Challenges in Trial design
As discussed above, the interpretation of the impact of emerging
therapies on respiratory function using aggregate data is confounded by
phenotypic variability where it may not be possible to determine if a
subject is evolving naturally with favorable or unfavorable respiratory
function. As well, the age group commonly evaluated combines subjects
whose VC may be rising with others whose VC is declining. When true
randomization is not possible, attempts to accommodate for this have
included propensity matching as in the ataluren study41. However, propensity score matching can only adjust
for measured covariates, and one cannot be assured that all confounding
factors have been eliminated 52.
Becuase of the somewhat unpredictable evolution of respiratory function
in any given individual which may confound study results for a treatment
intervention, consideration might be given to an N of 1, crossover
design where subjects are compared with themselves under control and
experimental conditions and crossed over at intervals from one to the
other 53. The duration of recent studies and the
suggestion of improvements due to treatment as early as 26 weeks47, indicate that these investigations might lend
themselves to such an N of 1 design, e.g. randomization to 6 to 12
months of intervention or placebo, followed by crossover for the
remaining 6 to 12 months. Such a design would help address the issues
with regard to randomization and phenotypic variability.
Future studies should employ outcome measures which take in to account
the dynamic of respiratory function evolution at different ages and
which are sensitive to change. A number of parameters should be followed
including those which are sensitive to respiratory muscle strength (MIP,
SNIP) and absolute VC which, particularly in an N of 1 design, would
account for individual variations in age at maximal VC and rate of
decline. The utilization and impact of standard therapies must also be
considered in fully understanding the effect of emerging therapies.