A phase 1/2 study of carfilzomib in Japanese patients with relapsed and/or refractory multiple myeloma

Summary

We conducted a phase 1/2 study of single‐agent carfilzomib in Japanese patients with relapsed/refractory multiple myeloma. Safety, pharmacokinetics and pharmacodynamics of carfilzomib were examined in phase 1. The primary endpoint in phase 2 was the overall response rate (ORR). Carfilzomib was administered in a twice‐weekly, consecutive‐day dosing schedule. In Phase 1, doses of 15 or 20 mg/m² were administered on this schedule or 20 mg/m² on Days 1 and 2 of Cycle 1 and then 27 mg/m² in the 20/27 mg/m² cohort. Patients had a median of five prior therapies, including bortezomib and an immunomodulatory agent. The dose level did not reach the maximum tolerated dose. The most common adverse events were haematological. Notably, carfilzomib either induced new hypertension (n = 4) or aggravated previously existing hypertension (n = 6) in 10 of 50 patients. Four of the eight patients who previously experienced peripheral neuropathy (PN) experienced a recurrence with carfilzomib use, but no new cases of PN occurred. The ORR of the 20/27 mg/m² 40 patient cohort was similar to that in the pivotal US study. The dose was efficacious and tolerable in heavily pre‐treated Japanese patients; however, meticulous control of hypertension may be necessary for further carfilzomib use.

Multiple myeloma (MM) is characterized by the abnormal accumulation of clonal plasma cells in the bone marrow. In the United States (US), an estimated 21 700 new cases of MM and 10710 deaths were predicted in 2012 (American Cancer Society 2012); in Japan, an estimated 6860 new cases of MM and 4066 deaths (http://ganjoho.jp/reg_stat/statistics/dl/) were predicted. Over the past decade, the introduction of immunomodulators, such as thalidomide and lenalidomide, and a proteasome inhibitor (PI) bortezomib (Velcade^®, Millennium Pharmaceuticals, Cambridge, MA, USA) has significantly improved the overall survival (OS) of MM patients (Jemal et al, 2010).

Carfilzomib (Kyprolis^®, Onyx Pharmaceuticals, an Amgen subsidiary, South San Francisco, CA, USA) is one of the second‐generation epoxyketone PIs that can irreversibly inhibit chymotrypsin‐like activity (Demo et al, 2007; Parlati et al, 2009). Carfilzomib induced apoptosis in myeloma cell lines and primary myeloma cells from patients whose disease was resistant to available therapies, including bortezomib (Kuhn et al, 2007). In addition, carfilzomib has fewer off‐target effects, which may explain a lack of neurodegeneration in vitro, and less neurotoxicity in animal studies (Arastu‐Kapur et al, 2011). Therefore, carfilzomib was expected to circumvent some of the clinical adverse events (AEs) of bortezomib, particularly peripheral neuropathy (PN) (Bruna et al, 2010; Arastu‐Kapur et al, 2011).

In fact, PN did not appear to be clinically limiting for extended carfilzomib treatment (Vij et al, 2012a,b). An alternative subcutaneous route of administration for bortezomib reduced the incidence of PN without reducing efficacy (Moreau et al, 2011). Despite these facts, the subcutaneous administration of bortezomib has still caused gastrointestinal (GI) AEs in patients with MM (Moreau et al, 2011). Carfilzomib has also caused GI AEs; however, those occurring with carfilzomib use were mild and manageable with routine supportive care (Siegel et al, 2012; Vij et al, 2012b).

Almost all patients with MM will eventually relapse, and new treatment options are needed. Accordingly, to change the intrinsic natural behaviour of MM, it was suggested that the use of higher doses, combination regimens and early treatment of the disease course without AEs are required for the new treatment using carfilzomib (Reece, 2012). However, discontinuation or dose reduction was inevitable because of increased toxicities in some of the carfilzomib trials using combination therapies (Bringhen et al, 2014; Sonneveld et al, 2015).

In an open‐label, phase 2 pilot study (PX‐171‐003‐A0) of carfilzomib in 46 patients with relapsed/refreactory MM (RRMM), patients treated with carfilzomib 20 mg/m² achieved an overall response rate (ORR) of 16·7%, with manageable toxicities (Jaganath et al, 2012). The study was subsequently amended to include an expanded dosing cohort with a scheduled dose escalation from 20 to 27 mg/m² beginning in the second cycle (PX‐171‐003‐A1 [003‐A1]). In 003‐A1, which included 266 patients with RRMM, the ORR was 23·7%; these results were the basis for the accelerated approval of carfilzomib by the US Food and Drug Administration (Siegel et al, 2012). In the present study, we used the same dose escalation schedule as 003‐A1, with one exception: 27 mg/m² was started on Day 8 of Cycle 1. With this method, we carefully characterized the AEs of carfilzomib in a phase 1/2 single‐agent carfilzomib study in Japanese patients with RRMM.

Methods

Study design

This was a multicentre, open‐label phase 1/2 study (ONO‐7057‐01) in Japanese patients with RRMM. The safety, tolerability, efficacy, pharmacokinetics (PK) and pharmacodynamics of carfilzomib were examined in phase 1 for intravenously administered carfilzomib at doses of 15, 20 and 20/27 mg/m². Phase 2 examined the safety and efficacy of carfilzomib at the recommended dose determined in phase 1. The primary endpoint in phase 2 was the ORR. Secondary endpoints included the duration of response (DOR), Progression‐free survival (PFS) and OS. It was planned that three or six patients for each cohort in phase 1 and 24 patients in phase 2 were enrolled. If dose‐limiting toxicity (DLT) occurred in one of three patients, an additional three patients were enrolled; if the incidence of DLT was two of three patients or three of six patients, the previous dose level was used as the recommended dose in phase 2.

