Mechanisms and Neuroimaging Patterns of Hypereosinophilia-Related Ischemic Stroke: A Narrative Review through Three Cases

Round 1

Reviewer 1 Report

This is an exciting research paper. 

However, a few suggestions are placed to further improve the manuscript.

 

 

Figure 1,2 and 3: findings in the figures should be marked with arrows etc

 

Case 1: rather than writing, ‘treated with high dose prednisolone’, it is advised to write the exact does )as in case 2 and 3) 

Author Response

First of all, we would thank the reviewer for the appretiation of our paper and for the suggestions. 

We modified the figures 1, 2 and 3 adding asterisks to label the main findings. 

We also updated the description of the treatment in patients 1:

"The patient was treated with high-dose intravenous glucocorticoid therapy (methylprednisolone 1000 mg each day for five days, followed by slow tapering off) and discharged on oral prednisone 25 mg, with a significant reduction in eosinophil count in the peripheral blood and clinical improvement."

Reviewer 2 Report

Abstract –

In a review of only three cases, you can’t state that ‘borderzone pattern is quite suggestive’

Background –

Best to state which classification criteria you are referring to in text, rather than forcing readers to go to the refs for this.

Please give more detail what ‘not uncommon’ means in the context of neurologic involvement.

It would be good to include details regarding the types of cancer associated with HES and also under what conditions reactive HES occurs.

Results –

First case – please clarify if the head CT that showed the ischaemic lesion was a repeat examination related to the second presentation two days later.

The word arterial is redundant from ‘Arterial CTA’

Please clarify why the identified allergies were unlikely to be the cause of the eosinophilia, what range of eosinophil counts would be expected in this scenario?

Second case – Given the range of tests performed for this patient (and for patient 1), I think a table detailing the commonest causes of eosinophilia would be helpful.

The relevance of CT body findings should be added.

The statement ‘No significant vascular occlusion’ should be modified, vascular occlusion is always significant. I think you mean vascular stenosis or obstruction.

Third case – Will need to reflect in Discussion that this patient didn’t strictly meet the criteria for HES.

Discussion –

In reference to the previous comment about case 3, the first sentence of the discussion is somewhat contradictory.

Given the importance of normal arterial supply here (since borderzone infarcts are commonly found in patients with proximal stenosis of the major vessels supplying brain and you are associating these borderzone infarcts instead with HES), please include details of angiography in Table 1.  

The discussion of borderzone infarcts is probably too detailed for a paper on HES and in places repetitive. Do you mean to imply in paragraph 5 that borderzone and watershed infarcts are different phenomena?   These are synonyms but you should be consistent throughout. You need to simplify this section to describe the different types and commonest causes of borderzone infarcts before saying how the situation differs in patients with HES.

Conclusion –

I think you should caveat this statement with ‘when symmetrically involving bilateral borderzones in the absence of hypoperfusion’ or similar.  

Figure 1 –

I don’t think these are a ‘cluster of ischaemic lesions’ but a peripheral wedge of infarcted tissue. In foot note please define ACA and clarify that the major or proximal arteries are patent, as distal branches not clearly seen.

Figure 2 –

You mean ‘multiple scattered hyperintense lesions’

Table 1 –

Was NIHSS really 0 for two patients?

Author Response

I would thank the reviewer for the suggestions and we tried to address them in order to improve the quality of our manuscript. The main aim of our paper is to highlight the neuroimaging pattern in HES patients and to discuss the pathophysiology of it detailing how hypereosinophilia may produce it. 

• In a review of only three cases, you can’t state that ‘borderzone pattern is quite suggestive’

We agree with the reviewer's comment and changed the text accordingly: "border zones pattern without large artery steno-occlusion on neuroimaging may help to raise the suspicion in the neurovascular diagnostic pathway"

• Best to state which classification criteria you are referring to in text, rather than forcing readers to go to the refs for this.

We tried to make easier the reading of the definition and classification of HES, accordingly with the reviewer's suggestion.

