Infliximab biosimilar CT-P13 is interchangeable with its originator for patients with inflammatory bowel disease in real world practice

1. Patients

Japanese patients with IBD were prospectively enrolled in the PMS during a 28-month enrollment period starting in November 2014. The target number for analysis was 300 patients, who were treated at least for 4 months, including 100 patients with CD and 100 with UC. This sample size provides a power of 95% or more to detect 1 patient reporting an adverse event (AE) at an incidence of 1%.

The PMS was required by the Ministry of Health, Labour and Welfare (MHLW) as a condition for approval of CT-P13, and conducted as a centralized registration system of all-case surveillance. The protocol of PMS was approved by the MHLW, and all PMS procedures complied with the Good Post-Marketing Study Practice Ordinance of MHLW. Therefore, formal approval by an ethics committee was not required.

2. Treatment

The approved dosage of CT-P13 is 5 mg/kg on weeks 0, 2, and 6, and every 8 weeks thereafter. Dose escalation up to 10 mg/kg and/or interval reduction down to 4 weeks is allowed for CD patients with diminished response after week 6. In the PMS, therefore, naïve patients initially received induction therapy at a dose of 5 mg/kg, the patients switched for nonmedical reasons received their first dose of biosimilar during maintenance therapy at intervals of 8 weeks, and some CD patients received higher-dose treatment.

3. Surveillance Methods

The duration of follow-up was 2 years, and case report forms (CRFs) were collected from each patient 4 months, 1 year, and 2 years after the start of treatment. In addition to patient characteristics and disease status, CT-P13 regimen, previous therapy, concomitant drugs, and ADR-related information, including subjective/objective findings and laboratory test data, were collected. Reported ADRs were coded by System Organ Class and Preferred Term, as listed in the Medical Dictionary for Regulatory Activities (MedDRA/J; version 20.0). Incidences of increased ALT and ALP were evaluated in accordance with the diagnostic criteria for drug-induced liver injury set by the Japan Society of Hepatology. Specifically, ALT more than twice the upper limit of normal (ULN), and ALP more than the ULN at each clinical site, were defined as increased ALT and ALP, respectively. When baseline values exceeded the ULN, the increase in ALT or ALP during CT-P13 treatment was used for this definition.

Efficacy was periodically assessed by the physician as effective, partially effective, ineffective, progressed, or not evaluable. The disease status evaluated by endoscopic examination was reported as severe, moderate, mild, or in remission. In addition to physicians’ assessments, changes in CDAI, partial Mayo score, and CRP were used as efficacy parameters. The data obtained on the nearest day to the scheduled administration time points (on weeks 2, 6, and 14, and every 8 weeks thereafter for naïve patients and patients switched for medical reasons; and every 8 weeks for patients switched for nonmedical reasons) were used as the representative values.

The response of each CD patient to CT-P13 was evaluated by comparing the CDAI on week 14 to baseline, in naïve patients and those switched for medical reasons; and CDAI on week 16 to baseline, in those switched for nonmedical reasons. Responders were defined as showing decreases of more than 70 from baseline, or scores less than 150 in patients with baseline scores above 150. Consistent responders were defined as having scores less than 175 throughout the relevant period, or showing increases of less than 70 in patients with baseline scores below 150. Nonresponders were defined as showing score increases from below 150 at baseline to above 175, or having scores above 150 throughout the relevant period without decreases of 70 or more. Partial Mayo scores were employed for the evaluation of each UC patient in the same manner, with a cutoff score of 3.

4. Database Search

Clinical data, including ALT and ALP, for patients prescribed IFX during the same period as the PMS were obtained from EBM Provider, a Japanese hospital-based administrative database created by Medical Data Vision Co., Ltd. (Tokyo, Japan), as reference data for comparison with CT-P13 [6]. Incidences of increases in ALT/ALP were assessed using the common standard laboratory values proposed by the Japanese Committee for Clinical Laboratory Standards.

5. Statistical Analysis

To identify risk factors that affect the occurrence of ADR and infusion reaction (IR), patient factors with P<0.2 in the Fisher exact probability test or the chi-square test were selected in a stepwise manner to conduct multivariate analysis using a logistic regression model.

The duration of treatment was plotted by the Kaplan-Meier method, with discontinuation of treatment as an event. Patients with whom treatment was discontinued within 7 days after the start of treatment were excluded. Differences between patient groups were analyzed using the generalized Wilcoxon test.