Eligibility

The main patient inclusion criteria were age ≥20 years and Eastern Cooperative Oncology Group (ECOG) performance status (PS) of 0–2. Patients were required to have relapsed myeloma and measurable disease (either serum M protein level of ≥5·0 g/l or urinary M protein of ≥0∙2 g/24 h) responsive to at least one previous therapeutic regimen (≥25% of reduction in M or total protein) and refractory to their most recent therapy (disease progression either during treatment or within 60 days after therapy completion). Patients were to be exposed to at least three prior treatments, including bortezomib, an immunomodulatory agent (lenalidomide and/or thalidomide), an alkylating agent, a corticosteroid and anthracycline (except for patients ineligible or clinically unsuitable for transplantation). Patients were excluded if they had Grade ≥3 or Grade 2 PN with pain or a past history of interstitial lung disease (ILD), congestive heart failure (CHF) of New York Heart Association class ≥III, symptomatic myocardial ischaemia or uncontrolled conduction abnormalities.

This study was conducted in compliance with the Good Clinical Practice guidelines. The study protocol was approved by the Institutional Review Board of each institution, and written informed consent was obtained from each patient enrolled in this study.

Dose‐limiting toxicity definition

DLT was defined as any of the following AEs in Cycle 1 that were at least possibly related to carfilzomib and met one of the following criteria: Grade 3 or 4 PN or Grade 2 PN with pain; Grade ≥3 non‐haematological toxicities; Grade ≥3 nausea, vomiting or diarrhoea that was uncontrolled after an adequate administration of anti‐emetic or anti‐diarrhoeal medications; febrile neutropenia; Grade 4 neutropenia persisting for >8 days without using granulocyte‐colony stimulating factor (G‐CSF) for supportive therapy and Grade 4 thrombocytopenia that required platelet transfusion or was accompanied by bleeding. Administration of G‐CSF was not permitted during the DLT evaluation period.

Treatment

Carfilzomib was intravenously administered for 10 min at doses of 15, 20 and 20/27 mg/m² on Days 1, 2, 8, 9, 15 and 16 of each 28‐day cycle until withdrawal of consent, disease progression or the occurrence of unacceptable toxic effects. For the 20/27 mg/m² dosage, 20 mg/m² was dosed on Days 1 and 2 of Cycle 1 and escalated to 27 mg/m² on Day 8 of Cycle 1 and thereafter.

Oral or intravenous dexamethasone (4 mg) was administered before each dose of carfilzomib in Cycle 1 and thereafter if necessary as pre‐medication to prevent infusion reactions. Intravenous and oral hydrations were also required during Cycle 1 and in subsequent cycles as needed. In Cycle 1, all patients were required to receive prophylactic antibiotics, and patients with a medical history of herpes infection received acyclovir.

Assessment of response and safety

The efficacy analysis set comprised all patients who received at least one dose of carfilzomib and had at least one assessment of efficacy or pharmacodynamics. The primary endpoint of phase 2 was the ORR based on central laboratory data according to the International Myeloma Working Group (IMWG) Uniform Response Criteria (Durie et al, 2006), including the minimal response (MR), as defined by the European Group for Blood and Marrow Transplantation criteria (Bladé et al, 1998). The ORR with 95% confidence interval (CI) was determined for each dose level. The investigational period ended when Cycle 6 was completed for all patients enrolled in the study, and subsequent cycles were included in the analysis for patients whose therapy lasted more than six cycles.

The safety analysis set comprised all patients who received at least one dose of carfilzomib, and all AEs were graded according to the National Cancer Institute Common Terminology Criteria for Adverse Events version 4∙0 (http://evs.nci.nih.gov/ftp1/CTCAE/CTCAE_4.03_2010-06-14_QuickReference_8.5x11.pdf).

Pharmacokinetics and pharmacodynamics

Samples for determining the plasma concentrations of carfilzomib were collected on Days 1 and 16 of Cycle 1 before administration, 5 min after the start of administration, immediately before the completion of administration and 5, 15 and 30 min and 1, 2 and 4 h after administration was completed. Moreover, whole blood and peripheral blood mononuclear cells (PBMCs) were collected on Days 1, 2 and 8 of Cycle 1 and Day 1 of Cycle 2 before administration as well as at 1 h after administration was completed to analyse proteasome activity.

Statistical analyses

All statistical analyses were performed using SAS^® version 9∙3 (SAS Institute Inc., Cary, NC, USA). Two‐sided 95% CI of the best ORR was determined according to Willson (1927) for evaluable patients whose best response was classified as stringent complete response (sCR), complete response (CR), very good partial response (VGPR) and partial response (PR).

The analysis of the ORR was performed in subgroups, defined by the patient baseline characteristics of age, sex, cytogenetics/fluorescence in situ hybridization (FISH) and International Staging System (ISS) for MM stage (Greipp et al, 2005) as exploratory analyses. Patients were classified as having standard‐risk or high‐risk cytogenetic abnormalities, as defined by IMWG criteria (Munshi et al, 2011). High‐risk cytogenetic markers included either del 13 or hypodiploidy by metaphase cytogenetic analysis and/or del 17p13, t(4;14), t(14;16) by interphase FISH. Hence, patients without these abnormalities were considered to be standard risk.

