Abstract: A phase 3 study (MMY-3021) in relapsed/refractory multiple myeloma (MM) demonstrated non-inferior efficacy and an improved systemic safety profile with subcutaneous (SC) versus intravenous (IV) bortezomib. The present retrospective observational study used population-level data from the iKnowMed electronic health records database within McKesson Specialty Health | The US Oncology Network to evaluate the impact of bortezomib administration route on treatment outcomes in patients with previously untreated MM. Baseline characteristics were similar between patients who received solely IV or SC bortezomib (436 and 379 patients, respectively), except in rate of general comorbidities. A shorter median duration of treatment (116 vs 142 days; P = 0.019), a trend for more dose reductions at 16 weeks (13% vs 11%; P = 0.27), and a shorter median time to dose reduction (49 vs 56 days; P = 0.30) were observed with IV versus SC administration. Similar 2-year survival rates were observed (78% vs 79%). Overall, our findings suggest similar treatment outcomes and the potential for improved tolerability with SC versus IV bortezomib as initial therapy in the clinical setting.
Bortezomib is the first-in-class proteasome inhibitor approved for the treatment of multiple myeloma (MM) in the United States and European Union.1,2 The phase 3, international MMY-3021 study in relapsed/refractory MM demonstrated non-inferior efficacy between subcutaneous (SC) and intravenous (IV) bortezomib on the measures of response rate after 4 cycles, time to progression, progression-free survival (PFS), and overall survival (OS). An improved systemic safety profile with SC bortezomib was demonstrated, with significantly lower rates of peripheral neuropathy compared with IV administration.3,4 This resulted in the approval of SC injection as an additional route of bortezomib administration in 2012.
The improved systemic safety profile of SC bortezomib demonstrated in the MMY-3021 study4 was further supported in a recent review of clinical data on SC bortezomib by Petrucci et al.5 Notably, several studies have shown that twice-weekly or weekly SC administration of bortezomib is generally well tolerated in newly diagnosed as well as relapsed/refractory MM patients,6-18 including elderly patients who are typically more sensitive to the side effects of treatment.10,13,19 SC bortezomib is also preferred compared with IV administration by a majority of patients20,21 and is also easier to administer.6,7,20,21 Several individual studies have also generally supported the association of a low rate of peripheral neuropathy with SC administration.6,8-12,14,18,22-24
The non-inferior efficacy and better tolerability with SC administration of bortezomib, as reported to date, have the potential to impact treatment outcomes in the clinical setting. However, there is a paucity of comparative effectiveness data on SC versus IV bortezomib as initial therapy in MM outside of the clinical trial setting. Therefore, the main objective of this study was to determine the impact of route of bortezomib administration on treatment outcomes in previously untreated MM patients in the community oncology setting.
Methods
This was a non-interventional, retrospective, observational cohort study using population-level data from an electronic health records (EHR) database. Key inclusion criteria for the target patient population were: newly diagnosed MM (ICD-9: 203.00); age ≥ 18 years; no prior chemotherapy before initiation of bortezomib-based treatment; initiation of bortezomib-based treatment during the study identification period (January 1, 2011 to November 30, 2012); and ≥ 2 MM-related visits. In order to reflect routine clinical practice in MM, the analysis excluded patients enrolled in randomized clinical trials and those with other cancer diagnoses. Patients were originally followed through May 31, 2013 to allow a minimum 6-month observational window; however, some patients were lost to follow-up prior to 6 months (range, 0.1–44 months). The follow-up period was updated through July 31, 2014. This study was conducted in accordance with the principles of the Declaration of Helsinki and with approval of the central Institutional Review Board at the US Oncology Network (USON).
Data Source
Data were collected via programmatic data queries of the iKnowMed (iKM) EHR database within the McKesson Specialty Health | USON (MSH | USON). iKM captures data on outpatient medical oncology care for patients treated across 19 states in the United States. Overall, the system captures data on approximately 10% of newly diagnosed cancer patients in the United States. Data collected from iKM EHR for this study included patient demographics and clinical characteristics at baseline and data on bortezomib treatment exposure, including route(s) of administration and dose. Payer status and data on claims for services provided within the MSH | USON were obtained from US Oncology’s financial data warehouse.