1. Patient Characteristics

In order to investigate the safety and efficacy of a biosimilar of IFX, we initiated PMS, and registered 700 patients during the enrollment period. In this interim analysis, 523 patients whose CRFs at month 4 had been collected were included in the safety and efficacy analysis set (Fig. 1A). Further reported data in 270 collected CRFs at 1 year and in 56 final CRFs at the end of 2-year follow-up period were fixed and analyzed. As shown in Fig. 1B, the ratios of biologic-naïve patients to patients switched for nonmedical reasons from the originator IFX were quite different between the CD and UC groups. The number of patients switched to CT-P13 for medical reasons was less than 20% in both CD and UC patients. This patient group with medical switch consisted of (a) patients switched from the originator IFX to CT-P13 due to medical reasons such as AEs or loss of response, (b) patients received CT-P13 after adalimumab (ADA), and (c) patients retreated with CT-P13 for relapse after discontinuation of IFX due to remission of disease.

Fig. 1.

Patient demography (A) and patient classification based on prior biologic therapy and reasons for switching to CT-P13 (B). Patients were classified into 3 groups: naïve patients who had no previous anti-TNF-α antibody treatment, patients switched for nonmedical reasons such as institutional policy or economic reason (nonmedical switch), and patients switched for medical reasons (medical switch). The medical switch group was further devided into 3 subgroups. ^aSwitching to CT-P13 from infliximab (IFX) due to adverse events or insufficient efficacy; ^bSwitching from adalimumab; ^cRetreatment with CT-P13 for relapse after discontinuation of IFX due to remission of disease.

The characteristics of patients and the disease statuses in these groups of patients with CD or UC were markedly different (Table 1). The duration of disease in naïve patients was shortest, and most of the naïve patients had severe or moderate disease. Patients who were switched for nonmedical reasons showed stable disease conditions, as indicated by decreased CDAI or partial Mayo scores due to previous treatment with IFX. They included lower proportions of patients with external fistulas or anal lesions in CD patients, and with steroid dependence in UC patients. The patients in the medical switch group had the longest disease duration, but the disease severity was comparable to that in the biologic-naïve patient group. Concomitant use of steroids was less frequent in the nonmedical switch group, and the rate was higher in UC patients than CD patients. The proportions of patients concomitantly administered immunomodulators showed no clear differences between the patient groups.

Table 1.

Patient Characteristics and Disease Status in Each of the 3 Groups of Patients with CD and UC

2. Treatment Status

The median initial dose of CT-P13 in this PMS was 5 mg/kg, which was same as the approved standard dose. While almost all patients with UC were administered the usual dosage less than 8 mg/kg, higher doses were used in 61 patients (22.8%) with CD, for whom dosage up to 10 mg/kg is permitted when response become low. However, the incidences of ADRs and IR with higher dose of CT-P13 were not high as compared with lower standard doses, and no serious ADR was reported.

The range of infusion duration was 60 to 600 minutes, and more than 50% of patients received 120-minute infusion. The incidences of ADR and IR were independent of the infusion duration.

3. Incidence of ADRs

Of the 523 patients in the safety analysis set, a total of 144 ADRs were reported in 106 patients (20.27% by number of patients, 26.64% per 100 person-years), and the most common ADR was IR reported in 49 patients (9.37% and 12.31%), as listed by System Organ Class in Table 2.

Table 2.

Incidence of ADRs and Severe ADRs to CT-P13 in 523 IBD Patients in the Safety Analysis Set during the Observation Period of 397.9 Person-Years

Twenty-nine serious ADRs were reported in 22 patients (4.21% and 5.53%). Although 20 patients recovered, 1 patient who experienced 3 serious ADRs, died of retroperitoneal hemorrhage and pneumonia following pancytopenia due to reactivation of cytomegalovirus. This patient showed a complicated clinical course after long-term use of steroid, and CT-P13 was not the sole cause for the severe ADRs, although involvement of CT-P13 cannot be ruled out. Recovery from the remaining one serious ADR of tuberculous meningitis could not be confirmed due to transfer to another hospital. A total of 35 patients, including these 2 patients, were withdrawn from CT-P13 therapy due to ADRs, of which 23 cases were IR.

Six cases of hepatobiliary disorders were reported by physicians. In addition, increased ALT and/or ALP fulfilling the diagnostic criteria for drug-induced liver injury set by the Japan Society of Hepatology was observed in 66 patients (16.3%). However, no patient discontinued CT-P13 treatment due to the increase in ALT/ALP, although 9 patients were withdrawn from treatment for other reasons such as IR, alopecia, insufficient efficacy, and personal reasons. The marker levels decreased to within normal range under treatment with CT-P13 in 45 patients (recovery in 12 patients was not confirmed, because subsequent CRFs were not collected), showing that the increases in liver function markers were transient, and did not become serious in any patient.

The incidences of ADRs and IRs in patient groups according to the prior biologics were compared in Table 3. In naïve patient group, ADRs were reported in 24.9% of patients, and the incidence of IRs was 9.7%. The incidences of ADRs and IRs were lower in patients who were switched for nonmedical reasons than other patient groups. On the contrary, ADRs were reported in 32.2% of the patients in medical switch group, and IRs was frequently observed in patients pretreated with originator IFX or ADA.

Table 3.

