Verification of a false positive in a two-year rat carcinogenicity study using dual control groups

There is sometimes controversy over whether or not statistically significant responses produced in carcinogenicity studies have biologically significance. Ambiguous results from our previous two-year oral carcinogenicity study on acotiamide hydrochloride hydrate (acotiamide-HH), a prokinetic drug for functional dyspepsia, in rats made it unclear whether the drug may exhibit uterine carcinogenicity. To check this finding, we performed a second long-term carcinogenicity study using two identical control groups to more accurately evaluate uterine carcinogenesis by considering the incidence of spontaneous neoplasms. Female Fischer 344 rats were divided into three groups: the two control groups (control 1 and 2) were administered vehicle (0.5% w/v methylcellulose) and the acotiamide-HH-treated group was administered 2,000 mg/kg/day of acotiamide-HH by oral gavage for two years. Among all groups, the incidence of endometrial adenocarcinoma (EmA) was highest in the control 2 group, followed by the acotiamide-HH-treated group and the control 1 group. Moreover, acotiamide-HH did not affect the incidence of precursor lesions of EmA. In cases where an ambiguous difference is observed, the use of two control groups allows for a more informed interpretation of the findings in the drug-treated groups. The outcomes in this study strongly support the hypothesis that the increase in EmA in rats treated with acotiamide-HH in our previous study is unrelated to administration of the drug.

Safety of betaine as a novel food pursuant to Regulation (EC) No 258/97

Following a request from the European Commission, the EFSA Panel on Dietetic Products, Nutrition and Allergies (NDA) was asked to deliver an opinion on betaine as a novel food (NF) pursuant to Regulation (EC) No 258/97. The information provided on the composition, the specifications, the batch-to-batch variability, stability and production process of the NF is sufficient and does not raise concerns about the safety of the NF. The NF is proposed to be used in foods intended to meet additional requirements for intense muscular effort with a maximum intake of 2.5 g/day of betaine for sports people above 10 years of age. Based on the lowest BMDL₀₅, which was derived from a chronic toxicity study in rats in which a dose-related increase in platelet counts was observed, and the anticipated daily intake of the NF in the target population, the Margins of Exposure are 3.6 and 5, which are generally regarded as not sufficient. However, the total exposure to betaine from the diet (about 830 mg/day) is not known to be associated with adverse effects. Moreover, no adverse effects on platelet counts were noted in human intervention studies with exposure levels of 4 g/day of betaine for up to 6 months. A significant increase in total and low-density lipoprotein (LDL)-cholesterol concentrations was noted at intakes of 4 g/day of betaine in overweight subjects with metabolic syndrome but not in healthy subjects, nor at intakes of 3 g/day. Thus, considering 4 g/day of betaine as a reference point and applying an uncertainty factor of 10 to account for interindividual variability, an amount of 400 mg/day of betaine (i.e. 6 mg/kg body weight (bw) per day for adults) in addition to the background exposure is considered as safe. The Panel considers that the NF is safe to be used at maximum intake of 6 mg/kg bw per day in the target population.

Survival trial design and monitoring using historical controls

In this paper, we propose a multistage group sequential procedure to design survival trials using historical controls. The formula for the number of events required for historical control trial designs is derived. Furthermore, a transformed information time is proposed for trial monitoring. An example is given to illustrate the application of the proposed methods to survival trial designs using historical controls. Copyright © 2016 John Wiley & Sons, Ltd.

Hormone Receptor Expression in Spontaneous Uterine Adenocarcinoma in Fischer 344 Rats

Most uterine cancers, the most common gynecological malignancies in women in developed countries, are hormone-dependent endometrial adenocarcinomas (EACs) that express estrogen and progesterone receptors. Although rat strains exist with a high spontaneous incidence of EAC, the Fischer 344 (F344) strain, previously one of the most commonly used strains in carcinogenicity testing, is not a high-incidence strain. To better understand the biology of this neoplasm, we assessed estrogen receptor α (ER), progesterone receptor (PR), and Ki-67 expression using immunohistochemistry in spontaneous EAC in 18 F344 rats used as control animals in 2-year National Toxicology Program bioassays. Of the 18 tumors, 9 were well-differentiated and 9 were poorly differentiated. Most tumors, 7/18, were ER+PR+, as observed in women. Of the remainder, 6/18 were ER+PR-, 2/18 were ER-PR+, and 3/18 were ER-PR-. Well-differentiated tumors were ER+ (8/9) more often than poorly differentiated tumors (5/9). The percentage of ER+ tumors (72%) in rats was similar to that seen in women, but rats less frequently had PR+ (50%) tumors than women. The heterogeneous estrogen and progesterone receptor immunophenotypes observed in F344 rats in this study highlight the importance of evaluating hormone receptor expression in animal models used for chemical evaluations.

[Background data of spontaneous tumors in F344/DuCrlCrlj rats]

INTRODUCTION

We investigated the 2-year survival rate and incidence of spontaneous tumors in F344/DuCrlCrlj rats used in carcinogenicity studies of chemical substances. Records for animals used in the control groups of carcinogenicity studies which were conducted during the last 10 years were obtained from the database of the Japan Bioassay Research Center (JBRC). Six hundred ninety-nine males and 550 females were used in 14 and 11 inhalation studies, respectively, and 500 animals of each sex were used in 10 male and 10 female oral studies.

METHODS

In each study, SPF (specific pathogen free) animals were housed for 2 years (104 weeks) as control groups in the carcinogenicity studies. All animals underwent necropsy and histopathological examination. Each study was conducted in accordance with the Good Laboratory Practice.

