Reply to the Letter to the Editor: “Non-hypervascular hepatobiliary phase hypointense nodules on gadoxetic acid-enhanced MR can help determine the treatment method for HCC”
by Dong Ho Lee, Jeong Min Lee (jmsh@snu.ac.kr)
Non-hypervascular hepatobiliary phase hypointense nodules on gadoxetic acid-enhanced MR can help determine the treatment method for HCCDear Editor,
First of all, we would like to thank Dr. Wang and his colleagues for their interest in our recent study published in European Radiology [1]. They pointed out important errors in our Figure legend 4a and 4b, and we would like to correct these errors. Figure 4a demonstrates the difference in Kaplan-Meier estimation of the recurrence-free survival (RFS) between 105 patients without non-hypervascular hepatobiliary phase (HBP) hypointense nodule after hepatic resection (HR) and 159 patients without non-hypervascular HBP hypointense nodule after radiofrequency ablation (RFA). Apparently, Figure 4b shows the comparison of RFS in 18 patients with non-hypervascular HBP hypointense nodules after HR to RFS in 63 patients with non-hypervascular HBP hypointense nodules after RFA. Indeed, the 1-year RFS rate in 63 patients with non-hypervascular HBP hypointense nodules after RFA is ‘84.1%’, and not ‘4.1%’. We apologize for these important errors in our figure legends and result and would like to correct them.
Regarding the overall survival (OS) rates, the presence of non-hypervascular HBP hypointense nodules was not a significant affecting factor for OS after both HR (p = 0.110) and RFA (p = 0.760), and this result was well correlated with the result of our previous study [2]. In our study, there was no statistically significant difference in the cumulative incidence of local tumor progression (LTP), intrahepatic distant recurrence (IDR), and extrahepatic metastasis (EM) between HR and RFA in patients with non-hypervascular HBP hypointense nodule, as Wang et al. inferred.
As Wang et al. pointed out, other imaging modalities such as contrast enhanced CT or contrast enhanced US can reveal the hypovascular nodule in the liver. However, in our study, we focused on nodule showing hyposignal intensity on HBP of gadoxetic acid-enhanced liver MR without arterial phase hyper-enhancement (APHE), and this nodule does not mean that it is hypovascular. Iso-enhancement during the arterial phase or portal venous phase can also be seen for non-hypervascular HBP hypointense nodules on gadoxetic acid-enhanced MR. According to the results of previous studies, non-hypervascular HBP hypointense nodule turned out to be either early HCC or dysplastic nodule (DN) on histopathologic examination [3-5], and gadoxetic acid enhanced liver MR is the best imaging modality to detect these borderline hepatocellular nodules including DN, by providing hepatobiliary phase: this was the reason why we chose gadoxetic acid enhanced liver MR over other imaging modalities. Several previous studies reported that non-hypervascular HBP hypointense nodule which could represent borderline hepatocellular nodule, could progress into hypervascular HCC during the follow-up period with the conversion rate of 26.5%-35.0% [2, 6-8]. Subsequently, it means that about 70% of non-hypervascular HBP hypointense nodules could remain stable, and do not progress into hypervascular HCC during the follow-up. Therefore, immediate treatment of non-hypervascular HBP hypointense nodules found on gadoxetic acid-enhanced liver MR would inevitably raise the concern of overtreatment issues as all treatment including HR and RFA, has its own complication and cost. Indeed, according to the recent HCC management guideline proposed by EASL and AASLD, non-hypervascular HBP hypointense nodules should not be diagnosed as HCC due to the lack of APHE [9, 10]. Up to now, the only method to accurately diagnose the non-hypervascular HBP hypointense nodule is liver biopsy with histopathologic examination. However, non-hypervascular HBP hypointense nodules found on gadoxetic acid-enhanced liver MR is usually small, and the mean size of non-hypervascular HBP hypointense nodules in this study was only 8.2 mm. The accurate diagnosis via liver biopsy for this small size nodule is quite difficult, and therefore, liver biopsy for all non-hypervascular HBP hypointense nodules found on gadoxetic acid enhanced liver MR is hardly possible in routine clinical practice.
There have been several studies regarding the impact of immediate treatment of concomitant non-hypervascular HBP hypointense nodules during the HCC treatment on clinical outcomes, and the study results are controversial. Midorikawa et al. reported that there was no significant difference in OS between careful observation and immediate resection for hypovascular tumor associated with HCC, and concluded that immediate treatment of hypovascular tumor was not necessary [11]. In contrast to that, Matsuda et al. reported that concomitant resection of non-hypervascular HBP hypointense nodules during HCC resection could significantly improve RFS compared to close observation even though there was no significant difference in OS [12]. Therefore, we would think that there has been no strong evidence supporting immediate treatment for non-hypervascular HBP hypointense nodules until now. We agree with Wang, and his colleagues’ opinion that non-hypervascular HBP hypointense nodules should be taken into account in clinical practice as the presence of non-hypervascular HBP hypointense nodules is significant predictive of tumor recurrence after treatment for HCC. However, for solid conclusion and treatment guideline about the management for non-hypervascular HBP hypointense nodules, we still need more sophisticated studies with prospective design and strong evidence. In this regard, risk stratification of non-hypervascular HBP hypointense nodules would be important, as it could guide the management plan, including liver biopsy, for an accurate diagnosis. For instance, according to the results of previous studies, T2 hyper-intensity of nodule and diffusion restriction within nodule were significantly associated with the progression into hypervascular HCC during the follow-up [6, 7]. Therefore, when we encounter non-hypervascular HBP hypointense nodules with T2 hyper-intensity or diffusion restriction, we would actively consider the liver biopsy for accurate diagnosis and subsequent treatment.
In summary, the presence of non-hypervascular HBP hypointense nodules on gadoxetic acid-enhanced liver MR is a significant determinant factor for tumor recurrence after HCC treatment. However, for management guideline for these nodules, further studies including risk stratification as well as treatment of these nodules, are warranted.