Tumor incidence rates

The statistic X2/T will be large when there is evidence of a dose-related increase or decrease in the tumor incidence rates, and small when there is little difference in the tumor incidence between groups or when group differences are not dose related. Under the null hypothesis of no differences between groups, X2lT has approximately a chi-squared distribution with one degree of freedom. 

Table 26.1 Mammary tumor incidence rates in five 2-year dosing studies with atrazine in female Sprague-Dawley ratsa...

The studies of Nishimoto et al. (1988), Yamada (1974) and Inada et al., (1978) provide strong evidence for a causal link between chemical agent exposure and cancer however, because the workers were exposed to multiple chemicals, it is not possible to state conclusively that the cancers were due solely to sulfur mustard. Furthermore, it should be noted that several possible confounding factors, such as tobacco smoking habits, preexisting health conditions, and post-exposure occupational histories of the workers, were not evaluated. In addition, SMRs themselves may not provide an accurate estimate of relative cancer risk if they do not correlate with tumor incidence rates in exposed and control groups (i.e., if social/economic or other differences between control and exposed groups result in differences in health care which affect survival rates). 

In the toxicity study only those animals tested and observed long enough to exceed the demonstrated minimum latency period for first tumor appearance (70 days post-exposure, see Table 4) were included in the data analysis. The resulting tumor incidence rates are shown in Table 7. 

Statistical analysis is performed on all parameters in the study. Its most fundamental objective is to determine whether administration of the test agent results in an increase in tumor incidence rates as compared to those in unexposed controls. Various statistical methods can be used. Tests for increased tumor occurrence rates between dosages may be based on pair-wise comparisons, such as the Chi-square test for 2x2 tables, the Fisher s exact test, or the Cochan-Armitage test. These tests are most appropriate when survival rates do not differ appreciably in the various dose groups. 

Portier CJ, Kopp-Schneider A, Sherman CD. Calculating tumor incidence rates in stochastic models of carcinogenesis. Math Biosci 1996 135 129 16. 

J.G. Gurney, et al.. The contribution of nonmalignant tumors to CNS tumor incidence rates among children in the United States, Cancer Causes Control 10 (2) (1999) 101-105. 

Reliable correlations were found between lead concentrations in the soil and the incidence rates in the population of Dnepropetrovsk. In the adults, the pollution levels were associated with complications in pregnancy and delivery, diseases of urogenital, nervous and bone-and-muscular system, blood problems and tumors. In the children, the blood, blood forming organs, respiratory system and the muscles and bones were most frequently affected. 

The single most important statistical consideration in the design of bioassays in the past was based on the point of view that what was being observed and evaluated was a simple quantal response (cancer occurred or it didn t), and that a sufficient number of animals needed to be used to have reasonable expectations of detecting such an effect. Though the single fact of whether or not the simple incidence of neoplastic tumors is increased due to an agent of concern is of interest, a much more complex model must now be considered. The time-to-tumor, patterns of tumor incidence, effects on survival rate, and age of first tumor all must now be captured in a bioassay and included in an evaluation of the relevant risk to humans. 

Another approach to controlling the false positive rate in carcinogenicity studies was proposed by Haseman (1983). Under this rule, a compound would be declared a carcinogen if it produced an increase significant at the 1% level in a common tumor or an increase significant at the 5% level in a rear tumor. A rare neoplasm was defined as a neoplasm that occurred with a frequency of less than 1% in control animals. The overall false positive rate associated with this decision rule was found to be not more that 7-8%, based on control tumor incidences from NTP studies in rats and mice. This false positive rate compares favorably with the expected rate of 5%, which is the probability at which one would erroneously... 

IDEC was interested in NHL, in 1990, because NHL, specihcally the low-grade or follicular type of the disease, remains an incurable tumor for the vast majority of pahents. It has a large and expanding patient base. NHL is the hfth most-frequently diagnosed cancer in the United States, the sixth leading cause of cancer death, and the second fastest-growing cancer (in incident rate). NHL mainly afflicts people over 50 years of age. As with most scien-hfic research, however, there was a fair bit of serendipity involved in our choice of target. 

Osborne-Mendel rats were fed technical-grade heptachlor (73%) males received TWA doses of 1.94 and 3.9 mg/kg/day and females received TWA doses of 1.28 and 2.56 mg/kg/day for 80 weeks (NCI 1977). The results of this study showed a statistically significant increase in follicular cell neoplasms in the thyroid (adenomas and carcinomas) in females fed the high dose compared to controls. This finding was discounted by the investigators, however, because the incidence rates were low and are known to be variable in the control rat population. Rates of tumor incidences in males were not increased. 

Figure 3 demonstrates the tumor incidence results obtained in these studies when just the three dietary regimens were compared, i.e. LF (2%), HPF (20% stripped corn oil) and HSF (18% stripped corn oil 2% linoleic acid). As can be seen in this figure, the animals consuming the HPF diet routinely develop a 97-100% tumor incidence. Those on the HSF diet develop a 55-70% incidence and those on the LF develop only a 20-30% incidence of mammary tumors. These values represent three sets of experiments with 30 rats per group or 90 total rats per each diet represented. There is a 10-15% variability in the incidence rates between experiments performed under identical conditions whether done concurrently or in a sequence of times. 

Studies for carcinogenicity in rats and mice should last at least 24 months. This rule avoids problems of acceptance at the international health authorities. There are other aspects which can influence the duration of a carcinogenicity study, e.g. survival rate. In case survival after a certain time of treatment is below 25 % health authorities should be informed with the request to sacrifice the remaining animals. In other cases survival will be higher than 75 % after 24 months of treatment. With this result, it can be considered to prolong the exposure period. The duration (and therefore the age of the animals) of a carcinogenicity study has a definite influence on the tumor incidence in the animals. The interpretation of the results with regard to tumor incidences has to be corrected for the age of the test animals. 

The estimated incidence rate of fatal tumors in each time period and the prevalence rate of nonfatal tumors in each time period. 

Dermal exposure to benzene did not induce skin tumors in mice (Bull et al. 1986). No papillomas developed in mice that were given a 2-week, 800 mg/kg/day topical application of benzene as the initiator and a 1 pg topical application of 12-o-tetradecanoylphorbol-13-acetate 3 times a week for 20 weeks and observed for 52 weeks (Bull et al. 1986). The authors concluded that it is difficult to estimate benzene-induced tumor incidence after dermal exposure, and that mouse skin may not be the optimal study system. This is because of the high rate of false-negative responses to chemicals, like benzene, with recognized carcinogenic activity. 

---