Historically, autosomal recessive 5q-linked spinal muscular atrophy (SMA) has been the leading inherited cause of infant death. SMA is caused by the absence of the SMN1 gene, and SMN1 gene replacement therapy, onasemnogene abeparvovec-xioi, was Food and Drug Administration approved in May 2019. Approval included all children with SMA age <2 years without end-stage weakness. However, gene transfer with onasemnogene abeparvovec-xioi has been only studied in children age ≤8 months.
In this article, we report key safety and early outcome data from the first 21 children (age 1–23 months) treated in the state of Ohio.
In children ≤6 months, gene transfer was well tolerated. In this young group, serum transaminase (aspartate aminotransferase and alanine aminotransferase) elevations were modest and not associated with γ glutamyl transpeptidase elevations. Initial prednisolone administration matched that given in the clinical trials. In older children, elevations in aspartate aminotransferase, alanine aminotransferase and γ glutamyl transpeptidase were more common and required a higher dose of prednisolone, but all were without clinical symptoms. Nineteen of 21 (90%) children experienced an asymptomatic drop in platelets in the first week after treatment that recovered without intervention. Of the 19 children with repeated outcome assessments, 11% (n = 2) experienced stabilization and 89% (n = 17) experienced improvement in motor function.
In this population, with thorough screening and careful post–gene transfer management, replacement therapy with onasemnogene abeparvovec-xioi is safe and shows promise for early efficacy.
Spinal muscular atrophy is an inherited neurodegenerative neuromuscular disease. There are now two safe and highly effective Food and Drug Administration–approved genetic therapies. However, the Food and Drug Administration indications were much broader than those assessed in clinical trials.
In this retrospective review, we report the safety and early efficacy of onasemnogene abeparvovec-xioi in all spinal muscular atrophy patients aged <2 years within Ohio regardless of phenotype.
5q-linked spinal muscular atrophy (SMA) is an autosomal recessive degenerative neuromuscular disease caused by the absence of the SMN1 gene and insufficient survival motor neuron protein.1,2 Both severity of disease and age of onset are predicted by the copy number of a second gene, SMN2, which is also inherited in an autosomal recessive fashion.3 In its most common and severe form, type 1, children either die or are dependent on mechanical ventilation by 2 years of age.1,4 However, the approval of 2 therapies since late 2016 has markedly changed the course of this disease.5,6
One of these therapies uses an adeno-associated virus serotype 9 (AAV9) vector to replace the missing SMN1 gene. Researchers of initial studies in murine models demonstrated SMN protein expression in motor neurons and peripheral tissues and an extended life span.7–10 Researchers of additional studies in a porcine model showed similar efficacy.11 After the successful SMA phenotype reversal in these models, a human trial in 15 children with type 1 SMA <8 months of age demonstrated safety and efficacy.6 Subsequently, SMN1 gene replacement therapy, onasemnogene abeparvovec-xioi, was Food and Drug Administration approved on May 24, 2019, for any patient age <2 years without end-stage disease. Thus, the safety and important features of drug administration in older (and likely heavier) individuals has not been well studied or reported.
In this article, we review the Ohio experience of treating an additional 21 children with a mean age of 10 ± 7 months. We provide key safety laboratory values and early outcomes for all individuals who have been dosed after the original research trial. These include children who received AVXS-101 gene transfer via a free managed and expanded access program (before commercial approval) or onasemnogene abeparvovec-xioi (commercially approved product) between December 2018 and February 2020 and who have completed the prednisolone course for immunosuppression.
Children were included from four children’s hospitals in Ohio: Nationwide Children’s Hospital, Rainbow Babies and Children’s Hospital, Cincinnati Children’s Hospital, and Akron Children’s Hospital. All individuals were approved for inclusion in this report by each institution’s Institutional Review Board. Anyone who received either AVXS-101 through the managed access program or onasemnogene abeparvovec-xioi commercially and completed their prednisolone course and taper by April 30, 2020, were included. Children were identified in two ways. The first group was presymptomatic and discovered after confirmation of a positive newborn screen result (n = 5). The second genetically confirmed group were symptomatic children without end-stage disease (n = 16).
