Evaluating the Economic Burden and Health Care Utilization Following First-Line Therapy for Diffuse Large B-Cell Lymphoma Patients in the US Medicare Population

Introduction: Diffuse Large B-Cell Lymphoma (DLBCL) is the most common sub-type of non-Hodgkin lymphoma. Despite advances in immunochemotherapy, a significant portion of individuals with DLBCL are refractory or relapse following initial treatment and eventually succumb to the disease. Real world studies on DLBCL treatment patterns beyond first-line therapy are limited, and the health care costs associated with relapsed DLBCL are unknown. Using 100% Medicare claims data, our study set out to measure health care utilization following completion of first-line therapy for DLBCL.

Methods: This retrospective study used 100% Medicare claims data (Inpatient, outpatient, Part D, SNF, HHA, Hospice, DME) to identify older adults (≥65 years) diagnosed with DLBCL (ICD9-CM diagnosis code: 200.7x) between Jan 1^st 2010 - June 30^th2014. Individuals were required to have prescription claims for a DLBCL chemotherapy treatment regimen 90 days pre-, or up to 1 year post-initial DLBCL claim. To limit our initial cohort selection to treatment-naïve patients, we excluded those with a DLBCL claim or prescription for a DLBCL related treatment at any time prior to the initial diagnosis date or treatment initiation date. All included patients were required to have continuous medical enrollment for 12 months before treatment initiation and receive at least one full cycle of valid DLBCL regimen. We then created our post first-line therapy study cohort by defining the end of first-line treatment by a gap of ≥60 days in therapy. Beneficiaries who initiated a second-line regimen during the follow-up period composed our relapsed group, whereas, individuals who completed first-line therapy without initiating any other chemotherapy treatment composed our non-relapse group. The index date for our health care utilization analysis was defined as the first-line treatment end date plus 60 days. Beneficiaries in both groups were required to have continuous medical benefits ≥1 year after this date or until death. Our primary outcome of health care resource utilization following first-line therapy were calculated as per patient per month and compared between our relapse and non-relapse groups using the chi-square test and t-test.

Results:We identified 5,909 Medicare beneficiaries who completed first-line treatment for DLBCL, of which 1,552 had claims indicating second-line therapy during follow up (relapsed group). The mean age for the relapsed group was 77 years compared to 76 years for the non-relapse group (p=0.006). Other baseline characteristics were similar between our relapsed versus non-relapsed cohort. Compared to the non-relapse group, patients in our relapsed cohort were less likely to receive R-CHOP as first-line therapy (56.3% vs 80.5%, p<0.001), but more likely to receive rituximab monotherapy (20.8% vs. 6.0%, p<0.001), bendamustine-rituximab (9.0% vs. 3.4%, p<0.001), and CVP regimens (7.9% vs. 4.9%, p<0.001). The mean (median) follow-up time after first-line therapy was similar between groups with 915 (863) and 929 (893) days for the relapsed and non-relapsed cohorts, respectively. In the period following first-line therapy, our relapsed cohort had higher healthcare utilization with more patients having claims for hospital admissions (60.7% vs. 41.1%, p<0.001), emergency room visits (51.7% vs. 43.0%, p<0.001), use of skilled nursing facilities (19.3% vs. 12.5%, p<0.001), home health agency (35.5% vs. 23.3%, p<0.001), and hospice services (19.9% vs. 6.3%, p<0.001) compared to non-relapse patients. Major cost drivers among relapsed patients were inpatient costs ($1,147 vs. $511, p<0.001), outpatient office visits ($3,036 vs. $691, p<0.001), and ambulatory costs ($1,730 vs. $343, p<0.001). Consequently, higher total all-cause health care costs per patient per month were observed in the relapse group ($6,566 vs. $1,951, p<0.001).

Conclusions: Our analysis found a significant portion of Medicare beneficiaries treated for DLBCL receive therapy beyond the first-line setting. Complete Medicare claims allowed us to capture comprehensive health care utilization, including end-of-life and supportive care not typically included in cost estimates. Our data confirms the expectations of higher health care utilization with relapsed DLBCL and suggest that improvements in first-line DLBCL therapy in older adults may offer significant health care savings in addition to improved clinical outcomes.