First FDA Approval of Dual Anti-HER2 Regimen: Pertuzumab in Combination with Trastuzumab and Docetaxel for HER2-Positive Metastatic Breast Cancer

Metastatic breast cancer (MBC) is a serious and life-threatening condition, causing 39,250 deaths in the United States in 2011 (1). The HER2/neu receptor is overexpressed in 15% to 30% of MBC and is associated with a poor prognosis and an aggressive phenotype. With the incorporation of trastuzumab-based therapy for the treatment of early breast cancer (2–4) and for MBC (5), the prognosis of HER2-positive breast cancer has improved. However, new therapies are needed to prolong life, substantially delay disease progression, and alleviate cancer-related symptoms.

Pertuzumab is administered in conjunction with trastuzumab, another IgG1κ humanized antibody. Both antibodies target the HER2 receptor extracellular domain (ECD). Trastuzumab binds to subdomain IV of the HER2 ECD, whereas pertuzumab binds to subdomain II (6). HER2 is one of 4 members of the human EGFR family, which also includes EGFR (HER1), HER3, and HER4. HER signaling is known to contribute to neoplastic cell growth, malignant transformation, and resistance to chemotherapy (7, 8). As a proposed primary mechanism of action, pertuzumab blocks heterodimerization of HER2 with other members of the family and diminishes resultant ligand-activated signaling. In contrast, trastuzumab disrupts ligand-independent HER2 activity and associated downstream signaling. As would be expected of IgG1κ antibodies, both pertuzumab and trastuzumab can also mediate antibody dependent cell-mediated cytotoxicity (6, 7).

Before approval of the pertuzumab biologic license application (BLA), there were 2 drugs available in the United States for the treatment of HER2-positive MBC (Table 1). Trastuzumab was approved for first-line MBC in 1998 in combination with paclitaxel based on a time-to-progression (TTP) endpoint, and later showed improved overall survival (OS). Trastuzumab was also granted a refractory MBC indication based on objective response rates (ORR) in a single-arm trial. In 2007, the U.S. Food and Drug Administration (FDA) approved lapatinib (Tykerb, GSK), a tyrosine kinase inhibitor, in combination with capecitabine for refractory HER2-positive MBC based on TTP.

Pertuzumab is produced using recombinant DNA technology in a Chinese hamster ovary cell line. During the preapproval inspection of the drug substance manufacturing facility, it became apparent that a high failure rate of the working cell bank (WCB) thaw and subsequent propagation of the cells used to manufacture pertuzumab was being experienced in the ongoing Q1/Q2 2012 manufacturing campaign. This indicated that the process was not in a state of control during this manufacturing campaign and was suggestive of WCB instability. The failure rates observed were inconsistent with previous pertuzumab manufacturing experience at the same facility in 2010 and with other antibodies manufactured at the facility. On the basis of these growth issues, the major concern was whether the applicant could consistently manufacture pertuzumab with product quality characteristics comparable with that used in their clinical trials (9).

An investigation into the cell growth issues was ongoing at the time of the FDA inspection and frequent teleconferences were held with the applicant subsequent to the inspection. Following discussions with FDA, the Applicant initiated 3 concurrent plans to address these concerns: (i) confirmation of stability and manufacturing of pertuzumab from the master cell bank; (ii) development of a new WCB; and (iii) extended characterization of drug substance produced in the Q1/Q2 campaign of 2012 to evaluate product quality (9).

In view of the cell growth failures observed upon thaw and subsequent propagation of the WCB, FDA determined that the applicant did not, in 2012, have a consistent process that would ensure continued supply of commercial material. However, qualification lots of pertuzumab were manufactured in 2010 by a consistent process and the applicant estimated that adequate supply of these lots existed for patients. Lots from the 2010 campaign met all applicable requirements with respect to product safety, purity, and potency and there were no GMP or process consistency issues associated with these lots. Therefore, only drug product from the 2010 manufacturing campaign was approved for marketing. To address issues about future supply, the FDA approval letter outlined postmarketing commitments to ensure a consistent pertuzumab manufacturing process. Some of the postmarketing commitments were to be conducted under third-party oversight. The marketing of pertuzumab generated in campaigns other than the 2010 campaign is subject to successful fulfillment of these postmarketing commitments (9).

