Associations of 9p21 deletions on chemotherapy outcomes in pancreatic ductal adenocarcinoma.

Abstract

783 Background: Pancreatic ductal adenocarcinoma (PDAC) is the third leading cause of cancer death in the United States, with a 5-year survival rate of ~13%. The loss of tumor suppressor CDKN2A/B is well described broadly from molecular profiling of advanced PDAC, representing ~20% of the overall population. Furthermore, the chromosomal loss more broadly at 9p21 can lead to concurrent methylthioadenosine phosphorylase (MTAP) gene loss, for which ongoing strategies are considering the use of PMRT5 inhibitors. Here, we analyze the frequency of 9p21 chromosomal loss in PDAC specifically as stratified by initial clinical presentation and outcomes to first-line chemotherapy between FOLFIRINOX (FFX) and gemcitabine-nab/paclitaxel (G/nPAC). Methods: A retrospective analysis was conducted utilizing PDAC patients (n = 498) from the UW Precision Medicine Molecular Tumor Board. Criteria included histologically confirmed PDAC with available comprehensive genomic profiling. Electronic medical records were queried for baseline demographics and clinical outcomes by treatment regimen. Pathogenic deletions in CDKN2A/B defined 9p21 status. Chi-square testing was used to evaluate variant enrichment in localized versus metastatic disease. Log-rank Kaplan-Meier analysis with censoring was performed to calculate overall survival (OS) defined from initial staging. Results: The population included 52.4% (n=261) with loco-regional disease and 47.6% (n=237) with metastatic disease. Baseline demographics included a median age of 66.5 years and 51.8% female (n=258). 9p21 deletion was seen in 23.5% (n=117) of patients with enrichment in metastatic disease as compared to loco-regional disease (32.1% v. 15.7%, χ 2 =9.68, p<0.005). 9p21 loss in metastatic PDAC did not confer significant differences in median OS (p=0.13). However, 9p21 loss in loco-regional disease was associated with inferior OS as compared to wildtype (18.3±6.4mo v. 25.4±5.7mo, p<0.001). In loco-regional PDAC with 9p21 wildtype status, FFX showed improved OS as compared to G/nPAC (36.8±11.6 v. 23.6±9.6, p=0.012). However, in 9p21 loss PDAC, there were no differences in OS by initial chemotherapy (FFX v. G/nPAC) for either metastatic (p=0.93) or loco-regional disease (p=0.93). Conclusions: 9p21 loss in PDAC is prognostic for inferior survival in loco-regional PDAC. FFX confers improved OS in 9p21 wildtype PDAC, despite no OS improvements as compared to G/nPAC in the 9p21 loss cohort. These findings underline 9p21 deletion as critical in resistance to the limited options of systemic chemotherapy. As a retrospective cohort, these findings require further validation in larger multicenter studies that consider additional clinical variables. Alternative therapeutic strategies, including the role of PMRT5 inhibitors, should be explored further, specifically in loco-regional disease.