Efficacy and safety of RGB-02, a pegfilgrastim biosimilar to prevent chemotherapy-induced neutropenia: results of a randomized, double-blind phase III clinical study vs. reference pegfilgrastim in patients with breast cancer receiving chemotherapy

The study population included chemotherapy-naïve women ≥18 and ≤ 65 years of age with invasive breast cancer (Stage IIB and III) appropriate for combined treatment with doxorubicin/ docetaxel in the neo−/adjuvant treatment setting. Additional inclusion criteria were Eastern Cooperative Oncology Group (ECOG) performance status 0 or 1 and adequate bone marrow function, as indicated by absolute neutrophil count (ANC) ≥ 1.5 × 10⁹/L, platelet count ≥100 × 10⁹/L, and Hemoglobin > 8 g/dL; adequate renal and hepatic function with an estimated creatinine clearance ≥50 mL/min (Cockcroft-Gault method), bilirubin, aspartate transaminase, alanine transaminase < 1.5 x upper limit of normal (ULN), and Alkaline phosphatase < 2.5 x ULN. Women with childbearing potential had to present a negative urine pregnancy test and had to agree to use 2 reliable methods of contraception during treatment period and for 3 months thereafter. Written informed consent had to be given prior to any study-related procedure. Main exclusion criteria were any other malignancy within 5 years prior to randomization, with the exception of cervical carcinoma in situ, non-melanoma skin cancer, or superficial bladder tumors (Ta, Tis, or T1) that had been successfully and curatively treated; active infection or systemic anti-infective treatment, radiation therapy within 4 weeks prior to randomization; past exposure to any G-CSFs; concurrent anti-cancer therapy, concomitant treatment with bisphosphonates; prior bone marrow or stem cell transplantation; history or presence of sickle cell disease, and significant cardiovascular disease (caution especially for doxorubicin).

Eligible patients were randomly assigned to receive the study drugs (6 mg s.c. of either RGB-02 or reference) in a 1:1 ratio via an interactive voice/web response system (IXRS). On day 1 of each treatment cycle, all patients received 60 mg/m² doxorubicin followed 1 h later by an intravenous infusion of 75 mg/m² docetaxel. Chemotherapy was to be repeated every 3 weeks for up to 6 cycles. Patients were dosed with the study drugs approximately 24 h after chemotherapy was initiated for each cycle.

The study design was set up in a double blind fashion for the first 2 treatment cycles to demonstrate confirmatory efficacy followed by an open-label safety assessment during treatment the subsequent cycles. Patients in the reference arm were switched to open-label RGB-02 starting with cycle 3. After the initial 3-weeks-screening period and baseline visits, patients were scheduled for 4 chemotherapy cycles (and 2 additional if deemed necessary) each requiring 12 study visits within 3 weeks followed by a final safety assessment 6 months after treatment start. Pre-defined hematology blood samplings to determine the absolute neutrophil count for efficacy assessment were performed on Days − 1, 1, 2, 3, daily from Day 5 to Day10 and on Days 14 and 18. All patients underwent a baseline clinical examination that included physical examination, pregnancy testing, safety monitoring including hematology data (Hemoglobin; WBC count, lymphocytes (absolute), neutrophils (absolute), platelets); testing of hepatic function (aspartate aminotransferase, alanine aminotransferase, and bilirubin), lactate dehydrogenase, albumin, blood urea nitrogen, uric acid, creatinine clearance.

Blood sampling for immunogenicity assessments of pegfilgrastim was performed at Day − 1 of Cycles 1, 3 and 4, before any study treatment was administered and at the end of Cycle 4 and at follow-up. A stepwise approach was established for immunogenicity assessment, namely all samples were analyzed with screening assays for anti-RGB-02 and anti Neulasta® antibodies. The samples assessed positive with the screening assays were to be analyzed with confirmatory assays and the confirmed positive samples would have been analyzed with a cell-based neutralizing assay. Immunogenicity assays used were developed and validated in line with the applicable guidelines and recommendations [14, 19, 20].

