Diagnosis of Portal Hypertension

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This is a fantastic overview of HVPG measurement and alternative methods to assess portal hypertension.  I learnt a lot.

In clinical practice I often see patients with alcohol dependent cirrhosis who have stopped drinking alcohol for many years including some who have had a TIPs inserted.  SOme of these pre-date fibroscan, but many do not.  If you undertake a fibroscan, it can sometimes be normal or less than 10kPa, and their platelet count may be normal.  Some, despite unequivocal evidence of prior cirrhosis, have a liver biopsy which shows continued cirrhosis, and some it may be normal suggesting resolution of cirrhosis.  My question to the authors is what is known about the correlation between HVPG and fibroscan findings in ALD following prolonged alcohol abstinence?  Are there any data?  Can I assume that if the liver stiffness has decreased from 44kPA to 10kPa that endoscopy is no longer required?  As that is what I do in practice.  Am I wrong?

Just as I was about to ask what are the pitfalls of measuring HVPG, the authors list them.  The authors are one of the foremost experts at measurement of HVPG.  Therefore I suggest there is  a checklist for radiologists or others to follow when measuring HVPG.  Such a checklist can then be used to consolidate future guidelines for measurement of HVPG.  I realise that the Baverno guidelines have a list (Baveno VII Technical / Procedural Standards for HVPG) but is it a checklist - I am not that familiar so this may be a mute point.

Finally the title is awful and would not get my attention in a search.  Keep it simple and broad.  Maybe just "Diagnosis

Author Response

Reviewer 1

This is a fantastic overview of HVPG measurement and alternative methods to assess portal hypertension.  I learnt a lot.

In clinical practice I often see patients with alcohol dependent cirrhosis who have stopped drinking alcohol for many years including some who have had a TIPs inserted.  SOme of these pre-date fibroscan, but many do not.  If you undertake a fibroscan, it can sometimes be normal or less than 10kPa, and their platelet count may be normal.  Some, despite unequivocal evidence of prior cirrhosis, have a liver biopsy which shows continued cirrhosis, and some it may be normal suggesting resolution of cirrhosis.  My question to the authors is what is known about the correlation between HVPG and fibroscan findings in ALD following prolonged alcohol abstinence?  Are there any data?  Can I assume that if the liver stiffness has decreased from 44kPA to 10kPa that endoscopy is no longer required?  As that is what I do in practice.  Am I wrong?

Authors answer: We would like to thank the Reviewer for the kind and constructive questions.

This is a very relevant and pragmatic question, although the data are limited in ALD. As already stated in the manuscript, HVPG remains the gold standard for assessing portal pressure and clinically significant portal hypertension (CSPH). Fibroscan primarily estimates liver stiffness, which largely reflects fibrosis, congestion, and eventually inflammation, but not directly the portal pressure. Therefore, the relationship between liver stiffness and HVPG is non-linear and has significant limitations, especially at higher levels of portlal hypertension (PH). As PH becomes more severe, extra-hepatic factors such as splanchnic vasodilatation, hyperdynamic circulation, and dynamic changes contribute more to the pressure gradient than pure intrahepatic resistance. Several authors caution that liver stiffness may saturate or lose discriminative power once HVPG is high. Salavrakos et al (J Clin Gastro 2019:772) found that fibroscan reliably ruled out advanced fibrosis and significant portal hypertension in ALD patients. In relation to the amount of fibrosis they found that TE cutoffs: ≥ 11.7 kPa for F2, ≥ 15.2 kPa for F3, and ≥ 21.2 kPa for F4. A TE value of 30.6 kPa predicted HVPG > 10 mmHg with a specificity of 94%. TE had a negative predictive value of 84% for ruling out varices on endoscopy and after 2 weeks of abstinence, the mean liver stiffness decreased by ~2.7 kPa, improving agreement between TE and histology.

