Diffuse large B-cell lymphoma : population-based studies of relapse and cardiac complications

Abstract

Diffuse large B-cell lymphoma (DLBCL) is an aggressive life-threatening disease, but it is often possible to cure with immunochemotherapy. Not responding to primary treatment or having a relapse is associated with poor prognosis. However, it has not been well described how large the proportion of patients with relapsed/refractory disease is. There is also a lack of knowledge regarding the outcome for these patients in a population-based setting. Standard primary treatment for DLBCL includes anthracyclines, which has been associated with an increased risk of heart failure in several studies. Most patients with DLBCL are also older (>70 years) and many have comorbidities that are associated with cardiovascular disease, but their rate of cardiovascular events such as acute myocardial infarction (AMI) is not well studied. We performed nation-wide studies to assess the cumulative incidence of relapsed/refractory disease, the outcome for the relapsed/refractory patients and patients with central nervous system (CNS) relapse in particular and also studied the AMI rate among DLBCL patients compared to the general population.

In study I we identified all newly diagnosed DLBCL patients (during 2007-2014, n=4243) in the Swedish Lymphoma Register, complemented with information on relapsed/refractory disease through a medical record review. Patients were followed from the time of DLBCL diagnosis until relapse or death of any cause. Their median age was 71 years (range 18-105) and 84% (n=3550) received primary treatment with curative intent resulting in 5-year overall survival (OS) of 65% (95% confidence interval (CI): 64-67). Fourteen percent of the DLBCL patients (median age: 84) did not receive curative intent treatment and their median OS was 2.9 months. The cumulative incidence of relapsed/refractory disease at any site among curatively treated patients, when assessed in the presence of the competing risk of death, was 23% (95% CI: 22-25) after 5 years. The cumulative incidence of CNS relapse at two years was 3% (95% CI: 2-4) overall and 8% (95% CI: 6-11) among high-risk patients with CNS IPI 4-6.

In study II the relapsed/refractory DLBCL patients identified in study I (n=736) were followed from the time of relapse/refractoriness until death of any cause. For the whole group the median OS was 6.6 months (95% CI: 5.8-7.9) and for patients below and above 70 years it was 9.6 and 4.9 months respectively. Having relapse within 12 months from diagnosis was associated with worse outcome. Among patients who were 70 years or younger at the time of relapse/refractoriness, two thirds (63%) received standard second-line intensive therapy and one third (35%) were consolidated with autologous stem cell transplantation (ASCT). For patients with early relapse (≤12 months) who received ASCT, 2-year OS was 40% (95% CI: 26-53) from the time of transplant, whereas for patients with late relapse (>12 months) 2-year OS was 66% (95% CI: 54-76). When applying inclusion/exclusion criteria commonly used in several clinical trials studying chimeric antigen receptor (CAR) T-cell therapy, in total 35% of patients ≤76 years fitted trial criteria and their survival was not much longer than those not eligible for CAR T (median progression-free survival (PFS): 4.8 months (95% CI: 2.9– 6.3) vs 3.1 months (95% CI: 2.6– 3.8)).

Study III included patients with CNS relapse identified in study II (n=145) who were followed from the time of CNS relapse until death of any cause with the aim to assess their survival and eligibility to CAR T-cell therapy. A majority had their CNS relapse at the time of their first relapse (81%, n=118) and most had isolated CNS-involvement (68%). Two year-OS for patients with CNS involvement at first relapse was 12% (95% CI: 7-18) and median OS was 3 months (95% CI: 3-4). Patients who could receive a second-line regimen containing high-dose methotrexate had 2-year OS of 18% (95% CI: 8-32). One third of the CNS relapse patients fitted the CAR T trial criteria and their median OS was 5 months (95% CI: 3-6).

In study IV the curatively treated DLBCL patients that were identified in study I (n=3548) and comparators from the general population matched by age and sex (n=35 474) were followed from the time of the DLBCL diagnosis to assess the rate of AMI. The cohort was linked to the National Patient Register and the Swedish Cause of Death Register to identify cases of AMI and to SWEDEHEART (Swedish web-system for enhancement and development of evidence-based care in heart disease evaluated according to recommended therapies) to obtain detailed information regarding the AMIs and their treatment. The DLBCL patients had a 33% excess rate of AMI compared to the general population (hazard ratio (HR) 1.33, 95% CI: 1.14-1.55) over the whole study period. The highest rate was observed during the first year following DLBCL diagnosis, but after two years there was no evident difference between DLBCL patients and comparators. Among older DLBCL patients, with mild or moderate comorbidity such as hypertension or diabetes, the excess rate of AMI was 61% (HR 1.61, 95% CI: 1.10-2.35). There was no significant difference in AMI characteristics, clinical management of the AMI or 30-day survival among DLBCL patients and comparators.

In summary we found that relapsed/refractory DLBCL was less common than previously described affecting approximately one in four curatively treated patients. However, the patients who experienced it had very poor prognosis, especially those with early relapse and those with relapse involving the CNS. DLBCL patients had an increased rate of AMI that motivates monitoring of other cardiovascular risk factors during and shortly after the DLBCL treatment.