Summary Basis of Decision for Mozobil ™

3.3.1 Pharmacodynamics

The effects of plerixafor on CD34+ mobilization was determined in healthy subjects and oncology patients.

In healthy subjects, the administration of a single dose of plerixafor (0.04 to 0.24 mg/kg) with no G-CSF led to dose-proportional increases from baseline of peripheral blood (PB) CD34+ cell counts, while the response to the 0.32 mg/kg dose was similar to that obtained with the 0.16 mg/kg dose. The pharmacodynamic response to plerixafor 0.24 mg/kg (no G-CSF) in healthy subjects occurred 6 to 10 hours after dosing. The median peak fold-increase was 15.8 over baseline.

Following 4 days of pre-treatment with G-CSF in healthy volunteers, the administration of plerixafor (0.24 mg/kg) in conjunction with G-CSF produced a sustained elevation in the PB CD34+ cell counts from 4 to 18 hours after plerixafor administration, with a peak response between 10 to 14 hours. The administration of a lower dose of plerixafor (0.16 mg/kg) in conjunction with G-CSF produced higher peak PB CD34+ cell counts compared to the treatment with either plerixafor (0.16 mg/kg) alone or G-CSF alone. These results showed a 3.8-fold increase, versus a 3.2-fold increase, versus a 1.2-fold increase, respectively, over the baseline response achieved on Day 4 with G-CSF alone.

In combination with G-CSF, patients with multiple myeloma (MM) in general, had higher responses than patients with non-Hodgkin’s lymphoma (NHL) in mobilizing CD34+ cells. In the NHL group, patients with higher baseline concentrations of PB CD34+ cell counts had better responses than those with lower baseline PB CD34+ cell counts. Plerixafor increased the PB CD34+ count by 3- to 6-fold over the pre-plerixafor dose level after the first injection, which was similar to the 3- to 4-fold increase observed in healthy subjects.

Cumulatively, pharmacodynamic studies showed that in healthy subjects, the plerixafor dose of 0.24 mg/kg elicited a higher and later peak response compared with the 0.16 mg/kg dose. The increase in PB CD34+ cells with plerixafor following 4 days of pre-treatment with G-CSF was higher than with plerixafor or G-CSF alone. When added to a dosing regimen of G-CSF in healthy subjects, 0.16 mg/kg and 0.24 mg/kg plerixafor had similar magnitudes of fold-increases in PB CD34+ cells. In patients with MM and NHL, the 0.24 mg/kg dose with G-CSF elicited a greater response (greater fold-increase in apheresis yields) than the 0.16 mg/kg dose with G-CSF. Based upon the above data and given the difference in response rates of patients with MM and NHL, the recommended dose of Mozobil (plerixafor injection) is 0.24 mg/kg body weight by SC injection.

3.3.2 Pharmacokinetics

Absorption

Plerixafor was rapidly absorbed following SC injection. Maximum plasma concentrations occurred in approximately 30-60 minutes.

Distribution

The apparent volume of distribution of plerixafor in humans is 0.3 L/kg, suggesting that plerixafor is largely confined to, but not limited to, the extravascular fluid space.

Plerixafor was moderately bound to human plasma proteins; up to 58%.

Metabolism

Plerixafor was not metabolized in vitro using human liver microsomes or human primary hepatocytes and did not exhibit inhibitory activity in vitro towards the major drug metabolizing CYP450 enzymes (1A2, 2A6, 2B6, 2C8, 2C9, 2C19, 2D6, 2E1, and 3A4/5). In the in vitro studies with human hepatocytes, plerixafor did not induce CYP1A2, CYP2B6, and CYP3A4 enzymes. These findings suggest that plerixafor has a low potential for involvement in P450-dependent drug-drug interactions.

Excretion

Renal excretion was the primary route of elimination. Following a 0.24 mg/kg dose in healthy volunteers with normal renal function, approximately 70% of the dose was excreted in the urine as the parent drug during the first 24 hours following administration. The median terminal half-life in plasma was 4.6 hours.