The ORR was estimated within each subgroup along with its 95% Wilson CI. The clinical benefit rate (CBR) is the percentage of patients whose best response was classified as CR, VGPR, PR and MR. The CBR was estimated along with its 95% Wilson CI. Analysis for time‐to‐event (PFS and OS) was performed by preparing Kaplan–Meier estimates of the median and plotting Kaplan–Meier curves. In addition, two‐sided 95% CIs for the medians were estimated.

Results

Patients and characteristics

Fifty patients were enrolled from 15 centres in Japan between August 2011 and January 2014; patient characteristics are shown in Table 1. The median age was 67 years (range, 48–81 years), and the median time from initial diagnosis to study entry was 4∙7 years (range 1∙6–12∙6 years). Most patients had either immunoglobulin G‐ (70%) or immunoglobulin A‐type (16%) myeloma, and 58% had ISS stage II or III at diagnosis. A significant proportion of patients (32%) had poor/unfavourable karyotypes, as determined by FISH analysis. The majority of patients (70%) had Grade 1 or 2 PN at baseline, and 40 of the 50 enrolled patients had past medical history of PN.

Table 1.

Patients' baseline characteristics

Characteristic	15 mg/m2		20 mg/m2		20/27 mg/m2		Total
	(n = 4)		(n = 6)		(n = 40)		(N = 50)
Sex, n (%)
Male	3	(75)	5	(83)	18	(45)	26	(52)
Median age, years (range)	67	(57–80)	71	(59–80)	66	(48–81)	67	(48–81)
ECOG PS, n (%)
0	3	(75)	3	(50)	23	(58)	29	(58)
1	1	(25)	3	(50)	17	(43)	21	(42)
ISS stagea, n (%)
I	2	(50)	2	(33)	12	(30)	16	(32)
II	1	(25)	3	(50)	14	(35)	18	(36)
III	1	(25)	1	(17)	9	(23)	11	(22)
Unknown	0	(0)	0	(0)	5	(13)	5	(10)
Ig subtype, n (%)
IgG	3	(75)	1	(17)	31	(78)	35	(70)
IgA	1	(25)	5	(83)	2	(5)	8	(16)
IgD	0	(0)	0	(0)	2	(5)	2	(4)
Bence–Jones	0	(0)	0	(0)	5	(13)	5	(10)
Cytogenetic/FISH prognostic markers, n (%)
Normal/Favourable	1	(25)	5	(83)	25	(63)	31	(62)
Unfavourableb	3	(75)	1	(17)	12	(30)	16	(32)
Unknown	0	(0)	0	(0)	3	(8)	3	(6)
Peripheral neuropathy, n (%)
Grade 0	0	(0)	1	(17)	14	(35)	15	(30)
Grade 1	2	(50)	4	(67)	20	(50)	26	(52)
Grade 2	2	(50)	1	(17)	6	(15)	9	(18)
Previous lines of therapy, median (range)	5	(4–7)	6	(3–8)	5	(3–10)	5	(3–10)
≥6, n (%)	2	(50)	3	(50)	16	(40)	21	(42)
Baseline anti‐hypertensive therapy, n (%)
Yes	3	(75)	1	(17)	14	(35)	18	(36)
No	1	(25)	5	(83)	26	(65)	32	(64)
Previous therapy, n (%)
Bortezomib	4	(100)	6	(100)	40	(100)	50	(100)
Immunomodulatory agent
Lenalidomide	4	(100)	6	(100)	33	(83)	43	(86)
Thalidomide	2	(50)	4	(67)	23	(58)	29	(58)
Corticosteroid	4	(100)	6	(100)	40	(100)	50	(100)
Alkylating agent	4	(100)	6	(100)	40	(100)	50	(100)
Anthracycline	2	(50)	4	(67)	26	(65)	32	(64)
Stem cell transplantation	1	(25)	2	(33)	17	(43)	20	(40)

ECOG PS, Eastern Cooperative Oncology Group performance status; FISH, fluorescence in situ hybridization; Ig, immunoglobulin; ISS, International Staging System.

^aAt diagnosis.

^bIncludes either t(4;14), t(14;16) or del (17p).

The patients had previously received a median of five (range 3–10) therapies, and 42% had previously received at least six therapies. All patients had received bortezomib and an immunomodulatory agent in previous regimens, and 48% had received at least two lines of bortezomib‐containing regimens. Twenty (40%) patients had undergone autologous stem cell transplantation (ASCT) (Table 1).

Dose escalation

Seventeen of the 50 patients enrolled in the study were enrolled in phase 1. Four patients, including one patient who was not evaluable for DLT, were enrolled in the 15 mg/m² cohort; no DLT was observed.

One of the first three patients experienced DLT (thrombotic microangiopathy, cardiomyopathy, hepatic disorder and sensorimotor disorder) in the 20 mg/m² cohort; therefore, an additional three patients were enrolled at this level. No further DLT was observed in the three patients; subsequently, no DLT was observed in a total of seven patients (including one who was not evaluable for DLT) enrolled in the 20/27 mg/m² cohort, thereby suggesting that a higher dose could reasonably be tested. However, 20/27 mg/m² was determined to be the recommended dose in phase 2 of this Japanese study at that time, considering the results in the previous carfilzomib studies conducted overseas (Siegel et al, 2012).

Efficacy

Fifty patients were included in the efficacy analysis set; the ORR was 20∙0% and the CBR was 28∙0% (Table 2). In the 40 patients who received the 20/27 mg/m² dose, the ORR was 22∙5% and the CBR was 27∙5%. Subgroup analysis of the 20/27 mg/m² group demonstrated that the ORR was not affected by age and ISS stage (Table 3). The comparison of the 20/27 mg/m² group in this study with that in the 003‐A1 study (Siegel et al, 2012) showed that the results were similar (22∙5% vs. 23∙7%) (Table 3). In the 20/27 mg/m² group, the median DOR was not reached (95% CI, 2∙3 months–not reached), and the median PFS was 5∙1 months (95% CI: 2∙8–7∙0 months), whereas the median OS was not reached (95% CI: 7∙4 months–not reached) at the time of the data cut‐off. The median follow‐up times for PFS and OS were 6∙0 months (95% CI: 5∙8–6∙7 months) and 6∙5 months [95% CI: 6∙0–7∙2 months], respectively.

