Bortezomib retreatment for relapsed and refractory multiple myeloma in real‐world clinical practice

Abstract Aims Studies have shown that bortezomib retreatment is effective in relapsed/refractory multiple myeloma (MM). The observational, prospective electronic VELCADE® OBservational Study (eVOBS) study assessed bortezomib‐based therapies for patients with MM in everyday practice. Here, we report on those patients receiving retreatment with bortezomib. Methods Consenting adults scheduled to receive bortezomib for MM were enrolled at 162 sites across Europe, Canada, Brazil, Russia, and Turkey between 2006 and 2010. Retrospective data on prior therapies and prospective observational data after bortezomib initiation were captured electronically at baseline, after every bortezomib cycle, and every 12 weeks after discontinuation or progression. Investigator‐assessed responses and adverse events (AEs) were evaluated. Results Ninety‐six of 873 patients enrolled to eVOBS received bortezomib as first retreatment for progressive disease during the prospective observation period. Median age was 62 years, 53% were male, and median number of prior therapies at retreatment was 4. Overall, 41% of patients initiated bortezomib retreatment in combination with dexamethasone, 16% in combination with lenalidomide, and 21% received monotherapy. Rate of partial response or better (≥PR) was 75% at initial bortezomib therapy, including 44% complete response (CR)/near CR (nCR); at retreatment, ≥PR rate was 46%, including 15% CR/nCR. Median progression‐free survival was 11.4 months (95% confidence interval [CI]: 9.1‐12.7) from start of initial bortezomib treatment and 6.4 months (95% CI: 4.4‐7.2) from start of retreatment. Median overall survival from start of retreatment was 17.6 months (95% CI: 14.4‐23.5). Of the 96 patients retreated with bortezomib, 77% reported an AE. Peripheral neuropathy during bortezomib retreatment occurred in 49% of patients, including 10% grade 3/4. Conclusion These data suggest that retreatment with bortezomib is a feasible option for patients with relapsed/refractory MM.


| INTRODUCTION
Multiple myeloma (MM) is an incurable disease with a high incidence rate in elderly people. 1 The disease typically follows a relapsing course, with many patients requiring multiple lines of therapy. 2 The choice of treatment for relapsed and/or refractory (RR) MM may be influenced by several factors, including patients' prior regimen(s), comorbidities, disease characteristics at relapse, prior treatmentrelated toxicities, and duration of prior remission. [3][4][5] One of the mainstays of treatment for RRMM is the proteasome inhibitor bortezomib, with numerous phase 2 and 3 studies clearly showing therapeutic effectiveness in this patient population. [6][7][8][9][10][11][12][13][14][15][16][17][18][19][20] In Europe, bortezomib is currently approved for progressive MM in patients who have received ≥1 prior therapy (alone or in combination with pegylated liposomal doxorubicin or dexamethasone) and who have already undergone or are unsuitable for hematopoietic stem cell transplantation (HSCT). It is also approved as a treatment for patients with previously untreated MM who are ineligible for high-dose chemotherapy with HSCT (in combination with melphalan and prednisone), or as induction treatment prior to high-dose chemotherapy with HSCT (in combination with dexamethasone, or with dexamethasone and thalidomide). 21 Bortezomib is a recommended treatment option for RRMM. 22 For patients with MM who receive a finite course of bortezomib (ie, not receiving maintenance treatment), their disease may remain sensitive to bortezomib-based therapy at relapse. Retreatment with bortezomib is, therefore, a viable option for patients with progressive disease (PD), either as a subsequent or later line of therapy after initial bortezomib treatment. A number of retrospective studies, [23][24][25][26][27][28][29][30][31] prospective clinical trials, 32,33 and a recent meta-analysis 34 have demonstrated the viability of retreatment with bortezomib, all showing bortezomib-based retreatment to be efficacious and tolerable. 35 On the basis of the prospective clinical study by Petrucci et al, 32 the indication for bortezomib in the US was expanded in late 2014 to include retreatment in patients who have previously responded to bortezomib and who have relapsed at least 6 months after completing prior bortezomib treatment. 36 While the efficacy and safety of bortezomib-based therapies for retreatment have been shown in the highly controlled clinical trial setting, these findings may not reflect those observed in routine medical practice, where the patient population can differ substantially from that selected by strict clinical trial entry criteria. To date, however, data on the use of bortezomib retreatment in the "realworld" oncology practice setting are limited. To address this gap, we conducted a sub-analysis of the prospective, international, noninterventional, electronic VELCADE OBservational Study (eVOBS) that was designed to study the efficacy and safety of bortezomibbased therapies for MM in real-world medical practice. 37 In our sub-analysis, we examined the efficacy and safety of bortezomibbased retreatment for relapsed MM during the monitoring period of eVOBS.