Special Populations and Conditions

Renal Impairment

Renal impairment did not affect the maximum plasma levels of plerixafor (C_max) but did increase exposure due to impaired clearance in patients with varying degrees of renal impairment.

A population pharmacokinetic analysis simulated the effect of creatinine clearance (CrCl), as determined by the Cockcroft-Gault formula, on the plasma clearance of plerixafor. These results support a dose reduction to 0.16 mg/kg in patients with a CrCl of 20-50 mL/min to reduce the increased exposure in these patients comparable to patients with a CrCl >50 mL/min receiving a 0.24 mg/kg dose of Mozobil. There are insufficient clinical data to make dosing recommendations for patients with a creatinine clearance <20 mL/min or for patients on dialysis. Tissue accumulation of plerixafor in patients with renal impairment was not studied.

Hepatic Impairment

Studies in patients with hepatic impairment have not been conducted. Plerixafor is not metabolized by the liver, and there is no recommendation for dose adjustment.

Paediatrics

The safety and efficacy of Mozobil in paediatric patients has not been established in controlled clinical studies; approximately 20 patients 8-18 years of age have been treated in compassionate use programs. As certain types of NHLs occur in children, a paediatric study is being conducted.

Pregnant Women

Use of Mozobil is not recommended during pregnancy due to its mechanism of action and results of reproductive toxicology studies. Because CXCR4 plays an essential role in foetal development and plerixafor is a selective antagonist of CXCR4, Mozobil is suggested to cause congenital malformations when administered during pregnancy. Studies in animals have shown teratogenicity. There are no adequate and well-controlled studies in pregnant women using Mozobil. Women of childbearing potential are advised to use effective contraception during treatment.

3.3.3 Clinical Efficacy

Data from a total of 345 adult oncology patients [165 NHL, 158 MM, and 22 Hodgkin’s lymphoma (HL)] treated with G-CSF in combination with Mozobil in two Phase II studies and two Phase III studies [Study 1 (NHL) and Study 2 (MM)] were used to demonstrate the efficacy of Mozobil.

The two pivotal Phase III studies were multi-centre, randomized, double-blind, parallel group, placebo-controlled studies. On the evening of Day 4 of daily morning doses of G-CSF 10 μg/kg, the first dose of assigned study treatment, either Mozobil 0.24 mg/kg or placebo was administered. On Day 5, patients received a morning dose of G-CSF 10 μg/kg and underwent apheresis approximately 10 to 11 hours after the first dose of study treatment (within 60 minutes after administration of G-CSF). Patients continued to receive an evening dose of study treatment followed the next day by a morning dose of G-CSF and apheresis for up to a maximum of 4 aphereses or until the target collection of CD34+ haematopoietic stem cells (HSCs) was achieved.

In both studies, patients who failed to collect ≥0.8 times (x) 10⁶ CD34+ cells/kg after 2 days of apheresis or at least 2 x 10⁶ CD34+ cells/kg in 4 or fewer days of apheresis had the option of entering an open-label rescue procedure. After a minimum 7-day rest period, they received another 4-day course of G-CSF followed by a course of Mozobil (0.24 mg/kg) in combination with G-CSF for repeat mobilization.

Following the last apheresis, patients underwent a rest period, then pre-transplant ablative chemotherapy followed by autologous transplantation within 5 weeks after the last apheresis. Transplantation was performed according to standard of care at each study centre.

Patients received G-CSF (5 μg/kg once daily) beginning on the fifth or sixth day after transplantation and continuing until the absolute neutrophil count (ANC) was ≥0.5 x 10⁹/L for 3 days or ≥1.0 x 10⁹/L for 1 day. Platelet (PLT) engraftment was defined as a PLT count ≥20 x 10⁹/L without transfusion for the preceding 7 days.

Graft durability was assessed at 100 days, 6 months, and 12 months post-transplantation.

