Prognosis value of early diffusion MRI in Legg Perthes Calvé disease

C Baunin1, D Sanmartin-Viron1, F Accadbled2

  • 1Department of Paediatric Imaging, hôpital des Enfants, CHU de Toulouse, 31059 Toulouse, France.

Insights

Diffusion MRI can predict Legg Calvé Perthes disease prognosis. An apparent diffusion coefficient (ADC) ratio of the femoral metaphysis above 1.63 indicates a poor outcome, offering early insights before traditional methods.

Area of Science:

  • Orthopedic imaging
  • Pediatric radiology
  • Biomedical engineering

Background:

  • Legg Calvé Perthes (LCP) disease affects the proximal femoral epiphysis in children.
  • Accurate prognosis is crucial for effective treatment planning.
  • Current prognostic indicators may have limitations in early stages.

Purpose of the Study:

  • To assess the utility of diffusion MRI in predicting the prognosis of LCP disease.
  • To evaluate the apparent diffusion coefficient (ADC) of the proximal femoral epiphysis and metaphysis.
  • To correlate ADC measurements with established prognostic classifications.

Main Methods:

  • Prospective study of 31 children with unilateral LCP disease.
  • MRI scans performed during condensation or fragmentation stages.
  • Bilateral measurement of ADC in femoral epiphysis and metaphysis, calculating ADC ratios.
  • Comparison of ADC ratios with Herring classification.

Main Results:

  • Increased ADC in affected hips compared to unaffected sides (epiphysis P<0.001, metaphysis P<0.0001).
  • Femoral metaphysis ADC ratio > 1.63 correlated with poor prognosis (Herring B-C or C) with 89% sensitivity and 58% specificity.
  • Excellent interobserver reliability for ADC measurements.
  • The 1.63 threshold was identifiable as early as the condensation stage.

Conclusions:

  • Diffusion MRI is a non-invasive, non-radiating imaging technique.
  • ADC ratio of the femoral metaphysis shows promise as an early prognostic factor for LCP disease.
  • This method may offer prognostic information prior to the applicability of Herring classification.
Abstract