| Objectives and endpoints
The overall objective of eVOBS was to evaluate the clinical outcomes associated with bortezomib-based therapies in real-world medical practice. 37 The objective of the present sub-analysis was to evaluate the efficacy and safety of bortezomib-based retreatment in patients with progressive MM following initial bortezomib-based treatment during the prospective observational period of eVOBS. Data on the following endpoints were collected prospectively: response rates

| Statistical analyses
All time-to-event endpoints were analyzed using Kaplan-Meier and Cox proportional hazards regression analyses. Patients lost to followup or who discontinued bortezomib treatment without a reason were censored in all time-to-event analyses. Kaplan-Meier analyses were stratified according to baseline characteristics (including age, MM stage, line of therapy, creatinine clearance, and baseline albumin) and best response to bortezomib. The two-sided log-rank test was used to assess the significance of any differences between the stratified data; the conventional significance threshold of 0.05 was used across all analyses. As described in Terpos et al, missing data were not substituted nor imputed. 37 Statistical analysis was performed using SAS version 9.2.  Demographics and baseline characteristics for the 96 retreated patients are summarized in Table 1 In the retreatment population, the median age (62 years) and proportion of males (53%) were both slightly lower than in the overall population (65 years and 58%, respectively). In remission 6 (6) 4 (4) Autologous stem cell transplantation 6 (6) 1 (1) Death 0 8 (9) Other c 12 (13) 13 (14) Abbreviation: AE, adverse event.
a Percentages may not equal 100% due to rounding. b 95 of 96 retreated patients had discontinued bortezomib-based therapy at data cut-off. c Includes patient withdrawal, lost to follow-up, and those with unreported reasons.

| Bortezomib retreatment
The majority of patients undergoing bortezomib retreatment received this therapy as their fourth or fifth line ( Table 2). The most common bortezomib-based regimens received at initial treatment and at retreatment were bortezomib plus dexamethasone (53% and 41%, respectively) and bortezomib monotherapy (each 21%) ( Table 2).
There was a notable increase in the use of bortezomib plus lenalidomide combination therapies between initial bortezomib treatment (2%) and retreatment (20%).
In total, 91 (95%) patients had received an alternative treatment for MM between initial bortezomib and bortezomib retreatment.  We investigated factors that may influence response to retreatment. In total, 68% of patients who had achieved CR/nCR with initial bortezomib achieved ≥PR on retreatment; this figure was 39% in patients who had achieved PR and 20% in patients who had achieved ≤MR with initial bortezomib (P = 0.0022; two-sided log-rank test). In patients who had received one or two prior therapies, 70% achieved ≥PR at retreatment, compared with 48% for three prior therapies and 39% for four or more prior therapies (Figure 1;

| Survival outcomes
In bortezomib-retreated patients, median follow-up from the start of initial bortezomib was 35 months (range 9-60), and from the start of There was no statistically significant difference in OS (P = 0.0723; two-sided log-rank test), or PFS (P = 0.3062; two-sided log-rank test),      Of the 96 patients included in this study who were eventually retreated with bortezomib, 75% achieved ≥PR after their initial bortezomib treatment, including 44% CR/nCR. These rates are slightly higher than the 69% ≥PR rate and 37% CR/nCR rate observed in the overall eVOBS study population following initial bortezomib-based treatment. 37 Although the populations were generally similar, differences in patient and baseline disease characteristics between the overall study population and the retreated population may have contributed to these observed results. 37 It is also likely that patients who had initially responded well to bortezomib were preferentially chosen for retreatment, resulting in a population with a higher initial response rate.

| Peripheral neuropathy
In our retreated cohort, 46% of patients achieved ≥PR with bortezomib retreatment, including 15% CR/nCR. The ≥PR rate with bortezomib retreatment reported here is comparable with overall response rates reported with bortezomib retreatment in previous prospective clinical trials, 32,33 a retrospective case series, 27 and a metaanalysis, 34 but is slightly lower than the 60% reported by Ahn et al, in their retrospective study. 31 In the latter study, however, patients had received a median of two prior therapies, and only those who had relapsed or progressed ≥6 months after the previous bortezomib therapy were included. 31 Despite a higher response rate, median PFS was comparable, at 5.5 months (95% CI: 4.2-6.8). Generally, though, due to differences in study design (eg, different criteria for response assessment and patient inclusion criteria/patient populations), interstudy comparisons of response rates should be interpreted with caution.
The ≥PR rates observed with bortezomib retreatment in this study are encouraging, considering the advanced disease stage and heavily pretreated nature of the retreated population. This is consistent with previous studies demonstrating that bortezomib retreatment is feasible in later lines of therapy and can produce responses in a considerable proportion of patients. 24,25,27,33,41 The observed decrease in ≥PR rate between initial bortezomib and bortezomib retreatment is consistent with the progressive nature of MM. 2 Notably, 95% of patients retreated with bortezomib in eVOBS had received at least one alternative treatment for MM between initial bortezomib and bortezomib retreatment, which may have impacted on the observed ≥PR rate. Although infrequent in clinical practice, 12 patients in our population received a second retreatment with bortezomib after PD, with two patients going on to achieve ≥PR.
Consistent with previous prospective 32 and retrospective studies, 23  The safety profile observed with bortezomib retreatment in this study is consistent with that known for bortezomib in RRMM, 12,13 and with prior clinical studies of bortezomib retreatment, 25,28,32 and in a bortezomib retreatment meta-analysis. 34 The most common AEs reported with bortezomib retreatment in this study were hematologic-, gastrointestinal-, and neurologic-related toxicities.
Notably, there was no apparent increase in PN incidence at  47 The role of second-generation proteasome inhibitors in retreatment after an initial bortezomib course should also be investigated, as well as the potential clinical contribution of other combinatory agents in that setting, such as immunomodulatory therapies, histone deacetylase inhibitors, and chemotherapy.
Healthcare plc, for writing assistance during the development of this manuscript, which was funded by Millennium Pharmaceuticals, Inc.
and Janssen Global Services, LLC.

FUNDING INFORMATION
This analysis was supported by Janssen Research and Development.

CONFLICTS OF INTEREST
None declared.