Homologous Recombination Deficiency in Patients With Pancreatic Ductal Adenocarcinoma and Response to Chemotherapy

Safi Shahda; Kirsten M Timms; Ashley A Ibrahim; Julia E Reid; Harvey M Cramer; Milan Radovich; Sulfikar Ibrahim; Brian Allen; Bert H O'Neil

doi:10.1200/PO.17.00087

Homologous Recombination Deficiency in Patients With Pancreatic Ductal Adenocarcinoma and Response to Chemotherapy

JCO Precis Oncol. 2018 Nov:2:1-11. doi: 10.1200/PO.17.00087.

Authors

Safi Shahda¹, Kirsten M Timms¹, Ashley A Ibrahim¹, Julia E Reid¹, Harvey M Cramer¹, Milan Radovich¹, Sulfikar Ibrahim¹, Brian Allen¹, Bert H O'Neil¹

Affiliation

¹ Safi Shahda, Harvey M. Cramer, Milan Radovich, and Bert H. O'Neil, Indiana University School of Medicine, Indianapolis; Ashley A. Ibrahim and Sulfikar Ibrahim, Indiana University Health Ball Memorial Hospital, Muncie, IN; and Kirsten M. Timms, Julia E. Reid, and Brian Allen, Myriad Genetics, Salt Lake City, UT.

PMID: 35135100
DOI: 10.1200/PO.17.00087

Abstract

Purpose: Mutations or copy number abnormalities of genes involved in homologous recombination (HR) occur in pancreatic ductal adenocarcinoma (PDAC). DNA-based measures of HR deficiency (HRD) have been developed and may help identify tumors with better response to DNA-damaging agents. This study aimed to describe the HR pathway mutations and HRD status and determine their association with treatment response and outcome in patients with PDAC.

Patients and methods: We performed a retrospective analysis of tumor samples from patients treated at Indiana University for locally advanced or metastatic PDAC. Patients were included if they received gemcitabine plus nanoparticle albumin-bound paclitaxel (control) or fluorouracil, oxaliplatin, leucovorin, and irinotecan (FOLFIRINOX) and had adequate follow-up to assess survival and response to therapy. Tumor analysis generated a three-biomarker HRD score and mutation data for 44 genes.

Results: Ninety-one samples met inclusion criteria, and 78 samples (formalin-fixed paraffin-embedded, n = 15; fine-needle aspiration, n = 63) generated mutation data. HRD analysis was successful for 57 samples (HRD score: median, 18; range, 5 to 61); the primary cause of failure was low tumor cellularity. Six BRCA1/2 mutations were detected, four with HRD scores in the top decile (P = .011). There was no statistically significant correlation between HRD score and radiographic response (odds ratio per interquartile range, 1.40; P = .32 adjusted for treatment) in either treatment group. In patients treated with FOLFIRINOX, HRD score dichotomized at the median was not associated with progression-free survival (median, 5.3 v 9.4 months for low v high HRD score, respectively; P = .083) or overall survival (median, 11.9 v 10.7 months for low v high HRD score, respectively; P = .76).

Conclusion: Mutations in DNA repair genes occur in PDAC, and HRD scores can be generated in the majority of patients. The HRD score was not significantly associated with higher response rate or prolonged survival in patients treated with FOLFIRINOX.