The primary objective of Study 1 was to determine if NHL patients mobilized with G-CSF plus Mozobil 0.24 mg/kg were more likely to achieve a target number of ≥5 x 10⁶ CD34+ cells/kg in 4 or fewer days of apheresis than NHL patients mobilized with G-CSF plus placebo. The primary objective of Study 2 was to determine if MM patients mobilized with G-CSF plus Mozobil 0.24 mg/kg were more likely to achieve a target number of ≥6 x 10⁶ CD34+ cells/kg in 2 or fewer days of apheresis than MM patients mobilized with G-CSF plus placebo.

Secondary efficacy objectives common to both studies were to compare the two treatment arms with respect to the number of patients who achieved a minimum of 2 x 10⁶ CD34+ cells/kg (minimum number required for transplantation) in 4 or fewer days of apheresis, the number of days of apheresis required to reach target cell numbers, the time to engraftment of polymorphonuclear leukocytes (PMNs) and PLTs, and the durability of the graft at pre-specified times post-transplantation. A secondary objective unique to Study 2 was to compare the two treatment arms with respect to the number of MM patients who achieved the target number of cells in 4 or fewer apheresis days.

G-CSF plus Mozobil was a superior mobilizing regimen compared to G-CSF plus placebo in NHL and MM, as demonstrated in the two Phase III studies. There was achievement of statistically significant target HSC collections (primary efficacy endpoint) as follows: ≥5 x 10⁶ CD34+ cells/kg in 4 or fewer days of apheresis in NHL patients [59.3% versus (vs.) 19.6%, G-CSF plus Mozobil vs. G-CSF plus placebo, respectively, (p <0.001)]; and ≥6 x 10⁶ CD34+ cells/kg in 2 or fewer days of apheresis in MM patients [71.6% vs.  34.4%, G-CSF plus Mozobil vs. G-CSF plus placebo, respectively (p <0.001)]. An additional secondary efficacy endpoint in MM patients demonstrated a statistically significant achievement of target HSC collection in 4 or fewer days of apheresis [75.7% vs. 51.3%, G-CSF plus Mozobil vs. G-CSF plus placebo, respectively (p <0.001)].

Additional benefits were seen in the results of the secondary efficacy endpoints. There was a statistically significant achievement of the minimum HSC required for autologous transplantation (≥2 x 10⁶ CD34+ cells/kg) in 4 or fewer days of apheresis in both NHL patients [86.7% vs. 47.3%, G-CSF plus Mozobil vs. G-CSF plus placebo, respectively (p <0.001)] and MM patients [95.3% vs. 88.3%, G-CSF plus Mozobil vs. G-CSF plus placebo, respectively (p = 0.031)]. Target HSC collections were all achieved in fewer days with G-CSF plus Mozobil compared with G-CSF plus placebo (median number of days to reach target HSC collection was 3 days for the plerixafor group and not evaluable for the placebo group in Study 1 and 1 day for the plerixafor group compared with 4 days for the placebo group in Study 2). For transplanted patients, time to neutrophil engraftment (10-11 days) and platelet engraftment (18-20 days) were similar across the treatment groups. Based on an adjusted analysis which used laboratory measurements and clinical criteria to assess graft durability, results were similar in both treatment groups. For transplanted patients, the frequency of graft failure was low in the Phase III studies; three events in Mozobil-treated NHL patients and none in MM patients. None of these graft failures were considered by the investigator as related to Mozobil.

Some patients with NHL and MM met minimal and target HSC collection criteria with G-CSF alone, thus avoiding any additional toxicity that might result from the addition of Mozobil. In Study 1 (NHL), 47.3% of G-CSF plus placebo patients achieved minimum CD34+ stem cell count in 4 or fewer days of apheresis and 19.6% of G-CSF plus placebo patients achieved target CD34+ stem cell count in 4 or fewer days of apheresis, while in Study 2 (MM), 88.3% of G-CSF plus placebo patients achieved minimum CD34+ stem cell count in 4 or fewer days of apheresis and 55.9% of G-CSF plus placebo patients achieved target CD34+ stem cell count in 4 or fewer days of apheresis.

Overall survival (OS) or other clinical benefit endpoints were not used to demonstrate efficacy in the Phase III studies. Information required for progression- or relapse-free survival analyses was not captured. As an exploratory endpoint, OS data were collected and summarized for patients who went to transplantation. The OS data at 12 months post-transplantation were similar in both treatment groups in both pivotal studies.

