Abstract IA29: Mice host selection for patient-derived xenograft (PDX) model development and other critical factors for success

The Jackson Laboratory PDX program engages 26 staff members, 5 PIs, and runs thousands of PDX models per year in pre-clinical experiments both for internal scientific experiments and external projects. Over the last 6 years of operation of the PDX program, we have gained a significant amount of experience in all aspects of the process. This experience includes the genomic analysis of the tumors and the establishment of an extensive database annotating many of the 446 PDX models in our US inventory. Herein, we describe some of the characteristics of the system that enhances successful experimentation in this platform. Several factors significantly improve the engraftment rate of tumors 1 : 1) the degree of immunodeficiency the host mouse - the NOD.Cg- Prkdc scid IL2rg tm1wjl /SzJ (aka, NSG), the NOD.CB17- Prkdc scid /J (aka,NOD- scid ), or the NOD.Cg- Rag1 tm1mom IL2rg tm1wjl /SzJ (aka, NRG) engrafting better than the beige- scid and athymic nude mice; 2) the greater the amount of tissue engrafted, 3) the late or metastatic nature of the tumor, 4) the shorter time from surgery to implantation, 5) the absence of enzymatic dissociation, and 6) orthotopic implantation. Sequencing and expression analysis show the maintenance of the core genomic configuration between PDX and the primary tumors though some genetic differences are noted of indeterminate significance. Human stromal cells, however, tend to be replaced by murine stroma after the first passage. Tumor genetic heterogeneity is maintained through the fourth passage in NSG PDX models as confirmed by deep sequencing of bulk tumor and of individual progenitor clones. Most importantly, however, the tumor responses to systemic therapies appear to reflect the patient response 2 . Therefore, the use of PDX models to test novel agents may speed pre-clinical drug development. The most important use of PDX systems may be in immuno-oncology given that “humanized” PDX models where implantation of a primary human tumor is implanted in a mouse with a reconstituted human cellular immune system provides a powerful preclinical experimental platform to test new immune modulators in human cancers. We, and others, have shown that humanized NSG mice bearing human tumor xenografts exhibit dramatic responses to immune checkpoint inhibitors that are immune cell and drug dependent 3 . Immune reconstitution is more complex because of the requirement of human cytokines that are not substituted by their murine counterparts. Engineered mice expressing human cytokines, e.g., IL3 , CSF2 ( GM-CSF ), and KITLG (stem cell factor) in the NSG background (aka, NSG-SGM3 mice), and those expressing CSF1 (M-CSF), IL3, CSF2 , and THPO in the C;129S4- Rag2 tm1.1Flv IL2rg tm1.1Flv /J background (aka, MITRG mice) after engraftment with human hematopoietic stem cells have been shown to support myeloid cells 4 including macrophages absent in standard NSG 5 . It is anticipated that these “next generation” humanized mice will provide a more nuanced picture of the tumor-immune system interaction 6 . It is important to understand the challenges and limitations of the PDX platform that can be mitigated to a degree by quality control and study design. There are simple caveats. ~5 % of engrafted solid tumors give rise to EBV positive lymphomas and not the primary tumor. Moreover, ~5-10% of PDX tumors are overgrown by a transformed murine cell especially in late passaged. Thus, stringent histological quality control is necessary which includes the assessment of human cytokeratin, which provides an assessment of murine cancer incursion of solid human cancers. Another concern is that drug dosing for PDX experiments is often very different from that used in human studies. The NSG mice are more sensitive to certain DNA damaging agents and to radiation than NRG or Rag1 null mice. Therefore, the structuring of combination studies using genotoxic agents is complicated and should be interpreted with appropriate care. In terms of study design, we have found that each PDX model from an individual patient will give rise to individual tumors in an NSG cohort with significant growth and response variations. Thus, for each treatment arm we have calculated that between 6-8 animals is the minimal number required to attain statistical power of 95-99% to identify efficacy between arms. Moreover, in