Real-world safety of nusinersen in Japan: results from an interim analysis of a post-marketing surveillance and safety database

Abstract Purpose Nusinersen is the first disease-modifying therapy to treat spinal muscular atrophy (SMA). This report describes the safety and effectiveness of nusinersen in Japanese clinical use using two data sources: an ongoing Japanese post-marketing surveillance (PMS) and the safety database of the marketing authorisation holder, Biogen . Materials and Methods The PMS is evaluating the safety and effectiveness of nusinersen in all patients treated with nusinersen in Japan between August 2017 and August 2025; this interim analysis included data up to May 30, 2019. Biogen safety database data up to June 30, 2019 were also included to capture adverse events (AEs) from after the interim analysis cutoff date. Collected data included medical history, dosage and administration, and AEs. Safety assessment included AEs and serious AEs (SAEs). Effectiveness analyses included motor function assessments and clinical global impressions of improvement. Results Of 271 patients in the PMS population, 94 had SMA type I (34.7%), and 177 had SMA types II–IV (65.3%). AEs occurred in 67 patients (24.7%) and SAEs in 23 patients (8.5%). The Biogen safety database contained reports of 345 AEs; the most common were pneumonia, headache, and pyrexia, consistent with symptoms of SMA and lumbar puncture. In the analysis set, 26.2% of patients receiving nusinersen showed motor function improvements and 99.6–100.0% showed overall improvement. Conclusion In this interim analysis of the PMS and Biogen safety database, nusinersen had a favourable benefit–risk profile in Japanese patients with SMA.


Introduction
Spinal muscular atrophy (SMA) is a leading genetic cause of death among infants [1,2], and the second most common fatal autosomal recessive disease globally after cystic fibrosis [3]. The estimated annual incidence is 1 in 10,000 live births [2,4,5], and approximately 1 in 50 people are SMA carriers [2]. The majority of SMA cases (~99%) result from homozygous disruption of the survival motor neuron (SMN) 1 gene, which causes insufficient expression of SMN protein [3]. Patients with SMA also have a variable number of copies of the SMN2 gene, which is very similar to the SMN1 gene but produces a truncated form of the SMN protein that is unstable and is therefore unable to compensate for the lack of SMN1 [1]. SMA is broadly classified into five clinical subtypes (types 0-IV) depending on age at first symptom onset and highest motor milestone achieved. Signs of type 0 are usually apparent before birth, whereas signs/symptoms of type I appear at birth or within the first 6 months, type II at 7-18 months, type III at >18 months, and type IV generally after the age of 35 years [1,6]. Motor development and life expectancy differ between the subtypes. The most severe form is type 0, in which patients achieve no motor milestones and rarely live past 1 month, patients with type 1 never sit, those with type 2 never stand, and the least severe is type IV, which is associated with a normal life span and low levels of debility [6]. Within each subtype, disease severity is related to the number of SMN2 copies the patient has -a higher number of SMN2 copies usually confers less marked disease severity [7]. Until the development of nusinersen, there were almost no treatment options, other than supportive care, for patients with SMA.
Nusinersen binds to the intronic splicing silencer (ISS)-N1 site on intron 7 of the precursor mRNA transcript for SMN2 [8]. This results in a greater inclusion of exon 7 in SMN2 mRNA transcripts, which in turn increases production of full-length SMN2 mRNA and subsequently full-length SMN protein [9][10][11]. The safety and effectiveness of nusinersen were demonstrated in two randomised, double-blind, sham-controlled clinical trials -ENDEAR (in patients with infantile-onset SMA) [12] and CHERISH (in patients with later-onset SMA) [13]. These studies demonstrated that intrathecal administration of nusinersen led to a significant improvement in motor development, motor function, and survival compared with sham-procedure in infants and children with, or most likely to develop, SMA type I, II or III [12,13]. The most common adverse events (AEs) in these studies were fever and respiratory tract infections, including pneumonia [12,14]. Based on these trials, nusinersen received marketing authorisation approval in the United States (approved in 2016), the European Union (2017), and Japan (2017), and has since been approved for use in many other countries.