Nonclinical studies in support of the BLA application investigated the pharmacology, pharmacokinetics, and toxicology of pertuzumab. The antitumor activity of pertuzumab was studied in multiple tumor xenograft models, including one model that is resistant to trastuzumab; tumor growth in this model was inhibited by pertuzumab, but not by trastuzumab treatment. In vitro and in vivo studies showed that although pertuzumab and trastuzumab as single agents could significantly inhibit the growth of HER2-overexpressing tumor cells, the combination of both agents had greater antitumor activity when compared with the single-agent effects.

The Cynomolgus monkey was the appropriate model for nonclinical evaluation as the amino acid homology of human and monkey ErbB2 is 99% and binding affinities for pertuzumab were comparable. The toxicologic profile of pertuzumab in monkeys suggests that it is well tolerated, with toxicities limited to diarrhea. Despite the clinical evidence of cardiovascular toxicity observed with trastuzumab treatment, safety pharmacology and general toxicology testing did not reveal any remarkable cardiovascular findings in cynomolgus monkeys treated with pertuzumab up to 6 months.

Administration of pertuzumab by intravenous injection to monkeys during gestation was associated with fetal lethality and embryo-fetal effects including oligohydramnios, decreased relative lung and kidney weights, and microscopic evidence of renal hypoplasia consistent with delayed renal development. Various external, visceral, and skeletal abnormalities were also observed and were generally considered to be secondary to intrauterine restriction resulting from the oligohydramnios. Ratios of fetal to maternal pertuzumab exposure levels were comparable, indicating that pertuzumab did cross the placenta, resulting in significant fetal exposure. Thus, administration of pertuzumab during pregnancy is expected to pose a risk to the human fetus. Because this risk is likely outweighed by the potential for clinical benefit, pertuzumab was designated pregnancy category D.

Pertuzumab showed linear pharmacokinetics at a dose range of 2 to 25 mg/kg. On the basis of a population pharmacokinetic analysis that included 481 patients, the median clearance of pertuzumab was 0.24 L per day and the median half-life was 18 days. With the approved intravenous dosing regimen (an initial dose of 840 mg followed by a maintenance dose of 420 mg every 3 weeks thereafter), the steady-state concentration of pertuzumab was reached after the first maintenance dose. Population pharmacokinetic analysis suggested no pharmacokinetic differences based on age or gender. No pharmacokinetic differences were observed between Japanese and non–Japanese populations. No dose adjustments on the basis body weight or baseline albumin level are needed.

No drug–drug interactions were observed between pertuzumab and trastuzumab or between pertuzumab and docetaxel. Dose adjustments of pertuzumab are not needed in patients with mild or moderate renal impairment. No large changes in the mean QT interval (i.e., > 20 ms) were detected for the approved dosing regimen.

Antitherapeutic antibodies (ATA) to pertuzumab were detected in patients, but no known association with hypersensitivity reactions or anaphylaxis was determined, and the benefit of pertuzumab treatment seemed to be preserved within both ATA-positive and ATA-negative subgroups.

The BLA was supported by results from a randomized, phase III, multinational, multicenter, double-blind, and placebo-controlled study (Cleopatra/WO20698; ref. 10). Patients with MBC were required to have HER2-positive disease, defined as 3⁺ by immunohistochemistry or ≥2.0 amplification by FISH, determined at a central laboratory. Those patients who received previous adjuvant or neoadjuvant therapy were required to have a disease-free interval of more than 12 months, before trial enrollment. Patients were randomized 1:1 to receive either pertuzumab by intravenous infusion (840 mg initial dose, 420 mg every 3 weeks thereafter) or matched placebo as an add-on to the standard-of-care: trastuzumab (8 mg/kg initial i.v. dose, 6 mg/kg i.v. every 3 weeks thereafter) and docetaxel (75 mg/m² i.v. every 3 weeks for at least 6 cycles, with the option of dose escalation to 100 mg/m²).

In addition to the phase III trial, 2 supportive clinical trials were submitted and reviewed: Neosphere/TOC 4,129 g (a four-arm randomized phase II trial in neoadjuvant early breast cancer; ref. 11) and BO17929 (a single-arm phase II study in refractory disease; ref. 12).