Data Analysis
Analyses were conducted on a per-protocol basis to evaluate any direct association between the route of administration and treatment exposure/outcomes. The per-protocol population comprised those patients who received bortezomib solely through IV or SC administration, and excluded any patients who switched route of administration.
An algorithm was employed to standardize data on initial bortezomib doses per the bortezomib prescribing information:2 first doses of ≥ 1.5 mg/m2 were adjusted to the 1.5 mg/m2 dose level; doses of ≥ 1.3 mg/m2 but < 1.5 mg/m2 were adjusted to the 1.3 mg/m2 dose level; doses of ≥ 1.0 mg/m2 but < 1.3 mg/m2 were adjusted to the 1.0 mg/m2 dose level; and doses of < 1.0 mg/m2 were adjusted to the 0.7 mg/m2 dose level. Among patients with dose reductions, patients’ subsequent doses with at least a 10% change from initial dose during treatment were adjusted to the standardized dose range of 0.7 to 1.5 mg/m2. Dose reductions and time to dose reduction data were captured at three time points: 16 weeks, 24 weeks, and 32 weeks. These time points were selected to correspond with a bortezomib cycle length of 3–5 weeks and with patient follow-up visits, which typically occur monthly.
Between-group comparisons were conducted using the chi-square test for categorical variables and using the Fisher exact test for sparse data. For comparison of continuous variables, t tests or nonparametric Wilcoxon-Mann-Whitney tests were used. Univariate and multivariate Cox proportional hazards regression model of time to dose reduction was used to identify statistically significant covariates (including route of administration, age, gender, body mass index, practice region, payer status, International Staging System stage, MM subtype, baseline score on the Karnofsky Performance Status Scale, receipt of stem-cell transplant, initial dose, baseline comorbidity, and chemotherapy backbone). Time to event data, including time to dose reduction and OS, were estimated in the overall population using the Kaplan-Meier method.
Results
A total of 1058 patients diagnosed with MM who had no previous chemotherapy and who initiated IV or SC bortezomib during the study identification period (January 1, 2011 to November 30, 2012) were included. Of these patients, 243 switched route of bortezomib administration during treatment (216 received IV then switched to SC; 27 received SC then switched to IV) and were excluded. Of the 815 patients who did not switch route of administration, 436 (53%) received IV bortezomib and 379 (47%) received SC bortezomib (Figure 1).
At baseline, patient demographics and clinical characteristics were broadly similar between the IV and SC groups (Table 1). However, the rate of general comorbidities at baseline, which included weight loss, anorexia, dehydration, diarrhea, dyspnea, hypervolemia, malaise, nausea, pain, and sepsis, was significantly higher in the IV than in the SC group (22% vs 15%; P = 0.0047). Similar proportions of patients receiving IV bortezomib and SC bortezomib underwent stem-cell transplant at any time (n = 158 [36%] and n = 134 [35%], respectively).
Bortezomib was most commonly administered in a doublet regimen of bortezomib plus dexamethasone (61% of IV and 49% of SC patients). Overall, 38% of IV and 50% of SC patients received triplet regimens, most commonly bortezomib-lenalidomide-dexamethasone (25% and 27%, respectively), and bortezomib-cyclophosphamide-dexamethasone (7% and 17%, respectively). Only ~1% of patients in each group received quadruplet regimens (including bortezomib-cyclophosphamide-lenalidomide-dexamethasone and bortezomib-cyclophosphamide-dexamethasone+pegylated liposomal doxorubicin).
Similar initial doses of bortezomib were observed with both routes of administration at medians of 1.29 mg/m2 (range, 0.47–1.96) in the IV group and 1.30 mg/m2 (range, 0.69–1.64) in the SC group (P = 0.0062). Median durations of treatment were 116 days (range, 1–1268) in the IV group and 142 days (range, 1–1053) in the SC group (P = 0.019).