To determine eligibility for treatment, every child who met the Food and Drug Administration criteria had screening laboratory studies done, including complete blood count with platelets, comprehensive metabolic panel, γ glutamyl transpeptidase (GGT), prothrombin, troponin I, HIV, hepatitis B and C screening, and AAV9 antibody status (<1:50 required). If the results of the above studies were within normal range, children were eligible for treatment. An additional stipulation at Nationwide Children’s Hospital was implemented after one managed access patient developed clinically significant thrombocytopenia (platelets <50 K/µL) on day 7 after gene transfer. This child had received gene transfer 1 month after a nusinersen dose, an antisense oligonucleotide with known potential to lower platelets (package insert, Biogen), and was noted to have a concurrent parainfluenza 3 infection. Subsequent candidates were required to wait 3 months after last nusinersen dosing before receiving gene therapy. On the basis of this occurrence, we believe there is a theoretical risk of thrombocytopenia and potentially other unknown complications when switching from nusinersen to onasemnogene abeparvovec-xioi and believed a longer waiting time between last dose of nusinersen and the switch may minimize this risk. Children could not be ill at the time of gene transfer. One child recovering from an acute illness was treated as an inpatient. If a child had been ill before gene transfer, at least 9 days were required from resolution of illness before dosing. At that time, repeat laboratory values, including AAV9 antibodies, were obtained and were normal before proceeding with dosing. Prednisolone (1 mg/kg per day) was started the day before gene transfer to suppress response to the AAV.6 Children were dosed either as “outpatient in a bed” or the infusion center and were discharged ∼4 hours after the infusion (except for one child who was treated while inpatient). Follow-up laboratory studies included platelets, aspartate aminotransferase (AST), alanine aminotransferase (ALT), GGT, prothrombin, bilirubin, and troponin I and were obtained weekly for ≥4 weeks. If at any point the AST, ALT, or GGT increased to >2 times normal, the prednisolone dose was increased to 2 mg/kg per day. No child required an increase beyond 2 mg/kg per day prednisolone dosing. Although the AST, ALT, and GGT did continue to rise in some children, the absence of an increase in bilirubin, prothrombin time, or symptoms led to the decision to keep the dose at 2 mg/kg per day in these children. The 2 mg/kg per day dose was maintained until the transaminases returned to ∼2 times the upper limit of normal and was then decreased slowly while close clinical and laboratory monitoring continued.
Twenty-one children are included in Table 1, providing age at treatment, SMN2 copy number, previous nusinersen use, laboratory values, and prednisone course. One patient was dosed under a single-patient Investigational New Drug application, 3 patients were dosed via the managed access program (2 sites), and the remaining 17 were dosed commercially (3 sites). Seven of the 9 children who were ≤6 months of age had no noteworthy elevation in AST, ALT, or GGT (Figs 1–3) after gene transfer. Five of these children were identified from newborn screening and were considered presymptomatic. Two of these 5 children had elevations in AST and ALT with or without a concurrent elevation in GGT. One had been treated initially with nusinersen and then transitioned to onasemnogene abeparvovec-xioi. In both cases, these elevations were related to suboptimal prednisolone administration. One child had difficulty taking prednisolone and better tolerated dissolvable prednisone tablets to maintain immunosuppression. The second had noncompliance, with abrupt discontinuations in prednisolone dosing on two occasions interrupting the needed immunosuppression. Overall, gene transfer was well tolerated in those 6 months of age or younger with the ability to ensure appropriate prednisolone administration as the key safety factor.
In the older and heavier children, the liver impact as assessed by transaminases was greater. Of the 12 children 8 months of age or older, 8 (67%) had elevations of ≥2 times the upper limit of normal in AST and/or ALT, and of these, 6 children also had elevations in GGT >1 times normal. Of the 12 children who weighed ≥8 kg, 10 (83%) had elevations of ≥2 times the upper limit of normal in AST and/or ALT, and of these, 4 children also had elevations in GGT >1 times normal. However, all patients remained clinically well and had no abnormalities indicative of symptomatic liver dysfunction (data not shown).
Regardless of age or weight, a decline in platelet count was present in nearly all patients at the time of the first posttransfer laboratory assessment (day 7 after gene transfer). Nineteen children (90%) had a decline in platelet count on day 7 after gene transfer, with 73% (n = 14) <200 K/µL and 16% (n = 3) <100 K/µL. None were symptomatic or required treatment and all returned to a value of ≥150 K/µL by day 14, and subsequent platelet counts remained normal.
Nineteen children completed at least two functional assessments, and all had stabilization or improvement in functional outcomes. Seventeen (89%) had objective improvements of at least 1 point in functional outcome scores by 4 months (Fig 4, Table 2). Twelve of these (70%) demonstrated a ≥3 point improvement by 4 months (mean: 8.1 ± 4.7, range: 3–20 points).
Sixteen of the 17 children who were feeding orally before gene transfer maintained this ability after gene transfer. One child with SMA type 1, dosed at age 5 months, was symptomatic before treatment and developed a significant respiratory illness 2 weeks after gene transfer. She subsequently required a tracheostomy and gastric feeding tube. Another 2 children with SMA, dosed at age 4.5 and 8 months, were unable to feed by mouth before gene transfer and are now able to take some purees by mouth. Of the 4 children who were feeding partially by mouth before gene transfer, all have been able to continue to make progress in oral feeding since gene transfer, but none are able to exclusively feed orally at this time.
Eight children were on bilevel positive airway pressure (BiPAP) at night before gene transfer. After gene transfer, 7 remain on BiPAP, 1 has transitioned to intermittent use only when ill or after a particularly tiring day, and 1 child no longer requires BiPAP. One child, as noted above, was not receiving any breathing support before gene transfer but required ventilator support after gene transfer because of a severe respiratory illness.
