Serum Level of Glypican-3 in Patients with Hepatocellular Carcinoma and Advanced Chronic Liver Disease: A Pilot Study

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Authors have done lot of work for this project, however the findings interpreted needs to be presented with proper cut-off values when comparing to existing tutor markers

As mentioned in the comments, authors should address why the cut-off values for AFP was taken ?

Did the authors use any validation cohort with this findings to confirm the findings?

Studies on GPC3 have been published before so this is not novel, authors should mention this.

 

Comments for author File: Comments.pdf

Author Response

Comment 1: Guidelines do not recommend use of all these tumor markers . PLease correct

Response 1: I agree with your comment and will correct the introduction accordingly. More information is added between line 42-53 of the paper.

The AFP + AFP-L3 + DCP combination are discussed in Japanese protocol for the diagnosis of HCC.  

“In Japan, liver ultrasonography (US) plus alpha-fetoprotein (AFP), des-gamma-carboxy prothrombin (DCP) and AFP-L3 testing are recommended every six months for the high-risk group (HBV/HCV infection, cirrhosis of other etiologies) and every three to four months for the extremely high-risk group (HBV/HCV-related cirrhosis)”

Cho Y, Kim BH, Park JW. Overview of Asian clinical practice guidelines for the management of hepatocellular carcinoma: An Asian perspective comparison. Clin Mol Hepatol. 2023 Apr;29(2):252-262. doi: 10.3350/cmh.2023.0099.

Comment 2: Amend this, as this is not new, use of GPC3 in HCC has been published

Response 2: I agree with your comment and will correct the introduction promptly (line 75-76).

Comment 3: why SEM was used rather than SD

Response 3: Due to small sample size; SD shows a very high range of variation.

Comment 4: why is this level of 5.5 chosen as cut-off, is this recommended by any guidelines ? does the Specificity and sensitivity of this level of AFP comparable to published literature ?

Response 4: The level of 5.2 ng/ml of AFP showed the best sensitivity for HCC detection, according to our patients’ group analysis. I will add data on other studies cut-off values (extended data in line 192-210)

Comment 5: why is 6 ng/ml used here while above table the cut-off is 5.2

Response 5: I agree with your suggestion and looked at our table data again. The transition of level of 6 ng/ml to 5.2 ng/ml does not change the proportions in table 4. Therefore, the 5.2 ng/ml will be used.

Comment 6: Did the authors use any validation cohort with this findings to confirm the findings?

Response 6: No, but we have a plan to do validation in the near feature.

 

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

As a pilot study, the Authors explore serum GPC3 levels specifically in patients with hepatocellular carcinoma (HCC) and advanced chronic liver disease, contributing new preliminary data to the field. The study utilizes serum GPC3 levels, which are less invasive than tissue biopsies and potentially aid in disease monitoring and early detection.

But there are numerous observations at different levels.

Being a pilot study, the sample size is minimal, limiting the findings' statistical power and generalizability. As an observational pilot, it may lack longitudinal follow-up to assess changes over time or causality. The study may not compare GPC3's performance directly against established markers like alpha-fetoprotein (AFP), restricting the assessment of its relative effectiveness.

The most serious concern is the potential for confounding factors: factors such as liver inflammation, underlying liver diseases like cirrhosis, viral hepatitis, or other comorbidities influence serum GPC3 levels. The authors did not perform any controls.

Furthermore, establishing a valid link between a biological object, such as GPC3, and a disease like HCC requires proof of its involvement in disease pathogenesis—in other words, demonstrating a mechanistic or causal relationship—that goes beyond the simple observation of elevated levels in patients. In this study, the authors primarily report serum GPC3 levels in patients with HCC and advanced liver disease, which are observational data. They do not establish a causal role or mechanistic involvement of GPC3 in developing HCC. Therefore, referring to GPC3 as a "marker" is based on its potential diagnostic utility rather than implying its direct participation in disease causation.

The authors do not claim GPC3 is causally involved in HCC; only that its serum levels may aid in diagnosis is a common and accepted practice in biomarker research.

It's essential for scientific rigor to differentiate between association and causation. The authors' statement claiming GPC3 as a marker is based on observed associations and is valid within the diagnostic utility context. Still, it does not establish causality or direct involvement in hepatocarcinogenesis.

