Observer reliability in assessing placental maturity by histology

Mapping cell-to-tissue graphs across human placenta histology whole slide images using deep learning with HAPPY

Accurate placenta pathology assessment is essential for managing maternal and newborn health, but the placenta's heterogeneity and temporal variability pose challenges for histology analysis. To address this issue, we developed the 'Histology Analysis Pipeline.PY' (HAPPY), a deep learning hierarchical method for quantifying the variability of cells and micro-anatomical tissue structures across placenta histology whole slide images. HAPPY differs from patch-based features or segmentation approaches by following an interpretable biological hierarchy, representing cells and cellular communities within tissues at a single-cell resolution across whole slide images. We present a set of quantitative metrics from healthy term placentas as a baseline for future assessments of placenta health and we show how these metrics deviate in placentas with clinically significant placental infarction. HAPPY's cell and tissue predictions closely replicate those from independent clinical experts and placental biology literature.

A validation of placental pathology reports by ethnicity in New Zealand, through systematic analysis of histological slides

INTRODUCTION

Reliability studies of placental examination have shown differing interobserver agreement for certain pathological features, a lack of uniform reporting criteria and variable experience among pathologists. In previous analyses we have shown that placental pathology differs by ethnicity. This validation study was performed to investigate whether bias related to ethnicity is a feature of placental pathology reporting in New Zealand (NZ).

METHODS

199 of 1726 eligible perinatal death cases between 2008 and 2017 were selected at random for this audit-type study, including 51 cases from South Asian, Māori and NZ European ethnicity and 46 cases from Pacific mothers. Stored histology slides were blinded and re-examined by an experienced perinatal pathologist, and linked to the corresponding original pathology report. Interobserver agreement (overall, by ethnicity and by gestational age) was described by proportional differences and kappa coefficients.

RESULTS

Total interobserver agreement between original placental reporting and the validation review was 89.7 %, which differed by pathological feature. There was generally more underreporting than overreporting (3.6 % and 6.7 %, respectively). There was little disagreement by ethnicity (decidual vasculopathy [p = 0.03]), although there were more differences by gestational age (villous morphology [p < 0.01], chorioamnionitis [p = 0.03], high-grade villitis of unknown etiology [p < 0.01], and placental haemorrhage [p = 0.03]).

DISCUSSION

No systematic bias in placental pathology reporting in NZ was identified by ethnicity or gestational age, as observed differences could be related to the underlying prevalence of pathology. We identified more underreporting than overreporting of pathology in the original reports, emphasizing the importance of placental investigation by specialised perinatal pathologists.

Placental contribution to neonatal encephalopathy

Placental assessment, although currently underused, can inform our understanding of the etiology and timing of Neonatal Encephalopathy (NE). We review our current understanding of the links between placental dysfunction and NE and how this information may inform clinical decisions, now and in the future, emphasizing the four major placental lesions associated with NE. In addition, we discuss maternal and fetal factors that are hypothesized to contribute to specific placental pathologies, especially innate or acquired thrombophilias. We outline the importance of assessing placenta across trimesters and after delivery. As this field continues to evolve, currently available placental histopathological examination methods may need to be combined with advanced prenatal molecular and imaging assessments of placenta and be applied in well-designed studies in large representative populations to better define the links between placental dysfunction and NE.

Disorders of placental villous maturation are present in one-third of cases with spontaneous preterm labor

OBJECTIVES

Spontaneous preterm labor is an obstetrical syndrome accounting for approximately 65-70% of preterm births, the latter being the most frequent cause of neonatal death and the second most frequent cause of death in children less than five years of age worldwide. The purpose of this study was to determine and compare to uncomplicated pregnancies (1) the frequency of placental disorders of villous maturation in spontaneous preterm labor; (2) the frequency of other placental morphologic characteristics associated with the preterm labor syndrome; and (3) the distribution of these lesions according to gestational age at delivery and their severity.

METHODS

A case-control study of singleton pregnant women was conducted that included (1) uncomplicated pregnancies (controls, n=944) and (2) pregnancies with spontaneous preterm labor (cases, n=438). All placentas underwent histopathologic examination. Patients with chronic maternal diseases (e.g., chronic hypertension, diabetes mellitus, renal disease, thyroid disease, asthma, autoimmune disease, and coagulopathies), fetal malformations, chromosomal abnormalities, multifetal gestation, preeclampsia, eclampsia, preterm prelabor rupture of the fetal membranes, gestational hypertension, gestational diabetes mellitus, and HELLP (hemolysis, elevated liver enzymes and low platelet count) syndrome were excluded from the study.

