%, respectively. They concluded that older rats were more susceptible to VC-induced hemangiosarcomas, and that females were more sensitive than males. Furthermore, they suggest that bioassay studies be modified by using older animals and exposing for a shorter duration. The data reported here do not support the conclusions of Groth et al. ( 198 1). We concur that females are more sensitive to VC; however, we strongly disagree with the conclusion that older animals are more susceptible to VC induced neoplasms. Groth et al. (198 1) only reported the incidence of hemangiosarcomas, a tumor that develops late in an animal’s life, regardless of when the exposure occurred. Since animals which were exposed at a younger age were also killed at a younger age, it is probable that the animals did not live long enough to develop readily observable hemangiosarcomas. In contrast, all the animals in our study were allowed to live out their natural life span, being killed only when found moribund. Other differences between the study of Groth ei al. (198 1) and ours include the fact that they used a different strain of rat (Sprague-Dawley), a higher concentration of
AGE
AND
EXPOSURE
DURATION
VERSUS
VC, and four different lots of animals. The most plausible explanation for the markedly different results in the two studies is the fact that in the study of Groth et al. (198 1) the younger rats were not allowed to reach the age where the hemangiosarcomas would have been readily observable. The data reported here indicate that inhalation carcinogenesis studies should begin early in life. Maltoni et al. (1982) also demonstrated that animals exposed early in life are more susceptible to the carcinogenic effects of VC. They showed that l-day-old rats exposed to VC had a much higher incidence of hemangiosarcomas and hepatomas than rats exposed to a similar dosage regimen at 11 weeks of age. When Sinha and Dao (1980) gave rats a single iv injection of 7,12-dimethylbenz[a]anthracene (DMBA), 80 to 90% of rats exposed at 60 to 70 days of age, and 40% of rats exposed at 90 days of age developed tumors. However, animals given DMBA at 120 days of age were refractory to tumor induction. In summary, the generally accepted notion that young animals are more sensitive
CANCER
129
INDUCTION
to the carcinogenic effects of most chemicals, including VC, is still valid. It is interesting to note that the time required for development of hemangiosarcoma generally remained the same, regardless of the rodent age at initiation of exposure (Table 5). B6C3Fl mice exposed to VC for 6 months after a holding period of 0, 6, or 12 months died with hemangiosarcoma an average of 343 days after exposure began. Swiss mice and rats showed a similar trend. In hamsters, the low incidence of hemangiosarcomas and the fact that deaths occurred earlier precluded making such an assessment. In Swiss mice where the toxicity of VC reduced the life span so that it was shorter than the time to tumor, a marked reduction in tumor incidence occurred. This observation confirms that doses that result in marked dimunition of life span should be avoided in screening for potential carcinogens. These data clearly show that limited exposures to VC will demonstrate carcinogenicity. For this carcinogen, a 1-year exposure appears as effective as a 2-year exposure for
TABLE
5
THETIMEREOUIREDTOINDUCEDEATHFROMHEMANGIOSARCOMAS B6C3Fl Exposure period (months)
Incidence
female
mice
CD-l
Induction time”
Incidence
Swiss female
mice
Female
Induction time’
Incidence
rats
Induction time’
0
4169
O-6 o-12 O-18 O-24
46167 32144 37146
343 t313 2 313 + -
8 5 4
29167 30147 20145
369 f 12 350 * 9 350 k 10
-
4176 12156 15155 24155
O-6 6-12 12-18 18-24
46167 27142 3015 1 -
343 -+ 8 344 * 10 343 f 12
29167 1 l/49 5153
-
369 f 12 340 k 15 226 + 39 -
4176 2153 o/53 o/53
716 -+ 60 613 + 34 -
o-12 6-18 12-24
32144 30148 29/48
313 * 319 + 304*
30147 1 J/46 3150
350 k 9 323 k 16 124 + 35
12156 5155 2150
671 -t 48 537 k 29 390 + 87
’ Survival
-
in days from
l/J1
Fischer-344
-
-
5 6 5
the date each animal
received
2/l 12
the first exposure
to vinyl
chloride
(i f SE).