RESULTS

The incidence of interstitial cell tumors was highest in both inhalation studies and oral studies (inhalation studies 86.1%, oral studies 68.6%). Tumors which had an incidence of 6% or higher were adenoma of the pituitary, C-cell adenoma of the thyroid, and mononuclear cell leukemia (LGL leukemia) of the spleen in male and female rats; fibroma of the subcutaneous tissue, adrenal pheochromocytoma, and islet cell adenoma of the pancreas in male rats; and endometrial stromal polyps and fibroadenoma of the mammary gland in female rats. Tumors other than the above had rare incidence rates. A clear difference in the incidence of spontaneous tumors was not observed between the inhalation and oral studies. The incidences of spontaneous tumors in control groups of previous oral studies are similar to our findings. There are no other reports of the spontaneous tumor incidence in the control groups of inhalation studies using F344/DuCrlCrlj rats. The 2-year survival rate was about 77% in both the inhalation and oral studies, and a gender difference was not observed. The F344/DuCrlCrlj rats used at JBRC had a higher 2-year survival rate than F344/N rats. This difference is possibly due to the low incidence of LGL leukemia in the F344/DuCrlCrlj rat.

CONCLUSIONS

The incidences of spontaneous tumors in F344/DuCrlCrlj rats used in control groups of both inhalation and oral studies during the last 10 years at JBRC are similar to each other and similar to those reported in other studies. This is the first report on the incidence of spontaneous tumors in inhalation studies and contributes to the toxicological evaluation of studies using F344/DuCrlCrlj rats.

Occurrence of Spontaneous Tumors in the Central Nervous System (CNS) of F344 and SD Rats

In order to accurately assess the carcinogenicity of chemicals with regard to rare tumors such as rat CNS tumors, sufficient information about spontaneous tumors are very important. This paper presents the data on the type, incidence and detected age of CNS tumors in F344/DuCrlCrlj (a total of 1363 males and 1363 females) and Crl:CD(SD) rats (a total of 1650 males and 1705 females) collected from in-house background data-collection studies and control groups of carcinogenicity studies at our laboratory, together with those previously reported in F344 and SD rats. The present data on F344/DuCrlCrlj rats (F344 rats) and Crl:CD(SD) rats (SD rats) clarified the following. (1) The incidences of all CNS tumors observed in F344 rats were less than 1%. (2) The incidences of malignant astrocytoma and granular cell tumor were higher in male SD rats than in female SD rats. (3) The incidences of astrocytoma and granular cell tumor were higher in SD rats than in F344 rats. (4) Among astrocytoma, oligodendroglioma and granular cell tumor, oligodendroglioma was detected at the youngest age, followed by astrocytoma, and ultimately, granular cell tumor developed in both strains. The incidences observed in our study were almost consistent with those previously reported in F344 and SD rats.

Transition of historical control data for high incidence tumors in f344 rats

Historical control data of tumor incidence were collected from the control groups (215 animals of each sex) in four recent carcinogenicity studies that were started between 2005 to 2009 (terminally sacrificed between 2007 and 2011) at BoZo Research Center Inc. (Gotemba, Shizuoka, Japan) using Fischer 344 rats (F344/DuCrlCrlj). These data were compared to the previous historical control data (from 1990 to 2004, previously reported) in the same facility. In the results, the incidence of C-cell adenoma in the thyroid tended to increase in both sexes in recent years (30.8% for males and 24.4% for females in 2005-2009) as compared with the previous data (17.4% and 20.1% for males and 11.5% and 11.8% for females in 1990–1999 and 2000–2004, respectively). In addition, the incidences of pancreatic islet cell adenoma in males and uterine adenocarcinoma tended to increase from around 2000 and remained high in recent years (incidences of islet cell adenoma in males of 10.5%, 17.1% and 20.5% in 1990–1999, 2000–2004 and 2005–2009; incidences of uterine adenocarcinoma of 3.3%, 12.0% and 13.5% in 1990–1999, 2000–2004 and 2005–2009, respectively). There was no apparent difference in the incidence of other tumors.

Calculating sample size in trials using historical controls

BACKGROUND

Makuch and Simon [Sample size considerations for non-randomised comparative studies. J Chronic Dis 1980; 33: 175-81.] developed a sample size formula for historical control trials. When assessing power, they assumed the true control treatment effect to be equal to the observed effect from the historical control group. Many researchers have pointed out that the Makuch-Simon approach does not preserve the nominal power and type I error when considering the uncertainty in the true historical control treatment effect.

PURPOSE

To develop a sample size formula that properly accounts for the underlying randomness in the observations from the historical control group.

METHODS

We reveal the extremely skewed nature in the distributions of power and type I error, obtained over all the random realizations of the historical control data. The skewness motivates us to derive a sample size formula that controls the percentiles, instead of the means, of the power and type I error.

RESULTS

A closed-form sample size formula is developed to control arbitrary percentiles of power and type I error for historical control trials. A simulation study further demonstrates that this approach preserves the operational characteristics in a more realistic scenario where the population variances are unknown and replaced by sample variances.

LIMITATIONS

The closed-form sample size formula is derived for continuous outcomes. The formula is more complicated for binary or survival time outcomes.

CONCLUSIONS

We have derived a closed-form sample size formula that controls the percentiles instead of means of power and type I error in historical control trials, which have extremely skewed distributions over all the possible realizations of historical control data.