Progression rate of ALSFRS-R at time of diagnosis predicts survival time in ALS
Paul H.Gordon, Columbia University, Neurological Institute, 9th Floor, 710 West 168th St, New York, NY 10032phg8@columbia.edu
Ying Kuen Cheung
Submitted May 24, 2006
We read with interest the article by Kimura et al [1] describing
assessment of progression rate at time of diagnosis using the ALS
Functional Rating Scale (ALSFRS-R). It is difficult to measure progression
in ALS trials because there are no biomarkers, and the standard outcomes
are clinical.
Six months are needed to detect changes in the ALSFRS-R
because of variability, due principally to differing rates of progression
among patients. Stratified enrollment lowers variability by reducing
heterogeneity in the treatment arms. While site of onset and riluzole
treatment may impart modest effects, the person’s rate of progression is
the most important predictor of outcome. [2] It is theoretically possible
to assign strata using historical information on progression at the
baseline visit of a trial using the DeltaFS. [1]
We measured the DeltaFS for our clinic patients to find the cutoff
value that best dichotomizes the population into two groups for an
upcoming phase II trial. We used DeltaFS = (48-"baseline" ALSFRS-R) /
time from onset to "baseline" (months).
We took first visit to clinic as baseline. Unlike Kimura et al who
considered survival as primary endpoint, our primary outcome is the 6-
month change in ALSFRS-R. Of 442 patients with information on DeltaFS
(median=0.55, inter-quartile range=[0.269, 1.11]), 112 patients had ASLFRS
-R scores at baseline and 6 months later. The reduction in variance was
maximized when the cutoff 0.50 per month was used to separate fast and
slow progression: The mean of the 6-month score in the fast progression
group was 4.11 points lower than that in the slow progression group
(p<0.0001). Using the cutoff suggested by Kimura et al (0.67),
variance reduction was significant (p=0.0005) with mean for fast
progression 3.67 lower than that for slow progression.
The DeltaFS is an excellent measure to determine rate of progression
at first encounter, and can be used for stratification in clinical trials.
Both 0.50 and 0.67 are acceptable points of dichotomization in terms of
reducing heterogeneity in the study population, although 0.50 provides
slightly better reduction and is slightly easier arithmetically.
We
recommend that individual studies choose a dichotomization cutoff based on
their data, as the value could change slightly from region to region.
Further analysis may provide a global value of dichotomization for
stratification in clinical trials.
References
1. Kimura F, Fujimura C, Ishida S, et al. Progression rate of ALSFRS-R at
time of diagnosis predicts survival time in ALS. Neurology 2006;66:265-
267.
2. Armon C, Moses D. Linear estimates of rates of disease progression as
predictors of survival in patients with ALS entering clinical trials. J
Neurol Sci 1998;160:S37-41.
Disclosure: The authors report no conflicts of interest.
We read with interest the article by Kimura et al [1] describing assessment of progression rate at time of diagnosis using the ALS Functional Rating Scale (ALSFRS-R). It is difficult to measure progression in ALS trials because there are no biomarkers, and the standard outcomes are clinical.
Six months are needed to detect changes in the ALSFRS-R because of variability, due principally to differing rates of progression among patients. Stratified enrollment lowers variability by reducing heterogeneity in the treatment arms. While site of onset and riluzole treatment may impart modest effects, the person’s rate of progression is the most important predictor of outcome. [2] It is theoretically possible to assign strata using historical information on progression at the baseline visit of a trial using the DeltaFS. [1]
We measured the DeltaFS for our clinic patients to find the cutoff value that best dichotomizes the population into two groups for an upcoming phase II trial. We used DeltaFS = (48-"baseline" ALSFRS-R) / time from onset to "baseline" (months). We took first visit to clinic as baseline. Unlike Kimura et al who considered survival as primary endpoint, our primary outcome is the 6- month change in ALSFRS-R. Of 442 patients with information on DeltaFS (median=0.55, inter-quartile range=[0.269, 1.11]), 112 patients had ASLFRS -R scores at baseline and 6 months later. The reduction in variance was maximized when the cutoff 0.50 per month was used to separate fast and slow progression: The mean of the 6-month score in the fast progression group was 4.11 points lower than that in the slow progression group (p<0.0001). Using the cutoff suggested by Kimura et al (0.67), variance reduction was significant (p=0.0005) with mean for fast progression 3.67 lower than that for slow progression.
The DeltaFS is an excellent measure to determine rate of progression at first encounter, and can be used for stratification in clinical trials. Both 0.50 and 0.67 are acceptable points of dichotomization in terms of reducing heterogeneity in the study population, although 0.50 provides slightly better reduction and is slightly easier arithmetically.
We recommend that individual studies choose a dichotomization cutoff based on their data, as the value could change slightly from region to region. Further analysis may provide a global value of dichotomization for stratification in clinical trials.
References
1. Kimura F, Fujimura C, Ishida S, et al. Progression rate of ALSFRS-R at time of diagnosis predicts survival time in ALS. Neurology 2006;66:265- 267.
2. Armon C, Moses D. Linear estimates of rates of disease progression as predictors of survival in patients with ALS entering clinical trials. J Neurol Sci 1998;160:S37-41.
Disclosure: The authors report no conflicts of interest.