The Prognostic Significance of Serum Beta-2 Microglobulin (sβ2m) Levels in Patients with Hodgkin Lymphoma (HL): Final Analysis on 915 Patients Treated with ABVD or Equivalent (ABVDeq) Chemotherapy or Combined Modality Therapy (CT/CMT) Focusing to the Determination of Optimal Cut-Offs

Abstract

Objective/Aim: Sβ2m is a well- established prognostic factor for several hematologic malignancies, but its role in HL is still controversial. We aimed to investigate the prognostic significance of sβ2m levels in a large series of homogenously treated HL patients. Methodology: We analyzed 915 patients with HL treated with ABVDeq (1990-2019) and selected solely based on the availability of pretreatment sβ2m levels. Sβ2m levels (upper normal limit 2.4mg/L) were analyzed in relation to other baseline features and the outcome. Freedom From Progression (FFP) was defined as time between treatment initiation and treatment failure (toxic death, primary refractoriness, PR with switch to alternative chemotherapy or relapse); deaths of unrelated causes were censored. Overall Survival (OS) and Disease-Specific Survival (DSS) were measured from treatment initiation to death of any cause or HL-related causes respectively. Sequential cut-offs (1.8-3.5 by 0.1mg/L increments) were used to explore the potential impact of sβ2m on outcomes. Results: The median serum β2m levels were 2.20mg/L (IQR 1.80-3.00, range 0.50-14.40) and 383/915 patients (42%) had elevated levels (>2.4mg/L). Sβ2m correlated strongly with all baseline features. Univariate Analysis: FFP was significantly inferior in patients with higher β2m at all tested cut-offs. At 2.4 mg/L ("normal versus high") the 10-year FFP was 80% vs 70% (p=0.001) after a median follow-up of 81.1 months. However, the best cut-off was 2.0mg/L [10-year FFP 83% vs 70% (p=0.001)]. A dose-response relationship was seen across quartiles Q1-4 with 10-year FFP 84%,78%,73% and 68% (p=0.001). In early stages (IA/IIA), statistically significant results were obtained at cut-offs between 1.8 and 2.1mg/L. The best cut-off was 1.9mg/L [10-year FFP 88% vs 78% (p=0.003)]. In advanced stages, none of the cut-offs yielded statistically significant results (borderline at 2.0mg/L; 10-year FFP 74% vs 64%, p=0.09). Multivariate Analysis: Sβ2m levels remained significant for FFP after adjustment for IPS factors, ESR and B-symptoms at 2.0mg/L [hazard ratio (HR) 1.55, p=0.01) in the whole series of 915 patients; it was not significant at the "normal versus high" comparison at 2.4mg/L. In early stages, sβ2m was a significant predictor of FFP at both cut-offs of 2.0 mg/L and 2.4 mg/L (HR 1.65, p=0.034 and 1.67, p=0.038). In advanced stages, sβ2m was an independent prognostic factor for FFP at 2.0mg/L (HR 1.44, p=0.098 despite the lack of univariate significance), but not at 2.4mg/L. The 10-year OS and DSS was lower in patients with elevated β2m levels (10-year rates 90% vs 77%, p<0.001 and 93% vs 86%, p=0.002). Similar results to FFP were obtained by multivariate analysis of DSS for all 915 and early-stage patients. Conclusion: Higher sβ2m was a significant independent predictor of FFP in HL but the optimal cut-off appears to lie within the normal limits, performing better than a "normal versus high" evaluation (cut-off 2.4 mg/L) and being 2.0mg/L for the whole series and 1.9 mg/L for early-stage patients. The prognostic significance in advanced stages was weaker (best cut-off 2.0 mg/L). Serum β2m was also highly predictive of OS and DSS. An analysis of corrected serum β2m according to renal function is ongoing.