RESULTS

Compared to the controls, the most prevalent placental lesions among the cases were the disorders of villous maturation (31.8% [106/333] including delayed villous maturation 18.6% [62/333] vs. 1.4% [6/442], q<0.0001, prevalence ratio 13.7; and accelerated villous maturation 13.2% [44/333] vs. 0% [0/442], q<0.001). Other lesions in decreasing order of prevalence included hypercapillarized villi (15.6% [68/435] vs. 3.5% [33/938], q<0.001, prevalence ratio 4.4); nucleated red blood cells (1.1% [5/437] vs. 0% [0/938], q<0.01); chronic inflammatory lesions (47.9% [210/438] vs. 29.9% [282/944], q<0.0001, prevalence ratio 1.6); fetal inflammatory response (30.1% [132/438] vs. 23.2% [219/944], q<0.05, prevalence ratio 1.3); maternal inflammatory response (45.5% [195/438] vs. 36.1% [341/944], q<0.01, prevalence ratio 1.2); and maternal vascular malperfusion (44.5% [195/438] vs. 35.7% [337/944], q<0.01, prevalence ratio 1.2). Accelerated villous maturation did not show gestational age-dependent association with any other placental lesion while delayed villous maturation showed a gestational age-dependent association with acute placental inflammation (q-value=0.005).

CONCLUSIONS

Disorders of villous maturation are present in nearly one-third of the cases of spontaneous preterm labor.

Placental villous hypermaturation is associated with improved neonatal outcomes

INTRODUCTION

Accelerated placental maturation is considered a sign of maternal vascular malperfusion, and is often interpreted as an adaptive, compensatory response by the placenta. We tested this interpretation by comparing outcomes in pregnancies with and without accelerated maturation.

METHODS

Using data from the National Collaborative Perinatal Project, we categorized preterm placentas (24-34 weeks, inclusive; 2525 births) by whether they showed placental villous hypermaturation (PVH), i.e., had the appearance of a placenta of 37 weeks or over upon microscopic examination. We assessed whether PVH was associated with maternal race, maternal BMI, fetal sex, type of preterm birth, preeclampsia, signs of infection or inflammation or placental abruption. We also assessed whether placentas showing PVH were associated with improved outcomes in terms of survival, Apgar score, or oxygen use.

RESULTS

PVH was more common in preeclamptic pregnancies and less common in pregnancies complicated by placental abruption or showing signs of placental infection or inflammation. Adjusting for potentially confounding factors, PVH was associated with reduced odds of fetal death, death between birth and 120 days of age, low Apgar scores and oxygen use. PVH was also associated with higher birthweights for gestational age. When correcting for the effect of birthweight, the association between PVH and reduced fetal and neonatal death remained significant.

DISCUSSION

Accelerated placental maturation, as manifested by PVH, is associated with improved outcomes. Our work therefore supports the hypothesis that accelerated maturation is a compensatory response by the placenta to improve its transport capacity in specific pregnancy complications.

Re-view and view on maturation disorders in the placenta

Until delivery, the placenta plays an important mediator role between mother and fetus. This unit is affected by peristatic conditions, such as acute or chronic maternal diseases, malnutrition, drugs, and others. But also genetic factors and fetal malformations due to embryonic developmental disorders may contribute to macroscopically visible changes and functional disorders of the placenta. In a constantly ongoing progress of maturation, the placenta records and saves changes due to fetal distress partly as maturation disorders. Understanding of maturation disorders might, therefore, be an important contribution to a better understanding of influences on villous differentiation and might improve follow up and fetal outcome to reduce recurrence risk. However, an internationally unified classification system of maturation disorders does not exist. In this review, terminology, trials, and classifications of villous maturation disorders are summed up and compared, to pinpoint the need of agreement on an international unified and reproducible classification of maturation disorders.

