NeoOPTIMIZE: Phase II trial of adaptive switching of neoadjuvant FOLFIRINOX (FFX) to gemcitabine/nab-paclitaxel (GA) resectable/borderline resectable (BR)/locally advanced (LA) pancreatic adenocarcinoma (PDAC).

• DOI: 10.1200/jco.2025.43.16_suppl.4190
• Publication Date: 2025-05-28T14:43:35Z
• Authors: Lyndsey Sandow; Justin Grenet; Adel Kardosh; Emerson Yu-sheng ...; Guillaume Joe Pegna; Skye C. Mayo; Robert Eil; Patrick Joseph Worth; Flavio G. Rocha; Nima Nabavizadeh; Aaron Grossberg; Alexander Guimaraes; Bryan Foster; Shaun M. Goodyear; Erin Taber; Dove Keith; Gordon Brent Mills; Rosalie C. Sears; Brett C. Sheppard; Charles D. Lopez

4190

Background: Neoadjuvant chemotherapy (NAC) for localized PDAC may improve R0 resection. FFX and GA are used but lack of predictive biomarkers remains a barrier to optimal NAC. The angiotensin-II receptor blocker (ARB), losartan, remodels vascular perfusion to enhance NAC efficacy. We established an experimental platform for dynamic switching of NAC +/- radiation therapy for resectable/BR PDAC and an exploratory cohort of LA PDAC. Methods: Patients (pts) received 4 cycles of FFX (ox 85 mg/m2; LV 400 mg/m2; iri 150 mg/m2; 5-FU 2400 mg/m2), then were restaged in a multi-disciplinary tumor board (Restage I). Pts without progression (CT and CA19-9 increase <30% from baseline) completed 4 more cycles of FFX. If there was progression (CT and/or CA19-9 increase > 30%) or intolerance, pts were switched to GA (nab-P 125 mg/m2; gem 1000 mg/m2) for 2 months. After a total of 4 months NAC, pts were re-staged (Restage 2) and had surgery or chemoRT (if vascular involvement) then surgery. Losartan (50mg QD) was given throughout NAC/chemoRT. The primary endpoint was the proportion of resectable/BR with R0 resection. Results: Of 43 patients screened,16 were resectable, 21 BR, 5 LA and 1 ineligible. Median age was 65 years (range: 34-80), 49% male, 84% Caucasian white. Head of pancreas primary was 84%. Mean baseline CA19-9 96 ng/mL. 1 pt had progressive disease prior to Restage 1. Of the 36 pts at Restage 1, 31 continued FFX and 5 switched to GA. 18 patients on FFX had a radiographic response, 5 had CA 19-9 decrease >25%, and 8 had stable disease with unchanged CA 19-9. 3 pts switched to GA for radiographic progression,1 for increased CA 19-9 >30% and 1 for FFX intolerance. Of 34 pts evaluated at Restage 2, 2 continued NAC, 13 had pre-op chemoRT (12 BR and 1 resectable), and 19 proceeded to surgery. 24/27 (88%) pts on FFX had R0 resections; 4/4 (100%) pts switching to GA had R0 resections. Overall, 77% pts completing NAC FFX had R0 resections; 80% pts switched to GA had R0 resections. Grade > 3 toxicities 7% FFX and 5% GA. Conclusions: Early switching to GA in pts progressing on FFX led to an equivalent R0 resection rate. Optimization of NAC made it possible to undergo curative intent surgery that would not have occurred. Secondary endpoints of DFS, OS and multiomic based studies of blood and tumor tissue via our Serial Measurements of Molecular and Architectural Responses to Therapy (SMMART) platform are underway to identify molecular markers to predict response and determine whether a switch in treatment is indicated.

Clinical trial information: NCT04539808

.