Table 2.

Best overall response

	15 mg/m2		20 mg/m2		20/27 mg/m2		Total
	(n = 4)		(n = 6)		(n = 40)		(N = 50)
Best response, n (%)
CR	0	(0)	0	(0)	0	(0)	0	(0)
VGPR	0	(0)	0	(0)	2	(5∙0)	2	(4∙0)
PR	1	(25∙0)	0	(0)	7	(17∙5)	8	(16∙0)
MR	0	(0)	2	(33∙3)	2	(5∙0)	4	(8∙0)
SD	1	(25∙0)	1	(16∙7)	16	(40∙0)	18	(36∙0)
PD	0	(0)	1	(16∙7)	9	(22∙5)	10	(20∙0)
NE	2	(50∙0)	2	(33∙3)	4	(10∙0)	8	(16∙0)
ORR (≥PR), n (%)	1	(25∙0)	0	(0)	9	(22∙5)	10	(20∙0)
CBR (≥MR), n (%)	1	(25∙0)	2	(33∙3)	11	(27∙5)	14	(28∙0)
DOR, median (95% CI), months	9∙5	(NR–NR)	–	–	NR	(2∙3–NR)	9∙5	(2∙3–9∙5)
PFS, median (95% CI), months	2∙8	(1∙7–15∙9)	11∙1	(0∙9–11∙1)	5∙1	(2∙8–7∙0)	5∙1	(2∙8–7∙0)
OS, median (95% CI), months	17∙9	(3∙0–NR)	17∙8	(3∙4–23∙4)	NR	(7∙4–NR)	23∙4	(10∙3–NR)

CR, complete response; VGPR, very good partial response; PR, partial response; MR, minimal response; SD, stable disease; PD, progressive disease; NE, not evaluable; ORR, overall response rate; CBR, clinical benefit rate; DOR, duration of response; NR, not reached; PFS, progression‐free survival; CI, confidence interval; OS, overall survival; NR, not reached.

Table 3.

Comparison of overall response rate in the 20/27 mg/m² cohort according to baseline characteristics

Characteristic	ONO‐7057‐01a			PX‐171‐003‐A1b			Odds ratio	95% CI	P‐valuec
	n	ORR	95% CI	n	ORR	95% CI
Overall	40	22∙5	12∙3–37∙5	257	23∙7	18∙7–29∙4	0∙93	0∙42–2∙07	0∙8640
Age
<65 years	12	25∙0	8∙9–53∙2	139	25∙2	18∙2–33∙2	0∙99	0∙25–3∙87	0∙9890
≥65 years	28	21∙4	10∙2–39∙5	118	22∙0	14∙9–30∙6	0∙97	0∙35–2∙63	0∙9445
Sex
Female	22	13∙6	4∙7–33∙3	108	29∙6	21∙2–39∙2	0∙38	0∙10–1∙36	0∙1232
Male	18	33∙3	16∙3–56∙3	149	19∙5	13∙4–26∙7	2∙07	0∙72–5∙98	0∙1720
Cytogenetics/FISH prognostic markers
Normal/Favourable	25	28∙0	14∙3–47∙6	158	22∙8	16∙5–30∙1	1∙32	0∙51–3∙40	0∙5677
Unfavourable	12	16∙7	4∙7–44∙8	71	29∙6	19∙3–41∙6	0∙48	0∙10–2∙36	0∙3554
ISS stage
I	12	25∙0	8∙9–53∙2	76	31∙6	21∙4–43∙3	0∙72	0∙18–2∙91	0∙6461
II	14	7∙1	1∙3–31∙5	96	24∙0	15∙8–33∙7	0∙24	0∙03–1∙97	0∙1547
III	9	33∙3	12∙1–64∙6	78	17∙9	10∙2–28∙3	2∙29	0∙51–10∙26	0∙2704

ORR, overall response rate; CI, confidence interval; FISH, fluorescence in situ hybridization; ISS, International Staging System.

^aPresent study. For the 20/27 mg/m² cohort, 20 mg/m² was dosed on Days 1 and 2 of Cycle 1 and escalated to 27 mg/m² on Day 8 of Cycle 1 and thereafter.

^bSiegel et al (2012). The dose for Cycle 1 was 20 mg/m², which was escalated to 27 mg/m² on Day 1 of Cycle 2 and thereafter.

^cχ² test.

Safety

Fifty patients who received at least one dose of carfilzomib were included in the safety population. All patients experienced at least one AE, and 88% had at least one AE of Grade ≥3. All AEs encountered in ≥20% of the patients are shown in Table SI. The most commonly observed AEs were haematological toxicities, including lymphopenia (86%), thrombocytopenia (68%), anaemia (58%), neutropenia (56%) and leucopenia (50%). The most commonly observed AEs of Grade ≥3 were lymphopenia (68%), neutropenia (38%), anaemia (30%), thrombocytopenia (26%) and leucopenia (26%).

Although PN was observed in eight patients (16%), none were of Grade ≥3, and four of the eight patients had PN of Grade 1 or 2 at the baseline of the study. Moreover, all eight patients had a past history of PN before enrolment in the study (Table 4). According to the detailed analysis of the history of PN, 45 (90%) of 50 patients had experienced PN before they were enrolled in the study; however, PN in 10 of the 45 patients resolved before enrolment. Subsequently, four of the 10 patients encountered PN again after carfilzomib treatment (Patients 2, 3, 5 and 7 in Table 4). In total, eight (18%) of the 45 patients developed PN again after carfilzomib treatment and, of the 35 patients who had PN at baseline (Table 1), carfilzomib exacerbated PN in three patients (6% of 50 enrolled patients, 8∙6% of the 35 patients) (Patients 1, 4 and 6 in Table 4). In contrast, the remaining five patients who had never experienced PN before enrolment into the carfilzomib study did not develop PN after carfilzomib treatment. One patient who had Grade 2 PN of the lower extremities at baseline newly developed a trigeminal nerve disorder of Grade 1 during carfilzomib treatment, whereas the pre‐existing PN was not aggravated by carfilzomib (Table 4). No ILD was observed.

Table 4.

Patients with peripheral neuropathy under study treatment

Patient No.	Dose level (mg/m2)	Preferred terminology (Grade)	Baseline PN Grade	Worst PN Grade before the study
1	20	Peripheral neuropathy (2)	1	3
2	20/27	Peripheral neuropathy (1)	0	2
3	20/27	Peripheral sensory neuropathy (1)	0	2
4	20/27	Peripheral sensory neuropathy (2)	1	3
5	20/27	Peripheral sensory neuropathy (1)	0	2
6	20/27	Peripheral sensory neuropathy (2)	1	3
7	20/27	Peripheral sensory neuropathy (1)	0	1
8a	20/27	Trigeminal nerve disorder (1)	2	2

PN, peripheral neuropathy.

^aDeveloped trigeminal nerve disorder during carfilzomib treatment. Patient had Grade 2 PN of the lower extremities at baseline, which was not aggravated by carfilzomib.

We particularly highlighted cardiovascular and infectious AEs in this study, and the details are presented in Table 5. In total, the occurrence rate of hypertension (HT) was low and similar to that reported in the previous study (17%) (Grade ≥3; 6%) (Vij et al, 2012a). In the present study, HT (10/50 patients; 20%) was relatively common among the cardiovascular AEs. Although HT (8%) of Grade ≥3 and cardiomyopathy (2%) of Grade ≥3 were observed, severe CHF was not reported. The AEs considered to be autonomic are also shown in Table SII, although they were mild, except HT. Among AEs of any grade, HT was the most commonly noted; moreover, as observed in four patients, HT was the only Grade ≥3 AE that was attributed to autonomic neuropathy.

Table 5.

Adverse events related to cardiovascular disorders and infections of all grades or ≥Grade 3

Dose (mg/m2)	15 mg/m2		20 mg/m2		20/27 mg/m2		Total
Patients	(n = 4)		(n = 6)		(n = 40)		(N = 50)
Grade	All Grades	≥Grade 3	All Grades	≥Grade 3	All Grades	≥Grade 3	All Grades	≥Grade 3
Cardiovascular disorder
Hypertension	2	0	2	0	6	4	10	4
Congestive Heart failure	0	0	0	0	2	0	2	0
Vascular pain	0	0	0	0	2	0	2	0
Hot flush	0	0	0	0	2	0	2	0
Atrioventricular block first degree	0	0	0	0	1	0	1	0
Palpitations	0	0	0	0	1	0	1	0
Dyspnoeaa	0	0	0	0	1	0	1	0
Vasculitis	0	0	0	0	1	0	1	0
Troponin T increased	0	0	0	0	1	0	1	0
Orthostatic hypotension	1	0	0	0	0	0	1	0
Cardiomyopathy	0	0	1	1	0	0	1	1
Infections
Nasopharyngitis	0	0	1	0	10	0	11	0
Pharyngitis	1	0	0	0	5	0	6	0
Gingivitis	1	0	0	0	1	0	2	0
Pneumonia	0	0	1	1	1	0	2	1
Bronchopneumonia	0	0	0	0	1	1	1	1
Influenza	0	0	0	0	1	0	1	0
Viral pneumonia	0	0	0	0	1	1	1	1
Upper respiratory tract infection	0	0	0	0	1	0	1	0
Enteritis infection	0	0	0	0	1	0	1	0
Lip infection	0	0	0	0	1	0	1	0
Oropharyngeal candidiasis	0	0	1	0	0	0	1	0
Staphylococcal infection	0	0	1	1	0	0	1	1
Herpes virus infection	0	0	1	1	0	0	1	1

^aDyspnoea is classified as a respiratory adverse event according to National Cancer Institute Common Terminology Criteria for Adverse Events (NCI‐CTCAE) version 4∙0 (http://evs.nci.nih.gov/ftp1/CTCAE/CTCAE_4.03_2010-06-14_QuickReference_8.5x11.pdf).

With respect to infectious AEs, it is notable that nasopharyngitis and pharyngitis were relatively common, but the incidence of other infectious AEs was low, a finding similar to those for the upper respiratory tract documented previously in patients treated with bortezomib (Shah et al, 2004; Teh et al, 2014a,b) and carfilzomib (31–34%) (Vij et al, 2012a,b). Regarding AEs of Grade ≥3, pneumonia, bronchopneumonia, viral pneumonia, staphylococcal infection and herpes virus infection were observed in one patient each in the study (Table 5).

