Case Report: Glioblastoma in cohabiting partners: a molecularly characterized temporal cluster.

BACKGROUND: Glioblastoma is the most lethal primary brain tumor of adults, with an age-standardized incidence of 5-7 per 100,000 person-years and a median survival of approximately 15 months despite maximal safe resection, radiotherapy, and temozolomide. Its occurrence in two unrelated cohabiting individuals within a short interval is an uncommon event that inevitably raises the question of shared causation, one that available retrospective methods rarely resolve. CASE PRESENTATION: Two unrelated adults from northern Italy, partners sharing a household since 2018 and a workplace since 2016, developed glioblastoma approximately 20 months apart. Case 1 was a 47-year-old woman with giant cell glioblastoma, IDH-wildtype, central nervous system (CNS) WHO grade 4, O6-methylguanine-DNA methyltransferase (MGMT) promoter unmethylated and telomerase reverse transcriptase promoter (pTERT)-negative. She developed recurrence with gliosarcomatous features and acquired homozygous cyclin-dependent kinase inhibitor 2A/B (CDKN2A/B) deletion, and died following craniospinal leptomeningeal dissemination. Case 2 was a 51-year-old man with biopsy-proven glioblastoma, IDH-wildtype, CNS WHO grade 4, meeting WHO 2021 criteria on both histologic grounds (focal necrosis) and molecular grounds (pTERT C250T mutation). EXPOSURE ASSESSMENT: A structured retrospective exposure-history questionnaire was administered to Case 2 during his diagnostic hospital admission. The assessment documented a shared office-based occupational and residential context without identifying a specific carcinogenic exposure. Multiple relevant domains, including radon, indoor air quality, household and water chemistry, remained unmeasured and should not be interpreted as negative findings. CONCLUSION: This report documents a temporal cluster without implying shared causation. Its primary contribution is the detailed molecular characterization of both cases, including recurrence profiling in Case 1, and a transparent, reproducible exposure-history framework that distinguishes absent from unmeasured data.