Since its approval in Japan for the treatment of SMA, information on the safety and effectiveness of nusinersen has been collected in a post-marketing surveillance (PMS), which is mandatory under Japanese regulatory law. The PMS is designed to assess the safety and effectiveness of nusinersen treatment in clinical practice from the time of nusinersen launch. In addition to the PMS, the marketing authorisation holder of nusinersen (Biogen) collects and analyses all reports of AEs reported to Biogen safety database.
This report describes the safety data collected on nusinersen during the initial years of use from these two sources: an interim safety analysis of data from the Japanese PMS between August 2017 and May 2019 and the safety information collected from Japanese patients and entered into the Biogen safety database between August 2017 and June 2019. Effectiveness data from the PMS are also reported.

PMS
The ongoing PMS was started in August 2017 (the initial market launch date in Japan) and will continue until August 2025 (8 years from the launch). The PMS includes all patients who receive nusinersen treatment in routine clinical practice (i.e. not during a clinical trial) during the enrolment period, August 2017 to August 2025; there are no exclusion criteria and no target sample size of patients. Patients are followed from PMS registration until death, termination of nusinersen administration, or the end of the investigation period (8 years), whichever occurs earliest. This report presents the safety outcomes reported in the Japanese PMS between August 2017 and 30 th May 2019. The Japanese segment of the PMS is being conducted by Biogen Japan, Ltd. All procedures in the PMS are consistent with the Declaration of Helsinki and Good Post-Marketing Study Practice (GPSP) in Japan. Under Japanese GPSP regulations, patients are not required to provide informed consent for PMS studies, but we obtained consent from patients and/ or their caregivers to use anonymised data in research publications.
Participating physicians enrolled patients into the PMS and completed a paper or electronic case report form at 3 months after the start of nusinersen administration, and annually thereafter. The following data were captured: patient medical history and allergy history; gestational age at birth; prior medications for SMA; information on SMA (SMA type, age at symptom onset, family history, genetic test for SMN1 deletion/ mutation, and copy number of SMN2); details of nusinersen administration (age at first injection, administration date, dosage, administration time, administration site, removal of cerebrospinal fluid, treatments at the time of nusinersen administration, height, head circumference, bodyweight at the time of administration, lot number of nusinersen); symptoms at the time of administration; clinical global impressions of improvement (CGI-I); development of any AEs and details of these events; concomitant medications and therapies; rehabilitation information; and clinical laboratory data. Comorbidities that may affect assessment of AEs were also recorded, specifically renal and hepatic dysfunction, which have been previously reported as possible side effects of antisense oligonucleotides [15].
The PMS safety analysis will include all patients who receive nusinersen in Japan between August 2017 and August 2025. The current analysis was limited to registered patients who had completed at least 3 months of observation, and whose baseline and follow-up questionnaires had been completed by the physician, submitted for analysis, and were evaluable for analysis by May 30 th 2019. The safety outcome measures were the occurrence of AEs, infectious diseases, and serious AEs (SAEs).
AEs and SAEs were coded using the Medical Dictionary for Regulatory Activities Terminology (MedDRA)/J version 22.0 and classified by system organ class (SOC) and preferred term (PT).
Effectiveness outcome measures in the PMS were time to death; time to permanent assisted ventilation (defined as need for tracheostomy, tracheal intubation, or ≥16 h/day of non-invasive ventilatory support used continuously for >21 days); motor function assessment; and overall improvement scales. Motor function was assessed using one or more of the following, depending on age, SMA type and functional ability: the Hammersmith Infant Neurological Examination (HINE) [16]; expanded World Health Organization (WHO) Motor Milestones [17]; expanded Hammersmith functional Motor Scale [18]; and the 6-min walk test [19]. Treatment was considered "effective" on the HINE if the number of endpoints that increased by ≥1 point on the 4-point scale (or increased by ≥2 points or to the highest score for "kick") was greater than the number that decreased by ≥1 point (or decreased by ≥2 points or to the lowest score for "kick"), since preliminary data suggest that an increase of between 0.4 and 0.7 is clinically meaningful [20]; on the WHO Motor Milestones score if motor milestones at baseline other than "consciousness during assessment" and "excitability during assessment" were maintained and ≥1 new milestone was achieved; on the Hammersmith functional Motor Scale (HfMS) if the total score of all endpoints improved by ≥3 points (maximum total possible score on this scale is 40 and >1-point increase is considered meaningful [21]); and on the 6-min walk test if there was improvement by ≥30 m (the minimum detectable change in 90% of patients [MDC 90 ] is 24 m in patients with SMA [19]). Nusinersen was considered effective if any one of the motor function assessments showed improvement using the above criteria at any given assessment after baseline, since data were not collected at prespecified timepoints.
