Changes over time in the course of advanced pancreatic cancer treatment with systemic chemotherapy: a pooled analysis of five clinical trials from two decades of the German AIO study group

Background: Over the past two decades, our group has conducted five multicenter trials focusing on first-line systemic therapy for patients with advanced pancreatic cancer. The current pooled analysis was designed to evaluate prognosis over time and the impact of clinical characteristics on survival.

Patients and methods: Individual patient data were derived from five prospective, controlled, multicenter trials conducted by the 'Arbeitsgemeinschaft Internistische Onkologie' (AIO): 'Gem/Cis', 'Ro96', 'RC57', 'ACCEPT' and 'RASH', which recruited patients between December 1997 and January 2017.

Results: Overall, 912 patients were included. The median overall survival (OS) for all assessable patients was 7.1 months. OS significantly improved over time, with a median OS of 8.6 months for patients treated from 2012 to 2017 compared with 7.0 months from 1997 to 2006 [hazard ratio (HR) 1.06; P < 0.004]. Eastern Cooperative Oncology Group performance status (HR 1.48; P < 0.001), use of second-line treatment (HR 1.51; P < 0.001), and Union for International Cancer Control (UICC) stage (III versus IV) (HR 1.34, P = 0.002) had a significant impact on OS. By contrast, no influence of age and gender on OS was detectable. Comparing combination therapy with single-agent chemotherapy did not demonstrate a survival benefit, nor did regimens containing epidermal growth factor receptor-tyrosine kinase inhibitors (EGFR-TKIs) such as afatinib or erlotinib, compared with chemotherapy-only arms. Patients with early-onset pancreatic cancer (age at study entry of ≤50 years, n = 102) had a similar OS compared with those >50 years (7.1 versus 7.0 months; HR 1.13; P = 0.273). The use of a platinum-containing regimen was not associated with better outcomes in patients with early-onset pancreatic cancer.

Conclusions: Within this selected group of patients treated within prospective clinical trials, survival has shown improvement over two decades. This effect is likely attributable to the availability of more effective combination therapies and treatment lines, rather than to any specific regimen, such as those containing EGFR-TKIs. In addition, concerning age and sex subgroups, the dataset did not provide evidence for distinct clinical behavior.