BackgroundThe role of the nuclear protein survivin in non-small cell lung cancer (NSCLC) remains controversial. To clarify the association of survivin with survival in NSCLC, we performed a meta-analysis of the literature with meta-analysis. MethodsTrials were selected for further analysis if they provided an independent assessment of nuclear survivin in NSCLC and reported the survival data in the context of survivin status. A total of 7 trials, which comprised 823 patients, were included in the meta-analysis. Sensitivity analyses were conducted using the patient's disease stage, IHC cutoff value, and ethnicity. ResultsThe combined hazard ratio (HR) of 1.54 [95% CI, 0.79–3.02; test for heterogeneity p<0.001] suggests that high nuclear survivin expression has no impact on patient survival. However, when the studies were restricted to Caucasian patients, high levels of nuclear survivin expression were correlated with reduced survival (HR 2.38, 95% CI=1.60–3.52; p=0.189 for heterogeneity). In addition, the heterogeneity disappeared when the analysis was restricted to Caucasians. ConclusionNuclear survivin expression is associated with poor prognosis for Caucasian NSCLC patients.