"

On the basis of the pathogenic mechanism underlying the increased production of eosinophils, HES can be further classified as:

- primary or neoplastic HES,

- secondary or reactive HES,

- idiopathic HES (when the other causes have been excluded) and

- other conditions and syndromes accompanied by hypereosinophilia."

• Please give more detail what ‘not uncommon’ means in the context of neurologic involvement.

We would thank the reviewer for this comment, which allows us to add more detail in the text, as follows: "Neurologic involvement is not uncommon (65% of cases) in HES and it typically consists in peripheral neuropathy, encephalopathy, and stroke [3]. More in detail, 15% of patients with HES have central nervous system involvement, mainly as an encephalopathy characterized by behavioral disturbances and upper motor neuron signs, but 12% of patients have cerebrovascular events with an embolic pattern; 52% of patients have a peripheral neuropathy, more commonly a sensory polyneuropathy [3]"

 

• It would be good to include details regarding the types of cancer associated with HES and also under what conditions reactive HES occurs.

Many thanks for the comment. We updated the text adding the requested informations:

"

In primary HES the most involved neoplasms are myeloid neoplasms: chronic myeloid leukemia, other myeloproliferative neoplasms (MPN), distinct variants of acute myeloid leukemia, rare forms of myelodysplastic syndromes (MDSs), some MDS/MPN overlap disorders, and a subset of patients with (advanced) systemic mastocytosis (SM) [1].

Reactive or secondary HES is associated with an underlying inflammatory condition (e.g. inflammatory bowel diseases) or solid tumors, but also some hematological neoplasms (Hodgkin lymphoma, T-cell lymphoma, or B-lymphoblastic leukemia/lymphoma). Other subsets of reactive HES is related to helminth infections, allergic reactions, atopic diseases and drug reactions."

•  First case – please clarify if the head CT that showed the ischaemic lesion was a repeat examination related to the second presentation two days later.

We added the suggested information: "A second head CT at two days from the symptom onset "

• The word arterial is redundant from ‘Arterial CTA

We deleted the word "arterial"

• Please clarify why the identified allergies were unlikely to be the cause of the eosinophilia, what range of eosinophil counts would be expected in this scenario?

Thanks for rising this point. We furtherly detailed our diagnostic reasoning adding the following sentence: "Moreover, BP is one of the most common contact allergens and reactions to BP are typically seen as contact eczema or dermatitis both for topical contact and oral intake with several systemic complications and the reported patient did not have any of these manifestations."

• Second case – Given the range of tests performed for this patient (and for patient 1), I think a table detailing the commonest causes of eosinophilia would be helpful.

We would thank the reviewer for this comment. The main aim of our narrative review is to use three cases to show the cerebrovascular pattern of HES-related stroke, so the diagnostic pathway of each case or the comparison among the cases from this point of view was considered outside the aim of the paper. If the reviewer thinks that a summary table with all diagnostic finding is mandatory, we will arrange it, but the lack of it is due to the fact that we are not proposing a case series about the causes or diagnostic pathway of HES but to use three case to highlight the cerebrovascular involvement in HES starting from acute stroke. 

 

•   The relevance of CT body findings should be added.

We added a short sentence detailing this issue: "The main findings of the whole body CT scan are the lack of solid neoplasms and the evidence of splenic infarcts. Both lungs and lymph nodes changes have been considered not specific in a bedridden patient with delirium."

• The statement ‘No significant vascular occlusion’ should be modified, vascular occlusion is always significant. I think you mean vascular stenosis or obstruction.

Thanks for noticing this point. We changed the sentence in "no stenosis or occlusion in the head and neck vessels was detected "

• Third case – Will need to reflect in Discussion that this patient didn’t strictly meet the criteria for HES

We wrote it in the description of the case but we slightly changed the sentence in order to better explasin that the need to start a treatment because of the severe and worsening condition of the patients from the neurological point of view induced to start steroid treatment, so we were not able to see an eventual further increase in eosinophil count. 

