Efficacy, Safety, and Immunogenicity of HLX02 Compared with Reference Trastuzumab in Patients with Recurrent or Metastatic HER2-Positive Breast Cancer: A Randomized Phase III Equivalence Trial

This randomized, multicenter, double-blind phase III equivalence study was designed to compare the efficacy and safety of HLX02 with reference trastuzumab in adult patients with HER2-positive recurrent or metastatic breast cancer. Patients were recruited from 89 centers in China, the Philippines, Poland, and Ukraine (Table 1 in the Electronic Supplementary Material [ESM]).

Eligible patients were aged ≥ 18 years, had histologically or cytologically confirmed breast adenocarcinoma, and an Eastern Cooperative Oncology Group (ECOG) performance status score of 0–1. Other key inclusion criteria were HER2 positivity (defined as fluorescence in situ hybridization amplification ratio ≥ 2 or immunohistochemistry score 3+), known estrogen-receptor (ER) and progesterone-receptor (PgR) status at study entry, measurable disease assessed by central imaging review (CIR), normal (within institutional range of normal) left ventricular ejection fraction (LVEF), and adequate hematologic, hepatic, and renal function. Key exclusion criteria were previously or on-treated (with systemic chemotherapy, biological, or targeted agent, or any other anticancer agent except hormonal therapy) metastatic breast cancer, symptomatic or untreated brain metastasis or any other central nervous system metastases, uncontrolled systemic disease that in the investigator’s opinion made the administration of study drug hazardous, prior exposure to doxorubicin (> 360 mg/m² or equivalent), and residual nonhematologic grade 2 or higher toxicity from prior therapies. Full inclusion and exclusion criteria are listed in Table 2 in the ESM.

The study protocol was reviewed and approved by the relevant independent ethics review board at each study site. All patients provided written informed consent before inclusion. This trial was conducted in accordance with the principles outlined in the Declaration of Helsinki, good clinical practice guidelines, and all applicable local regulatory requirements. Study design details are illustrated in Fig. 1 in the ESM.

After confirmation of eligibility, patients were randomized 1:1 to receive either HLX02 or EU-trastuzumab in combination with docetaxel. Randomization was conducted using a block randomization scheme and stratified by ER/PgR status, prior neo-/adjuvant therapy with trastuzumab, and ethnicity. An interactive web response system was used to assign patients to study groups as per a predefined randomization code. Randomization codes were not revealed to study participants, investigators, or study site personnel until all final clinical data had been entered into the database and the database had been locked and released for analysis. ORR and other outcomes were also assessed by blinded reviewers.

The study consisted of a 28-day screening period and a treatment period. In the treatment period, patients received HLX02 or EU-trastuzumab at an initial dose of 8 mg/kg over 90-min intravenous infusion on day 1, cycle 1, followed by 6 mg/kg study drugs once every 3 weeks for a maximum of 12 months. Docetaxel 75 mg/m² was administered over 60-min intravenous infusion on day 2 of cycle 1 and then 60 min after the infusion of HLX02 or EU-trastuzumab in the following cycles at the investigator’s discretion for a maximum of 12 months. Infusions were administered in line with site-specific protocols, local guidelines, and product information for reference trastuzumab.

The primary efficacy endpoint of this study was ORR up to week 24 (ORR₂₄), defined as the proportion of patients with a best response of complete response (CR) or partial response (PR) from the first assessment up to week 24. Secondary efficacy endpoints included ORR at weeks 6, 12, 18, and 24; disease control rate (DCR; the proportion of patients who achieved CR, PR, or stable disease [SD] for ≥ 12 weeks); clinical benefit rate (CBR; the proportion of patients who achieved CR, PR, or durable SD [SD sustained for ≥ 24 weeks]); duration of response (DoR); 12-month progression-free survival (PFS) rate; and 12-, 24-, and 36-month overall survival (OS) rate. Tumor response was evaluated based on Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1 by blinded CIR until week 24, after which it was evaluated by the principal investigator. Tumor assessments were performed at screening, weeks 6, 12, 18, and 24, and then every 9 weeks with computed tomography (CT) scan or magnetic resonance imaging (MRI). The method used was consistent throughout the entire study. Bone scans or X-rays and brain CT scan/MRI were performed at screening and during the treatment period if clinically indicated.

The safety and tolerability of HLX02 or EU-trastuzumab were evaluated in this study by recording the incidence, severity, and causality of treatment-emergent adverse events (TEAEs), serious TEAEs, serious adverse events (SAEs) and adverse events (AEs) of special interest (AESIs). TEAEs were defined as AEs that began or worsened in severity during or following the first administration of study medication and ≤ 30 days (± 2) following the last dose of study medication. As the most common safety issue with trastuzumab, cardiac function was monitored by echocardiogram (ECG) or multigated acquisition scan at screening (within 42 days before randomization) and after every 3 cycles (or more frequently if clinically indicated). Patients who permanently discontinued the study drug because of a drop in LVEF (for a persistent [> 8 weeks] decline of LVEF or for suspension of HLX02/trastuzumab dosing on more than three occasions for cardiomyopathy [21]) continued to undergo assessments until the LVEF values returned to ≥ 50%.

