Morphological study of the placenta in deliveries with pre‐eclampsia: Results from a prospective, observational study in India and Pakistan (PURPOSe)

To compare placental findings in women with and without pre‐eclampsia.


| I N TRODUC T ION
Hypertensive disorders of pregnancy (HDP) are multisystem disorders that include pre-eclampsia, gestational hypertension, chronic hypertension and superimposed pre-eclampsia.These conditions are estimated to contribute to approximately 14% of maternal mortality worldwide. 1ommunity-based studies have shown that these conditions are associated with about 10% of all pregnancies in low-and middle-income countries (LMICs). 2 They are characterised by the presence of high blood pressure (>140/90 mmHg), and sometimes proteinuria or convulsions.HDP are also related to adverse fetal and neonatal outcomes, including fetal growth restriction, low birthweight, low Apgar scores, need for resuscitation, neonatal intensive care unit admissions, stillbirths and neonatal deaths. 3][6] Placental abnormalities have also been implicated in a range of adverse pregnancy outcomes, but research has been limited by inconsistent definitions across studies.Furthermore, few studies have been conducted in LMICs, which have a disproportionate burden of adverse pregnancy outcomes.
To better understand causes of neonatal death and stillbirth in LMICs, including the relationship of maternal conditions such as HDP, we conducted a prospective observational study in India and Pakistan, known as the PURPOSe study. 7n PURPOSe, all placentas of women who were enrolled were evaluated using the Amsterdam Criteria, a standardised approach to characterise placental lesions. 8In this paper, we further explore the relationship of HDP and placental conditions among women with preterm births or stillbirth.We present the morphological characteristics of placentas of mothers with pre-eclampsia (defined as hypertension and proteinuria) and compare these findings with the placentas of women with no hypertension and no proteinuria.

| M ET HODS
The PURPOSe study was a prospective, observational study to determine the causes of neonatal death among preterm infants and stillbirths.In this sub-study we were interested in characterising the placental findings in women with preeclampsia-defined as women with documented hypertension and proteinuria, compared with women without pre-eclampsia-defined as those women with no hypertension and no proteinuria.The study was conducted in two sites in south Asia.One was based in Davanagere, India (three public and private hospitals) and one in Karachi, Pakistan (one public maternity hospital and one children's referral hospital).

| Data collection
The medical and obstetric history, and data from the current delivery were collected.The presence of hypertension was determined from the hospital records.Reports of urine protein testing were collected and recorded.Cases defined as pre-eclampsia required both a clinical diagnosis of hypertension (blood pressure of ≥140/90 mmHg) and proteinuria.Women without pre-eclampsia were defined as those with neither clinical hypertension nor proteinuria.Gestational age was determined based on ultrasound when available, or otherwise by using last menstrual period and clinical assessment by the clinician.
The placentas were collected, immediately stored in 10% formalin and transported to pathology departments in study hospitals for gross and microscopic examinations.All placental evaluations were conducted by trained study pathologists who used the Amsterdam Consensus criteria to define placental lesions. 8The specific histological findings needed for the diagnoses of maternal vascular malperfusion and fetal vascular malperfusion are shown in Figure 1.Inflammatory lesions included chorioamnionitis, funisitis, villitis and intervillitis and were also summarised as 'any inflammatory lesion'.

| Statistical analyses
Data were collected, reviewed, and edited at each study site by the research team.Additional edits and the study pre-eclampsia compared with 33.8% without pre-eclampsia.We also evaluated the inflammatory lesions by whether the mother had or did not have pre-eclampsia.When all inflammatory lesions were considered, women with pre-eclampsia had significantly fewer inflammatory lesions than those women without pre-eclampsia (17.1% versus 23.6% p = 0.001).Each of the specific inflammatory lesions was less common in placentas of women with pre-eclampsia than those with chorioamnionitis (16.1% versus 21.9%, p = 0.004) and funisitis (1.5% versus.5.1%, p = 0.0004).Conclusions: Of placental lesions in women with pre-eclampsia, maternal vascular malperfusion was the most common.Inflammatory lesions were less common in women with pre-eclampsia.

K E Y W O R D S
inflammatory lesions, maternal vascular malperfusion, placenta, pre-eclampsia analyses were performed centrally at the Research Triangle Institute (RTI).We restricted the analyses to singleton births in which the placenta was available and a histological evaluation had been completed.As a result, there were two groups, cases with hypertension and urine protein present, classified as pre-eclampsia, and those with neither hypertension nor urine protein present.We compared placental findings for each group using descriptive statistics.Analyses were done overall.All statistical analyses were performed using SAS (v.7.1; Cary, NC, USA).

| Ethics approvals
The study was reviewed and approved by the ethics review committees at the Aga Khan University (Karachi, Pakistan), KLE Academy of Higher Education and Research (Belagavi, India), J.J.M. Medical College (Davangere, India), and RTI International (Durham, NC USA).All women provided informed written consent before participating in the study.

