Association between cerebral palsy and microscopically verified placental infarction in extremely preterm infants

Association of Placental Pathology with Physical and Neuronal Development of Infants: A Narrative Review and Reclassification of the Literature by the Consensus Statement of the Amsterdam Placental Workshop Group

The placenta is the largest fetal organ, which connects the mother to the fetus and supports most aspects of organogenesis through the transport of nutrients and gases. However, further studies are needed to assess placental pathology as a reliable predictor of long-term physical growth or neural development in newborns. The Consensus Statement of the Amsterdam Placental Workshop Group (APWGCS) on the sampling and definition of placental lesions has resulted in diagnostic uniformity in describing the most common pathological lesions of the placenta and contributed to the international standardization of descriptions of placental pathology. In this narrative review, we reclassified descriptions of placental pathology from previously published papers according to the APWGCS criteria and comparatively assessed the relationship with infantile physical and/or neural development. After reclassification and reevaluation, placental pathology of maternal vascular malperfusion, one of the APWGCS criteria, emerged as a promising candidate as a universal predictor of negative infantile neurodevelopmental outcomes, not only in term and preterm deliveries but also in high-risk groups of very low birthweight newborns. However, there are few studies that examined placental pathology according to the full categories of APWGCS and also included low-risk general infants. It is necessary to incorporate the assessment of placental pathology utilizing APWGCS in the design of future birth cohort studies as well as in follow-up investigations of high-risk infants.

Tissue Contamination Challenges the Credibility of Machine Learning Models in Real World Digital Pathology

Machine learning (ML) models are poised to transform surgical pathology practice. The most successful use attention mechanisms to examine whole slides, identify which areas of tissue are diagnostic, and use them to guide diagnosis. Tissue contaminants, such as floaters, represent unexpected tissue. Although human pathologists are extensively trained to consider and detect tissue contaminants, we examined their impact on ML models. We trained 4 whole-slide models. Three operate in placenta for the following functions: (1) detection of decidual arteriopathy, (2) estimation of gestational age, and (3) classification of macroscopic placental lesions. We also developed a model to detect prostate cancer in needle biopsies. We designed experiments wherein patches of contaminant tissue are randomly sampled from known slides and digitally added to patient slides and measured model performance. We measured the proportion of attention given to contaminants and examined the impact of contaminants in the t-distributed stochastic neighbor embedding feature space. Every model showed performance degradation in response to one or more tissue contaminants. Decidual arteriopathy detection--balanced accuracy decreased from 0.74 to 0.69 ± 0.01 with addition of 1 patch of prostate tissue for every 100 patches of placenta (1% contaminant). Bladder, added at 10% contaminant, raised the mean absolute error in estimating gestational age from 1.626 weeks to 2.371 ± 0.003 weeks. Blood, incorporated into placental sections, induced false-negative diagnoses of intervillous thrombi. Addition of bladder to prostate cancer needle biopsies induced false positives, a selection of high-attention patches, representing 0.033 mm2, and resulted in a 97% false-positive rate when added to needle biopsies. Contaminant patches received attention at or above the rate of the average patch of patient tissue. Tissue contaminants induce errors in modern ML models. The high level of attention given to contaminants indicates a failure to encode biological phenomena. Practitioners should move to quantify and ameliorate this problem.

Tissue contamination challenges the credibility of machine learning models in real world digital pathology

Machine learning (ML) models are poised to transform surgical pathology practice. The most successful use attention mechanisms to examine whole slides, identify which areas of tissue are diagnostic, and use them to guide diagnosis. Tissue contaminants, such as floaters, represent unexpected tissue. While human pathologists are extensively trained to consider and detect tissue contaminants, we examined their impact on ML models. We trained 4 whole slide models. Three operate in placenta for 1) detection of decidual arteriopathy (DA), 2) estimation of gestational age (GA), and 3) classification of macroscopic placental lesions. We also developed a model to detect prostate cancer in needle biopsies. We designed experiments wherein patches of contaminant tissue are randomly sampled from known slides and digitally added to patient slides and measured model performance. We measured the proportion of attention given to contaminants and examined the impact of contaminants in T-distributed Stochastic Neighbor Embedding (tSNE) feature space. Every model showed performance degradation in response to one or more tissue contaminants. DA detection balanced accuracy decreased from 0.74 to 0.69 +/- 0.01 with addition of 1 patch of prostate tissue for every 100 patches of placenta (1% contaminant). Bladder, added at 10% contaminant raised the mean absolute error in estimating gestation age from 1.626 weeks to 2.371 +/ 0.003 weeks. Blood, incorporated into placental sections, induced false negative diagnoses of intervillous thrombi. Addition of bladder to prostate cancer needle biopsies induced false positives, a selection of high-attention patches, representing 0.033mm2, resulted in a 97% false positive rate when added to needle biopsies. Contaminant patches received attention at or above the rate of the average patch of patient tissue. Tissue contaminants induce errors in modern ML models. The high level of attention given to contaminants indicates a failure to encode biological phenomena. Practitioners should move to quantify and ameliorate this problem.

