A case series of paediatric patients with spinal muscular atrophy type I undergoing scoliosis correction surgery

Summary

Spinal muscular atrophy is a neuromuscular disorder with degeneration of spinal motor neurons. Type I is a severe variant that was recently shown to be amenable to treatment with the antisense oligonucleotide nusinersen. As a result of increased life expectancy with this treatment, more children with spinal muscular atrophy type I are presenting for spinal correction surgery. In this case series, we present four such patients who underwent spinal surgery at our institution over the course of one year. Pre‐operative assessment showed evidence of reduced respiratory function requiring nocturnal non‐invasive ventilation in all four patients. A difficult airway was encountered in two of the four patients. Postoperative complications were ubiquitous and included CSF leak, poor wound healing, metal frame exposure, frame instability and wound infection. There were no postoperative respiratory complications and all four children returned to their respiratory baseline postoperatively. All patients underwent successful lumbar puncture and intrathecal nusinersen injection following their spinal surgeries. Given the risk of complications and prolonged recovery following spinal surgery, a detailed family discussion is advisable.

Introduction

Spinal muscular atrophy (SMA) is a neuromuscular disorder that causes progressive muscular weakness. The condition is caused by homozygous deletion or mutation of the survival motor neuron (SMN)1 gene. It has an incidence of 1 in 11,000 live births [1]. The SMN1 gene produces SMN protein, which is required for the upkeep of motor neurons. A paralogous gene, SMN2, can also produce SMN protein but 95% of the translated protein is non‐functional due to aberrant splicing [2]. Nusinersen, an antisense oligonucleotide drug administered by the intrathecal route, targets the pre‐mRNA splicing of SMN2 and results in increased production of functional SMN protein.

Spinal muscular atrophy is classified from SMA type 0 to IV according to the age of onset, with SMA 0 being prenatal and SMA IV adult onset. Spinal muscular atrophy type I, also known as Werdnig‐Hoffmann disease, accounts for 60% of all cases. It is diagnosed between birth and six months of age. Patients exhibit weakness with failure to meet motor milestones such as being unable to sit unsupported. The median life expectancy before nusinersen was 2 y [3].

Nusinersen has been shown to improve muscle control and survival rates in patients with SMA I [2]. As a result, these patients now live longer and present with complications not previously seen in this cohort, including kyphoscoliosis. There remains a lack of published literature concerning SMA I and kyphoscoliosis correction surgery. In this case series, we present four patients with SMA I who underwent corrective spinal surgery at our institution.

Report

Four patients (designated A–D for the purpose of this report) with SMA I underwent kyphoscoliosis correction between July 2020 and June 2021. Pre‐, intra‐ and postoperative characteristics are presented in Table 1. The airway management and anaesthesia technique were at the discretion of the anaesthetist caring for the patient. Rod positioning was carefully planned to facilitate ongoing intrathecal nusinersen injection. All patients underwent successful postoperative nusinersen injection with fluoroscopy and Entonox® for analgesia and sedation.

Table 1.

Peri‐operative characteristics.

		Patient A	Patient B	Patient C	Patient D
Pre‐operative	Age (years)	3	4	5	7
	Weight (kg)	15	13	18	16
	Respiratory Function	Nocturnal BiPAP a 14/6 1x LRTI b in last year	Nocturnal BiPAP 16/6 1x LRTI last year	Nocturnal BiPAP 20/8 3x LRTI in last year	Nocturnal BiPAP 10/4 Recurrent lobar collapse
	Physiotherapy	Cough assist, Home nebuliser	Cough assist, Home nebuliser	Cough assist, Home nebuliser	Cough assist, Home nebuliser
	Previous PICU admission	Yes, LRTI	No	Yes, LRTI	Yes, lobar lung collapse
	Cardiac Function	TTE c : good function, No evidence of PHTN d	TTE: good function, No evidence of PHTN	TTE: good function, No evidence of PHTN	TTE: good function, No evidence of PHTN
	Gastrointestinal Function	Percutaneous Endoscopic Gastrostomy fed	Percutaneous Endoscopic Gastrostomy fed	Percutaneous Endoscopic Gastrostomy fed	Percutaneous Endoscopic Gastrostomy fed
	Latest CHOP e score	52/64	39/64	42/64	50/64
Intra‐operative	Induction of Anaesthesia	Inhalational with Sevoflurane/Air/Oxygen	Intravenous induction with Propofol	Inhalational with Sevoflurane/Air/Oxygen	Inhalational with Sevoflurane/Air/Oxygen
	Maintenance of Anaesthesia	Propofol Targeted control infusion (4–4.5 μg.ml‐1) and Remifentanil infusion 0–0.15 μg.kg‐1.min‐1)	Sevoflurane/Air/Oxygen	Sevoflurane/Air/Oxygen	Propofol Targeted control infusion (3–4.5 μg.ml‐1) and Remifentanil infusion 0–0.3 μg.kg‐1.min‐1)
	Tracheal tube used	Cuffed 4.5 mm	Cuffed 4.5 mm	Cuffed 5.5 mm	Cuffed 5.0 mm
	Laryngoscope used	C‐MAC® Macintosh size 1 (Karl Storz, Tuttlingen, Germany)	Glidoscope® size 2.5 (Verathon, Bothell, USA)	Glidoscope® size 2.5 (Verathon, Bothell, USA)	Macgrath™ size 2 (Medtronic, Minneapolis, USA)
Postoperative	PICU length of stay (days)	1	1	1	2
	Postoperative analgesia	Paracetamol, NSAID f , Morphine NCA g	Paracetamol, NSAID, Morphine NCA	Paracetamol, NSAID, Morphine NCA	Paracetamol, NSAID, Morphine NCA
	Discharge Home	Postoperative Day 7	Postoperative Day 6	Postoperative Day 11	Postoperative Day 7