Patients with NHL received considerable doses of prior cytotoxic chemotherapy which may have negatively impacted HSC mobilisation and collection. In Study 1, 52 NHL patients in the G-CSF plus placebo group entered into the rescue procedure with Mozobil and G-CSF. Of these patients, 60% (31 out of 52) mobilized ≥2 x10⁶/kg CD34+ cells and had successful engraftment. In Study 2, 7 MM patients in the G-CSF plus placebo group entered the rescue procedure, all of whom mobilized ≥2 x10⁶/kg CD34+ cells and had successful engraftment.

There were insufficient data to support use of G-CSF plus Mozobil with rituximab in NHL or use of G-CSF plus Mozobil in the indication of tandem transplantation in MM.

Patients with HL were not included in pivotal Study 1. Efficacy results from a supportive uncontrolled open-label Phase II study are promising but limited and did not support inclusion of HL in the indication.

Mozobil was not studied in controlled clinical studies in patients with severe renal impairment or in patients on dialysis. A total of 60 patients with an estimated creatinine clearance (CrCl) 51-80 mL/min, 11 patients with CrCl 31-50 mL/min, and none with CrCl ≤30 mL/min were enrolled in the pivotal studies and received at least one dose of Mozobil (0.24 mg/kg body weight subcutaneously).

Mozobil dose and treatment of patients weighing more than 175% of ideal body weight was not investigated. Based on increasing exposure with increasing body weight, the Mozobil dose should not exceed 40 mg/day for patients with a CrCl >50 mL/min and 27 mg/day for patients with a CrCl 20-50 mL/min. The Sponsor has initiated a fixed dose (20 mg) vs. weight based dose study in lower weight NHL patients to investigate the effects of weight on pharmacokinetics and pharmacodynamics. This study has been restricted to patients weighing 70 kg or less.

3.3.4 Clinical Safety

The clinical safety evaluation of Mozobil was based on safety data obtained from 25 clinical studies including a thorough QT/QTc study, as well as the North American Compassionate Use Programme. Data from over 1,400 patients who received at least 1 dose of Mozobil in the clinical development programme were used to establish the safety profile. Safety data for G-CSF plus Mozobil were obtained from two Phase III studies (301 patients) and 10 uncontrolled studies (242 patients). Patients were primarily treated with Mozobil at daily doses of 0.24 mg/kg by SC injection. Median exposure to Mozobil in these studies was 2 days (range 1 to 7 days).

The adverse reactions reported in oncology patients who received Mozobil in the controlled Phase III studies and uncontrolled studies, including a Phase II study of Mozobil as monotherapy for HSC mobilization, were similar. No notable differences in the incidence of adverse reactions were observed for oncology patients by disease, age, or sex.

In both Phase III studies, the majority of adverse events (AEs) were reported during mobilization and apheresis and were mild to moderate in severity. The AEs that occurred during mobilization and apheresis in ≥5% of the pooled Phase III results from patients who received Mozobil, regardless of causality and more frequent with Mozobil than placebo, were as follows: diarrhea; nausea; vomiting; flatulence; injection site reactions; fatigue; arthralgia; headache; dizziness; and insomnia. The incidence of anxiety during HSC mobilization and apheresis was 5.3% vs. 4.5%, Mozobil vs. placebo, respectively. Paresthesia was considered an AE related to study treatment in 7.0% of patients in the Mozobil group and 5.1% of patients in the placebo group.

Hypokalaemia and hypomagnesaemia were reported as treatment-related AEs more frequently with Mozobil than with placebo during mobilization and apheresis in the pooled Phase III data. The administration of G-CSF plus Mozobil caused leukocytosis. Thrombocytopaenia was also observed in patients receiving Mozobil.

The effect of Mozobil on spleen size in patients has not been specifically evaluated in clinical studies. The possibility that Mozobil in conjunction with G-CSF can cause splenic enlargement cannot be excluded.

