Skip to Content
MDPIMDPI

Advanced

by
  • Mayumi Higashi 1,*,
  • Masahiro Tanabe 1 and
  • Katsuyoshi Ito 1
  • et al.

Reviewer 1: Koji Takahashi Reviewer 2: Francesco Giangregorio

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This retrospective study evaluates the temporal shrinkage of the ablated area volume (AAV) after radiofrequency ablation (RFA) for hepatocellular carcinoma (HCC) using three-dimensional (3D) volumetric CT analysis. The authors demonstrated that a lower AAV reduction rate at approximately 6 months post-RFA was significantly associated with impaired liver function (assessed by Child-Pugh class, mALBI grade, albumin, and bilirubin) and advanced liver fibrosis (assessed by the FIB-4 index). They also found that lesion location (medial and anterior segments) affected the reduction rate. The clinical implication—that the shrinkage of the necrotic zone reflects the regenerative capacity of the surrounding liver parenchyma—is physiologically plausible and of clinical interest. However, several critical methodological and statistical issues must be addressed before this manuscript can be considered for publication.
Major 
1. The AAV was measured "in consensus" by one radiologist and one third-year medical student. Measuring the volume of coagulative necrosis at 6 months can be highly subjective due to ill-defined borders, surrounding parenchymal changes, or fibrosis. Relying solely on a consensus reading without reporting inter-observer variability is a significant methodological weakness. The authors should have two observers perform the segmentations independently and report the intraclass correlation coefficient (ICC) to demonstrate the reproducibility and robustness of their 3D volumetric measurements.
2. The study analyzed 53 lesions in 41 patients, meaning some patients contributed multiple lesions. This introduces a "cluster effect" (non-independence of data). Although the authors attempted to address this in the multivariable analysis using generalized estimating equations (GEE), the simple linear regression analyses (Table 2) and subgroup comparisons (Figures 2, 3, 5, and 6) appear to treat all 53 lesions as completely independent observations. Analyzing nested data as independent data artificially inflates the degrees of freedom and biases the p-values. The authors must strictly apply cluster-robust standard errors or mixed-effects models for all statistical comparisons, or perform a patient-level analysis (e.g., using the average reduction rate per patient) to ensure statistical validity.
3. While patients undergoing concurrent TACE were excluded, the authors did not clarify whether patients had a history of prior hepatectomy, TACE, or RFA in the same liver segment or lobe. Prior local treatments can significantly impair regional portal blood flow and induce localized fibrosis, which directly affects local liver regeneration and AAV shrinkage. The presence of prior treatments must be evaluated, adjusted for in the multivariable models, or at least discussed as a major limitation with detailed baseline data.
Minor
4. Table 1 presents clinical and laboratory data for "All patients (n=41)," but the footnote states that laboratory data were recorded at the time of each RFA session (which would correspond to 53 sessions or a different denominator if some lesions were treated simultaneously). If patients underwent multiple RFA sessions at different time points, summarizing their baseline laboratory data strictly at the "patient level (n=41)" is statistically problematic. The authors must clarify how they handled duplicate laboratory values for patients with multiple sessions in Table 1.
5. In Figure 6 and the text, the authors grouped the medial and anterior segments together, showing they had lower reduction rates compared to the lateral segment. While they attribute this to middle hepatic vein (MHV) drainage and susceptibility to congestion, the medial segment (S4) and anterior segments (S5/S8) are anatomically distinct. The authors should present the reduction rates for each segment individually, or provide a stronger, evidence-based justification for grouping these specific segments together.
6. The interval between the baseline CT and the follow-up CT ranged from 136 to 220 days (approximately 4.5 to 7.3 months). This is a wide variation that could directly confound the "reduction rate," as more shrinkage is expected at 220 days than at 136 days. The authors must perform a correlation analysis between the actual follow-up interval (in days) and the AAV reduction rate to confirm whether the variation in follow-up timing acted as a confounding factor.
7. With only 53 lesions (and effectively 41 independent subjects), entering multiple variables into the multivariable regression or GEE model risks overfitting. The authors should report the variance inflation factors (VIF) to rule out multicollinearity, especially since albumin and T-Bil are highly correlated and both are components of the mALBI/Child-Pugh scores.

