Retrospective Analysis of Prognostic Factors for Waldenström Macroglobulinemia: A Multicenter Cooperative Study in Japan

Background: Analysis of prognostic factors and clinical trials of novel agents for Waldenstrӧm macroglobulinemia (WM) are ongoing in Western countries, but few studies of WM have been performed in Japan. As a step toward future investigations, we retrospectively analyzed clinical features and prognostic factors in Japanese patients with WM.

Methods: We retrospectively analyzed clinical and laboratory characteristics, treatment and outcomes of 110 patients with WM, IgM-MGUS or lymphoplasmacytic lymphoma (LPL) diagnosed from January 2001 to March 2013 at 12 institutes. Overall survival (OS) was analyzed using Kaplan-Meier methods and survival was compared using log-rank testing. Several clinical characteristics at diagnosis were assessed by Cox regression for uni- and multivariate analysis for OS.

Results: Median age at diagnosis was 69 (range, 41-96) years, 73.6% were male, 12.0% had an ECOG performance status 2-4 and 6.4% presented with B-symptoms. Hyperviscosity, peripheral neuropathy, amyloidosis, cryoglobulinemia and cold agglutinin disease were shown in 9.1%, 4.5%, 1.8%, 4.5% and 2.7%, respectively. In 94 patients with available CT findings at diagnosis, lymphadenopathy, hepatosplenomegaly, pleural effusion, lung involvement, bone involvement and skin involvement were shown in 41.5%, 14.9%, 8.5%, 4.3%, 4.3% and 6.4%, respectively. Median serum monoclonal protein level was 2.62 g/dl (range, 0.70-9.35 g/dl). Symptomatic WM was present in 76 patients, asymptomatic WM in 23 and IgM-MGUS in 2 according to criteria of the Second International Workshop on WM. Seven patients showed IgG- or IgA-secreting LPL and 2 showed LPL without bone marrow infiltration. In patients with symptomatic WM, international prognostic scoring system for WM (ISSWM) was low in 9.2%, intermediate in 34.2%, high in 39.5% and unknown in 17.1%. Among patients with asymptomatic and symptomatic WM, watchful waiting was performed in 91.3% and 40.0%, respectively, with 61.9% and 36.7% remaining untreated, respectively. Median time to treatment from diagnosis of asymptomatic or symptomatic WM was 240 days (range, 3-1238 days) and 31 days (range, 0-2011 days), respectively. Oral alkylating agents were administered to 34.7% of patients with WM, 19.4% were treated with CHOP or CHOP-like regimen with or without rituximab, 8.2% received fludarabine mono- or combination therapy and 6.1% received rituximab monotherapy. Rituximab-containing therapy was administered as the initial treatment in 33.8% of patients who received treatment. Overall response rate (ORR) (complete + partial response rate) was 48.6%, and patients treated with rituximab-containing therapy displayed higher ORR (64.0%) compared to those with non-rituximab therapy (40.8%). Plasmapheresis was performed in 3.7% of patients. Three patients (2.7%) showed transformation to diffuse large B-cell lymphoma, and 7 (6.4%) developed second primary malignancies. Median follow-up was 38 months, 5-year OS rate for all patients was 74.9% (95% confidence interval (CI) 62.5-83.7) and rates for those with symptomatic WM, asymptomatic WM and other LPL were 66.0% (95%CI 50.6-77.6), 100% and 88.9% (95%CI 43.3-98.4), respectively. Significant differences in survival between risk groups of ISSWM in patients with symptomatic WM were not seen (5-year OS: high, 62.4%; intermediate, 64.3%; low, 75.0%; p=0.86). Although no significant difference in OS was observed compared to initial treatment (p=0.265), patients treated with rituximab during the observation period showed significantly prolonged OS compared to those treated without rituximab (5-year OS rates: 78.9% vs. 45.6%, p=0.036). In univariate analysis, age, pleural effusion, serum albumin, C-reactive protein and serum IgM levels were poor prognostic factors for OS. In multivariate analysis, age >65 years (hazard ratio (HR)=3.294; 95%CI 1.097-9.888, p=0.0336) and pleural effusion (HR=4.55; 95%CI 1.602-12.930, p=0.0045) were identified as significant prognostic factors for OS.

Conclusion: Prognostic factors for WM in Western countries may not be applicable to Japanese patients. This study suggested presence of pleural effusion at diagnosis is associated with poor clinical outcomes. Further investigations including histopathological examinations and molecular analyses are required to elucidate prognostic factors in Japan.