P.262Crossing thresholds and changing rates of respiratory function decline are predictive of clinical outcomes in Duchenne muscular dystrophy (DMD)

Abstract

Falling below the forced vital capacity (FVC) threshold of 1 liter (L) has been shown to predict mortality in DMD. No systematic analyses have assessed the impact of crossing other respiratory function thresholds, or on the predictive value of the respiratory function decline rate in relation to clinically relevant outcomes. The CINRG Natural History Study was used to investigate these hypotheses using FVC or peak expiratory flow (PEF) as predictive measures. Crossing threshold analysis: Conducted in patients with at least one assessment above and below the FVC threshold of interest. Thresholds were scanned in small decrements until <5 events occurred. Kaplan-Meier method was used to derive 5- and 2-year risk estimates of death and start of assisted ventilation. Changing rates analysis: Conducted in patients with at least one assessment between 80–50% predicted (%p). Linear regression model was fitted for each patient to estimate average annual declines. Patients were categorized as fast (>5% per year) or slow (<5% per year) decliners. Time from crossing 50% PEF%p or FVC%p to death or start of ventilation was estimated with Kaplan-Meier method; sensitivity analyses were conducted using 4%, 6% and 7% declines and from crossing thresholds of 45%, 55% and 60%. 5-year survival after crossing an FVC of 1L was 86%. Similar survival was seen after PEF of 96 L/Min. Differences in rates between fast (mean: 9.17% per year) and slow (mean: 1.41% per year) decliners resulted in a 3-year delay in starting ventilation. Similarly, 5-year survival increased from 70% to 93% in fast vs slow decliners. These analyses demonstrate that crossing thresholds of FVC and PEF, and the rate of FVC%p and PEF%p decline are predictive of time to needing assisted ventilation and of survival. PEF and FVC are equally good predictors of clinical outcomes in DMD.