All AEs, physical examinations, vital signs, ECGs, and laboratory tests were graded according to National Cancer Institute Common Terminology Criteria for Adverse Events (NCI-CTCAE) version 4.03 and classified according to the Medical Dictionary for Regulatory Activities (MedDRA) version 21.1. Cardiac AEs were collected up to 12 months after randomization, in line with LVEF calculations. Routine laboratory tests were performed by the local laboratory at screening, during treatment, at the end of the study, and 30 days after the end of the last administration.

Pharmacokinetic blood samples were collected from all patients at cycle 1 (within 7 days prior to infusion) and every 3 cycles starting at cycle 3 (cycles 3, 6, 9, 12, and 15). Extended pharmacokinetic collections were collected from all patients in cycle 1 (at the end of infusion) and cycles 4 (prior to infusion) and 8 (prior to and after infusion). Immunogenicity (as assessed by antidrug antibody [ADA] and neutralizing ADA [NADA]) was evaluated with ADA and NADA at screening, cycles 3, 6, 9, 12, 15, and at the safety follow-up visit. The samples were submitted to the central laboratory (WuXi AppTec bioanalytical services department, Shanghai, China) and measured using validated assays.

Assuming a 5% dropout rate, a sample size of 608 was required to ensure that 578 patients (289 in each group) randomly received treatments to evaluate the equivalence between HLX02 and EU-trastuzumab with approximately 84% power. Equivalence was defined as the 95% confidence interval (CI) for the treatment difference being fully contained in the margins of ± 13.5%. These margins were derived by reviewing historical data from two randomized studies of trastuzumab in patients with HER2-positive metastatic breast cancer [8, 10] and estimated with the Der Simonian–Laird [22] random effect.

The intention-to-treat (ITT) set was defined as all patients randomly allocated to study drug, regardless of whether a dose of study drug was given. The per-protocol (PP) set comprised all patients who received ≥ 8 cycles of treatment and had one or more tumor assessment after treatment, or who discontinued treatment early because of disease progression or intolerable toxicity without major protocol deviations. The safety set included all patients who were randomly allocated and received at least one dose of study drug (HLX02 or EU-trastuzumab). The pharmacokinetic set comprised all patients who received study drug and had at least one measured concentration at a scheduled pharmacokinetic time point with no protocol deviations or other pharmacokinetics-affecting events.

The ITT and PP population sets were used for efficacy analyses. Per ITT set, patients with missing ORR assessments were regarded as nonresponders in the primary analysis. The difference (95% CI) in ORR₂₄ between the treatment groups was assessed for statistical significance with a chi-squared test. Sensitivity analyses of ORR between the two treatment groups (HLX02 and EU-trastuzumab) were conducted using a stratified Cochran–Mantel–Haenszel (CMH) test (95% Wald CI). The stratification factors for CMH tests were ER/PgR status, prior neo-/adjuvant therapy with EU-trastuzumab, and ethnicity. The CMH test used for the primary efficacy analysis was repeated for the PP set.

For the secondary efficacy endpoints, ORR, CBR, and DCR were analyzed using the same method as for the primary efficacy endpoint. Other secondary efficacy endpoints, such as DoR; PFS up to 12 months; OS rate at 12, 24, and 36 months; and OS at the cut-off date in the two treatment groups were presented graphically using Kaplan–Meier curves along with a summary of associated statistics (i.e., the probability of being event free) and the corresponding two-sided 95% CIs. Furthermore, the treatment difference, which was characterized by the “HLX02/EU-trastuzumab” hazard ratio (HR), was calculated using a stratified Cox proportional hazards model with ER/PgR status, prior (neo)/adjuvant therapy with trastuzumab, and ethnicity as covariates.

All analyses, summaries, and listings were calculated using SAS^® software (version 9.4 or later, SAS Institute Inc., Cary, NC, USA). This study and all data were monitored by an independent monitoring committee.

Between 11 November 2016 and 10 July 2019, a total of 1046 patients were screened, of whom 649 were randomized to receive HLX02 (n = 324) or EU-trastuzumab (n = 325) and included in the ITT set (Fig. 1). The PP set comprised 616 patients: 310 in the HLX02 group and 306 in the EU-trastuzumab group (Fig. 1). A total of 12 patients had one or more major protocol deviations: four (1.2%) in the HLX02 group and eight (2.5%) in the EU-trastuzumab group. Major protocol deviations in each group are listed in detail in Table 3 in the ESM.