The sample size of 111 evaluable patients per treatment arm was determined based on an equivalence test of means using two 1-sided tests on data from a parallel-group design in order to achieve 90% power at 5% significance level when the true difference between the means was assumed to be 0.25, the standard deviation (std) was assumed to be 1.70, and the equivalence limits were − 1.00 and 1.00 days.

The primary efficacy variable was the duration of severe neutropenia, defined as ANC < 0.5 × 10⁹/L, in the first cycle of chemotherapy. The difference in mean duration of severe neutropenia between the 2 treatment arms and the 2-sided 95% confidence interval (CI) for the difference between means was calculated using an analysis of covariance (ANCOVA) model with treatment, country, chemotherapy treatment setting (neoadjuvant or adjuvant) as factors, and baseline ANC value (value at Day − 1, Cycle 1) as covariate in the model.

If the upper limit of the 95% CI for the difference in means was ≤1 day and the lower bound of the CI for the difference in means was ≥ − 1 day, then the means in the 2 arms were to be considered equivalent. A similar analysis was performed for Cycle 2.

The duration of severe neutropenia in Cycles 3 and 4 as well as the depth of ANC nadir in Cycles 1 and 2 were summarized using descriptive statistics. An ANCOVA analysis was also performed for the difference in depth of ANC nadir.

The difference in the incidence of patients with febrile neutropenia in Cycles 1 and 2 between the 2 treatment arms with associated 95% CI was presented. Time to ANC recovery in Cycles 1 and 2 was analyzed using Kaplan-Meier life table methods. The analyses were performed using the protocol definition for ANC recovery (number of days from any ANC value < 0.5 × 10⁹/L to ANC ≥ 2 × 10⁹/L) and repeated for the alternative definition (number of days from the date of the lowest measured ANC value to ANC ≥ 2 × 10⁹/L). The primary data set for efficacy analysis was the per-protocol (PP) population; the full analysis set (FAS) was analysed in addition for demonstrating robustness of data. All patients who received at least one dose of a study medication were included in the safety analysis. Safety variables were summarized by treatment arm. All statistical analyses were performed using SAS 9.2 software.

Therapeutic equivalence could be demonstrated between RGB-02 and the reference pegfilgrastim with 95% CIs within the predefined margins of ±1 day confirming equivalence.

The mean duration of severe neutropenia (DSN, defined as the number of days from the first ANC value < 0.5 × 10⁹/L until increasing back ≥0.5 × 10⁹/L) in the PP collective during cycle 1 was comparable in the RGB-02 arm (1.7 ± 1.14 days) and the reference arm (1.6 ± 1.31 days). The LS Means (95% CI) were 1.5 (1.2, 1.8) and 1.4 (1.1, 1.7) days, respectively. The LS Mean for the difference in duration of severe neutropenia was 0.1 days (95% CI: -0.2, 0.4), identical to that observed in the FAS population. In cycle 2 the DSN declined equally to 0.7 days for both compounds (Table 2). The mean duration of severe neutropenia in Cycles 3 and 4, after patients in the reference arm switched to RGB-02, was < 1 day in the RGB-02 arm (0.9 days) and the reference arm switched to RGB-02 (0.6 days), indicating that the switch from reference to RGB-02 treatment did not increase the patient’s risk to develop longer lasting grade 4 neutropenia.

The similarity of results in the primary efficacy variable between the PP population and the FAS further supports the observed equivalent effect and robustness of data.

Neither statistically significant nor clinically relevant differences were detected in secondary endpoints between treatment groups.

The mean daily ANC values for both groups in Cycle 1 were almost identical (Fig. 2).