The primary endpoint for the phase III trial was progression-free survival (PFS) based on tumor assessments by an independent review facility (IRF). The study featured a "real-time" PFS analysis by the IRF. If the investigator determined progressive disease (PD), but the IRF disagreed, patients would continue with every 9-week tumor assessments until the IRF determined PD. In addition, once PD was determined, the blind was maintained without cross-over. This design reduced informative censoring and confounding variables in the PFS and OS analyses. Secondary endpoints in the phase III trial included OS, PFS based on investigator assessment, ORR, and safety.

The study was designed to have 80% power to detect a 33% improvement in median PFS (14 months on the pertuzumab-treatment arm and 10.5 months on the placebo-treatment arm, corresponding to a HR of 0.75) at a two-sided α of 0.05. The planned sample size was 800, with 381 planned PFS events. For OS, the study was designed to have 80% power to detect a 33% increase in median survival (48 months on the pertuzumab-treatment arm and 36 months on the placebo-treatment arm, corresponding to a HR of 0.75) at a two-sided α of 0.05. One interim analysis for OS was planned at the time of final PFS analysis with Lan–Demets α spending based on the O'Brien-Fleming boundary. PFS and OS were compared between treatment arms by log-rank test stratified by prior treatment status and region.

In the phase III trial, 808 patients were randomized, 402 to the pertuzumab-treatment arm and 406 to the placebo-treatment arm. Overall baseline demographic and tumor characteristics were balanced between treatment arms. The median age was 54, 59% of patients were white, 32% of patients were Asian, and 78% had visceral metastases. Approximately 47% of patients previously received adjuvant or neoadjuvant therapy, including a small number of patients (11%) who received trastuzumab previously. Forty-eight percent of patients were hormone receptor-positive, of whom roughly half (52%) received adjuvant hormonal therapy.

In May 2011, the prespecified number of PFS events occurred, showing a statistically significant prolongation of IRF-PFS in the pertuzumab arm [HR, 0.62; 95% confidence interval (CI), 0.51–0.75; P < 0.0001] as shown in Fig. 1. The median PFS was 18.5 months in the pertuzumab arm and 12.4 months in the placebo arm.

At the final PFS analysis, a planned interim OS analysis was conducted, comprising 43% of events planned at the final OS analysis. This interim analysis indicated an OS advantage for the patients allocated to pertuzumab (HR, 0.64; 95% CI, 0.47–0.88; P = 0.0053). However, the HR and P value for the interim analysis of OS did not meet the predefined stopping boundary.

During the pertuzumab review, FDA evaluated efficacy in various subgroups of interest. Importantly, in the subset of patients who received prior trastuzumab in the adjuvant setting (which best reflects the U.S. population in which most patients are diagnosed with early breast cancer), the PFS benefit of pertuzumab was maintained (HR, 0.62; 95% CI, 0.35–1.07). The pertuzumab PFS benefit in patients with hormone receptor-positive MBC (HR, 0.72; 95% CI, 0.55–0.95) and in patients with disease confined to nonvisceral metastasis (HR, 0.96; 95% CI, 0.61–1.52) seemed to be less compared with the benefit observed in the overall population.

In the phase III trial, patients were to be treated with docetaxel in combination with trastuzumab and either pertuzumab or placebo for a minimum of 6 cycles, in the absence of unacceptable toxicity or disease progression. Trastuzumab and pertuzumab/placebo were to be continued until progression.

The safety population consists of 397 patients on the placebo-treatment arm and 407 on the pertuzumab-treatment arm. The median duration of study treatment was 18.1 months in the pertuzumab-treated group and 11.8 months in the placebo-treated group.

Table 2 summarizes the safety overview. Adverse events more than 5% more common in the pertuzumab-treated group included mucosal inflammation, rash, dry skin, diarrhea, and febrile neutropenia. Grade 3 to 4 adverse events more than 2% more common in the pertuzumab-treatment group included diarrhea, neutropenia, and febrile neutropenia. Among Asian patients, the incidence of febrile neutropenia was particularly high in the pertuzumab-treated group (26%) compared with the placebo-treated group (12%). These differences were not explained by differences in supportive care between regions or higher doses of or higher exposure to docetaxel. There were no differences in rates of left ventricular systolic dysfunction between treatment arms.

