BackgroundDespite receiving alectinib therapy, patients with ALK-rearranged non-small-cell lung cancer (NSCLC) remain at risk of central nervous system (CNS) progression. Our retrospective study aimed to identify baseline clinical and molecular factors associated with the risk of CNS progression in this patient subset. MethodsWe analyzed the clinical, molecular, and imaging data of 318 patients with ALK-positive advanced NSCLC who received alectinib as first-line (1L-alectinib) or second-line (2L-alectinib) therapy at baseline (1L, n=183; 2L, n=135) and at disease progression (1L, n=57; 2L, n=75). ResultsThe incidence rates of CNS progression were 12.3% after 1L-alectinib and 30.7% after 2L-alectinib. Compared with patients who received 1L-alectinib, CNS progression was significantly higher in patients who received 2L-alectinib (p=0.04), particularly those who had oligoprogression (p=0.01). Oligoprogression was detected in 55.0% (44/80) of patients who progressed after first-line alectinib, with the remaining 45.0% (36/80) had non-oligoprogression. Univariate and multivariate analyses and stepwise regression analyses consistently identified a higher likelihood of CNS progression among (i) patients who received 2L-alectinib than 1L-alectinib; (ii) patients with non-3a/b variant ALK fusion than those with EML4-ALK variant 3a/b; and (iii) patients with PD-L1 tumor proportion score (TPS) ≥50% than PD-L1 TPS<50%. ConclusionOur study provided real-world evidence that patients who harbor PD-L1 TPS≥50% were more likely to experience CNS progression during alectinib therapy. The association between CNS progression and breakpoint variants warrants further investigation. Our findings suggest that close monitoring and prompt intervention are crucial in prolonging the quality of life of this patient subset.