Juvenile spondylolysis: a comparative analysis of CT, SPECT and MRI

Skeletal Radiol. 2005 Feb;34(2):63-73. doi: 10.1007/s00256-004-0878-3. Epub 2004 Nov 25.

Abstract

Objective: To evaluate whether MRI correlates with CT and SPECT imaging for the diagnosis of juvenile spondylolysis, and to determine whether MRI can be used as an exclusive image modality.

Design and patients: Juveniles and young adults with a history of extension low back pain were evaluated by MRI, CT and SPECT imaging. All images were reviewed blindly. Correlative analyses included CT vs MRI for morphological grading and SPECT vs MRI for functional grading. Finally, an overall grading system compared MRI vs CT and SPECT combined. Statistical analysis was performed using the kappa statistic.

Results: Seventy-two patients (mean age 16 years) were recruited. Forty pars defects were identified in 22 patients (31%), of which 25 were chronic non-union, five acute complete defects and ten acute incomplete fractures. Kappa scores demonstrated a high level of agreement for all comparative analyses. MRI vs SPECT (kappa: 0.794), MRI vs CT (kappa: 0.829) and MRI vs CT/SPECT (kappa: 0.786). The main causes of discrepancy were between MRI and SPECT for the diagnosis of stress reaction in the absence of overt fracture, and distinguishing incomplete fractures from intact pars or complete defects.

Conclusions: MRI can be used as an effective and reliable first-line image modality for diagnosis of juvenile spondylolysis. However, localised CT is recommended as a supplementary examination in selected cases as a baseline for assessment of healing and for evaluation of indeterminate cases.

Publication types

  • Comparative Study
  • Evaluation Study

MeSH terms

  • Adolescent
  • Adult
  • Child
  • Female
  • Humans
  • Image Interpretation, Computer-Assisted
  • Image Processing, Computer-Assisted
  • Lumbar Vertebrae / diagnostic imaging
  • Lumbar Vertebrae / pathology
  • Magnetic Resonance Imaging
  • Male
  • Spondylolysis / diagnosis*
  • Tomography, Emission-Computed, Single-Photon
  • Tomography, X-Ray Computed