No AEs led to death during the administration period of carfilzomib or within 30 days after the final administration of carfilzomib. Eight patients (16%) discontinued treatment, and dosing was interrupted or reduced in 24 patients (48%) because of AEs. It is noteworthy that infection was the most frequent reason for the interruption of carfilzomib treatment. Of the 24 patients whose treatment was interrupted, 11 (46%) experienced viral or upper respiratory diseases, including one patient with flu, one with viral disease, eight with upper respiratory diseases and one with fever who was given an anti‐inflammatory drug commonly used as a medicine for cold. In addition, three patients were believed to have infections leading to treatment interruption; these three patients included one patient with pneumonia and two with fever (of three events), for which levofloxacin or acetaminophen was prescribed. The treatment was interrupted in five other patients who developed neutropenia (of seven events).

Our findings indicate that carfilzomib 20/27 mg/m² is feasible for Japanese patients with RRMM.

Pharmacokinetics and pharmacodynamics

PK analyses were performed in a total of 17 patients in phase 1. The PK parameters for carfilzomib are shown in Table SIII. The plasma carfilzomib concentration showed a rapid decrease after intravenous administration with terminal half‐lives (T_1/2) of 0∙424–0∙706 h. In the dose range of 15–27 mg/m², the area under the plasma concentration–time curve from time 0 to the time of last quantifiable concentration (AUC_last) and maximum plasma concentration (C_max) increased in a dose‐dependent manner. There was no trend toward increasing or decreasing the clearance and volume of distribution at steady state over the dose range. Following repeated doses of carfilzomib at 15 and 20 mg/m², the PK parameters were similar on Days 1 and 16. Although C_max was measured, the results were not compared with those of previous studies because the duration of intravenous administration of carfilzomib was 10 min in the present study and 2–10 min in the overseas studies (PX‐171‐007; Papadopoulos et al, 2013) Therefore, we concluded that there was no remarkable ethnic difference in the PK parameters of carfilzomib compared with the AUC_last and T_1/2 in PX‐171‐007.

For all dosing levels of carfilzomib, the proteasome activities in whole blood and PBMCs were reduced 1 h after administration on Days 1, 2 and 8 of Cycle 1 and on Day 1 of Cycle 2 with ≥80% inhibition. Furthermore, although proteasome activity in whole blood before the administration of carfilzomib on Day 1 of Cycle 2 slightly recovered (≥70% inhibition) because drug interruption was longer during this period than during other parts of the administration period, administration of carfilzomib resulted in a similar level of inhibition of proteasome activity. The level of inhibition was ≥80%, which was similar to that obtained in the overseas studies (Alsina et al, 2012).

Discussion

The present study aimed to evaluate the efficacy and safety of single‐agent carfilzomib in Japanese patients with RRMM. The dose level did not reach the MTD, but the recommended dose in phase 2 was determined to be 20/27 mg/m² on the basis of the results of phase 1. The results of single‐agent carfilzomib at a 20/27 mg/m² dose showed good responses in heavily pre‐treated patients, with an ORR of 22∙5% and a median PFS of 5∙1 months.

A comparison of the 20/27 mg/m² group in this study with that in the pivotal US study (003‐A1) (Siegel et al, 2012) showed that the results of the ORR were similar (22∙5% and 23∙7%, respectively), the median number of previous lines of therapy of the 20/27 mg/m² group of each study were the same and the proportion of patients with poor prognosis according to cytogenetic abnormalities was 30% in this study (Table 1) and 28% in 003‐A1. However, the median PFS was better in this study than in 003‐A1 (5∙1 months vs. 3∙7 months, respectively). The following factors account for the difference: 1) no patients with ECOG PS 2 were enrolled in this study, whereas 13% of the enrolled patients in 003‐A1 were PS 2, and 2) the median cumulative carfilzomib dose in this study was 796 mg/m² (range 80–1363 mg/m²), which was much higher than the 470 mg/m² (range 20–2647 mg/m²) in 003‐A1. On the other hand, ethnic differences in the efficacy of carfilzomib did not seem to be significant, and both studies demonstrated good efficacy in patients with RRMM.

The incidences of PN were similar: PN of any grade occurred in 73/526 (13∙9%) patients; Grade ≥3 PN occurred in seven (1∙3%) patients in the integrated analysis of four phase 2 studies of single‐agent carfilzomib (Siegel et al, 2013), whereas PN of any grade was encountered in eight (16∙0%) patients and no patient developed PN of Grade ≥3 in the present study. In addition, of the 35 patients who had PN at baseline, carfilzomib aggravated PN in three patients (8∙6%), which contrasts with the results of a previous study (Vij et al, 2012b). However, none of the patients who had not previously experienced PN developed new PN. There may be some patients who are prone to develop PN induced by PIs (Broyl et al, 2010; Corthals et al, 2011; Watanabe et al, 2013), and the choice of carfilzomib among PIs decreases the chance of encountering PN that hinders patients with MM from continuing to receive currently efficacious treatment or future treatment for RRMM.

Adverse effects of particular interest have been cardiac events, previously reported for single‐agent carfilzomib treatment (Siegel et al, 2012; 2013). Aggregated cardiac AEs, including arrhythmia, CHF, ischaemic heart disease and cardiomyopathy, have been reported in 116 of 526 patients (22∙1%), with 50 patients (9∙5%) being Grade ≥3, in the integrated analysis (Siegel et al, 2013). However, in the present study, regarding Grade ≥3 cardiac AEs, only one (2∙0%) patient in the 20 mg/m² cohort had cardiomyopathy, and no deaths occurred. Cardiotoxicities have been unexpectedly induced by PIs (Voortman & Giaccone, 2006; Orciuolo et al, 2007). In addition, unexplained deaths have been reported in the single‐agent bortezomib study, which were attributed to CHF and sudden death, although they were regarded as probably not related to bortezomib (Richardson et al, 2009).