for CGI-I, the degree of improvement from the previous administration of nusinersen was evaluated by physicians on a 7-point scale, where marked improvement = +3, moderate improvement = +2, mild improvement = +1, unchanged = 0, mild deterioration = −1, moderate deterioration = −2 and marked deterioration = −3 [22]. Two analyses of CGI-I were undertaken: (1) nusinersen was considered to be effective if the degree of improvement was "unchanged" or better at any time point during the survey period, since clinical stabilization and improvement are both meaningful outcomes in a condition characterized by progressive deterioration; and (2) nusinersen was considered to be effective if the total of the values at each assessment was ≥0.
Since HINE is the only one of these endpoints which is applicable to all SMA subtypes, we also analysed HINE data in patient subgroups based on SMA subtype.
Nusinersen was administered at the discretion of the patients' own physicians. The Japanese package insert for nusinersen specifies that the drug is administered by the intrathecal route over 1-3 min. After the first dose, the approved administration regimen for patients with infantile-onset SMA (type 1) is to receive subsequent doses at 2, 4 and 9 weeks, and at intervals of 4 months thereafter, and for patients with SMA of other types to receive subsequent doses at 4 and 12 weeks, and then every 6 months thereafter. The approved nusinersen dose depends on the patient's age and is 9.6 mg in 4 mL in infants aged 0-90 days; 10.3 mg in 4.3 mL in those aged 91-180 days; 10.8 mg in 4.5 mL in those aged 181-365 days; 11.3 mg in 4.7 mL in children aged 1 to 2 years, and 12 mg in 5 mL in those aged over 2 years. The prescribing information specifies that the dosage in preterm infants be adjusted according to their gestational age.

Biogen safety database
Biogen safety database is the central pharmacovigilance repository for safety reports collected from any sources on SMA patients treated with nusinersen in Japan. The procedure for collecting safety reports followed Good Pharmacovigilance Practice (GVP) regulations. This report presents the safety reports recorded in the safety database between August 2017 and June 2019 (23 months, cutoff dated 30-Jun-2019).

Statistical analysis
Descriptive statistics were used to determine frequencies and percentages in the cohorts.

Patients in the PMSC
Three hundred and fifty-three patients were registered in our PMS between August 2017 and 30th May 2019. Among those patients registered, 271 cases had confirmed data suitable for analysis. All 271 patients were included in the safety analysis and effectiveness analysis sets: 94 (35%) with SMA type I, 113 (42%) with type II, 63 (23%) with type III, and 1 (0.4%) with type IV (Table 1). One hundred and fifty-two patients (56%) were male; 236 of 271 patients (87%) were aged >24 months at first nusinersen injection (84 [31%] were aged between 16 and 64 years). forty patients (15%) had a family member with SMA; the family member was a sibling in 38/40 cases. Comorbidities of any type were present in 114 patients (42%). One patient had renal dysfunction (mild) and nine had hepatic dysfunction (mild in eight and moderate in one). Other treatments patients were receiving for SMA are shown in Supplementary Table S1.

