A Phase 3 Study of UX003 Recombinant Human Betaglucuronidase (rhGUS) Enzyme Replacement Therapy in Patients With Mucopolysaccharidosis Type 7 (MPS 7)

Baseline was defined as the average of all assessments prior to or on the date of cross-over to active treatment with UX003. (This provides the valid assessment, as placebo effect was subtracted from the treatment effect estimate providing a more conservative assessment of the true treatment effect.) Percent change from baseline in uGAG DS was analyzed by generalized estimating equation (GEE) modeling based on observed data. The GEE model included included baseline value, and the UX003 treatment week as a categorical variable. The covariance structure within subjects was assumed to be exchangeable.

In the United States (US), this was considered a secondary outcome measure. Per guidance from the Food and Drug Administration (FDA), no primary efficacy variable was declared in the US. Efficacy was to be based on the totality of the clinical data on a per participant basis.

MDRI score, calculated as the total response score at UX003 Treatment Week 24 across 6 domains: 6-Minute Walk Test, forced vital capacity predicted value, shoulder flexion, visual acuity, and Bruininks-Oseretsky Test of Motor Proficiency fine motor and gross motor capacity. For each domain, a minimally important difference (MID) was pre-specified. Changes from before treatment (baseline) to 24 weeks after treatment in each domain variable were scored against pre-specified MIDs. (This provides the valid assessment, as placebo effect would be subtracted from the treatment effect estimate providing a more conservative assessment of the true treatment effect.) An improvement or decline ≥ MID was scored either as a +1 or -1, respectively, and a change <MID was scored as 0. The integration of benefit occurred by summing the responses (-1, +1, 0) across all 6 domain variables to derive the MDRI score, with a range of -6 (greatest possible decline) to +6 (greatest possible improvement).

The total distance walked (in meters) in a 6-minute period was measured. Baseline was defined as the last assessment prior to or on the date of cross-over to active treatment with UX003. (This provides the valid assessment, as placebo effect would be subtracted from the treatment effect estimate providing a more conservative assessment of the true treatment effect.) A positive change from Baseline indicates improvement. Change from baseline in 6MWT was analyzed by GEE modeling based on observed data. The GEE model included all participants who had non-missing baseline and at least one post-baseline value during the 24 weeks of UX003 treatment. The model included baseline value, and the UX003 treatment week as a categorical variable. The covariance structure within participants was assumed to be exchangeable.

Spirometry was administered to participants who did not require invasive ventilatory support or have a tracheostomy and measured percentage of predicted FVC. The percent predicted values were calculated after testing using published normative data. Baseline was defined as the last assessment prior to or on the date of cross-over to active treatment with UX003. (This provides the valid assessment, as placebo effect would be subtracted from the treatment effect estimate providing a more conservative assessment of the true treatment effect.) No GEE analysis was performed for FVC due to the limitation of the sample size.

Spirometry was administered to participants who did not require invasive ventilatory support or have a tracheostomy to measure MVV. The percent predicted values were calculated after testing using published normative data. Baseline was defined as the last assessment prior to or on the date of cross-over to active treatment with UX003. (This provides the valid assessment, as placebo effect would be subtracted from the treatment effect estimate providing a more conservative assessment of the true treatment effect.)

Goniometry was used to measure (in degrees) the maximum passive shoulder range of motion in both flexion and extension. Baseline was defined as the last assessment prior to or on the date of cross-over to active treatment with UX003. (This provides the valid assessment, as placebo effect would be subtracted from the treatment effect estimate providing a more conservative assessment of the true treatment effect.) Change from baseline in shoulder flexion-left was analyzed by GEE modeling based on observed data. The GEE model included all participants who had non-missing baseline and at least one post-baseline value during the 24 weeks of UX003 treatment. The model included baseline value, and the UX003 treatment week as a categorical variable. The covariance structure within participants was assumed to be exchangeable.

Visual acuity was measured (corrected and uncorrected) using a standard eye chart and recorded for each eye independently. Baseline was defined as the last assessment prior to or on the date of cross-over to active treatment with UX003. The change in the number of lines from pre-treatment baseline to 24 weeks of treatment was evaluated. (This provides the valid assessment, as placebo effect would be subtracted from the treatment effect estimate providing a more conservative assessment of the true treatment effect.) A positive change from baseline indicates improvement. Change from baseline in uncorrected visual acuity was analyzed by GEE modeling based on observed data. The GEE model included all participants who had non-missing baseline and at least one post-baseline value during the 24 weeks of UX003 treatment. The model included baseline value, and the UX003 treatment week as a categorical variable. The covariance structure within participants was assumed to be exchangeable.

BOT-2 was administered to evaluate treatment-related changes in 4 domains assessing both fine and gross motor function: balance (score 0 to 37), fine motor precision (score 0 to 41), manual dexterity (score 0 to 45), and running speed/agility (score 0 to 52). Higher scores indicate more motor proficiency; a positive change from baseline indicates improvement. Baseline was defined as the last assessment prior to or on the date of cross-over to active treatment with UX003. (This provides the valid assessment, as placebo effect would be subtracted from the treatment effect estimate providing a more conservative assessment of the true treatment effect.) Change from baseline in BOT-2 was analyzed by GEE modeling based on observed data. The GEE model included baseline value, and the UX003 treatment week as a categorical variable. The covariance structure within participants was assumed to be exchangeable.

The PedsQL 18-item scale is comprised of 3 dimensions: general fatigue (6 items), sleep/rest fatigue (6 items) and cognitive fatigue (6 items). Each item has a 5-point Likert response scale that is reverse scored and transformed to a 0 to 100 scale with higher scores indicating less fatigue. Baseline was defined as the last assessment prior to or on the date of cross-over to active treatment with UX003. (This provides the valid assessment, as placebo effect would be subtracted from the treatment effect estimate providing a more conservative assessment of the true treatment effect.) Change from baseline in PedsQL total fatigue score was analyzed by GEE modeling based on observed data. The GEE model included all participants who had non-missing baseline and ≥1 post-baseline value during the 24 weeks of UX003 treatment. The model included baseline value, and the UX003 treatment week as a categorical variable. The covariance structure within participants was assumed to be exchangeable.

Percentage of participants who were ICR responders based on MID criteria at Week 24. At the Randomization visit, the physician queried the participant or parent/caregiver about signs and symptoms of MPS VII that interfered most with the participant's daily life. Answers were mapped to an appropriate clinical outcome measure (e.g., difficulty walking could map to the 6MWT; breathing problems to FVC). The clinical outcome ranked with the highest impact on daily life that could be reliably completed by the participant and met a threshold level of impairment was selected as the ICR for that participant. ICR response was assessed based on a positive change (according to pre-specified MID criteria) of each participant's ICR. Agresti-Coull confidence interval with nominal coverage ≥ 95%.

The 3 most impactful clinical problems as reported by the subject/parent/caregiver during the Clinical Problem Evaluation were scored on a Likert scale from 1 (very little problem) to 7 (an extreme amount) at randomization and post-randomization visits. At post-randomization visits, each clinical problem was again scored for impact on daily activities. Total scores ranged from 3 to 21; lower scores reflect less impact on daily life. Baseline was defined as the last assessment prior to or on the date of cross-over to active treatment with UX003. (This provides the valid assessment, as placebo effect would be subtracted from the treatment effect estimate providing a more conservative assessment of the true treatment effect.) The change from baseline up to UX003 Treatment Week 24 were analyzed by GEE modeling, including baseline value, and the post-UX003 initiation treatment week as a categorical variable. The covariance structure within participants is assumed to be exchangeable.