Effect of phenotypic heterogeneity
Another important factor potentially confounding the interpretation of aggregate results of intervention studies on respiratory outcomes is phenotypic heterogeneity. This variability has been recognized for years18 and can manifest as markedly different timing and value maximum VC as well as rate of decline. Neve, in a study of 33 CS naïve DMD subjects identified subphenotypes requiring NIV either prior to, or after age 17 40. The former was identified by an earlier (age 11.5 to 13) and lower peak VC and SNIP and more rapid decline. In a study using the STRIDE nonsense mutation (nmDMD) cohort, investigators found no correlation in 181 patients between disease severity and exon location of the nonsense mutation. There was a large degree of phenotypic variability despite of having identical dystrophin mutations 41. Birnkrant has identified highly discordant cardiopulmonary function in brothers with DMD and identical dystrophin mutations and identified 8 of 15 patients (53%) with a shared genotype who had discordant cardiopulmonary function23,42. Clinical manifestations were not related to the dystrophin mutation. In a cohort of over 200 patients, Magri, as well, noted the absence of correlation of pulmonary, cardiac or motor phenotype with dystrophin genotype, with the exception of neurocognitive function being related to more distal mutations 43.
Being unable to predict the natural evolution of cardiopulmonary function in a portion of the population despite common genotype could pose a significant confounding issue when assessing the impact of emerging therapies in DMD. Rather than being the true treatment effect, such beneficial or adverse observations might simply be due to more favorable or unfavorable phenotype.