Letter to the Editor: “Non-hypervascular hepatobiliary phase hypointense nodules on gadoxetic acid-enhanced MR can help determine the treatment method for HCC”

by Yining Wang, Danjun Song, Lingyu Tian, Xiaoying Wang

Non-hypervascular hepatobiliary phase hypointense nodules on gadoxetic acid-enhanced MR can help determine the treatment method for HCC

Dear Editor,

We read with interest the article by Dong Ho Lee et al. entitled “Non-hypervascular hepatobiliary phase hypointense nodules on gadoxetic acid-enhanced MR can help determine the treatment method for HCC” [1]. We would like to thank authors for serial work and contributions concerning hypovascular nodules or HCC but there are some limitations of critical importance that we are supposed to emphasize.

First, the legends of Figure 4a and 4b were inadvertently reversed. Figure 4a is supposed to demonstrate the Kaplan-Meier estimation of the RFS in 105 patients WITHOUT nonhypervascualr HBP hypointense nodules after HR compared with 159 patients WITHOUT nonhypointense nodules after RFA. Figure 4b is supposed to compare RFS in 18 patients WITH hypointense nodules after HR and 63 patients WITH hypointense nodules after RFA. In addition, the 1-year RFS rate is supposed to be “84.1%” not “4.1%” in 63 patients with hypointense nodules after RFA in the last paragraph of Results. For HCC concomitant with hypointense nodules, there is no significant difference in RFS between the HR group and RFA group. We infer there is no statistical difference in IDR, LTP, EM as well but these results were not reported in their article.

Second, for HCC concomitant with hypointense nodules, their study only included 18 patients after HR and 63 patients after RFA. The sample size is small to implement PSM so whether there is no prognostic difference between RFA and HR for HCC concomitant with hypovascular nodules is required to be further investigated. And the focus of this research is hypovascular nodules. Although gadoxetic acid–enhanced MRI is superior to triphasic dynamic enhanced multislice CT (MSCT) [2, 3] and contrast-enhanced ultrasonography (CEUS) [4] in the specificity, accuracy and consistency of HCC diagnosis especially for < 1cm HCC, this retrospective study did not include hypovascular nodules detected by other imaging measures, which may be one of causes of data loss.

Third, the significance of hypovascular nodules or HCC has been gradually proven and acknowledged. Nonhypervascular HBP hypointense nodules are thought to represent borderline hepatocellular nodules including high-grade dysplastic nodules or early HCCs which have the potential to progress into typical hypervascular HCCs [1, 5, 6]. Early studies have demonstrated a 26.5%-35.0% conversion rate of nonhypervascular HBP hypointense nodules progressing into typical hypervascular HCC [7-10]. In this study, they could not exclude a malignant risk of hypovascular nodules detected by gadoxetic acid–enhanced MRI. Therefore, it is not ethically rational without interventions. We suggest pathological examinations and preventive interventions for these hypovascular nodules rather than no intervention and follow-up observation adopted by Lee et al. But we consider the cohort of this study included consecutive patients from 2009 to 2014 when clinicians did not recognize hypovascular nodules as a significant predictor for survival and recurrence. In addition, this study did not show and compare the data regarding overall survival rate although earlier studies of Lee et al. concluded that concomitant hypovascular nodules did not impact the OS of patients with HCC [11], which could be up to 97.0% after RFA [12].

For these reasons, we should reconsider the clinical strategy of concurrent hypovascular nodules. It is required to be confirmed whether the prognosis of one-stage resection or RFA is better for HCC concomitant with hypovascular nodules. Although hypovascular nodules have been generally acknowledged as a poor prognostic factor [6, 8, 11, 13-15], the optimal therapeutic option is still controversial for less aggressive hypovascular HCCs until now [16, 17]. No significant difference in OS and RFS was validated between the surgery group and RFA group for small solitary hypovascular HCC ≤ 2 cm [18]. And the optimal treatment is supposed to be further proven for hypovascular HCC (2-3cm). Recently, Midorikawa et al. [19] concluded that early hypovascular HCC were not appropriate for surgery because HR for early hypovascular HCCs provided an only marginal survival benefit. Besides, even the necessity of transplantation for hypovascular HCC has been identified [20].

In summary, we think hypovascular nodules or HCC should be evaluated and taken seriously in clinical practice.