Patterns and Determinants of Cancer-Directed Drug Therapy in Medicare Beneficiaries with Multiple Myeloma

Introduction: The introduction of novel therapies for multiple myeloma (MM) such as immunomodulatory drugs and proteasome inhibitors has led to considerable improvements in clinical outcomes, but the disease remains incurable. Both conventional and to a lesser extent novel therapies come with toxicity and treatment decisions need to be individualized based on factors such as age and comorbidities. Little is known about the frequency of drug therapy use in MM in real-world clinical practice.

Objective: This study aimed to assess the prevalence of, and determinants associated with cancer-directed drug therapy among newly diagnosed symptomatic MM (NDMM) Medicare beneficiaries.

Methods: This retrospective cohort study used Surveillance, Epidemiology, and End Results (SEER) registry and linked Medicare Claims (SEER-Medicare) data. We identified patients (>65 years) with an incident diagnosis of MM from 2007-2011 and associated claims from 2006-2012. Eligible patients had continuous enrollment in Medicare Parts A and B (12 months prior to) and in Part D (two months prior to) through six months post-diagnosis, or death (among patients who died within 6 months post diagnosis). Patients were required to have evidence of CRAB symptoms (hypercalcemia, renal insufficiency, anemia and bone disease) based on intravenous bisphosphonate utilization and ICD-9 diagnosis codes found in claims from six months pre- to one month post-MM diagnosis. Patients were followed until death; or censoring due to non-continuous Parts A, B and D enrollment after six months post diagnosis. We identified receipt of MM-directed therapies from Medicare claims during the follow up period using NDC and HCPCS codes for the following agents: bendamustine, bortezomib, cyclophosphamide, (liposomal) doxorubicin, interferon alfa-2b, etoposide, lenalidomide, melphalan, thalidomide, vincristine, and vorinostat, excluding dexamethasone monotherapy. The distribution of frontline therapies received within 90 days of diagnosis was characterized descriptively. Charlson Comorbidity Index (CCI) was derived based on claims within one year pre-diagnosis using a validated algorithm. Baseline characteristics among those who received any MM-directed treatment were compared to those who did not using t-tests and Chi-square tests for continuous and categorical variables, respectively. We also compared treatment rates in those diagnosed in an early (2007-2009) versus late period (2010-2011).

Results: Of 3,391 patients with NDMM who met our inclusion criteria, 2,599 (76.6%) received MM-directed therapy during follow up. The median time until treatment initiation was 49 days (range: 0-2156). The treatment rate using a landmark analysis of patients who survived at least 6 months produced congruent results (77% treated patients). Those who initiated therapy were younger (mean age: 75.6 years) compared to untreated patients (mean age: 78.7 years, p<.001) and had a lower comorbidity burden based on the CCI (p<.001) (Table 1). Among those with CCI=0, 81.7% received treatment, while 79.0% of those with CCI=1 and 69.0% of those with CCI>1 received treatment (p<0.001). African American patients were less likely to receive therapy (68.6%) compared to Caucasians (77.7%) and those who were of another race/ethnicity (79.7%, p<0.001). The proportion of patients initiating treatment was highest in the West (78.8%) and Midwest (78.7%) versus the South (75.1%) and Northeast (72.6%) (p<0.001). Among patients receiving treatment, the following drug therapies were observed within three months of diagnosis: bortezomib (n=937 patients, 36.1%), lenalidomide (n=633,25.5%), melphalan (n=398, 15.9%), thalidomide (n=404, 15.5%) and cyclophosphomide (n=75, 2.8%). Of patients diagnosed in 2007-2009, 75.6% received MM-directed drug treatment compared to 78.0% of those diagnosed in 2010-2011 (p=0.12).

Conclusion: Among NDMM patients, 23% did not receive MM-directed therapy. Treatment was associated with younger age and a lower comorbidity burden. Racial disparities, with a lower proportion of African American treated patients, and geographic variation in treatment rates were also observed. Future research is needed to assess the extent to which these determinants reflect patient preferences versus other factors, such as barriers to treatment access.

View largeDownload slide

View largeDownload slide

Close modal