Imaging Biomarkers of the Physis: Cartilage Volume on MRI vs. Tract Volume and Length on Diffusion Tensor Imaging.

Abstract

Current methods to predict height and growth failure are imprecise. MRI measures of physeal cartilage are promising biomarkers for growth. In the physis, to assess how 3D MRI volume measurements, and diffusion tensor imaging (DTI) measurements (tract volume and length) correlate with growth parameters and detect differences in growth. We compared patients exposed to cis-retinoic acid, which causes physeal damage and growth failure, with normal subjects. Case-control. Twenty pediatric neuroblastoma survivors treated with cis-retinoic acid and 20 age- and sex-matched controls. 3T; DTI and 3D double-echo steady-state (DESS) sequences. On distal femoral MR studies, physeal 3D volume and DTI tract measurements were calculated and compared to height. We used partial Spearman correlation, analysis of covariance, logistic regression, Wald test, and the intraclass correlation coefficient (ICC). The height percentile correlated most strongly with DTI tract volumes (r = 0.74), followed by mean tract length (r = 0.53) and 3D volume (r = 0.40) (all P < 0.02). Only tract volumes and lengths correlated with annualized growth velocity. Relative to controls, patients showed smaller tract volumes (8.00 cc vs. 13.71 cc, P < 0.01), shorter tract lengths (5.92 mm vs 6.99 mm, P = 0.03), and smaller ratios of 3D cartilage volume to tract length; but no difference (4.51 cc vs 4.85 cc) in 3D MRI volumes. The 10 patients with the lowest height percentiles had smaller tract volumes (5.07 cc vs. 10.93 cc, P < 0.01), but not significantly different 3D MRI volumes. Tract volume is associated with abnormal growth, with an accuracy of 75%. DTI tract volume of the physis/metaphysis predicts abnormal growth better than physeal cartilage volumetric measurement and correlates best with height percentile and growth velocity. 2 TECHNICAL EFFICACY: Stage 2 J. Magn. Reson. Imaging 2020;52:544-551.