Abstract Background Ki-67 labeling index (LI) is widely used to quantify tumor proliferation in cancer. Although helpful in assessing proliferative activity across gliomas, its prognostic value in glioblastoma remains uncertain. Methods We conducted a systematic prognostic review and meta-analysis of studies reporting associations between Ki-67 LI and overall survival in adult glioblastomas. Random-effects models (restricted maximum likelihood with Knapp–Hartung adjustments) were used to pool hazard ratios (HRs). Risk of bias was assessed with the QUIPS tool. Results Fourteen retrospective cohort studies (n = 1,097) met the inclusion criteria. Across all studies, higher Ki-67 LI was associated with an increased hazard of death (univariable pooled HR 1.73, 95% confidence interval (CI) 1.33–2.25; P = .001; multivariable pooled HR 2.23, 95% CI 1.67–2.96; P = .0001). Prediction intervals were 0.99–3.02 (univariable) and 1.11–4.48 (multivariable), suggesting that future multivariable studies are more likely to observe HRs 1 in patients with higher Ki-67 LI. Between-study heterogeneity was low to moderate (I2 = 24.1% univariable; I2 = 40.3% multivariable). QUIPS assessments highlighted greatest concerns for selective study participation and heterogeneous Ki-67 cutoffs, while outcome measurement, confounder adjustment, and statistical reporting were generally lower risk. Conclusions Higher Ki-67 LI is associated with an increased hazard of death in patients with glioblastoma, with effects persisting after multivariable adjustment and limited between-study heterogeneity. Standardized Ki-67 assessment and clinically validated cutoffs are needed to improve comparability and enable robust prognostic stratification.
Nordahl et al. (Thu,) studied this question.