What Is the Correlation Among dGEMRIC, T1p, and T2*Quantitative MRI Cartilage Mapping Techniques in Developmental Hip Dysplasia?

Background Delayed gadolinium-enhanced MRI of cartilage (dGEMRIC) is a validated technique for evaluating cartilage health in developmental dysplasia of the hip (DDH), which can be a helpful prognosticator for the response to surgical treatments. dGEMRIC requires intravenous injection of gadolinium contrast, however, which adds time, expense, and possible adverse reactions to the imaging procedure. Newer MRI cartilage mapping techniques such as T1 rho (rho) and T2* have been performed in the hip without the need for any contrast, although it is unknown whether they are equivalent to dGEMRIC. Question/purpose In this study, our purpose was to determine the correlation between the relaxation values of three cartilage mapping techniques, dGEMRIC, T1 rho, and T2*, in patients with DDH. Methods Fifteen patients with DDH (three male, 12 female; mean age 29 +/- 9 years) scheduled for periacetabular osteotomy underwent preoperative dGEMRIC, T1 rho, and T2* MRI at 3T with quantitative cartilage mapping. The outcomes of dGEMRIC, T1 rho, and T2* mapping were calculated for three regions of interest (ROI) to analyze the weightbearing cartilage of the hip: global ROI, anterior and posterior ROI, and further subdivided into medial, intermediate, and lateral to generate six smaller ROIs. The correlation between the respective relaxation time values was evaluated using the Spearman correlation coefficient (r(S)) for each ROI, categorized as negligible, weak, moderate, strong, or very strong. The relaxation values within the subdivided ROIs were compared for each of the three cartilage mapping techniques using the Kruskal-Wallis test. Results There was a moderate correlation of T1 rho and T2* relaxation values with dGEMRIC relaxation values. For the global ROI, there was a moderate correlation between dGEMRIC and T2* (moderate; r(S) = 0.63; p = 0.01). For the anterior ROI, a moderate or strong correlation was found between dGEMRIC and both T1 rho and T2*: dGEMRIC and T1 rho (strong; r(S) = -0.71; p = 0.003) and dGEMRIC and T2* (moderate; r(S) = 0.69; p = 0.004). There were no correlations for the posterior ROI. The mean dGEMRIC, T1 rho, and T2* relaxation values were not different between the anterior and posterior ROIs nor between the subdivided six ROIs. Conclusion Quantitative T1 rho and T2* cartilage mapping demonstrated a moderate correlation with dGEMRIC, anteriorly and globally, respectively. However, the clinical relevance of such a correlation remains unclear. Further research investigating the correlation of these two noncontrast techniques with clinical function and outcome scores is needed before broad implementation in the preoperative investigation of DDH.