This suggests that LSM has a useful but imperfect association to HVPG in ALD, and that early abstinence may tend to normalize the liver with respect to edema and inflammation with some drop in stiffness. In a Korean cohort, Ryu et al (Clin Mol Hepatol 2021:197) reported a correlation between HVPG and liver stiffness in ALD of r = 0.542, and the optimal liver stiffness cutoff for predicting HVPG ≥ 10 mmHg was ~32.2 kPa (PPV ~94.5%), and for HVPG ≥ 12 mmHg was ~36.6 kPa (PPV ~91.0%).

These cutoff values are somewhat high, reflecting that alcoholic cirrhosis tends to show elevated stiffness values before achieving clinically relevant HVPG thresholds.

According to the recent noninvasive cACLD- Baveno guidelines criteria, follow-up of portal hypertension suggests that in patients with chronic liver disease, LSM < 10 kPa is considered a low-risk threshold (i.e. “portal hypertension unlikely”) whereas LSM ≥ 20–25 kPa suggests a higher risk / suspicion for CSPH (de Franchis, JHepatol 2022:959).This is in line with EASL / AASLD consensus/guidelines where a stable reduction of liver stiffness below 20 kPa following for example alcohol abstinence, the risk of liver-related events and portal hypertension may become negligible as the LSM drops further, especially <10 kPa (Thabut D. Clin Res Hep Gastroenterol 2022.Oct 46(8)). However, these are consensus-based thresholds rather than strictly validated in long-term prospective ALD cohorts.

Practically, your scenario is not unrealistic: a liver stiffness that high initially (44 kPa) suggests advanced fibrosis / cirrhosis and a high probability of PH, and a drop to 10 kPa is dramatic. But caution is needed before dropping surveillance or endoscopy solely on that basis.

In the revised version of our paper, we have added on p.8  that there is evidence of a correlation between fibroscan measurements and HVPG in ALD, but with important limitations such as abstinence. Therefore, fibroscan is an imperfect surrogate for HVPG.

 

Just as I was about to ask what are the pitfalls of measuring HVPG, the authors list them.  The authors are one of the foremost experts at measurement of HVPG.  Therefore I suggest there is  a checklist for radiologists or others to follow when measuring HVPG.  Such a checklist can then be used to consolidate future guidelines for measurement of HVPG.  I realise that the Baverno guidelines have a list (Baveno VII Technical / Procedural Standards for HVPG) but is it a checklist - I am not that familiar so this may be a mute point.

Authors answer: We follow the Reviewer´s proposal and have added a practice checklist in a new Table 2 inserted in section 4.1.

Finally the title is awful and would not get my attention in a search.  Keep it simple and broad.  Maybe just "Diagnosis.

Authors answer: As suggested by the Reviewer, the title of our paper now reads: “Diagnosis of portal hypertension”

Reviewer 2 Report

Comments and Suggestions for Authors

Recommendations

1. State the nature of this review: narrative or systematic.
2. Transparency regarding search algorithm should be provided: time-frame of the article searched, type of papers (preferably original articles/reviews); peer-reviewed only, no preprints and finally, database and keywords used for article screening.
3. A figure of chart with basic mechanisms of PH could be provided in section 2.
4. A new section with common etiologies could be added after section 2, or could be merged into a new section.
5. In etiology section i did not see any contribution of myeloproliferative neoplasms especially Policytemia Vera wich can cause PVT. see this: https://doi.org/10.3390/livers5030029 or duplication cysts: https://doi.org/10.3390/diagnostics14070675 
6. In section 3-Encephalopathy could benefit from a more broader description even a small table with clinical aspects and classification.
7. In section 5.3 some markers are not really assessing the PV pressure, so they can not be considered non-invasive markers, rather surrogate markers, I think this is a important distinction and you need to specify.
8. Invasive measurement of PVP is very rare in most of the centers, could you provide the clinical context (indications/recommendations) given the fact it can have serious complications-arrhythmia?
9. Platelet count below 150.000/mm3 has very low sensitivity for PH, could you ellaborate more on this?
10. Conclusion should be more concise.
11. Many instances of auto-citation- first and last author.

Comments on the Quality of English Language

Minor English language revisions.

