Abstract 1617: Analysis of single mouse tumor response results from the Pediatric Preclinical Testing Program (PPTP)

Autor: Peter J. Houghton, Brendan Murphy, Igor Dvorchik, Richard Gorlick, Anders Kolb, Patrick Reynolds, Han Yin, Jianrong Wu, Raushan T. Kurmasheva, John M. Maris, Malcolm A. Smith, Catherine A. Billups, Stephen T. Keir, Min H. Kang
Rok vydání: 2015
Předmět:
Zdroj: Cancer Research. 75:1617-1617
ISSN: 1538-7445
0008-5472
Popis: Traditional approaches to evaluating antitumor agents using human tumor xenograft models have usually used 8-10 mice in control and treatment groups against a limited panel of tumor models. However, with increasing evidence of molecular/genetic subsets of tumors within a histotype, there is a need to establish panels of tumors that accurately reflect such heterogeneity. The consequences of evaluating agents in a large number of models is that it is highly resource consuming. An alternative approach is to use fewer animals per tumor line, allowing a greater number of models to be evaluated that capture greater molecular/genetic heterogeneity of the cancer type. We analyzed 67 agents evaluated by the PPTP to determine whether a single mouse, chosen by a random number generating routine in Excel, predicted the median response for groups of mice (solid tumors n = 10; acute lymphoblastic leukemias n = 8). Agents evaluated included standard cytotoxic agents, tyrosine kinase inhibitors, receptor-binding antibodies, and anti-angiogenic agents. Data were analyzed to determine predictive value for 79 xenograft models, and to identify classes of antitumor agent where predictive value of single mouse experiments was poor. The individual tumor response was compared to the group median response using response criteria developed in the PPTP. A total of 2106 comparisons were made. The single tumor response accurately predicted the group median response in 1684 comparisons (79.34%). Models had a range for accurate prediction (0.583 - 0.918). Allowing for mis-prediction of + or - one response category, the overall prediction rate increased to 95.63%. Single tumor results accurately predicted objective response rates (ORR) determined by group response, in 66 of 67 studies. There was a single study where single mouse data incorrectly predicted an agent as ‘active’ (>20% ORR). For most tumor models the single mouse appears to give results similar to groups of 8-10 mice. Importantly, even allowing for slight inaccuracy (± 1 response category), the single mouse experiment identified all active agents, and in the worse case over-predicted activity in one tumor histotype. The advantage of the single mouse experimental design is that it allows for inclusion of models that encompass greater molecular heterogeneity, thus recapitulating the clinical heterogeneity more accurately. Potentially, this design also reduces the numbers of mice used for experimentation. However, to generate these patient derived xenograft models, particularly solid tumors at diagnosis and relapse will require national and international coordination through organizations such as the Children's Oncology Group. Citation Format: Brendan Murphy, Han Yin, John M. Maris, Anders E. Kolb, Richard Gorlick, Patrick C. Reynolds, Min Kang, Stephen T. Keir, Raushan Kurmasheva, Igor Dvorchik, Jianrong Wu, Catherine Billups, Malcolm A. Smith, Peter J. Houghton. Analysis of single mouse tumor response results from the Pediatric Preclinical Testing Program (PPTP). [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 1617. doi:10.1158/1538-7445.AM2015-1617
Databáze: OpenAIRE
Externí odkaz: