A prospective, international phase 2 study of bortezomib retreatment in patients with relapsed multiple myeloma

Authors


  • Prior presentations of this study
  • Petrucci MT, Blau, IW, Corradini P, et al Efficacy and safety of retreatment with bortezomib (VELCADE®) in patients with multiple myeloma: results from a prospective international phase II trial. Blood 2008;112:Abstract 3690. Poster presentation at the 2008 Annual Meeting of the American Society of Hematology (ASH).
  • Petrucci MT, Blau, IW, Corradini P, et al Efficacy and safety of retreatment with bortezomib in patients with multiple myeloma: Results from RETRIEVE, a prospective international phase II study. Blood 2009;114:Abstract 3866. Poster presentation at the 2009 Annual Meeting of the American Society of Hematology (ASH).
  • Petrucci MT, Blau I, Corradini P, et al Efficacy and safety of retreatment with bortezomib in patients with multiple myeloma: Interim results from RETRIEVE, a prospective international phase 2 study. Haematologica 2010;95(s2):Abstract 0377. Poster presentation at the 15th Congress of the European Haematology Association (EHA).

Correspondence: Dr Maria T. Petrucci, Department of Cellular Biotechnology and Haematology, Sapienza University of Rome, via Benevento 6, Rome 00161, Italy.

E-mail: petrucci@bce.uniroma1.it

Summary

Multiple myeloma (MM) typically follows a relapsing course with many patients requiring multiple therapies. This single-arm phase 2 study prospectively evaluated the efficacy and safety of bortezomib retreatment in MM patients who had relapsed after achieving at least a partial response (≥PR) to prior bortezomib-based therapy.

Patients aged ≥18 years, with measurable, secretory MM, who relapsed ≥6 months after prior bortezomib treatment were eligible. Patients received up to eight cycles of bortezomib (±dexamethasone). The primary endpoint was best confirmed response at retreatment; secondary endpoints included duration of response (DOR), time to progression (TTP), and safety. Adverse events (AEs) were graded by National Cancer Institute Common Terminology Criteria for Adverse Events version 3.0.

A total of 130 patients (median of two prior lines of therapy) were enrolled and received retreatment. At retreatment, 28% and 72% of patients received bortezomib and bortezomib-dexamethasone, respectively. Overall response rate was 40%. In patients who achieved ≥PR, median DOR and TTP were 6·5 and 8·4 months, respectively. Thrombocytopenia was the most common grade ≥3 AE (35%). Forty percent of patients experienced neuropathy events, which improved and resolved in a median of 1·5 and 8·9 months, respectively.

In conclusion, bortezomib retreatment was effective and tolerable in relapsed MM patients, with no evidence of cumulative toxicities.

Multiple myeloma (MM) is the second most common haematological malignancy (Dimopoulos & Terpos, 2010) and is a major cause of mortality in the European Union (Ferlay et al, 2010) and USA (Siegel et al, 2012). The use of novel agents (including thalidomide, lenalidomide, and bortezomib) for the front-line treatment of MM has resulted in improved responses and outcomes in patients with relapsed or refractory disease (Kastritis et al, 2009; Lonial, 2010); however, responses are often of limited duration (Kumar et al, 2004). MM typically follows a relapsing course with many patients requiring multiple lines of therapy (Kumar et al, 2004). Relapse is thought to be due, in part, to the changing tumour biology, and the evolution of drug-resistant phenotypes within the tumour (Kumar et al, 2004; Dimopoulos et al, 2011).

Factors influencing treatment choice in the relapsed setting include the duration of response (DOR) to prior therapy, and the associated toxicity profile (Dimopoulos & Terpos, 2010; Lonial, 2010). Re-challenge with the prior regimen may be suitable if a long DOR was obtained following a short course of treatment; this is reflected in the National Comprehensive Cancer Network clinical practice guidelines for MM, which suggest that patients may be retreated with their primary induction therapy if relapse occurs >6 months after completion of initial therapy available at http://www.nccn.org/professionals/physician_gls/f_guidelines.asp). Alternatively, switching to a different treatment may be recommended if only a short DOR (<6 months) was obtained after a long duration of initial therapy (Dimopoulos & Terpos, 2010). The toxicities associated with the prior therapy may also influence treatment choice at relapse; for example, switching to a non-neurotoxic agent may be recommended for patients who experienced peripheral neuropathy (PN) during prior lines of therapy.

Bortezomib (VELCADE®, Millennium Pharmaceuticals Inc. Cambridge, MA, USA) is approved in the USA for the treatment of patients with MM (http://www.velcade.com/Files/PDFs/VELCADE_PRESCRIBING_INFORMATION.pdf). In the European Union, VELCADE is approved for use in combination with melphalan and prednisone for the treatment of patients with previously untreated MM who are ineligible for high-dose therapy and bone marrow transplantation, and also for patients with progressive MM who have received ≥1 prior therapy and have already undergone or are unsuitable for bone marrow transplantation (http://www.ema.europa.eu/docs/en_GB/document_library/EPAR_-_Summary_for_the_public/human/000539/WC500048136.pdf). In combination therapy, bortezomib has demonstrated substantial efficacy in several phase 3 studies of patients with previously untreated MM (San Miguel et al, 2008; Cavo et al, 2010; Harousseau et al, 2010; Mateos et al, 2010; Sonneveld et al, 2012). In the relapsed MM setting, bortezomib has shown activity as a single agent (Richardson et al, 2003, 2005, 2007; Jagannath et al, 2004), in combination with dexamethasone (Richardson et al, 2003; Jagannath et al, 2004; Mikhael et al, 2009) or liposomal doxorubicin (Orlowski et al, 2007), and as part of multi-drug regimens (Terpos et al, 2008; Chanan-Khan et al, 2009; Kim et al, 2010; Lee et al, 2010; Offidani et al, 2011).

Several previous retrospective studies (Ciolli et al, 2007; Conner et al, 2008; Rubio-Martinez et al, 2008; Wolf et al, 2008; Hrusovsky et al, 2010; Taverna et al, 2012) and a small prospective phase 4 study in the USA (EVEREST) (Sood et al, 2009) have suggested that bortezomib retreatment is well tolerated and effective, resulting in substantial clinical response rates with no additional or cumulative toxicity. The feasibility of bortezomib retreatment is also supported in the long-term follow-up of one phase 2 trial (Berenson et al, 2005). This international, open-label, phase 2 trial prospectively evaluated the efficacy and safety of bortezomib retreatment in patients with MM, who had relapsed/progressed after achieving at least partial response (PR) to prior bortezomib-based therapy.

Methods

Study design

This prospective, single-arm, open-label, phase 2 study, conducted at 55 centres in Austria, Belgium, France, Germany, Greece, Italy, Luxembourg, Portugal and Spain, evaluated the efficacy and safety of retreatment with bortezomib ± dexamethasone in patients with MM who had relapsed after achieving ≥PR with initial bortezomib-based therapy.

Patients

Patients with measurable secretory MM were eligible for enrolment if they met the following criteria: ≥18 years of age; previously tolerated bortezomib at a dose of 1·0 or 1·3 mg/m2 alone or in combination with other agents; achieved complete response (CR) or PR upon completion of prior bortezomib therapy; had progressive disease (PD) after previously achieving PR or had relapsed from CR (as defined by European Group for Blood and Marrow Transplantation [EBMT] criteria (Blade et al, 1998)) and ≥6 months elapsed since the last dose of bortezomib; Karnofsky Performance Status (KPS) of ≥60%; life expectancy of >3 months; adequate haematological parameters and renal/hepatic function within 14 d before enrolment, including: platelet count ≥50 × 109/l; haemoglobin ≥75 g/l; absolute neutrophil count ≥0·75 × 109/l; corrected serum calcium <3·5 mmol/l; aspartate aminotransferase ≤2·5 × the upper limit of normal (ULN); alanine aminotransferase ≤2·5 × ULN; total bilirubin ≤1·5 × ULN; and calculated or measured creatinine clearance ≥20 ml/min.

Key exclusion criteria included: a history of PD, minimal response (MR) or stable disease after prior exposure to bortezomib; achievement of CR or PR followed by relapse while on bortezomib therapy; receipt of chemotherapy, radiotherapy, antibody immunotherapy, or experimental therapy for MM since last dose of bortezomib (except for maintenance therapy with thalidomide, interferon, or dexamethasone [or equivalent]); grade ≥2 PN or neuropathic pain as defined by the National Cancer Institute Common Terminology Criteria for Adverse Events (NCI-CTCAE) version 3.0 (http://ctep.cancer.gov/protocolDevelopment/electronic_applications/docs/ctcaev3.pdf).

All patients provided written informed consent, and ethics committees at all participating institutions approved the study, which was conducted according to the Declaration of Helsinki and International Conference on Harmonization Guidelines for Good Clinical Practice. The study was registered at ClinicalTrials.gov (NCT00431769) and EudraCT (2005-005819-26).

Study endpoints

The primary endpoint was best confirmed response to bortezomib retreatment by EBMT criteria (Blade et al, 1998), as reviewed by an Independent Data Monitoring Committee (IDMC). Secondary endpoints included: DOR and time to progression (TTP) for best confirmed response; comparison of best response to prior bortezomib-based therapy with single best response to bortezomib retreatment; comparison of best response by type of bortezomib retreatment (bortezomib ± dexamethasone); and safety.

Bortezomib retreatment

Patients received bortezomib on days 1, 4, 8, and 11 in 21-day cycles for a maximum of eight cycles. The last tolerated dose of bortezomib (1·0 or 1·3 mg/m2) was used as the starting dose for retreatment. Patients were escalated from the 1·0 mg/m2 to 1·3 mg/m2 dose at the investigator's discretion. Bortezomib was administered with or without dexamethasone at the investigator's discretion and per the investigator's standard of care. Dose interruptions and/or reductions were permitted based on the severity and duration of adverse events (AEs). For any haematological or grade ≥3 non-haematological AEs within a cycle, bortezomib was held for up to two weeks until symptoms resolved or improved to grade ≤2; bortezomib could then be re-initiated at a 25% reduced dose. The minimum dose of bortezomib permitted was 0·7 mg/m2.

Bortezomib retreatment was discontinued if the patient experienced PD after ≥2 cycles of treatment, unacceptable toxicity, or insufficient recovery from the previous treatment cycle that delayed the start of the next treatment cycle by more than two weeks (except where clinical benefit was observed and the sponsor approved continuation of treatment). Dose interruption or discontinuation was not required for lymphopenia of any grade. PN was managed using standard dose-modification criteria (Richardson et al, 2009).

Patients could receive selected medications and/or supportive therapies including granulocyte-colony stimulating factor, granulocyte macrophage-colony stimulating factor, erythropoietins, bisphosphonates, kyphoplasty, local radiation therapy, analgesics, and loperamide according to site standards and/or at the investigator's discretion. Prophylaxis for herpes zoster virus was not given as standard in this study.

Assessments

Response assessments

Serum and urine samples were collected for M-protein quantification and assessment of disease status every six weeks during treatment and six weeks after the last bortezomib dose. All responding patients were followed until documented PD. M-protein levels were assessed every eight weeks during follow-up. Responses were assessed according to EBMT criteria, and reviewed by an IDMC. CR was defined as 100% M-protein reduction by electrophoresis and negative serum and urine immunofixation. Analyses of time to response (TTR), DOR, and TTP were based on best confirmed responses.

Safety assessments

AEs were graded according to NCI-CTCAE version 3.0 http://ctep.cancer.gov/protocolDevelopment/electronic_applications/docs/ctcaev3.pdf). The Functional Assessment of Cancer Therapy/Gynecologic Oncology Group-Neurotoxicity (FACT/GOG-Ntx) questionnaire (Calhoun et al, 2003) was completed by all patients during the screening period, before dosing on day 1 of each treatment cycle, and at the end-of-treatment visit, and was used as a checklist to assist in the evaluation of neuropathy symptoms. For analyses of time to resolution and improvement of neuropathy, neuropathy AEs with start and stop dates one day apart were linked into one event.

Statistical methods

The sample size was determined based on an expected response rate of 38%, and with the precision of the 95% confidence interval (CI) to be approximately ±9% (i.e. 29% to 47%). The estimate of expected response rate was based on the findings of the phase 3 APEX study of single-agent bortezomib in patient with relapsed and/or refractory MM (Richardson et al, 2005). Under these conditions, a sample size of 112 evaluable patients was required. With a withdrawal rate of 10% assumed based on previous studies, it was planned to enrol a total of 125 patients.

All analyses were performed using response data reviewed by the IDMC. Kaplan–Meier analyses were used to estimate TTR, DOR, and TTP, and distribution of the time to resolution and time to improvement of neuropathy events. TTR was defined as the time from first study drug administration to the date of best confirmed response or censoring. DOR was defined as the time from date of best confirmed response to date of relapse/PD or censoring; TTP was defined as the time from first study drug administration to the date of relapse/PD or censoring. Patients were censored for discontinuation, receipt of alternative therapy, study end/end of treatment, or death.

Results

Patients

In total, 132 patients were screened for this study, and two patients were excluded as screening failures. The remaining 130 patients were enrolled, received at least one dose of bortezomib retreatment, and were included in the safety population. Of these 130 patients, 126 had undergone at least one post-baseline efficacy assessment and were included in the evaluable population; 4 patients were excluded due to discontinuation (= 2; patient request, AE) or withdrawal (= 2; major protocol violations) prior to undergoing a first post-baseline efficacy assessment. Patient baseline characteristics for the safety population are summarized in Table 1. Patients had a median age of 67 years (range 38–86), 57% were male, 98% were Caucasian, 61% had a KPS of ≥90%, and 52% had IgG disease. The median time from MM diagnosis was 4·5 years (range 0·8–13·9). Patients had received a median of two prior lines of therapy (range 1–7), including a bortezomib-based regimen; prior therapies included steroids (89%), alkylating agents (77%), anthracyclines (58%), and thalidomide (31%), or a combination of these agents. Thirty percent of patients had received prior high-dose chemotherapy and stem cell transplant.

Table 1. Patient demographics and baseline disease characteristics
Characteristic= 130
Median age, years (range)67 (38–86)
Male, n (%)74 (57)
Race, n (%)
Caucasian127 (98)
Other/not available3 (2)
Karnofsky Performance Status, n (%)
≤70%21 (16)
80%29 (22)
≥90%79 (61)
Missing1 (0·8)
Myeloma type, N (%)
IgG67 (52)
IgA34 (26)
IgM2 (2)
Light chain22 (17)
Unspecified5 (4)
Median time since diagnosis, years (range)4·5 (0·8–13·9)
Median no. of prior lines of therapy, including prior bortezomib, n (range)2 (1–7)
115 (12)
280 (62)
323 (18)
≥412 (9)
Other prior therapies received, n (%)
Steroids115 (89)
Alkylating agents100 (77)
Anthracyclines75 (58)
Thalidomide40 (31)
≥2 of the above115 (89)
≥3 of the above78 (60)
All of the above21 (16)
Prior high-dose therapy/stem cell transplant, n (%)39 (30)

Prior bortezomib treatment

Per protocol, all patients had received bortezomib-based treatment as their last prior line of therapy; 48 (37%) patients had received single-agent bortezomib and 82 (63%) had received bortezomib-dexamethasone (Table 2). One hundred and seventeen (90%) patients had received ≥4 treatment cycles of prior bortezomib-based therapy. The best response to prior bortezomib treatment was CR in 34 (26%) patients and PR in 96 (74%) patients. The median DOR to prior bortezomib for patients achieving ≥PR to bortezomib retreatment in the present study was 13·4 months (range 0–38·5; 95% CI: 10·6–16·0). The median time to relapse/disease progression after prior bortezomib therapy was 17·7 months (range 4·3–41·4).

Table 2. Prior exposure to bortezomib and bortezomib retreatment
 = 130
  1. a

    Time to relapse/disease progression was measured from start of prior bortezomib treatment.

  2. b

    One patient had a time from prior bortezomib retreatment of five months; with agreement of the sponsor, this patient was deemed to have fulfilled the inclusion criterion of a time from bortezomib retreatment of ≥6 months.

  3. c

    Dose intensity is equivalent to the total bortezomib dose administered (mg)/body surface area (m2).

Prior bortezomib treatment
Patients receiving ≥4 treatment cycles, n (%)117 (90)
Treatment received, n (%)
Single-agent bortezomib48 (37)
Bortezomib in combination with other agents82 (63)
Last tolerated dose, n (%)
1·0 mg/m238 (29)
1·3 mg/m281 (62)
Other11 (9)
Median duration of therapy, months (range)4·9 (1–23)
Median time to relapse/disease progression, months (range)a17·7 (4·3–41·4)
Bortezomib retreatment
Median time from prior bortezomib treatment, months (range)b13·9 (5–39)
Median number of treatment cycles, n (range)7 (1–8)
Patients completing all 8 cycles, n (%)56 (43)
Treatment received, n (%)
Single-agent bortezomib36 (28)
Bortezomib-dexamethasone94 (72)
Median dose intensity of bortezomib, mg/m2 (range)c28·2 (2·6–42·2)
Last dose administered on study, n (%)
0·7 mg/m216 (12)
1·0 mg/m245 (35)
1·3 mg/m269 (53)

Summary of bortezomib retreatment

Details of bortezomib retreatment are summarized in Table 2. The median time from prior bortezomib treatment was 13·9 months (range 5–39). Thirty-six (28%) patients received single-agent bortezomib, and 94 (72%) patients received bortezomib-dexamethasone. Patients received a median of seven cycles (range 1–8) of bortezomib retreatment. Fifty-six (43%) patients completed all eight treatment cycles; AEs (30%) and PD (31%) were the most common reasons cited for patients failing to complete all eight cycles.

Efficacy of bortezomib retreatment

The overall best confirmed response rate (ORR; CR + PR) to bortezomib retreatment in the evaluable population (= 126) was 40%; one (1%) patient achieved a CR, and 49 (39%) patients achieved a PR. Minimal response was achieved by 23 (18%) patients (Table 3). In 27 (21%) patients, best confirmed response status was not assessable or unknown.

Table 3. Best confirmed response to bortezomib retreatment by M-protein assessment
Best confirmed response= 126
  1. ORR, overall response rate; CR, complete response; PR, partial response; MR, minimal response; NC, no change; PD, progressive disease.

  2. a

    Patients who did not have a best confirmed response within the study period were included in this category; response was unknown or could not be assessed in nine patients, and 18 patients had a single best response (see Table 5) but no confirmatory response assessment.

ORR (CR + PR), N (%)50 (40)
CR1 (1)
PR49 (39)
ORR + MR73 (58)
MR23 (18)
NC, N (%)14 (11)
PD, N (%)12 (10)
Unknown/unable to assessa, n (%)27 (21)

Best confirmed response data were also analysed by patient subgroup, including type and dose of bortezomib administered, age, and number of prior therapies received (Table 4). Of note, the ORR appeared to decrease with increasing number of prior therapies, with 67%, 39%, 33%, and 25% of patients who received 1, 2, 3 and ≥4 prior lines of therapy, respectively, achieving ≥PR. Patients who had undergone prior treatment with thalidomide had a slightly lower ORR (33%) compared with the overall population (40%).

Table 4. Overall response rate (best confirmed response) to bortezomib retreatment by patient subgroup
Patient sub-group n ORR, n (%)
  1. ORR, overall response rate.

Bortezomib treatment
Single agent3411 (32)
Bortezomib-dexamethasone9239 (42)
Bortezomib dose
≤1·0 mg/m23412 (35)
1·3 mg/m29238 (41)
Age
≤65 years5324 (45)
>65 years7326 (36)
Number of prior lines of therapy
11510 (67)
27830 (39)
3217 (33)
≥4123 (25)
Prior thalidomide only248 (33)
Prior stem cell transplant only2410 (42)
Both therapies131 (8)
Neither therapy6531 (48)

A comparison of patients' single best response (confirmed or unconfirmed) to bortezomib retreatment and best response to prior bortezomib therapy is shown in Table 5. In the evaluable population, 69 (55%) patients achieved a single best response of CR (= 1) or PR (= 68). It appeared that a higher proportion of patients who had achieved CR versus PR to prior bortezomib therapy went on to achieve a single best response of ≥PR to bortezomib retreatment (63% vs. 52%, respectively). Further, a higher proportion of patients who previously achieved a best response of CR or PR with bortezomib combination therapy versus single-agent bortezomib went on to achieve ≥PR with bortezomib retreatment (59% vs. 48%) (Table 5).

Table 5. Single best response to bortezomib retreatment stratified by best confirmed response to prior bortezomib therapy
Best confirmed response to prior bortezomib therapySingle best response to bortezomib retreatment, n (%)a
CRPRMRNCPD
  1. CR, complete response; PR, partial response; MR, minimal response; NC, no change; PD, progressive disease.

  2. a

    Of the 126 patients in the evaluable population, 117 patients had a single best response recorded, and nine patients were classified as ‘Unknown/unable to assess’.

Any prior bortezomib treatment
CR (= 32)1 (3)19 (59)4 (13)6 (19)1 (3)
PR (= 94)049 (52)21 (22)10 (11)6 (6)
Single-agent bortezomib
CR (= 14)1 (7)6 (43)3 (21)4 (29)0
PR (= 32)015 (47)9 (28)2 (6)3 (9)
Bortezomib in combination
CR (= 18)013 (72)1 (6)2 (11)1 (6)
PR (= 62)034 (55)12 (19)8 (13)3 (5)

The median time to best confirmed response of ≥PR in the evaluable population was not estimable; 76 patients were censored due to discontinuation (= 47), end of treatment (= 25), receipt of alternative therapy (= 2), or death (= 2). In the 50 patients who achieved a best confirmed response of ≥PR to bortezomib retreatment, the median DOR was 6·5 months (range 0·6–19·3; 95% CI: 5·0–7·1) (Fig 1A); 16 patients were censored due to receipt of alternative therapy (= 6), discontinuation (= 5), death (= 3), or study end (= 2). The median DOR did not appear to be affected by the number of prior therapies received (5·6, 6·7, 4·9, and 7·1 months for 1, 2, 3, and ≥4 prior lines of therapy, respectively); however, the median DOR appeared longer in patients aged >65 years versus ≤65 years (7·0 vs. 5·4 months), and slightly longer in patients who received retreatment with single-agent bortezomib versus bortezomib-dexamethasone (7·1 vs. 6·5 months).

Figure 1.

Kaplan–Meier analyses of (A) duration of response, calculated from the time of best response and (B) time to progression, calculated from the time of the first study drug administration, in patients with a best confirmed response of ≥partial response during bortezomib retreatment. In both analyses, 16 patients were censored due to receipt of alternative therapy (= 6), discontinuation (= 5), death (= 3), or study end (= 2). 95% CI, 95% confidence interval.

Outcomes

The median TTP after prior bortezomib in patients who achieved a best confirmed response of ≥PR to bortezomib retreatment in the present study was 18·9 months (range 0–41·4; 95% CI: 15·4–21·6). TTP was 8·4 months (range 3·3–20·7; 95% CI: 7·9–9·7) in the 50 patients who achieved ≥PR during bortezomib retreatment (Fig 1B), with 16 patients being censored for the reasons described for DOR; however, methodological differences in TTP calculation precluded a direct comparison of TTP to prior bortezomib and to bortezomib retreatment. In patients achieving a best confirmed response of ≥PR with bortezomib retreatment, there was no apparent effect of the number of prior lines of therapy (8·4, 9·0, 6·7, and 8·7 months for 1, 2, 3, and ≥4 prior therapies, respectively), nor of the type of administered prior therapy (single-agent bortezomib: 8·4 months; bortezomib-dexamethasone: 8·4 months) on median TTP; however, the median TTP appeared longer in patients aged >65 vs. ≤65 years (8·7 vs. 7·9 months).

Safety

AEs associated with bortezomib retreatment are summarized in Table 6. Ninety-eight percent of patients experienced ≥1 treatment-emergent AE, and 91% of patients experienced ≥1 drug-related AE. The most common all-grade AEs occurring in ≥25% of patients were thrombocytopenia (55%), neuropathy (40%), anaemia (37%), diarrhoea (35%), and constipation (28%). Grade ≥3 AEs were reported by 60% of patients, the most common of which was thrombocytopenia (35%). Rates of new-onset grade 3/4 AEs overall, and grade 3/4 thrombocytopenia and grade 3/4 anaemia tended to be higher in earlier than in later treatment cycles (data not shown), indicating that these events were transient and reversible. There was no evidence for cumulative toxicities with increasing cycle number.

Table 6. Overview of adverse eventss during bortezomib retreatment
Number of patients with AEs, n (%)= 130
All-grade AEsaGrade ≥3 AEsb
  1. AE, adverse event; SAE, serious adverse event.

  2. a

    Most common all-grade AEs in ≥25% of patients are shown.

  3. b

    Most common grade ≥3 AEs in ≥5% of patients are shown.

  4. c

    Includes PN, motor neuropathy, sensory neuropathy, polyneuropathy, neuralgia, burning sensation, dysesthesia, hyperesthesia, hypoesthesia, and paraesthesia.

Any AE127 (98)78 (60)
Most common AEs
Thrombocytopenia71 (55)45 (35)
Neuropathyc52 (40)
Anaemia48 (37)7 (5)
Diarrhoea45 (35)9 (7)
Constipation36 (28)
Neutropenia9 (7)
Pneumonia8 (6)
Any SAE41 (32)
AEs leading to dose reduction28 (22)
AEs leading to treatment discontinuation27 (21)
Treatment-related AEs leading to treatment discontinuation17 (13)
AEs leading to death8 (6)

Severe AEs were reported by 41 (32%) patients, the most common of which were pyrexia (5%), pneumonia (5%), respiratory tract infections (5%), thrombocytopenia (4%), and diarrhoea (4%). Bortezomib dose reduction was required in 28 (22%) patients. Twenty-seven (21%) patients experienced AEs that resulted in treatment discontinuation; the most common reasons were neuropathy (6%) and diarrhoea (3%). Eight (6%) patients experienced treatment-emergent AEs leading to on-study death; of these, two (sepsis and stroke) were considered as possibly and probably related to treatment, respectively. The remaining six deaths (acute myocardial infarction (= 1), cholecystitis acute (= 1), decubitus ulcer (= 1), embolism (= 1), and progressive disease (= 2)) were not considered to be related to treatment.

Based on responses to the additional concerns section of the FACT/GOG-Ntx questionnaire, 94 (72%) patients reported symptoms of neuropathy (of any grade) at baseline; however, only 44 (34%) patients had confirmed symptoms of sensory or motor neuropathy at baseline neurological examination. During bortezomib retreatment, the incidence of neuropathy events of all grades was 40%, with peripheral sensory neuropathy being the most frequent (17%). Eleven (9%) patients experienced grade ≥3 neuropathy during treatment, of which the most frequent individual events were peripheral sensory neuropathy (3%), followed by peripheral neuropathy (2%), and neuralgia (2%). No grade 4 neuropathy was reported. No clinically relevant differences were observed in the proportion of patients who had neuropathy (by FACT/GOG-Ntx) at baseline between patients with or without neuropathy AEs during the treatment phase, and between patients with starting doses of ≤1·0 or 1·3 mg/m2 bortezomib.

In patients who experienced PN during the study period, the median time to onset of neuropathy for patients with a starting dose of ≤1·0 mg/m2 (reduced dose of bortezomib received due to prior PN) was 4·9 months (range 0·2–19·7); however, the median time to onset of neuropathy was not reached in patients with a starting dose of 1·3 mg/m2 bortezomib (range 0·1–23·5 months). The median time to improvement of treatment-emergent neuropathy was 1·5months (range 0·03–25·7), and the median time to resolution of neuropathy was 8·9 months (range 0·03–26·8). The median time to resolution or improvement of treatment-emergent neuropathy was shorter for patients with a starting dose of ≤1·0 mg/m2 bortezomib (5·2 months (range 0·2–26·8) and 1·1 months (range 0·1–25·7), respectively) compared with patients with a starting dose of 1·3 mg/m2 bortezomib (8·9 months (range 0·03–19·6) and 2·1 months (range 0·03–19·6) respectively).

Discussion

This large, multicentre, phase 2 study represents the first international prospective investigation of bortezomib retreatment in patients with relapsed MM. The results demonstrate the efficacy of bortezomib retreatment (± concomitant dexamethasone) in patients who had previously responded to, and relapsed ≥6 months after, completion of prior bortezomib therapy. Forty percent of patients achieved ≥PR in response to retreatment, and a further 18% of patients achieved MR, giving a combined CR + PR + MR rate of 58%.

The results presented here support those of the small US prospective study (EVEREST), in which 50% of patients achieved ≥PR to bortezomib retreatment (Sood et al, 2009). Several retrospective studies have also assessed the efficacy of bortezomib retreatment in patients achieving ≥PR to initial bortezomib treatment (Ciolli et al, 2007; Conner et al, 2008; Rubio-Martinez et al, 2008; Wolf et al, 2008; Hrusovsky et al, 2010), with ORRs (≥PR) in these studies ranging from 21–73%. The variation in response rates between different studies may be due to the different criteria for response assessment, the size of the patient cohorts, factors such as patient selection bias, and the use of bortezomib in combination with agents other than dexamethasone.

As response status was not assessable or unknown in 21% of patients after bortezomib retreatment, it is possible that the calculated ORR may have been an underestimation of the true response rate achieved after bortezomib retreatment. Further, the low CR rate observed with bortezomib retreatment in this study may be due to the stringent response assessment criteria used, and a lack of confirmatory response assessments in many patients. Indeed, immunofixation and bone marrow data were not collected for the majority of patients in this study; patients with incomplete CR assessment were therefore counted as having achieved PR, thus potentially leading to an underestimation of the true CR rate in this population.

