PURPOSE Mutations in TP53 , detected in over 20% of diffuse large B-cell lymphomas (DLBCLs), are associated with poor prognosis. However, clinical outcomes among patients with TP53 -mutant disease vary, with some patients showing treatment responses similar to those with wild-type TP53 . This study aims to understand the clinical and molecular determinants underlying poor outcomes in TP53 -mutant DLBCL. METHODS Clinical and molecular data for 3,091 patients were derived from 10 cohorts of patients with newly diagnosed DLBCL treated with frontline rituximab-based immunochemotherapy regimens. Targeted or whole-exome/whole-genome sequencing was available for all patients. Bulk RNA-seq was analyzed for 591 patient samples. The primary outcome measures were progression-free survival (PFS) and overall survival (OS). RESULTS TP53- mutant DLBCL differed from wild-type disease in pattern and number of genetic lesions, malignant B-cell expression states, and tumor microenvironment composition. TP53 mutations were 6-fold more prevalent than MYC/BCL2/BCL6 double-/triple-hit status, but conferred similar adverse prognostic risk. Among patients with TP53 -mutant disease, variant allele frequency (VAF) further stratified risk, with patients featuring VAF ≥ 75% (indicative of loss of heterozygosity) experiencing significantly inferior PFS/OS. Downregulation of interferon signaling and lower macrophage content were identified in TP53 -mutant samples derived from patients with poor outcomes or VAF ≥ 75%. TP53 mutations were adversely prognostic among patients with DLBCL assigned to specific LymphGen subtypes (EZB, MCD), malignant B-cell states (S1), and ecotypes (LE4, LE7, LE8), whereas outcomes were similar to wild-type disease within other molecular subtypes. In re-examination of the Phoenix trial data, addition of ibrutinib to R-CHOP improved PFS in patients with TP53 -mutant DLBCL and abrogated the deleterious impact of high VAF, irrespective of patients' age. CONCLUSION The poor prognosis of TP53 -mutant DLBCL is dependent on intrinsic features, such as VAF, and modulated by co-occurring genomic lesions or lymphoma cell-intrinsic or microenvironmental expression patterns.