Onasemnogene abeparvovec-xioi was well tolerated in all children treated in the state of Ohio. Our experience suggests that when a thorough screening process is completed, social isolation is implemented to minimize the risk of illness after gene transfer, and patients are monitored closely in the weeks to months after gene transfer, adverse events are few. We do recommend delaying gene transfer until complete resolution of preexisting viral illnesses. We also recommend a delay in live vaccinations until 4 weeks after the prednisolone course and taper have been completed.
In this Ohio population, there were no clinically significant adverse events after gene transfer. However, just over half of our patients required a prolonged prednisolone course compared with none of the 15 children in the phase 1 study.6 In children ≤6 months of age, the key to successful post–gene transfer clinical course is dependent on prednisolone administration. These young children are often still feeding by mouth and may resist taking this medication. Care must be taken to ensure that the prescribed dosage is received and given properly, with flexibility in delivery method (oral liquid, crushed tablets, dissolvable tablets) as required. The majority of children who required a prolonged prednisolone course were in the older age group, usually >8 months and/or >8 kg. These children had AST and ALT elevations with or without increases in GGT. The transaminase elevation may be due to total viral load received, as would be encountered in larger patients. However, as there were some children in this group without significant elevations, it may also be related to differing immune responses to AAV9. We conclude that regular post–gene transfer laboratory monitoring and clinical assessments are critical while taking prednisolone and until the course has been completed. It is also important to point out that all immune responses to AAV were controlled with prednisolone alone and did not require more aggressive immune suppression thought necessary by some investigators on the basis of preclinical studies,13 regardless of patient age or weight at dosing.
Interestingly, almost all children had a drop in platelet count on day 7 after gene transfer. This is likely complement mediated given the timing. Only one child had a clinically apparent concomitant viral infection. All of the children (n = 3) with a platelet drop below 50 K/µL had been transitioned from nusinersen therapy. However, platelet count drops did also occur in nusinersen-naive children, suggesting that nusinersen is not the sole contributing factor but may confer greater risk for a more significant decline. Nevertheless, no child had any clinical symptoms of thrombocytopenia, nor required additional treatment, and all platelet count normalized on day 14. All children should be monitored closely for thrombocytopenia after gene transfer, and extra consideration should be given for a child with lower platelets before gene transfer.
Finally, all symptomatic individuals experienced subjective and objective functional improvements in motor function. In the 5 children treated before symptoms, no signs of weakness characteristic of SMA have developed over follow-up periods of 2 to 8 months. Parents consistently reported improvements in motor skills for their children and objective testing by physical therapists showed that 89% have had improvements in their outcome measures. Because of the heterogeneous ages and baseline functional testing at infusion we cannot comment on the effect of age at dosing on expected improvement. We can, however, state that functional improvements were measured across the cohort regardless of age at infusion and previous treatment status. These data are currently limited because of variations in functional assessments and length of follow-up. However, these are promising results for single-dose gene delivery with onasemnogene abeparvovec-xioi in infants 1 to 23 months of age.
Drs Waldrop and Connolly conceptualized and designed the study, drafted the initial manuscript, conducted initial data analyses, and reviewed and revised the manuscript; Ms Karingada, Ms Powers, Dr Iammarino, Ms Miller, and Drs Alfano, Drs Noritz, Rossman, Ginsberg, Mosher, Broomall, Goldstein, Bass, Lowes, and Tsao coordinated and supervised data collection and critically reviewed the manuscript; Dr Mendell conducted data analysis and critically reviewed the manuscript for important intellectual content; and all authors approved the final manuscript as submitted.
FUNDING: No external funding.
POTENTIAL CONFLICT OF INTEREST: Dr Waldrop has served on advisory boards for Sarepta Therapeutics and Avexis Inc. Dr Storey has served on advisory boards for Avexis Inc and Sarepta Therapeutics. Dr Mosher has served on advisory boards for Sarepta and PTC Therapeutics. Drs Goldstein and Bass have served on speaker and advisory boards for Biogen. Dr Alfano has served on advisory boards for Genentech-Roche, Acceleron Pharma, and Audentes Therapeutics. Dr Mendell has received personal fees from Sarepta Therapeutics, Avexis Inc, and Vertex Pharmaceuticals. Dr Connolly has served on advisory boards for Sarepta Therapeutics, Avexis, and Genentech-Roche and serves on the DMSB for Catabasis; the other authors have indicated they have no potential conflicts of interest to disclose.
FINANCIAL DISCLOSURE: Dr Waldrop has served on advisory boards for Sarepta Therapeutics and Avexis Inc. Dr Storey has served on advisory boards for Avexis Inc and Sarepta Therapeutics. Dr Mendell has received personal fees from Sarepta Therapeutics, Avexis Inc, and Vertex Pharmaceuticals. Dr Connolly has served on advisory boards for Sarepta Therapeutics, Avexis, and Genentech-Roche and serves on the Data and Safety Monitoring Board for Catabasis; the other authors have indicated they have no financial relationships relevant to this article to disclose.