Fatty degeneration of the rotator cuff muscles on pre- and postoperative CT arthrography (CTA): is the Goutallier grading system reliable?

• DOI: 10.1007/s00256-013-1660-1
• Publication Date: 2013-06-20T06:26:32Z
• Authors: Eugene Lee; Jung-Ah Choi; Joo Han Oh; Soyeon Ahn; Sung Hwan Hong; Jee Won Chai; Heung Sik Kang
• Keywords: Hounsfield scale; Supraspinatus muscle; Grading (engineering)

To retrospectively evaluate fatty degeneration (FD) of rotator cuff muscles on CTA using Goutallier's grading system and quantitative measurements with comparison between pre- and postoperative states.IRB approval was obtained for this study. Two radiologists independently reviewed pre- and postoperative CTAs of 43 patients (24 males and 19 females, mean age, 58.1 years) with 46 shoulders confirmed as full-thickness tears with random distribution. FD of supraspinatus, infraspinatus/teres minor, and subscapularis was assessed using Goutallier's system and by quantitative measurements of Hounsfield units (HUs) on sagittal images. Changes in FD grades and HUs were compared between pre- and postoperative CTAs and analyzed with respect to preoperative tear size and postoperative cuff integrity. The correlations between qualitative grades and quantitative measurements and their inter-observer reliabilities were also assessed.There was statistically significant correlation between FD grades and HU measurements of all muscles on pre- and postoperative CTA (p < 0.05). Inter-observer reliability of fatty degeneration grades were excellent to substantial on both pre- and postoperative CTA in supraspinatus (0.8685 and 0.8535) and subscapularis muscles (0.7777 and 0.7972), but fair in infraspinatus/teres minor muscles (0.5791 and 0.5740); however, quantitative Hounsfield units measurements showed excellent reliability for all muscles (ICC: 0.7950 and 0.9346 for SST, 0.7922 and 0.8492 for SSC, and 0.9254 and 0.9052 for IST/TM). No muscle showed improvement of fatty degeneration after surgical repair on qualitative and quantitative assessments; there was no difference in changes of fatty degeneration after surgical repair according to preoperative tear size and post-operative cuff integrity (p > 0.05). The average dose-length product (DLP, mGy · cm) was 365.2 mGy · cm (range, 323.8-417.2 mGy · cm) and estimated average effective dose was 5.1 mSv.Goutallier grades correlated well with HUs of rotator cuff muscles. Reliability was excellent for both systems, except for FD grade of IST/TM muscles, which may be more reliably assessed using quantitative measurements.