| R E SU LTS
Women were recruited at delivery if they had a stillbirth at any gestational age, a liveborn preterm baby, or a term live birth (controls).Figure 2 summarises participant enrolment.In all, 5684 women were screened at delivery, 5160 (90.8%) were eligible, and 4652 completed the study.A total of 3067 births were singletons with a placental evaluation available.
Of these pregnancies, 733 women had hypertension, proteinuria, and a placental examination performed, and 2334 women had placental examination, but neither hypertension nor proteinuria.These two groups formed our study population.The placentas of these two groups were compared using the Amsterdam Criteria for both malperfusion and inflammatory lesions.
We first considered the malperfusion lesions, separated using the Amsterdam Criteria into maternal vascular malperfusion and fetal vascular malperfusion lesions (Table 1).There were large and significant differences in the rate of maternal vascular malperfusion between those mothers with (57.3%) and without (37.1%)pre-eclampsia (p < 0.0001).However, the small differences in fetal vascular hypertension between mothers with (17.1%) and without (14.8%)pre-eclampsia were not statistically significant (p = 0.6118).When placentas were classified by whether they were 'histologically normal' or not, if they were from pre-eclamptic pregnancies, 61.3% were classified as abnormal whereas if there was no pre-eclampsia, only 45.0% were classified as histologically abnormal (p < 0.0001).
We also considered rates of placental maternal vascular malperfusion in women with and without pre-eclampsia in women with a stillbirth, preterm neonatal death or term live birth (Table 2).In women at term with no pre-eclampsia, 16.7% of the placentas had features of maternal vascular malperfusion.Women with stillbirths with pre-eclampsia (79.9%) had more maternal vascular malperfusion lesions than those without pre-eclampsia (51.8%).The same was true for women with preterm live births with pre-eclampsia (49.7%) and without pre-eclampsia (33.8%).In addition to examining the overall categories of malperfusion lesions, we also evaluated the frequency of specific components of maternal and fetal vascular malperfusion in women with and without pre-eclampsia (Table 3).The most common component of the lesions of maternal vascular malperfusion in placentas of women with pre-eclampsia were infarcts, present in 49.6%, compared with 16.5% in placentas of women without pre-eclampsia (p < 0.0001).In both women with and women without pre-eclampsia, distal villous hypoplasia was the second most common placental lesion; 36.6% in placentas of women with pre-eclampsia, and 23.7% in women without pre-eclampsia.None of the lesions of fetal malperfusion were noted in more than 10% of placentas, and none were significantly different in placentas of women with and without pre-eclampsia.
We also evaluated inflammatory lesions by whether the mother had pre-eclampsia (Table 4).When all inflammatory lesions were considered, women with pre-eclampsia had significantly fewer inflammatory lesions than those without T A B L E 2 Maternal vascular malperfusion by pre-eclampsia status and birth outcome.pre-eclampsia (17.1% versus 23.6% p = 0.001).Each of the specific inflammatory lesions was less common in placentas of women with pre-eclampsia than in women without pre-eclampsia with chorioamnionitis (16.1% versus 21.9%, p = 0.004) and funisitis (1.5% versus 5.1%, p = 0.0004) significantly less so.

| DISCUS SION
In this PURPOSe sub-study, we described placental lesions in pregnancies with and without pre-eclampsia.We used the Amsterdam Criteria to define the lesions we evaluated, including placental maternal and fetal vascular malperfusion and inflammatory lesions consisting of chorioamnionitis, funisitis, villitis and intervillitis.The most important observation from this evaluation is that women with preeclampsia had significantly more maternal vascular malperfusion lesions, but not more fetal vascular malperfusion lesions.0][11] The findings related to fetal vascular malperfusion are justifiable because it is more associated with intrauterine growth restriction, fetal death, poor perinatal outcomes and neurodevelopmental sequlelae than with the development of pre-eclampsia. 12In women with pre-eclampsia, the prevalence of chorioamnionitis and funisitis were significantly lower than in those women without pre-eclampsia.][15] The most common maternal vascular malperfusion lesions in women with pre-eclampsia were infarcts (49.5%) and distal villous hypoplasia (36.6%).None of the lesions of fetal malperfusion were noted in more than 10% of the placentas.Inflammatory lesions were less common than malperfusion lesions.They occurred in 16.7% overall, and the most common inflammatory lesion was chorioamnionits (16.5%).There is evidence to show such placental lesions in pre-eclampsia. 16,17his study had a number of strengths and weaknesses.Strengths included the large number of placentas evaluated using the Amsterdam Criteria.The study was performed in two countries in south Asia, providing data about placental pathology in an area of the world where placentas are rarely evaluated histologically.The major weakness was the lack of a population-based sample.Instead, although all the women were clearly defined as having pre-eclampsia or not, women were eligible for the study because they had a stillbirth, preterm live birth, or a normal pregnancy at term.For this reason, we limited the analysis to those pregnancies with clearly defined characteristics of documented hypertension and the presence of proteinuria and a second group with documented evidence that neither hypertension nor proteinuria was present.

| CONCLUSIONS
There is substantial evidence that placental maternal vascular malperfusion is either causal for pre-eclampsia or at least on the causal pathway.This study provides additional evidence supporting the relationship between maternal vascular malperfusion and pre-eclampsia in south Asian populations.The study also suggests that there is not a relationship between fetal vascular malperfusion and pre-eclampsia.

AC K NO W L E D GE M E N T S
The authors wish to thank the staff in India and Pakistan who carried out the study and all the women and their families who provided consent for the study at a difficult and traumatic time in their lives.

AU T HOR C ON T R I BU T ION S
SY, VK, GG, SMD, SSG, EMM, SS and RLG conceptualised the study.KJ performed the data analysis.VK, ZU, NKG and IA performed the histological analyses and HY, SD, SG, SB, MS and SST managed the patient care.SY wrote the first draft with inputs from RLG, EMM, SSG and VK.All the authors approved the final draft.

DATA AVA I L A BI L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.

F I G U R E 1 F
Amsterdam criteria for placental evaluation.Placental morphology by pre-eclampsia status.
Features of maternal and fetal vascular malperfusion by pre-eclampsia status.
a Hypothesis test results: p values calculated from Cochran-Mantel-Haenszel test adjusted by site.

Overall p value Pre-eclampsia No pre-eclampsia
Placental inflammation by pre-eclampsia status.