Streptozotocin-based chemotherapy in advanced PanNETs: a systematic review and meta-analysis

BACKGROUND: Streptozotocin (STZ)-based chemotherapy remains a treatment option for advanced pancreatic neuroendocrine tumours (PanNETs), particularly when tumour shrinkage is a therapeutic priority. However, its activity has been reported across heterogeneous treatment schedules, patient populations, and response-assessment eras. We conducted a systematic review and meta-analysis to estimate objective response rate (ORR) and landmark progression-free survival (PFS). METHODS: We systematically reviewed randomised trials, prospective studies, and retrospective cohorts reporting ORR and/or PFS with STZ-containing regimens. Random-effects meta-analyses were conducted in a primary ELIGIBLE dataset restricted to advanced or metastatic pancreatic-only cohorts with standardised outcome assessment, and in a broader ALL dataset for contextual analyses. Comparisons across regimen families were exploratory. RESULTS: Twenty-seven studies including 1,744 patients were included. In the ELIGIBLE dataset, pooled ORR was 36.1% (95% CI 30.6-42.0), with a 95% prediction interval of 23.0-51.0%. Landmark PFS was 59.2% (50.6-67.2) at 12 months and 41.5% (33.4-50.1) at 24 months. Estimates from the ALL dataset were directionally consistent but more heterogeneous. No statistically significant differences were detected between regimen families. Certainty of evidence was very low, mainly because of risk of bias and inconsistency. CONCLUSION: In cohorts using standardised outcome assessment, STZ-containing regimens were associated with objective responses in approximately one-third of patients, supporting their consideration when tumour shrinkage is a major therapeutic objective. These single-arm pooled estimates do not establish superiority among STZ schedules or relative to other treatments, and the PFS benchmarks require cautious interpretation because of substantial clinical and methodological heterogeneity.