Machine learning classification of placental villous infarction, perivillous fibrin deposition, and intervillous thrombus

INTRODUCTION

Placental parenchymal lesions are commonly encountered and carry significant clinical associations. However, they are frequently missed or misclassified by general practice pathologists. Interpretation of pathology slides has emerged as one of the most successful applications of machine learning (ML) in medicine with applications ranging from cancer detection and prognostication to transplant medicine. The goal of this study was to use a whole-slide learning model to identify and classify placental parenchymal lesions including villous infarctions, intervillous thrombi (IVT), and perivillous fibrin deposition (PVFD).

METHODS

We generated whole slide images from placental discs examined at our institution with infarct, IVT, PVFD, or no macroscopic lesion. Slides were analyzed as a set of overlapping patches. We extracted feature vectors from each patch using a pretrained convolutional neural network (EfficientNetV2L). We trained a model to assign attention to each vector and used the attentions as weights to produce a pooled feature vector. The pooled vector was classified as normal or 1 of 3 lesions using a fully connected network. Patch attention was plotted to highlight informative areas of the slide.

RESULTS

Overall balanced accuracy in a test set of held-out slides was 0.86 with receiver-operator characteristic areas under the curve of 0.917-0.993. Cases of PVFD were frequently miscalled as normal or infarcts, the latter possibly due to the perivillous fibrin found at the periphery of infarctions. We used attention maps to further understand some errors, including one most likely due to poor tissue fixation and processing.

DISCUSSION

We used a whole-slide learning paradigm to train models to recognize three of the most common placental parenchymal lesions. We used attention maps to gain insight into model function, which differed from intuitive explanations.

Placental histology predicted adverse outcomes in extremely premature neonates in Norway-population-based study

AIM

We evaluated the role of placental pathology in predicting adverse outcomes for neonates born extremely preterm (EPT) before 28 weeks of gestation.

METHODS

This was a prospective observational study of 123 extremely preterm singletons born in a hospital in western Norway, and the placentas were classified according to the Amsterdam criteria. The associations between histologic chorioamnionitis (HCA), by the presence or the absence of a foetal inflammatory response (FIR+ or FIR-), maternal vascular malperfusion (MVM) as a whole and adverse neonatal outcomes were evaluated by logistic regression analyses. Adverse outcomes were defined as perinatal death, necrotising enterocolitis (NEC), bronchopulmonary dysplasia (BPD), brain pathology by magnetic resonance imaging at term-equivalent age, retinopathy of prematurity and early-onset neonatal sepsis. The results are reported as odds ratios (ORs) with 95% confidence intervals (CIs).

RESULTS

HCA was associated with NEC (OR 12.2, 95% CI 1.1 to 137.1). HCA/FIR+ was associated with BPD (OR 14.9, 95% CI 1.8-122.3) and brain pathology (OR 9.8, 95% CI 1.4-71.6), but HCA/FIR- was not. The only neonatal outcome that MVM was associated with was low birthweight.

CONCLUSION

Placental histology provided important information when assessing the risk of adverse neonatal outcomes following EPT birth.

Maternal-Fetal Inflammation in the Placenta and the Developmental Origins of Health and Disease

Events in fetal life impact long-term health outcomes. The placenta is the first organ to form and is the site of juxtaposition between the maternal and fetal circulations. Most diseases of pregnancy are caused by, impact, or are reflected in the placenta. The purpose of this review is to describe the main inflammatory processes in the placenta, discuss their immunology, and relate their short- and long-term disease associations. Acute placental inflammation (API), including maternal and fetal inflammatory responses corresponds to the clinical diagnosis of chorioamnionitis and is associated with respiratory and neurodevelopmental diseases. The chronic placental inflammatory pathologies (CPI), include chronic villitis of unknown etiology, chronic deciduitis, chronic chorionitis, eosinophilic T-cell vasculitis, and chronic histiocytic intervillositis. These diseases are less-well studied, but have complex immunology and show mechanistic impacts on the fetal immune system. Overall, much work remains to be done in describing the long-term impacts of placental inflammation on offspring health.

Is obesity in pregnancy associated with signs of chronic fetal hypoxia?

