Assessing Hepatic Fat with Transient Elastography or PCOS-HS Score in Youth with PCOS and Obesity

• Published in: Obesity (Silver Spring, Md.) Vol. 29; p. 59
• Main Authors: Sundararajan, Divya, Morelli, Nazeen, Garcia-Reyes, Yesenia, Ware, Meredith, Rahat, Haseeb, Sundaram, Shikha, Severn, Cameron, Pyle, Laura, Cree-Green, Melanie
• Format: Journal Article
• Language: English
• Published: Silver Spring Blackwell Publishing Ltd 01-12-2021
• Subjects: Obesity; Polycystic ovary syndrome
• ISSN: 1930-7381; 1930-739X

Background: Polycystic ovary syndrome (PCOS) affects 5-15% of reproductive aged women, and nonalcoholic fatty liver disease (NAFLD) is a common comorbidity. Diagnostic methods for NAFLD are expensive or invasive, and few lab-based predictive formulas are validated in adolescents. Recently, transient elastography has been used to diagnose hepatic steatosis (HS). We sought to determine if transient elastography accurately diagnosed HS as compared to MRI or PCOS-HS score in teens with PCOS + obesity. Methods: Teens with PCOS + obesity studied in 2 clinical trials were included. Liver fat was assessed with 3 methods: MRI proton density fat fraction, transient elastography controlled attenuation parameter (CAP) score and calculated PCOS-HS score, 1/(1+(exp(-(25.19 +(-0.3411· BMI percentile) + (0.06149· waist circumference (cm)) + (0.09374·alanine aminotransferase (U/L)) + (-0.07954·SHBG (nmo-l/L))). Scans and fasting measures for HS score were collected within a 24-hr period. HS was defined as MRI fat fraction >5.5%. CAP and PCOS-HS scores for those with and without HS by MRI were compared with Mann Whitney U tests. Receiver operator curves (ROC) for CAP and PCOS-HS were generated based on MRI determined HS. Results: Data from 41 girls (age 15.8 ± 1.7 years, BMI 36.2 ± 6.15 kg/m2) were included. Twenty-four had HS by MRI, with liver fat different by group (9.4 [7.4,14.9]% HS vs 3.8 [2.3,5.0]; p < 0.001). CAP scores were not different by MRI based HS status (300 [269,369] dB/m HS vs (269 [224,34]; p = 0.133], whereas the PCOS-HS was (0.59 [0.27,0.79] HS vs (0.13 [0.10,0.15]; p = 0.001). For CAP, ROC was not significant (r = 0.68, p = 0.128) and the Youden index had a sensitivity of 64% and a specificity of 62% at a score of 290 dB/m. For PCOS-HS, ROC was significant (r = 0.91, p = 0.002), with a sensitivity of 93% and specificity of 87% at a 0.16 cut-off. Conclusions: In teens with significant obesity, the transient elastography is not an accurate measure of HS compared to MRI. A CAP score of 290 dB/m to identify HS is higher than previously described in pediatrics, and it would be even higher for screening test requirements. The PCOS-HS score has good performance as a surrogate measure to identify those likely to have HS in this patient population. Additional studies in males and non-PCOS females with significant obesity are needed to evaluate these surrogate measures further.