Placental pathology in spontaneous and iatrogenic preterm birth: Different entities with unique pathologic features

Abstract

Introduction:

Placental pathology is an important contributor to the understanding of preterm birth and reveals major differences between spontaneous preterm birth (SPTB) and iatrogenic preterm birth (IPTB). The aim of this study was to investigate these relationships.

Methods:

Research midwives collected placentas from 1101 women with singleton pregnancies who were enrolled in the Safe Passage Study. Trained pathology technologists prepared and processed placenta specimens for macroscopic and microscopic examination by designated pathologists. Statistical analyses were done with STATISTICA version 13.

Results:

In SPTB we found more cases of accelerated villous maturation; however, the other features of maternal vascular malperfusion (MVM) were not present. The prevalence rate of funisitis was also increased. In IPTB, multiple features of MVM - accelerated villous maturation, distal villous hypoplasia, decidual arteriopathy, increased syncytial knots, increased perivillous fibrin, and prominent extravillous trophoblast were increased, as were features of fetal vascular malperfusion (FVM) - umbilical cord vessel thrombosis, avascular villi, and fetal vascular thrombosis. Increased syncytial knots were found in 26% of preterm stillbirths and in 29% of preterm infant demises as compared to 81% of IPTB infants alive at one year.

Discussion:

SPTB and IPTB differ. The detected “abnormal” accelerated villous maturation pattern in SPTB and preterm demises, suggests an inability of the placenta to adapt and may be a trigger for SPTB. Funisitis was the only inflammatory response significant for SPTB. MVM and FVM are implicated in IPTB, but not an inflammatory process.

1. Introduction

Preterm birth (PTB) has devastating maternal and neonatal consequences, accounting for 75% of perinatal mortalities and more than 50% of chronic infant morbidities [1]. Placental pathology associated with PTB may be useful in categorizing and determining its etiology.

Despite the attention to spontaneous preterm birth (SPTB) in medical research, the etiology remains largely unexplained. It is considered a syndrome caused by multiple pathologies such as infection, inflammation, malperfusion, haemorrhage, stress, uterine over-distention and other immunologically mediated processes [2].

Increasing evidence indicates that placental dysfunction, such as maternal or fetal vascular malperfusion, is an important risk factor for PTB [3,4], but information from low-income-countries is lacking. Macroscopic and microscopic placental features may illuminate the clinical understanding of preterm delivery, perhaps most demonstratable in communities with high prevalence rates [5-7]. The rate of PTB is 13.8% in a large low-to middle income community in Cape Town, South Africa [8] where this study was conducted to investigate the relationship between placental pathology and spontaneous or iatrogenic preterm birth (IPTB) and poor neonatal outcomes.

2. Methods

Women attending the antenatal clinics of Midwife Obstetric Units (MOUs) near Tygerberg Academic Hospital (TBH), Cape Town, were recruited (n = 7060) for the South African (SA) part of the Safe Passage Study (SPS) from August 2007 to January 2015 [9]. Forty percent, of the women enrolled before 24 weeks were randomly selected for an embedded study which included placental histology. The earliest ultrasound examination was used to calculate the gestational age (GA). Socio-economic data were obtained from a maternal history and demographics questionnaire. Maternal chart abstraction was done after delivery to establish pregnancy outcome. For this study we focused on women with singleton pregnancies who participated in the study for the first time, and who had placental histology available. Z-scores of the birthweights were calculated using INTERGROWTH-21st study references [10]. The study cohort was classified into SPTB (any spontaneous birth <37 weeks), IPTB and spontaneous term birth (STB).

After delivery, placentas were placed in a refrigerator at 4 °C. They were then transported to TBH and labelled with the participant’s study identity number. The only information the pathologist received was the GA at birth and whether it was a live or stillbirth.

The placentas were examined macroscopically, recorded on a standard template and the maternal and fetal surfaces photographed. Thereafter two sections of the umbilical cord (proximal and distal), a membrane roll and three full thickness sections of parenchyma (normal and lesions) were taken according to standardised protocols and processed into paraffin embedded wax blocks. Subsequently sections were cut, and haematoxylin and eosin-stained slides produced for histological examination by SA perinatal pathologists (CW/PS). For stillbirths, the slides were also examined by external perinatal pathologists. Quality control was ensured by checking inter-observer concordance between all pathologists with a standard of 90% for individual items.

