Susceptibility of BRCA Associated PDAC to Immunotherapy Checkpoint Inhibition

Introduction: Pancreatic ductal adenocarcinoma (PDAC) is one of the most lethal malignancies. Germline-BRCA1/2 mutation carriers are the most well defined DNA damage repair deficient subgroup and present up-to ~15% of PDAC in selected populations. DDR deficient cells display susceptibility to DNA damaging agents and PARP inhibition (PARPi). However, durable responses are limited and resistance evolves. Immunotherapy strategies have not shown significant clinical benefit in PDAC. DDR deficient PDAC features increased mutational burden, which may confer sensitivity to checkpoint inhibition. Compared to the general PDAC population, BRCA1/BRCA2-mutated cases have higher incidence of PD-1 and PD-L1 expression, respectively. We hypothesize that immuno-modulating strategies may have therapeutic value in BRCAmut PDAC, due to the higher, neo-antigen-encoding mutational burden.

Material and methods: We have developed unique patient-derived xenograft (PDX) models from metastatic PDAC patients (n=42). These models resemble the morphologic and genomic characteristics of primary PDAC as we described (Golan;Oncotarget;2017). 13/42 PDX models were established from germline BRCA-mutated patients obtained at distinct time points; before treatment and at progression. Correlation between disease course at tissue acquisition and response to PARPi/platinum was demonstrated in PDXs in-vivo (Golan;IJC;2018). 4/6 models were classified as `unstable`/`SDBR` (Waddell;Nature;2015, Connor;JAMAoncol;2017) by whole genome sequencing.

Results and discussion: Hematopoiesis-engrafted PDAC PDX model: 3-4week old NSG mice underwent sub-lethal irradiation and transplanted with CD34+ cells purified from umbilical–cord-blood. Human cell engraftment (hCD45) was detected from week 12 onwards. On week 18, cryopreserved tumor chunks from a PDX model established from a germline BRCA2 6174delT PDAC patient with high mutational (>32,000 SNVs) and neoantigene load was subcutaneously transplanted to engrafted and non-engrafted mice (control) and thereafter treated with pembrolizumab (10mg/kg; q3-4) for 5 weeks(n=5-6/group). Weight and tumor volume ((length*width^2)/2) were measured bi-weekly. Peripheral-blood was obtained every ~2-3-weeks. Decreased tumor growth rate was observed in the hCD45-engrafted pembrolizumab treated mice compared to hCD45-engrafted control treated mice, with one mice demonstrating complete response. hCD45, CD19, CD3, CD4 and CD8 cells were detected in splenocytes, bone-marrow and blood of engrafted mice. IHC analysis with HLA/CD45/CD3/CD4 and CD8 demonstrated tumor T cells infiltration. A trend of correlation between hCD45 engraftment and tumor growth inhibition was observed.

Conclusions: DDR-deficient PDAC tumors display increased mutational load and may serve as a favorable subgroup for development of effective (immuno)therapy strategies. Preliminary data demonstrated checkpoint inhibition efficacy in a humanized BRCAmut PDAC PDX model. The humanized model may serve as a platform for development of immunotherapy combinations for BRCAmut PDAC patients.