2957 - Predictors of Local and Distant Recurrence after Neoadjuvant Chemoradiation for Extended Criteria Borderline Resectable Pancreatic Cancer

C. Billena¹, S. Koro², S. Iyer³, A. Bommireddy², Z. S. Mayo¹, N. M. Woody², K. L. Stephans⁴, P. Pendyala⁵, J. A. Miller⁶, D. Joyce⁷, R. Simon⁷, M. Walsh⁷, N. Samer⁷, B. Estfan⁸, S. Kamath⁸, A. Khorana⁸, S. Krishnamurthi⁸, and E. H. Balagamwala⁴; ¹Department of Radiation Oncology, Cleveland Clinic Foundation, Cleveland, OH, ²Department of Radiation Oncology, Taussig Cancer Institute, Cleveland Clinic, Cleveland, OH, ³Case Western Reserve University School of Medicine, Cleveland, OH, ⁴Department of Radiation Oncology, Cleveland Clinic, Cleveland, OH, ⁵Department of Radiation Oncology, Cleveland Clinic Akron General, Akron, OH, ⁶Cleveland Clinic, Cleveland, OH, ⁷Department of Hepatobiliary Surgery, Cleveland Clinic, Cleveland, OH, ⁸Department of Medical Oncology, Taussig Cancer Center, Cleveland Clinic, Cleveland, OH

Purpose/Objective(s): PREOPANC showed improved overall survival with neoadjuvant chemoradiation (CRT) in borderline resectable pancreatic cancer (BR-PC). However, the definition of BR-PC is variable. The objective of this study was to report predictors and patterns of failure in extended criteria BR-PC at a high-volume pancreatic cancer center. Materials/

Methods: Charts and CT/MR imaging of 51 patients with BR-PC from 2010-2021 were reviewed from an institutional registry. Clinicopathologic features were summarized with descriptive statistics. Fisher’s exact test was used for categorical variables. Local (LF) and distant failure (DF) rates were assessed using cumulative incidence function and predictors were assessed with Gray’s test and competing risk regression.
Results: Median follow-up was 11.1 months. Median age was 69 years. All patients were deemed to be BR-PC by multidisciplinary review. Major arterial involvement (MAI) of the CA, SMA, or CHA was present in 13 (15%) patients. 3 patients had haziness and 10 patients had solid tumor contact, of which 3 had >180 encasement. Major venous involvement (MVI) of the MPV or SMV occurred in 15 (29%) patients. 1 had haziness and 14 had solid tumor contact, of which 3 had >180 encasement and 8 had venous contour irregularity. 11 patients (24%) had both MAI and MVI. 13 (25%) were unresectable per NCCN criteria. Among the 13 (27%) node positive patients, 12 were N1 and 1 was N2. Median dose of CRT was 50.4 Gy in 28 fractions with an SIB to 56 Gy and concurrent chemotherapy. Multi-agent perioperative chemotherapy was used in 78%. Of the 51 patients, surgical resection was performed in 18 (35%) patients. Negative margins were achieved in 94%. Treatment effect scores for CR, PR, and minimal response were 39%, 44%, and 17%, respectively. The reasons contributing to inability to resect included development of metastases in 19 (58%), un-resectability of the primary in 10 (30%), or clinical deterioration in 6 (18%). NCCN un-resectability criteria was not associated with receipt of surgery (31% unresectable vs 37% resectable, p=0.77) Incidence of LF was 24% and 39% at 1 and 3 years, respectively. Predictors of LF included MVI (p<0.01), MAI (p<0.01), and nodal status (p<0.01) but not receipt of surgery, margin, treatment effect score, LVSI, PNI, or treatment sequencing (p>0.05). On MVA, MVI (HR 11.5, p<0.01) and nodal status (HR 51.4, p<0.01) persisted. Incidence of DF was 49% and 64% at 1 and 3 years, respectively, which was associated with MAI (p=0.01), MVI (p<0.01), receipt of surgery (p<0.01), and treatment sequencing (p<0.01). On MVA, MVI (HR 2.21, p<0.01) and receipt of surgery (HR 0.43, p<0.01) remained predictors.
Conclusion: Of those patients that did not develop metastatic disease after neoadjuvant CRT and peri-operative multi-agent chemotherapy, 56% were down-staged from extended criteria BR. Although current management for locoregional disease centers around resectability, our data suggests major venous involvement portends significantly worse LF and DF risk.