Preliminary Outcomes Following Use of Concurrent Nab-Paclitaxel and Radiation Therapy—A Novel Regimen for Borderline Resectable or Unresectable Pancreatic Cancer

The optimal concurrent chemoradiotherapy (CRT) regimen for treatment of borderline resectable or unresectable pancreatic cancer is unknown. This study evaluates toxicity and tumor response outcomes with concurrent nab-paclitaxel and radiation therapy (RT). As tumor cells exhibit preferential uptake of nab-paclitaxel, we hypothesized that its use results in improved toxicity and local control compared to standard (5-FU or gemcitabine) CRT. Thirty-three consecutive patients treated for initially borderline resectable or unresectable pancreatic cancer at our institution from 2014-2016 were divided into two groups based on treatment with concurrent nab-paclitaxel (100-125mg/mˆ2 weekly)/RT, or with concurrent standard CRT. The primary endpoint was difference in toxicity between groups and secondary endpoints were cumulative incidence of local failure and conversion to resectability. Fisher’s exact test was used to determine toxicity differences between groups. Local failure according to treatment group was estimated using the cumulative incidence function and the Fine-Gray regression model with death as a competing risk. There were 17 patients in the nab-paclitaxel/RT group and 16 patients in the standard CRT group. One patient in each group had borderline resectable disease. Median follow-up from diagnosis was 16.8 months (standard deviation [SD] 7.2) for all patients. The nab-paclitaxel/RT and standard CRT groups received 2.3 months (SD 1.4) and 4.3 months (SD 4.3) of chemotherapy prior to concurrent CRT (p=0.13); mean total RT dose received by each group was 5371 cGy and 5470 cGy, respectively. Maximum grade 2 or higher toxicities were similar between groups (p=0.69). There were no statistically significant grade 2 or higher CNS (fatigue, neuropathy), GI (nausea/vomiting, abdominal pain, diarrhea), or hematologic (neutropenia, anemia, thrombocytopenia) toxicities (p > 0.4 for each). Grade 3 toxicities were only observed for hematologic toxicity and were higher in the nab-paclitaxel group (35.6% vs 6.7%, p=0.09). The nab-paclitaxel/RT group experienced a lower cumulative incidence of local failure compared to the standard CRT group, though this difference was not statistically significant (SHR=0.65, 95% CI [0.19, 2.19], p=0.49). Six of 17 (35.3%) patients in the nab-paclitaxel/RT group and 2/16 (12.5%) patients in the standard CRT group underwent surgery following CRT (p=0.23); 4/6 (66%) and 2/2 (100%) patients in each respective group had a complete (R0) resection. Concurrent nab-paclitaxel/RT appears to have similar toxicity as standard CRT in the treatment of borderline resectable or unresectable pancreatic cancer. Its use was also associated with a lower cumulative incidence of local failure and a higher conversion to resectability though these differences were not statistically significant, possibly due to small sample size. Further work is needed to validate these observations in a larger patient cohort.