Defined Clinical Classifications Are Associated with Outcome of Patients with Anatomically Resectable Pancreatic Adenocarcinoma Treated with Neoadjuvant Therapy

Abstract

Background We previously introduced a classification system for patients with localized pancreatic adenocarcinoma that integrates assessments of tumor anatomy, cancer biology, and patient physiology. By means of this system, we sought to analyze outcomes of patients with resectable anatomy but heterogeneous biology and physiology who were treated with neoadjuvant therapy. Methods We evaluated consecutive patients (2002–2007) with anatomically potentially resectable cancers treated with chemotherapy or chemoradiation before potential pancreatectomy. We compared clinical factors and outcomes of patients classified as having disease that was clinically resectable (CR; no extrapancreatic disease, preserved performance status); suspicion for extrapancreatic disease (BR-B); or marginal performance status or significant comorbidity (BR-C). Patients with borderline resectable anatomy (BR-A) were excluded. Results Resection rates for 138 CR, 41 BR-B, and 38 BR-C patients were 75, 46, and 37%, respectively (P < 0.001). Metastases, detected during treatment in 23% of patients, were the most common contraindication to resection among CR (15%) and BR-B (46%) patients. Performance status rarely precluded surgery except among BR-C (32%) patients. Factors associated with selection against surgery were older age, poor performance status, pain, and therapeutic complications (P < 0.05). The median overall survival of all patients was 21 months. Resected and unresected BR-B and BR-C patients had median overall survival durations similar to those of resected and unresected CR patients, respectively (P > 0.22). Conclusions This system describes discrete clinical subgroups of patients with pancreatic cancer who have similar, potentially resectable tumor anatomy but heterogeneous physiology and cancer biology. It may be used with neoadjuvant therapy to predict outcomes, individualize treatment algorithms, and optimize survival.