Evaluating Multiple Myeloma Treatment and Clinical-Decision Making with Real-World Data

Background. In the last two decades, frontline treatment of multiple myeloma (MM) has been revolutionized with the addition of immunomodulatory drugs (IMiDs), proteasome inhibitors (PIs) and monoclonal antibodies (mAbs), which have resulted in significantly improved outcomes for MM patients. Varying combinations of IMiDs, PIs and mAbs are used as frontline treatment for MM patients with some proceeding to autologous stem cell transplant (ASCT) after induction. These combinations can differ based on patient comorbidities, tolerance, finances and medication availability. For those that do not achieve better than a partial response (PR) to treatment, clinicians will often change therapy by completely replacing all medications in a regimen or by substituting one or more medications within the same class of drug, e.g. carfilzomib for bortezomib.

All4Cure is a continuously updated online network of patients with cancer, clinicians and researchers, layered on top of a comprehensive real-world database of treatments and outcomes collected in near real time. Our study aimed to use All4Cure data to further examine the outcomes of patients who do not achieve PR or better with frontline therapy, and whether complete replacement versus substitution of therapy results in better outcomes.

Methods. Eligible patients included those registered for All4Cure who had received triplet therapy in the frontline setting (± ASCT), achieved a best response to frontline therapy of PR or worse (including minimal response (MR), stable disease (SD), progressive disease (PD) and not applicable (NA)) as assessed by the All4Cure registry, and initiated second-line treatment. Patients were classified as Replacement (R) if all frontline agents (excluding dexamethasone) had been replaced in the second-line, and Substitution (S) if one or more (but not all) non-dexamethasone agents were replaced in the second-line. Patients with second-line therapies that did not qualify as R or S were excluded. Outcomes in the second-line setting included clinical response, time to discontinuation of second-line therapy (TTD), and time to next treatment (i.e., start of third-line therapy; TTNT). Outcomes were assessed with chi-square/Fisher exact test and Kaplan-Meier analysis, respectively.

Results. Of 135 eligible patients, 43 were categorized as R and 92 were categorized as S. Of the R group, 22 (51.2%) achieved VGPR or better in second-line, and 21 (48.8%) achieved PR or worse with NA included. Of the S group, 38 (41.3%) achieved VGPR or better in second-line, and 54 (58.7%) achieved PR or worse with NA included (test of independence ^χ2 statistic = 1.153, p = 0.283). Median TTD was 338 days for S vs. 587 days for R (log-rank ^χ2 statistic = 2.52, p = 0.113). Median TNTT was 1282 days for S vs. 1822 days for R (log-rank ^χ2 statistic = 1.15, p = 0.283).

Conclusion. While patients in the R group appeared to have better outcomes than patients in the S group, the difference between groups was not statistically significant. Interestingly, most patients are contained in the S group, which suggests that many clinicians believe patients may continue to benefit from elements of their existing therapy.

Although results were nonsignificant, we did observe a trend suggesting that R may be a superior strategy to S as evidenced by TTD and TTNT results. Further studies will analyze the pre-treatment differences between the R and S groups to facilitate elucidation of covariates that may influence results and later lines of treatment.