INTRODUCTION

The prevalence of obesity in pregnancy is increasing worldwide. Maternal obesity increases risks of severe fetal and neonatal complications. The underlying pathophysiological mechanisms are unclear. One possible contributing factor could be chronic fetal hypoxia. The aim of this study was to compare placentas from women with and without obesity with respect to placental lesions, which could reflect compensatory mechanisms in response to chronic fetal hypoxia as well as lesions possibly leading to chronic fetal hypoxia. In addition, levels of erythropoietin in cord blood were compared between offspring of lean and obese women.

MATERIAL AND METHODS

This cohort study included 180 women with uneventful, full-term, singleton pregnancies, out of which 91 lean women had a body mass index (BMI) of 18.5-24.9 kg/m² and 89 women had obesity (BMI ≥30 kg/m² ). Women were recruited at Södersjukhuset between 16 October 2018 and 2 December 2019. Placentas were investigated by two senior perinatal pathologists, who were blinded for maternal BMI. Cord blood was analyzed for levels of erythropoietin.

RESULTS

Levels of erythropoietin in cord blood increased with maternal BMI (P = .01, β = 0.97, 95% CI 0.27-1.68). There was no difference between placentas of obese and lean women in number of placental lesions reflecting chronic fetal hypoxia or in lesions that could possibly lead to chronic fetal hypoxia.

CONCLUSIONS

This study of term and uneventful pregnancies demonstrated a positive association between maternal obesity and concentrations of erythropoietin in cord blood at birth. This finding supports the hypothesis of chronic fetal hypoxia as a risk factor for complications in the pregnancies of obese women. There were no differences in lesions associated with hypoxia between placentas of obese and lean women.

Placental Pathology, Cerebral Blood Flow, and Intraventricular Hemorrhage in Preterm Infants: Is There a Link?

BACKGROUND

There is growing evidence to support an association between placental inflammation and neurological sequelae of preterm infants. The goal of this study is to evaluate the relationship between placental pathology, post-natal Doppler cerebral resistive indices (RI's), and intraventricular hemorrhage (IVH) in premature infants.

METHODS

In a retrospective cohort study, preterm infants born between 23 0/7 and 32 6/7 weeks' gestation at Parkland Hospital were examined with placental pathology and serial ultrasound Doppler to evaluate for the primary outcome of IVH and death.

RESULTS

A total of 255 infants were included, and 166 (65%) had at least one significant placental pathology, most commonly chorioamnionitis. Infants with placental pathologies were significantly more likely to have mothers with clinical chorioamnionitis and to have lower gestational ages. There was no observed association between placental pathology and IVH or death. Secondary analysis demonstrated that resistive indices obtained from the first and second head ultrasounds were not different in infants with IVH.

CONCLUSION

In this study, we observed a high rate of placental pathologies but no alterations in cerebral indices on ultrasound, or differences in rates of IVH or death. Additional studies are necessary to delineate the relationship between placental pathology, white matter brain injury, and outcomes.

Impact of multiple placental pathologies on neonatal death, bronchopulmonary dysplasia, and neurodevelopmental impairment in preterm infants

BACKGROUND

To determine the association of placental pathology, including multiple placental lesions, with the occurrence and severity of bronchopulmonary dysplasia (BPD), death, and neurodevelopmental impairment (NDI) in preterm infants.

METHOD

A retrospective cohort study of neonates <29 weeks gestational age (GA) born at Parkland Hospital from 08/2009 to 08/2012. Infants were stratified as follows: Group 1: no significant placental pathology; Group 2: single significant placental lesion; and Group 3: ≥2 placental lesions (multiple lesions). Primary outcome was death and/or BPD. Two-year neurodevelopmental follow-up was compared.

RESULTS

In all, 42% (100/241) of infants had one placental lesion, and 34% (82/241) ≥2 lesions. As the number of the pathologic lesions increased (no lesions vs. 1 vs. ≥2), the occurrence of death or BPD increased (25%, 37%, and 52%, respectively; P = 0.004). Moreover, infants with multiple pathologic lesions were more likely to have NDI (29%, 29%, and 46%, respectively; P = 0.03). After logistic regression, infants with multiple pathologic lesions were more likely to develop moderate-to-severe BPD [P < 0.01; OR 3.9 (1.5-10.1)] but not NDI.

CONCLUSION(S)

Neonates <29 weeks GA with multiple placental pathologic lesions have an increased risk for developing BPD, suggesting an interaction between placental inflammation and vascular pathology and the pathogenesis of BPD; however, the risk of NDI is not increased.

Epigenome-wide analysis in newborn blood spots from monozygotic twins discordant for cerebral palsy reveals consistent regional differences in DNA methylation

Background

Cerebral palsy (CP) is a clinical description for a group of motor disorders that are heterogeneous with respect to causes, symptoms and severity. A diagnosis of CP cannot usually be made at birth and in some cases may be delayed until 2–3 years of age. This limits opportunities for early intervention that could otherwise improve long-term outcomes. CP has been recorded in monozygotic twins discordant for the disorder, indicating a potential role of non-genetic factors such as intrauterine infection, hypoxia-ischaemia, haemorrhage and thrombosis. The aim of this exploratory study was to utilise the discordant monozygotic twin model to understand and measure epigenetic changes associated with the development of CP.