"Although her eosinophil count was a little less than 1500, not strictly meeting the criteria for hypereosinophilia, considering the significantly increased value and the co-occurrence of a very high D-dimer, a steroid treatment was started and an extensive diagnostic workup was performed," 

• In reference to the previous comment about case 3, the first sentence of the discussion is somewhat contradictory.

We hope that the clarification about the previous comment may help to understand the first sentence in the discussion. We are presenting three cases of stroke patients with hyperesosinophilia and two of them can be defined as HES. The third patient does not fullfill the criteria for HES but shares the same issue in neuroimaging and clinical course, so the proposed reasoning applies to her too. 

•  Given the importance of normal arterial supply here (since borderzone infarcts are commonly found in patients with proximal stenosis of the major vessels supplying brain and you are associating these borderzone infarcts instead with HES), please include details of angiography in Table 1.  

Thankf for teh suggestion. We added a line in the table 1 with negative findings of CTA and MRA in all patients. 

• The discussion of borderzone infarcts is probably too detailed for a paper on HES and in places repetitive. Do you mean to imply in paragraph 5 that borderzone and watershed infarcts are different phenomena?   These are synonyms but you should be consistent throughout. You need to simplify this section to describe the different types and commonest causes of borderzone infarcts before saying how the situation differs in patients with HES.

We would thank the reviewer for this comment and for the suggestion. The discussion about the pathophysiology and classification of border zone infarcts is paramount for the aim of this review, so we checked the whole manuscript for watershed/border zone terms in order to avoid misunderstanding. The two terms are used with the same meaning in the literature, so we cited several references with "watershed" in the title. Our aim is to give to the less experienced reader all the informations to understand the interplay of different mechanisms that converge in the composite ischemic pattern on neuroimaging of the presented patients. The main message is that the so called hemodynamic (border zone) pattern is not always hemodynamic, may occur without proximal arterial stenosis or occlusion and may be associated to cortical infarctions. In these cases, look to the blood test for eosinophil count and consider the related hypotheses in the diagnostic pathway. This part of the discussion is mandatory to interpret the findings of our patients and to give to the reader the tools to use the neuroimaigng findings as a red flagt in the clinical practice. 

• I think you should caveat this statement with ‘when symmetrically involving bilateral borderzones in the absence of hypoperfusion’ or similar.  

We rephrased the sentence as follows: "

In conclusion, HES should be considered in the diagnostic work-up of ischemic stroke and the suspicion may be raised and supported by the a neuroimaging pattern of lesions symmetrically involving the border zones, in the presence of a marked increase in eosinophils’ count. An extensive diagnostic work-up is hence necessary to establish the type of HES and the correct treatment which should be started promptly."

• cluster of ischaemic lesions’ but a peripheral wedge of infarcted tissue. In foot note please define ACA and clarify that the major or proximal arteries are patent, as distal branches not clearly seen.

We changed the capsion as follows: "

MRI of patient 1 showed the presence in the axial Fluid Attenuated Inversion Recovery (FLAIR) sequences of a wedge of cortical ischemic lesions in the left posterior parietal lobe (yellow asterisk) (A) with Diffusion Weighted (DWI) and Apparent Diffusion Coefficient (ADC) sequences pattern suggesting acute ischemic lesions (B, C). MRA (time-of-flight [TOF] reconstruction) shows fully patent proximal intracranial arteries and right A1 anterior cerebral artery (ACA) aplasia (D)."

• You mean ‘multiple scattered hyperintense lesions’

Thanks for the correction. We cahnged it in the text. 

• Was NIHSS really 0 for two patients?

Yes, but we refer to NIHSS score at admisison, so we added this detail in the table. Patient 3 had an uncountable number of strokes during the first two weeks of hospital staying, so the NIHSS at 7 days was clearly much more significant. 

 

 

Round 2

Reviewer 2 Report

Thank you for your prompt response to my comments.

I have only one further suggestion, that the conclusion in the main body of the paper matches the abstract, i.e. mention again the absence of arterial stenosis and occlusion.