any treatment arm, it is necessary to assess the response of each individual PDX bearing mouse since a few individual tumors in a cohort may be resistant to a drug whereas the average of the arm shows an overall response. Recently, Gao, et al 7 presented an alternative way to conduct PDX studies for drug development where only one mouse per PDX model was used per drug. The overall data (not whether a drug specifically was efficacious in a specific disease type) provided important strategic information in development. They raised an important point and that is that PDX preclinical studies should be structured differently from classical clinical trials to make best use of the platform. Not only the trial design, but even how to call a response should be reexamined. The partial responses seen in PDX experiments that can be precisely quantified but that do not qualify using RECIST criteria provide potentially important information about drug efficacy. Taken together, the PDX platform using severely immunodeficient mice is a powerful tool that can significantly accelerate the development of new therapeutics by dramatically facilitating the advancement of innovative therapies with a high likelihood for success 8 . References: 1 Shultz LD, Goodwin N, Ishikawa F, Hosur V, Lyons BL, Greiner DL. Human cancer growth and therapy in immunodeficient mouse models. Cold Spring Harbor Protoc. 2014 Jul 1;2014(7):694-708. 2 Garralda E, Paz K, Lopez-Casas PP, Jones S, Katz A, Kann LM, Lopez- Rios F, Sarno F, Al-Shahrour F, Vasquez D, Bruckheimer E, Angiuoli SV, Calles A, Diaz LA, Velculescu VE, Valencia A, Sidransky D, Hidalgo M. Integrated next-generation sequencing and avatar mouse models for personalized cancer treatment. Clin Cancer Res. 2014 May 1;20(9):2476-84 Epub 2014 Mar 14. 3 Wang M, Keck JG, Cheng M, Cai D, Shultz L, Palucka K, Banchereau J, Bult C, Huntress R. Patient-derived tumor xenografts in humanized NSG mice: a model to study immune responses in cancer therapy. [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 LB-050. 4 Billerbeck E, Barry WT, Mu K, Dorner M, Rice CM, Ploss A. Development of human CD4+FoxP3+ regulatory T cells in human stem cell factor-, granulocyte-macrophage colony-stimulating factor-, and interleukin-3- expressing NOD-SCID IL2Rγ(null) humanized mice. Blood. 2011 Mar 17;117(11):3076-86. 5 Rongvaux A, Willinger T, Martinek J, Strowig T, Gearty SV, Teichmann LL, Saito Y, Marches F, Halene S, Palucka AK, Manz MG, Flavell RA. Development and function of human innate immune cells in a humanized mouse model. Nat Biotechnol. 2014 Apr;32(4):364-72. 6 Shultz LD, Brehm MA, Garcia-Martinez JV, Greiner DL. Humanized mice for immune system investigation: progress, promise and challenges. Nat Rev Immunol. 2012 Nov;12(11):786-98. 7 Gao H, Korn JM, Ferretti S, Monahan JE, Wang Y, Singh M, Zhang C, Schnell C, Yang G, Zhang Y, Balbin OA, Barbe S, Cai H, Casey F, Chatterjee S, Chiang DY, Chuai S, Cogan SM, Collins SD, Dammassa E, Ebel N, Embry M, Green J, Kauffmann A, Kowal C, Leary RJ, Lehar J, Liang Y, Loo A, Lorenzana E, Robert McDonald E 3rd, McLaughlin ME, Merkin J, Meyer R, Naylor TL, Patawaran M, Reddy A, Roelli C, Ruddy DA, Salangsang F, Santacroce F, Singh AP, Tang Y, Tinetto W, Tobler S, Velazquez R, Venkatesan K, Von Arx F, Wang HQ, Wang Z, Wiesmann M, Wyss D, Xu F, Bitter H, Atadja P, Lees E, Hofmann F, Li E, Keen N, Cozens R, Jensen MR, Pryer NK, Williams JA, Sellers WR. High- throughput screening using patient-derived tumor xenografts to predict clinical trial drug response. Nat Med. 2015 Nov;21(11):1318-25. 8 Gandara DR, Mack PC, Bult C, Li T, Lara PN Jr, Riess JW, Astrow SH, Gandour-Edwards R, Cooke DT, Yoneda KY, Moore EH, Pan CX, Burich RA, David EA, Keck JG, Airhart S, Goodwin N, de Vere White RW, Liu ET. Bridging tumor genomics to patient outcomes through an integrated patient-derived xenograft platform. Clin Lung Cancer. 2015 May;16(3):165- 72. Citation Format: Edison T. Liu, Carol Bult, Jeff Chuang, Pooja Kumar, Mingshan Cheng, R. Krishna Murthy Karuturi, Vivek Philip, James Keck, Karolina Palucka, Larry Shultz. Mice host selection for patient-derived xenograft (PDX) model development and other critical factors for success. [abstract]. In: Proceedings of the AACR Special Conference: Patient-Derived Cancer Models: Present and Future Applications from Basic Science to the Clinic; Feb 11-14, 2016; New Orleans, LA. Philadelphia (PA): AACR; Clin Cancer Res 2016;22(16_Suppl):Abstract nr IA29.