Population-based analysis of treatment patterns and outcomes for pancreas cancer in Victoria.

Abstract

The Victorian Pancreas Cancer summit 2017 analysed state-wide data on management of Victorians with pancreas cancer between 2011 and 2015 to identify variations in care and outcomes. Pancreas cancer remains a formidable disease but systemic therapies are increasingly effective. Surgery remains essential but insufficient alone for cure. Understanding patterns of care and identifying variations in treatment is critical to improving outcomes. This population-based study analysed data collected prospectively by Department of Health and Human services (Victorian state government). Data were extracted from Victorian Cancer Registry (covering all Victorian cancer diagnoses), Victorian Admitted-Episodes Dataset (all inpatient data), Victorian Radiotherapy Minimum Dataset and Victorian Death Index providing demographics, tumour and treatment characteristics, age-standardized incidence, overall and median survival. Of 3962 Victorian patients with any form of pancreatic malignancy, 82% were ductal adenocarcinoma (PDAC), of whom 67% had metastases at diagnosis. One-year overall survival for PDAC was 30% (60% non-metastatic, 15% if metastatic). Median survival with metastases increased from 2.7 to 3.9 months, and from 13.3 to 15.9 months for non-metastatic PDAC between 2011 and 2015. Thirty-one percent of non-metastatic patients underwent pancreatectomy. About 1.5% were treated with neoadjuvant chemotherapy/chemoradiation. Of patients undergoing intended curative resection, 77% proceeded to adjuvant therapy. Fifty-one percent of metastatic PDAC patients never received anti-tumour therapy. Nearly one-fourth of surgically treated patients never received systemic therapy. More than two-thirds of non-metastatic patients never proceeded to surgery. Further consideration of neoadjuvant therapy should be given to borderline resectable patients. Most patients with PDAC still die soon after diagnosis, but median survival is increasing.