Letter to the editor: Change in serial liver stiffness measurement by magnetic resonance elastography and outcomes in NAFLD

To the editor, We read with interest the study by Gidener et al.1 on the value of repeated liver stiffness measurements (LSMs) by magnetic resonance elastography (MRE) for prediction of outcomes in NAFLD. We congratulate the researchers on their statistically sound approach to capture dynamics in LSM and would like to discuss some important aspects of this study. First, the lack of knowledge on the reason for repeating the MRE‐LSM combined with the retrospective design of the study may have introduced selection, confirmation, and survival bias. For example, there is no knowledge of the clinical reason for repeating the MRE measurement, making those selected for and/or survived to have a second LSM an indeed selected cohort. Also, no data on the total number of NAFLD patients who were potentially eligible for a repeated MRE‐LSM are given. With this regard, an inception cohort would allow to compare the prognostic ability of baseline MRE‐LSM with serial changes. Second, the small sample size limits generalizability of the results: The 100% rate of hepatic decompensation was derived from 3 patients from a total of only 29 patients with compensated cirrhosis at last MRE‐LSM, including 2 patients who developed jaundice as an incident event. Similarly, associations of progression with “incident cirrhosis” were derived from 1 in 7 progressors who did not have cirrhosis at last MRE‐LSM. It could be suspected that patients with increasing LSMs are more likely to undergo further investigations and thereby get diagnosed with “incidence cirrhosis.” Therefore, larger studies are needed to validate the relevance of “progression” that go beyond the 19% increase cutoff. Although this cutoff has been derived from the repeatability coefficient in MRE2 and was proposed to be corresponding to a “true” change in LSM, its clinical significance remains to be validated. Third, whereas only hypertension was associated with changes in LSM, further elaboration/discussion of this finding could limit speculations about unmeasured confounders given the limited pathophysiological evidence for a causal connection between arterial hypertension and fibrosis progression. Fourth, although the researchers assessed longitudinal changes of MRE‐LSM in NAFLD, it would be interesting to see the results of the subgroup of patients with steatohepatitis being prone to fibrosis progression and liver‐related outcomes. Also, 24 patients with decompensated cirrhosis were included in this study, but neither separately analyzed nor further discussed. Finally, although MRE‐LSM showed a higher accuracy for fibrosis staging compared to vibration‐controlled transient elastography (VCTE), current guidelines do not recommend its routine nor repeated use because of limited availability and unfavorable cost‐effectiveness. With this regard, similar studies investigating the prognostic relevance of dynamics in VCTE‐based LSM are warranted.