Clinical Characteristics and Prognostic Prediction of Secondary Solid Malignancies in Patients With Diffuse Large...

Secondary primary solid malignancies (SPSMs) significantly impact long-term outcomes in lymphoma patients. However, subtype-specific differences remain unclear, such as diffuse large B-cell lymphoma (DLBCL) and follicular lymphoma (FL).

This study collected 1,377 DLBCL and 489 FL patients, identifying 50 DLBCL and 31 FL cases with SPSMs. Clinical characteristics, SPSM types, and survival were compared in these 81 patients. Demographic, clinical, treatment-related variables (including radiotherapy for primary lymphoma, recorded as yes/no), and survival outcomes were collected. Overall survival (OS) was analyzed using Kaplan-Meier estimates and Cox proportional hazards models. The cumulative incidence of SPSMs was evaluated with competing risk analysis (death as a competing event), and group comparisons were performed with Gray's test. Prognostic factors identified in univariable analysis (p < 0.05) were included in a multivariable Cox model. A nomogram was developed, with discriminative ability assessed by the area under the receiver operating characteristic curve (ROC). Statistical significance was set at p < 0.05.

SPSMs were observed more frequently in FL (6.34%) than DLBCL (3.63%). Thyroid cancer predominated (22.2%, 18/81). DLBCL patients developed SPSMs earlier than FL patients (median 28.47 vs. 41.77 months, p = 0.031), though cumulative incidence accounting for competing risks did not differ significantly (Gray's test p = 0.34). DLBCL patients with SPSMs had inferior OS compared to FL patients (p = 0.04). Non-GCB DLBCL showed greater SPSM diversity and survival disadvantage. Multivariable analysis identified FL (vs. DLBCL) subtype (HR = 0.328, p = 0.018), bone marrow infiltration (HR = 2.815, p = 0.014), initial radiotherapy before SPSM diagnosis (HR = 3.475, p = 0.005), and shorter time to SPSM development (HR = 0.973 per month, p = 0.021) as independent prognostic factors for worse OS. The nomogram model showed acceptable discrimination, with an AUC of 0.761 in the time-dependent ROC analysis.

This study provides novel insights into SPSM characteristics and prognostic differences in DLBCL and FL patients within a Chinese cohort. SPSMs in FL patients were observed more frequently, while DLBCL patients, particularly those with non-GCB subtypes, experience earlier SPSM onset and poorer survival. The validated nomogram, incorporating lymphoma-related factors, enables personalized risk stratification. However, the single-center design, small sample size (n = 81), and lack of SPSM-specific data limit generalizability, necessitating multi-center studies with comprehensive SPSM characterization for validation.