"Delayed Villous Maturation" in Placental Reporting: Concordance among Consultant Pediatric Pathologists at a Single Specialist Center

Delayed villous maturation (DVM) has been associated with an increased risk of adverse pregnancy outcome, including stillbirth, in the late third trimester, but there are limited published data. Moreover, it is recognized that the assessment of villous maturation is subjective and hampered by both intraobserver and interobserver variability. This audit aims to assess concordance in the reporting of DVM among pediatric pathologists at a single specialist center to improve reproducibility of this potentially important diagnosis. This is a retrospective review of singleton placentas from pregnancies at 35 weeks gestation or greater submitted for histopathologic examination between June 2013 and December 2013. Placental slides were reviewed independently by 4 pediatric pathologists, blinded to the original report, apart from gestational age; villous maturation was assessed as appropriate, accelerated, or delayed for the stated gestational age. A total of 464 placental histopathology reports were reviewed, of which 164 were greater than 35 weeks gestation; of those, 42 (26%) were originally reported as DVM. Following the audit slide review, 38 cases (23%) were assessed to show DVM by at least 1 pathologist. Consensus, with at least 3 pathologists agreeing to a diagnosis of DVM, was achieved in only 14 cases (9% of all cases reviewed; 37% of all cases called DVM). However, the proportion of overall agreement between 2 of the pathologists was 0.92. Concordance for DVM is poor among pathologists and subject to much interobserver variability. Consistency may be improved by consensus histologic review of all the placentas in which the diagnosis of DVM is being considered and stringent application of the published diagnostic criteria.

Ultrastructure of Placenta of Gravidas with Gestational Diabetes Mellitus

Objectives. Gestational diabetes mellitus (GDM) leads to an abnormal placental environment which may cause some structural alterations of placenta and affect placental development and function. In this study, the ultrastructural appearances of term placentas from women with GDM and normal pregnancy were meticulously compared. Materials and Methods. The placenta tissues of term birth from 10 women with GDM and 10 women with normal pregnancy were applied with the signed informed consent. The morphology of fetomaternal interface of placenta was examined using light microscopy (LM) and transmission electron microscopy (TEM). Results. On LM, the following morphological changes in villous tissues were found in the GDM placentas when compared with the control placentas: edematous stroma, apparent increase in the number of syncytial knots, and perivillous fibrin deposition. On TEM, the distinct ultrastructural alterations indicating the degeneration of terminal villi were found in the GDM placentas as follows: thickening of the basal membrane (BM) of vasculosyncytial membrane (VSM) and the VSM itself, significantly fewer or even absent syncytiotrophoblastic microvilli, swollen or completely destroyed mitochondria and endoplasmic reticulum, and syncytiotrophoblasts with multiple vacuoles. Conclusion. Ultrastructural differences exist between GDM and control placentas. The differences of placenta ultrastructure are likely responsible for the impairment of placental barrier and function in GDM.

Placental hypoxia: the lesions of maternal malperfusion

The placental lesions classically ascribed to placental hypoxia, here denoted maternal malperfusion (MMP), are among the more significant that a placental pathologist may encounter. Yet the appearance of these lesions may be subtle, and the clinical implication of their diagnosis is frequently unclear. The aim of this review is to provide a more nuanced perspective on the clinical utility of placental pathology for the detection of MMP. The review will first detail MMP lesions in the placenta and discuss their associations with pregnancy complications. The review will then delve into the diagnostic and interpretive difficulties of these lesions. Finally, recent research findings that may aid in the development of better diagnostic tools will be briefly discussed.

Pathologic evaluation of normal and perfused term placental tissue

This study reports for the 1st time the incidence and interobserver variation of morphologic findings in a series of 34 term placentas from pregnancies with normal outcome used for perfusion studies. Histologic evaluation of placental tissue is challenging, especially when it comes to defining "normal tissue" versus "pathologic lesions." A scoring system for registration of abnormal morphologic findings was developed. Light microscopic examination was performed independently by 2 pathologists, and interobserver variation was analyzed. Findings in normal and perfused tissue were compared and selected findings were tested against success parameters from the perfusions. Finally, the criteria for frequent lesions with fair to poor interobserver variation in the nonperfused tissue were revised and reanalyzed. In the perfused tissue, the perfusion artefact "trophoblastic vacuolization," which is believed to represent dilated transtrophoblastic channels, was reproducible and significantly correlated to the perfusion marker "fetal leakage." In longer perfusions, microscopy of the perfused cotyledon revealed bacteria in the fetal vessels. This finding led to an adjustment in the perfusion protocol with addition of antibiotics to the medium. In the "normal" tissue, certain lesions were very frequent and showed only fair or poor interobserver agreement. Revised minimum criteria for these lesions were defined and found reproducible. This study has emphasized the value of pathologic examination as a supplement in placental perfusion models. Examination of the perfused cotyledon for trophoblastic vacuolization is recommended as an additional quality marker in perfusion models. The study also underlines the need for exact definitions of abnormality in frequent placental lesions.

