Outcomes of patients with primary refractory multiple myeloma in the era of triplet and quadruplet induction therapy

Multiple myeloma (MM) is a plasma cell malignancy with heterogenous responses to the initial treatment and survival outcomes.^1,2 Induction therapy includes combinations of 3 different drug classes, namely immunomodulators (IMiDs), proteasome inhibitors (PIs), and monoclonal antibodies (mAbs), with or without consolidative autologous stem cell transplantation (ASCT) and maintenance therapy until progression.^2,3 These therapies have substantially increased both progression-free survival (PFS) and overall survival (OS) over the last 10 years.^4-6 

Current risk stratification systems focus on cytogenetic abnormalities detected via fluorescence in situ hybridization (FISH) and factors related to tumor-host interactions (eg, β2 microglobulin and lactate dehydrogenase) to identify patients with poor prognosis.^7-10 However, a small proportion of patients do not respond to initial therapy (primary refractory disease) and exhibit particularly unfavorable outcomes in the absence of traditional high-risk markers.^11-13 However, the impact of primary refractory disease on survival outcomes has not been examined in the context of novel combination therapies or the use of mAbs in the upfront setting; thus, most available data are not applicable to the current standard-of-care therapies.

In this retrospective study, we sought to examine an academic center’s experience with the implications of hematologic responses achieved through the standard-of-care induction therapy in MM. In addition, we evaluated the impact of primary refractory disease on the PFS and OSas well as the effect of salvage chemotherapy on subsequent survival outcomes.

Between 1 January 2007 and 31 December 2019, we studied consecutive patients with MM treated with combination triplets or quadruplets (IMiD, PI, and mAb) as the induction therapy at Mayo Clinic, Rochester, Minnesota. Approval for this study was obtained from the Mayo Clinic Institutional Review Board, and informed consent for review of their electronic medical records was confirmed for all patients. Data were collected from the Mayo Clinic Dysproteinimia Database and manual chart review, as needed. Demographic information and MM-related variables included age at diagnosis, sex, International Staging System (ISS) and Revised- ISS (R-ISS) stages, FISH results, heavy/light chains involved, bone marrow (BM) plasma cells at diagnosis, and induction regimen. We used the mSMART classification to define high-risk (single-hit) status (t[4;14], t[14:16], t[14;20], deletion 17p, and gain/amplification 1q).^2,14 Double-hit MM was defined as the presence of ≥2 high-risk features in the diagnostic FISH. Response to induction was evaluated at 4 or 6 cycles after the initial treatment, based on the International Myeloma Working Group criteria.¹⁵ Primary refractory disease was defined as either progressive disease (PD) or stable disease (SD) after 4 or 6 cycles. Patients with very good partial response (VGPR) and complete response (CR) were analyzed as 1 group, given the lack of consistent BM evaluation at the time of response assessment. PFS was calculated from MM diagnosis to either progression, escalation/change in treatment, or death, and the OS was calculated from diagnosis to death from any cause. In the primary refractory group, the second PFS and OS were measured from the start of second-line therapy to the second progression or death, respectively. Data of patients without disease progression or who were alive on the last day of follow-up were censored for survival analyses.

The χ² test was used to compare categorical variables, and the t test was used to compare continuous variables, as appropriate. Kaplan-Meier estimates were used for survival probabilities in PFS and OS, and the long-rank test was used to compare the produced curves. The median follow-up was estimated using the reverse Kaplan-Meier method.¹⁶ The Cox proportional hazard model was used to estimate hazard ratios (HRs), with 95% confidence intervals (CI) for both the univariable and multivariable analyses. Two-sided P < .05 was considered statistically significant. All statistical analyses and graphs were performed using R version 4.2 (R Foundation for Statistical Computing).