Patient demographics and baseline characteristics were well-balanced between the treatment groups in the ITT set (Table 1) and the PP set (data not shown). Patients in the HLX02 and EU-trastuzumab groups had a mean age of 53.6 and 52.8 years, respectively. Around half of the patients were ER/PgR receptor positive (46 vs. 48%). The majority of patients in both treatment groups were Asian (> 75%). The proportion of patients aged > 50 years was slightly lower in the HLX02 group than in the EU-trastuzumab group (30.2 vs. 36.9%). The mean LVEF levels at baseline were 64.7 and 64% in the HLX02 and EU-trastuzumab groups, respectively.

The median follow-up durations were comparable in the HLX02 and EU-trastuzumab groups (457.0 vs. 455.0 days). Exposure to study drug and chemotherapy agent docetaxel were similar between the treatment groups (Tables 4 and 5 in the ESM). The mean number of treatment cycles completed were 12.4 and 11.8; mean days of treatment exposure were 264.6 and 253.4 in the HLX02 and EU-trastuzumab groups, respectively. Exposure to docetaxel was also similar (mean 8.2 vs. 8.1 cycles) between the two treatment groups.

In the ITT set, ORR₂₄ was 71.3 and 71.4% in the HLX02 and EU-trastuzumab groups, respectively, with an intergroup difference of − 0.1% (95% CI − 7 to 6.9) (Table 2), and the sensitivity analysis revealed a stratified intergroup difference of 0.1% (95% CI − 6.9 to 7). Both 95% CIs of intergroup differences fell completely in the predefined equivalence margins of ± 13.5%. Results in the PP set were comparable. The difference in ORR₂₄ of the HLX02 or EU-trastuzumab groups was 1.0% (95% CI − 6 to 7.9); the intergroup difference produced by a stratified CMH (sensitivity) analysis was 1.3% (95% CI − 5.7 to 8.2) in the PP set.

In addition, no statistically significant differences were observed between the two treatment groups in all secondary endpoint efficacy analyses and further sensitivity analyses (Table 2). At up to week 24, a similar proportion of patients in the HLX02 and EU-trastuzumab groups experienced a CR (5.2 vs. 3.7%) or PR (66 vs. 67.7%). ORRs (at weeks 6, 12, 18, and 24), DCRs, and CBRs were similar between the two treatment groups.

The median DoR (10.6 vs. 10.2 months; HR 0.79; p = 0.103) and median PFS (11.7 and 10.6 months; HR 0.83; p = 0.086) were comparable between the HLX02 and EU-trastuzumab groups (Fig. 2a, b). Median OS was not reached in the HLX02 group and was 28.5 months in the EU-trastuzumab group (Fig. 2c, HR 0.85; p = 0.388).

Overall, the safety profiles of HLX02 and EU-trastuzumab were similar (Table 3), with 98.8% of patients in the HLX02 and EU-trastuzumab groups experiencing one or more TEAEs. The most common TEAEs were decreased neutrophil count, decreased white blood cell count, and anemia. Grade 3 or higher TEAEs (85.8 vs. 86.4%) and serious TEAEs (23.8 vs. 24.9%) were reported in a similar proportion of patients in the HLX02 and EU-trastuzumab treatment groups. TEAEs leading to treatment discontinuation occurred in 3.1 and 3.4% of patients, and death due to TEAEs occurred in three patients in the HLX02 group (one case each due to lung infection, dyspnea, and pneumonia) and six patients in the EU-trastuzumab group (one case each due to dyspnea and cardiovascular event; one case due to electrolyte imbalance, arthralgia, and altered consciousness; and three cases due to general disorders and administration site conditions).

The most commonly reported AESIs were decreased white blood cell count (69.4 vs. 68.9%), decreased neutrophil count (66 vs. 64.3%), and anemia (37.7 and 40.9%) in the HLX02 and EU-trastuzumab groups, respectively. Cardiac disorders of special interest occurred in similar proportions of patients (4.9 vs. 5.2%) in each treatment group. Cardiac disorders of special interest classified by preferred terms are listed in Table 6 in the ESM. LVEF shifts from normal to abnormal were observed in 17 (5.3%) patients in the HLX02 treatment group and in 20 (6.2%) patients in the EU-trastuzumab group. LVEF did not show any significant changes from baseline to week 21 and week 48 in the two treatment groups (Table 4).

No notable differences in pharmacokinetic and immunogenicity endpoints were observed between the treatment groups. In the pharmacokinetic set, serum concentration–time profiles were comparable in the HLX02 (n = 321) and EU-trastuzumab (n = 324) groups (Fig. 2 in the ESM). In the safety set, seven (2.2%) patients in the HLX02 group and 17 (5.2%) patients in the EU-trastuzumab group were ADA positive prior to the initiation of treatments. Of these, four and six were NADA positive, respectively. Four (0.6%) patients (two patients in each treatment group) were considered overall ADA and NADA positive during the study (Tables 7 and 8 in the ESM). No SAEs were reported in these four patients.