Two recent cross-trial comparison analyses have attempted to compare treatment regimens using statistical matching methods (known as MAICs), but not a traditional randomized trial, in relapsed or refractory (R/R) multiple myeloma (MM).
In one, published in Leukemia & Lymphoma, researchers found longer progression-free survival (PFS) for the carfilzomib-dexamethasone-daratumumab (KdD) triplet than either the pomalidomide-bortezomib-dexamethasone (PVd) or daratumumab-pomalidomide-dexamethasone (DPd) triplets.1 In the other, also published in Leukemia & Lymphoma, researchers found that the B-cell maturation antigen (BCMA)/CD3 bispecific monoclonal antibody elranatamab produced a better response and PFS in BCMA-naïve patients than teclistamab, another such antibody.2
Both comparison studies were funded by the companies that manufacture the therapies found to be superior. Onyx, a subsidiary of study sponsor Amgen, manufactures carfilzomib, and study sponsor Pfizer manufactures elranatamab.
The analyses could offer some guidance to clinicians on how best to manage patients, although the researchers acknowledge that some caution is necessary in interpreting the results.
Both used matching-adjusted indirect comparisons (MAIC) to measure the results of a therapy in one trial against the results in another trial. In MAICs, researchers compare outcomes using individual patient-level data (IPD) from one study and results in the aggregate from the other study by matching the individual data to the inclusion criteria and baseline characteristics of the summarized study. Patients whose characteristics more closely resemble those in the other study are weighted more heavily and contribute more to the results.
In the KdD analysis, researchers first indirectly compared results from the lenalidomide-exposed KdD arm of the CANDOR trial, totaling 123 patients, with the 281-patient PVd arm of the OPTIMISMM trial. Before matching, there were considerable differences between these groups in the proportion of patients who were older than 75, had previously been exposed to bortezomib, were refractory to bortezomib, had Eastern Cooperative Oncology Group performance status (ECOG PS) scores of 0, were refractory to lenalidomide, and had two or more lines of prior therapy. However, the baseline characteristics were balanced after matching, researchers said.
In the matched populations, the PFS at 24 months was 48.5% for KdD and 28.8% for PVd. The hazard ratio (HR) for PFS was 0.60 in favor of KdD (95% CI 0.37-0.88).
There was a lower incidence of grade 3 or higher neutropenia in the KdD group, but the incidence of grade 3 or higher anemia and thrombocytopenia were slightly higher among patients receiving KdD, researchers found.
Researchers also indirectly compared results from the same arm of CANDOR with results from the DPd arm of the APOLLO trial. Before matching, there were considerable differences in several categories, including proportion of patients who were over 75 years old and who were refractory to proteasome-inhibitor therapy. After matching, baseline traits were balanced, they said.
After matching, the PFS at 24 months was 47.4% for KdD and 36.6% for DPd, which was numerically but not significantly higher, with a PFS HR of 0.77 favoring KdD (95% CI 0.50-1.08). Adverse event differences were similar to the KdD and PVd comparison.
“Our estimates suggest that KdD may be a highly effective lenalidomide-sparing treatment option at first relapse, especially for the growing number of patients who received lenalidomide as part of their frontline therapy,” said study author Katja Weisel, MD, of the University Medical Center Hamburg-Eppendorf in Hamburg, Germany. “The addition of anti-CD38 monoclonal antibodies, such as daratumumab, to previously existing standards of care has demonstrated favorable efficacy across a growing number of clinical trials.”
In the elranatamab comparison with teclistamab, researchers used elranatamab data from the 123-patient BCMA-naïve cohort of the MagnetisMM-3 trial and teclistamab data from the MajesTEC-1 study, totaling 165 patients. Adjustments had to be made to account for different exclusion criteria regarding ECOG PS scores and for definitions of extramedullary disease. Before matching, there were also differences in several characteristics, including International Staging System stage and penta-drug refractory status. Key baseline traits were balanced after matching, researchers said.
In the matched comparison, the elranatamab group had a significantly higher overall response rate — 75.3% versus 63.0% — with an odds ratio of 1.79 (95% CI 1.01-3.19). The elranatamab group also had a significantly higher PFS (HR=0.59; 95% CI 0.39-0.89).
“Although both elranatamab and teclistamab are BCMA/CD3 bispecifics, each antibody has a different structure, which might contribute to differences in clinical effects,” said study author Yannan Hu, PhD, of Cytel, Inc., which provides clinical trial design services, in Rotterdam, the Netherlands.
