Central nervous system involvement in Mantle Cell Lymphoma: A retrospective multicenter cohort study from the spanish geltamo group Free

Central nervous system (CNS) involvement in mantle cell lymphoma (MCL) is an uncommon but clinically significant event, often associated with poor prognosis. Prior studies have offered important insights, yet data from large, unselected real-world populations remain scarce. We report findings from one of the largest multicenter cohorts of MCL patients in Spain, aimed at characterizing CNS involvement and comparing clinical outcomes—including treatment response, progression-free survival (PFS), and overall survival (OS)—between patients with and without CNS infiltration.

This retrospective multicenter study included 1,162 MCL patients diagnosed in Spain (GELTAMO-MCL-2022). Among them, 60 patients (5.2%) had CNS involvement, either at diagnosis or during the course of the disease, while 1,102 had no evidence of CNS infiltration. Data from participating centers encompassed baseline characteristics, treatment approaches, response rates, and survival outcomes. PFS and OS were analyzed and compared between groups.

The crude incidence of CNS involvement was 5.2%, with 1.4% at the time of diagnosis. For patients who developed CNS relapse later (not at diagnosis), the median time from initial diagnosis to CNS involvement was 15.10 months (IQR 7.47–34.14). CNS localization was leptomeningeal in 70%, parenchymal in 16%, and both in 14%. Baseline characteristics, including established high-risk features, were compared between patients with and without CNS involvement. Sex distribution (male 73% in both; p=0.9), age >65 years (47% vs. 53%; p=0.2), and extranodal disease (p=0.3) were similar between groups. In contrast, high-risk baseline features such as high-risk MIPI (77% vs. 44%; p=0.001), blastoid/pleomorphic variant (45% vs. 16%; p=0.001), and Ki-67 >30% (80% vs. 49%; p=0.001) were significantly more frequent in patients with CNS involvement. First-line intensive regimens including high-dose cytarabine (HIDAC) were given to 48% with CNS involvement versus 36% without (p=ns). BTK inhibitors were used as a frontline treatment in 8% overall, but none of the CNS patients. Early progression of disease (ePOD) was more frequent with CNS involvement (72% vs. 51%; p=0.003).

In multivariate analysis, CNS involvement at diagnosis was an independent adverse factor for OS (HR 3.25; 95% CI: 1.62–6.52; p<0.001) and PFS (HR 2.09; 95% CI: 1.45–3.01; p=0.001). Other independent predictors of OS were age >65 (HR 2.04; p<0.001), high-risk MIPI (HR 1.72; p=0.003), and Ki-67 >30% (HR 1.63; p=0.001). For PFS, age (HR 1.39; p=0.016), high-risk MIPI (HR 1.45; p=0.014), and Ki-67 >30% (HR 1.39; p=0.007) were significant; blastoid morphology showed a trend toward significance (HR 1.30; p=0.056). Treatment with HIDAC was not an independent predictor of survival, neither for OS nor for PFS.

Median OS from diagnosis was 1.41 years (95% CI: 1.19–NA) for patients with CNS involvement at diagnosis (n=16) versus 7.5 years (95% CI: 6.92–8.4) for those without (p=0.001). Median PFS was 11.3 months (95% CI: 2.89–NA) versus 42.1 months (95% CI: 36.6–47.0), respectively (p=0.012).

In patients with CNS involvement at diagnosis, high-dose cytarabine (HIDAC) did not significantly improve OS or PFS. Median OS was 1.41 years with HIDAC and 1.30 years without it (HR 2.08; p=0.277), while median PFS was 16.4 vs 9.6 months, respectively (HR 1.51; p=0.504). Similarly, first-line consolidation with autologous stem cell transplantation had no significant impact on survival.

When calculated from the time of CNS involvement, patients with CNS disease at diagnosis had significantly worse outcomes compared to those with CNS relapse, with a median PFS of 3.5 vs 18.0 months (HR 3.00; 95% CI: 1.36–6.59; p=0.006) and median OS of 1.5 vs 0.19 years (HR 3.02; 95% CI: 1.36–6.69; p=0.006).

In this large Spanish multicenter cohort, CNS involvement in MCL—although a rare complication both at diagnosis and at relapse—was associated with markedly poorer PFS and OS. It remained an independent adverse prognostic factor alongside age, high-risk MIPI, and Ki-67. CNS infiltration at diagnosis was the strongest predictor of poor outcome, while occurrence during disease evolution conferred an even worse prognosis. These findings highlight the need for early identification and tailored therapeutic approaches for high-risk patients.