Placental maturity, hypertensive disorders of pregnancy and birth weight

OBJECTIVES

To compare maturity of placentas from women with hypertensive disorders with those from normotensive pregnancies and to determine the relationship between placental maturity (PM) and the diagnosis of small-for-gestational-age (SGA) in the newborns.

MATERIALS AND METHODS

We examined placental stained specimens from women with normotensive pregnancies (n = 100), diagnosis of gestational hypertension (n = 38), mild (n = 10), or severe preeclampsia (n = 34) in an optical microscope. Placental Maturity Index (PMI) was calculated as the number of vasculo-syncytial membranes (VSM) in 1 mm(2) divided by VSM thickness (µm). Hypermaturity was defined as >90th percentile of the PMI from placentas of normotensive pregnancies. Newborns were classified as SGA, adequate-for-gestational-age (AGA) or large-for-gestational-age (<10th, 10-90th, and >90th percentile from weight for gestational age reference tables, respectively).

RESULTS

PMI in preeclamptic women (taking together mild and severe preeclampsia, PMI = 43.4 ± 1.6) was significantly higher than in normotensive women (PMI = 36 ± 2, p = 0.045). Hypermaturity was more frequent (p < 0.05) in placentas from women with preeclampsia than in those from normotensive women only in preterm pregnancies (<37 weeks), but not in those at term (p = 0.41). The frequency of hypermaturity in placentas from women with gestational hypertension was not statistically different than in normotensive women. Hypermaturity was also more frequent in placentas from SGA (OR = 2.63, p < 0.05) than in AGA newborns.

CONCLUSION

The PMI was increased in preeclampsia, but not in gestational hypertension. Placental hypermaturity was also associated with the diagnosis of SGA in newborns. PM might have a role in the relationship between maternal factors and SGA.

Hypoxic patterns of placental injury: a review

CONTEXT

In utero hypoxia is an important cause of perinatal morbidity and mortality and can be evaluated retrospectively to explain perinatal outcomes, to assess recurrence risk in subsequent pregnancies, and to investigate for medicolegal purposes by identification of many hypoxic placental lesions. Definitions of some placental hypoxic lesions have been applied relatively liberally, and many of them are frequently underreported. Objectives To present a comprehensive assessment of the criteria for diagnosing acute and chronic histologic features, patterns, and lesions of placental and fetal hypoxia and to discuss clinicopathologic associations and limitations of the use thereof. The significance of lesions that have been described relatively recently and are not yet widely used, such as laminar necrosis; excessive, extravillous trophoblasts; decidual multinucleate extravillous trophoblasts; and, most important, the patterns of diffuse chronic hypoxic preuterine, uterine, and postuterine placental injury and placental maturation defect, will be discussed.

DATA SOURCES

Literature review.

CONCLUSIONS

The placenta does not respond in a single way to hypoxia, and various placental hypoxic features should be explained within a clinical context. Because the placenta has a large reserve capacity, hypoxic lesions may not result in poor fetal condition or outcome. On the other hand, very acute, in utero, hypoxic events, followed by prompt delivery, may not be associated with placental pathology, and many poor perinatal outcomes can be explained by an etiology other than hypoxia. Nevertheless, assessment of placental hypoxic lesions is helpful for retrospective explanations of complications in pregnancy and in medicolegal investigation.

Placental villous hypermaturation is associated with idiopathic preterm birth

OBJECTIVE

Pregnancy complications such as intra-amniotic infection, preeclampsia, and fetal intrauterine growth restriction (IUGR) account for most cases of preterm birth (PTB), but many spontaneous PTB cases do not have a clear etiology. We hypothesize that placental insufficiency may be a potential cause of idiopathic PTB.