A total of 1127 patients met the inclusion criteria. The median age at diagnosis was 62.6 years old (range, 23.9-90.4 years; 257 [22.8%] were older than 70 years, and 22 [2%] were older than 80 years), and 62% (N = 699) of the cohort were male. In terms of ISS staging, 382 (37.7%) had stage I disease, 365 (36%) had stage II disease, and 267 (26.3%) had stage III disease. For R-ISS, 246 (25.9%), 571 (60%), and 134 (14.1%) had stage I, II, and III disease, respectively. Only 1.7%¹⁷ of our patients had concurrent evidence of amyloidosis, and 15.5% (173) were treated using a clinical trial protocol. Most of our cohort (932 patients [82.7%]) were treated with bortezomib-lenalidomide-dexamethasone (VRd) as induction therapy, with 718 (23% missing) patients with available treatment scheduling (64.1% weekly, 33.1% bi-weekly, and 2.5% monthly bortezomib), followed by carfilzomib-Rd (KRd; 8.3%), quadruplets (addition of mAb to the PI/IMiD backbone; 7.2%), and other regimens (eg, elotuzumab-Rd; 1.8%). First-line ASCT was administered to 677 (61.3%) of the patients, with the rest of the cohort either not receiving ASCT or receiving it as salvage treatment. There were no statistically significant differences in the stem cell yield (CD34⁺ cells) and number of attempts between patients with primary refractory disease and the rest of the cohort. Regarding FISH stratification, 562 (54.4%) patients had no high-risk features, 355 (34.4%) had 1 abnormality, and 116 (11.2%) had 2 or more abnormalities in the initial plasma cell clones. The median percentage of clonal plasma cells in the BM was 50% (range = 0-100%). Patients with primary refractory disease were more likely to have traditional high-risk features (FISH, mSMART, and R-ISS) than primary responders.⁹ Table 1 summarizes and compares the baseline demographic and clinical characteristics of patients evaluable for response at induction (1086), and supplemental Table 1 summarizes the baseline data of the whole cohort (1127).

The initial response to treatment was assessed after 4 or 6 cycles of induction therapy. Information for response assessment was available for 1086 (96.3%) patients. Of these, 1013 (93.3%) attained at least a minimal response (MR), 304 (28%) achieved a partial response (PR), and 61.9% (672) achieved more than a VGPR. In contrast, 73 (6.7%) patients achieved either SD or PD during the initial induction cycles. The bar chart in Figure 1 shows the responses achieved based on the different induction regimens. There were no statistically significant differences in the baseline disease characteristics between patients treated with quadruplets and triplets (P is nonsignificant across all comparisons). All combinations had >60% of VGPR/CR rates at the time of assessment. More specifically, although no statistically significant differences were found between the responses achieved through the different regimens, patients treated with quadruplets had the highest rates of VGPR/CR (66.7% vs 63.7% for KRd and 60.9% for VRd) and lowest rates of primary refractory disease (2.5% vs 4.4% for KRd and 7.4% for VRd).

The estimated median duration of follow-up for the entire cohort was 68.1 months (95% CI, 65.9-69.6). The median PFS was 38.9 months (95% CI, 35.4-43), and the median estimated OS was 135.5 months (95% CI, 116.7 to not reached).

We compared the PFS and OS based on the hematologic responses after 4 or 6 induction cycles. We found that patients with primary refractory disease had significantly shorter PFS compared with patients with at least an MR at induction (4.2 months vs 9.7, 32.6, and 50.8 months for MR, PR, and VGPR/CR respectively; P < .01) and OS (51.3 months vs 141.8, 150.7, and 135.5 months, respectively; P < .01; Figure 2A-B). To avoid potential immortal time bias for patients with available follow-up from 4 to 6 cycles, we additionally performed a landmark analysis for OS at 6 months. Again, patients with primary refractory disease had significantly shorter OS than patients who responded to initial treatment (50.3 vs 135.8 months, respectively; P < .01; Figure 3). We then compared patients with primary refractory disease and those with early relapse (within 18 months of diagnosis). We found that the median OS of these groups was almost identical, indicating similar disease biology (51.3 vs 48.2 months, respectively; supplemental Figure 1). Regarding the FISH results, patients with double-hit MM had a significantly shorter PFS than patients with either 1 high-risk abnormality or standard-risk disease (23.1 months vs 31.7 vs 50.8 months, respectively; P < .01). The unfavorable effect of multiple aberrations was also observed in OS estimates, with a median OS of 72.5 months vs 116.6 and not reached, respectively; P < .01; Figure 4A-B).