^a Bilevel positive airway pressure.

^b Lower respiratory tract infection.

^c Transthoracic echocardiography.

^d Pulmonary Hypertension.

^e Children’s Hospital of Philadelphia Score.

^f Non‐steroid Anti‐inflammatory drug.

^g Nurse Controlled Analgesia. The Children’s Hospital of Philadelphia (CHOP) score reflects motor function at the time of assessment.

Patient A had a left growing rod inserted. There was no anticipated airway difficulty and, following an uneventful inhalational induction of anaesthesia and institution of intravenous access, tracheal intubation was performed with videolaryngoscopy (C‐MAC® Macintosh size 1, Karl Storz, Tuttlingen, Germany). Intra‐operatively, there was difficulty with lung ventilation which was relieved by endobronchial suctioning of respiratory secretions. Postoperatively, tracheal extubation occurred after two hours and non‐invasive ventilation (NIV) was initiated at the patient’s usual pressure settings. Oxygen supplementation was weaned from 40% to air before discharge to the ward. Two weeks following home discharge, the patient represented to hospital with a lumbar CSF leak. This required surgical revision with a three week hospitalisation. Ten months later, a right‐sided spinal growing rod was inserted with an uncomplicated peri‐operative course.

Patient B had bilateral spinal growing rods inserted. The patient had a history of previous difficult tracheal intubation. An intravenous induction of anaesthesia was performed with propofol. Rocuronium 1 mg.kg^‐1 was administered to facilitate airway intervention. Ventilation with a facemask was uneventful. Tracheal intubation using a videolaryngoscope with a hyperangulated blade (Glidoscope® size 2.5, Verathon, Bothell, USA) was successful on the first attempt. The patient required a 10 ml.kg^‐1 packed red cell transfusion intra‐operatively based on clinical status and a transfusion threshold of 8 g.dL^‐1. Postoperatively, tracheal extubation to baseline NIV settings occurred after three hours. The postoperative course was complicated with metal frame exposure, pain and spinal instability requiring three surgical revisions. Each operation necessitated a one day postoperative stay in the paediatric intensive care unit (PICU).

Patient C had insertion of a spinal rod construction. A previous history of difficult tracheal intubation with direct laryngoscopy but uneventful facemask ventilation was recorded. An inhalational induction was performed due to difficulty acquiring intravenous access. A videolaryngoscope with a hyperangulated blade (Glidoscope® size 2.5, Verathon, Bothell, USA) was used successfully to perform tracheal intubation on the first attempt. The intra‐operative course was uneventful and tracheal extubation to baseline NIV settings occurred 15 min following PICU admission. On the fifth postoperative day, a surgical revision for frame displacement was required. Intra‐operative findings included reduced bone condition and investigation of osteoporosis is ongoing. A total of six surgical revisions with resultant inpatient admissions occurred over a duration of five months following wound dehiscence, metal exposure and wound infection with Klebsiella oxytoca.

Patient D underwent bilateral spinal rod insertion and bilateral Achilles’ tendon release. There was no anticipated airway difficulty. An uneventful inhalational induction of anaesthesia with intravenous access institution was followed by tracheal intubation with videolaryngoscopy (Macgrath™ size 2 Medtronic, Minneapolis, USA). The intra‐operative course was uneventful. Postoperatively, tracheal extubation to NIV occurred before transfer to PICU. An increase in NIV support (10/4 cmH₂O–10/6 cmH₂O) was required for 24 h before returning to baseline. Following discharge home, the patient represented a month later with spinal rod protrusion and skin breakdown. A revision was performed with rod exchange. The patient was discharged home after eight days, following revision surgery with ongoing microbiological input.

