Climate-sensitive exposures and cancer outcomes: a systematic review and meta-analysis.

Climate change is intensifying environmental exposures - wildfire smoke, ambient air pollution, extreme heat, and weather-driven disasters - with plausible but incompletely quantified consequences for cancer, and no prior synthesis has jointly examined them across cancer incidence, mortality, survival, and recurrence. We systematically searched five databases for observational studies up to May 2026 of wildfire, ambient temperature, hurricane or multi-hazard disaster, precipitation anomaly, or ambient PM₂.₅ exposure and cancer outcomes, and pooled 25 effect estimates from 14 studies within ten exposure-outcome groups using random-effects models, with robust variance estimation (RVE) as the primary approach for groups containing dependent estimates from shared cohorts, alongside prespecified subgroup, meta-regression, sensitivity, and GRADE analyses. After accounting for within-cohort dependence the overall association was null (RVE 1.02 (0.93-1.11)); the uncorrected estimate (1.07 (1.02-1.12)) was inflated by correlated effects and extreme heterogeneity (I2 = 97%). Group-specific signals diverged: wildfire exposure showed a non-significant elevation in cancer mortality (RVE 1.11 (0.89-1.38)) and a borderline association with incidence from a single national cohort, ambient PM₂.₅ was associated with poorer survival, and higher temperature appeared protective for survival under the conventional model but not under RVE, most consistent with residual confounding. Certainty was Low for two groups and Very low for the remaining eight. This directionally concerning but statistically fragile, low-certainty evidence does not support a single overall effect; wildfire and ambient-PM₂.₅ carcinogenicity and disaster-related disruption of cancer care are the priorities for adequately powered, dependence-aware research within a planetary-health framework.