Most patients experienced severe neutropenia (defined as ANC < 0.5 × 10⁹/L) during cycle 1. The incidence of severe neutropenia decreased in cycle 2 compared to cycle 1 in both treatment groups with no statistical significant differences, for RGB-02 from 84.6% (99 patients) to 54.1% (60 patients) and for the comparator groups from 77.0% (87 patients) to 43.7% (45 patients) (Table 3).

During Cycle 1, the observed incidence and the overall incidence of febrile neutropenia including cases meeting the ESMO criteria and those who started systemic antibiotic treatment was similar in both groups with 5 patients [4.3%] and 10 patients [8.5%]) in the RGB-02 arm and 4 patients [3.5%] and 8 patients [7.1%]) in the reference arm. During Cycle 2, no febrile neutropenia was observed in any of the treatment arm except one overall febrile neutropenia case in the RGB-02 group (Table 4).

There were no significant differences between groups regarding mean time to ANC recovery (defined as recovery from any ANC value < 0.5 × 10⁹/L to ANC ≥ 2 × 10⁹/L). During Cycle 1, mean time to recovery was 3.4 ± 1.84 days in the RGB-02 arm and 3.7 ± 1.88 days in the reference arm; during Cycle 2, mean time to recovery was 2.8 ± 1.09 days and 3.4 ± 2.11 days, respectively. The recovery from the date of the lowest measured ANC value was comparable between groups, too (Fig. 2). Also, no clinically meaningful difference was found when comparing the mean depth of ANC nadir in Cycle 1 and 2.

When analysing the safety population one has to take into account that the reference drug was given only for the double-blind period of the first 2 Cycles. All patients from that group received thereafter RGB-02 for cycles 3–6. The safety population for the comparative safety analysis (Table 5) comprised 238 women (RGB-02 = 121, reference = 117) whereas altogether 234 women received 995 doses of RGB-02 throughout the course of the study. In total, 204/234 (87.2%) patients treated with RGB-02 had at least 1 adverse event (AE).

The cumulative incidence of adverse events was similar between both groups.

During Cycles 1 and 2, 80.2% treatment-emergent adverse events (TEAE) were reported in the RGB-02 arm (n = 97) compared to 93.2% in the reference arm (n = 109). Similarly, the number of patients with IMP-related AEs was marginally lower in the RGB-02 arm (17 patients [14.0%]) compared to the reference arm (27 patients [23.1%]). The most frequent pegfilgrastim-related AE was bone pain, slightly less frequently reported in the RGB-02 arm (14 patients [11.6%]) compared to the reference arm (20 patients [17.1%]).Other musculoskeletal and connective tissue disorders included arthralgia in the RGB-02 arm (2 patients [1.7%]. Myalgia (3 patients [2.6%], pain in extremity and spinal pain (2 patients [1.7%] each) were reported in the reference arm only.

Serious adverse events (SAE) were reported during the double-blind period at Cycles 1 and 2 in 8.3% of the RGB-02 arm and 6.8% of the reference arm cases, none of them related to pegfilgrastim (Table 6). The most frequent SAE was febrile neutropenia: 4.1% of patients in the RGB-02 arm and 5.1% of patients in the reference arm. There were 2 deaths reported in the RGB-02 arm, none of them were related to the investigational medicinal product (IMP). Causes for death were metastases to central nervous system and viral infection.

The screening assay test was negative in 96.1% of immunogenicity samples for the double blind treatment period. However, all positive screening tests were negative in the confirmatory test. Neutralising assay tests were not performed since none of the confirmatory tests were positive. In the open-label period including the 6-month follow - up, the screening assay test was negative in a range of 96.3–98.2% of all obtained immunogenicity samples at different time points. Again, positive screening tests turned out to be negative in the confirmatory test, therefore no neutralising assay tests were performed.

In conclusion, no patients of either treatment group had true positive immunogenicity results. The re-treatments did not increase the incidence of positive immune responses and the switch from reference treatment to RGB-02 did not provoke any immunogenic response.