Chemotherapy-induced neutropenia as a prognostic factor in pancreatic cancer: a systematic review and meta-analysis.

BACKGROUND

Chemotherapy-induced neutropenia (CIN) may reflect higher pharmacodynamic exposure and relate to improved outcomes, but its clinical relevance in pancreatic cancer remains uncertain.

METHODS

PROSPERO-registered, PRISMA-compliant systematic review and meta-analysis. PubMed, Embase, and CENTRAL were searched to 18 August 2025. Eligible studies compared OS in CIN ≥G3 vs <G3. Quantitative synthesis was limited to advanced/unresectable/metastatic disease. Log(HR) and SEs were pooled with random effects (DerSimonian - Laird); heterogeneity by Q and I^2. Sensitivity analyses included 95% prediction interval, Hartung - Knapp - Sidik - Jonkman, and leave-one-out. Publication bias was not assessed (k = 5).

RESULTS

Eight observational studies met inclusion; five (all advanced disease) were meta-analyzed. CIN ≥G3 was associated with better OS (grouped HR 0.50; 95% CI 0.40-0.62; I2 = 0%), with consistent direction across studies. Sensitivity analyses confirmed robustness. ROBINS-I rated four studies at serious and four at moderate risk of bias; GRADE certainty for the pooled OS effect was moderate.

CONCLUSIONS

In advanced pancreatic cancer, CIN ≥G3 is associated with improved OS, supporting its role as a cohort-level prognostic/pharmacodynamic marker. Generalizability is limited by predominantly Japanese cohorts. Prospective multicenter studies with time-dependent modeling and control of relative dose intensity.

REGISTRATION: www.crd.york.ac.uk/prospero, identifier CRD420251005347.