CSF Dynamics for Shunt Prognostication and Revision in Normal Pressure Hydrocephalus

Round 1

Reviewer 1 Report

This manuscript is a summary of extensive experience in CSF dynamics, with a well-described introduction to the complexities of the issue of clinical management of hydrocephalus and a clearly defined objective. The paper provides a good overview of the limitations and benefits of CSF dynamic testing. Particularly, the discussion regarding confirmation of shunt function in the context of assessing outcome makes a very relevant point, and this is not always adequately addressed in studies on INPH. Overall, it is a good review and I have no major concerns. However, I do have a number of more minor comments that could improve the clarity and quality of the manuscript:

• There are some formatting issues with the text and labels in several figures.
  • Some text appears to be missing in fig 1.
  • Figure 7 shows an interesting phenomenon but needs clearer labelling of the graphs axes and line 292 seems like it should be part of the legend for fig 7?
  • Figure 10 requires more explanation. What, if any, of the presented data stems from PET during infusion? Is the left and right middle panels data for INPH and healthy (i.e. INPH on the left and healthy on the right)? The parameter of SRoR requires a better explanation in the text or a reference.
  • For fig 11 and the corresponding text, it would be beneficial to state what the critical pressure for the investigated shunt type actually is (rather than simply saying it was exceeded in the investigation).
  • The gap in data in figure 12 C is strange and not explained.
• The levels of section headings are occasionally confusing. Section 3.8 and 3.9 seem to be in some sense subsections to 3.7, but this is not clearly marked, i.e. it is not directly clear that 3.10 begins a section that is not related to CBF/autoregulation. Section 3.11 is a single sentence and seems to be a higher level heading for section 3.12-13. Improved formatting of the headers would improve the structure and readability of the manuscript.
• The sentence in line 163-168 is unclear; it seems to refer to normal values for both baseline ICP and Rout. However, the parenthesis and wording then make it seem as if the Rout value of 13 only applies as a definition for the group of patients for which the value of 10 mmHg for baseline ICP was derived.
• The formatting of table 1 makes it difficult to read and some column could benefit from clearer titles, also the abbreviation LR is not explained. The “other main findings” column content could use some clearer explanations to be of much use to the reader
• The sentence on line 186-188 is unclear, seemingly implying that Rout was around 18-20 in healthy elderly in recent studies. I believe the intended meaning is likely that the upper limit for the normal range of Rout was raised to this level?
• Regarding AMP, it may be worth noting that the most positive evidence stem from overnight ICP monitoring. One of the publications from the Norwegian group (Eide & Sorteberg, Neurosurgery 2010) covers 100+ subjects with this methodology. While that is a retrospective study rather than a clinical trial, it may contradict the statement regarding a lack of larger studies with positive indications for the use of AMP to predict shunt response. While the present manuscript concerns CSF dynamic tests, it is not really clarified if the statement in question is meant to reflect only AMP recorded in this setting. Mention of the two different ways of assessing AMP would be a relevant addition to this section.
• Regarding the AMP-P relationship, the breakpoint in this relationship may be worth remarking upon (outside a figure legend), including the relationship between this and a low RAP.
• The time scale be clearer in several of the figures, to help the reader interpret the observations.
• The sentence on line 329-333 would benefit from revised language, as this seemingly very relevant reflection is difficult to understand.
• For clarity, the paragraph starting on line 413 should be joined with the previous paragraph.
• Section 3.13 references interesting results for 369 patients, but no reference is included and it is no clear if this analysis is published in any paper.

Author Response

1) Some text appears to be missing in fig 1: we have changed fig1 accordingly and adjusted the text so it would show on the figure

2) Figure 7 shows an interesting phenomenon but needs clearer labelling of the graphs axes and line 292 seems like it should be part of the legend for fig 7?:

We have amended the labelling and legend accordingly.

3) Figure 10 requires more explanation. What, if any, of the presented data stems from PET during infusion? Is the left and right middle panels data for INPH and healthy (i.e. INPH on the left and healthy on the right)? The parameter of SRoR requires a better explanation in the text or a reference.

We have amended Fig 10 to reflect those changes and have added text and reference for SRoR

4) For fig 11 and the corresponding text, it would be beneficial to state what the critical pressure for the investigated shunt type actually is (rather than simply saying it was exceeded in the investigation).

We have changed fig 11 accordingly

5) The gap in data in figure 12 C is strange and not explained.

We have changed figure 12C to explain the gap in the data

6) The levels of section headings are occasionally confusing. Section 3.8 and 3.9 seem to be in some sense subsections to 3.7, but this is not clearly marked, i.e. it is not directly clear that 3.10 begins a section that is not related to CBF/autoregulation. Section 3.11 is a single sentence and seems to be a higher level heading for section 3.12-13. Improved formatting of the headers would improve the structure and readability of the manuscript.

