Outcome of T-Cell/Histiocyte-Rich Large B-Cell Lymphoma in Denmark 2008-2022

Introduction

T-cell/histiocyte-rich large B-cell lymphoma (THRLBCL) is a rare subtype, comprising less than 10% of large B-cell lymphomas (LBCL), characterized by low lymphoma cells content and abundant numbers of T-cells and histiocytes. Prospective data on outcome is lacking. In retrospective studies overall outcomes were reported similar or superior compared to diffuse large B-cell lymphoma, not otherwise specified (DLBCL, NOS) when treated with R-CHOP. Here we describe the results of a Danish population based matched comparison cohort study comparing outcomes of THRLBCL and LBCL in the Rituximab era.

Methods

Patients with histologically confirmed THRLBCL diagnosed from January 1^st 2008 to December 31^st 2022 were identified in the Danish Pathology Register (DPR). Clinical data and outcomes on Danish patients diagnosed with LBCL, including patients with THRLBCL, diagnosed in the same time period were retrieved from the Danish National Lymphoma Registry (LYFO). THRLBCL patients and LBCL patients were matched 1:5 according to sex, age +/- 3 years, elevated LDH, PS, Ann Arbor stage, extranodal (EN) involvement (dichotomized into 0, 1, or ≥2 sites of involvement) and IPI. Kaplan-Meier curves were used for description of overall survival (OS). Cox proportional hazard regression adjusted and unadjusted was used to estimate the influence of clinical variables on survival.

Results

A total of 166 THRLBCL patients were identified in DPR of whom 131 could be identified in LYFO. Four patients were moved from the THRLBCL cohort to the LBCL cohort due to misclassification. A total of 6501 patients diagnosed with LBCL were identified in LYFO including the 127 patients with THRLBCL. Two hundred eighty-four patients were excluded due to missing IPI data. Ultimately, 122 THRLBCL patients and 6095 LBCL patients were available for statistical analysis. The incidence of THRLBCL in the total LBCL cohort was 2%.

Before matching, there was a slight predominance of younger patients (64.2% vs 67.3%) in the THRLBCL group. There were more pronounced differences in the THRLBCL versus LBCL group regarding the following characteristics: Male sex (68.9% vs 56.5%), Ann Arbor stage ≥ 3 (87.7% vs 62.7%), IPI 2-3 (84.9% vs 72.8%). The THRLBCL group was more likely to present with liver, spleen, and bone marrow involvement, as well as B-symptoms.

In the 1:5 matched cohort analysis (THRLBCL, n=121; LBCL, n=602) the following variables were used: sex, age +/- 3 years, elevated LDH, PS, Ann Arbor stage, EN involvement (divided into 0, 1, or ≥2 sites of involvement), and calculated IPI which we chose to divide into low risk (IPI 0-1), medium risk (IPI 2-3), and high risk (IPI 4-5). There was no suitable match for one THRLBCL patient. Following matching the mean follow-up time for patients with THRLBCL and patients with LBCL was 5.2 years in both groups. There was no statistical significant difference in 3-year and 5-year OS of THRLBCL versus matched LBCL patients (3-year OS 69.5% vs 73.5% and 5-year OS 64.2% vs 66.5%, all p>0.25). The more frequently observed liver and spleen involvement among THRLBCL patients persisted whereas bone marrow involvement did not reach statistical significance.

Conclusion

In this national Danish population based retrospective matched cohort analysis we found similar survival of THRLBCL and matched LBCL patients in line with previous large registry based studies from the Rituximab era. This is despite the fact that patients with THRLBCL more often present with stage III or IV. Although EN disease was not more frequent among THRLBCL patients, they were more likely to present with liver and spleen involvement, which persisted following matching. This is also in line with previous observations and does not appear to represent an additional adverse risk.