Patterns of Total Costs of Care over Time for Patients with Multiple Myeloma (MM)

Introduction. A previous study of patients (pts) with newly-diagnosed multiple myeloma (NDMM) showed that initial high monthly total costs declined over time until returning near initial levels upon pts advancing to later lines of treatment (Tx) at disease progression (Arikian, CMRO 2015). However, the longer-term economic implications of the choice of first-line Tx, regardless of subsequent Tx regimens, have not been described. We utilized an intent-to-treat approach to further evaluate and compare the total costs of NDMM patient cohorts initiated on Lenalidomide (LEN) or Bortezomib (BORT)-based Tx and followed these patients over 54 months.

Methods: A retrospective analysis was performed using a large US medical and pharmacy claims database, covering > 25 million commercial and Medicare lives annually. NDMM patients were defined as having ≥ 2 outpatient claims or ≥ 1 inpatient medical claim with a diagnosis of MM (ICD-9-CM code 203.0X), with the first such claim used to define the index date. Inclusion criteria required ≥ 12 months’ pre-index and ≥ 6 months’ post-index continuous enrollment from 2006 - 2013, and evidence of initiation of a subsequent line of therapy. Pts with claims for stem cell transplantation (SCT) or receiving concurrent LEN plus BORT as first-line Tx were excluded. The analysis focused on cohorts of NDMM pts receiving LEN or BORT-based first line Tx, who progressed to one or more subsequent lines of Tx, using time to next therapy (TTNT) as a proxy for progression. Pts receiving LEN or BORT-based first line Tx were randomly matched 1:1 on age (+/- 3 years at index), sex, baseline Charlson comorbidity score (+/- 1), and presence of renal disease. Using methods similar to those described by Gaultney(J Clin Pharm Ther, 2013), pts’ average monthly costs and standard errors were determined, including medical and pharmacy costs, and total cost patterns were calculated from first line Tx initiation until the end of patient eligibility or 54 months.

Results: 1,181 NDMM pts receiving LEN (N=444) or BORT-based (N=737) first line Tx with complete data available through initiation of subsequent line of Tx were identified. After matching, 856 NDMM patients remained (428 LEN, 428 BORT). Monthly total direct medical plus pharmacy costs for these pts were in excess of $12,000 at initiation. Total monthly costs declined quarterly for each cohort until the median TTNT was reached (20 months for BORT; 37 months for LEN).  As increasing numbers of pts in each cohort then advanced to subsequent lines of therapy, costs of those patients began to offset continued cost declines for pts still on first line of therapy. Over the full 54 month follow-up period, total monthly costs averaged $8,324 (SE = $370.30) for pts initiated on LEN vs. $10,728 (SE = $500.55) for pts initiated on BORT (p-value <0.001). Medical costs represented 54-57% of total monthly costs for each cohort. In each cohort, the most common Tx option upon disease progression was to switch to the other regimen. 

Figure 1: Direct monthly costs (medical and pharmacy) for LEN- and BORT-initiated patients and savings attributed to LEN-initiated patients

Conclusions: For a matched population of NDMM pts followed continuously for 54 months, initiation on LEN- versus BORT-based Tx without SCT was associated with an average of $2,404 lower monthly direct total costs. Following initiation of first-line therapy, total monthly costs per pt declined quarterly until initiation of the next line of Tx. Cumulative costs over the 54-month period were over $120,000 lower per pt in favor of the LEN cohort, influenced by more rapid advancement to subsequent Tx in the BORT cohort, and the associated higher costs. Based on second-line Tx patterns, this analysis also suggests that a sequence of LEN in first line followed by BORT in second line is associated with lower costs over time compared to the reverse sequence.  Due to the nature of retrospective claims, some underlying differences may remain between the two cohorts even after matching.