Recent advances in the treatment of Spinal Muscular Atrophy (SMA) promoted interest in the development of reliable assays for use in United States (US) newborn screening (NBS) programs. SMA is a progressive neuromuscular disease resulting from a deficiency of the Survival Motor Neuron (SMN) protein that is caused by bi-allelic pathogenic variants in the SMN1 gene; it is the leading genetic cause of death for infants. Approximately 95% of patients with SMA have a homozygous deletion of Exon 7 of the SMN1 gene. The clinical presentation of SMA is often modified by the number of copies of a paralog gene, SMN2, which differs from SMN1 by only five distinct base pairs.
Massachusetts and Utah were the first states in the US to offer statewide NBS for SMA. As early as 2015, Massachusetts had established an SMA working group including newborn screeners and clinician specialists to develop guidance for eventual implementation of statewide SMA NBS. Following the successful demonstration of a population-based SMA screening program in Taiwan [1
] and the availability of new treatment options targeting the modulation of the SMN2
], with other treatments such as gene therapy in the pipeline, the Massachusetts Newborn Screening Advisory Committee voted to offer SMA NBS. The Commonwealth of Massachusetts began offering statewide NBS for SMA in January 2018. In July 2018, the US Secretary of Health and Human Services approved the addition of the disorder, “SMA due to homozygous deletion of Exon 7 in SMN1
”, to the Recommended Uniform Screening Panel (RUSP) [3
]. As of 19 March 2021, 29 states were offering universal screening for SMA, accounting for approximately 60% of US births [4
Nine infants with SMA were identified by newborn screening and referred to a specialist before the end of their first week of life. Our tiered algorithm functioned well to identify all known affected infants in a timely manner without overwhelming nurseries, specialists, or primary care providers and without causing undue anxiety to families of unaffected infants.
Our choice of screening algorithm was attributable in part to the timing of our assay evaluations and our algorithm’s interface with our LIMS system. Additionally, we chose a conservative approach, not wanting to rely only on the LNA for sequence specificity within the high-throughput environment of NBS until more data accrued about this use. We continue to use independent assays for SCID and SMA. Until SMA screening is universally required in all states for which we provide screening services, we will continue to perform independent assays for SCID and for SMA. In the meantime, data we generate will be useful for methods comparison and harmonization with CDC proficiency testing panels and states that are performing multiplex SCID/SMA assays.
Our choice to provide SMN2
copy number data was driven by two factors: a need for universal access to such services and a need to improve turnaround time for these results. Turnaround time for diagnostic results including SMN2
copy number has greatly improved since we began screening in 2018. In addition, original treatment guidelines required that SMN2
copy number be no greater than 3 [12
]. More recent treatment guidelines [13
] do not require limits to the copy number of SMN2
for treatment, but in order to interpret both natural and treatment histories appropriately, a standardized copy number ascertainment is required. One of our infants was almost 6 months old before his first treatment, but he has 4 copies of SMN2
and was born prior to the 2020 guidelines recommending treatment. Our average age at first treatment for infants with 3 or fewer copies of SMN2
is 18 days.
At the time we began screening, we knew that if we had used Assay A of Tier 1 alone, we would generate a significant number of false-positive results. Some of these false positives would be due to the SMN1 hybrids that are identified by Assay B in Tier 2. We found that there is another subset of specimens that are not SMN1 hybrids (likely with blood collected from a line or using heparinized capillaries) that continue to show absent Exon 7 amplification in our Assay A in Tiers 1 and 2 but show amplification in Assay B, which are reconciled by our Tier 3 sequencing assay. The preponderance of specimens from NICU babies among those specimens prompting Tier 2 does suggest that there might be a PCR inhibitor that would have more implications for Assay A since NICU specimens are more likely to be collected from a line rather than a direct heel stick.
Our observed incidence of SMN1
hybrids is 1 in 17,947 [95% Confidence Interval (CI): 1/11,080–1/47,202], which is similar to the incidence of 1 in 24,053 reported by Chien et al. [1
] in the Taiwanese population. Historically, SMA has been reported to occur in approximately 1 in 10,000 births [14
]. To date, our observed incidence of SMA is 1 in 18,957 [95% CI: 1/12,061–1/57,519], which is lower than might be expected from projections of clinical presentation incidence [15
] that were observed prior to generalized but non-standardized availability of pre-pregnancy or prenatal SMA testing recommended by American College of Obstetricians and Gynecologists in 2017 [16
]. Whether the lower incidence in Massachusetts is due to a limited population sampling, reproductive choices, or a combination of the two remains to be answered.
Of the nine SMA infants identified, one had a prenatal diagnosis, three were born to parents who were both known to be carriers, and one was born to parents with only a single parent identified as a carrier (the other parent had not been tested). The clinical status reported by PCPs at the time of the NBS report was as follows: of the five high-risk infants (prenatal diagnosis or at least one parent known carriers), three were reported to be well, one had poor tone, and one was reported as already being under the care of a specialist. Of the four infants with usual risk (no reported prenatal or parent carrier testing), three were reported as well and one had low or poor tone and a dislocatable hip. One of the usual risk SMA-affected infants was in a SCN for reasons unrelated to overt signs of SMA (gestational age <37 weeks). Interestingly, the clinical status reported by PCPs did not always match the observations of the specialists and underscores the need for immediate clinical evaluation by a trained specialist. The specialists often noted more subtle clinical signs such as tongue fasciculations or axillary slip in infants who were reported to be well by PCPs.
All of our SMA-affected infants are alive and four of the nine are currently meeting developmental milestones expected for unaffected infants. The other five have experienced mild to moderate delays. There have been no documented harms introduced by treatment. SMA NBS is feasible and allows for the early detection and treatment of affected infants. We are optimistic that SMA NBS will help to maximize the likely benefits of early treatments.