The trimmed placental weight for gestation was considered small when <10th centile according to SPS centiles [11]. Macroscopic infarction was defined as firm solid area(s) extending from the basal plate with the apex towards the fetal surface. Any infarction in a preterm placenta and >5% infarction at term was considered pathological. Microscopic infarction was diagnosed as villi crowded together with trophoblastic smudging or complete necrosis. Distal villous hypoplasia was diagnosed when villi were small and stringy, elongated or appeared tiny, barely the size of a syncytial knot with increased intervillous space. Accelerated villous maturation indicates villi hyper-mature for GA and is usually associated with an increase in syncytial knots. Increased syncytial knots were reported if they were found in >33% of villi at term and >30% preterm [12]. Decidual arteriopathy included muscular hypertrophy and/or fibrinoid necrosis with or without atherosis. Increased perivillous fibrin (where the fibrin surrounds villi) and increased extravillous trophoblast (trophoblast fibrinoid islands) were also recorded.

Fetal vascular malperfusion (FVM) was defined per the Amsterdam International Consensus Criteria [13]. Avascular villi were defined as regional or diffuse and are terminal villi with loss of villous capillaries and hyalinized stroma. Villous stromal karyorrhexis was defined as regional or diffuse, demonstrating nuclear dust with extravasation of red cells into the villous stroma with preservation of the surrounding trophoblast. Chorionic plate/stem villus thrombosis was categorised into venous, arterial, both or indeterminate, focal or multifocal and included occlusive and non-occlusive thrombi. It also included intramural fibrin deposition. Microscopic thrombosis of the umbilical cord vessels was categorised into venous, arterial or both. Fetal vascular thrombosis was diagnosed when two or more of the four above-mentioned features (avascular villi, villous stromal karyorrhexis, chorionic plate/stem villus thrombosis or microscopic thrombosis of the cord) were present.

Macroscopic retroplacental haemorrhage (abruption) was defined as an indentation and/or haemorrhage involving >15% of the maternal surface of the placenta. The microscopic criteria for retroplacental haemorrhage were the presence of blood beneath the decidua and dissecting into the decidua and placental parenchyma, with congestion, and/or intravillous haemorrhage (haemorrhage into villous stroma) or associated dome-shaped infarction.

Vasculitis of the umbilical cord was recorded when one or more vessels had inflammation within the muscular wall, while funisitis of the umbilical cord was recorded when inflammation present was perivascular and constituted a fetal inflammatory response (FIR). These correspond to stage 1–2 or stage 3 per the Amsterdam Criteria [13] respectively.

Chorioamnionitis was defined when inflammatory cell types of membranes and chorionic plate was acute, chronic or eosinophillic and the location was subchorion, chorion or chorioamnion and constituted a maternal inflammatory response (MIR). Villitis was recorded when inflammatory cell types of parenchyma were chronic-specific (infective) or chronic-nonspecific (non-infective) also known as villitis of unknown etiology (VUE) where infectious causes were ruled out.

Statistical analyses utilised STATISTICA (Dell Inc. (2015). Dell Statistica (data analysis software system), version 13. software.dell.com). Continuous variables were compared versus the different groups of interest with either analysis of variance (ANOVA) or with repeated measures ANOVA with the assumption of compound symmetry i.e., equicorrelation among repetitions. Bonferroni corrections for multiple comparisons were used to identify significant differences among means. Maximum likelihood chi-square statistics were used to analyse associations between nominal variables in contingency tables. For 2x2 contingency tables odds ratios (ORs) with their 95% confidence intervals (CIs) were computed.

3. Results

This study (Fig. 1), consists of 1101 women of whom 99.8% were from mixed ancestry who had either a SPTB (n = 141 (12.8%)) or an IPTB (n = 84 (7.6%)) or a STB (n = 876 (79.6%)). The basic statistics of these three groups are summarised in Table 1.

Fig. 1.

Study profile.

Table 1.

Basic statistics of all groups.