ADRs and Infusion Reactions of CT-P13 According to Prior Biologics

4. Risk Factors for ADRs

To investigate the impact of patient factors on ADRs, multivariate analysis was performed using a logistic regression model (Table 4). IR and other systemic ADRs were analyzed separately, since those had different characteristics as well as apparent timing after administration of CT-P13.

Table 4.

Multivariate Logistic Analysis of Incidence of Infusion Reaction and Other ADRs to CT-P13 Based on Patient Factors

History of drug allergy was the most significant factor for IR (P<0.01). A particularly high incidence of IR was reported in patients who had allergy to IFX (31.8%, 7/22). The incidence of IR in patients who were allergic to ADA or 5-aminosalicylic acid were 28.6% (2/7) and 20.8% (5/24), respectively. Patients who had a history of allergies to other drugs or nonmedicinal substances did not show higher incidence of IR compared to patients without an allergy history.

Prior treatment with biologics was another significant factor for IR. The incidence of IRs in biologic-naïve patients without prior biological therapy was significantly higher than in patients switched for nonmedical reasons (P<0.01), but also significantly lower than in patients switched for medical reasons (P<0.01).

For ADRs other than IR, only switching from prior IFX for nonmedical reasons was a significant factor lowering their incidence. No other factors were significant, but principal disease (CD or UC) and concurrent disease affected the incidence of other ADRs (P=0.074 and P=0.068, respectively).

5. Treatment Continuation

For evaluation of the general usefulness of CT-P13, the treatment durations of CT-P13 were analyzed by the Kaplan-Meier method with mean observation period of 299 days. The treatment continuation rate was higher in CD than UC patients (79.8% vs. 62.1%, P<0.001; generalized Wilcoxon test). Patient withdrawal due to ADRs was lower in CD patients (5.2%) than in UC patients (11.3%), as was withdrawal due to insufficient efficacy (11.1% and 15.6% in CD and UC patients, respectively).

For both CD and UC patients, the treatment duration was significantly longer in patients switched for nonmedical reasons than in naïve patients or patients switched for medical reasons (Fig. 2). Although treatment continuation rate was higher in the naïve group than in the medical reasons group over time, no significant difference was shown due to limited number of patients in each group.

Fig. 2.

Kaplan-Meier plot of treatment period of CT-P13 in CD (A) and UC patients (B). Patients who ceased further treatment with CT-P13 earlier than 7 days from the initial administration were excluded from the plot. Statistically significant differences were analyzed by the generalized Wilcoxon test. ^aP<0.05, ^bP<0.01, ^cP<0.001.

6. Efficacy Analysis

Next, we evaluated the efficacy of CT-P13 on the basis of clinical scores (CDAI and partial Mayo score) and CRP, as well as assessment by primary-care physicians (Fig. 3). In patients switched for nonmedical reasons, lowered clinical scores and CRP by the previous treatment with the originator were maintained after switching to CT-P13, and the efficacy rate assessed by physicians to be “effective” or “partially effective” was maintained at 90% or more. In naïve patients, the effect of CT-P13 appeared immediately after initiation of treatment, as shown by decrease in CRP on week 2, followed by decreases in CDAI or partial Mayo score. These efficacy parameters remained in low ranges thereafter. In patients switched from original biologics for medical reasons, all the variables decreased, as seen in naïve patients, and the efficacy rate assessed by physicians increased and remained above 80%.

Fig. 3.

Changes in 3 efficacy variables in CD (A-C) and UC patients (D-F). Evaluated CDAI (A), partial Mayo score (D), and measured CRP (B, E) are expressed as mean±SD. Effective rate of clinician-reported efficacy (C, F) is based on assessment of the response to CT-P13 as effective or partially effective by primary care physicians.

Furthermore, we evaluated responses of each patient to CT-P13 on week 14 or 16 based on changes in CDAI or partial Mayo score (Fig. 4). Most of naïve patients were assessed as responders showing substantial decreases from baseline clinical score. On the contrary, in the nonmedical switch group, most patients had low clinical scores even at the start of treatment with CT-P13 due to previous IFX therapy, and were classified as consistent responders with scores that remained low. In the medical switch group, a high proportion of nonresponders showed increases from low baseline scores or no change from high baseline scores, but responders and consistent responders still made up the majority of patients (72% and 79% for CD and UC, respectively). The proportions of responders and consistent responders were higher in patients switched from ADA than in those switched from IFX (82.6% vs. 65.0%). The favorable response rate (77.8%) was observed in patients retreated with CT-P13 after discontinuation of originator IFX due to remission, as high efficacy rate was reported with IFX retreatment [7-9].

Fig. 4.

Efficacy of CT-P13 based on CDAI in CD (A) and partial Mayo score in UC patients (B). Efficacy of treatment with CT-P13 in each patient was assessed based on CDAI or partial Mayo score estimated at week 14/16 in comparison with baseline scores (see methodological details described in the text).