Placental Maternal Vascular Malperfusion Is Associated with Prepregnancy and Early Pregnancy Maternal Cardiovascular and Thrombotic Profiles

Round 1

Reviewer 1 Report

It was a great pleasure to read and review such a well written and elegant study. I have no major comments or corrections for the authors. 

Author Response

Thank you for your thoughtful and thorough comments to our manuscript, “Placental Maternal Vascular Malperfusion is Associated with Prepregnancy and Early Pregnancy Maternal

Cardiovascular and Thrombotic Profiles“. We have responded to each of your comments and concerns below, and made the relevant changes to the manuscript.

• We have clarified the number of placentas with features of each category of MVM lesion in Table 1. We have additionally added a section to Table 1 which identifies the number of placentas with each type of lesion, allowing for a better understanding of how the lesions are distributed across the 63 placentas evaluated in the study. The methods section has also been edited to more clearly specify how decidual arteriopathy was identified in lines 160-161.

The results section outlining cases of preeclampsia, lines 200-204, has been revised.

• We have clarified that the definition of DA required at least three features of DA, as defined in the Atlas of Placental Pathology Version 6. In response to your comments, our placental pathologist also reviewed slides of cases with identified DA, and stands by her identification. While DA is not common in large populations, within a small population that is selected for increased risk of preeclampsia (including prior preterm preeclamptics and nulliparous women), it is possible that the large number of placentas with identified features of MVM may be due in part to selection bias. However, as this study sought to identify associations between maternal characteristics and the occurrence of placental lesions, the increased frequency of these lesions within our population is acceptable. We have added a comment addressing this to our discussion (lines 333 to 337).
• Both the Amsterdam Criteria and the Atlas of Placental Pathology classify AVM and DVH as separate features of MVM. As this study was designed based on these published and widely accepted definitions, we disagree with the reviewer that the lesion types are the same, but do acknowledge that they commonly co-occur (see page 702 of the Amsterdam Criteria).
• As this study was not designed to redefine histopathological features of the placenta, but to employ published definitions from placental histopathology papers, the addition of images will detract from our findings. Images of the individual lesions can be retrieved from the published reference values.
• A second placental pathologist is not available as there are so few in the United States. Our institution, unlike many other research universities, employs a single board-certified pathologist specializing in placental pathology. The design of this study was to evaluate the presence of lesions of MVM, using pre-defined, published criteria. The pathologist was blinded to all clinical and outcome data at the time of assessment. We would also like to speak to the comment of agglutination as an outdated term, as Roberts and Polizzano continue to use this term in the most recent edition of the Atlas of Placental Pathology, released in August 2021. As they define, agglutination is a collection of micro-infarcts, and fibrous deposition alone is not included in the definition.
• We have revised the methods to clarify that 2 or more individual lesion types were required for MVM prediction in the logistic regression model (lines 164-168).
• Thank you for this comment. We reviewed the placental growth curves of Naeye, and they did not change the recognition of hypoplasia in those placentas <10% by Pinar. We have cited the Naeye growth curve as an additional placental growth curve reference, and added a comment to this effect (lines 163, 262-263).

Reviewer 2 Report

Application of maternal vascular malperfusion patterns rather than distinct placental histopathology like syncytial knots to clinical associations is an important step.  The study design and wealth of data generated are useful, but placental analysis has a number of problems.

1. The numbers of placentas examined do not match up with 63 cases: Methods lines 159-162 may explain the problem if the authors excluded decidual arteriopathy cases due to potential "bias".  This makes no sense.  If 19/27 placentas had 1 feature of MVM (70%) what happened to the other 36 cases?  Methods and Table 1 need to be cleaned up. Table 1 should list # of cases in the MVM diagnostic features as well as %; then # with 1 feature; # with 2; #3.   Note that PET section described 5/9 with preterm PET and then 5 more term cases (3/5 note in lines 202-03).
2. The nomenclature, definition, and application of "decidual vasculopathy/arteriopathy" needs to be cleaned up.  BE very careful. Fibrinoid necrosis in the membrane roll is a hypertension finding and should be clear described as not seen in the decidua basalis; therefore, it is a sign of high blood pressure, it is not a sign of early pathophysiology of spiral artery remodeling.  Un-remodeled spiral arteries or spiral arteries with retained muscle are a very weak diagnosis due to basal artery fragments in the maternal interface of the placenta.  A definition of Decidual Vasculopathy is most specific if it requires foamy macrophages, or smooth muscle medial hyperplasia in the placenta bed.  Use one term and one clear definition.  There is no way decidual vasculopathy is present in 40% of 63 cases-- it is rare to find it.  A much more sensitive feature is AVM and infarctions are specific. 
3. Distal villous hypoplasia is an unfortunate descriptor that was weakened in Amsterdam criteria because it is essentially seen only in early onset severe IUGR and few cases of early onset PET. It was replaced largely by AVM, which a number of studies have shown is present in about 2/3rds of MVM diagnoses-- it is more likely 32% AVM + 17% "DVH" are the same thing in this study.
4. NO FIGURES PROVIDED to illustrate pathologic diagnoses is an issue. 
5. A second experienced placental pathologist is needed to provide kappa statistic for all the diagnostic categories. For example, what is "agglutination" is that perivillous fibrinoid seen in AVM? An infarction?  This is an old term that should be avoided going forward. 
6. Should be clear if you are defining MVM as 1 feature, 2 features, or 3 features, maybe suggest potential "grade" mild, mod, severe for multivariate analysis.   Certainly preterm PET and IUGR will have >1 feature and most likely 3 features.
7. Trimmed placental weight relative to gestational age is the subject of growing controversy.  Some placental pathologists are using the Pinar table cited in this manuscript.  This reviewer recommends additional thought about using this table. It has limited cohort size and was created during the current obesity epidemic in the USA using placentas submitted to pathology.  Please consider the Table by Naeye 1987 that employed tens of thousands of placentas measured before the obesity epidemic and was the data relied on by Dr. Bernirschke for decades.  It will likely not change the [1] placenta classified as small in this paper, but it is a more reliable reference for the general audience. 