Safety in the PMS
Overall, 67 patients (24.7%) in the safety analysis set developed one or more AEs during treatment with nusinersen. There were 21 cases (7.8%) of "infections and infestations" ( Table 3). The overall incidence of AEs was 26.6% in patients with infantile-onset SMA (type I) and 23.7% in patients with later-onset SMA (types II-IV) ( Table 4). SAEs occurred in 23 patients (8.5%) in the safety analysis set during the reporting period (Table 3). By type of SAEs, there were 10 cases (3.7%) of "infections and infestations", six cases (2.2%) of "respiratory, thoracic and mediastinal disorders", three cases (1.1%) of "nervous system disorders" and "injury, poisoning, and procedural complications", two cases (0.7%) of   "gastrointestinal disorders", and one case (0.4%) each of "hepatobiliary disorders", "musculoskeletal and connective tissue disorders", "congenital, familial and genetic disorders", and "investigations".

Safety in the biogen safety database
A total of 345 AEs were recorded in the Biogen safety database ( Table 5). The most common AEs were pyrexia (34 events), pneumonia (30 events) and headache (25 events). Although most AEs recorded in the safety database were non-serious, some SAEs were reported. These SAEs were mostly infections (39 SAEs), respiratory disorders (30 SAEs), gastrointestinal disorders (8 SAEs) and nervous system disorders (7 SAEs). One of these serious nervous system events was hydrocephalus, which developed in a 3-month-old infant after receiving nusinersen. The infant underwent ventriculo-peritoneal shunt (VPS) placement and nusinersen was discontinued, but the hydrocephalus had not resolved at the time of reporting.

Effectiveness in the PMS
During the PMS interim analysis period, there were no deaths in the effectiveness analysis set. Permanent assisted ventilatory support was initiated in two patients with Type I SMA; in these patients, the time to permanent assisted ventilation from treatment initiation was 7 and 9 days, respectively.      atreatment considered "effective" if any one of assessments b-e was considered effective. b treatment considered "effective" if the number of endpoints that increased by ≥1 point (or increased by ≥2 points or to the highest score for "kick") was greater than the number that decreased by ≥1 point (or decreased by ≥2 points or to the lowest score for "kick"). c treatment considered "effective" if motor milestones at baseline other than "consciousness during assessment" and "excitability during assessment" were maintained and ≥1 new milestone was achieved. dtreatment considered "effective" if the total score of all endpoints improved by ≥3 points. e treatment considered "effective" if there was improvement by ≥30 m. f Degree of improvement from previous nusinersen administration was evaluated by physicians on a 7-point scale (marked improvement = +3, moderate improvement = +2, mild improvement = +1, unchanged = 0, mild deterioration = -1, moderate deterioration = -2, and marked deterioration = -3). g treatment considered "effective" if the degree of improvement was "unchanged" or better at least once during the survey period. h treatment considered "effective" if the total score of all assessments was ≥0. HIne, Hammersmith Infant neurological examination; sMa, spinal muscular atrophy; WHo, World Health organization.
Motor function assessment was performed in 172 patients in the effectiveness analysis set. Of these patients, nusinersen was considered to be effective based on improvements in at least one of the motor function assessments in 45 patients (26.2%; Table 6).
CGI-I was assessed in 268 patients in the effectiveness analysis set. In all patients on nusinersen, the degree of improvement was "unchanged" or better than "unchanged" at least once during the survey period (i.e. met the first definition of 'effective'). The total score of at each assessments was ≥0 (i.e. met the second definition of 'effective') in 267 of 268 patients (99.6%; Table 6).
In the analysis of HINE data in patient subgroups based on SMA subtype, nusinersen was found to be effective in 8/55 patients with SMA type I (14.5%), 11/36 with type II (30.6%) and 3/9 with type III (33.3%) ( Table 6).