Comments on the Quality of English Language

As I am not a native English speaker, I am not qualified to evaluate this item.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This review analyzes the manuscript "Factors affecting temporal changes in ablated liver volume after radiofrequency ablation for hepatocellular carcinoma evaluated by three-dimensional volumetric computed tomography" (Higashi et al.).

Conversational Analysis and Review

Let’s dive into what this paper brings to the table. In interventional oncology, we often focus intensely on the "immediate win"—achieving that perfect 5–10 mm safety margin right after the procedure to prevent local tumor progression (LTP). However, Higashi and colleagues have taken a fascinating "long-view" approach. They aren't just looking at the ablation zone's birth, but its involution over six months.

The authors used 3D volumetric CT analysis to track how the ablated area volume (AAV) shrinks. This is technically superior to old-school 2D measurements, which are notoriously subjective due to tissue deformation and patient positioning. Their key finding is both intuitive and scientifically grounded: ablation zones shrink slower in patients with poor liver function (Child-Pugh B or high ALBI grades) and advanced fibrosis (high FIB-4 index).

Why does this matter? It’s all about the hepatic regenerative capacity. A healthy liver actively heals and regenerates parenchyma around the "scar" of the ablation, causing rapid shrinkage. In a cirrhotic liver, this process is stunted. This isn't just academic; it has huge implications for how we read follow-up scans. If we don’t account for this "lazy shrinkage" in sick livers, we might misinterpret a large, persistent ablation zone as a sign of something more sinister, or conversely, underestimate the risk of LTP if we expect rapid involution that never comes.

Suggestions for Amelioration

While the study is solid, there are several areas where it could be strengthened to align with current international trends and guidelines:

  1. Address the "Margin Underestimation Dilemma": Recent research suggests that because ablation zones shrink mono-exponentially, an adequate margin measured at one month may appear inadequate at six months. The authors should discuss whether the rate of shrinkage they observed could lead to false-positive assessments of treatment failure in cirrhotic patients.
  2. Compare with Microwave Ablation (MWA): The study focuses on RFA, but current BCLC 2022 guidelines and recent reviews often favor MWA for its faster heating and reduced "heat-sink" effect. Since the authors mention MWA in their discussion, a brief sub-analysis or more detailed comparison regarding the involution kinetics of MWA vs. RFA (where MWA zones often persist longer) would add significant value.
  3. Refine Image Registration Details: The authors used manual tracing. To truly modernize the manuscript, they should discuss the role of deformable image registration (DIR). DIR is now the gold standard for comparing pre- and post-ablation volumes because it accounts for the liver’s non-rigid deformations.
  4. Biological Correlates: Mentioning the role of Neutrophil Extracellular Traps (NETs) or inflammatory markers beyond just basic LFTs could humanize the "why" behind the clinical data. Incomplete involution and chronic inflammation in the periablative zone can actually drive metastasis via inflammatory feedback.
  5. Technical Success vs. Involution: The authors should clarify if "technical success" (complete coverage) influenced the shrinkage rate. Did a larger initial AAV (to compensate for a difficult location) shrink differently than a standard one?

