Management of Critically Ill Patients With Spinal Muscular Atrophy Admitted With Acute Respiratory Failure

Abstract

Background:

Spinal muscular atrophy (SMA) is a neurodegenerative disease that results in progressive muscular atrophy and weakness. The primary cause of morbidity and mortality in these children is pulmonary disease due to poor airway clearance that leads to acute respiratory failure. There is a paucity of literature on the treatment of children with SMA and acute respiratory failure.

Methods:

We performed a retrospective chart review of pediatric patients with SMA type I or II admitted to the PICU in a tertiary-care children's hospital with acute respiratory failure who required mechanical ventilation and/or aggressive airway clearance.

Results:

The study included 300 unique encounters among 137 unique subjects. All the subjects received airway clearance at home before admission, and 257 encounters (85.7%) were supported with noninvasive ventilation (NIV) before admission. Sixty-eight subjects (49.6%) required endotracheal intubation on admission or at some point during their PICU stay. The median (interquartile range [IQR]) time to intubation was 0 (0–20) h, and the median (IQR) duration of mechanical ventilation was 2 (1–5) d on invasive mechanical ventilation with no statistical difference between type I and type II (P = .89). Of those, 65 (91.1%) were successfully extubated to NIV on the first attempt, whereas 4 subjects (5.8%) required multiple extubation attempts and 3 subjects (4.4%) required subsequent tracheostomy. For the subjects who were intubated, both PICU and hospital length of stay were longer (P < .001) when compared with the subjects managed by NIV alone. The subjects with SMA type I had a longer PICU length of stay, with a median (IQR) of 5 (3–11) d versus 4 (2–7) d (P = .002). The hospital length of stay and duration of invasive mechanical ventilation were not statistically different (P = .055 and P = .068, respectively).

Conclusions:

The subjects with SMA types I or II can be treated successfully with NIV and aggressive airway clearance during acute respiratory failure. Similarly, when intubation is required, successful extubation can be achieved with NIV transitional support combined with aggressive airway clearance maneuvers.

Introduction

Spinal muscular atrophy (SMA) is a neurodegenerative disease that results in progressive muscular atrophy and weakness.¹ Pulmonary involvement is characterized by progressive intercostal muscle weakness that results in ineffective coughing and the inability to clear lower airway secretions.² The International Spinal Muscular Atrophy Consortium has defined type I as the most severe form of SMA, with an age of onset between 0 and 6 months.^3,4 Although there have been significant improvements in the natural history of SMA due to proactive approaches, including the introduction of noninvasive ventilation (NIV) (ie, aggressive airway clearance therapies, gene therapy, and early enteral nutrition), morbidity and mortality in these patients remain high.^3,5,6

During an acute illness, an inability to clear the increased lower airway secretions and the resulting worsening hypoventilation lead to acute-on-chronic respiratory failure and the potential need for endotracheal intubation. Limited studies in adults with neuromuscular disease have shown improved outcomes in subjects treated with NIV and cough assist.⁷ However, there currently is a paucity of literature on treating children with SMA by using NIV with aggressive secretion mobilization and clearance techniques during an acute respiratory illness.^2,7 There is even less literature that describes the use of NIV and airway clearance during acute-on-chronic respiratory failure in these patients or the role of NIV in liberation from invasive ventilation. This study examined and described a single center’s experience with the use of NIV and invasive mechanical ventilation combined with aggressive airway clearance in children with SMA types I or II who were admitted to a PICU with acute respiratory failure over the span of 15 years.

Quick Look.

Current Knowledge

Spinal muscular atrophy (SMA) is a neurodegenerative disease that results in progressive muscular atrophy and weakness. There currently is a paucity of literature on the treatment of children with SMA during an acute respiratory illness.

What This Paper Contributes to Our Knowledge

The majority of children with SMA types I or II admitted to the PICU with acute respiratory failure were treated successfully with noninvasive ventilation and aggressive airway clearance. The subjects who were supported with noninvasive ventilation at home or who had a higher body mass index percentile were at higher risk for endotracheal intubation.

