Gemcitabine response prediction in the adjuvant treatment of resected pancreatic ductal adenocarcinoma using an AI histopathology platform

• Published in: Journal of clinical oncology Vol. 40; no. 16_suppl; p. e16295
• Main Authors: Krishna, Viswesh, Nimgaonkar, Vivek, Tiu, Ekin, Krishna, Vrishab, Bhambhvani, Hriday, Cook, Stephen, Miller, Daniel, Vrabac, Damir, Joshi, Anirudh, Singhi, Aatur D., Hendifar, Andrew Eugene, Rajpurkar, Pranav, Collisson, Eric Andrew
• Format: Journal Article
• Language: English
• Published: 01-06-2022
• ISSN: 0732-183X; 1527-7755
• DOI: 10.1200/JCO.2022.40.16_suppl.e16295

e16295
Background: Adjuvant chemotherapy improves survival following resection of pancreatic ductal adenocarcinoma (PDAC). A modified fluorouracil/irinotecan/oxaliplatin regimen (mFOLFIRINOX) has demonstrated improved disease free survival and overall survival, though gemcitabine-based monotherapy and gemcitabine plus capecitabine are alternatives in less fit patients. Though there are several proposed biomarkers to guide treatment decisions (GATA6, hENT1, and GemPred), no biomarker is used to guide treatment selection in clinical practice. Consequently, we sought to develop an artificial intelligence-derived signature of features from digital images of routine histopathology specimens that could identify patients susceptible to routine chemotherapeutic agents. Methods: 139 whole-slide digitized histological slides corresponding to 102 resected PDAC tumors from TCGA-PAAD were used in this study. This dataset corresponded to patients that had received either gemcitabine-backbone or 5 FU-backbone chemotherapy as their first-line adjuvant treatment. We extracted nuclei images from tissue regions using segmentation models and computed geometric features of these nuclei which we then correlated with Disease Specific Survival (DSS) in order to construct a signature associated with treatment benefit. This signature was compared against two board certified pathologists using the grade of the digital slides images to classify patients into above or below average DSS buckets. Results: Among quantitative geometric features, a set of area and ellipse features describing nuclei geometry correlated most with response to gemcitabine (R̃0.4). The cox proportional hazards model using these geometric nuclei features was found to be predictive of response to gemcitabine and achieved a C-index (95% CI) of 0.69 (0.58, 0.79). The pathologist-based baseline model for above and below average DSS had a median DSS of 443 and 461 days respectively. Using the average expected lifetime as the threshold, the model divides patients receiving gemcitabine into two histological subtypes with median DSS of 586 and 394 days respectively (p < 0.05). The model appeared specific to gemcitabine. Among patients receiving 5-FU (n = 10) there was no statistical significance in median DSS between the subtypes and a c-index of 0.63 (0.27, 1.0). Conclusions: An artificial intelligence approach utilizing only routine histopathology can identify features that correlate with treatment outcomes in PDAC with classification performance (c-index:0.69) superior to the validated AJCC treatment prediction tool (0.59).