Natural history of infantile-onset spinal muscular atrophy

Abstract

Objective: Infantile-onset spinal muscular atrophy (SMA) is the most common genetic cause of infant mortality, typically resulting in death preceding age 2. Clinical trials in this population require an understanding of disease progression and identification of meaningful biomarkers to hasten therapeutic development and predict outcomes.

Methods: A longitudinal, multicenter, prospective natural history study enrolled 26 SMA infants and 27 control infants aged <6 months. Recruitment occurred at 14 centers over 21 months within the NINDS-sponsored NeuroNEXT (National Network for Excellence in Neuroscience Clinical Trials) Network. Infant motor function scales (Test of Infant Motor Performance Screening Items [TIMPSI], The Children's Hospital of Philadelphia Infant Test for Neuromuscular Disorders, and Alberta Infant Motor Score) and putative physiological and molecular biomarkers were assessed preceding age 6 months and at 6, 9, 12, 18, and 24 months with progression, correlations between motor function and biomarkers, and hazard ratios analyzed.

Results: Motor function scores (MFS) and compound muscle action potential (CMAP) decreased rapidly in SMA infants, whereas MFS in all healthy infants rapidly increased. Correlations were identified between TIMPSI and CMAP in SMA infants. TIMPSI at first study visit was associated with risk of combined endpoint of death or permanent invasive ventilation in SMA infants. Post-hoc analysis of survival to combined endpoint in SMA infants with 2 copies of SMN2 indicated a median age of 8 months at death (95% confidence interval, 6, 17).

Interpretation: These data of SMA and control outcome measures delineates meaningful change in clinical trials in infantile-onset SMA. The power and utility of NeuroNEXT to provide "real-world," prospective natural history data sets to accelerate public and private drug development programs for rare disease is demonstrated. Ann Neurol 2017;82:883-891.

Figure 1. Retention in SMA cohort

Age of infants at the time of death (filled circle), completion of the study (filled square) or loss to follow up (censored event, open circle). Age of scheduled study visits denoted on the x axis. Approximate age of permanent endotracheal intubation for respiratory support (closed triangle) is denoted for two infants. Bar colors denote SMN2 copy number; 4 copies = Grey, 3 copies = Blue, 2 copies = Red, Unknown copies = White.

Figure 2. Progression of Outcomes

Longitudinal average weight and motor function test results in first two years of life for healthy infants (blue), SMA infants where SMN2 copy number = 2 or is unknown (red) and SMA infants where SMN2 copy number > 2 (grey). A) Average weight in kilograms at each study visit. B) Average Test of Infant Motor Performance Screening Items (TIMPSI) score, C) The Children’s Hospital of Philadelphia Infant Test for Neuromuscular Disorders (CHOP-INTEND) score and D) Alberta Infant Motor Score (AIMS) score at each study visit. E) Average ulnar CMAP peak area (mV/s). Study visits linked to infant age (in months, +/− 2 weeks for visits 6, 12, 18 and 24 months). Shaded areas describe the standard deviation for each mean at each study visit. F) Kaplan-Meier curve of time to death or endotracheal tube placement plotted separately for the subgroup of SMA infants with SMN2 copy number equal to 3 or 4 (solid red line, n = 6) and for the subgroup of SMA infants with SMN2 copy number equal to 2 or unknown (dashed red line, n = 20). Circles represent censored events that occurred when participants left the study before observing either event in the combined endpoint.

Figure 3. Change in CHOP-INTEND in SMA Infants

Change in CHOP-INTEND score from the score obtained on the 6-month study visit for each infant in the SMA cohort (black lines). A mixed effects model was fit to model change for each time period and was adjusted for the 6-month CHOP-INTEND score (red line). Dashed black lines denote infants who received invasive ventilatory support. All differences in CHOP-INTEND from 6-months on were statistically significant. A test for linear trend was performed using orthogonal polynomials and the observed trend in differences was not linear (p = 0.1161).