Regulatory Forum Commentary* Counterpoint: Dose Selection for Tg.rasH2 Mouse Carcinogenicity Studies

Toxicologic Pathology, 43: 621-627, 2015 Copyright # 2015 by The Author(s) ISSN: 0192-6233 print / 1533-1601 online DOI: 10.1177/0192623315587722 Regulatory Forum Commentary* Counterpoint: Dose Selection for Tg.rasH2 Mouse Carcinogenicity Studies JARIG DARBES 1,FRANK D. SISTARE 2, AND JOSEPH J. DEGEORGE 2 1 Laboratoires Merck Sharp and Dohme Chibret, Riom, France 2 Merck and Co., Inc., West Point, Pennsylvania, USA ABSTRACT High-dose selection for 6-month carcinogenicity studies of pharmaceutical candidates in Tg.rasH2-transgenic mice currently primarily relies on (1) estimation of a maximum tolerated dose (MTD) from the results of a 1-month range-finding study, (2) determination of the maximum dose administrable to the animals (maximum feasible dose [MFD]), (3) demonstration of a plateau in systemic exposure, and (4) use of a limit dose of 1,500 mg/kg/day for products with human daily not exceeding 500 mg. Eleven 6-month Tg.rasH2 carcinogenicity studies and their corresponding 1-month range-finding studies conducted at Merck were reviewed. High were set by estimation of the MTD in 6, by plateau of exposure in 3, and by MFD in 2 cases. For 4 of 6 studies where MTD was used for high-dose selection, the 1-month study accurately predicted the 6-month study tolerability whereas in the remaining 2 studies the high showed poorer tolerability than expected. The use of 3 or more drugtreated dose levels proved useful to ensure that a study would successfully and unambiguously demonstrate that a drug candidate was adequately evaluated for carcinogenicity at a minimally toxic high dose level, especially when the high dose may be found to exceed the MTD. Keywords: Carcinogenicity; mouse; preclinical safety-assessment/risk management; Tg.rasH2; maximum tolerated dose. INTRODUCTION Paranjpe et al. (2014) have published a summary of their experience with 29 six-month carcinogenicity studies conducted with Tg.rasH2 mice at BioReliance between 4 and 2013. The authors, in sharing their extensive and valuable experience, seek to propose further optimization of Tg.rasH2 mouse dose rangefinding study and 6-month study design, conduct, and interpretations. They conclude from their retrospective analyses that high-dose groups exceeded the maximum tolerated dose (MTD) in 22 studies for male mice and 13 studies for female mice based on body weight gain (BWG), mortality, and tumor incidence. *This is an opinion article submitted to the Regulatory Forum and does not constitute an official position of the Society of Toxicologic Pathology or the journal Toxicologic Pathology. The views expressed in this article are those of the authors and do not necessarily represent the policies, positions, or opinions of their respective agencies and organizations. The Regulatory Forum is designed to stimulate broad discussion of topics relevant to regulatory issues in toxicologic pathology. Readers of Toxicologic Pathology are encouraged to send their thoughts on these articles or ideas for new topics to regulatoryforum@toxpath.org. The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article. The author(s) received no financial support for the research, authorship, and/or publication of this article. Address correspondence to: Jarig Darbes, Laboratoires Merck Sharp and Dohme Chibret, Centre de Recherche de Mirabel, Route de Marsat Riom, 63963 Clermont-Ferrand Cedex 9, France; e-mail: jarig_darbes@merck.com. Abbreviations: BW, body weight; BWG, body weight gain; CAC, Carcinogenicity Assessment Committee; CDER, Center for Drug Evaluation and Research; FDA, United States Food and Drug Administration; HRAS, Harvey rat sarcoma viral oncogene homolog; ICH, International Conference on Harmonization of Technical Requirements for Registration of Pharmaceuticals for Human Use; MFD, maximum feasible dose; MTD, maximum tolerated dose; TK, toxicokinetic. However, information in the following areas was absent from their analyses: (1) a study-by-study breakdown of which criteria were specifically associated with the conclusion that the high dose was excessive, (2) the results from the dose range-finding studies that were used to select for the 6-month studies, (3) the criteria used by the sponsor for selecting high on a study-by-study basis, and (4) input from dialog between sponsors and regulatory authorities on how for the 6-month study were ultimately chosen. The authors recommend that when a high dose is chosen, investigators first apply the standard criteria based on MTD in the 1-month dose range-finding study, then multiply that estimated dose by 50% for the males