METHODS

Secondary analysis of 82 placental samples from women with PTB obtained from a multicenter trial of repeat versus single antenatal corticosteroids. Samples were centrally reviewed by a single placental pathologist masked to clinical outcomes. The histopathologic criterion for infection was the presence of acute chorioamnionitis defined as neutrophils marginating into the chorionic plate. Placental villous hypermaturation (PVH) was defined as a predominance of terminal villi (similar to term placenta) with extensive syncytial knotting. Idiopathic PTB comprised a group without another known etiology such as preeclampsia, IUGR or infection.

RESULTS

Acute chorioamnionitis was observed in 33/82 (40%) cases. Other known causes of PTB were reported in 18/82 (22%). The remaining 31/82 (38%) were idiopathic. The frequency of PVH in idiopathic PTB (26/31 = 84%) was similar to cases with IUGR or preeclampsia (16/18 = 89%), but significantly more common than PVH in the group with acute chorioamnionitis (10/33 = 30%) (p < 0.001).

CONCLUSIONS

PVH, which is a histologic marker of relative placental insufficiency, is a common finding in idiopathic PTB.

Delayed villous maturation of the placenta: quantitative assessment in different cohorts

Placental villous maturation is maximal in the 3rd trimester, with an abundance of terminal villi. Delayed villous maturation (DVM) of the placenta is associated with chromosomal abnormalities, gestational diabetes, and an adverse outcome. This study compares quantitative assessment of vasculo-syncytial membranes (VSM) in cases of liveborn infants, perinatal deaths, and controls. Cases were selected as follows: (1) liveborn infants with a qualitative diagnosis of DVM (n  =  15); (2) controls matched for gestational age whose placentas did not have DVM (n  =  15); (3) stillbirths (SB)/neonatal deaths (NND) showing DVM (n  =  13); and (4) SB from autopsies in which DVM was felt to be the cause of death (COD) (n  =  12). Vasculo-syncytial membranes were counted in 10 terminal villi in each of 10 consecutive high-power fields on 3 slides. Data analysis was carried out using SPSS. Liveborn cases with DVM showed statistically significantly less VSM than controls (mean 1.01 vs 2.42, P < 0.0001). The SB/NND group also showed significantly less VSM than the control group (mean 0.46 vs 2.42, P < 0.0001) and less than the liveborn DVM group (mean 0.46 vs 1.01, P  =  0.001). The COD group was significantly different from the control group (mean 0.42 vs 2.42, P < 0.0001) and the liveborn DVM group (mean 0.42 vs 1.01, P < 0.0001) but not significantly different from the SB/NND group. There is a quantitative reduction in VSM in cases of DVM compared to controls.

Diagnosing placental membrane hypoxic lesions increases the sensitivity of placental examination

CONTEXT

Two relatively unknown and recently described placental membrane hypoxic lesions (laminar necrosis and microscopic chorionic pseudocysts) have never been compared with time-honored, focal (infarction), and diffuse hypoxic lesions of placental parenchyma.

OBJECTIVE

To compare the effect on placental diagnosis of the above placental membrane hypoxic lesions and chorionic disc hypoxic lesions (infarctions and global hypoxic pattern of placental injury).

DESIGN

Twenty-three clinical (maternal and fetal) and 32 gross and microscopic placental features were retrospectively compared in 4590 placentas from a placental database built during a 13-year period: 168 placentas with at least one hypoxic disc lesion (infarct or global hypoxia) and at least one membrane lesion (microscopic chorionic pseudocysts or laminar necrosis (group 1), 750 placentas with at least one hypoxic villous lesion but no membrane lesion (group 2), 480 placentas with at least one membrane lesion but no villous lesion (group 3), and 3192 placentas with no hypoxic villous or membrane lesions (group 4).

RESULTS

Several clinical and fetal conditions and placental features known to be associated with in utero hypoxia had a statistically significant correlation with the index hypoxic placental lesions, both villous and membranous. Of placentas from patients associated with clinical conditions at risk for hypoxia, 15% featured only hypoxic membrane lesions without a chorionic disc hypoxic lesion.

CONCLUSIONS

Recognizing placental membrane hypoxic lesions increases the sensitivity of placental examination in diagnosing placental hypoxia by at least 15%. The risk of in utero hypoxia is increased when microscopic chorionic pseudocysts and laminar necrosis occur in conjunction with villous hypoxic lesions.

