Abstract
BACKGROUND: Although primarily reported in solid organ transplant recipients and patients undergoing chimeric antigen receptor T-cell immunotherapy (CAR-T), non-hepatic hyperammonemia (NHHA) is a rare but lethal complication in the broader context of post- chemo-immunotherapy hematologic malignancies. It often presents with unexplained encephalopathy that mimics primary central nervous system (CNS) progression, leading to diagnostic delays. With the expanding use of bispecific antibodies (e.g., glofitamab), the etiology of NHHA, particularly the complex interplay between opportunistic infections and potential metabolic susceptibility, remains poorly understood. CASE PRESENTATION: We report a fatal case of NHHA in a 58-year-old male with diffuse large B-cell lymphoma (DLBCL) following glofitamab-based chemo-immunotherapy. The patient developed sudden onset altered mental status with extreme hyperammonemia (peak blood ammonia 638.9 µmol/L) despite preserved liver function. Metagenomic next-generation sequencing (mNGS) of bronchoalveolar lavage fluid identified Ureaplasma urealyticum. Furthermore, post-mortem whole-exome sequencing (WES) identified a heterozygous variant of SLC25A13 (NM_014251.3:c.2 T > C). As biochemical confirmation of citrin deficiency was not available, the clinical significance of this variant remains uncertain, though it may represent a contributory metabolic susceptibility factor. Despite aggressive ammonia-lowering strategies, including continuous renal replacement therapy (CRRT) and targeted antibiotics, the patient succumbed to fulminant cerebral edema. CONCLUSION: This case highlights the Ureaplasma urealyticum infection as a critical precipitant of fatal NHHA following glofitamab therapy, occurring in the background of possible genetic metabolic susceptibility (an unverified heterozygous SLC25A13 variant of uncertain functional significance). These findings underscore the critical need for early blood ammonia monitoring and rapid mNGS screening in immunocompromised patients with unexplained encephalopathy. We propose a structured diagnostic algorithm to expedite the recognition and management of this reversible yet life-threatening condition.