and by 67% for the females, and use these lower calculated as the actual 6-month study high. The authors also recommend that when the MTD is estimated based on an effect of the test article on BWG, investigators should use a dose estimate associated with inhibition of 5% BWG rather than inhibition of 10% BWG. Further, the authors recommend that only 2 rather than 3 drug-treated dose groups are needed. The authors of this article have reviewed the results of 22 Tg.rasH2 dose range-finding studies (all conducted internally at Merck), 11 of which are associated with matching subsequent 6-mo carcinogenicity studies. Reflections on the conclusions of this analysis are presented, compared, and discussed in regard to the recommendations of Paranjpe et al. METHODS Animals Wild type (wt/wt) or hemizygous (tg/wt) Harvey rat sarcoma viral oncogene homolog (HRAS)-transgenic mice 621

622 DARBES ET AL. TOXICOLOGIC PATHOLOGY (CByB6F1-Tg(HRAS)2Jic) were obtained from CLEA Japan, Inc., Tokyo, Japan, or from Taconic Farms, Inc., Germantown, New York. Mice used in the range-finding and toxicokinetic (TK) studies were wt/wt and approximately 6 to 9 weeks old at study start. Mice used in the 6-month carcinogenicity studies were hemizygous and approximately 8 to 9 weeks old at study start. The animals were group housed in plastic boxes with contact bedding under environmentally controlled conditions with a 12-hr light/dark cycle. The institutional Animal Care and Use Committee at Merck Research Laboratories, West Point, Pennsylvania, or the Committee for Ethics and Animal Welfare at Merck Research Laboratories, Mirabel France, approved all procedures in these studies. General Dose Selection Criteria for the 6-month Carcinogenicity Studies The highest for 11 six-month carcinogenicity studies were based on the results from the 1-month range-finding studies and the decisions made using one of the following criteria detailed in the International Conference on Harmonization of Technical Requirements for Registration of Pharmaceuticals for Human Use (ICH) S1C(R2) guidance (ICH Guidance for Industry: S1C(R2) 8): Demonstration of an MFD limited by dosing volume, physical, and chemical characteristics of the dose form. TK data showing that a further increase of dose would not result in meaningful increases in exposure (plateau of exposure). Estimation of a maximum tolerated dose for a 6- month study not resulting in increased mortality and more specifically in the early loss of the dose group. The data from the 1-month range-finding study evaluated for this purpose typically are the following, listed in order of decreasing importance: mortality, Body weight (BW) losses, Decreased BWGs (generally 10 20% is considered necessary, but also taking into account transient changes at study start), Demonstration of significant target organ toxicity based on clinical pathology and/or target organ histopathology, Clinical signs (significant clinical signs being typically seen only in the presence of the above-mentioned effects). This ranking is not to be considered absolute, as the criteria are used in combination and, as in specific cases, lowerranked criteria could prevail over the absence of mortality or BW effects (e.g., notable neurologic clinical signs, increasing with time). The limit dose rationale (high dose of 1,500 mg/kg/day for compounds not exceeding a daily intake of 500 mg/day in humans) was not used in any of the studies presented here but is also considered acceptable for high-dose selection in a transgenic mouse carcinogenicity study. In general, when the lowest tested dose at which the MTD is exceeded based on the above criteria, one-third of this dose is generally used to predict the maximally tolerated high dose for the 6-month study. There may be mitigating factors to select two-thirds or half, rather than one-third of this dose such as when there is high TK variability, where exposure is very nonlinear, exposure margins are low or uncertain at the time of study conduct, or there are concerns regarding adequate metabolite coverage, and so on. RESULTS An overview of study results from 1-month range-finding and matching 6-month Tg.rasH2 carcinogenicity studies pertaining to the assessment of dose selection is presented in Table 1. There are 4 (compounds A, B, C, and D) out of 6 cases where the MTD based on tolerability in the 1-month dose rangefinding study predicted well the resulting lack of effect on BW or mortality for the selected for the 6-month study, including the poor tolerability of the initial high dose in the study with compound B. In addition, in agreement with Paranjpe et al., there were 2 (compounds E and F) out of 6 cases where