At the 16-week time point, 97 of 815 patients had required bortezomib dose reductions; a greater number of patients required dose reductions in the IV group (57 of 436 patients [13%]) than in the SC group (40 of 379 patients [11%]). A similar pattern of more frequent dose reductions in the IV group than the SC group was observed at the 24- and 32-week time points: 66 of 436 patients (15%) and 45 of 379 patients (12%) in the IV group and SC group, respectively, required dose reductions at 24 weeks, and 69 of 436 patients (16%) and 49 of 379 patients (13%) in the IV group and SC group, respectively, required dose reductions at 32 weeks. At the 16-week time point, the most common dose reductions (based on adjusted first dose and subsequent dose data) were from 1.3 to 1.0 mg/m2.
Among patients who had a dose reduction at 16 weeks, the median time to dose reduction was numerically shorter in the IV group (49 days [range, 7–112]) than in the SC group (56 days [range, 3–105]; mean [standard deviation], 50.2 [27.6] days vs 53.4 [29.4] days for the IV and SC groups, respectively; P = 0.30). Among patients who had a dose reduction by 24 or 32 weeks, the median times to dose reduction were similar between the two groups (24 weeks: 56 days [range, 7–161] and 56 days [range, 3–161] in the IV and SC groups, respectively; 32 weeks: 62 days [range, 7–203] and 63 days [range, 3–217] in the IV and SC groups, respectively).
Across the whole per-protocol population, median time to dose reduction was not reached; however, although the difference was not statistically significant, the overall trend suggests patients receiving IV bortezomib are more likely to require dose reduction than those receiving SC bortezomib (Figure 2). In the univariate regression analysis using the Cox proportional hazards model, no variable was associated with time to dose reduction at 16, 24, and 32 weeks. In the multivariate analysis, initial bortezomib dose was identified as the only statistically significant covariate of time to dose reduction at 16 weeks (hazard ratio [HR], 3.883 [95% confidence interval (CI), 1.052–14.328]; P = 0.0417); at 24 weeks (HR, 4.298 (95% CI, 1.349–13.695); P = 0.0136), and at 32 weeks (HR, 4.879 (95% CI, 1.589–14.98); P = 0.0056).
Data on reasons for treatment discontinuation were available for 311 patients, including 161 patients in the IV group and 150 patients in the SC group. Among these patients, the most common reasons for treatment discontinuation in the IV and SC groups included treatment completion (79 patients [49%] and 63 patients [42%], respectively); progression or relapse (20 patients [12%] and 23 patients [15%], respectively); toxicity (medically required; 23 patients [14%] and 20 patients [13%], respectively); and patient-specific reasons (15 patients [9%] and 15 patients [10%], respectively).
After a median follow-up time of 29.9 months (range, 0.1–44.0) in the IV group and of 22.4 months (range, 0.1–37.1) in the SC group (P < 0.0001), the median OS was not reached (Figure 3). However, OS prospects remained similar in the IV and SC groups, with 1-year survival rates of 84% and 87%, respectively, and 2-year survival rates of 78% and 79%, respectively.