The AE reporting beyond mobilization and apheresis was limited by protocol. The majority of serious adverse events (SAEs) occurred after the mobilization and apheresis period, and were considered unrelated to study treatment. In the pooled Phase III data, the incidence of bactaeremia, lung infections and febrile neutropaenia was higher in the Mozobil group than in the placebo group; however, the majority of these events followed myeloablative chemotherapy and most were considered by the investigator to be unrelated to Mozobil administration. The incidence, cause, and timing of deaths, as well as the incidence of study or treatment discontinuations due to AEs were similar in both treatment groups (Mozobil and placebo). The majority of deaths occurred post-engraftment.

In the Phase III studies, examination of all clinical cardiovascular adverse events did not identify any rhythm-related cardiac safety signals attributable to Mozobil treatment in the populations studied. However, in a randomized, double-blind, placebo-controlled crossover study in healthy subjects, Mozobil was associated with an asymptomatic shortening of the PR interval and therefore, caution is recommended in patients with pre-excitation syndromes. Additional cardiovascular events, such as decreases in blood pressure, vasovagal reactions, and myocardial infarctions were also included in the Warnings and Precautions section of the Product Monograph. An increased incidence of hypotension occurred during mobilization and apheresis in the Phase III studies for patients in the G-CSF plus Mozobil group compared to the G-CSF plus placebo group. In the Mozobil oncology and non-oncology clinical studies, 0.8% of the subjects experienced vasovagal reactions (the majority within 1 hour of Mozobil administration). In the clinical studies, 0.9% of oncology patients in the G-CSF plus Mozobil group experienced myocardial infarctions (MIs) after mobilization as compared to 0.3% of oncology patients in the G-CSF plus placebo group after mobilization. All MIs occurred at least 14 days after the last Mozobil administration and in patients with known cardiac risk factors for MI and/or previous exposure to cardiotoxic chemotherapy.

In the Mozobil oncology clinical studies, 5/752 (0.7%) of patients experienced mild or moderate systemic reactions within approximately 30 minutes after Mozobil administration. Events included one or more of the following: urticaria [(n) = 2]; periorbital swelling (n = 2); dyspnoea (n = 1); or hypoxia (n = 1). Symptoms generally responded to treatments or resolved spontaneously.

In the Phase III studies, the incidence of psychiatric disorders during mobilization and apheresis was 14.8% in the G-CSF plus Mozobil group compared to 10.2% in the G-CSF plus placebo group. Insomnia and anxiety were the most common events.

G-CSF mobilization is recognized to have the potential to mobilize tumour cells from the marrow; however, the effect of reinfusion of tumour cells with the apheresis product has not been well-studied and only limited data are available. It is also known that chemotherapy, with or without growth factors, results in the mobilization of tumour cells in patients with haematologic malignancies resulting in tumour cell contamination of the graft. When Mozobil and G-CSF have been administered to patients with acute myelogenous leukemia and plasma cell leukemia, some patients experienced an increase in the number of circulating leukemia cells. Therefore, Mozobil should not be used for HSC mobilization and harvest in patients with leukemia.

There exists a potential for tumour cell mobilization in patients with MM or NHL receiving treatment with Mozobil for the purpose of stem cell mobilization. Furthermore, Mozobil and G-CSF mobilize HSC via different mechanisms; therefore, they might theoretically result in a broader potential for mobilization of tumour cells when used in combination. Tumor cell mobilization by Mozobil has not been thoroughly studied. Nonclinical studies are lacking. Based on laboratory investigations conducted in clinical studies of patients with NHL and MM, an increase in mobilization of tumour cells above that which occurs with G-CSF mobilization alone has not been observed with Mozobil, although limitations in assays utilized and patient numbers examined are important to acknowledge. The severity and nature of the risk associated with potential enhancement of tumour cell mobilization with Mozobil in combination with G-CSF when used for stem cell mobilization in patients with MM or NHL remains uncertain at present. However, by excluding patients with plasma cell and other leukemias from Mozobil stem cell mobilization, the risk and safety concern is considered to be acceptable.