Methods

This was a single-center retrospective study. The institutional review board of the University of Wisconsin-Madison approved the study protocol with a waiver of informed consent. A retrospective chart review was conducted for all pediatric patients with SMA types I or II who were admitted between January 2002 and December 2009 and between January 2010 and December 2018 (data from 2008 were not available). We included only patients in acute respiratory failure who were admitted to the PICU. Subjects were identified by using the International Classification of Diseases -9 or -10 codes, depending on the year of the encounter.

General information collected included subjects’ age, sex, SMA type (I or II), date of PICU admission, PICU length of stay, hospital length of stay, admission diagnosis, etiology of respiratory failure (if known), home NIV settings, and mortality. PICU admission data collected included admission ${\text{F}}_{{\text{IO}}_{\text{2}}}$ requirement, duration of supplemental oxygen requirement, maximum NIV settings, endotracheal intubation (yes or no), time from admission to endotracheal intubation, supplemental oxygen requirement at the time of endotracheal intubation, duration of endotracheal intubation, number of failed extubations, and extubation to NIV (yes or no).

All subjects with SMA admitted to the PICU were scored daily on 11 clinical parameters (breathing frequency, heart rate, room air peripheral oxygen saturation, oxygen requirement, breath sounds, color, most recent chest radiograph findings, work of breathing, secretions, cough, and mental status) per the locally developed protocol. The sum score determined the frequency of airway clearance and the estimated need for NIV (Supplementary document A [see the supplementary materials at http://www.rcjournal.com]). A pediatric respiratory therapist determined a triage score for each subject every morning 1 h after completing his or her airway clearance treatment. If possible, the subjects were assessed while they were on room air and without respiratory support. If the subject was using NIV due to hypoxemia when on room air and could not be taken off respiratory support, then he or she was scored at the highest level and received airway clearance every 2 h.

An airway clearance treatment consisted of mechanical insufflation-exsufflation (MI-E), followed by a secretion mobilization technique (percussion chest physiotherapy and postural drainage or intrapulmonary percussive ventilation [IPV], high-frequency chest wall compressions), followed by MI-E, followed by postural drainage, followed by MI-E again. Nasal and oral secretions were suctioned as needed. MI-E was performed either via face mask or endotracheal tube at 30–40 cm H₂O inspiratory to −30 to −40 cm H₂O expiratory pressure, with a minimum of 4 sets of 5 breaths given or until secretions were cleared each time it was used.

Statistical Analysis

Comparisons of categorical and continuous variables for all encounters were analyzed by using the chi-square test, Fisher exact test, and Kruskal-Wallis test as appropriate (cross-sectional analysis). To account for correlations within the subjects, repeated measures analysis was conducted (longitudinal analysis) by using the SAS Glimmix procedure (SAS Institute, Cary, North Carolina). Logit link function for binary data and cumulative logit for binomial data were used. The LSMEANS statement (SAS) was specified to compute predicted means and 95% confidence limits for longitudinal analysis. P < .05 was considered statistically significant. SAS software version 9.4 was used for all statistical analyses. An encounter for a postoperation admission (1) or admissions with tracheostomy (24) were excluded from the subanalysis.

To identify and examine the risk factors of endotracheal intubation, the time to event analysis by using the SAS procedure Phreg was performed. Variables potentially linked to an increased risk of intubation were included. Stepwise selection (with option SLENTRY = 0.25 and option SLSTAY = 0.15 specified) was used to identify important prognostic factors from these variables. A variable has to be significant at the 0.25 level before it can be entered into the model, whereas a variable in the model has to be significant at the 0.15 level for it to remain in the model. The stepwise selection process results in a model with 6 explanatory variables (use of home NIV, weight, body mass index, pH, airway clearance with IPV, and number of hospital admissions).