Modeling the variability of shapes of a human placenta

BACKGROUND

Placentas are generally round/oval in shape, but "irregular" shapes are common. In the Collaborative Perinatal Project data, irregular shapes were associated with lower birth weight for placental weight, suggesting variably shaped placentas have altered function.

METHODS

(I) Using a 3D one-parameter model of placental vascular growth based on Diffusion Limited Aggregation (an accepted model for generating highly branched fractals), models were run with a branching density growth parameter either fixed or perturbed at either 5-7% or 50% of model growth. (II) In a data set with detailed measures of 1207 placental perimeters, radial standard deviations of placental shapes were calculated from the umbilical cord insertion, and from the centroid of the shape (a biologically arbitrary point). These two were compared to the difference between the observed scaling exponent and the Kleiber scaling exponent (0.75), considered optimal for vascular fractal transport systems. Spearman's rank correlation considered p<0.05 significant.

RESULTS

(I) Unperturbed, random values of the growth parameter created round/oval fractal shapes. Perturbation at 5-7% of model growth created multilobate shapes, while perturbation at 50% of model growth created "star-shaped" fractals. (II) The radial standard deviation of the perimeter from the umbilical cord (but not from the centroid) was associated with differences from the Kleiber exponent (p=0.006).

CONCLUSIONS

A dynamical DLA model recapitulates multilobate and "star" placental shapes via changing fractal branching density. We suggest that (1) irregular placental outlines reflect deformation of the underlying placental fractal vascular network, (2) such irregularities in placental outline indicate sub-optimal branching structure of the vascular tree, and (3) this accounts for the lower birth weight observed in non-round/oval placentas in the Collaborative Perinatal Project.

Fetal vascular obstructive lesions: nosology and reproducibility of placental reaction patterns

The purpose of this study was to assemble and test the reliability of a complete set of the placental reaction patterns seen with chronic fetal vascular obstruction in the hope that this might provide a standardized diagnostic framework useful for practicing pathologists. Study cases (14 with fetal vascular obstructive lesions, 6 controls) were reviewed blindly by seven pathologists after agreement on a standard set of diagnostic criteria. Majority vote served as the gold standard and 80% of the 180 diagnoses rendered (9 diagnoses each for 20 cases) were agreed upon by at least six of the seven scores. The sensitivity of individual diagnosis relative to the group consensus averaged 83% (range, 69-100%) and specificity averaged 91% (range, 86-100%). Reproducibility was measured by unweighted kappa-values and interpreted as follows: < 0.2, poor; 0.2-0.6, fair/moderate; > 0.6, substantial. Kappa values for lesions of distal villi were generally superior to those for lesions involving large fetal vessels: avascular villi (0.49), villous stromal-vascular karyorrhexis (0.58), and villitis of unknown etiology (VUE) with stem villitis and avascular villi (0.65) versus large vessel thrombi (any vessel, 0.34; chorionic plate vessel, 0.40) and intimal fibrin cushions (recent, 0.47; remote, 0.78). Reproducibility for a global impression of any villous change consistent with chronic fetal vascular obstruction was substantial (0.63), while that for a more severe subgroup was moderate (0.44). Three points are worthy of emphasis. Our system separately recognizes, but later combines, uniformly avascular villi and villous stromal-vascular karyorrhexis as manifestations of the same underlying process. We propose that this combined group of villous lesions be dichotomized with the terms fetal thrombotic vasculopathy or extensive avascular villi (and/or villous stromal-vascular karyorrhexis) being reserved for the group with 15 or more affected terminal villi per section. Scattered foci of avascular villi (and/or villous stromal-vascular karyorrhexis) could be used to describe less severe cases. Finally, we distinguish VUE with stem villitis and avascular villi (obliterative fetal vasculopathy) as a distinct process with substantial perinatal morbidity.