Discussion
The results from this retrospective analysis of EHR data in the US community oncology setting indicate that SC bortezomib is associated with a prolonged duration of treatment, numerically fewer dose reductions, a trend for longer time to dose reduction, and similar 1-year and 2-year OS rates compared with IV bortezomib. In addition, initial bortezomib dose was identified as the only statistically significant covariate of time to dose reduction in multivariate analysis. Together, these findings support our hypothesis that outcomes associated with SC administration of bortezomib are as good as those with IV bortezomib in newly diagnosed MM patients in the clinical setting and reflect the results of the phase 3 MMY-3021 study after prolonged follow-up.3
Notably, the prolonged treatment duration with SC versus IV administration (142 vs 116 days), considered together with the similar median bortezomib dose per month (5.3 and 5.5 mg/m2 in the SC and IV groups, respectively), suggests a potentially clinically significant, higher overall cumulative dose of bortezomib with SC administration and better treatment outcomes. In relation to this, an analysis of data on bortezomib-melphalan-prednisone from the phase 3 VELCADE® as Initial Standard Therapy in Multiple Myeloma: Assessment with Melphalan and Prednisone (VISTA) trial25 has shown that higher cumulative bortezomib dose (due to prolonged treatment duration and/or dose intensity) is associated with improved survival. Prolonged follow-up is required to determine whether the data from this study in a community oncology setting may support these findings from VISTA. This point has become increasingly important as MM is increasingly being considered a continuous therapy disease; randomized studies26 and a meta-analysis27 have reported benefits from maintenance or continuous therapy with novel agents such as bortezomib, lenalidomide, and thalidomide, in terms of increased response rate, PFS, and OS. Indeed, as these community clinical practice data continue to mature, further analyses will be required to evaluate the impact of bortezomib administration route on OS as well as other treatment outcomes such as PFS, time to treatment failure, and time to next therapy.
Nevertheless, in the context of the similar efficacy seen in previous studies with both routes of bortezomib administration,6-18 use of SC versus IV bortezomib may be important in terms of maintaining patients on active therapy for longer. Two survey studies have demonstrated that a majority of patients prefer SC over IV bortezomib,20,21 and one of these studies highlighted other potential contributory factors of improved patient convenience and reduced patient burden with SC administration, as evidenced by reductions in chair time and infusion center visit time compared with IV bortezomib.21 Although SC administration of bortezomib has been associated with injection-site reactions in a number of studies and reports,6,11,21,28,29 including the original phase 3 study,3,4 these have typically been only mild to moderate in severity and have generally resolved rapidly.4,6,11,29 Notably, the incidence and severity of such reactions may be improved through optimizing SC administration techniques to further enhance the tolerability of this route of administration.7,20,30
This study has several limitations. Data were collected for clinical and not research purposes; thus, associations, but not causality, can be detected. Patients and physicians are heterogeneous; therefore, demographics and practice behaviors may differ between clinics. Data collection may not be as complete across the entire population. Missing values may result in exclusion from the study or analysis. Patient treatment history outside the USON may not be fully captured in the iKM EHR. Adverse events are not graded in the iKM EHR. The USON encourages the use of evidence-based treatment guidelines; therefore, results may differ from other community-based practices. In the absence of randomization, the SC bortezomib cohort, on average, appeared to have a lower level of general baseline comorbidities than the IV bortezomib cohort, although the reasons for this difference are not clear. Multivariate regressions cannot fully adjust for the differences between the two arms. Additionally, due to the date of the US Food and Drug Administration approval of SC administration of bortezomib, in January 2012, there was an asymmetric follow-up between patients receiving IV versus SC bortezomib who were included in this study; thus, there is the possibility that a higher proportion of patients in the SC group may have still been receiving bortezomib therapy at the end of the study follow-up period (July 31, 2014), and so a prolonged follow-up period is desirable. The absence of comorbidity data does not indicate absence of the comorbidity; rather, it is the absence of documentation of the comorbidity. Although data-quality checks were conducted, some variables of interest were not complete or available from the data source, and further patient chart reviews will be necessary to warrant research on treatment outcomes in terms of response, PFS, and safety, including peripheral neuropathy. As for comorbidities, the absence of adverse-event data does not indicate absence of adverse events.
The per-protocol analysis approach, excluding patients who switched routes of administration during treatment, may potentially have confounded the investigation of the “true” tolerability of each route of administration. Notably, there was a substantially higher number of patients who switched from IV to SC administration (n = 216) than who switched from SC to IV administration (n = 27). Given the demonstrated improvement in the systemic safety profile of bortezomib with SC versus IV administration in the phase 3 MMY-3021 study,3,4 these “switchers” from IV to SC administration may have comprised those patients who were not tolerating IV bortezomib as well. As a result, there may have been a selection bias for the per-protocol population used for these analyses towards those who better tolerated IV bortezomib. Additionally, the imbalance between the IV and SC groups in terms of use of doublet or triplet regimens may also have affected the rate of dose reductions; a higher proportion of patients in the SC group received triplet therapy, which, compared with doublet therapy, may have required additional dose reductions due to the added toxicity profile of a third agent.