the MTD based on tolerability in the 1-month dose range-finding study did not predict the unexpected mortality or BW effects seen in the selected high dose of the subsequent 6-month study. For one (study with compound K) of the other 5 studies (with compounds G, H, I, J, and K) where the high dose was selected based on MFD or saturation of exposure, an impact on BWG of at least approximately 20% was observed in the 6-month study while no such effects were seen in the dose range-finding study. For compound K, BW effects, however, were not unexpected based on knowledge of pharmacologic target, were not deemed a tolerability issue, and for such a drug class normal weight reduction limits are routinely expected to be exceeded as part of testing requirements. Nevertheless, the 1-month study overall accurately predicted high with no unexpected poor tolerability or BW effects observed on 8 of 11 studies. In 3 of 11 studies, tolerability or BW effects were observed at high of the 6-month study despite expectations that minimal effects would be seen based on the 1-month study results. For carcinogenicity studies where a clear increase in mortality in both genders in the corresponding range-finding studies had been the only critical driver for high-dose selection (3 studies, compounds A, B, and E), the definitive high chosen for the carcinogenicity studies were approximately one-third of the lethal dose observed in the corresponding range-finding study, except for compound B where for reasons beyond the scope of this discussion, a decision was made to push the high dose to >50% of the lethal dose from the range-finding study. For compound B, the dose level chosen at risk (350 mg/kg/day) was indeed subsequently and not surprisingly found to be

article 1-M study (mg/kg/day a ) A 75 500 1,000 B 25 100 500! 250 1000 C 75 150 450 900 D 2 10 2 50 2 150 2 E 50 750 2,500 F 10 30 100 500 2,250 TABLE 1. Dose selection: Key findings from 1-M studies and subsequent 6-M studies. 1-M study results and dose selection for 6-M studies Results of High-dose rationale for dose proposed by sponsor Definitive b,c Dose adjustment during study deaths Morbidity/mortality or negative BW effects tumors Value of 1-M study to inform dose selection MTD: mortality at 1,000 mg/ kg/day 75 500 50 (0.4) 150 (1) (2) MTD: mortality at 500 mg/ kg/day 75 125 250 100 d (1) d (1 to 2) 350!250 d 350! 250 At 350/250 mg/kg/day; dose terminated early in study week 9 Unspecific clinical signs in moribund animals at 350/250 mg/kg/day only (earlyterminated dose); no morbidity or BW effects at other No Poor tolerability of high dose predicted; tolerability of other predicted well MTD: Mortality at 900 and in males no BW gain and minimal mortality at 450 mg/kg/day MTD: Based on target organ toxicity at 2 mg/kg/day 75 150 75 (0.3/0.6) e 150 (0.5/2) e (7/7) e NA 2 50 2 150 None None At mg/kg/day: #BWG ( 80%/ 61%) e ; final BW within 10% of vehicle control No BWG effects at high dose not expected tolerability of other predicted well MTD Based on mortality at 2500 mg/kg/day 50 750 75 (1 [5]) f 250 (2 [13]) f 750 (3 [27]) f None At 750 mg/kg/day 48% mortality in females At 750 mg/kg/day: Unspecific clinical signs in moribund animals. No BWG effects at any dose No Deaths at high dose predicted poorly; tolerability of other groups predicted well MTD Based on Mortality at 2,250 mg/kg/day and BW losses associated with stress-related postmortem findings at 500 mg/kg/day 10 40 250 10 (2/5) e 30 (7/18) e 100 (50/99) e,g 250! 150 250!150 At 250/150 mg/kg/day dose terminated early in study week 10 250/150 mg/kg/day: #BW in both genders; unspecific clinical signs in moribund animals and significant cardiopulmonary postmortem findings 100 mg/kg/day: #BWG ( 46%/ 46%) e, 30 mg/kg/day: #BWG ( 22%/ 21%) e ; final BW at 30 and 100 mg/kg/day within 10% of control No Mortality of highdose and notable BWG effects at mid-and low predicted poorly; decreased BWGs at mid- and low unexpected, but final BW acceptably within 10% relative to control (continued)

