Abstract
Despite advancements in antiretroviral therapy, AIDS-related diffuse large B-cell lymphoma (AR-DLBCL) remains a major cause of morbidity and mortality. Compared to non-HIV-infected individuals, AR-DLBCL presents with considerable disease heterogeneity, which impairs the accuracy of current prognostic tools. This study aims to develop a novel prognostic model to enhance risk assessment for AR-DLBCL. We retrospectively analyzed 90 AR-DLBCL cases using univariate and multivariate analyses to identify clinical factors affecting overall survival (OS) and progression-free survival (PFS). A nomogram was created based on independent OS risk factors. The cohort had a median age of 43 years (range: 22-75), with 96.5% male patients. The median follow-up was 30 months (range: 1-139), with 5-year OS and PFS rates of 60.7% and 58.7%, respectively. Key prognostic factors for OS included decreased absolute lymphocyte count (p = 0.002), extranodal involvement (p = 0.005), reduced hemoglobin (Hb) levels (p = 0.004), Epstein-Barr virus (EBV) infection (p = 0.005), and elevated lactate dehydrogenase (LDH) levels (p = 0.018). The nomogram demonstrated robust predictive performance, with a 5-year receiver operating characteristic curve area under the curve of 0.949. Its C-index of 0.849 surpassed the International Prognostic Index (IPI) and age-adjusted IPI (aaIPI), which had C-index of 0.708 and 0.693, respectively. Additionally, the nomogram identified significant OS differences among low risk, intermediate-low risk, intermediate-high risk, and high-risk groups, with 5-year survival rates of 100%, 88%, 56%, and 8%, respectively. The model offers a personalized risk assessment for AR-DLBCL patients, facilitating precise prognosis prediction and informing individualized treatment strategies.
Introduction
Diffuse large B-cell lymphoma (DLBCL) is the most common form of AIDS-related lymphoma. While antiretroviral therapy (ART) has significantly improved outcomes, AIDS-related DLBCL (AR-DLBCL) remains highly aggressive and heterogeneous. Standard prognostic tools like the International Prognostic Index (IPI) and the age-adjusted IPI (aaIPI) were developed for the general population and often fail to account for the unique biological landscape of HIV patients, such as chronic immune dysfunction and viral co-infections. This study introduces a specialized prognostic tool designed to address these gaps by incorporating HIV-specific clinical markers.
Methodology
The researchers conducted a retrospective cohort study of 90 patients diagnosed with AR-DLBCL at Beijing Youan Hospital, Capital Medical University, between 2012 and 2024. The study identified independent risk factors for Overall Survival (OS) using univariate and multivariate Cox proportional hazards regression. These factors—absolute lymphocyte count (ALC), hemoglobin (Hb), Epstein-Barr virus (EBV) status, lactate dehydrogenase (LDH), and extranodal involvement—were used to construct a visual nomogram for predicting 1-, 3-, and 5-year survival probabilities. The model was internally validated using 1,000 bootstrap samples and compared against traditional indices using the Concordance index (C-index), Receiver Operating Characteristic (ROC) curves, and Decision Curve Analysis (DCA).
Results and Model Performance
The developed nomogram demonstrated exceptional predictive power, significantly outperforming traditional benchmarks. The model achieved a C-index of 0.849, compared to 0.708 for the IPI and 0.693 for the aaIPI. The 5-year OS area under the curve (AUC) reached 0.949, indicating high reliability in long-term predictions.
As shown in Figure 6, the nomogram integrates multiple clinical variables to provide a personalized risk score. Decision Curve Analysis (DCA) further confirmed that the nomogram provides a higher "net benefit" than traditional staging, making it a more effective tool for guiding clinical interventions across various risk thresholds. The model successfully stratified patients into four risk groups with 5-year survival rates ranging from 100% (low-risk) to just 8% (high-risk).
Discussion and Clinical Implications
A key finding of this study is the significance of the host's immune status. Reduced ALC emerged as a primary predictor of poor survival, highlighting that immune competence is as critical as tumor characteristics in AR-DLBCL. Furthermore, the inclusion of EBV status is vital, as EBV is found in a much higher proportion of AR-DLBCL cases (30–60%) than in non-HIV cases (~10%). By refining the IPI with markers like anemia (Hb) and ALC, the nomogram provides a more nuanced risk profile tailored to this specific demographic, allowing for clearer clinical decision-making regarding treatment intensity.
Limitations and Future Outlook
The study is limited by its retrospective nature and a relatively small sample size of 90 patients from a single center in Northern China. While internal validation was robust, the model requires testing on independent, external cohorts to confirm its generalizability across different global populations and HIV subtypes. Future research should also explore the integration of molecular genetic markers and psychosocial factors to further enhance the precision medicine approach for AR-DLBCL.
Summary updated