The previous integrated analysis reported that HT was documented in 14∙3% and that more than half of those had a history of HT (Siegel et al, 2013). Although HT was more frequently recorded in our study (10 of 50 enrolled patients; 20%) than in the previous studies, four of the 10 patients were newly induced; however, in the remaining six patients who were prescribed hypertensive drugs before enrolment (Table SIV), HT was aggravated after carfilzomib treatment. As PIs have a potential to exacerbate impaired hypertensive states, blood pressure should be carefully monitored during the treatment, particularly in those who have a history of HT, and should be strictly controlled with additional anti‐hypertensive drugs during carfilzomib treatment. It is noteworthy that there was a case reported in which the female patient developed severe CHF after bortezomib treatment, for which HT was the sole cardiovascular risk factor (Bockorny et al, 2012); hence, we highlight this point. Moreover, intriguingly, in pressure‐overload hearts of mice, it has been shown that proteasome activities in cardiomyocytes were depressed, resulting from cardiomyocyte apoptosis through the accumulation of pro‐apoptotic proteins caused by impaired degradation, before the onset of cardiac dysfunction (Tsukamoto et al, 2006). Therefore, there is a great need for the pre‐control of HT to mitigate the risk of cardiac toxicity, including heart failure, and control of HT is likely to be an important component of the successful management of MM patients treated with PIs.

GI disorders are caused by bortezomib; however, the rate of GI disorders was 21% lower for the subcutaneous administration of bortezomib than for intravenous administration, of which the incidence of diarrhoea was 12% lower (Moreau et al, 2011). The GI disorders may be ascribed to autonomic neuropathy (Mele et al, 2015) because it is a well‐known fact that autonomic neuropathy is induced by bortezomib (Shah et al, 2004; Giannoccaro et al, 2011; Stratogianni et al, 2012; Mele et al, 2015); therefore, autonomic neuropathy may also account for PI‐induced HT.

Although 222 (42∙2%) and 67 (1∙7%) patients experienced dyspnoea and pneumonia of any grade, respectively, in the integrated analysis (Siegel et al, 2013), only one (2∙0%) and two (4∙0%) of the patients in the present study experienced dyspnoea and pneumonia of any grade, respectively (Table 5). The aetiology of dyspnoea remained unknown (Siegel et al, 2012) and has been debated (Siegel et al, 2013) because ILD was scarcely reported with regard to carfilzomib. Therefore, although dyspnoea was considered as pulmonary toxicity in the previous report (Siegel et al, 2013), it is more likely a symptom caused by pulmonary oedema owing to reversible acute left ventricular failure induced by PIs because it has been reported as a transient symptom that appeared on the day of or the day after carfilzomib dosing (Siegel et al, 2013). Consequently, in the present study, similar to a recent report (Sonneveld et al, 2015), we assumed dyspnoea to be a cardiovascular disorder (Table 5) so that they would not be underestimated.

If carfilzomib can replace bortezomib as the mainstay of triple combination therapy (Jakubowiak et al, 2012; Niesvizky et al, 2013; Wang et al, 2013; Bringhen et al, 2014; Sonneveld et al, 2015; Stewart et al, 2015), it may be necessary to optimize the use of carfilzomib, although carfilzomib and lenalidomide combined with dexamethasone has an extremely compelling efficacy and is well tolerated (Stewart et al, 2015). To explain this, illustrative results of phase 1 or 2 carfilzomib trials using combination therapies were as follows: 1) a total of 33% of patients required carfilzomib dose reduction and 20% discontinued treatment because of AEs in combination with cyclophosphamide and dexamethasone for patients with newly diagnosed MM (Bringhen et al, 2014), 2) notably, a total of 31% of the patients enrolled in a dose‐escalating study (up to 56 mg/m²) of carfilzomib experienced at least one Grade ≥3 dyspnoea when combined with 300 mg/m² of cyclophosphamide and low‐dose dexamethasone (Bensinger et al, 2014) and 3) furthermore, the rate of any cardiac‐related AEs increased up to 19% with 5% Grade 3 after consolidation therapy following autologous stem cell transplantation in a phase 2 study, in which combination of carfilzomib, thalidomide and dexamethasone was used (Sonneveld et al, 2015). In this study, notably, only 59% of the patients were able to complete the original treatment schedule without either delays, reductions, interruptions or premature stoppage of carfilzomib during the induction therapy. Furthermore, a slower (30 min) infusion of carfilzomib was better tolerated and permitted the administration of higher doses (20/45 mg/m² or 20/56 mg/m²) according to the dissociation constant (Kd); however, higher incidences of AEs were reported, including cough (40∙9%), dyspnoea (31∙8%) and HT (31∙8%), with 13∙6% of Grade ≥3as the most common non‐haematological Grade ≥3 AE (Badros et al, 2013).

Moreover, because our study showed that lymphopenia was the most common (Grade ≥3, 68%), haematological AE (Table SI), in addition to upper respiratory disease being the most common reason for interruption of the treatment, additional care should be taken should carfilzomib be introduced into combination therapy in the future, particularly with pomalidomide and dexamethasone.