Discussion
This report presents safety and effectiveness data for nusinersen from 271 Japanese patients with SMA types I-IV who were enrolled in this PMS between August 2017 and May 30, 2019 and safety data from the Japanese Biogen safety database that records all reported (identified, notified) potentially adverse safety events. The PMS and the Biogen safety database are effective in collecting safety data in this vulnerable Data derived from the safety information notice of sPInraZa intrathecal injection 12 mg that was disclosed to prescribers (https://smaone.togetherinsma.jp/ja-jp/home/treatment-support/spinraza-safety.html. [in Japanese]). aes are coded according to MedDra/J version (22.0). ae, adverse event; csf, cerebrospinal fluid; ecG, electrocardiogram; soc, system organ class. patient group, including the identification of rare but potentially serious AEs (e.g. hydrocephalus). While the PMS provides an overview of the incidence of AEs occurring during clinical practice, the Biogen safety database collects all AEs possibly associated with nusinersen (the spectrum of AEs) in terms of the number of AEs recorded. Such pharmacovigilance databases are not designed to calculate AE incidence, because they are subject to selection bias and do not record the denominator of total treated patients, but rather to identify patterns of reporting that may identify rare events and explore their relationship to nusinersen [23,24]. In the PMS, AEs occurred in 67 patients and SAEs occurred in 23 patients; the most common AE was pyrexia. Our findings are generally similar to those of the ENDEAR and CHERISH studies [12,14], and integrated analyses of safety data from the clinical trials of nusinersen in patients with infantile-onset and later-onset SMA [25,26]. The most recent (2019) integrated safety analysis includes data from seven nusinersen trials in infants and children with SMA and has demonstrated a positive benefit-risk profile [25]. The type and incidence of AEs during nusinersen treatment were consistent with the natural history of SMA itself or with the lumbar puncture procedure [25]. A lower incidence of AEs was reported in our PMS compared with the randomised studies [12,13], but this is not unexpected since the period of observation in this PMS report is relatively short compared to clinical studies and studies with a protocol-defined schedule for AE assessment tend to identify more AEs compared with observational studies in a real-world population [27]. Importantly, the most common types of AEs in the clinical studies and in both the PMS and the Biogen safety database were generally similar. The integrated safety analyses from the clinical trials with nusinersen found that most AEs were not related or unlikely to be related to the treatment [25,26], and the most common AEs occurring during nusinersen treatment were respiratory infections and lumbar puncture reactions, such as vomiting and (in those with later-onset SMA) headache [25,26]. Our study also found that respiratory infection or upper respiratory tract inflammation were common. Nusinersen is delivered via lumbar puncture and intrathecal injection, and the frequency of AEs related to lumbar puncture in SMA patients receiving nusinersen was similar to that previously reported in patients undergoing lumbar puncture for other indications [28]. Only one patient in our PMS analysis developed vomiting, which sometimes occurs with lumbar puncture [29], but other lumbar puncture-related events were reported, including cerebrospinal fluid (CSf) leakage in three patients. In the safety database, four events of CSf leakage and five events of vomiting were reported. Cases of meningitis associated with the lumbar puncture procedure, hydrocephalus, and hypersensitivity have been reported during the use of nusinersen in the real-world clinical setting [30], including one case of hydrocephalus in an infant recorded in the current analysis of safety data. The occurrence of hydrocephalus in untreated children with SMA is not well understood.
Communicating hydrocephalus not related to meningitis or bleeding has been reported previously in patients, including children, treated with nusinersen in the post-market setting (Biogen, data on file). Some of these patients, including the infant reported here, have been managed with placement of a VPS. During VPS surgery, a catheter is inserted into the ventricle and the CSf is diverted to the peritoneal space for absorption. All reported cases of communicating hydrocephalus in patients treated with nusinersen occurred in the real-world setting, and individual treatment decisions were managed by treating physicians. However, aside from one published case report, in which the patient was effectively treated with nusinersen several years after undergoing VPS placement [31], there is limited information on the continued effectiveness of nusinersen after VPS placement. Physicians should therefore closely monitor and assess patients who continue to receive nusinersen following placement of a VPS, and the patients and/or their caregivers should be informed that the risks and benefits of nusinersen treatment in patients with VPS are unknown.
Outside Japan, serious infections associated with lumbar puncture and aseptic meningitis have been classified as identified risks (Biogen, data on file). To address this, the prescribing information for nusinersen has been updated to include risk minimisation measures, including a recommendation for the use of an aseptic technique. In Japan, hydrocephalus has been identified as an important risk associated with nusinersen treatment, based on the patient described here, and therefore the potential risk of hydrocephalus continues to be monitored through routine pharmacovigilance practices in Japan and internationally.