TABLE 1. (continued) 1-M study results and dose selection for 6-M studies Results of article 1-M study (mg/kg/day a ) High-dose rationale for dose proposed by sponsor Definitive b,c Dose adjustment during study deaths Morbidity/mortality or negative BW effects tumors Value of 1-M study to inform dose selection G 40 160 380 600 PL 20 80 380 20 (0.3/0.6) e 80 (0.6/1) e 380 (3/1) e H 30 100 450 PL 10 60 10 (6) 60 (16) (36) I 400 800 1, PL NA 400 800 1, J 100 h h 600 h 1,344 h MFD 25 50 125 650 25 (2) 50 (8) (41) 650 (88) K 17 70 356 MFD 100 400 100 (3) (6) 400 (16) None None Pharmacologic #BWG: 100 mg/kg/day: 33%/ 44% e mg/kg/day: 26%/ 22% e NE Predicted well 400 mg/kg/day: 50%/ 25% e Final BW of all groups within 10% relative to vehicle control Note: 1-M ¼ 1-month range-finding; 6-M ¼ 6-month carcinogenicity; #BWG ¼ decreased body weight gain relative to controls; #BW ¼ body weight losses relative to pretest; BW ¼ body weight; BWG ¼ body weight gain; FDA ¼ United States Food and Drug Administration; MFD ¼ maximum feasible dose; MTD ¼ maximum tolerated dose; NA ¼ not applicable (no request for special protocol assessment submitted); NE ¼ Drug treated animals from final necropsy not examined histomorphologically due to discontinuation of the compound; PL ¼ plateau of exposure. a Italicized : dose-limiting toxicity based on mortality or significant toxicity (including negative BW effects). b Consensus after FDA feedback, where applicable and not noted otherwise; in parentheses: multiples of human exposure; additional may have been added by the sponsor. c Italicized : dose group not evaluated for carcinogenicity due to early termination. d Dose level higher than suggested in FDA feedback. e In males/females. f Bracketed values: margins based on metabolite exposure. g Additional dose level added due to Agency s suggestion. h Dosing with preliminary vehicle; the 650-mg/kg/day dose in the final vehicle (used in the 6-month study) resulted in equivalent to slightly higher systemic exposures compared to the 1,344-mg/kg/day dose from the 1-month study.

Vol. 43, No. 5, 2015 TG.RASH2 MOUSE CARCINOGENICITY DOSE SELECTION 625 poorly tolerated, while the dose of mg/kg/day (0.4 of the lethal dose of 500 mg/kg/day in the 1-month study) was well tolerated in the 6-month study. In general and where applicable (where a request for special protocol assessment was submitted), the final dose selection was confirmed or adjusted based on feedback of the U.S. Food and Drug Administration (FDA) s Executive Carcinogenicity Assessment Committee (CAC). It is of note that, where the MTD had been the rationale for high dose selection and where FDA s feedback on the dose selection was requested (studies with compounds A, B, C, E, and F), the Agency s feedback clearly proposed lowering of the high dose in 2 cases and a mitigation strategy for the potential loss of the high dose in 1 case. The Agency was also consistent in its feedback in that it had advocated high of less than or equal to approximately one-third of a lethal dose in the range-finding studies, for the carcinogenicity studies in all 3 cases, even where the sponsor s initial proposal had been higher (compounds A and B). Low tolerability was poorly predicted at the high dose of 750 mg/kg/day with compound E, resulting in poor survival in high-dose females. The reason for the lower tolerability of the compound in females is unknown and was not predicted from the results of the range-finding study. For the study where (marginally) increased mortality and negative BW effects in one gender (the gender with the lower systemic exposure) in the range-finding study were the criteria for the high-dose selection in the carcinogenicity study (compound C), the high dose of mg/kg/day chosen (and agreed to by the Agency) was two-thirds of the non-tolerated dose from the 1-month range-finding study of 450 mg/kg/day, and yielded approximately 30% to 50% of the plasma drug exposure seen at 450 mg/kg/day. In the 6-month study, mg/kg/day of compound C still resulted in some degree of reduction in BWG (see subsequently). In the study where negative BW effects and associated stress-related postmortem findings in the range-finding study had been used for the high-dose selection in the carcinogenicity study (compound F), the high dose chosen was half of the dose associated with BW losses in the range-finding study. This dose was not tolerated in the 6-month study where this group was terminated early (see subsequently). It is of note that the Agency concurred with this high-dose selection but also suggested increasing the mid-dose to hedge the study against the potential loss of the high dose. However, the decision was made by the sponsor to add an additional intermediate dose group instead. Additional adjustments of mid and/or low based on Agency feedback were made on several studies (compounds A, E, and F) where MTD had been the rationale for the high-dose selection. Negative BW effects in the range-finding studies had been critical for the high-dose selection in the studies with compounds C and F. For the study with compound F, the unequivocal and significant BW losses in the range-finding study, even not associated with mortality, retrospectively proved to be predictive that anything above approximately one-third of this excessive dose would result not only