Timing of Disease Progression and Impact on Survival in Mantle Cell Lymphoma: A Novel Illness-Death Model Study from the Swedish Mclcomplete Project

It has been established that progression of disease (POD) within 24 months in mantle cell lymphoma (MCL) patients is associated with poor prognosis. However, comparisons with progression-free (PF) patients are scarce, and the cut-off at 24 months potentially over-simplifies a more complex patient trajectory. This population-based study aimed to assess the impact of POD, and specifically timing thereof, on overall survival (OS) by first-line treatment.

A total of 1,185 MCL patients diagnosed 2006-2018 treated with systemic first-line therapy, were identified in the Swedish Lymphoma Register. POD was defined as either progressive disease as best response to first-line treatment, or relapse following initial response. Patients not (yet) fulfilling any of these were considered PF. Hazard ratios (HRs) with 95% confidence intervals (CIs), comparing all-cause mortality between patients experiencing POD and PF patients, were estimated using Cox proportional hazards models, with POD handled as a time-varying covariate. Additionally, an illness-death model was defined to provide estimates on the absolute scale, both the cumulative risk of POD accounting for death prior to progression, and the impact of progression on 5-year OS, conditional on time of POD/PF.

Among all included patients, 387 (33%) received BR (rituximab bendamustine), 351 (30%) Nordic MCL2 (dose-escalated R-CHOP [rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone] alternating with R-cytarabine for six courses and high-dose treatment with autologous stem cell transplantation as consolidation), and 171 (14%) R-CHOP or R-CHOP-like treatment. The median age at diagnosis was 70 years (inter-quartile range [IQR]: 64-78) overall, and 75 (IQR: 70-80) for BR, 62 (IQR: 56-66) for Nordic MCL2, and 72 (IQR: 65-79) for R-CHOP. The proportion of male patients was 74%. Over a median follow-up of 3.7 years (IQR: 1.2-6.2), almost half of the patients (571, 48%) experienced POD, of which 70 (12%) occurred immediately after first-line treatment (i.e., progressive disease as best response). The cumulative risk of POD within 5 years was 44% (95% CI: 39-50) in BR, 39% (95% CI: 32-47) in Nordic MCL2, and 51% (95% CI: 44-58) in R-CHOP.

Regardless of timing and treatment protocol, patients with POD experienced a significantly increased all-cause mortality compared to patients who remained PF (adjusted HR=7.45, 95% CI: 6.24-8.90). Stratified by treatment and time interval of progression, adjusted HRs ranged from 2.25 (95% CI: 1.25-4.05) among R-CHOP treated patients with POD >24 months, to 34.2 (95% CI: 21.2-55.3) among Nordic MCL2 treated patients with POD <12 months.

The difference in 5-year OS between patients with POD and PF patients was most pronounced when the progression occurred shortly after first-line treatment. Among BR treated patients with POD at one year, the 5-year OS was 4.7% (95% CI: 1.8-7.2) compared to 50% (95% CI: 46-56) if they had remained PF. For patients treated with Nordic MCL2 the 5-year OS was 14% (95% CI: 9.0-22) if POD at one year and 74% (95% CI: 70-78) if still PF, and for R-CHOP 13% (95% CI: 7.9-22) and 48% (95% CI: 40-56). The dismal prognosis for patients experiencing POD persisted throughout follow-up. To illustrate, the 5-year OS estimates for patients who experienced POD versus remained PF at six years were 33% (95% CI: 21-47) and 69% (95% CI: 60-77) for BR, 59% (95% CI: 46-71) and 87% (95% CI: 81-91) for Nordic MCL2, and 57% (95% CI: 39-73) and 66% (95% CI: 52-78) for R-CHOP.

Taken together, POD is associated with worse outcome regardless of when it occurs, although very early POD is especially serious. Survival continues to be inferior among patients with POD also beyond the initial 24-month mark, particularly among BR and Nordic MCL2 treated patients. This underlines the importance of not limiting to a dichotomised comparison (<24/>24 months) when studying patient prognosis and eligibility for new treatments within or outside of clinical trials. The comparison of survival between POD patients and those who remain PF incorporating the full patient pathway over a continuous timescale adds further nuance to our understanding of the disease course.

These results stress the need for improved first-line regimens to sustain remission and optimise outcomes for MCL patients. Future research should elucidate factors influencing disease progression and OS in MCL in the evolving treatment landscape.