Conclusion
Our findings suggest similar treatment outcomes and an overall trend for prolonged treatment and fewer dose reductions with SC bortezomib compared with IV bortezomib as initial therapy in previously untreated MM patients in the community oncology setting. Overall, these findings suggest similar treatment outcomes and the potential for improved tolerability with SC versus IV bortezomib as initial therapy in routine clinical practice.
References
1. Velcade (bortezomib) for injection [product characteristics]. Beerse, Belgium: Janssen-Cilag International NV; 2009. https://www.ema.europa.eu/docs/en_GB/document_library/EPAR_-_Product_Information/human/000539/WC500048471.pdf. Accessed January 12, 2015.
2. Velcade (bortezomib) [prescribing information]. Cambridge, MA: Millennium Pharmaceuticals Inc; 2015. https://www.velcade.com/files/PDFs/VELCADE_PRESCRIBING_INFORMATION.pdf. Accessed January 12, 2015.
3. Arnulf B, Pylypenko H, Grosicki S, et al. Updated survival analysis of a randomized phase III study of subcutaneous versus intravenous bortezomib in patients with relapsed multiple myeloma. Haematologica. 2012;97(12):1925-1928.
4. Moreau P, Pylypenko H, Grosicki S, et al. Subcutaneous versus intravenous administration of bortezomib in patients with relapsed multiple myeloma: a randomised, phase 3, non-inferiority study. Lancet Oncol. 2011;12(5):431-440.
5. Petrucci MT, Finsinger P, Chisini M, Gentilini F. Subcutaneous bortezomib for multiple myeloma treatment: patients’ benefits. Patient Prefer Adherence. 2014;8:939-946.
6. Lamm W, Drach-Schauer B, Eder S, Drach J. Bortezomib administered subcutaneously is well tolerated in bortezomib-based combination regimens used in patients with multiple myeloma. Oncology. 2013;85(4):223-227.
7. Adam J-P, Lemieux-Blanchard E, Lemieux B, Letarte N. First year experience of subcutaneous bortezomib use in a university teaching hospital. Blood. 2013;122(21):5610.
8. Shah G, Kaul E, Fallo S, et al. Bortezomib subcutaneous injection in combination regimens for myeloma or systemic light-chain amyloidosis: a retrospective chart review of response rates and toxicity in newly diagnosed patients. Clin Ther. 2013;35(10):1614-1620.
9. Sidana S, Faiman B, Elson P, et al. Neuropathy and efficacy of weekly subcutaneous bortezomib in myeloma and AL amyloidosis. Blood. 2013;122(21):1975.
10. Simpson D, McCulloch R, Liang J, et al. Weekly subcutaneous bortezomib is well tolerated and effective as initial therapy of symptomatic myeloma. Blood. 2013;122(21):3229.
11. Takamatsu Y, Imamura Y, Nakazato T, et al. Once-weekly subcutaneous bortezomib plus oral cyclophosphamide and dexamethasone therapy for relapsed or refractory multiple myeloma. Blood. 2013;122(21):5382.
12. Takezako N, Sekiguchi N, Nagata A, et al. Subcutaneous injection of bortezomib, combined with doxorubicin and dexamethasone (sPAD) is effective and safety induction therapy for untreated multiple myeloma patients. Blood. 2013;122(21):5391.
13. Herbaux C, Joao CM, Plocque A, et al. Subcutaneous bortezomib, melphalan and prednisone in elderly newly diagnosed multiple myeloma patients. Blood. 2014;124(21):3481.
14. Merz M, Salwender H, Haenel M, et al. Subcutaneous versus intravenous bortezomib in two different induction therapies for newly diagnosed multiple myeloma: an interim analysis from the prospective GMMG-MM5 trial. Haematologica. 2015;100(7):964-969.