Variable	N	Mean	Std Dev	Median	Minimum	Maximum
Spontaneous preterm birth group (141)
Gestational age at enrolment (days)	141	118	37	118	49	203
Maternal age (years)	141	25.0	5.8	24.0	16.0	42.0
Maternal arm circumference	138	268	44	259	194	429
Body mass index (kg/m2)	139	24.4	5.9	22.7	15.3	43.8
Gravidity	141	2.3	1.4	2.0	1.0	7.0
Parity	141	1.1	1.2	1.0	0.0	6.0
Education (years)	141	9.8	1.7	10.0	4.0	13.0
People per room	123	1.6	1.1	1.3	0.5	8.0
Household income (ZAR per month)	83	934	663	750	167	3000
Gestational age at delivery (days)	141	229	30	241	149	258
Birthweight (grams)	138	1923	772	2105	360	3500
Birthweight Z-score	129	−0.3	1.0	−0.3	−6.3	1.8
Trimmed placenta weight (grams)	141	313	108	319	82	565
Iatrogenic preterm birth group (84)
Gestational age at enrolment (days)	84	121	39	125	53	217
Maternal age (years)	84	25.7	6.4	25.0	16.0	43.0
Maternal arm circumference	82	278	48	270	196	475
Body mass index (kg/m2)	81	25.7	6.1	24.7	15.4	45.7
Gravidity	84	2.1	1.3	2.0	1.0	6.0
Parity	84	0.9	1.2	0.0	0.0	5.0
Education (years)	83	10.0	1.8	10.0	6.0	13.0
People per room	72	1.5	0.8	1.3	0.3	5.0
Household income (ZAR per month)	56	992	638	929	100	3000
Gestational age at delivery (days)	84	217	33	222	146	258
Birthweight (grams)	81	1500	867	1440	190	4170
Birthweight Z-score	75	−1.0	1.4	−0.8	−5.7	3.1
Trimmed placenta weight (grams)	84	253	132	237	50	740
Spontaneous term birth group (876)
Gestational age at enrolment (days)	876	119	32	121	42	256
Maternal age (years)	876	24.4	5.8	24.0	16.0	41.0
Maternal arm circumference	856	273	44	265	190	458
Body mass index (kg/m2)	853	24.8	5.5	23.4	15.8	52.3
Gravidity	875	2.0	1.2	2.0	1.0	8.0
Parity	875	0.9	1.1	1.0	0.0	6.0
Education (years)	875	10.1	1.7	10.0	3.0	13.0
People per room	866	1.6	0.9	1.4	0.3	8.0
Household income (ZAR per month)	655	896	612	750	50	6000
Gestational age at delivery (days)	876	276	8	276	259	301
Birthweight (grams)	876	3099	439	3100	1160	4600
Birthweight Z-score	875	−0.4	1.0	−0.3	−4.3	2.8
Trimmed placenta weight (grams)	875	427	90	420	158	877

ZAR = South African Rand (1$ = 15.2 ZAR/1 EURO = 18.5 ZAR), Std Dev = Standard deviation.

Indications for IPTB were 31 (36.9%) for severe hypertension, pre-eclampsia, or eclampsia; 21 (25.0%) for an intrauterine demise (IUD); 13 (15.5%) for preterm premature/prelabour rupture of membranes (PPROM); 5 (6.0%) each for intrauterine growth restriction (IUGR) and congenital anomalies; 2 (2.4%) each for antepartum haemorrhage or placental abruption, placenta praevia, fetal distress, inevitable miscarriage; and 1 (1.2%) for nephrotic syndrome. The pathological causes of IUD were placental abruption 8 (38%), fetal flow restriction of the umbilical cord 4 (19%), uteroplacental malperfusion 3 (14%), undetermined 2 (10%), maternal systemic disease, fetal ascending infection, severe congenital anomalies, and severe maternal hypotension 1 (5%) each. We assessed different complications for IPTB in three GA categories. The three main indications for each group were: IUD (9, 39%), hypertension/pre-eclampsia (6, 26%) and congenital anomalies (5, 22%) for group GA 20w0 to 27w6 (23, 27%); hypertension/pre-eclampsia (13, 42%), IUD (11, 35%), and PPROM (3, 10%) for group GA 28w0 to 33w6 (31, 37%); and hypertension/pre-eclampsia (12, 40%), PPROM (9, 30%), and IUGR (3, 10%) for group GA 34w0 to 36w6 (30, 36%).

Decidual arteriopathy was identified in 100 (9.1%) placentas, thrombosis of cord vessels in 22 (2.0%), avascular villi in 237 (21.5%), chorionic plate/stem villus thrombosis in 110 (10.0%), villous-stromal karyorrhexis in 293 (26.6%), chorioamnionitis in 361 (32.8%), villitis in 249 (22.6%), funisitis in 56 (5.1%), and vasculitis of umbilical cord vessels in 75 (6.8%). (Table 2).

