The impact of neoadjuvant therapy in patients with left-sided resectable pancreatic cancer: an international multicenter study

Background: Left-sided pancreatic cancer is associated with worse overall survival (OS) compared with right-sided pancreatic cancer. Although neoadjuvant therapy is currently seen as not effective in patients with resectable pancreatic cancer (RPC), current randomized trials included mostly patients with right-sided RPC. The purpose of this study was to assess the association between neoadjuvant therapy and OS in patients with left-sided RPC compared with upfront surgery. Patients and methods: This was an international multicenter retrospective study including consecutive patients after left-sided pancreatic resection for pathology-proven RPC, either after neoadjuvant therapy or upfront surgery in 76 centers from 18 countries on 4 continents (2013-2019). The primary endpoint was OS from diagnosis. Time-dependent Cox regression analysis was carried out to investigate the association of neoadjuvant therapy with OS, adjusting for confounders at the time of diagnosis. Adjusted OS probabilities were calculated. Results: Overall, 2282 patients after left-sided pancreatic resection for RPC were included of whom 290 patients (13%) received neoadjuvant therapy. The most common neoadjuvant regimens were (m)FOLFIRINOX (38%) and gemcitabine-nab-paclitaxel (22%). After upfront surgery, 72% of patients received adjuvant chemotherapy, mostly a single-agent regimen (74%). Neoadjuvant therapy was associated with prolonged OS compared with upfront surgery (adjusted hazard ratio 0.69, 95% confidence interval 0.58-0.83) with an adjusted median OS of 53 versus 37 months (P = 0.0003) and adjusted 5-year OS rates of 47% versus 35% (P = 0.0001) compared with upfront surgery. Interaction analysis demonstrated a stronger effect of neoadjuvant therapy in patients with a larger tumor (Pinteraction = 0.003) and higher serum carbohydrate antigen 19-9 (CA19-9; Pinteraction = 0.005). In contrast, the effect of neoadjuvant therapy was not enhanced for splenic artery (Pinteraction = 0.43), splenic vein (Pinteraction = 0.30), retroperitoneal (Pinteraction = 0.84), and multivisceral (Pinteraction = 0.96) involvement. Conclusions: Neoadjuvant therapy in patients with left-sided RPC was associated with improved OS compared with upfront surgery. The impact of neoadjuvant therapy increased with larger tumor size and higher serum CA19-9 at diagnosis. Randomized controlled trials on neoadjuvant therapy specifically in patients with left-sided RPC are needed.