ESMO 2017: Immunogenomics of Lethal Prostate Cancer

Madrid, Spain (UroToday.com) Dr. Johann de Bono provided an in-depth discussion on the immunogenomics of lethal prostate cancer at this morning’s Immunobiology in Prostate Cancer symposium at ESMO 2017 in Madrid, Spain. Dr. de Bono and his team hypothesized that a subset of lethal prostate cancers have higher mutational loads and that these are associated with DNA repair defects, with tumor genomics impacting antitumor immune response.

As part of the SU2C research dream team’s initiative, sequencing data has certainly demonstrated that advanced prostate tumors have a large subset of DNA repair defects. Among these 335 patients, additional genomic signatures included DNA repair defect mutations focusing on the MMR genes and mutational load and MMR signatures. Transcriptome profiling of these patients noted expression of 30 immune checkpoint transcripts and expression of 760 selected immune transcripts.

Using two different methods (Bayesian and non-Bayesian) for the SU2C cohort, their group was able to identify and analyze four different mutational types: MMR6, MMR26, HRD3, and AGING1. Transcriptomes of these MMR_signa mCRPC tumors had mRNA signatures reported to associate with MMR loss of function with decreased MMR gene expression, as well as higher mutational load. MMR_signa, but not the degree of mutational load, was associated with inferred immune infiltrate based on high macrophage detection from mCRPC biopsies. Furthermore, there appears to be an association between immune checkpoint expression and mutation load.

Among 306 samples from the Royal Marsden cohort (179 hormone sensitive and 127 mCRPC), targeted NGS with a 113-gene panel was performed to assess mutational load, MMR gene mutations, and microsatellite instability by NGS. Immunohistochemistry (IHC) for MMR proteins (MSH2, MSH6, MLH1, and PMS2) was performed, as was immune cell multi-color immunofluroesence and PD-L1 immunohistochemistry. For both hormone sensitive and mCRPC samples, MMR protein levels were associated with microsatellite instability by NGS. In clinical models, MMR_IHC mCRPC patients have a worse prognosis compared to MMR proficient tumors when comparing OS from diagnosis (log-rank p-value 0.07) and OS from start of LHRH (log-rank p-value 0.01). Furthermore, MMR_IHC associates with PD-L1 in mCRPC and immune cell infiltration based on myeloid-derived suppressor cell markers and T-cell markers.

Certainly, the immunogenomic research space has implications for clinical trials. As reported at last year’s ESMO 2016 meeting [1], KEYNOTE-028 assessed in a phase Ib setting the role of pembrolizumab in PD-L1 positive patients with CRPC. Among 23 patients, with a median follow-up of 33 weeks, three patients had a confirmed partial response, for an ORR of 13% and median duration of response of 59 weeks. Stable disease was noted in 39% of patients, although 20% of patients had increased tumor burden. As such, patient selection is crucial. Ultimately, immunotherapy trials must interrogate assays for determining RNAseq profile, immune checkpoint expression, and immune transcript expression in order to determine the optimal predictive biomarker model.

Speaker: Johann de Bono, Institute of Cancer Research and Royal Marsden, London, United Kingdom

Written By: Zachary Klaassen, MD, Urologic Oncology Fellow, University of Toronto, Princess Margaret Cancer Centre, Twitter: @zklaassen_md at the European Society for Medical Oncology Annual Congress - September 8 - 12, 2017 - Madrid, Spain

References:

Hansen A, Massard C, Ott PA, et al. Pembrolizumab for patients with advanced prostate adenocarcinoma: Preliminary results from the KEYNOTE-028 study. ESMO 2016 abstr 725.