in significant BW effects but even in mortality, in the 6-month study. Half of the dose (and approximately half of the systemic exposure) associated with BW losses in the range-finding study with compound F resulted in excessive mortality in the corresponding 6-month carcinogenicity study. Additionally, even 30 mg/kg/day, that is, a dose of less than one-tenth of the lowest poorly tolerated dose from the range-finding study (500 mg/kg/day), resulted in notable effects on BWG in the 6-month study. However, despite some effects on BWG, the of 30 and 100 mg/ kg/day in the 6-month study resulted in final BWs that were still no more than 10% below control BWs and considered acceptable. The high dose in the carcinogenicity study with compound C was two-thirds of a dose (and corresponded to approximately half of the systemic exposure) associated with no BWG and a marginal increase in mortality in males in the range-finding study. It resulted in notable decreases in BWG ( 60% to 80%) in the 6-month study, but the decreases in final relative BW were <10% compared to controls and therefore within the limits of what is generally considered acceptable. For compound C, a dose of one-third of the dose of 450 mg/kg/day in the dose range-finding study where significant BW effects were seen would be 150 mg/kg/day. The 150 mg/kg/day in the 1-month study resulted in no effects on BWG and were well tolerated in the 6-month study. The MFD or a plateau of exposure, in the absence of significant toxic effects in the corresponding range-finding study, had been the decisive criterion for the high-dose selection in 5 six-month carcinogenicity studies (compounds G to K). Where a CAC package had been submitted and thus Agency feedback on the dose selection had been requested (compounds G, H, J, and K), there was general agreement between the Agency and the sponsor on the proposed and the rationale for the high-dose selection and the Agency feedback translated in the adjustment (increase) of an intermediate dose in a single study (compound J), only. Not surprisingly, test articles were generally well tolerated at all in these studies and test effects of note were limited to some nondose-related decreases in BWG (approximately 20 40%) at all drug-treated with compound K. These effects, however, were not unexpected based on knowledge of the pharmacologic target, and were not deemed a tolerability issue, as for such a drug class normal weight reduction limits are routinely expected to be potentially exceeded as part of testing requirements. Moreover, the decreases in final BWs compared to concurrent vehicle controls were well below 10%. For these compounds, exposure margins (where applicable; i.e., for compounds G, H, J, and K) achieved at well-tolerated dose levels were approximately 2-, 36-, 88-, and 16-fold. Taken together, the 1-month study overall accurately predicted high with no unexpected poor tolerability or excessive relative BW effects observed on 9 of the 11 studies. In 2 of the 11 studies, poor tolerability was observed at high of the 6-month study, despite expectations that minimal effects would be seen based on the 1-month study results. There were 11 other dose range-finding studies for which no 6-month studies were conducted (data not shown). Based on the results from these studies, the high in the carcinogenicity

626 DARBES ET AL. TOXICOLOGIC PATHOLOGY studies would have been based on MTD in 8 cases, and limit dose in 3 cases. Of the 22 total dose range-finding studies conducted, 14 of them or 63% would have been based on MTD. DISCUSSION There are various reasons why a sponsor may want to maximize the tested in a Tg.rasH2 mouse carcinogenicity study. One in particular is that, in contrast to the classic 2-year rodent studies, the concept of margins over human exposure as the determining factor for the high-dose selection is not yet accepted by the regulatory agencies. It is reasonable for regulatory agencies to request sponsors to demonstrate that they have made reasonable attempts to reach MTD and to maximize exposure to the test article. There may also be reasons pertaining to the development strategy such as uncertainty about the definitive clinical therapeutic (especially if the Tg.rasH2 mouse carcinogenicity study is conducted early as part of a de-risking strategy). In this context, the most relevant questions raised by Paranjpe et al. seem to pertain more to the estimation and definition of an acceptable MTD in the 6-month Tg.rasH2 mouse study than to the systematic reduction of the estimated MTDs. Our experience suggests that whereas certain paradigms seem to predict MTD in the 6-month carcinogenicity bioassays generally well (e.g., using one-third of a lethal dose from a range-finding study, in the absence of significant toxicity or pronounced BW effects at lower, as the high dose), a residual risk for exaggerated mortality or significant toxicity and therefore for a potential impact still occasionally persists. It is valuable to be aware of this likelihood as a rationale for determining the lower dose levels. The design of the International Life Sciences Institute - Health and Environmental Sciences Institute (ILSI-HESI) validation studies conducted in the late 1990s (Robinson and MacDonald 1) emphasized the need to use maximally tolerated in the transgenic models for positive responses to become evident with known carcinogens, and in fact, some human carcinogens were positive in the model only at the highest tested, some of which had been set at the MTD (Usui et al. 1). The authors feel that mid- and low-dose groups provide assurances that any tumor risk would be identified if the high dose is too aggressive and fails due to potential confounding tolerability effects or excessive mortality. These lower would adequately assess the tumor risk. A 3-dose-level paradigm is likely to shield the compound development from this risk in most cases as our examples show. If the high dose selected from dose range-finding studies is found to exceed the MTD in the 6-month carcinogenicity study, the sponsor and regulatory agencies would have the next without potential confounding effects to assure confidence in study interpretation conclusions. In 5 of the 6 cases where the high dose was selected based on estimation of the MTD, the next 2 were well tolerated without effects on BWG, were generally one-third to one-fifth of the excessively high dose, and clearly served as an unconfounded and certain test of carcinogenicity potential using the Tg.rasH2 mouse model. In all 6 cases, 2 dose levels were provided where relative BWs were within 10% difference to concurrent vehicle controls and well tolerated. Data from Table 1 does suggest that in some cases FDA reviewers are expressing concern for overdosing and provide such feedback by suggesting that high be lowered. Sponsors need to weigh this advice against the risk that results may be challenged at some later point by regulatory authorities if high are not well justified to maximize human exposure margins while appropriately demonstrated to be minimally toxic so as not to confound study interpretation. For 4 (A, B, C, and E) of the 5 compounds where the high dose in the 6-month carcinogenicity studies was set by mortality (alone or in combination with BW effects) in the 1-month range-finding studies, the exposure margins achieved in the 6-month studies were 2-fold, 1- to 2-fold, 0.5-/2-fold (males/females), and 2-fold, respectively, for the highest dose levels where no effects on BWG were seen. If these lower dose levels had been chosen as the high by default, the tumor-negative results of these studies might have been challenged by regulatory authorities since these lower were without notable toxicity and represented narrow margins over anticipated clinical exposure. Regarding Paranjpe et al. s proposal to use 2 dose levels, this recommendation would appear risky for a compound such as compound F. For compound F, a 6-month study was conducted with 4 dose levels because of uncertain tolerability concerns that surfaced in the 1-month dose range finding study. For this compound, a high dose (250 mg/kg/day) was chosen between a welltolerated (100 mg/kg/day) and non-tolerated (500 mg/kg/day) dose level based on the 1-month study findings. In the 6- month study the 250 mg/kg/day dose was clearly lethal in the 6-month study setting and had to be terminated early. The dose of 100 mg/kg/day and even the 30 mg/kg/day dose (groups) were associated with some negative effects on BWG but did not result in final BW decreased by >10% relative to concurrent vehicle controls. If a study is conducted resulting in only a single-dose level without effects on relative BW or morbidity, one may argue that this would still be an adequate assessment of carcinogenic risk but could also be open to challenge. When the spread in a meaningful dose selection is narrow due to low human exposure multiples being achievable, it may seem more reasonable to use only 2 dose levels to conserve on animal use. For compound B (see previously), for example, a dose of one-third of the lethal dose in the range-finding study would yield a dose with an exposure margin of less than 2-fold over the clinical exposure and the actual low dose was already only 1 projected clinical exposure. However, a high-dose selection based on the criteria proposed by Paranjpe may have resulted in studied dose levels at even lower exposure margins. Furthermore, by using a third dose level for compound G, the objective was achieved of definitively demonstrating and removing any doubts that maximal were evaluated and that no carcinogenic risk was inadvertently missed. In our experience, vehicle-treated control Tg.rasH2 transgenic mice show limited BWGs over the course of a 6-mo study (generally <5 g for final BWs of typically approximately