The pertuzumab BLA marks the first time FDA approved a molecularly targeted monoclonal antibody to be used in combination with another monoclonal antibody. In its decision, the agency considered the large magnitude of PFS improvement (6.1 months difference in medians), the consistency of findings across relevant subgroups, the interim OS analysis favoring the pertuzumab arm (which on a subsequent analysis showed statistical significance; ref. 13), and the supportive evidence of high antitumor response (80% ORR) in the phase III trial as well as in the 2 supportive phase II trials.

An exploratory analysis of subgroups in the phase III trial revealed less efficacy in the hormone receptor-positive subgroup compared with the overall population. This was consistent with results in the neoadjuvant randomized phase II trial (11), in which rates of pathologic complete response were lower in patients with hormone receptor-positive disease as compared with patients with hormone receptor-negative disease. Therefore, the applicant agreed as a postmarketing commitment to submit the results of a randomized phase II study (14), testing whether the addition of pertuzumab to the combination of trastuzumab and an aromatase inhibitor can enhance antitumor activity in HER2-positive, hormone receptor-positive MBC.

The safety of pertuzumab in combination with trastuzumab and docetaxel was acceptable, with no evidence of additive cardiac toxicity in the phase III trial. This was in contrast with findings from early clinical development, which suggested that pertuzumab might increase the cardiotoxicity of trastuzumab (15). Trials in neoadjuvant- and adjuvant-breast cancer will further clarify the cardiac risk in a treatment-naive, early breast cancer population.

The cell growth problems associated with pertuzumab manufacturing identified during the 2012 preapproval inspection of the drug substance manufacturing facility proved to be a critical issue during BLA review. FDA requires demonstration of a consistent manufacturing process before approval of a BLA. In this case, FDA took the unusual step of approving only product containing drug substance manufactured during the applicant's 2010 manufacturing campaign, which was produced using a consistent process and of which an adequate supply existed. Marketing of product produced from future manufacturing campaigns, including that produced during the Q1/Q2 2012 campaign, would be subject to further approval.

It was FDA's determination that a compelling exigent public health need outweighed the risk of future interruption in the drug supply to patients. The justification for approving the pertuzumab application to allow only release of the 2010 lots was based on: (i) FDA's determination that product from the 2010 campaign met all applicable requirements with respect to safety, purity, and potency; (ii) the cell expansion problems of the Q1/Q2 2012 campaign were not observed with the 2010 campaign; (iii) a consistent and controlled process was in place at the time of the 2010 manufacturing campaign; (iv) the Applicant's commitment to undertake several steps to resolve expeditiously the cell expansion problem; and (v) the Applicant's commitment to reduce and mitigate the risk of a drug shortage. At the time of approval, the manufacturer agreed to several postmarketing commitments that would be conducted under third-party oversight to address the process consistency concerns and allow approval of a viable commercial process.

In summary, the approval of the pertuzumab BLA marked the licensing of the first dual anti-HER2 regimen for treatment of breast cancer in the United States. This represents another significant advance in the treatment of HER2-positive breast cancer, which began with the approval of trastuzumab for MBC in 1998. The safety and efficacy of pertuzumab in early breast cancer is currently being tested in a large phase III adjuvant trial (16).

No potential conflicts of interest were disclosed.

Conception and design: G.M. Blumenthal, P. Cortazar, R. Pazdur

Development of methodology: G.M. Blumenthal, P. Cortazar

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): G.M. Blumenthal

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): G.M. Blumenthal, N.S. Scher, P. Cortazar, S. Chattopadhyay, S. Tang, P. Song, Q. Liu, K. Ringgold, A.M. Pilaro, K.E. King, L. Graham, B.L. Rellahan, W.C. Weinberg, B. Chi, C. Thomas, P. Hughes, R. Pazdur

Writing, review, and/or revision of the manuscript: G.M. Blumenthal, N.S. Scher, P. Cortazar, S. Chattopadhyay, S. Tang, P. Song, Q. Liu, K. Ringgold, A. Tilley, K.E. King, L. Graham, B.L. Rellahan, W.C. Weinberg, B. Chi, C. Thomas, P. Hughes, R. Pazdur

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): G.M. Blumenthal, P. Cortazar, R. Pazdur