S. A. Becker1, Y. Zhu1, C. Wang4, A. W. Cross2, E. Curl2, K. E. Armeson5, K. E. Hurst1, D. N. Lewin2, G. Warren6, B. J. Hoffman4, E. G. Hill5, V. J. Findlay2, E. R. Camp1 1Medical University Of South Carolina,Department Of Surgery,Charleston, SC, USA 2Medical University Of South Carolina,Department Of Pathology,Charleston, SC, USA 3Medical University Of South Carolina,Charleston, SC, USA 4Medical University Of South Carolina,Department Of Medicine,Charleston, SC, USA 5Medical University Of South Carolina,Department Of Public Health Sciences,Charleston, SC, USA 6Medical University Of South Carolina,Department Of Radiation Oncology,Charleston, SC, USA
Introduction:
Neoadjuvant 5-fluorouracil chemoradiation (5FU/RT) has established benefits in locally advanced rectal cancer (LARC) patients; however, therapeutic resistance is observed in ~70% of the cases. To advance care, predictors of 5FU/RT response still need to be identified and novel personalized approaches are desperately needed. Our multi-disciplinary research team hypothesized that a unique “bedside-to-bench” LARC patient-derived xenograft (PDX) model established from pre-neoadjuvant 5FU/RT LARC endoscopic biopsy samples could serve as a therapy predictive platform used for testing future personalization strategies.
Methods:
Mouse xenografts were created from pre-neoadjuvant 5FU/RT LARC endoscopic biopsy samples (IRB approved protocol). Dissociated cancer cells were mixed with Matrigel (1:1 ratio) and subcutaneously (SQ) injected in NSG mice to yield the passage 0 (P0) tumors. Dissociated tumor cells were passaged in an expanded number of mice to create subsequent passages for investigation. P2-3 xenografts were used for histologic characterization and for in vivo therapy experiments. Therapy groups include 1) DMSO control; or 2) 5FU (40 mg/kg ip X2) + RT (1Gy daily for 5 days) 3) Cetuximab (1mg/dose ip twice weekly for 3 weeks) +5FU/RT. Xenograft response, as measured by tumor weight and serial volume measurements, was compared to the corresponding human LARC.
Results:
After optimization, we achieved an outgrowth (human tumor to P0) rate of 73% (n= 19/26) and engraftment (P0 to P1) of 85%. Clinical factors did not predict outgrowth but tumors with distant metastasis demonstrated increased xenograft doubling times. Xenograft histology was conserved in 88% of cases compared with the human LARC tumor. Xenograft 5FU/RT response based on tumor volume correlated with clinical tumor regression grading response to neoadjuvant (5FU/RT) (Figure 1). KRAS wild type tumors demonstrated response to Cetuximab in addition to 5FU/RT (p<0.05).
Conclusion:
A novel LARC patient-derived xenograft (PDX) platform mirror images the biology, architecture, and therapy heterogeneity observed in LARC patients suggesting our model represents an effective platform to develop future personalization therapeutic strategies.