To determine the factors associated with poor OS in the entire cohort, we performed univariable and multivariable analyses. In the univariable analysis, patients with primary refractory disease had an increased risk for shorter OS (HR, 3.5; 95% CI, 2.5-4.9). Other significant factors in univariable analysis were included in the multivariable model (supplemental Table 2). We found that age (10-year intervals) (HR,1.3; 95% CI, 1-1.5), ISS stage 3 (HR, 1.6; 95% CI, 1.1-2.3), Eastern Cooperative Oncology Group performance status (HR, 1.7; 95% CI, 1.3-2.3), single-hit and double-hit high-risk FISH (HR, 2.1 [95% CI, 1.5-2.9] and HR, 4.4 [95% CI, 3-6.4], respectively), and primary refractory disease (HR, 4.3; 95% CI, 2.6-6.9) were independently associated with shorter survival (Figure 5).

In a subgroup analysis of only the patients with primary refractory disease, 5 patients received no salvage therapy and died within 4 months of the first progression, and 1 patient did not have available information on the next treatment. The median second PFS and second OS from the first relapse were 11.9 months (95% CI, 8.7-17.3) and 45.1 (95% CI, 31.4-75.9), respectively. Clinical refractoriness was noted in 24 patients, and biochemical refractoriness in 49 patients. Although the second PFS (5.4 months vs 12.1 months, respectively) and second OS (21.7 vs 48.3 months, respectively) of clinical PD were numerically shorter, they did not reach our statistical significance threshold, probably because of the small number of patients (supplemental Figures 2 and 3). Of the patients with known subsequent treatment, 33 (49.3%) received salvage chemotherapy and ASCT as the second-line therapy, whereas 23 (34.2%)¹⁸ received daratumumab as part of their second-line therapy. For patients who underwent salvage ASCT, no statistically significant differences were observed in the baseline and clinical characteristics compared with those who did not, although the median age was 60.9 vs 67.2 years, respectively (P = .51; Table 2). Of these, 4 patients directly underwent ASCT without any change in induction therapy, 10 patients received the Velcade, Thalidomide, Dexamethasone, Cisplatin, Doxorubicin, Cyclophosphamide, Etoposide regimen, 5 received a daratumumab-containing regimen, and the rest received combinations of a PI, an IMiD, and cyclophosphamide. Importantly, patients who received second-line ASCT had increased second PFS (20.9 vs 8.1 months, respectively; P < .01) and second OS (74.7 vs 31.3 months, respectively; P = .02) compared with patients who did not (Figure 6A-B). Although the number of patients was small, (25 vs 13), the benefit of second-line ASCT was also shown via a subgroup analysis of younger patients (65 years), with a median PFS of 15.5 months and a median OS of 74.7 months, compared with 4.1 months and 48.3 months, respectively. Patients who received daratumumab as the second-line did not have any significant survival differences compared with those who did not; P = NS. We then created 2 groups based on the year of MM diagnosis and found no statistically significant differences in the PFS and OS for patients with primary refractory disease between 2007 and 2014 and patients treated from 2015 to 2019; P = NS (supplemental Figures 4 and 5).

In this study, we evaluated the incidence of primary refractory disease in MM and assessed its effect on survival outcomes in patients treated with a standard-of-care multidrug induction regimen. We showed that primary refractory disease occurs in ∼7% of patients with MM and is associated with significantly shorter PFS and OS, independent of other high-risk disease features. We also observed a trend of better response rates for patients treated with mAbs in addition to the PI/IMiD backbone and decreased odds for suboptimal responses. In the primary refractory cohort, we found that the median PFS after the first relapse was just under a year, and the median OS was 3 and a half years. Importantly, we showed that patients who were candidates for and received ASCT as their second-line had significantly prolonged second PFS and second OS and thus could potentially benefit from this approach.

