Letter to the Editor: “Hepatocellular adenomas: is there additional value in using Gd-EOB-enhanced MRI for subtype differentiation?”

by Edouard Reizine, Maxime Ronot , Alain Luciani (ereizine@gmail.com)

Hepatocellular adenomas: is there additional value in using Gd-EOB-enhanced MRI for subtype differentiation?

Dear Editor,

We have read with great interest the article entitled “Hepatocellular adenomas: is there additional value in using Gd-EOB-enhanced MRI for subtype differentiation?” by Auer and colleagues published recently in European Radiology [1]. We would first like to compliment the authors on this well designed and interesting study, including a large cohort of hepatocellular adenomas (HCA) focusing on the added value of Gd-EOB MRI for subtype differentiation. However, we do believe that this study raises some issues which should be shared with readers of this article.

The authors included 48 patients with a total of 79 pathologically proven HCAs. They reported that 50.5% (40/79) of all HCA were considered as significantly enhancing on HBP and that inflammatory HCA (IHCA) was the subtype associated with the higher rate of uptake on HBP (89% of the lesions with more than 50% intralesional Gd-EOB uptake in the HBP were IHCA). Hence, the authors reported that combining Gd-EOB-specific and established MRI qualitative features improved accuracy for HCA subtyping, particularly for the diagnostic of ß-catenin mutated ones (bHCA), as the combination of significant uptake and a central scar resulted in 80% sensitivity and 92% specificity for the identification of this subtype.

We agree with the authors that signal hyperintensity on HBP is not uncommon in HCA, especially in IHCA, and previous studies have already reported rates similar to that reported by authors [2, 3]. However, we are convinced that hyperintensity on HBP should be analyzed with caution in HCA, particularly in case of hepatic steatosis. In our experience, inflammatory HCA showing hyperintensity on HBP are encountered almost exclusively in patients with underlying hepatic steatosis [4]. Therefore, we strongly believe that uptake assessment should necessarily consider both the signal of the lesion and of the liver on pre-contrast sequences, whether the uptake is analyzed visually [5] or quantitatively [6]. As a consequence, it is difficult to fully apprehend their data without information on the signal of the lesions and of the liver on pre-contrast images, together with the presence of hepatic steatosis in the studied population. In addition, reproducibility of the technique used for the assessment of signal change on HBP could be challenging in clinical practice.

Second, authors decided to focus on the atoll sign for the diagnosis of IHCA. However, several studies showed that IHCA are characterized by a combination of two main features: marked hyperintensity on T2W and persistent delayed enhancement [7, 8]. Interestingly, a recent study showed that the persistent delayed enhancement may be missing using EOB, leading to lower subtyping abilities with EOB [9]. Therefore, since authors used EOB-DTPA, they could not apply the main features. Although adding HBP analysis may improve the diagnosis of IHCA the diagnostic performance of EOB DTPA was not compared with those obtained with extracellular contrast agent or Gd-BOPTA, that may very well be similar.

Third, identifying HCA with high ß-catenin activation is a major challenge as this subtype is known to have a high risk of malignancy into hepatocellular carcinoma [10], and results of this current study are promising. However, as pointed out by authors, several ß-catenin mutations have been described in HCA [10]. Importantly, those mutations are associated with different level of ß-catenin activation, and different profiles on immunohistochemistry (IHC) have been described according to the different mutations [11]. Hence, it would also be interesting for readers if IHC subtyping could be provided if possible. As for the value of the ‘central scar’, its absence is frequently reported in bHCA [12] and in addition, as reported by the authors, it is also encountered in IHCA [13]. Hence the combination of uptake on HBP, as measured in this study, and a central scar is expected to have insufficient performance for the diagnosis of bHCA.