Vol. 43, No. 5, 2015 TG.RASH2 MOUSE CARCINOGENICITY DOSE SELECTION 627 30 g in males and 23 g in females) and BWGs tend to vary between studies. As a result, what appears as a huge BWG reduction of 50% to 60%, for example (approximately 2.5 to 3 g), when expressed as percentage BWG often results in relatively small differences in final BW of approximately 10% or less. Furthermore, if a dose resulting in a decrement of 5% BWG (approximately 250 mg) were to be invoked, the relative final BW would be expected to result within 1% of control. Definition of MTD (and more importantly, decisions of exceeding it) based on BW effects may therefore be more confidently concluded when based on percentage differences in final BW than percentage BWG. We have observed differences in percentage BWG that appear to exceed 10% between 2 similarly treated vehicle control groups of the same sex in the same study (data not shown). Our experience also indicates (data not shown) that there is too much spontaneous variability in BWG over the course of a 1-month range-finding study in Tg.rasH2 wt/wt mice to allow for a reliable prediction of an MTD in a carcinogenicity study based on a 5% decrease in BWG. We advocate using clear BWG effects (at least 10%) in conjunction with other relevant data (mortality at higher ; other significant toxic effects) to estimate MTD in the completed 6-month carcinogenicity studies. A 10% lower relative BW observed in the high-dose group at the end of a 6-month transgenic mouse study, or at the end of a 2-year rat or mouse carcinogenicity study, is viewed as acceptable and is not considered to have altered drug-induced tumor patterns (personal communication from Dr. Abigail Jacobs, FDA Center for Drug Evaluation and Research [CDER]). On the other hand and as shown in one example presented here, clear BW losses in 1-month rangefinding studies should be considered as almost predictive as mortality for the determination of the MTD in the corresponding 6-month carcinogenicity study. In summary, while we recognize the value of the considerations raised by Paranjpe et al., we conclude based on the authors experience that traditional dose selection criteria applied to 2-year bioassays as outlined in ICH S1 guidance are reasonably extrapolated to 6-month Tg.rasH2 transgenic mouse assay. We would like to see all the dose selection options specified in the ICH S1C guidance (ICH Guidance for Industry: S1C[R2] 8) to be made available, including an exposurebased approach but as yet this is not the case. AUTHOR CONTRIBUTION Authors contributed to conception or design (JD, FS, JJD); data acquisition, analysis, or interpretation (JD, FS); drafting the manuscript (JD); and critically revising the manuscript (FS, JJD). All authors gave final approval and agreed to be accountable for all aspects of work in ensuring that questions relating to the accuracy or integrity of any part of the work are appropriately investigated and resolved. REFERENCES ICH Guidance for Industry: S1C(R2) Dose Selection for Carcinogenicity Studies. (8). U.S. Department of Health and Human Services, Food and Drug Administration, Center for Drug Evaluation and Research, Center for Biologics Evaluation and Research. Retrieved April 17th 2015 from the U.S. Food and Drug Administration Website: http://www.fda. gov/ucm/groups/fdagov-public/@fdagov-drugs-gen/documents/documen t/ucm074919.pdf. Paranjpe,M.G.,Denton,M.D.,Vidmar,T.J.,andElbekai,R.H.(2014). Regulatory forum opinion piece*: Retrospective evaluation of in the 26-week Tg.rasH2 mice carcinogenicity studies: Recommendation to eliminate high at Maximum Tolerated Dose (MTD) in future studies. Toxicol Pathol. doi:10.1177/0192623314557526. Robinson, D. E., and MacDonald, J. S. (1). Background and framework for ILSI s collaborative evaluation program on alternative models for carcinogenicity assessment. International Life Sciences Institute. Toxicol Pathol 29, 13 19. Usui, T., Mutai, M., Hisada, S., Takoaka, M., Soper, K. A., McCullough, B., and Alden, C. (1). CB6F1-rasH2 mouse: Overview of available data. Toxicol Pathol 29, 90 108. For reprints and permissions queries, please visit SAGE s Web site at http://www.sagepub.com/journalspermissions.nav.