The impact of a lack of response to initial treatment on survival outcomes has also been studied in prior cohorts.^{11-13,17,19-22} However, most of these studies were published before the advent of novel therapies and, thus, did not include patients treated with the current recommendations of triplets/quadruplets as frontline therapy. Our group previously published a series of 816 patients treated mostly with novel agents and found that 17% had primary refractory disease, compared with 6.7% in this cohort. It should be noted that <10% of the patients were treated with combination therapies in that study, and no patients were exposed to mAbs as the frontline therapy. The median survival for the entire cohort was longer in this study (135.5 vs 68.4 months, respectively), and the median OS for the primary refractory subgroup was shorter (43.4 vs 50.4, respectively). These data show that although the percentage of nonresponders has substantially decreased over the years, the impact on survival is likely to be more pronounced, given the heavier treatment in the induction setting.¹¹ A recent multi-institutional study also reported the outcomes of 85 patients with primary refractory disease treated with novel agents (PI + IMiD). Similar to our study, they reported a median second PFS of 21.6 months and a median second OS of 35.6 months.¹⁹ Furthermore, they also found improved second PFS and second OS for patients treated with ASCT as the second-line consolidation compared with those who did not. As a result, for patients who are fit enough to undergo such interventions, ASCT as a second-line therapy could portend an effective option. Finally, Blocka et al found that salvage therapy could be omitted in select patients with PD who were able to undergo transplantation, because no significant PFS and OS benefits were observed when compared with patients who directly underwent transplantation.²² These patients should not remain off-treatment for prolonged periods, because this has been shown to predict worse outcomes, especially in patients with suboptimal responses to induction.²³ 

The use of CAR T cells and bispecific therapies could also be considered in earlier lines of therapy, especially for patients with primary refractory disease who could be refractory to many drugs from the first-line.^18,24 Although correlated with traditional high-risk markers (FISH, mSMART, and R-ISS), patients at a functional high risk are often overlooked in clinical trials, and studies that will include this population are needed. Recent results from the CARTITUDE-2 trial (12-month PFS rate of 90%) showed promising efficacy of ciltacabtagene autoleucel in patients receiving 1 line of prior therapy, and future trials are eagerly anticipated for this population.²⁵ 

Our study has several limitations. Firstly, the retrospective nature of the study design lends itself vulnerable to potential biases. Although the study was homogenous with respect to induction therapy, the long recruitment period (2007-2019) and changes in salvage treatment options could have introduced possible biases in survival outcomes. Finally, although ASCT was beneficial for survival in our cohort and previous studies, it should be noted that these patients were probably fitter and could undergo the procedure in the first place. Thus, although our data can be used for hypothesis generation, randomized trials are needed to further delineate the role and applicability of ASCT in this population.

We conclude that the incidence of primary refractory disease has markedly decreased in the current treatment landscape of MM. However, early progression remains a significant factor for shorter PFS and OS, even after adjusting for common high-risk features. Finally, although further investigation of the most effective therapies is needed, ASCT could be the most beneficial option for prolonging survival.

The authors acknowledge the Mayo Clinic Hematological Malignancies Program.

Contribution: S.K.K. and C.C. designed the study, collected and analyzed the data, and wrote the manuscript, and U.G., P.K., M.B., F.K.B., J.C., D.D., A.D., A.L.F., M.A.G., W.G., S.R.H., M.A.H., Y.L.H., T.K., M.Q.L., N.L., Y.L., R.W., R.A.K., S.V.R., and S.K.K. were involved in patient management, revised the manuscript, and approved the final version of the manuscript.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Shaji Kumar, Division of Hematology, Mayo Clinic, 200 First St SW, Rochester, MN 55906; e-mail: kumar.shaji@mayo.edu.