Postmarketing safety and effectiveness of recombinant factor IX (nonacog alfa) in Japanese patients with haemophilia B

Abstract Introduction In 2010, nonacog alfa became the first recombinant factor IX (rFIX) available in Japan for patients with haemophilia B. Aim To determine real‐world safety (adverse events, incidence of inhibitors) and effectiveness of nonacog alfa in Japan. Methods This multicentre, prospective, observational, postmarketing surveillance study enrolled previously treated and untreated patients (PTPs and PUPs, respectively) who were observed for 1 and 2 years, respectively, after initiating nonacog alfa therapy. Safety and effectiveness were assessed for each treatment type. Annualized bleeding rate (ABR) and incremental recovery of rFIX were also evaluated. Results Overall, 312 of 314 patients enrolled from 173 sites were eligible for the safety analysis set (PTPs, 281; PUPs, 28; other, 3). Mean age was 25.4 (PTPs) and 14.8 (PUPs) years. Haemophilic severity ranged from mild to severe, and 133 (42.6%) patients had haemophilic arthropathy. Of 285 patients (PTPs, 257; PUPs, 28) in the effectiveness set, 112 received on‐demand treatment for 1161 bleeding episodes (effectiveness rate, 93.7%) and 185 received routine prophylaxis (effectiveness rate, 95.5%). No spontaneous bleeding was observed in 52.4% of patients during prophylactic treatment. Median ABR was lower during routine prophylaxis (2.0) vs the rest of the observation period (8.3). A weak negative correlation was found between body weight and the reciprocal of rFIX recovery. Eleven adverse drug reactions occurred in 7 PTPs (2.2% [7/312]); recurrence of inhibitor was observed in 1 patient, but no new inhibitor developed in PTPs or PUPs. Conclusion Nonacog alfa therapy is safe and effective in the real‐world scenario in Japan.


| INTRODUC TI ON
Haemophilia B is a hereditary coagulation disorder caused by deficiency or dysfunction of blood-clotting factor IX (FIX). In Japan, 1097 patients were reported to have haemophilia B in 2016. 1 Bleeding episodes in patients with haemophilia B are treated by FIX replacement. Routine prophylaxis is also widely adopted and involves regular FIX replacement therapy, aiming to maintain FIX levels to prevent bleeding. [2][3][4][5][6] Nonacog alfa (BENEFIX ® , Pfizer, New York, NY, USA), a coagulation FIX product, is a purified protein produced by recombinant DNA technology for use in haemophilia B therapy. It was approved in the United States and Europe in 1997 and is currently marketed in more than 50 countries worldwide. In Japan, Wyeth Pharmaceuticals Inc. In the last couple of years, several extended half-life blood coagulation factor products gained regulatory approval for the treatment of haemophilia B. 7 However, access to this most advanced treatment option remains limited to developed countries, 8  The objectives of this study were, therefore, to determine the real-world safety, including occurrence of adverse events (AEs) and incidence of inhibitors, and effectiveness of nonacog alfa among patients with haemophilia B in Japan.

| Study design and patients
This multicentre, prospective, observational, PMS study was conducted in accordance with the Japanese regulatory requirements stipulated in the Good Post-Marketing Study Practice (GPSP). The enrolment period was from January 2010 to December 2014 (5 years after product launch), during which all patients treated with nonacog alfa across Japan were to be enrolled into the study. All enrolled patients were followed up by their physicians in routine practice for 1 year after initiating nonacog alfa therapy for previously treated patients (PTPs) and 2 years for previously untreated patients (PUPs, ie patients who had previously received blood coagulation FIX products other than nonacog alfa for no longer than 3 exposure days). The target number of patients was 300, including 20 PUPs. Safety, effectiveness and FIX recovery data were collected every 6 months for each patient using case report forms completed by the physician investigators.
There were no specified activities, visits or investigations in the study.

| Safety assessment
The primary analysis was summarized using the number and incidence rate of adverse drug reactions (ADRs) according to the severity, action taken, seriousness, outcome and causality of AE to nonacog alfa. The safety analysis set (SAS) comprised patients who satisfied the eligibility criteria and received at least one dose of nonacog alfa. AEs and ADRs were tabulated according to system organ class and preferred terms of the Medical Dictionary for Regulatory Activities/Japanese (MedDRA/J) version 17.0.

| Effectiveness assessment
The effectiveness analysis set (EAS) comprised patients from the SAS with effectiveness evaluated at least once after infusion with nonacog alfa, excluding those with protocol violation, no visit after the first day of treatment, other than target disease and non-assessable effectiveness.
Effectiveness of nonacog alfa infusion used for routine prophylaxis, ondemand treatment and surgical prophylaxis was rated by the physicians vs their expectations using a 4-point scale of 'excellent' , 'good' , 'moderate' or 'no response' (Appendix S1). The annualized bleeding rates (ABRs) were also analysed to objectively assess the effectiveness of nonacog alfa.

| rFIX recovery
The incremental recovery of rFIX ([IU/dL]/[IU/kg]) was calculated as the observed postinfusion increase of FIX activity (IU/mL) in plasma divided by the rFIX dosage. The recommended conditions for assessment of recovery were evaluation of FIX levels in the non-haemorrhagic period at first-time rFIX replacement, a washout period of ≥4 days and 2 measurements that were recorded before and 30 minutes after infusion.

| Correlation coefficient of FIX level
The correlation of body weight and age with reciprocal of observed incremental FIX recovery was assessed using a scatter plot with a non-parametric regression based on locally weighted scatter plot smoothing. In each plot, Spearman's rank correlation coefficient was calculated and the test for non-correlation was performed.

| Effectiveness
Clinical effectiveness rate, with the corresponding exact 2-sided 95% confidence interval, was defined as the percentage of evalua- Patients were stratified by haemophilic severity, and the ABR was calculated for those with and without haemophilic arthropathy after dividing their observation period into routine prophylaxis period and rest of the observation period. Any divided treatment period lasting for <7 days was excluded from the ABR calculation. (42.6%) patients had haemophilic arthropathy; the most common joints involved were ankles, followed by knees and elbows. More than 60% of patients had arthropathy in the age groups of ≥20 years ( Figure 2).

| Nonacog alfa administration
Overall, there were 18 292 nonacog alfa infusions during the study pe-
One serious ADR of recurrence of anti-FIX antibody positive was observed in a PTP (0.32%), and 1 unexpected ADR of epilepsy was observed in a PTP (0.32%). Other non-serious ADRs were headache, dizziness, cough, dyspnoea, nausea, rash, urticaria and injection site pruritus. No thrombotic events were observed. No ADRs including new inhibitor development were observed among the 28 PUPs despite their high mean number (>100) of infusions ( during ITI with nonacog alfa after 4 months of inhibitor absence. In this patient, the inhibitor titre was >5 BU/mL when he was treated with plasma-derived FIX before enrolment. At baseline, the titre was 1 BU/mL, and this patient had history of allergy such as skin pruritus and urticaria but did not have history of anaphylaxis or nephrosis.
During the course of the ITI with nonacog alfa, after the recurrence of anti-FIX antibody positive, the inhibitor titre was elevated to 11 BU/mL at the maximum.

| Surgical prophylaxis (including dental procedures)
In the EAS, the effectiveness of surgical prophylaxis (including dental procedures) was evaluated in 28 surgeries in 21 patients (PTPs, 12; PUPs, 9). Effectiveness outcomes, with respect to blood loss, were deemed excellent in 10 events (35.7%) and good in 18 (64.3%). The overall clinical effectiveness rate was 100.0%. Six events (tooth extraction, hepatic resection, pulmonary artery plasty, laparoscopic cholecystectomy, artificial joint replacement in the left knee and artificial femoral head implantation) required blood transfusions.
F I G U R E 3 Nonacog alfa dose (IU/ kg; mean ± SD) in on-demand treatment (A) and routine prophylaxis (B). A, Of 113 patients who underwent on-demand treatment, eight were excluded from calculation of mean single dosage because there was no record of dosage (n = 1), body weight (n = 6) or age (n = 1). B, Of 186 patients who underwent routine prophylaxis, four patients were excluded from calculation of mean single dosage because there was no record of body weight. Eight patients with unknown severity were included in the overall group but excluded from mean dosage stratified by severity. SD, standard deviation

| rFIX recovery
The increased value of FIX was measured under recommended conditions in 129 events; corresponding median and mean (SD) reciprocal of recovery values were 1.284 and 1.449 (0.748), respectively.
Correlation between reciprocal of observed incremental FIX recovery and age was not significant (P = 0.055; Figure 6A). However, a weak, significant (negative) correlation was observed between the reciprocal of observed incremental FIX recovery and body weight; increase in body weight led to decrease in recovery (P < 0.001; Figure 6B). F I G U R E 5 Annualized bleeding rate summarized by existence of haemophilic arthropathy and severity of haemophilia (effectiveness analysis set). Any treatment period (routine or other) lasting for <7 days was excluded from the calculation of annualized bleeding rate. a Surgical prophylaxis, prophylaxis at chance, shortterm prophylaxis and FIX recovery test. FIX, factor IX plasma FIX level above 1% was implemented. 13  Our results demonstrated that nonacog alfa has a favourable risk-benefit profile in the control and prevention of bleeding events in patients with haemophilia B in Japan. As such, we suppose that nonacog alfa is still an important treatment option especially to maintain surgical operations and for patients with intense activity or those who need routine prophylaxis when access to extended half-life products is limited. Nonacog alfa can also be a significant alternative to plasma-derived FIX products, as it is a recombinant protein product and thus can minimize the risk of transmission of blood-borne pathogens.

| Limitations
Although results of this study should reflect real-world use of nonacog alfa, interpretation of this study is limited as it was implemented as a GPSP-compliant, all-patient, non-interventional, observational study. Also, the ABR between the routine prophylaxis period and rest of the observation period was compared in a non-randomized manner, and the study was not designed to assess the differences in ABR.
F I G U R E 6 Correlation between reciprocal of observed incremental FIX recovery and age (A)/body weight (B). FIX, factor IX; Loess, locally weighted scatter plot smooth

| CON CLUS ION
Results of this PMS study show that nonacog alfa therapy is safe and effective in the real-world scenario in Japan. The results suggest that nonacog alfa was well tolerated and appropriately used under routine clinical practice as on-demand, routine prophylaxis and in perioperative management of haemophilia B patients.

ACK N OWLED G EM ENTS
This study was funded and conducted by Pfizer Japan Inc. We would like to thank all of the physicians and staff members who cooperated in conducting this study at 173 institutions nationwide in Japan. of Pfizer R&D Japan GK. TK is a full-time employee of Pfizer Japan.

AUTH O R CO NTR I B UTI O N
KF was involved in study design, the interpretation of data and manuscript preparation. MT, TM and MS were involved in study design, the interpretation of data and revision of the manuscript. AT, HY and TK were involved in data analysis and manuscript preparation. All the authors read and approved the final version of the paper.
Editorial support was provided by Dr. Annirudha Chillar of Cactus Communications, and funded by Pfizer Japan.

DATA AVA I L A B I L I T Y
The data sets analysed during the current study are not available because data sharing with third parties was not included in the contract with study sites.