Patient subgroup analyses were suggestive of a trend for higher PR rates to bortezomib retreatment in patients who achieved a CR, rather than PR, to prior bortezomib therapy; a result consistent with that of the RESERVE study (Hrusovsky et al, 2010). A trend towards improved response was also observed in patients who received 1·3 mg/m2 versus ≤1·0 mg/m2 bortezomib (consistent with the phase 2 CREST study (Jagannath et al, 2004)), and in patients who received bortezomib-dexamethasone versus single-agent bortezomib. Although the EVEREST prospective (Sood et al, 2009) and RESERVE retrospective studies (Hrusovsky et al, 2010) reported similar response rates after retreatment with bortezomib or bortezomib-dexamethasone, concomitant dexamethasone was associated with improved responses in 18% of patients who had previously achieved a suboptimal response to bortezomib monotherapy in the SUMMIT study (Richardson et al, 2003), and added dexamethasone was associated with improved responses in 23% of patients in a phase 3 study of subcutaneous versus intravenous bortezomib in relapsed and/or refractory MM (Arnulf et al, 2012). Reflecting this apparent improved activity with addition of another agent, more extensive combination regimens based on bortezomib-dexamethasone are showing utility in the relapsed and/or refractory setting (Offidani et al, 2011; Ahn et al, 2012; Berenson et al, 2012; Garderet et al, 2012; Jimenez-Zepeda et al, 2012). The observed decrease in ORR with increasing number of prior therapies (excluding prior bortezomib) in the present study is consistent with the progressive nature of MM (Kumar et al, 2004). Prior thalidomide appeared to be associated with poorer ORR compared with patients who had not received prior thalidomide treatment, supporting the results of a sub-analysis of the phase 3 APEX study (Vogl et al, 2009).

Although methodological differences precluded a direct comparison of DOR and TTP with prior bortezomib therapy and bortezomib retreatment, median DOR and TTP both appeared shorter with bortezomib retreatment. These observations are also consistent with the progressive nature of MM (Kumar et al, 2004) and have been confirmed in the EVEREST prospective (Sood et al, 2009) and RESERVE retrospective studies (Hrusovsky et al, 2010). Despite being shorter than the median DOR and TTP to prior bortezomib, a median DOR and TTP of 6·5 months and 8·4 months following retreatment, respectively, could potentially offer a survival benefit to MM patients who have already received multiple prior lines of therapy. These data are also consistent with a TTP of 6·2 months observed in the phase 3 APEX study of bortezomib-naïve patients with relapsed/refractory MM (Richardson et al, 2005) and support the hypothesis that bortezomib retreatment does not negatively influence TTP (e.g. via selection of bortezomib-resistant tumour cell clones).

Bortezomib retreatment appeared to be well tolerated, with patients receiving a median of seven treatment cycles. The safety profile was consistent with that seen in phase 2 and 3 studies of bortezomib in patients with relapsed MM (Richardson et al, 2003, 2005; Jagannath et al, 2004) and previously untreated MM (San Miguel et al, 2008). The most frequent AEs were thrombocytopenia and neuropathy, both of which are commonly associated with bortezomib use; AEs were manageable and transient, in line with previous reports (Richardson et al, 2005, 2006, 2009; Jagannath et al, 2008; Lonial et al, 2008; San Miguel et al, 2008). Notably, the presence of baseline neuropathy symptoms (as assessed by FACT/GOG-Ntx) appeared to have no effect on rates of PN AEs during bortezomib retreatment; therefore, retreatment with bortezomib is a feasible treatment option for patients who may be prone to developing bortezomib-associated PN.

In summary, these data suggest that retreatment with bortezomib (± dexamethasone) represents a feasible and effective treatment option for patients with relapsed MM who previously responded to bortezomib, with no evidence for cumulative toxicities, thus representing a potential alternative to initiating subsequent line therapy with a different class of agents.

Acknowledgements

The authors would like to acknowledge the writing assistance of Jane Saunders and Emma Landers of FireKite in the development of this manuscript, which was funded by Millennium Pharmaceuticals, Inc., and Janssen Global Services.

Funding

This work was supported by Janssen Pharmaceutica N.V. J.B would like to acknowledge funding from the Instituto Carlos III (RD06/0020/0005).

Author contributions

MTP, PG, PC, AT, MAD, IWB, JD, and JB performed the research. MTP, RA, and JB designed the research study. MTP, RA, NA, EB, and VM analysed the data. MTP, RA, NA, and EB drafted the paper. All authors reviewed the paper and revised it critically, and approved the final submitted version of the paper.

Disclosure of conflicts of interest

MTP: Honoraria (Janssen, Celgene); PC: Honoraria (Janssen, Celgene, Novartis); MAD: Consultancy (Millennium, Ortho-Biotech), Honoraria (Ortho-Biotech); JD: Honoraria (Janssen, Celgene); RA: Employment (Janssen), Stock options (Janssen); EB: Employment (Janssen); JB: Honoraria (Janssen, Celgene, Merck, Onyx Pharmaceuticals), Advisory board (Janssen, Celgene), Grant support (Janssen, Celgene). All remaining authors have declared no conflicts of interest.

Ancillary