15. Minarik J, Pavlicek P, Pour L, et al. Subcutaneous bortezomib in multiple myeloma patients induces similar therapeutic response rates as intravenous application but it does not reduce the incidence of peripheral neuropathy. PLoS One. 2015;10(4):e0123866.
16. Ohgiya D, Tsuchiya T, Suyama T, et al. An acceptable incidence of infusion site reactions after subcutaneous bortezomib administration in the upper arm in Japanese patients with multiple myeloma. Acta Haematol. 2015;133(1):29-30.
17. Ong SY, Ng HY, Surendran S, et al. Subcutaneous bortezomib combined with weekly cyclophosphamide and dexamethasone is an efficient and well tolerated regime in newly diagnosed multiple myeloma. Br J Haematol. 2015;169(5):754-756.
18. Popat R, Cavenagh JD, Owen RG, et al. Subcutaneous PAD as induction therapy for patients with newly diagnosed myeloma: a phase 2 trial assessing the impact of minimal residual disease (MRD) in patients with deferred autologous stem cell transplantation (PADIMAC). Blood. 2014;124(21):4745.
19. Larocca A, Cavallo F, Magarotto V, et al. Reduced dose-intensity subcutaneous bortezomib plus prednisone (VP) or plus cyclophosfamide (VCP) or plus melphalan (VMP) for newly diagnosed multiple myeloma patients older than 75 years of age. Blood. 2013;122(21):539.
20. Martin JR, Beegle NL, Zhu Y, Agretelis JM. Subcutaneous (SC) administration of bortezomib: a pilot survey of oncology nurses. Blood. 2013;122(21):5585.
21. Barbee MS, Harvey RD, Lonial S, et al. Subcutaneous versus intravenous bortezomib: efficiency practice variables and patient preferences. Ann Pharmacother. 2013;47(9):1136-1142.
22. Brioli A, Zannetti BA, Zamagni E, et al. Peripheral neuropathy induced by subcutaneous bortezomib-based induction therapy for newly diagnosed multiple myeloma. Haematologica. 2014;99(11):242-243.
23. Koh Y, Lee SY, Kim I, et al. Bortezomib-associated peripheral neuropathy requiring medical treatment is decreased by administering the medication by subcutaneous injection in Korean multiple myeloma patients. Cancer Chemother Pharmacol. 2014;74(3):653-657.
24. Lok A, Mocquard J, Bourcier J, et al. Subcutaneous bortezomib incorporated into the bortezomib-thalidomide-dexamethasone regimen as part of front-line therapy in the context of autologous stem cell transplantation for multiple myeloma. Haematologica. 2014;99(3):e33-e34.
25. Mateos MV, Richardson PG, Dimopoulos MA, et al. Effect of cumulative bortezomib dose on survival in multiple myeloma patients receiving bortezomib-melphalan-prednisone in the phase III VISTA study. Am J Hematol. 2015;90(4):314-319.
26. Vande Broek I, Jacobs P. Continuous treatment in multiple myeloma: the future? Transfus Apher Sci. 2013;49(2):147-150.
27. Zou Y, Sheng Z, Lu H, Yu J. Continuous treatment with new agents for newly diagnosed multiple myeloma. Anticancer Drugs. 2013;24(5):527-533.
28. Kamimura T, Miyamoto T, Yokota N, et al. High incidence and severity of injection site reactions in the first cycle compared with subsequent cycles of subcutaneous bortezomib. Int J Hematol. 2013;98(6):694-701.
29. Kamimura T, Miyamoto T, Yokota N, et al. Higher incidence of injection site reactions after subcutaneous bortezomib administration on the thigh compared with the abdomen. Eur J Haematol. 2013;90(2):157-161.
30. Kurtin S, Knop CS, Milliron T. Subcutaneous administration of bortezomib: strategies to reduce injection site reactions. J Adv Pract Oncol. 2012;3(6):406-410.