Lessons have been learned regarding the optimal administration of PIs from experiences with the use of the first‐generation PI, bortezomib. Supportive care to avoid or prevent AEs induced by bortezomib and carfilzomib is important in continuous treatment with PIs (Delforge et al, 2010; Siegel, 2013). MM eventually develops resistance to all existing available therapies, and patients succumb to the disease (Kumar et al, 2012). Therefore, it is important to judiciously use PIs to reduce toxicities and to maintain the drug efficacy against currently existing MM in patients through a consecutive treatment of patients with MM through their life. To optimize the dose of carfilzomib, prescribing prophylactic drugs in advance for potential AEs in its earliest stages when toxicities are anticipated will be crucial for patients with MM to continue carfilzomib treatment and achieve more profound responses, which should prolong survival (Chanan‐Khan & Giralt, 2010; Gay et al, 2011; Martínez‐López et al, 2013).

In conclusion, in terms of safety, we did not find any clinically important ethnic differences in safety when carfilzomib was administered in a 20/27 mg/m² dosing regimen in Japanese MM patients. Although the PN rates with carfilzomib are low, it may worsen pre‐existing PN. Cardiotoxicities were the major concern in previous carfilzomib studies, but they were less frequently observed in the present study; hypertensive status seemed to be exacerbated by the administration of carfilzomib and bortezomib, an affect that may be caused by PI‐induced autonomic neuropathy. Therefore, we highlight the importance of managing AEs, including HT, by early treatment to alleviate PI‐induced AEs so that PI treatment can continue. With respect to efficacy, Japanese patients with RRMM achieved relatively longer PFS after higher total doses of carfilzomib than those administered in previous studies.

Author contributions

T. Watanabe performed the research and wrote this manuscript. Y. Shumiya and T. Kashihara designed the protocol and wrote the draft manuscript. K. Tobinai, M. Matsumoto, K. Suzuki, K. Sunami, T. Ishida, K. Ando, T. Chou, S. Ozaki, M. Taniwaki, N. Uike, H. Shibayama, K. Hatake, K. Izutsu, T. Ishikawa and S. Iida performed the research and contributed to the final version of the manuscript.

Conflict of interest

The authors declare the following: TW: personal fees from Celgene. K.K., Janssen Pharmaceutical K.K., Takeda Pharmaceutical Co., Ltd., Nippon Kayaku Co., Ltd., Eisai Co., Ltd., Nippon Shinyaku Co., Ltd., Daiichi Sankyo Co., Ltd., Zenyaku Kogyo Co., Ltd., outside the submitted work. KT: grants from Ono Pharmaceutical Co., Ltd., during the conduct of the study; grants and other funding from Eisai and Takeda; grants from Chugai Pharma, Kyowa Hakko Kirin, Celgene, Janssen Pharmaceuticals, GlaxoSmithKline, Mundipharma, SERVIER, Abbvie; other funding from Zenyaku Kogyo and Spectrum Pharmaceuticals, outside the submitted work. MM: personal fees from Celgene K.K., Janssen Pharmaceutical K.K. and Ono Pharmaceutical Co.,LTD, outside the submitted work. KS: No relevant financial relationship(s) to disclose. TI: grants and personal fees from TAKEDA, personal fees from CELGENE and JANSSEN, outside the submitted work. KA: No relevant financial relationship(s) to disclose. TC: Honoraria, lecture fee from Jansen Japan Pharmaceutical Co., Ltd., Celgene Japan Pharmaceutical Co., Ltd., BMS Japan Pharmaceutical Co., Ltd., Takeda Japan Pharmaceutical Co., Ltd., Chugai Japan Pharmaceutical Co., Ltd. SO: No relevant financial relationship(s) to disclose. MT: grants from Kyowa Hakko Kirin, Chugai Pharma, Janssen Pharma, Novartis, Bristol‐Myers Squibb, Celgene, Pfizer Inc, Takeda Pharma, Asahikasei Pharma and Dainippon Sumitomo Pharma, outside the submitted work. NU: No relevant financial relationship(s) to disclose. HS: grants from Ono Pharmaceutical Company, during the conduct of the study; grants and personal fees from Celgene K.K. and Takeda Pharmaceutical Co. Ltd., personal fees from Janssen Pharmaceutical Co. Ltd. and grants from Bristol‐Meyers Squibb Company, outside the submitted work. KH: No relevant financial relationship(s) to disclose. KI: grants from Ono Pharmaceutical Co. Ltd., during the conduct of the study; personal fees from Janssen Pharmaceutical K. K., Eisai Co., Ltd., Kyowa Hakko Kirin Co., Ltd., Takeda Pharmaceutical Co., Ltd., Genzyme Japan K. K., Celgene K. K., Shionogi & Co., Ltd., MSD K. K., Eli Lilly Japan K. K., Chugai Pharmaceutical Co., Ltd., Sumitomo Dainippon Pharma Co., Ltd., Nihon Servier Co. Ltd. and Zenyaku Kogyo Co., Ltd., grants and personal fees from Ono Pharmaceutical Co. Ltd., outside the submitted work. TI: No relevant financial relationship(s) to disclose. YS: Ono Pharmaceutical Co.,Ltd. employee (sponsor company). TK: Ono Pharmaceutical Co.,Ltd. employee (sponsor company). SI: grants from Ono Pharmaceutical Co. Ltd., during the conduct of the study; grants and personal fees from Celgene K.K., Ono Pharmaceutical Co. Ltd. and Chugai Pharmaceutical Co. Ltd., personal fees from Janssen Pharmaceutical Company, Kyowa Hakko Kirin Inc., Eli Lilly Japan K.K., Bristol‐Myers Squibb Company, Taiho Pharmaceutical Co., Ltd and Nippon Kayaku Co. Ltd., outside the submitted work.

Supporting information

Table SI. Adverse events (≥20%) of all grades or ≥Grade 3.

Table SII. Adverse events attributable to autonomic neuropathy.

Table SIII. Pharmacokinetic parameters.

Table SIV. Adverse event of hypertension.