As treatment outcomes improve in diffuse large B-cell lymphoma (DLBCL), attention is increasingly shifting toward optimizing disease management across the continuum, from frontline therapy selection to relapse detection and long-term risk assessment. Three recent analyses highlight progress in refining surveillance tools, understanding evolving prognostic factors, and improving treatment strategies in both newly diagnosed and high-risk populations.
One key area of focus is improving post-remission surveillance. Despite achieving complete remission with first-line therapy, a substantial proportion of patients with DLBCL experience relapse, often detected only after the disease burden has significantly increased. A systematic review and meta-analysis evaluated the use of circulating tumor DNA (ctDNA) as a minimally invasive tool for detecting relapse earlier.1
Across 3 studies including 157 patients, ctDNA demonstrated high specificity (93.9%) for detecting relapse, meaning that a positive result strongly indicates disease recurrence.1 However, sensitivity was more limited at 58.6%, indicating that a significant proportion of relapses may not be detected using ctDNA alone. Importantly, ctDNA positivity often preceded or coincided with clinically recognized relapse, suggesting a potential utility for earlier intervention. These findings support ctDNA as a valuable adjunct to surveillance, although not as a replacement for imaging and clinical assessment.1
Risk stratification is also evolving, particularly in patients with extranodal disease. A large population-based study examining more than 30,000 patients found that the prognostic impact of primary extranodal site changes over time.2 Although certain sites may appear favorable at diagnosis, their relative risk can shift as patients survive longer.
Central nervous system (CNS) involvement consistently emerged as a high-risk feature (12-month hazard ratio [HR], 2.46; 95% CI, 2.29-2.65), conferring more than double the risk of death compared with gastrointestinal involvement as reference, both at diagnosis and at subsequent landmark analyses.2 In contrast, the prognostic significance of other sites, such as bone, head and neck, breast, and skin, evolved over time, with some becoming more relevant in longer-term survivors. These findings underscore the importance of dynamic, time-dependent risk assessment rather than relying solely on baseline characteristics.2
At the same time, efforts continue to optimize frontline therapy for patients with high-risk disease. A large systematic review and meta-analysis of 66 studies and 3633 patients compared intensified immunochemotherapy regimens, such as rituximab, etoposide, prednisone, vincristine, cyclophosphamide, and doxorubicin (R-EPOCH), with standard rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP) in high-risk DLBCL populations.3
Although overall response rates were similar between regimens, intensified therapy was associated with higher complete response rates (R-EPOCH, 70% vs R-CHOP, 62%; P=.044) and improved survival outcomes.3 Specifically, R-EPOCH was linked to better progression-free (HR, 0.48; 95% CI, 0.39-0.60) and overall survival (HR, 0.60; 95% CI, 0.42-0.83), with notable improvements in 2-year outcomes. These benefits appeared particularly pronounced in patients with high-risk disease features, although limitations in available data and potential selection bias highlight the need for prospective validation.3
Together, these studies reflect a broader shift toward more precise and adaptive management in DLBCL. Advances in ctDNA monitoring offer the potential for earlier detection of relapse, whereas evolving insights into extranodal disease highlight the importance of time-dependent prognostication. Meanwhile, intensified treatment strategies may offer improved outcomes for patients with high-risk disease, although optimal selection remains an open question.
As these approaches continue to evolve, integrating improved surveillance tools, dynamic risk assessment, and tailored treatment strategies will be essential to further improving outcomes in DLBCL.