Table 2.

Placenta microscopic assessment in 1101 participants.

Features of maternal vascular malperfusion
Decidual arteriopathy	Muscular hypertrophy	Fibrinoid necrosis	Fibrinoid necrosis with atherosis
100 (9.1%)	57 (57.0%)	43 (43.0%)	0 (0.0%)
Features of fetal vascular malperfusion
Thrombosis of cord vessels	Artery	Vein	Both
22 (2.0%)	6 (27.3%)	15 (68.2%)	1 (4.5%)
Avascular villi	Regional	Diffuse
237 (21.5%)	229 (96.6%)	8 (3.4%)
Chorionic plate thrombosis	Artery	Vein	Both	Indeterminate
110 (10.0%)	3 (2.7%)	20 (18.2%)	29 (26.4%)	58 (52.7%)
Villous-stromal karyorrhexis	Regional	Diffuse
293 (26.6%)	268 (91.5%)	25 (8.5%)
Inflammatory response
Chorioamnionitis	Acute	Chronic	Eosinophilic
361 (32.8%)	333 (92.2%)	28 (7.8%)	0 (0.0%)
	Subchorion	Chorion	Chorioamnion	Indeterminate
	232 (64.3%)	68 (18.8%)	60 (16.6%)	1 (0.3%)
Villitis	Acute	Chronic specific	VUE
249 (22.6%)	2 (0.8%)	7 (2.8%)	240 (96.4%)
Funisitis	Perivascular	Superficial	Both
56 (5.1%)	54 (96.4%)	2 (3.6%)	0 (0.0%)
Vasculitis	One vessel	Two vessels	Three vessels
75 (6.8%)	46 (61.3%)	9 (12.0%)	20 (26.7%)
	Artery	One	Two
	33 (44.0%)	13 (39.4%)	20 (60.6%)

VUE villitis of unknown etiology (chronic non-specific).

The ORs for SPTB and IPTB as compared to STB are given in Table 3. For SPTB the following features of maternal vascular malperfusion (MVM) were associated with an increased risk: accelerated villous maturation (OR 17.94), macroscopic pathological infarction (OR 3.35), and microscopic pathological infarction (OR 1.85). Reduced risks for SPTB were associated with increased syncytial knots (OR 0.37), increased perivillous fibrin (OR 0.42), and increased extravillous trophoblast/fibrinoid islands (OR 0.55). For IPTB features associated with an increased risk were accelerated villous maturation (OR 69.9), macroscopic pathological infarction (OR 5.80), distal villous hypoplasia (OR 5.55), microscopic pathological infarction (OR 5.01) and decidual arteriopathy (OR 4.66).

Table 3.

Risks of placental features for spontaneous preterm and iatrogenic preterm birth compared to spontaneous term birth.