Results

Our study population included 300 unique encounters among 137 unique subjects. The trend in subject encounters admitted over the study period is illustrated in Figure 1. The general characteristics of the admissions can be found in Table 1. Eighty-nine subjects (65%) had the diagnosis of SMA type I and 48 (35%) had SMA type II. The subjects with type I accounted for 199 admissions (66.3%). Thirty-four percent of the subjects with type I and those with type II were boys (P = .67). The subjects with SMA type II were older than the subjects with SMA type I, with a median (IQR) age of 4.0 (4.1–6.0) y and 1.0 (0.7–5.0) y, respectively (P < .001). Of the 137 subjects admitted, 49 had multiple admissions during the study period, 39 of those subjects had SMA type I. Twenty-six subjects were admitted > 3 times each during the study period, 19 of those subjects (73%) had SMA type I. The subjects were malnourished, with a median (IQR) body mass index of 17 (1–62) percentile, with SMA type I having a lower median (IQR) body mass index percentile, of 10 (1–44) versus 38 (3–77), for the subjects with SMA type II (P = .003). Similarly, the subjects with SMA type I were more often admitted with gastric tubes (P = .007) (Table 2).

Fig. 1.

Hospital encounters for subjects admitted with spinal muscular atrophy (SMA) type I or II over the study period.

Table 1.

Subject Demographics (N = 137)

Table 2.

SMA Type I vs Type II Encounters (cross-sectional analysis)

Two hundred fifty-seven encounters (85.7%) were supported with home NIV before admission. Of those, 166 encounters (83.4%) were in the subjects with SMA type I and 91 of the subjects (90.1%) had SMA type II (P = .12). However, 14 (7.1%) SMA type I subjects' were admitted with tracheostomy (P = .003). The most common admission diagnosis was pneumonia, followed by mucus plugging. Admission pH, ${\text{P}}_{{\text{aCO}}_{\text{2}}}$, or admission diagnoses were not statistically significant between the subjects with SMA type I and type II (P = .17, P = .07, and P = .29, respectively). The overall mortality rate was 12.4%.

Intubation and Mechanical Ventilation

Sixty-eight subjects were intubated on admission or during their stay in the PICU: 65% with SMA type I and 35% with SMA type II (P = .89). The median (IQR) time to intubation was 0 (0–20) h, and the median (IQR) duration of invasive mechanical ventilation was 2 (10–5) d, with no statistical difference between type I and type II (P = .89). The subjects who required endotracheal intubation had higher Pediatric Risk of Mortality score III scores and worse hypoxemia (P = .002 and P = .004, respectively). The risk of intubation was higher in the subjects who were supported with NIV at home (hazard ratio 6.46, 95% CI 1.19–35.12; P = .042) and had a higher body mass index percentile (hazard ratio 1.13, 95% CI 1.001–1.267; P = .045). Of those, 62 (91.2%) were successfully extubated to NIV on the first attempt, whereas 4 subjects (5.9%) required multiple extubation attempts, and 3 subjects (4.4%) ultimately required tracheostomy.

In the group managed with NIV but without endotracheal intubation, 7 subjects (3.4%) received elective tracheostomy before discharge. For the subjects who required endotracheal intubation, both PICU and the total length of stay were longer (P < .001) when compared with the subjects who did not require endotracheal intubation. Admission ${\text{F}}_{{\text{IO}}_{\text{2}}}$, ${\text{P}}_{{\text{aCO}}_{\text{2}}}$, pH, Pediatric Risk of Mortality score III, and time to wean to ${\text{F}}_{{\text{IO}}_{\text{2}}}$ = 0.21 are shown in Table 3. When comparing the subjects with SMA type I and those with SMA type II, the subjects with type I had longer PICU stay with a median (IQR) of 5 (30–11) d versus 4 (20–7) d (P = .002). However, the hospital length of stay and duration of invasive mechanical ventilation were not statistically different (P = .055 and P = .068, respectively).

Table 3.

