Gemcitabine (G) fixed-dose-rate infusion (FDR) plus erlotinib (E) in patients with advanced pancreatic cancer (APC).

304 Background: G (30-minute infusion) plus E improves survival in patients with APC compared with G alone. In a recent phase III trial, G-FDR showed a trend to better OS compared with standard G (6.2 vs. 4.9 months, HR 0.83, p=0.04), although the study was underpowered to detect great difference in OS. Based on our previous experience with G-FDR, we decided to evaluate the combination of G-FDR plus E, after E approval for APC. Methods: Patients with previously untreated pathologically confirmed APC, locally advanced (LAPC) or metastatic (MPC), and ECOG PS 0-2 were included. G 1500 mg/m 2 was given by 150-min infusion (10 mg/m 2 /min) on days 1, 8, and 15 every 28 days combined with E 100 mg/day orally. Treatment modifications for G-FDR were planned according with previously Tempero's phase II trial, and as described in prescribing information for E. Results: 62 pts were included (36M/26F), with a median age of 63 y-o (range 37-78). ECOG PS 0/1/2: 19/40/3. LAPC/MPC: 16/46. All except one had measurable disease. ORR was 13% (8 PR), 95% CI: 4.7-21.3, and there were 34 (55%) SD. Mean relative dose intensity for G was 0.76 and 0.90 for E. Main hematologic toxicities 3/4 per pt: anaemia 12/0, thrombocytopenia 7/4, neutropenia 18/7. Acneiform rash 1/2/3 occurred in 16/16/3 pts. Other relevant adverse events were (grade 2/3/4): diarrhoea 18/3/0, mucositis 5/1/0, infection 9/8/1, thrombosis 1/4/1 and vomiting 6/4/0. There were three treatment-related deaths (septic shock, cholangitis, and bilateral pulmonary embolism). Ten pts (all LAPC) received RT after ≤ 6 cycles, all with concomitant capecitabine 825 mg/m 2 bid. In 4 pts salvage surgery were performed: 2 R0, 1 R1 and 1 R2. Median PFS was 4.9 months (95% CI: 3-6.7), 7.9 m for LAPC and 2.5 m for MPC (p = 0.004). Median OS was 10 months (95% CI: 7.1-12.9), 17.5 m for LAPC and 7 m for MPC (p = 0.019). OS was significantly shorter in males (p = 0.01) and in pts taking major opioids (p = 0.027). There was a trend to better OS in pts who developed skin rash grade ≥ 2 (p = 0.078). Conclusions: In this noncomparative study, G-FDR plus E is a feasible regimen in APC with an acceptable toxicity and notable activity. G-FDR seems to increase hematologic toxicity compared with standard infusion. No significant financial relationships to disclose.