Thumar, Jaykumar

PURPOSE SWOG S1815 was a randomized, open label phase III trial, evaluating gemcitabine, nab-paclitaxel, and cisplatin (GAP) versus gemcitabine and cisplatin (GC) in patients with newly diagnosed advanced biliary tract cancers (BTCs). METHODS Patients with newly diagnosed locally advanced unresectable or metastatic BTC, including intrahepatic cholangiocarcinoma (ICC) and extrahepatic cholangiocarcinoma (ECC) and gallbladder carcinoma (GBC), were randomly assigned 2:1 to either GAP (gemcitabine 800 mg/m 2 , cisplatin 25 mg/m 2 , and nab-paclitaxel 100 mg/m 2 intravenously once per day on days 1 and 8 of a 21-day cycle) or GC (gemcitabine 1,000 mg/m 2 and cisplatin 25 mg/m 2 intravenously once per day on days 1 and 8 of a 21-day cycle). RESULTS Among 452 randomly assigned participants, 441 were eligible and analyzable, 67% with ICC, 16% with GBC, and 17% with ECC. There was no significant difference in overall survival (OS) between GAP versus GC. Median OS with GAP was 14.0 months (95% CI, 12.4 to 16.1) and 13.6 months with GC (95% CI, 9.7 to 16.6); hazard ratio (HR), 0.91 (95% CI, 0.72 to 1.14); P = .41. Median progression-free survival (PFS) was similar between groups with median PFS for GAP being 7.5 months (95% CI, 6.4 to 8.5) versus 6.3 months for GC (95% CI, 4.4 to 8.2); HR, 0.89 (95% CI, 0.71 to 1.12); P = .32. In exploratory subset analyses, the OS and PFS benefits of GAP versus GC treatment were greater in locally advanced disease compared with metastatic disease, although not statistically significant (interaction P = .14 for OS and P = .17 for PFS). Moreover, GAP versus GC showed greater improvement in PFS among participants with GBC than those with ICC or ECC (interaction P = .01), but not OS (interaction P = .28). CONCLUSION The addition of a taxane in the GAP regimen to the standard gemcitabine-cisplatin regimen did not improve OS in newly diagnosed BTC. More toxicity was encountered with GAP versus GC.

LBA490 Background: Biliary tract cancers (BTCs) are a heterogeneous group of malignancies with a dismal prognosis. Gemcitabine-based regimens are the standard of care in advanced disease, but median overall survival (OS) is roughly 12 months. The addition of albumin-bound paclitaxel to gemcitabine and cisplatin (GAP) demonstrated promising efficacy in a 60 patient, single-arm phase II study (Shroff et al, JAMA Oncol 2019), with observed median OS of 19.2 months. Methods: SWOG 1815 is a randomized, open-label, phase III trial comparing GAP to gemcitabine/cisplatin (GC). The study included newly diagnosed advanced BTC patients (pts), randomized 2:1 to GAP vs. GC. GAP included gemcitabine at 800 mg/m2, cisplatin at 25 mg/m2 and albumin-bound paclitaxel at 100 mg/m2 on days 1 and 8 of a 21-day cycle. GC included standard dosing of gemcitabine at 1000 mg/m2 and cisplatin at 25 mg/m2 on days 1 and 8 of a 21-day cycle. Pts were treated until progression. The primary endpoint was overall survival (OS) with a target hazard ratio of 0.7 with 90% power and a 1-sided alpha of 0.025; randomization was stratified by disease site (intrahepatic cholangiocarcinoma [CCA] vs gallbladder adenocarcinoma [GBC] vs extrahepatic CCA), disease stage (locally advanced vs metastatic), and Zubrod PS 0 vs 1. Results: Of 441 eligible pts randomized, 55% were female. 67% of patients had intrahepatic CCA, 16% had GBC and 17% had extrahepatic CCA. Most pts had metastases (73%). Median OS with GAP vs. GC was 14 vs. 12.7 mo respectively (HR 0.93, 95% CI 0.74-1.19, p=0.58), ORR (confirmed and unconfirmed) 34% vs 25% (p=0.11) and median PFS 8.2 vs 6.4 mo (HR 0.92, 95% CI 0.72-1.16, p=0.47), respectively. Grade 3 and 4 treatment related adverse events (TRAEs) in ≥10% of pts for GAP and GC were anemia, neutropenia, and thrombocytopenia. GAP had more ≥ grade 3 hematologic AEs compared to the GC arm (60% vs. 45%, p=0.003). Discontinuation due to toxicity was at 24% vs 19% (p=0.26) with GAP vs GC. In exploratory subset analyses, GAP vs GC improved OS in pts with locally advanced disease (medians 19.2 vs 13.7 mo; HR 0.67, 95% CI 0.42-1.06, p=0.09) and in GBC pts (medians 17.0 vs 9.3 mo; HR 0.74, 95% CI 0.41-1.35, p=0.33). ORR for GAP vs GC in GBC was 50% vs 24% (p=0.09) and for locally advanced disease 28 vs 21% p=0.74. Conclusions: SWOG 1815 did not result in a statistically significant improvement in median OS with GAP vs. GC. The GAP regimen had higher rates of TRAEs without a statistically significant difference in discontinuation rates. Pts with locally advanced disease and GBC may benefit from the use of GAP. Further analyses are ongoing to understand potential benefit of GAP in subsets of BTC pts. Funding: NIH/National Cancer Institute grants CA180888, CA180819, CA180820, CA180821, and CA180868; and in part by Celgene Corporation, Summit, NJ (subsidiary of Bristol Myer Squibb). Clinical trial information: NCT#03768414 .