Maternal vascular underperfusion: nosology and reproducibility of placental reaction patterns

Placental examination can be a useful tool for specifying the etiology, prognosis, and recurrence risk of pregnancy disorders. The purpose of this study was to test the reliability of a predetermined set of placental reaction patterns seen with maternal vascular underperfusion in the hope that this might provide a useful diagnostic framework for practicing pathologists. Study cases (14 with clinical and pathologic evidence of maternal underperfusion plus 6 controls) were evaluated for the presence or absence of 11 lesions by eight perinatal pathologists. After analysis of initial results, diagnostic criteria were refined and a second, overlapping set of cases was reviewed. The collective sensitivity, specificity, and efficiency of individual assessments for the 11 lesions relative to the group consensus ranged from 74-93% (22/33 > 90%). Reproducibility was measured by unweighted kappa-values and interpreted as follows: < 0.2 poor, 0.2-0.6 fair/moderate, > 0.6 substantial. Kappa values for lesions affecting villi and the intervillous space were increased syncytial knots (any -0.42, severe -0.50), villous agglutination (0.42), increased intervillous fibrin (0.25), and distal villous hypoplasia (0.57). Individual estimates of percent involvement for syncytial knots, intervillous fibrin, and distal villous hypoplasia were correlated with placental and fetal weight for gestational age. Extent of increased intervillous fibrin showed the strongest correlation with both placental weight ( R = -0.64) and fetal weight ( R = -0.45). Kappa values for lesions affecting maternal vessels and the implantation site were acute atherosis (0.50), mural hypertrophy of membrane arterioles (0.43), muscularized basal plate arteries (0.48), increased placental site giant cells (0.54), and immature intermediate trophoblast (0.36). Correlation of maternal vessel and implantation site lesions with the clinical diagnosis of preeclampsia showed that excessive placental site giant cells and immature intermediate trophoblast were more sensitive and efficient predictors, whereas atherosis and muscularized basal plate arteries were more specific. Kappa value for a thin umbilical cord, a possible indicator of fetal volume depletion, was 0.61. Reproducibility for a global impression of maternal vascular underperfusion, taking into account all of the above lesions, was moderate (kappa 0.54) and improved after inclusion of additional pathologic and clinical data (kappa 0.68). Adoption of this clearly defined, clinically relevant, and pathologically reproducible terminology could enhance clinicopathologic correlation and provide a more objective framework for future clinical research.

Amniotic infection syndrome: nosology and reproducibility of placental reaction patterns

Clinically responsive placental examination seeks to provide useful information regarding the etiology, prognosis, and recurrence risk of pregnancy disorders. The purpose of this study was to assemble and validate a complete set of the placental reaction patterns seen with amniotic fluid infection in the hope that this might provide a standardized diagnostic framework useful for practicing pathologists. Study cases (14 with amniotic fluid infection, 6 controls) were reviewed blindly by six pathologists after agreement on a standard set of diagnostic criteria. After analysis of initial results, criteria were refined and a second, overlapping set of cases were reviewed. Majority vote served as the gold standard. Grading and staging of maternal and fetal inflammatory responses was found to be more reproducible using a two- versus three-tiered grading system than a three- versus five-tiered staging system (overall agreement 81% vs. 71%). Sensitivity, specificity, and efficiency for individual observations ranged from 67-100% (24/30 > 90%). Reproducibility was measured by unweighted kappa values and interpreted as follows: < 0.2, poor; 0.2-0.6, fair/moderate; > 0.6, substantial. Kappa values for the 12 lesions evaluated in 20 cases by the six pathologists were: acute chorioamnionitis/maternal inflammatory response (any, 0.93; severe 0.76; advanced stage, 0.49); chronic (subacute) chorioamnionitis (0.25); acute chorioamnionitis/fetal inflammatory response (any, 0.90; severe, 0.55; advanced stage, 0.52); chorionic vessel thrombi (0.37); peripheral funisitis (0.84); acute villitis (0.90); acute intervillositis/intervillous abscesses (0.65), and decidual plasma cells (0.30). Adoption of this clearly defined, clinically relevant, and pathologically reproducible terminology could enhance clinicopathologic correlation and provide a framework for future clinical research.

Discrepancy in pathologic diagnosis of placental lesions

CONTEXT

Placentas are routinely examined by surgical pathologists, but peer review of placental diagnosis is rarely performed.

OBJECTIVE

To determine the frequency of discrepant placental diagnosis between general surgical pathologists and a pediatric pathologist.

DESIGN

One hundred fourteen placentas from infants with intrauterine growth restriction (IUGR) and 170 placentas from infants appropriate for gestational age (AGA) were reviewed for 10 lesion types using standardized criteria. The review diagnosis was compared with original reports.