Methods

We performed a genome-wide analysis of DNA methylation using the Illumina Infinium Human Methylation 450 BeadChip array with DNA from newborn blood spots of 15 monozygotic twin pairs who later became discordant for CP. Quality control and data preprocessing were undertaken using the minfi R package. Differential methylation analysis was performed using the remove unwanted variation (RUVm) method, taking twin pairing into account in order to identify CP-specific differentially methylated probes (DMPs), and bumphunter was performed to identify differentially methylated regions (DMRs).

Results

We identified 33 top-ranked DMPs based on a nominal p value cut-off of p < 1 × 10⁻⁴ and two DMRs (p < 1 × 10⁻³) associated with CP. The top-ranked probes related to 25 genes including HNRNPL, RASSF5, CD3D and KALRN involved in immune signalling pathways, in addition to TBC1D24, FBXO9 and VIPR2 previously linked to epileptic encephalopathy. Gene ontology and pathway analysis of top-ranked DMP-associated genes revealed enrichment of inflammatory signalling pathways, regulation of cytokine secretion and regulation of leukocyte-mediated immunity. We also identified two top-ranked DMRs including one on chromosome 6 within the promoter region of LTA gene encoding tumour necrosis factor-beta (TNF-β), an important regulator of inflammation and brain development. The second was within the transcription start site of the LIME1 gene, which plays a key role in inflammatory pathways such as MAPK signalling. CP-specific differential DNA methylation within one of our two top DMRs was validated using an independent platform, MassArray EpiTyper.

Conclusions

Ours is the first epigenome-wide association study of CP in disease-discordant monozygotic twin pairs and suggests a potential role for immune dysfunction in this condition.

Electronic supplementary material

The online version of this article (10.1186/s13148-018-0457-4) contains supplementary material, which is available to authorized users.

Placental examination: prognosis after delivery of the growth-restricted fetus

PURPOSE OF REVIEW

This article describes the role of placental examination in the prognostic evaluation of fetal growth restriction (FGR) infants.

RECENT FINDINGS

A new comprehensive placental classification system was reported. Maternal underperfusion, fetal thrombotic vasculopathy (FTV), villitis (including villitis of unknown etiology and infectious villitis), inflammation, and immature/dysmature villi are important factors affecting FGR prognosis, whereas genomic imprinting is a key factor affecting growth and diseases, as well as placental abnormality.

SUMMARY

We discuss the role of placental examination in determining FGR prognosis. Maternal underperfusion, fetal thrombotic vasculopathy, and villitis (including villitis of unknown etiology and infectious villitis) are the most important findings affecting FGR prognosis. Although limited, data have suggested an association of inflammation and immature/dysmature villi with postnatal growth in FGR infants. Placental size also contributes postnatally through fetal programming. In addition, placental imprinting can be a key of pre and postnatal growth and diseases, including imprinting disorders, as well as placental abnormalities such as placental mesenchymal dysplasia.

Platelet reactivity in twin pregnancies

BACKGROUND

Gestational thrombocytopenia is more likely to occur in twin than singleton pregnancies. However, it is unclear whether platelets are more reactive in twin than singleton pregnancies.

METHODS

Changes in spontaneous platelet aggregation and concomitant fall in platelet count were examined over 90min after blood sampling in 171 and 52 citrated whole blood (CWB) samples from 59 and 17 women with singleton and twin pregnancies, respectively. Soluble P-selectin (sP-selectin) levels in the plasma were also determined.

RESULTS

CWB 60min after blood sampling during 2nd trimester exhibited significantly larger numbers of platelet aggregates (1297±1600 vs. 497±432/μl, P=0.040) concomitant with significantly greater net decrease in platelet count (152±55 vs. 115±45×10(9)/μl, P=0.036) in twin than singleton pregnancies, respectively. This was followed by significantly lower 3rd trimester platelet count (181±43 vs. 229±62×10(9)/l, P=0.009) with significantly greater mean platelet volume (8.0±1.2 vs. 7.1±1.1fl, P=0.021) in twin than singleton pregnancies, respectively. The 3rd trimester sP-selectin per platelet was significantly higher in twin than singleton pregnancies.

CONCLUSIONS

Platelets were more reactive in the 2nd trimester of twin than singleton pregnancies. This enhanced platelet reactivity may explain the decreased platelet count in the 3rd trimester of twin pregnancy.