Prognostic impact of comorbidity burden on overall survival in patients with Waldenström macroglobulinemia in the United States.

Abstract

7083 Background: Waldenström macroglobulinemia (WM), also known as lymphoplasmacytic lymphoma (LPL), is a rare indolent Non-Hodgkin B-cell lymphoma affecting approximately 1500 patients annually in the US. Existing prognostic models: IPSS-WM, revised IPSS-WM, and MSS-WM focus primarily on disease specific factors including age, albumin, hemoglobin, platelet count, β2-microglobulin, LDH, and monoclonal IgM at diagnosis, and were derived from small cohorts. However, comorbidity burden is not included in these models, despite data showing that mortality is frequently attributed to comorbidities unrelated to lymphoma. We conducted this analysis to identify prognostic factors with emphasis on the impact of comorbidity burden on overall survival in WM. Methods: In this IRB approved retrospective study, we analyzed 17130 patients with histologically confirmed WM (ICD-O-3 9761/3) from the NCDB (2004–2023). Univariable and multivariable Cox regression identified independent prognostic factors for overall survival. Age stratified sub-analysis assessed the prognostic impact of comorbidity within each age group. Analyses were performed using SAS 9.4. Results: On multivariable Cox regression with backward elimination, Charlson comorbidity score (CC) ≥1 emerged as a significant independent predictor of mortality (HR 1.7 95% CI 1.58–1.75 p < 0.0001). Age-stratified analysis confirmed CC ≥1 significance in all age groups (HR 1.52-2.07, p<0.0001). Seven additional prognostic factors were identified: advanced age was associated with progressively worse survival (55≤age<65: HR 1.7; 65≤age<75: HR 2.3; age≥75: HR 5.5; all p<0.0001), while female sex was protective (HR 0.86; p<0.0001). Lower median household income was associated with worse outcomes (<$40,227: HR 1.38; $40,227–$50,353: HR 1.26; $50,354–$63,332: HR 1.1; all p<0.0001). Private insurance was associated with better OS compared with uninsured (HR 1.7; p<0.0001) and governmental insurance (HR 1.3; p<0.0001). Treatment at non-academic facilities (HR 1.13; p<0.0001) and multiple primary malignancies (HR 1.19; p<0.0001) were associated with inferior survival. No difference was observed between multi- and single-agent chemotherapy (HR 1.0; p=0.96), while patients not requiring chemotherapy had superior survival (HR 0.80; p<0.0001). Conclusions: In the largest WM cohort to date, we identified several prognostic factors impacting survival WM patients in the US. Among these factors, CC ≥1 was a strong independent predictor of mortality. Given that half of deaths in WM patients are due to comorbidities, comorbidity assessment is essential for accurate prognostication and individualized treatment planning. Managing comorbidities may be as important as treating the disease itself. CC or equivalent should be integrated into future prognostic models to improve clinical decision-making in WM.