Figure 1 summarizes the analysis groups. After institutional review board (IRB) approval, all patients at our institution who underwent curative intent surgery between 2000 and 2012 for pancreatic adenosquamous or adenocarcinoma without neoadjuvant chemotherapy or radiation therapy were identified (“upfront resection” group). Tumors were deemed resectable by the surgeon and not assigned a clinical stage (TX) or were American Joint Committee on Cancer (AJCC) 7th edition stage I–IIB.
Our IRB approved registry contained 159 patients from 2010 to 2014 who received neoadjuvant or definitive chemotherapy and SBRT for the treatment of BRPC or LAPC, AJCC clinical stage IIA–III. 110 BRPC patients received neoadjuvant therapy and 56 (51%) underwent resection. An additional 5 (10%) of 49 LAPC patients underwent resection after sufficient response to the neoadjuvant treatment regimen. Demographics and outcomes for all BRPC and LAPC patients are reported separately (14). As our goal was to analyze the outcomes of surgically resected pancreatic cancer with or without neoadjuvant therapy, patients who did not undergo a curative intent resection were excluded from all analyses except the “intention-to-treat” survival analysis. Resected BRPC and LAPC patients comprise the “neoadjuvant therapy” group.
Staging methods at our institution have been previously described (19). BRPC and LAPC patients were initially judged medically fit to tolerate resection without probable distant metastases. Patients routinely undergo physical examination, standard blood chemistries including CA19-9, multidetector thin-slice pancreatic protocol computed tomography (CT) scan, positron emission tomography (PET) scan, endoscopic ultrasound (EUS), specialist pathology review, and presentation at multi-disciplinary tumor board. The highest stage identified by any imaging modality determined the clinical stage. In the upfront resection group, imaging methods were not standardized, but contrast CT and PET/CT were routinely performed.
Neoadjuvant treatment and restaging
Chemotherapy and SBRT protocols have been described in detail (20). In summary, patients typically received chemotherapy with three 21-day cycles of gemcitabine, docetaxel, and capecitabine (GTX) for BRPC or six 14-day cycles of oxaliplatin, irinotecan, fluorouracil, and leucovorin (FOLFIRINOX) for LAPC. Other chemotherapies were occasionally given at the discretion of the medical oncologist.
Before radiation planning, 2 to 4 fiducial markers were placed into the tumor under EUS guidance (21). A tumor motion study was then performed to gauge the amplitude of tumor motion by fluoroscopic tracking of the markers during respiration with an applied abdominal compression device. After immobilization by BodyFix cradle (Elekta, Stockholm, Sweden), isocenter was determined on non-contrast CT simulation scan followed by repeat 4-dimensional CT with intravenous and oral contrast. For tumor motion <1 cm, internal target volume (ITV) and planning organ-at-risk volumes (PRV) encompassing the entire respiratory cycle were used. If tumor motion was ≥1 cm, radiation treatment required respiratory gating by the Real-Time Position Management system (Varian Medical Systems Inc., Palo Alto, CA, USA), and treatment was planned and delivered during the 40–60% respiratory phases with voluntary breath-hold technique.
Dose painting intensity modulated radiation therapy was used to plan 5 fractions daily delivery totaling 30 Gy to the planning tumor volume (PTV) and concurrently up to 50 Gy to the high dose PTV (hdPTV) as limited by PRV constraints. The PTV consists of 3–5 mm expansion of the ITV generated from gross tumor volume within the pancreas plus motion. The hdPTV is delineated based on CT and EUS to include areas of tumor involvement of vascular structures that limit resectability: celiac axis, SMA, common hepatic artery, gastroduodenal artery, superior mesenteric vein (SMV), PV, or inferior vena cava. Highest priority is given to PRV constraints as follows: each of duodenum, small bowel, and stomach mean <20 Gy, volume receiving 30 Gy (V30) <2 cc, V35 Gy <0.5 cc; each kidney mean <10 Gy, spinal cord maximum 20 Gy. SBRT delivered by Varian Truebeam or Trilogy linear accelerator began at least 7 days after chemotherapy.
After completion of neoadjuvant treatment, BRPC and LAPC patients underwent restaging by repeat examination, pancreas protocol CT, and PET/CT. Resectability was again determined at multidisciplinary tumor board. The target time to surgery was 1–2 months after SBRT, and all patients except one underwent surgery within 3 months after SBRT.