Author Response

Reviewer 2

We would like to thank the Reviewer for the helpful and constructive comments to our paper.

 

Recommendations

1. State the nature of this review: narrative or systematic.

Authors answer: This review is primarily a narrative review based on the most recent literature. According to the guidelines, the reference list is restricted to 60 numbers. This has now been specified in the revised version of our paper at p.2.

 

1. Transparency regarding search algorithm should be provided: time-frame of the article searched, type of papers (preferably original articles/reviews); peer-reviewed only, no preprints and finally, database and keywords used for article screening.

Authors answer: Please see the answer to the above-raised question.

1. A figure of chart with basic mechanisms of PH could be provided in section 2.

Authors answer: As recommended by the Reviewer, we have now incorporated a new Figure 1 summarizing the pathophysiologic players of PH in section 2.

1. A new section with common etiologies could be added after section 2, or could be merged into a new section.

Authors answer: As suggested by the Reviewer, we have added the most common etiologies after section 2 in the former Figure 1, which is now the new Figure 2.

1. In etiology section i did not see any contribution of myeloproliferative neoplasms especially Policytemia Vera wich can cause PVT. see this: https://doi.org/10.3390/livers5030029 or duplication cysts: https://doi.org/10.3390/diagnostics14070675.

Authors answer: As proposed by the Reviewer, we have added these etiologies to the new Figure 2.

 

6. In section 5.3-Encephalopathy could benefit from a more broader description even a small table with clinical aspects and classification.

Authors answer: We think it is beyond the scope of this paper to go into details with hepatic encephalopathy apart from what has already been mentioned at p. 1, 3, 10, and 11.

1. In section 5.3 some markers are not really assessing the PV pressure, so they can not be considered non-invasive markers, rather surrogate markers, I think this is a important distinction and you need to specify.

Authors answer: We agree with the Reviewer, and we have now specified that some of the non-invasive markers may correlate variably with the portal pressure. These may to a lesser extent be related to the pathophysiology of portal. This has now been specified at p.9.

1. Invasive measurement of PVP is very rare in most of the centers, could you provide the clinical context (indications/recommendations) given the fact it can have serious complications-arrhythmia?

Authors answer: In general, we consider measurement of the portal pressure as safe with few complications. It is recommended as an important part of the evaluation of patients with PH as supported by the Baveno and EASL guidelines. Indications for measurement of PH are listed in Table 1 and in the text at p.3. Moreover, we have made a new Table 2 with a checklist.

1. Platelet count below 150.000/mm3 has very low sensitivity for PH, could you ellaborate more on this?

Authors answer: Please see the answer to point nr. 7.

1. Conclusion should be more concise

Authors answer: The Reviewer is correct and in the revised paper we have now shortened and strengthened the conclusion at p.11.

1. Many instances of auto-citation- first and last author.

12. Authors answer: We would prefer to maintain the reference list with its relatively few numbers at the Editor´s discretion.

 

 

Reviewer 3 Report

Comments and Suggestions for Authors

I suggest to add a table with the different causes of portal hypertension, specifically in the etiologies of intra-hepatic portal hypertension; It must be distinguished between pre-sinusoidal and sinusoidal causes. I suggest another table resuming the interpretation of portal pressure measurements. You could clarify better the concept of mechanobiology or you can delete it from the paragraph "pathophysiology". 

Author Response

Reviewer 3.

I suggest to add a table with the different causes of portal hypertension, specifically in the etiologies of intra-hepatic portal hypertension; It must be distinguished between pre-sinusoidal and sinusoidal causes. I suggest another table resuming the interpretation of portal pressure measurements. You could clarify better the concept of mechanobiology or you can delete it from the paragraph "pathophysiology". 

We thank the Reviewer for the useful and constructive comments to our paper. As recommended by the Reviewer we have now revised a new Figure 2 and inserted etiologies of PH according to topography. In order not to expand the space of the paper we have, at the Editor´s discretion, chosen not to expand the mentioning of mechanobiology further. We feel that it is beyond the scope of this paper to go into details with this topic.