Thank you indeed for observing this, we have changed the headings and subheadings to reflect your suggestions.

7) The sentence in line 163-168 is unclear; it seems to refer to normal values for both baseline ICP and Rout. However, the parenthesis and wording then make it seem as if the Rout value of 13 only applies as a definition for the group of patients for which the value of 10 mmHg for baseline ICP was derived.

We have changed the wording of this sentence to make it more clear. This sentence has now moved to lines 173-175 due to previous corrections.

8) The formatting of table 1 makes it difficult to read and some column could benefit from clearer titles, also the abbreviation LR is not explained. The “other main findings” column content could use some clearer explanations to be of much use to the reader

We have amended Table 1 to include an improved title, further explanations of findings and LR

9) The sentence on line 186-188 is unclear, seemingly implying that Rout was around 18-20 in healthy elderly in recent studies. I believe the intended meaning is likely that the upper limit for the normal range of Rout was raised to this level?

Sentence corrected accordingly (now lines 199-200)

10) Regarding AMP, it may be worth noting that the most positive evidence stem from overnight ICP monitoring. One of the publications from the Norwegian group (Eide & Sorteberg, Neurosurgery 2010) covers 100+ subjects with this methodology. While that is a retrospective study rather than a clinical trial, it may contradict the statement regarding a lack of larger studies with positive indications for the use of AMP to predict shunt response. While the present manuscript concerns CSF dynamic tests, it is not really clarified if the statement in question is meant to reflect only AMP recorded in this setting. Mention of the two different ways of assessing AMP would be a relevant addition to this section.

We have corrected the corresponding line under AMP paragraph (now lines 260-261) to "randomised trials". We have also added a relevant segment on the two different methods of calculating AMP.

11) Regarding the AMP-P relationship, the breakpoint in this relationship may be worth remarking upon (outside a figure legend), including the relationship between this and a low RAP.

We have added text explanation for the AMP-P relationship in lines 269-275.

12) The time scale be clearer in several of the figures, to help the reader interpret the observations.

13) The sentence on line 329-333 would benefit from revised language, as this seemingly very relevant reflection is difficult to understand.

We have gone through the figures and added timescales accordingly

14) For clarity, the paragraph starting on line 413 should be joined with the previous paragraph.

We have moved the paragraph "the coefficient Mz... result in impaired CBF", now in row 338.

15) Section 3.13 references interesting results for 369 patients, but no reference is included and it is no clear if this analysis is published in any paper.

Thank you, we have added the corresponding reference.

Reviewer 2 Report

The authors have well illustrated and articulated the problems related to the management of hydrocephalus at normal pressure.The paper is truly complete:the result was to obtain a flow-chart wich is essential for the neurosurgeon in setting the correct indication for treatment and  in subsequent management of problems related to shunt dysfunction.Another important aspect for the neurosurgeon and that  the authors could if possible based on their experience illustrate in their work,predict the type of valve to be implanted based on dynamic CSF flow studies. 

Author Response

It looks like there are no corrections suggested

Reviewer 3 Report

This review by Lalou et al. nicely summarize the role of CSF hydrodynamic tests in NPH. 

I have only few minor notes:

1. Set up of Table 1 – rows are blurry?

2. Lines 415-417: Please give more data (in relation to ref 197) why aspiration of CSF from shunt (mimicking tap test) is considered potentially dangerous? I suggest that the shunt infusion test (despite of pressure limit) can be potentilly dangerous as well.

3. Consider proving example of functioning shunt in figure 12?

 

Author Response

1. Set up of Table 1 – rows are blurry?

We have reformatted Table 1

2. Lines 415-417: Please give more data (in relation to ref 197) why aspiration of CSF from shunt (mimicking tap test) is considered potentially dangerous? I suggest that the shunt infusion test (despite of pressure limit) can be potentilly dangerous as well.

We have changed the paragraph "empirical assessments of shunt patency ...", now in lines 479-483, to explain a few more details about the aforementioned dangers. Recogniseable, there would be limitations in proving how "safe" is the 40 mmHg safety threshold we mentioned, but the main point we have stressed is that infusion tests, unlike aspirating blindly or pumping blindly, are done after strict control and recording of any raise or drop in pressure induced and in the presence of a working shunt, no pressure effects will be seen, whereas in a non-functioning shunt, the pressure effects temporarily noted from the infusion tests would be nothing compared to the dangers of a non-functioning shunt that has been allowing these pressure rises chronically

We have referred to the literature reporting the safety of infusion tests

3. Consider proving example of functioning shunt in figure 12?

Thank you for your useful suggestion, we have added an example of a functioning shunt in figure 11 (now split into 11A and 11B), to avoid overcrowding figure 12