Endotracheal Intubation vs Noninvasive Ventilation Encounters, Excluding Subjects Admitted With Tracheostomy

Airway Clearance Therapies

All the subjects received airway clearance at home before admission as part of their chronic care (Supplementary Table 1, see the supplementary materials at http://www.rcjournal.com). In the PICU, there was no significant difference in the duration of airway clearance frequency between the subjects with type I and those with type II (Table 2). The duration of airway clearance ordered and received at the frequencies of every 2, 4, and 6 h were significantly higher in the subjects who were intubated (Table 3), as was the use of IPV and MI-E devices (P < .001) (Table 4). IPV, high-frequency chest wall compressions therapy, and percussion chest physiotherapy and postural drainage were more commonly used in the subjects with SMA type I (Supplementary Table 2, see the supplementary materials at http://www.rcjournal.com).

Table 4.

Airway Clearance Therapy Encounters During Admission While on Invasive Mechanical Ventilation vs Noninvasive Ventilation

Discussion

In this study, we reported that the majority of the subjects with SMA type I or II admitted to the PICU with acute or acute-on-chronic respiratory failure could be treated successfully with NIV and aggressive airway clearance. Similarly, when endotracheal intubation was required, successful extubation could be achieved with NIV transitional support combined with aggressive airway clearance maneuvers. To the best of our knowledge, this study was the largest clinical report that describes the use of both NIV and invasive mechanical ventilation combined with aggressive airway clearance in subjects with SMA type I or II.

Children with SMA are at higher risk for acute respiratory decompensation due to impaired secretion clearance, community-acquired pneumonia, and aspiration.⁸ This is primarily due to respiratory muscle weakness and the inability to clear secretions. Therefore, NIV and MI-E techniques are essential for managing these children during episodes of acute respiratory failure.^2,9 The current literature and guidelines focus on escalating airway clearance, NIV, and nutrition to avoid endotracheal intubation.^3,4 Lemoine et al¹⁰ described the use of early, proactive NIV support and saw an increased survival in this population, although the mortality rates in study were higher than in ours. We believe this is because of our unique and aggressive but highly standardized approach to airway clearance. Our local airway clearance protocol consists of MI-E, IPV, or percussion chest physiotherapy, and postural drainage according to a scheduled scoring system. This aggressive treatment is performed every 2–4 h based on our clinical scoring system, and as needed, to clear secretions and mucus plugs, prevent mucus plugging, open atelectatic alveoli, and minimize further atelectasis.

That said, endotracheal intubation is not always avoidable, even with such an aggressive management strategy. Bach et al¹¹ described their experience with 106 subjects with SMA type I when using both NIV and tracheostomy, and found that younger subjects were more frequently treated with endotracheal intubation and recommended that approach until patients are old enough to cooperate with NIV and MI-E. In our experience, endotracheal intubation is required in < 25% of admissions and mainly in older children. Also, we found that the majority of subjects could be managed successfully with NIV and aggressive airway clearance alone during an acute respiratory illness. Moreover, in our cohort, only 17 of the subjects with SMA type I (12.4%) referred to our institution required tracheostomy during the study period compared with 25.47% in the cohort in the study by Bach et al.¹¹

Limited evidence-based guidelines are available to guide extubation in this population. The 2018 SMA diagnosis and management clearance guidelines recommend aggressive use of airway clearance to re-expand lungs before extubation and NIV as transitional support after extubation for all patients, which is consistent with our current local guidelines for the past 15 or more years.⁸ The only published extubation protocol for patients with SMA is the Bach protocol, which has a success rate of 83% compared with 91.1% in our study.^12,13 This protocol does not require the child to have a respiratory effort before extubation. It also requires that patients do not have an oxygen requirement at the time of extubation. If one includes the “conventional plus” (extubated to NIV plus supplemental oxygen) group in the 2002 study by Bach et al¹³ with the “protocol” group, the success rate of extubation in that protocol decreases to 75%.The extubation practice in our institution differs significantly from this protocol in that we do not require the child to be on room air, but they must be spontaneously breathing before extubation, and we provide scheduled, aggressive airway clearance instead of “only as needed” treatments.

