Epidemiological aspects of scoliosis in a cohort of Japanese patients with Prader-Willi syndrome.

Background Context: The prevalence of scoliosis in Prader-Willi syndrome (PWS) is high; however, the prevalence of PWS is rare, with one person in 10,000 to 20,000 affected. The etiology and characteristics of scoliosis associated with PWS remain unidentified. In addition, it is believed that the speedup of growth associated with growth hormone (GH) supplement treatment may influence the deterioration of scoliosis in PWS.

Purpose: To investigate scoliosis in Japanese patients with PWS.

Study Design: Retrospective observational study.

Patient Sample: We investigated 101 patients (67 men, 34 women) who were followed up from November 2002 to January 2008. All patients were diagnosed using fluorescence in situ hybridization or the methylation test. Of the 101 patients, 80 had an inherited deletion of chromosome 15q11-13 (deletion) and 21 patients had no deletion, including those with uniparental disomy.

Outcome Measures: 1) Prevalence of scoliosis; 2) association of scoliosis with GH treatments; 3) association of scoliosis with genotype; 4) clarification of PWS scoliosis characteristics; and 5) analysis of severe PWS scoliosis patients (Cobb angle greater than 40 degrees ). Scoliosis for our study was defined as scoliosis with a Cobb angle greater than 10 degrees .

Methods: To investigate PWS-associated scoliosis, we used spinal X-ray examinations. The pattern of scoliosis was classified into three types: primary single lumbar or thoracolumbar curve (Type 1), double curve (Type 2), and primary single thoracic curve (Type 3). For statistical analysis, chi-square tests for the distribution of patients were used (p<.05).

Results: 1) Scoliosis was found in 38.6% (39/101) of patients with PWS. 2) There was no statistical difference in the prevalence of scoliosis between the GH treatment group (32.8%) and the GH nontreatment group (group with no GH treatments) (46.5%) (p=.16, chi-square test). 3) There was no statistical difference in the prevalence of scoliosis between the deletion group (38.8%) and the nondeletion group (38.1%) (p=.84, chi-square test). 4) Scoliosis was classified into three types, according to single or double curve scoliosis and position of scoliosis. The prevalence of these groups was 61.5% for Type 1 (primary single lumbar and thoracolumbar curve), 48.7% for lumbar curve convex on the left side, 28.2% for Type 2 (double curve), and 10.3% for Type 3 (primary single thoracic curve). 5) Severe scoliosis was found in nine patients (8.9%, 9/101). Type 2 was found in 66.7% (6/9) of patients with severe scoliosis. During the follow-up period, two patients changed from Type 1 to Type 2.

Conclusions: Scoliosis in PWS can be classified into three types. A lumbar curve convex on the left side was found in most patients. In addition, severe deterioration of scoliosis was found in Type 2 patients. Therefore we recommend careful, ongoing observations for patients showing double curve tendencies.