PURPOSE Pegylated recombinant human hyaluronidase (PEGPH20) degrades hyaluronan (HA) and, in combination with chemotherapy, prolongs survival in preclinical models. The activity of PEGPH20 with modified fluorouracil, leucovorin, irinotecan, and oxaliplatin (mFOLFIRINOX) was evaluated in patients with metastatic pancreatic cancer (mPC). MATERIALS AND METHODS Patients had untreated mPC, a performance status of 0 to 1, and adequate organ function. Tumor HA status was not required for eligibility. After a phase Ib dose-finding study of mFOLFIRINOX plus PEGPH20, the phase II open-label study randomly assigned patients (1:1) to the combination arm or to mFOLFIRINOX alone (n = 138). The primary end point was overall survival (OS). RESULTS PEGPH20 dosages of 3 µg/kg every 2 weeks were more tolerable than twice-weekly dosages used in the phase I study, so 3 µg/kg every 2 weeks was the phase II dosage. An amendment instituted enoxaparin prophylaxis in the PEGPH20 combination arm as a result of increased thromboembolic (TE) events. The planned interim futility analysis when 35 deaths (of 103 analyzable patients) occurred resulted in an OS hazard ratio (HR) of 2.07 that favored the control arm, and the study was closed to accrual. The treatment-related grade 3 to 4 toxicity was significantly increased in the PEGPH20 combination arm relative to control (odds ratio, 2.7; 95% CI, 1.1 to 7.1). The median OS in the mFOLFIRINOX arm was 14.4 months (95% CI, 10.1 to 15.7 months) versus 7.7 months (95% CI, 4.6 to 9.3 months) in the PEGPH20 combination arm. CONCLUSION Addition of PEGPH20 to mFOLFIRINOX seems to be detrimental in patients unselected for tumor HA status. This combination caused increased toxicity (mostly GI and TE events) and resulted in decreased treatment duration compared with mFOLFIRINOX alone. The median OS in the mFOLFIRINOX control arm (14.4 months) is, to our knowledge, the longest yet reported and can be considered for patients with good PS.

208 Background: PEGPH20 degrades hyaluronan (HA), a major component of the stroma, increases delivery of gemcitabine and prolongs survival in preclinical models. We evaluated the activity of PEGPH20 in combination with mFFOX in mPC , unselected for tumor HA. Methods: Pertinent eligibility: untreated mPC, PS of 0-1 and adequate organ function. Standard FFOX was modified to add prophylactic growth factor support and omit bolus 5FU. Due to increased thromboembolic (TE) events with PEGPH20, an amendment instituted LMWH prophylaxis in the PEGPH20 arm only. Following a dose finding cohort of mFFOX + PEGPH20, the Phase II study randomized patients (1:1) to the combination arm or mFFOX alone (n = 138). The primary endpoint was overall survival (OS), with a null median OS of 10 mo and an alternative of 15 mo (1-sided type 1 error 0.1, 80% power). Results: PEGPH20 at 3 mcg/kg, q2 weeks was more tolerable than twice weekly dosing. The randomized phase II study began May 2015. The planned interim futility analysis when 35 deaths occurred gave a HR of 0.44 favoring the standard arm, thus the study was closed to accrual (March 2017). As of August 22, 2017 (n = 111), the estimated median OS in the mFFOX arm was 15.1 mo (95% CI 10.1-15.7) vs 7.6 mo (95% CI 4.6-9.2) in the PEGPH20 arm. Conclusions: OS in the mFFOX control arm (15.1 mo) is longest yet reported. Addition of PEGPH20 to mFFOX is not recommended for further study and appears to be detrimental (HR = 0.48). The impact of PEGPH20 on OS was unexpected and in contrast to favorable results reported for the combination of gemcitabine/nab-paclitaxel + PEGPH20 especially in the HA high cohort (Hingorani S et al. A 4008, ASCO 2017). PEGPH20 with mFFOX caused increased toxicity (mostly GI and TE events) and decreased treatment duration compared to mFFOX alone. Tumor HA content will be analyzed. Funding: NIH/NCI grants CA180888, CA180819; and in part by Halozyme Inc. Clinical trial information: 01959139. [Table: see text]