In brief, this is an important study with an interesting hypothesis, but the methods, analysis, and presentation require clarification. 

Author Response

Thank you for your thoughtful and thorough comments to our manuscript, “Placental Maternal Vascular Malperfusion is Associated with Prepregnancy and Early Pregnancy Maternal

Cardiovascular and Thrombotic Profiles“. We have responded to each of your comments and concerns below, and made the relevant changes to the manuscript.

• We have clarified the number of placentas with features of each category of MVM lesion in Table 1. We have additionally added a section to Table 1 which identifies the number of placentas with each type of lesion, allowing for a better understanding of how the lesions are distributed across the 63 placentas evaluated in the study. The methods section has also been edited to more clearly specify how decidual arteriopathy was identified in lines 160-161.

The results section outlining cases of preeclampsia, lines 200-204, has been revised.

• We have clarified that the definition of DA required at least three features of DA, as defined in the Atlas of Placental Pathology Version 6. In response to your comments, our placental pathologist also reviewed slides of cases with identified DA, and stands by her identification. While DA is not common in large populations, within a small population that is selected for increased risk of preeclampsia (including prior preterm preeclamptics and nulliparous women), it is possible that the large number of placentas with identified features of MVM may be due in part to selection bias. However, as this study sought to identify associations between maternal characteristics and the occurrence of placental lesions, the increased frequency of these lesions within our population is acceptable. We have added a comment addressing this to our discussion (lines 333 to 337).
• Both the Amsterdam Criteria and the Atlas of Placental Pathology classify AVM and DVH as separate features of MVM. As this study was designed based on these published and widely accepted definitions, we disagree with the reviewer that the lesion types are the same, but do acknowledge that they commonly co-occur (see page 702 of the Amsterdam Criteria).
• As this study was not designed to redefine histopathological features of the placenta, but to employ published definitions from placental histopathology papers, the addition of images will detract from our findings. Images of the individual lesions can be retrieved from the published reference values.
• A second placental pathologist is not available as there are so few in the United States. Our institution, unlike many other research universities, employs a single board-certified pathologist specializing in placental pathology. The design of this study was to evaluate the presence of lesions of MVM, using pre-defined, published criteria. The pathologist was blinded to all clinical and outcome data at the time of assessment. We would also like to speak to the comment of agglutination as an outdated term, as Roberts and Polizzano continue to use this term in the most recent edition of the Atlas of Placental Pathology, released in August 2021. As they define, agglutination is a collection of micro-infarcts, and fibrous deposition alone is not included in the definition.
• We have revised the methods to clarify that 2 or more individual lesion types were required for MVM prediction in the logistic regression model (lines 164-168).
• Thank you for this comment. We reviewed the placental growth curves of Naeye, and they did not change the recognition of hypoplasia in those placentas <10% by Pinar. We have cited the Naeye growth curve as an additional placental growth curve reference, and added a comment to this effect (lines 163, 262-263).

Reviewer 3 Report

Well written article. There is growing evidence concerning the role of maternal cardiovascular status plus adaptive potential as the origin of placental-mediated complications in pregnancy, such as pre-eclampsia. This article (study) thoroughly assesses maternal health characteristics impact on pregnancy outcome. It is also interesting that different stages of pregnancy are evaluated, from the prepregnancy state through delivery. Statistical analysis is adequate and review of literature is complete.

Author Response

We thank you for your comments. We have reviewed our manuscript, “Placental Maternal Vascular Malperfusion is Associated with Prepregnancy and Early Pregnancy Maternal Cardiovascular and Thrombotic Profiles” for minor edits to grammar and spelling.

Round 2

Reviewer 2 Report

Thank you for the careful revision.