Biom-01. prognostic implications of small cell lung cancer (sclc) transcriptional subtyping for central nervous system (cns) metastases

Abstract BACKGROUND Small cell lung cancer (SCLC) is characterized by a high propensity for brain metastases (BM) and is associated with a poor prognosis. Subtypes characterized by high versus low neuroendocrine states (expressing ASCL1 or NEUROD1) have been identified. This study aims to evaluate the influence of transcription factor-defined SCLC subtypes on the incidence and outcomes of BM. METHODS We identified 164 patients with SCLC based on their ASCL1 and NEUROD1 expression status by immunohistochemistry. Four transcriptional subtypes were defined: ASCL1+/NEUROD1- (A+/N-), ASCL1+/NEUROD1+ (A+/N+), ASCL1-/NEUROD1+ (A-/N+), and ASCL1-/NEUROD1- (A-/N-). Cumulative incidence competing risk analyses examined the incidence of central nervous system (CNS) progression. Cox proportional hazards models were utilized to assess both overall survival (OS) and CNS progression-free survival (CNS-PFS). RESULTS Out of 164 patients, the majority fell into either A+/N- or A+/N+ subtypes (n = 62, n = 63, respectively). BM were detected at the time of diagnosis in 12% (20/164) of patients, while 18% (29/164) developed BM later. In two distinct cohorts of patients (initially free from BM and with de novo BM at diagnosis), the 12-month cumulative incidence of CNS progression was numerically highest for the A+/N- subtype (HR = 11.8, 95%CI: 4.7-22.4, and HR = 44, 95%CI: 10.5-74.7, respectively). OS in both A+/N- and A-/N+ subtypes was worse compared to the A+/N+ subtype (HR = 1.62, 95%CI: 1.03-2.56, and HR = 1.63, 95%CI: 1.06-2.60, respectively). Similarly, CNS-PFS was poorer in same A+/N- and A-/N+ subtypes (HR = 3.10, 95%CI: 1.39-6.93, and HR = 3.38. 95% CI: 1.49-7.67) when compared to A+/N+. CONCLUSIONS This study is the first to explore CNS-specific outcomes based on transcription factor SCLC subtypes. Among patients without BM at SCLC diagnosis, those exclusively expressing ASCL1 or NEUROD1 demonstrated poorer outcomes than those with co-expression. Further investigation is warranted to elucidate the implications of SCLC subtyping, ultimately facilitating tailored disease management.