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The authors have addressed the reviewers' statistical and methodological concerns in a sincere and logical manner, significantly enhancing the scientific validity of the manuscript.
However, the Abstract still reflects the outdated analysis methods and results from the previous version. Please address the following omissions regarding necessary revisions:
1. Abstract - Methods:
Current text: "...using two sample t-test and one-way analysis of variance, and was correlated with clinical data using simple linear and multiple regression analyses."
Proposed revision: Please rewrite this section to state that inter-group comparisons and univariable/multivariable analyses were performed using Generalized Estimating Equations (GEE)—accounting for within-patient correlation—to align with the actual analysis performed.
2. Abstract - Results:
Regarding the comparison between mALBI grade 2b and grade 1 or 2a, the p-values ​​have been corrected to "p<0.001" and "p=0.004" in Fig. 3 (Version 2), yet the Abstract still lists "p<0.01." Please ensure the statistical values ​​or notation are consistent.
3. Abstract - Results:
Current text: "Simple linear regression showed significant correlations... Multiple regression identified albumin as..."
Proposed revision: Please revise the wording to reflect the GEE analysis—for example, "Univariable GEE analyses showed..." and "Multivariable GEE analysis demonstrated that both albumin and T-Bil remained independently associated..."
4. Abstract - Results:
The p-value for the comparison between FIB-4 index groups is listed as "p=0.040," whereas the revised Section 4.5 and Fig. 5 state "p<0.001." Please update the Abstract to align with the latest p-value derived from the GEE analysis.

Comments on the Quality of English Language

As I am not a native English speaker, I am not qualified to evaluate this item.

Author Response

Dear reviewer 1,

We greatly appreciate your review of our manuscript and your helpful suggestions. Below are our responses to the reviewer’s comments, with a description of the changes made to the manuscript.

  1. Abstract - Methods:
    Current text: "...using two sample t-test and one-way analysis of variance, and was correlated with clinical data using simple linear and multiple regression analyses."
    Proposed revision: Please rewrite this section to state that inter-group comparisons and univariable/multivariable analyses were performed using Generalized Estimating Equations (GEE)—accounting for within-patient correlation—to align with the actual analysis performed.

Response;

Thank you for pointing out these inconsistencies in the Abstract. We have revised the Methods section of the Abstract to reflect the revised statistical analyses by describing the use of generalized estimating equations (GEE) for group comparisons and univariable and multivariable analyses, as follows:

The AAV reduction rate was compared among Child-Pugh classification, modified albumin-bilirubin (mALBI) grades, FIB-4 index categories, and lesion locations using generalized estimating equations (GEE). Univariable and multivariable GEE analyses were performed to evaluate associations between the AAV reduction rate and clinical parameters.

  1. Abstract - Results:
    Regarding the comparison between mALBI grade 2b and grade 1 or 2a, the p-values ​​have been corrected to "p<0.001" and "p=0.004" in Fig. 3 (Version 2), yet the Abstract still lists "p<0.01." Please ensure the statistical values ​​or notation are consistent.
  2. Abstract - Results:
    Current text: "Simple linear regression showed significant correlations... Multiple regression identified albumin as..."
    Proposed revision: Please revise the wording to reflect the GEE analysis—for example, "Univariable GEE analyses showed..." and "Multivariable GEE analysis demonstrated that both albumin and T-Bil remained independently associated..."
    4. Abstract - Results:
    The p-value for the comparison between FIB-4 index groups is listed as "p=0.040," whereas the revised Section 4.5 and Fig. 5 state "p<0.001." Please update the Abstract to align with the latest p-value derived from the GEE analysis.

Response;

We have revised the Results section of the Abstract to ensure consistency with the updated GEE analyses presented in the revised manuscript. Specifically, we corrected the reported p-values for the mALBI grade and FIB-4 index comparisons and revised the descriptions of the statistical analyses to reflect the use of univariable and multivariable GEE analyses, as follows:

The AAV reduction rate was significantly lower in Child–Pugh class B than class A (p < 0.001) and in mALBI grade 2b than grade 1 (p < 0.001) or grade 2a (p = 0.004). Significant differences were also observed among FIB-4 index groups (p < 0.001) and among lesion locations, with lower AAV reduction rates in medial and anterior segments than in lateral (p < 0.001) and posterior (p = 0.017) segments. Univariable GEE analyses showed significant associations between AAV reduction rate and cholinesterase, albumin, total bilirubin (T-Bil), prothrombin time, platelet count, and FIB-4 index (p < 0.05). Multivariable GEE analysis demonstrated that both albumin and T-Bil remained independently associated with AAV reduction rate (p < 0.001).