A risk of hypersensitivity was also newly identified during the reporting period from post-marketing reports outside Japan. The relationship between hypersensitivity and nusinersen treatment could not be ruled out based on available data. In Japan, hypersensitivity was also already recognised as an important potential risk and listed as a potential AE in the nusinersen product information at the time of initial approval, so no further action was required during the reporting period. There were no reports of hypersensitivity to nusinersen in the current interim analysis period of the PMS.
There are emerging published real-world data on nusinersen treatment from other countries, including Germany [32,33], Italy [34,35], the UK and Ireland [36] showing a favourable efficacy and safety profile. The largest of these analyses included data from 139 adult SMA patients participating in the SMArtCARE registry in Germany [32]. That study recorded significant increases in the mean HfMS score at 6 months, 10 months and 14 months (all p < 0.001) compared with baseline, and 40% of patients had a clinically meaningful improvement in HfMS score at 14 months [32]. The most commonly reported AEs were headache (35%), back pain (22%) and nausea (11%) [32]. A real-world study on nusinersen conducted in the UK and Ireland reported acceptable safety of intrathecal nusinersen in 28 paediatric patients [36]. An observational study was conducted in Italy in 85 children (age 2 months to 15 years) with type I SMA who received nusinersen for 12 months [34]. Nusinersen was associated with a significant improvement in neurological endpoints (such as HINE score), but the authors did not report safety outcomes in these patients [32]. In an Italian retrospective study of 116 adult patients with type II or III SMA, nusinersen was associated with significant improvements in functional assessments (such as expanded Hammersmith functional Motor Scale) in patients with type III SMA, but not in those with type II SMA [35]. This study confirmed the safety of nusinersen administration, with most AEs being related to the lumbar puncture procedure rather than the drug itself [35]. While differences in study design, patient populations and follow-up duration preclude direct comparison between our results and those from other countries, the type of AEs in our study are consistent with these international reports. Moreover, while the treatment duration in our study was shorter than in the other real-world studies, the direction of change in functional assessment parameters in our study was consistent with previous reports from outside of Japan.
In the PMS effectiveness analysis set, there were no deaths and only two patients initiated permanent ventilatory support. furthermore, nusinersen was considered to be effective in 26.2% of patients based on improvements in motor function, and in almost all patients based on the two CGI-I analyses ( Table 6). These effectiveness findings show a tendency towards improvements in motor function over the PMS period.
However, it should be noted that these analyses were not sequential assessments at prespecified timepoints, as in clinical studies, but rather a single analysis undertaken at the physician's discretion. The high rate of improvement in the CGI-I analysis likely reflects a high proportion of patients achieving clinical stabilization, since these analyses included "no change" as a measure of improvement. We cannot conclusively state that this is a treatment effect, since it may represent a period of clinical stability unrelated to nusinersen treatment, particularly in patients with the later-onset subtypes of SMA. However, the findings are supportive of data from the other effectiveness analyses in the current study and from other studies [12,13]. for example, the subgroup analyses of the HINE scale revealed that nusinersen was effective among patients with all subtypes of SMA, with effectiveness ranging from 14.5% (type I) to 33% (type III) ( Table 6). There was a trend in our study for nusinersen to show greater effectiveness in patients with less severe SMA (types II and III) than in those with type I SMA. In the ENDEAR study of nusinersen-treated patients with type I SMA, 51% achieved a motor-milestone response on HINE-2 (vs 0% with sham control; p < 0.001) [12], and in the CHERISH study, nusinersen provided significantly greater improvements from baseline in the expanded HfMS at 15 months compared with placebo (least mean squares difference 5.9 points; p < 0.001) in children with type > I SMA [13]. There could be a number of reasons for the trend towards greater improvement in patients with type II or III SMA in our study. It could have been a chance finding since patient numbers were small and the treatment duration was short (~100 days). It could also be related to the age of nusinersen initiation, since age at the start of nusinersen has been shown to be a significant predictor of functional improvements in patients with SMA across subtypes I to II [37][38][39][40]. The high rate of effectiveness in SMA type I patients in the ENDEAR study compared with our results is likely related to the fact that all patients in ENDEAR were aged <6 months [12] compared with only 15/271 patients (5.5%) in our study. Another potential reason is that HINE scores may reflect a period of clinical stability unrelated to nusinersen treatment in patients with SMA types II or III. further study is needed to confirm the long-term effectiveness of nusinersen in patients with SMA. The current ongoing PMS will continue to evaluate the safety and effectiveness of nusinersen over time; therefore, it is possible that the final data from the PMS will differ from the data in this interim analysis.
