Introduction
Diffuse Large B-Cell Lymphoma (DLBCL) is the most common aggressive non-Hodgkin lymphoma. While standard R-CHOP therapy (rituximab plus chemotherapy) cures approximately 60% of patients, the remaining 30–40% develop relapsed or refractory (r/r) disease. For these patients, the prognosis is often poor, and traditional tools like the International Prognostic Index (IPI) are less effective in the modern post-rituximab era. This study aimed to identify key predictors and construct a prognostic nomogram for risk stratification in r/r DLBCL patients who failed initial R-CHOP treatment.
Methods
The researchers conducted a retrospective study of 296 r/r DLBCL patients treated at the Cancer Hospital, Chinese Academy of Medical Sciences, between 2006 and 2017. A total of 231 eligible patients were randomly assigned to a primary cohort (n=161) for model construction and a validation cohort (n=70) for testing. Multivariate Cox proportional hazards regression was used to identify independent prognostic factors for Overall Survival (OS). Additionally, ordinal regression analysis was employed to identify factors predicting the efficacy of second-line therapy.
Key Results and Nomogram Development
The final prognostic model integrated four independent clinical variables:
- Rituximab in second-line treatment: Surprisingly, including rituximab in salvage therapy remained beneficial even after initial R-CHOP failure.
- Initial ECOG Performance Status (PS): Baseline physical status was a significant predictor of long-term outcomes.
- Response to front-line treatment: Patients who achieved Complete or Partial Remission (CR/PR) initially had significantly better survival than primary refractory patients (SD/PD).
- Invasion pattern on recurrence: Patients with cancer appearing in entirely new sites fared better than those with recurrence in original sites or both.
The resulting nomogram demonstrated solid predictive accuracy, achieving a C-index of 0.70 and AUC values of 0.73 and 0.71 in the primary and validation cohorts, respectively.
Risk Stratification and Survival Outcomes
Using the nomogram scores, patients were stratified into three distinct risk groups with significantly different survival profiles:
- Low Risk: Median Overall Survival (OS) of 38.0 months.
- Intermediate Risk: Median OS of 25.0 months.
- High Risk: Median OS of only 7.0 months.
Furthermore, the study identified that a prior positive response to front-line treatment and the absence of bulky disease at diagnosis were the two strongest independent predictors of second-line treatment success.
Discussion and Limitations
While the nomogram provides a valuable tool for risk stratification, the study has limitations. As a retrospective, single-center study, the findings may be influenced by selection bias. Additionally, the rate of Autologous Stem Cell Transplant (ASCT) was lower than in many Western cohorts, which may affect generalizability. Future research should focus on validating these findings in larger, multi-center trials and incorporating molecular biomarkers to further refine the model's accuracy.
Conclusion
This novel nomogram offers a practical roadmap for personalized treatment in the post-rituximab era, helping clinicians identify high-risk patients who may require more aggressive or experimental therapeutic interventions.
Summary updated