A randomized, phase II trial of oral azacitidine (CC-486) in patients with resected pancreatic adenocarcinoma at high risk for recurrence.

BACKGROUND: Of the only 20% of patients with resectable pancreatic ductal adenocarcinoma (rPDA), cancer recurs in 80% of cases. Epigenetic dysregulation is an early hallmark of cancer cells acquiring metastatic potential, and epigenetic modulators may reactivate tumor suppressor genes, delay recurrence, and sensitize PDA to future chemotherapy. METHODS: This was a randomized phase II study (NCT01845805) of CC-486 (oral DNA methyltransferase inhibitor azacitidine) vs. observation (OBS) in rPDA patients harboring high-risk features (stage pN1-2, R1 margins, or elevated CA 19-9 level) with no evidence of disease following standard adjuvant therapy. Patients were randomized to oral CC-486 treatment (300 mg daily on days 1-21 on a 28-day cycle) or OBS for up to 12 cycles or until disease relapse/unacceptable toxicities. Following recurrence, records of next-line therapies, imaging, and survival were obtained. The primary endpoint was progression-free survival (PFS)-time from randomization to recurrence (imaging/biopsy confirmed or death). Secondary endpoints included OS and PFS and ORR and metastatic PFS with subsequent next-line systemic therapy in metastatic setting. RESULTS: Forty-nine patients (24 in CC-486 arm, 25 in OBS arm) were randomized: median age 66 (range 36-81), 53% male, 73% node positive, 49% elevated CA 19-9, 20% R1 resection, 63% and 100% received perioperative concurrent chemoradiation and chemotherapy, respectively. Median time from surgery to randomization was 9.6 mo (range 2.9-36.8). For the CC-486 arm, median treatment duration was 5.6 mo (range 1.3 to 12.8) with 14 treatment-related grade 3 or 4 AEs among 5 patients (22%) resulting in dose-reduction. Four patients (17%) discontinued therapy due to AEs. With median follow-up of 20.3mo (IQR 12.8, 41.4), 38 (79%) of evaluable patients recurred (34 imaging-confirmed, 4 clinically). Median PFS in imagining-confirmed cases was 9.2 and 8.9mo (HR 0.94, 95% CI 0.46-1.87, p = 0.85) for CC-486 and OBS patients, respectively. Median OS (2-yr OS%) was 33.8 (50%) and 26.4 mo (61%) in CC-486 and OBS patients, respectively. (HR 0.98, 95% CI 0.46-2.05, p = 0.96). ORR with subsequent chemotherapy in the metastatic setting was minimal in both arms. CONCLUSIONS: Treatment with CC-486 following adjuvant therapy did not prolong time-to-relapse in patients with high-risk rPDA or improve disease response on 1st-line metastatic therapy.

Integrated preclinical and clinical development of S-trans, trans-Farnesylthiosalicylic Acid (FTS, Salirasib) in pancreatic cancer.

PURPOSE: S-trans,trans-Farnesylthiosalicylic Acid (FTS, salirasib) inhibits Ras-dependent cell growth by dislodging all isoforms of Ras, including mutant Ras, from the plasma membrane. This study evaluated the activity, safety, and toxicity of salirasib in preclinical models and patients with metastatic pancreatic adenocarcinoma (PDA). PATIENTS AND METHODS: In the preclinical study, salirasib was tested, alone and in combination with gemcitabine, in patient derived xenografts (PDX) of PDA. In the clinical study, treatment-naïve patients with advanced, metastatic PDA were treated with a standard dose schedule of gemcitabine and salirasib 200-800 mg orally (PO) twice daily (bid) for 21 days every 28 days. Tissue from preclinical models and patients' biopsies were collected pre-treatment and on Cycle (C) 1, Day (D) 9 to characterize the effect of gemcitabine and salirasib on activated Ras protein levels. Plasma samples for pharmacokinetics were collected for salirasib administered alone and in combination. RESULTS: Salirasib inhibited the growth of 2/14 PDX models of PDA and modulated Ras signaling in these tumors. Nineteen patients were enrolled. No DLTs occurred. Common adverse events included hematologic and gastrointestinal toxicities and fatigue. The median overall survival was 6.2 months and the 1 year survival 37 %. In 2 patients in whom paired tissue biopsies were available, Ras and KRas protein levels were decreased on C1D9. Salirasib exposure was not altered by gemcitabine and did not correlate with PD outcomes. CONCLUSION: The combination of gemcitabine and salirasib appears well-tolerated, with no alteration of salirasib exposure, and exerted clinical and PD activity in PDA.