Dr. Weisel said cross-trial comparisons give a “robust estimation of relative outcomes,” but the researchers of both studies caution that adjustments can be made only for the differences that are observed but not for unobserved differences that might affect outcomes.
Dr. Weisel said randomized clinical trials offer “invaluable” information but “may require considerably large numbers of patients, incur significant costs, and take several years to read out.”
Although head-to-heads are the “gold standard,” they can sometimes be impractical, Dr. Hu said.
“If the difference in efficacy between two treatments is unknown or highly uncertain, then the number of patients required for a trial to have sufficient statistical power to detect differences between treatment arms may not be operationally realistic,” she said. “Indirect treatment comparisons, while they have their limitations, can provide helpful evidence in an expedient manner to inform health care decision-making.”
Any conflicts of interest declared by the authors can be found in the original article.
References
- Weisel K, Dimopoulos MA, Beksac M, et al. Carfilzomib, daratumumab, and dexamethasone (KdD) vs. lenalidomide-sparing pomalidomide-containing triplet regimens for relapsed/refractory multiple myeloma: an indirect treatment comparison. Leuk Lymphoma. 2024;65(4):481-492.
- Mol I, Hu Y, LeBlanc TW, et al. A matching-adjusted indirect comparison of the efficacy of elranatamab versus teclistamab in patients with triple-class exposed/refractory multiple myeloma [published online ahead of print, 2024 February 12]. Leuk Lymphoma. doi: 10.1080/10428194.2024.2313628.
Perspectives
I have often given talks where I create charts of outcomes from different clinical trials because the drug comparisons I want have not been officially studied. With these presentations, I remind the audience that cross-trial comparisons can lead to unfounded conclusions regarding comparative efficacy. So what are we to make of publications that seek to make these same comparisons of trials that were never designed to be compared to one another?
The National Institute for Health and Care Excellence (NICE), comprised of experts in evidence-based best practices, suggests that unanchored indirect comparisons can be conducted if any and all potential prognostic factors or modifiers are taken into consideration.1 NICE further cautions that these assumptions are “largely impossible to meet.” NICE also makes recommendations regarding the specific mathematic calculations for conducting such an analysis, but the details of such calculations are beyond my understanding as a non-biostatistician.
So what can we make of these two recently published unanchored indirect comparison analyses, one between recently approved anti-BCMA bispecific antibodies elranatamab and teclistamab using phase II data2 and the other of KDd versus PVd using randomized controlled trial data?3 Both of these publications, funded by the respective pharmaceutical companies for elranatamab (Pfizer) and carfilzomib (Amgen), try to adjust for variability between studies to allow for comparative efficacy. I find the recommendations of ISPOR, a professional society designed to promote good practices in health economics and outcomes research, to be most helpful. According to ISPOR, a decision-maker assumes “the risk of making the wrong decision and therefore losing health benefits by relying on lower quality evidence”4 when looking to such indirect comparisons.
Ultimately, I’m not sure I would make any treatment decisions specifically based on the level of data provided through unanchored comparison analyses, particularly when other factors like convenience, prior toxicities, patient preferences, and the personal experience of a treating physician can have increased importance when the evidence for differences in clinical efficacy is low. However, such analyses are likely better evidence than the charts I create that make no attempt to account for trial differences and can help contribute to the overall body of data surrounding drug choices.
Brea C. Lipe, MD
Associate Editor, ASH Clinical News
University of Rochester Medical Center
Rochester, New York
References
- Phillippo DM, Ades AE, Dias S, et al. NICE DSU technical support document 18: methods for population-adjusted indirect comparisons in submission to NICE. December 2016. Accessed March 11, 2024. https://www.sheffield.ac.uk/nice-dsu/tsds/population-adjusted.
- Mol I, Hu Y, LeBlanc TW, et al. A matching-adjusted indirect comparison of the efficacy of elranatamab versus teclistamab in patients with triple-class exposed/refractory multiple myeloma [published online ahead of print, 2024 February 12]. Leuk Lymphoma.doi: 10.1080/10428194.2024.2313628.
- Weisel K, Dimopoulos MA, Beksac M, et al. Carfilzomib, daratumumab, and dexamethasone (KdD) vs. lenalidomide-sparing pomalidomide-containing triplet regimens for relapsed/refractory multiple myeloma: an indirect treatment comparison. Leuk Lymphoma. 2024;65(4):481-492.
- Jansen JP, Fleurence R, Devine B, et al. Interpreting indirect treatment comparisons and network meta-analysis for health-care decision making: report of the ISPOR Task Force on Indirect Treatment Comparisons Good Research Practices: part 1. Value Health. 2011;14(4):417-28.