NUTRITIONAL PARAMETERS AS PROGNOSTIC MARKERS IN OLDER PATIENTS WITH DIFFUSE LARGE B-CELL LYMPHOMA: A POST-HOC ANALYSIS OF THE ELDERLY PROJECT BY FONDAZIONE ITALIANA LINFOMI
Salvatrice Mancuso et al
University of Palermo, Palermo, Italy
Background
- Diffuse Large B Cell Lymphoma (DLBCL) is the most common lymphoma in the elderly, with variable response to immunochemotherapy.
- Several prognostic systems have been developed to guide treatment. In this setting, comprehensive geriatric assessment (CGA) enables risk stratification and correlates with outcomes, with nutritional status representing a key evaluated component.Albumin, Body Mass index (BMI) and Geriatric Nutritional Risk Index (GNRI) are indices that reflect nutritional status and have been variably employed across different disease settings to explore their prognostic significance.
- To date, few studies have investigated the prognostic potential of these parameters in the elderly population undergoing treatment for DLBCL.
- The aim of the study is to evaluate the predictive value of albumin, BMI and GNRI on outcomes of patients aged > 65 years undergoing first-line therapy for DLBCL.
Methods
- This is an ancillary analysis of Elderly Project of the Fondazione Italiana Linfomi, a prospective multicenter observational study aimed at validating a simplified geriatric assessment tool and a new prognostic score, the EPI. In this study, we selected patients whose complete anthropometric and laboratory parameters were available at baseline.
- The primary endpoint was the correlation between serum albumin levels, BMI and GNRI and overall survival.
Results
- Based on the inclusion criteria, 688 out of total of 1207 patients (59%) were deemed eligible, with median age of 75 years. In univariate analysis, albumin showed statistical correlation with 3-yr OS (HR, 0.43; 95% CI, 0.34-0.55; P<0.001). In addition, patients with serum albumin concentration<3.5 g/dl presented 3-yr OS of 56% vs 73% of subjects with values >3,5 gr/dl (P<0.001). In a multivariate Cox proportional hazard regression, we observed a HR of 0.58 (95% CI, 0.44-0.75, P<0.001). As regards BMI, it correlates with 3-yr OS (p-value: 0.003, HR: 0.95, 0.91-0.96 CI).
- With a cut-off value of 25, patients with BMI < 25 showed worst survival compared with BMI ≥25 (p 0.025). In a multivariate Cox regression analysis, BMI was an independent prognostic factor with an HR of 0.96 (95% CI, 0.92-0.99; P= 0.002).
- The authors have identified four groups of patients stratified according to GNRI values in relationship to risk of death: none risk (GNRI >98), low risk (GNRI 92-98), moderate (GNRI 82-92), high risk (GNRI <82). Regarding GNRI as continuous variable, there was a significant correlation with 3-yr OS (P<0.001, HR 0.65, 95% CI 0.58-0.74).
- A correlation was also observed between different GNRI patients groups and 3-yr OS (Tab.1). Finally, high statistical association of albumin, BMI and GNRI with EPI was observed.
Conclusion
- In the current therapeutic landscape, treatment personalization requires accurate risk stratification. In elderly patients with DLBCL undergoing antineoplastic treatment, albumin, BMI and GNRI represent independent prognostic parameters for survival, which are simple and easily applicable