Prognostic Impact of Clinical and Tumor Associated Variables in a Population-Based Cohort of Mantle Cell Lymphomas in the Stockholm Region Between 1998–2010

Abstract

1592 Introduction: Mantle cell lymphoma (MCL) constitutes 3–10% of non-Hodgkin lymphomas and affects predominantly middle-aged to elderly men. The median survival is 3–5 years and seems to improve with new therapeutic regimens. The MCL International Prognostic Index (MIPI) has been proven useful for predicting survival in MCL patients included in clinical trials, but its value in unselected population based MCL cohorts is less well known. Biological markers are increasingly used for prognostication of MCL patients, especially for defining indolent cases. Material and Methods: All 186 patients diagnosed with MCL, confirmed by IHC for cyclinD1 and/or by FISH for t(11;14), between January 1998 and June 2010 in the Stockholm region, were included in a retrospective analysis. Clinical data from patient files, diagnostic biopsies and flow cytometry data were reviewed. Last follow-up was in May 2011. The prognostic value of the following variables, evaluated at the time of diagnosis, were analyzed: age, sex, Ann Arbor stage, ECOG, B-symptoms, Hb, LDH, albumin, lymphocytosis, leukocytosis, splenomegaly, nodal, extranodal and bone marrow involvement, blastoid morphology, expression of CD23, light chain, Ki 67, p53 and nuclear SOX11. Results: The median age at diagnosis was 68.8 years (range 36.2 – 89.9); 67.4 in males and 72.1 in females, respectively. The male: female ratio was 2. Thirty patients had a known malignancy of other type before the MCL diagnosis and 12 acquired a cancer later. In 13 patients the other malignancy was the cause of death. Median overall survival (OS) time was 43 months in the whole cohort and 38 months, when excluding 39 patients receiving ASCT as part of first-line therapy. No statistically significant difference in OS was seen with respect to whether the lymphoma was diagnosed before or after 2005. In the non-transplanted patients (n=149), univariate analysis showed the following clinical variables to be negatively correlated to overall survival: age >65 years, B-symptoms, splenomegaly, ECOG >2, low albumin, and high LDH. The median survival was not reached in the low risk MIPI group, and was 79 and 34 months, in the middle and high risk MIPI group, respectively. Blastoid morphology and p53 positivity (>20%), were negatively correlated to overall survival (both with p<0.0001), as was increasing tumor cell proliferation (measured as a continous variable or using the cut-offs >50%, both with with p<0.0001), but not with cut-off >30% (p=0.061), while SOX11 positivity was related to a prolonged survival (p=0.015). Multivariate analyses showed that age >65 (HR 6.1, p<0,002), ECOG >2 (HR 63, p<0.001), high LD (HR 3.7, p< 0.001), and p53 positivity (HR 5.6, p< 0.0001) remained significant. Clinically indolent MCL, defined as in retrospect not requiring treatment within two years from diagnosis, was seen in 17 patients. In two of these patients the proliferation was >30%, in one >50%, two had a p53 expression >20% and two were SOX11 negative. Therapy was never required in 9 of these initially indolent patients and only one had an autologous transplantion later in the disease course. The median OS was 72 months for the 17 indolent MCL compared with 34 months in patients requiring treatment earlier in their disease (p=0.003). The follow-up time did not differ significantly between the two groups. Conclusions: Compared to data from published clinical trials of advanced MCL, our population-based cohort of 186 cyclin D1 positive MCL patients were diagnosed at an older age, which may contribute to a shorter overall survival. Certain well-established prognostic variables seem to loose significance outside study populations. In the group of 147 non-transplanted patients multivariate analysis showed that only age, ECOG, LDH and p53 positivity were independently associated with overall survival. Leukocytosis as a variable of MIPI had no impact. Neither SOX11, CD23 or other biological markers applied at the time of diagnosis could predict for clinically indolent disease. Disclosures: No relevant conflicts of interest to declare.