Background: No validated biomarkers exist to direct treatment decisions in PDAC. Decisions regarding tx choices are based on age, organ function, and performance status. Defining predictors of efficacy to F (alone or in combination) or G (alone or in combination) in PDAC are urgently needed. Histologic subtype by immunohistochemistry (IHC); pancreatobiliary type (PB), intestinal type (I), gastric type (Ga) and intestinal/gastric type (I/G); may predict benefit to G or F. Methods: Charts of PDAC pts from 2 institutions from 2007 to 2013 having both pathology specimen and a qualifying tx (FOLFIRINOX, FOLFOX, FOLFIRI, Capecitabine or G or G and nab-paclitaxel) were reviewed. IHC phenotypes were defined by staining >25% CDX2 = I, >25% MUC5 = Ga, >25% CDX2 and MUC5 = I/G, >25% MUC1 = PB. Results: 51 pts were identified. 4 cases had Stage II disease without recurrence after adjuvant tx and were excluded. 5 tumors were PB, 1 I, 3 I/G, 38 Ga. 41 (87%) had Ga or I/G type. Due to lack of statistical power for the other histologic subtypes, response to F or G was only analyzed in the 35 metastatic pts with the Ga subtype. 14 were metastatic at diagnosis. 24 males. Median age 57. 21 Hispanic, 14 Non-Hisp. The PFS for Ga treated with F based tx (>12 m) versus G (~9 m) based tx in the 1st line approached statistical significance (p=0.050). In multivariable analysis with adjustments for clinical stage, ECOG, age at dx, there was a significant lower risk of progression in pts treated with F based tx versus G based tx (HR=0.278, p=0.0315). Conclusions: There was near statistical significance favoring F based tx over G based tx in PDAC Ga subtype (p=0.05). Historically, G has proven superior to F as a single agent but our findings suggest F in combination may be superior to G in combination, at least for the treatment of Ga subtype.