Based on market data, the number of patients treated globally with nusinersen from May 31, 2018 to November 30, 2018 was estimated to be 2,101, corresponding to an estimated 175.1 person-years of treatment. A multinational registry of SMA patients called the International SMA Consortium Registry was also recently initiated and is collecting data on the safety and effectiveness of nusinersen [41]. In addition, the SMArtCARE registry is underway in Germany, Austria and france [42], with preliminary data already available [32], and a National SMA Registry is collecting data in Canada [43]. Three interventional clinical trials using nusinersen [ISIS 396443-CS11 (SHINE), 232SM202 (EMBRACE), and 232SM201 (NURTURE)] were also ongoing during the period of May 31, 2018 to November 30, 2018, involving 352 patients who had received at least one dose of nusinersen, and no new safety signals had been identified at the time of preparing this report (Biogen, data on file). Drug use surveys during the reporting period (May 31, 2018 -November 30, 2018) describe seven patients with pneumonia (one event each), and one case (one event) each of headache, hydrocephalus, respiratory failure, nausea, upper gastrointestinal haemorrhage, and lumbar puncture post-paracentesis syndrome (Biogen, data on file).
The nusinersen package inserts throughout the world are being regularly updated with any new AEs confirmed after marketing. It is likely that these newly identified risks of hypersensitivity and hydrocephalus can be adequately addressed by following the instructions in the nusinersen package inserts.
Nusinersen has the potential to address a critical unmet medical need in patients with SMA, for whom treatment options are limited. Nevertheless, it is important with any new treatment to be able to inform patients of the expected risks and benefits, but this can be challenging with treatments for rare diseases, since clinical trial populations and real-world cohorts tend to be small. The current study adds to the cumulative treatment experience with nusinersen, and indicates that the safety of nusinersen in Japanese patients is consistent with international data, with the most frequent AEs being infections, headache and pyrexia. Although short-term, the effectiveness data indicate benefits that are also consistent with international data. This information will help Japanese physicians to have informed discussions with their patients about the benefit-risk profile of nusinersen in the treatment of SMA.

Limitations
Limitations inherent to real-world studies apply to our PMS, namely the lack of a comparator group and no blinding. This analysis includes data from 271 patients which may not be sufficient to identify new safety signals, and ongoing surveillance is necessary. The key reason for the small sample size is that SMA is a comparatively rare disease. Ongoing surveillance in this PMS will collect additional information in a larger cohort of patients and over a longer time frame. As a PMS, we did not record the likely relationship of AEs to treatment, or the relationship between AEs and treatment discontinuation. In addition, our observational study design did not require effectiveness analyses to be undertaken at prespecified time intervals. The CGI-I scale that was used in the PMS effective analysis has not been validated in SMA and we did not assess inter-rater variability. furthermore, to our knowledge, the two dichotomized definitions of overall improvement using the CGI-I scale have not been validated.

Conclusion
This Japanese PMS and Biogen safety database analysis demonstrated that nusinersen has an acceptable safety profile, and the AEs and SAEs seen in these patients are consistent with events typically seen in patients with SMA and/or those undergoing lumbar puncture. In addition, the analysis of effectiveness, while short-term, show benefits on functional outcomes that are similar to those seen in previous studies. Overall, the benefit-risk profile of nusinersen treatment in patients with SMA remains favourable, but ongoing pharmacovigilance and real-world research will continue to collect data on the safety and effectiveness of nusinersen in patients with SMA.