Discussion

With the development of nusinersen, patients with SMA I are living longer and are presenting for procedures that would have been previously considered only in the SMA II and III cohorts. A pre‐nusinersen 27‐year retrospective review of all patients with SMA showed only one spinal instrumentation among 20 procedures conducted on patients with SMA I [4]. In one year our institution has recorded four patients undergoing a total of 15 separate spinal surgeries.

The anaesthesia principles when caring for patients with neuromuscular disease apply to the SMA I cohort. This includes a careful pre‐operative evaluation with particular emphasis on cardiorespiratory function when undergoing spinal correction procedures. Succinylcholine is avoided due to the potential risk of hyperkalaemia [3]. Inhalational and intravenous forms of general anaesthesia have both been safely used in patients with SMA I [4], as reflected in our case series. As with other spinal correction procedures, blood loss remains a possibility. Attentive postoperative care is required as complication rates (respiratory failure, surgical wound infection, urinary tract infection) are higher with neuromuscular disease‐related scoliosis repair compared with idiopathic scoliosis repair [5].

Difficult airway management has been reported in patients with SMA [4, 6]. Two out of the four patients in this series (B and C) had a history of difficult tracheal intubation with direct laryngoscopy. This difficulty was overcome in both cases with the use of hyperangulated videolaryngoscopy as the first‐line approach. In both patients, facemask oxygenation and ventilation remained possible. A proposed mechanism for the difficult airway in children with SMA is atrophy of the masseter and other muscles of mastication [4]. A further explanation could be due to reduced neck movement secondary to contractures. It has also been noted that as patients with SMA get older, mouth opening reduces [3]. A further consideration is the reduced bulbar function seen in this cohort. Pooling of oral secretions may blur the glottic view, particularly during videolaryngoscopy.

The respiratory function of patients with SMA I needs careful peri‐operative evaluation as these patients will often present with baseline ventilatory support requirements [3]. However, pulmonary function tests are limited by patient age and cooperation. All four patients in this case series required non‐invasive ventilatory support at night. All four had also required hospitalisation in the prior year for respiratory issues. Of those admissions, three out of the four patients required PICU care. Intra‐operative ventilatory challenges can be encountered, as seen with patient A. In this instance, difficult mechanical ventilation was relieved following endobronchial suctioning of respiratory secretions. Postoperative care can also be difficult with the risk of muscular weakening following prolonged mechanical ventilation, though we did not witness any postoperative pulmonary complications in our series. All four patients had a successful tracheal extubation and returned to their baseline NIV settings within 24 h. All were monitored in a PICU setting immediately postoperatively, and early physiotherapy involvement may have helped prevent respiratory compromise. This observation may however be coincidental, reflective of the small number of patients in this case series.

Cardiac manifestations of SMA I include congenital cardiac defects (hypoplastic left heart and septal defects) and autonomic dysfunction in older patients [3]. The presence of scoliosis is associated with restrictive lung disease and pulmonary hypertension. All four patients underwent cardiac evaluation with transthoracic echocardiography review pre‐operatively. We did not detect any evidence of any cardiac structural defect nor any pulmonary hypertension.

All four patients in our case series unfortunately had postoperative surgical complications requiring further intervention. Postoperative complications are not uncommon [7]. There is increasing evidence that patients with SMA have underlying abnormal bone density [8]. Bulbar dysfunction leads to feeding difficulties with the risk of nutritional deficiencies including vitamin D. Bone strength is further diminished by the lower muscle mass and reduced weight bearing function. This was highlighted with patient C with slipping of the bone anchoring sites and the requirement for multiple revisions. Endocrinology input needs to be considered pre‐operatively if there is any concern for osteopenia.

Another consideration is the requirement of maintenance nusinersen injection following scoliosis correction surgery. The placement of spinal rods distorts the anatomy and may make lumbar puncture more challenging. The aid of interventional radiology to facilitate delivery has previously been documented [9], and fluoroscopy successfully facilitated this procedure in all four patients. Other methods include the use of computed tomography guidance or the use of an intrathecal reservoir.

Another novel therapy gaining traction includes onasemnogene abeparvovec. This is a single‐dose gene therapy treatment that has shown improved motor and respiratory function 18 months post‐injection [10]. Whether this has any effect on scoliosis progression has yet to be seen.

As more children with SMA I receive nusinersen treatment, more will present for scoliosis correction surgery. With the risk of a prolonged recovery, surgical and anaesthetic complications, a detailed discussion regarding risks and benefits must be had with the guardian before undertaking such procedures.