Livelli elevati di AFP sono presenti in circa il 70% dei pazienti con HCC, e un aumento basale dell'AFP è spesso associato a una biologia tumorale, un carico e una sopravvivenza peggiori. Tuttavia, non tutti i tumori HCC presentano un'elevazione dell'AFP, e alcuni pazienti possono avere livelli normali o bassi di AFP.

 

The AFP metabolic world is connected to approximately 150-200 HCC-associated genes, but these HCC-associated genes are also related to HVB (hepatitis B virus), sharing approximately 25-30% of their genes. Cirrhosis-associated genes are also present among the HCC- and HVB-associated genes, and all are present and connected in the AFP metabolic world. All of this means that these conditions are interconnected at the molecular level. Specifically, the fact that a significant portion of the genes shared between HCC and HVB (25%) and that cirrhosis genes are also present in HCC patients indicates that patients diagnosed with HCC often have underlying cirrhosis or viral components. To be clear, the observed association between GPC3 and hepatocellular carcinoma (HCC) may be influenced by confounding factors such as cirrhosis and hepatitis B virus (HBV) infection, rather than being specific to HCC itself. In the patient population, both viral and cirrhotic factors could affect this association. Furthermore, the small sample size complicates the analysis even more. Therefore, GPC3 may be a marker for underlying liver damage or viral activity, increasing the risk of developing HCC. Its elevated levels might correlate with the progression to HCC, without being the direct cause of this progression. As a result, the link between GPC3 and HCC should be reconsidered. It may primarily indicate liver injury, cirrhosis, or viral infection, which are common in patients with HCC, rather than acting as a specific marker for malignant transformation.

 The studio needs other research to control for these underlying conditions and demonstrate an independent predictive ability to establish GPC3 as a specific marker for HCC. Consequently, based on the current data, the connection between GPC3 and HCC appears to be observational and possibly indirect, reflecting broader liver pathology rather than being unique to HCC itself.

Author Response

Dear Reviewer,

Thanks for the proper remarks! I agree with most of them considering the small number of patients, no direct comparison to AFP and about the fact that our control group are realistically exposed to risk for HCC development, rather than healthy people. In the revised document more data was added about weak or absent GPC3 expression in hepatitis or cirrhosis (line 219-230 part of Discussion) as well as a discussion of recent published studies on GPC3 performance for HCC detection. However, the tables and figures cannot be corrected in their design.

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript has been improved. It suggests serum GPC3 levels could serve as a marker for hepatocellular carcinoma (HCC), likely in conjunction with alpha-fetoprotein (AFP). Your work is a pilot study and should be regarded as such.

Having studied HCC for many years, I remain skeptical that current approaches will lead to a cure or reliable progression control. The AFP itself has a reliability of 70%. I want to share some observations that may help you understand what I mean and the current state of research in this field. In one of my articles (in press), I selected the most significant proteins and pathways associated with HCC and created a graph (a network) using interactomics. The resulting graph illustrates what is known about this cancer at the molecular and causal levels.

This analysis was based on an AI review of over 10,000 articles on PubMed, which identified more than 30,000 interactions and 9,085 enriched functional terms. The complex network structure reflects cellular location, biological processes, and molecular functions, enabling the functional characterization of these proteins and their roles in cancer.

However, when I selected the most significant experimentally validated interactions (i.e., biophysically/biochemically validated), I found that out of the tens of thousands of interactions identified, only 1.61% had been validated. This means that very few were confirmed to be reliable and trustworthy, while the rest remain unconfirmed hypotheses. The foundation of science relies solely on experiments, not on indirect data gathered through digital algorithms, which are never validated.

This significant discrepancy explains why, after decades of research and hundreds of thousands of scientific articles, we still lack minimal reliable knowledge and effective cures for HCC. Moreover, anyone who tries to interpret their results by comparing them to others—assuming those comparisons are precise and trustworthy—only immerses themselves in a vast bubble of scientific speculation. Conflicts within this bubble often coexist without solid experimental designs to support them.

I hope this information is helpful. Good luck with your work!