Functional Outcome Measures to Optimize Drug Development in Spinal and Bulbar Muscular Atrophy: Results From a Meta-Analysis of the Global SBMA Dataset

Abstract

Background and objectives: Spinal and bulbar muscular atrophy (SBMA) is a rare, slowly progressive, and debilitating disease without effective treatments available. Lack of reliable biomarkers and sensitive outcome measures makes clinical research conduct challenging. The primary objective of this study was to identify clinically meaningful and statistically sensitive outcome measures enabling the evaluation of therapeutic interventions in late-stage clinical trials.

Methods: This study was a meta-analysis of SBMA patient-level data from 6 observational studies conducted in Italy, South Korea, Denmark, United Kingdom, Japan, and United States. Patients with confirmed SBMA genetic diagnosis and differing severity were enrolled following individual site protocols. Routine assessments were performed longitudinally for approximately 3 years, including one or more clinical outcomes, such as SBMA functional rating scale (SBMAFRS), 6-minute walk test (6MWT), quantitative muscle testing (QMT), and Adult Myopathy Assessment Tool (AMAT). A modified scale, m-SBMAFRS, was derived by including only lower limb and trunk subscales having lower variability and larger effect size compared with the others. Changes from baseline at follow-up time points were calculated for all measures, and percent changes using random slope models were calculated to compare clinical measure performances. A survey conducted on 196 patients by the Coordination of Rare Diseases at Sanford (CoRDS), elucidating the impact of specific disease aspects on patients' lives, was also evaluated to corroborate these research outcomes.

Results: This global SBMA dataset analyzed data from 278 men (mean age = 59.7 ± 10.8 years, mean disease duration = 17.7 ± 11.9 years). Patients progressed on SBMAFRS (-4.7 ± 6.2 points after 38 months with 1-year standard response mean [SRM] = 0.6) and 6MWT (distance walked decreased by -53.2 ± 87.0 meters after 26 months with 1-year SRM = 0.5). These measures showed lower variability and larger effect size than AMAT and QMT (1-year SRM = 0.1 and -0.2, respectively) and confirmed SBMA linear progression across a range of disease stages. The m-SBMAFRS also showed a significant yearly decline of 0.9 ± 1.5 points (SRM = 0.6) and more consistent performance with less variability across clinical sites. The CoRDS survey confirmed the relevance of lower limb strength and mobility, which correlated with higher quality-of-life metrics and were reported by patients as predominant disease issues.

Discussion: We generated a comprehensive global SBMA dataset, enabling the identification of sensitive functional end points for clinical trials. Possible limitations relate to data collection nuances across sites that a single study protocol could override.

Figure 1. Linear Changes in SBMA Functional Outcome Measures

Boxplots illustrating the progressive decline in the SBMA functional rating scale (SBMAFRS) and 6-minute walk test (6MWT) scores among patients with SBMA. These plots show changes from baseline with lower (25th percentile) and upper (75th percentile) hinges and whiskers extending to 1.5 times the interquartile range. The horizontal line at the center and black dot within each boxplot represent the median and mean values, respectively, while error bars depict standard errors. SBMA, spinal and bulbar muscular atrophy.

Figure 2. Comparing Longitudinal Progression Using SBMA Functional Outcome Measures

Spaghetti plots displaying changes in quantitative muscle testing (QMT), Adult Myopathy Assessment Tool (AMAT), 6MWT, SBMAFRS, and m-SBMAFRS. Each plot includes a best-fitting linear regression line and a shaded 95% CI. Random slope model statistics are reported, including sample size (N), t-statistic (t), slope coefficient (B), subject-specific SD, p value (P), and Pearson correlation coefficient (R). SBMA, spinal and bulbar muscular atrophy.

Figure 3. Changes in SBMAFRS Subscales and Subscale Items Over Time

Boxplots showing longitudinal changes in (A) SBMAFRS subscales and (B) specific items within each subscale, arranged to emphasize those with significant mean deterioration. Each boxplot marks the 25th and 75th percentiles at the lower and upper bounds, respectively. Whiskers extend 1.5 times the interquartile range from the hinges. Median scores are indicated by horizontal lines, with black dots and error bars representing means and standard errors. The m-SBMAFRS (C) was derived by including lower limb and trunk subscales, which show higher effect size and lower variability. SBMAFRS, SBMA functional rating scale.

Figure 4. Age at Onset of SBMA Symptoms in Patients

Line plots depicting the average onset age of individual symptoms among patients with SBMA, comparing self-reported symptoms from the CoRDS survey and clinical assessments from the Italian SBMA cohort. Data points, color-coded by symptom frequency, show the mean and standard error. Percentages above points indicate the prevalence of each symptom within the CoRDS and Italian cohorts. CoRDS, Coordination of Rare Diseases at Sanford; SBMA, spinal and bulbar muscular atrophy.

Figure 5. Patient Perspective on Kennedy Disease Impact

(A) Lollipop plot representing responses to the survey question about the primary problem experienced due to Kennedy disease. Categories are derived from key terms in open-ended responses: breathing, upper limbs, falling, bulbar, walking and mobility, weakness, and muscle wasting. (B) Scatter plot correlating SBMAFRS scores with self-reported quality-of-life (QoL) metrics. QoL is rated on a scale from 0 (poor) to 4 (excellent). Pearson correlation coefficients (R) and p values (P) are included. Note: 1 item in the SBMAFRS (bowing) was not assessed. Points are jittered for clearer visualization.