716 671 643 666
+ + 4 If;
60 48 16 11
130
DREW
producing carcinogen-related tumors, suggesting a potential for reducing both cost and the need for chamber space. However, it cannot be assumed that all inhaled carcinogens will behave in a similar manner. Whether or not there is a correlation between carcinogen potency and the ability of a l-year exposure interval to produce neoplasms remains to be determined. Additional studies are needed to learn if exposures for 1 year are adequate to predict the lifetime effect for potent carcinogens in general, and the extent to which this exposure duration can assess the activity of weak carcinogens. Whether the results reported for VC-induced neoplasms also apply to tumor induction by other routes also needs to be determined. In summary, the data clearly demonstrate that: (1) regardless of VC exposure duration, a higher incidence of neoplasms is observed when exposures are started early in life [in direct contrast to the conclusions of Groth et al. (198 l)]; (2) in spite of high mortality, both mice and hamsters will develop VC-induced neoplasms after only 6 months of exposure; and (3) in rats, a year of exposure was sufficient to demonstrate increased incidences of all major types of VC-induced neoplasms. If these principles can be demonstrated for other carcinogens by various exposure routes, it may be possible to significantly shorten the duration of their exposure. However, if exposure durations are shortened, the animals should be observed throughout their natural lifetime. ACKNOWLEDGMENTS The exposures for this study were carried out by the Becton Dickinson Research Center under contract to the NIEHS. The authors are grateful to Dr. Finis L. Cavender and Dr. Connie J. Stone and their staff.
REFERENCES
ET
AL.
duction of hemangiosarcoma. Environ.
Health
Per-
spect. 41, 53-57.
HEHIR, R. M., MCNAMARA, B. P., MCLAUGHLIN, J., WILLIGAN, D. A., BIERBOWER, G., AND HARDISTY, J. F. (1981). Cancer induction following single and multiple exposures to a constant amount of vinyl chloride monomer. Environ. Health Perspect. 41, 63-72. HINNERS, R. G., BURKART, J. K., AND PUNTE, C. L. (1968). Animal inhalation exposure chambers. Arch. Environ.
Health
16, 194-206.
KEPLINGER, M. L., GOODE, J. W., GORDON, D. E., AND CALANDARA, J. C. (1975). Interim results of exposure of rats, hamsters, and mice to vinyl chloride. Ann. N. Y. Acad. Sci. 246, 2 19-224.
MALTONI, C. (1977) Vinyl chloride carcinogenicity: An experimental model for cam&genesis studies. In Origins of Human Cancer (H. W. Hiatt, J. D. Watson, and J. A. Winsten. eds.), p. 119-146. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. MALTONI, C., AND LEFEMINE, G. (1974). Carcinogenicity bioassaysof vinyl chloride. 1. Research plans and early results. Environ. Rex 41, 387-405. MALTONI, C., LEFEMINE, G., CILIBERTI, A., COTTI, G., AND CARETTI, D. (198 1). Carcinogenicity bioassays of vinyl chloride monomers. A model of risk assessment on an experimental basis. Environ. Health Perspect. 41, 3-30.
MALTONI, C., LEFEMINE, G., CILIBERTI, A., Co~-rr, G., AND CARE~I, D. (1982). Vinyl chloride: A model carcinogen for risk assessment. In Genotoxic E&cts of Airborne Agents (R. R. Tice, D. L. Costa, and K. M. Schaich, eds.), p. 329-344. Plenum, New York. MANTEL, N., AND HAENSZEL, W. (1959). Statistical aspects of the analysis of data from retrospective studies of disease. J. Nat. Cancer Inst. 22, 719-758. PETO, R.. PIKE, M., DAY, N., GRAY, R., LEE, P., PARISH, S.. PETO, J., RICHARDS, S., AND WAHRENDORF, J. (1980). Guidelines for simple, sensitive signihcance tests for carcinogenic effects in long term animal experiments. In IARC Monographs on the Long Term and Short Term Screening Assays for Carcinogenesis: Critical Appraisal. WHO, Geneva, Suppl. 2. pp. 321.
A
3 I l-
SINHA, D. K., AND DAO, T. L. (1980). Induction of mammary tumors in aging rats by 7,12-dimethylbenz(a)anthracene: Role of DNA synthesis during carcinogenesis. J. Nat. Cancer Inst. 64, 519-521. SWENBERG, J. A., KERNS, W. D., MITCHELL, R. I., GRALLA, E. J., AND PAVKOV, K. L. (1980). Induction of squamous cell carcinoma of the rat nasal cavity by inhalation exposure to formaldehyde vapor. Cancer Rex 40, 3398-3402.
Cox, D. R. (1972). Regression models and life tables. J. R. Stat. Sot. 34, 187-220. GROTH, D. H., COATE, W. B., ULLAND, B. M., AND HORNUNG, R. W. (1981). Effects of aging on the in-
WARD, J. M., GOODMAN, D. G., SQUIRES,R. A., CHU, K. C., AND LINHART, M. S. (1979). Neoplastic and nonneoplastic lesions in aging (C57BL/6NX C3H/Hen) F, (B6C3Fl) mice. J. Nat. Cancer Insf. 63, 849-854.