Although there are no published protocols for extubation criteria and ventilator liberation in our PICU for patients with a neuromuscular disorder, we have guidelines for extubation readiness. Our extubation readiness standard of care requires that the patient tolerate a PEEP of 5 cm H₂O, that the ventilator rate is equal to the patient's home NIV rate, that the patient tolerates pressure support sufficient to compensate for the resistance of smaller endotracheal tube sizes that deliver physiologic tidal volumes, and that the patient tolerates an ${\text{F}}_{{\text{IO}}_{\text{2}}}$ ≤ 0.4 to achieve ${\text{S}}_{{\text{pO}}_{\text{2}}}$ > 90–92%.

These criteria are consistent with widely used mechanical ventilation liberation protocols, except for the ventilatory rate specifically for children with SMA.¹⁴ Furthermore, the patient must also be afebrile, have less than a moderate amount of thin, clear-to-white secretions, and be tolerant of airway clearance treatments every 4 h. Conditioning of the respiratory muscles by decreasing the breathing frequency is begun 2–3 d before the expected extubation date. Patients are placed on a breathing frequency equal to their home NIV machine breathing frequency overnight to rest. This practice is because children with SMA fatigue more easily while asleep and because resting the respiratory muscles overnight allows more energy expenditure during the day. Pressure support trials are not conducted because the child is being extubated to NIV with a set machine breathing frequency and because families have reported that their children fatigue quickly and become tachypneic with these trials.

Both MI-E and IPV are performed immediately before and after extubation. All patients are extubated to NIV, and the airway clearance frequency is changed to every 2 h initially and then spaced out as the child tolerates and per his or her protocol score (see the supplementary materials at http://www.rcjournal.com). Although the majority of these weaning and extubation practices are not in our institution's PICU extubation readiness guidelines, they have been discussed by the PICU physicians and pulmonology division over the years and are considered the standard of care in all our patients with a neuromuscular disorder.

Once extubated, if the child is having difficulty synchronizing with the noninvasive ventilator, then the backup rate may be adjusted to equal the patient’s actual breathing frequency or at an interval that better synchronizes with it in an attempt to capture his or her respiratory pattern. Additional settings, when applicable, such as inspiratory time, rise time, and flow patterns, may also be adjusted if there is difficulty with synchronization. We try to use the child's home NIV mask (nasal or full face) and machine to make him or her more comfortable. If a child using the nasal NIV interface is having significant air leakage from the mouth, then a chin strap may be used. Oronasal masks are also an option in the event of leakage alarms.

Finally, during our study period, we noticed a peak in encounters in 2014, followed by a significant decline in the following years. We believe that is, in part, due to the emergence of survival motor neuron (SMN) gene disease–modifying treatments over the past decade for patients with SMA and the potential change in the natural history of the disease, particularly in patients with SMA type I.¹⁵ However, this study did not evaluate the effects of such therapies on our population.

There were limitations to this study. This was a single center, retrospective observational study and had no control group. In addition, our protocol scoring tool and determination of the frequency of airway clearance maneuvers has not been validated, is subjective, and depends on the experience, skill, and clinical observations of a respiratory therapist at one point in time. The respiratory therapists at our institution are experienced at evaluating this subset of patients due to the large numbers that we treat, but this may be a more difficult task in PICUs that infrequently care for these children. Also, the frequency of the airway clearance is often at the discretion of the physician provider, thus there are values for the treatments performed every 3 h when the triage scoring only permits the frequency at every 2 or 4 h. Patients will also often get treatments at differing frequencies during the day and night hours as they improve.

Conclusions

Children with SMA type I or II can be treated successfully with NIV and aggressive airway clearance during acute respiratory failure. Similarly, when intubation was required, successful extubation could be achieved with NIV transitional support combined with aggressive airway clearance maneuvers. A multi-centered prospective study that evaluates the use of NIV and types of airway clearance during acute respiratory failure is needed.