Multi-omics profiling reveals homologous recombination repair hyperactivation as a therapeutic vulnerability in neuroendocrine cervical carcinoma

Neuroendocrine cervical carcinoma (NECC) is an aggressive and rare malignancy comprising 0.9-1.5% of cervical cancers. Its molecular basis remains poorly understood, limiting treatment options. A national multicenter cohort survival analysis showed that NECC patients had a 5.17-fold higher death risk than non-NECC patients and demonstrated broad chemotherapy resistance, which was further validated in both NECC cell lines and patient-derived organoids. To identify therapeutic targets, multi‑omics profiling of NECC tumors and matched normal adjacent tissues revealed pronounced DNA replication stress and hyperactivation of homologous recombination repair (HRR). These findings were validated in cells and patient-derived organoids using DR-GFP reporter and pRPA32 assays, suggesting that NECC may be susceptible to PARP inhibitor therapy. NECC cells exhibited moderate PARP inhibitor sensitivity, greater than that of resistant non‑NECC lines but less than that of highly sensitive A2780 cells. However, combining PARP inhibitors with cisplatin/carboplatin did not enhance efficacy, prompting the exploration of alternative combinatorial targets. Multi-omics analysis and in vitro validation identified Bloom syndrome helicase (BLM) as a key factor overexpressed in NECC that modulates PARP inhibitor sensitivity. Further in vitro and in vivo experiments confirmed that combining PARP and BLM inhibitors achieved potent anticancer effects. Mechanistically, cancer-associated fibroblasts (CAFs) infiltration in the tumor microenvironment upregulated the transcription factor E2F1 via IL6, with ChIP‒qPCR verifying that activated E2F1 directly binds the BLM promoter (hg38, chr15:90,715,346-90,717,346) to drive its expression, suppressing PARP inhibitor sensitivity. Our findings establish dual PARP and BLM inhibition as a promising therapeutic strategy for NECC by blocking HRR pathway compensation.