Mixed Results With MRI for Liver Triglycerides in NAFLD

This story was originally published July 25, 2018. As part of MedPage Today’s year-end review of 2018’s top stories, we are republishing it along with an update on what has happened since with radiological diagnosis of NAFLD and NASH.

Proton density fat fraction on MRI (MRI-PDFF) remains a useful tool for the noninvasive detection of nonalcoholic fatty liver disease (NAFLD), and quantification of steatosis over time, according to a retrospective study.

MRIs were performed at 44 clinical sites in the U.S. and Australia and histological data were available for 170 patients (45.9%). The mean baseline PDFF was 17.4% and the median was 16.0% (interquartile range 10.3-23.48). Baseline PDFF values correlated with several histopathology parameters, including positively for steatosis grade (r=0.78, P<0.001), NAFLD activity score (NAS; r=0.54, P<0.001), and negatively for fibrosis stage (r= –0.59, P<0.001), reported Benjamin M. Wildman-Tobriner, MD, of Duke University Medical Center in Durham, North Carolina, and colleagues.

But this fat-mapping technique was not sufficiently accurate to distinguish between NAFLD and nonalcoholic steatohepatitis (NASH), either with early or advanced fibrosis and cannot replace the gold standard of percutaneous biopsy, they noted in Gastroenterology.

“Imaging techniques such as MRI-PDFF can quantify liver fat levels in patients with NAFLD and NASH, though our data suggest that liver biopsy remains a necessary part of work-up in order to fully characterize the extent of disease activity,” Wildman-Tobriner told MedPage Today. “Combining MRI-PDFF with other noninvasive markers of disease may help reduce the need for biopsy in the future, but further studies are needed.”

The investigators looked at pooled data from three recent phase II trials of pharmacotherapy for NAFLD or NASH, assessing the relationship between PDFF values and liver histologic findings. All three parent trials required pre-treatment MRI-PDFF estimation, with the minimum PDFF thresholds for trial inclusion ranging from 6% to 15%. Patients with other chronic liver diseases or uncontrolled diabetes were excluded.

The 370 adult patients from the combined trials had a mean age of 51. They were 60.1% female and 88.4% white with a mean BMI 33.3. Sex, ethnicity, and baseline PDFF values were similar across the three trials, while small but statistically significant differences existed between trials in age, BMI, race, and some laboratory and histological parameters.

For accurately identifying patients with NAS ≥4, the area under receiver operating curve (AUROC) was 0.72 (95% CI 0.59-0.84). The optimal PDFF threshold (Youden index) for predicting an NAS ≥4 was 12.4%, which resulted in a sensitivity of 0.75 (95% CI 0.35-0.97), a specificity of 0.75 (95% CI 0.67-0.81), a positive predictive value (PPV) of 0.13 (95% CI 0.05-0.26), and a negative predictive value (NPV) of 0.98 (95% CI 0.94-1).

For identifying more advanced fibrosis stages of ≥3, the AUROC was 0.66 (95 % CI 0.57-0.74). The optimal threshold for predicting stage ≥3 was 20.4%, for a sensitivity of 0.53 (95% CI 0.44-0.62), a specificity of 0.17 (95% CI 0.08-0.31), a PPV 0.63 (95% CI 0.53-0.73), and an NPV 0.14 (95% CI 0.06-0.22). Baseline PDFF did not significantly correlate with ballooning or lobular inflammation.

Finding a substantial overlap in PDFF values between patients with NAFLD versus NASH, as well as between NAFLD patients with versus without fibrosis, “PDFF does not appear to be a strong predictor of advanced liver fibrosis or NASH for an individual patient, but may be more useful for risk stratification,” the authors stated.

While the study found MRI-PDFF correlated positively with steatosis grade and NASH and inversely with fibrosis stage, it was suboptimal in identifying patients with higher NAFLD activity scores (NAS) of ≥4 or with advanced fibrosis. The NAS, ranging from 0 to 8, is the sum of scores for steatosis and lobular inflammation and ballooning.

“Although MRI-PDFF is an important imaging biomarker for continued evaluation of this patient population, liver biopsy analysis is still necessary,” Wildman-Tobriner’s group wrote. They noted that NAFLD currently affects an estimated 20% of the U.S. population and NASH an estimated 3%-12%, highlighting the need for accurate noninvasive assessment measures.

The authors noted that early-stage clinical trials accepted steatosis improvement as measured by MRI-PDFF as a surrogate endpoint for NASH improvement. “However, our data suggest it remains to be seen whether histological response to drug therapy will be well represented by PDFF changes,” they wrote.

In 2017, co-author Michael M. Middleton, MD, PhD, of the University of California San Diego, and colleagues reported that MRI-PDFF accurately classified grades and changes in hepatic steatosis in relation to reference biopsy.

The current authors said techniques other than, or combined with, MRI-PDFF such as ultrasound-based, MRI elastography, or liver biopsy remain necessary for assessing hepatic parenchymal changes other than steatosis.

“Consequently, an MRI performed to assess for hepatic steatosis may be helpful if negative, but an examination demonstrating abnormally high fat content does not imply NASH or the presence of liver fibrosis, and further assessment using non-invasive tools or biopsy may be needed,” they wrote.

Study limitations included its retrospective nature and the lack of biopsy results and complete laboratory data for some trial patients. Also, the findings may not be generalizable to patients with cirrhosis, who were excluded from the three parent trials. Finally, the histopathological assessment of liver iron was available in only a very small number of patients, as were MR elastography values.