RESULTS

The review identified 333 lesions, 168 in the IUGR group and 165 in AGA group. Discrepant diagnosis occurred in 137 lesions (41.1%). There was no significant difference in the frequency of discrepant diagnosis between the IUGR (44.7%) and AGA groups (37.6%) (P >.05). Most discrepancies (92.7%) were due to underdiagnosis (identified on review but not mentioned in original diagnosis), but a few (7.3%) were due to misdiagnosis (mentioned in original report but disagreed on review). The common underdiagnoses with their corresponding rates were as follows: hemorrhagic endovasculitis (84.6%), fetal thrombotic vasculopathy (75%), massive perivillous fibrin deposition (68.4%), maternal floor infarction (66.7%), retroplacental hemorrhage (60.6%), intervillous thrombus (57.1%), decidual angiopathy (33.3%), placental infarction (25.4%), acute chorioamnionitis (22.7%), and chronic villitis (21.7%). Misdiagnosis was found in 10 cases: 5 cases of infarction (review diagnosis was perivillous fibrin deposits in 4, intervillous thrombus in 1), 3 cases of acute chorioamnionitis, and 2 cases of decidual angiopathy. Among the 8 general surgical pathologists involved, the frequency of discrepant diagnosis ranged from 31.5% to 58.6% (P >.05). The intraobserver discrepancy rate for the reviewer was 4.8%, significantly lower than the discrepancy rate for the 8 general surgical pathologists.

CONCLUSION

It is common for general surgical pathologists not to recognize placental lesions, which may have clinical significance. Awareness of this deficiency, standardization of diagnostic criteria, and increased knowledge in placental pathology may improve the quality of diagnosis in this area.

Why is the placenta being ignored?

The relationship between the frequency of published recommended indications for placental pathological examination and the frequency of requests for such examination in a population-based study of term newborn encephalopathy was examined. Only 11.2% of placentas among 276 case infants and 0.7% of placentas among 564 term control infants were examined. Using the criteria set out in a consensus statement by the American College of Pathologists, all 276 cases fulfilled multiple maternal, fetal and placental indications for placental examination. Furthermore 43.3% of control infants fulfilled at least one criterion. Of the 25 case placentas that underwent pathological review, 16 were reported as having no diagnostic abnormality Six cases (24%) showed clinically important findings: four had evidence of infection, one had multiple chorangiomata and one had thrombosis and rupture of the umbilical vein. Of the three remaining placentas, one showed funisitis, one showed minor lymphohistiocytic villitis and one was from monochorionic twins. To our knowledge there are no agreed Australian guidelines for when a placenta should be submitted for pathological examination. We suggest that until guidelines based on properly designed studies are developed it may be appropriate to store all placentas for at least 72 hours. If the infant develops neurological symptoms or requires unexpected admission to a neonatal intensive care unit then placental examination may reveal important aetiological diagnostic and prognostic information.

Reliability of placental histology using archived specimens

Archived placental materials provide a resource for retrospective studies of perinatal outcomes. Using archived specimens, we evaluated inter-rater reliability in the assessment of placental histological features possibly associated with neonatal illness and long-term neurological outcome. Five expert placental pathologists independently reviewed archived placental specimens on 30 subjects born during 1983-85 in six different hospitals. Moderate to substantial agreement among the raters was observed for a variety of indicators of inflammation, presence of macrophages with pigment and indicators of villous maturity, increased syncytial knots and maternal vasculopathy. Existing pathology reports for these specimens from routine histology examinations conducted at the time of delivery were in substantial agreement with expert review for the presence of any subchorionitis/chorionitis/chorioamnionitis. In conclusion, considerable inter-rater agreement was observed for several relatively common and potentially important placental findings. Archived placental materials and pathology reports may be useful in retrospective studies.

From delivery suite to laboratory: optimizing returns from placental examination in medico-legal defence

While examinations of the placenta may reveal significant lesions that are helpful in a medico-legal context, its value maybe diminished by a lack of attention to detail in the period between delivery of the baby and receipt of the placenta in the pathology department. These fall generally into 2 groups: initial description and handling of the placenta and lack of relevant clinical information. This communication describes some of these pitfalls. By observing these details, the onus then falls on the pathologist to provide a meaningful assessment of the placenta.