Screening for Sleep Disordered Breathing in Children With Spinal Muscular Atrophy
—In children with spinal muscular atrophy, oximetry had value in predicting abnormal polysomnography, but transcutaneous carbon dioxide (TcCO2) was not useful.
Respiratory insufficiency is common in non-sitters and sitters with spinal muscular atrophy (SMA). A sleep study is recommended for children with SMA when there is a suspicion of nocturnal hypoventilation to determine the need for non-invasive ventilation (NIV) for those who show respiratory insufficiency.2 For those without symptoms, pulse oximetry to measure O2 saturation and measures of either end tidal CO2 or transcutaneous carbon dioxide (TcCO2) are recommended as screening tools.2
However, according to the authors of a new study, “There is currently no evidence comparing PSG with pulse oximetry and TcCO2 for ability as diagnostic or screening test(s) in SMA.”
The advantage of pulse oximeters and TcCO2 monitors, which measure the partial pressure of CO2 in arterial blood, is that they can be used at home without the need for a clinical laboratory specializing in polysomnography.
To determine whether pulse oximeters or TcCO2 monitors could have screening value as alternatives to laboratory PSG to identify sleep disordered breathing in children with SMA, the researchers assessed data for oxygen saturation and desaturation and arterial TcCO2 obtained from PSG before initiating disease-modifying therapy (DMT).
Study design
The study was conducted between June 2018 and November 2022 at the Queensland Children’s Hospital in Brisbane, Australia, and included all patients with SMA 1-3. The primary analysis was of data from overnight PSG before patients received DMT. PSG data evaluated included TcCO2, oxygen saturation, and pulse rate. For patients who were already receiving nocturnal NIV, PSG data were analyzed through the end of 2 rapid eye movement sleep cycles in the absence of NIV.
Abnormal PSG was defined as an apnea-hypopnea index of at least 10 events per hour for children with SMA aged 3 months and younger and at least 5 events per hour for children younger than age 3 months. The researchers used cutoff values for an abnormal apnea-hypopnea index that were low compared with the mean values for healthy children so that abnormal values would not be missed.3
Hypoventilation was defined by the American Academy of Sleep Medicine criteria for children based on the fraction of sleep time spent with elevated arterial partial pressure of CO2.4
The researchers defined 3 criteria as indications for need for NIV: 1) abnormal apnea-hypopnea index, 2) TcCO2 greater than 50 mm Hg for more than 2% of total sleep time on PSG, or 3) more than 2 lower respiratory tract infections that required intensive care admissions with NIV.
Variables evaluated for their use as screening measures included mean O2 saturation, mean CO2, number of desaturations below 90%, oxygen desaturation index 4 (ODI4, the hourly number of oxygen desaturation episodes with a decrease of at least 4%), maximal TcCO2, and McGill score, which takes into account the ODI4 and the depth of desaturation.
Participant characteristics
There were 47 children in the study who underwent PSG before initiating DMT. Median age was 1.5 years (range, 0.2-7.7) for 13 children with SMA type 1 and 7.8 years (range, 1.1-18.8) for 21 children with SMA type 2. Another 13 children had SMA type 3, with a median age of 9.6 years (range, 2.9-15.0).
Nocturnal NIV was already used in some participants, but none used NIV when awake or invasive ventilation. Few children with SMA type 1 (2 [15.4%]) or SMA type 3 (1 [7.7%]) used NIV. However, 8 of 21 (38.1%) with SMA type 2 used NIV.
Evaluation of potential screening tools
PSG was abnormal in 11 children (85%) with SMA type 1, 8 (42%) with SMA type 2, and 4 (31%) with SMA type 3. Despite the use of NIV by 11 children, none of the 47 children in the study met the criteria for hypoventilation.
Of the variables analyzed for use as a screening tool, the researchers found that only an ODI4 of at least 20 events per hour had any useful agreement with abnormal PSG. Nineteen of the 23 children with abnormal PSG also had an OD14 of at least 20 events per hour.
OD14 of at least 20 events per hour had a sensitivity of 82.6% (95% confidence interval [CI], 61.2%-95.0%), a specificity of 58.3% (95% CI, 36.6%-77.9%), a positive predictive value of 65.5% (95% CI, 45.7%-82.1%), and a negative predictive value of 77.8% (95% CI, 52.4%-93.6%).
All 3 children who died during the study had an ODI4 of at least 20 events per hour of McGill ≥2.
Combinations of ODI4 of at least 20 events per hour with other measures, such as TcCO2 of at least 50 mm Hg, were evaluated, but none had better sensitivity or specificity than ODI4 of at least 20 events per hour alone.
To see whether ODI4 of at least 20 events per hour agreed with abnormal PSG in treated patients, the researchers analyzed PSG results for 36 children after initiation of DMT. ODI4 did not reach 20 events per hour for any of the children treated regardless of SMA type and, therefore, was not useful.
A potentially useful screening tool—for some patients
The researchers concluded that TcCO2 was not a useful screening tool, likely because none of the children had hypoventilation. Close monitoring may be why children with SMA followed at the University of Queensland did not develop hypoventilation. At the researchers’ institution, all children with SMA undergo annual PSG. “When [sleep disordered breathing] is detected, children are offered NIV prior to hypoventilation development,” the researchers wrote.
Given the high non-predictive value (77.8%) of ODI4 of at least 20 events per hour, a negative result could not rule out the need for NIV. According to the researchers, the measure may still be useful in predicting which children need NIV before treatment for positive results. Similar predictive and non-predictive values are found with screening by oximetry for obstructive sleep apnea in children.5
“Unlike the current SMA respiratory management consensus guidelines, in our untreated patient group, as 11/13 of children with type 1 had abnormal PSG, we would suggest consideration of PSG with conversion to NIV/PSG guided NIV titration study in all type 1; screening using ODI4 =20 events/hour can be considered in type 2 and 3,” the researchers suggested in their paper.
Patients receiving treatment, the authors said, would still need PSG to screen for sleep disordered breathing because the ODI4 values were all low in this population. Limitations of the study they noted were use of a sleep laboratory with conditions different from home and lack of home screening tests to compare to PSG. Future studies should be performed with a larger sample size, if possible.
Published:
References
)