Switching from active vitamin D and phosphate supplementation to burosumab significantly corrects lower limb malalignment in pediatric X-linked hypophosphatemia

Abstract

X-linked hypophosphatemia (XLH) is a rare disorder of renal phosphate wasting and dysregulated active vitamin D metabolism, ultimately presenting as rickets and osteomalacia, among other manifestations. Lower extremity deformity (genu valgum and/or varum) is frequent in this pediatric population. Despite prompt active vitamin D and phosphate supplementation (active D/Pi), many patients require corrective surgery for lower limb malformation. Burosumab has demonstrated improvements in lower limb malalignment in children with XLH in several studies. We expand on those reports by assessing mechanical femoral tibial angle (mFTA) change in patients enrolled in the XLH Disease Monitoring Program (DMP), (NCT03651505) to determine the impact of initiating burosumab treatment after a history of active D/Pi. Included patients had either switched from active D/Pi to burosumab treatment at the discretion of their treating physician or as part of a burosumab clinical trial, or remained on active D/Pi through Year 3 of the DMP. Year 3 radiographs were compared with baseline to assess mFTA change and gauge improvement. Additional multivariate factor analysis examined 24 attributes to determine which had the greatest association with mFTA change. Change in mFTA was assessed for each limb independently. A greater proportion of limbs of patients switching from active D/Pi to burosumab had improved mFTA compared with those remaining on active D/Pi (p < .023). Odds ratios comparing limbs that improved to those that did not showed that switching to burosumab yields a significantly greater chance of improvement than continuing active D/Pi (OR [95% CI]: 4.38 [1.09-17.50]; p = .0469). Factor analysis identified younger age at burosumab initiation (p = .001) and lower baseline height Z-score (p = .006) as being significantly associated with greater change in mFTA Z-score. This study shows that switching to burosumab significantly improves lower limb malalignment in children with XLH over benefits conferred by active D/Pi, with early burosumab initiation providing the greatest benefit.

Keywords: X-linked hypophosphatemia; active vitamin D; burosumab; genu valgum; genu varum; lower limb malalignment; phosphate.

Figure 1

Mechanical femoral tibial angle (mFTA) measurement. mFTA measurements were performed at each time point (baseline and DMP Year 3) by a single reader masked to the subject identification number and treatment group. The reader reported the mFTA of each limb independently, measuring from the center of the femoral head to the midpoint of the knee joint, and from the midpoint of the knee joint to the midpoint of the ankle joint. mFTA Z-scores greater than +2.0 (ie, more than 2.0 SD above the healthy average) were considered varus (see left), while those less than −2.0 (ie, more than 2.0 SD below the healthy average) were considered valgus (see right).

Figure 2

Definitions of normal and change from baseline. mFTA Z-score changes from baseline at DMP year 3 were categorized as stable (normal/abnormal), improving, or worsening based on initial and final measures. Any baseline score of “normal” (ie, within 2 SD of the healthy average) with a corresponding final value outside of the normal range, or a baseline score outside of the normal range deviating further from the normal range by at least 1.0 SD at final measure, was considered “worsening.” Any baseline score of “abnormal” (ie, beyond 2.0 SD of the healthy average in either direction) with a corresponding final value within the normal range, or with a final value at least 1.0 SD improvement towards the normal average, was considered “improving.” Scores that remained within the normal range of ±2.0 SD of the age-matched healthy average at baseline and follow-up were considered stable normal; those that changed less than 1.0 SD in the presence of both baseline and follow-up values that were outside of the normal range were considered stable abnormal.

Figure 3

Proportion of patients (A) and limbs (B) with lower limb malalignment at baseline and the corresponding mean Z-scores for Varus and valgus deformities. mFTA Z-scores were categorized by baseline deformity (valgus, varus, or one of each [windswept]) at the patient (A) and limb (B) levels grouped by treatment.

Figure 4

mFTA change from baseline at DMP Year 3. The change in mFTA Z-score from baseline at DMP Year 3 was used to group limbs as improved, stabilized, or worsened within each treatment group. The proportion of limbs in each grouping is reported. Fisher’s exact test comparing the proportions of patients within each treatment group that improved, with those that worsened, yielded a p-value of .023. *Mean burosumab duration 4.0 ± 1.8 yr, including exposure during prior clinical trials.

Figure 5

Radiographic images of mFTA outcomes. Standing, full-length lower extremity radiographs were obtained at baseline and DMP Year 3 according to protocol specifications: bilateral, anterior-posterior views in the upright (standing) position, with single images containing both legs from above the iliac crests to below the ankle. Subject preparation and positioning were performed per the site’s institutional standard of care. Image quality standards were established to ensure quality control. Representative images of improvement (top), stabilization (middle), and worsening (bottom) are provided.

Figure 6

Mean change in mFTA Z-score by time on active D/Pi between baseline and Year 3. Mean change in mFTA Z-score is presented with limbs grouped by duration of prior active D/Pi treatment. Here, the vertical limits of the box represent quartiles 1 and 3, the whiskers represent the minimum/maximum values, the inner horizontal line represents the median, the “X” represents the mean, and the circles represent outliers. Mean (SD) mFTA Z-scores at baseline and Year 3 are reported below (table).