Pathology variable	OR SPTB vs STB	CI lower SPTB vs STB	CI upper SPTB vs STB	SPTB (141) n (% YES) of variable	STB (876) n (% YES) of variable	IPTB (84) n (% YES) of variable	OR IPTB vs STB	CI lower IPTB vs STB	CI upper IPTB vs STB
Features of maternal vascular malperfusion
Small trimmed placenta for GA	0.79	0.55	1.12	66 (47%)	456 (53%)	49 (59%)	1.29	0.81	2.03
Macroscopic pathological infarction	3.25	1.43	7.39	9 (6%)	18 (2%)	9 (11%)	5.80	2.52	13.36
Microscopic pathological infarction	1.85	1.09	3.14	20 (14%)	72 (8%)	26 (31%)	5.01	2.97	8.43
Decidual arteriopathy	1.68	0.94	3.00	16 (11%)	62 (7%)	22 (26%)	4.66	2.69	8.08
Accelerated villous maturation	17.94	9.09	35.42	30 (21%)	13 (1%)	43 (51%)	69.62	34.75	139.50
Increased syncytial knots	0.37	0.25	0.55	42 (30%)	466 (53%)	52 (62%)	1.43	0.90	2.26
Distal villous hypoplasia	0.98	0.52	1.85	12 (9%)	76 (9%)	29 (35%)	5.55	3.34	9.22
Increased extravillous trophoblast	0.55	0.35	0.86	26 (18%)	255 (29%)	28 (33%)	1.22	0.76	1.96
Increased perivillous fibrin	0.42	0.29	0.62	42 (30%)	439 (50%)	46 (55%)	1.21	0.77	1.89
Features of abruption
Macroscopic abruption	1.59	0.88	2.88	15 (11%)	61 (7%)	16 (19%)	3.14	1.72	5.75
Microscopic abruption	2.19	1.37	3.50	28 (20%)	89 (10%)	20 (24%)	2.76	1.60	4.78
Acute placental abruption (both)	2.08	1.00	4.34	10 (7%)	31 (4%)	11 (13%)	4.11	1.98	8.51
Features of fetal vascular malperfusion
Thrombosis of cord vessels (microscopic)	0.88	0.20	3.93	2 (1%)	14 (2%)	6 (7%)	4.85	1.81	12.98
Avascular villi	0.77	0.48	1.22	24 (17%)	185 (21%)	28 (33%)	1.87	1.15	3.02
Chorionic plate/stem villus thrombosis	0.75	0.39	1.44	11 (8%)	89 (10%)	10 (12%)	1.19	0.60	2.40
Villous-stromal karyorrhexis	0.47	0.29	0.76	22 (16%)	246 (28%)	25 (30%)	1.09	0.66	1.77
Fetal vascular thrombosis	0.57	0.31	1.04	13 (9%)	132 (15%)	24 (29%)	2.25	1.35	3.74
Vascular necrosis	991.00	992.00	993.00	0 (0%)	8 (1%)	1 (1%)	1.34	0.17	10.82
Inflammatory response
Chorioamnionitis (MIR)	1.40	0.97	2.02	56 (40%)	280 (32%)	25 (30%)	0.90	0.55	1.47
Villitis (MIR)	1.33	0.89	2.00	38 (27%)	190 (22%)	19 (23%)	1.06	0.62	1.80
Funisitis of cord (FIR)	2.24	1.19	4.23	14 (10%)	41 (5%)	1 (1%)	0.25	0.03	1.85
Vasculitis of cord vessels (FIR)	1.58	0.86	2.92	14 (10%)	57(7%)	4 (5%)	0.74	0.26	2.08
Outcome
Intrauterine growth restriction	0.45	0.24	0.83	12 (9%)	163 (19%)	25 (33%)	2.18	1.31	3.63
Low birthweight	30.16	19.32	47.10	99 (72%)	68 (8%)	69 (85%)	68.32	35.27	132.34
Stillbirths	20.51	9.68	43.48	27 (19%)	10 (1%)	26 (31%)	38.82	17.86	84.37
Infant demises before one year	9.48	3.55	25.33	10 (7%)	7 (1%)	4 (5%)	6.21	1.78	21.66

SPTB, spontaneous preterm birth; IPTB, iatrogenic preterm birth; STB, spontaneous term birth; OR, odds ratio; CI, confidence interval; Bold indicates significant OR; MIR, maternal inflammatory response; FIR, fetal inflammatory response; vs, versus.

For SPTB microscopic abruption was associated with an increased risk (OR 2.19), while for IPTB acute placental abruption (both macroscopic and microscopic) (OR 4.11), macroscopic abruption (OR 3.14), and microscopic abruption (OR 2.76) were associated with an increased risk.

Villous-stromal karyorrhexis was associated with a reduced risk for SPTB (OR 0.47), while the following features of FVM were associated with increased risk for IPTB: thrombosis of cord vessels (OR 4.85), fetal vascular thrombosis (OR 2.25), and avascular villi (OR 1.87).

For SPTB perivascular funisitis of the umbilical cord (OR 2.24) was associated with a greater risk. Inflammatory response was not associated with increased risks for IPTB.

In the SPTB group, IUGR (OR 0.45) occurred less frequently but low birthweight (OR 30.16), stillbirth (OR 20.15) and demise before the age of one year (OR 9.48) more frequently. In the IPTB group, IUGR (OR 2.18), low birthweight (OR 68.32), stillbirth (OR 38.82), and demise before the age of one year (OR 6.21) carried increased risks (Table 3).

When comparing the placental histopathologic features of SPTB directly to the features of IPTB, features of MVM and FVM were significantly greater in the IPTB group. Only perivascular funisitis of the umbilical cord (FIR stage 3) occurred significantly more often in the SPTB group (Table 4).

Table 4.

Placenta pathology and spontaneous preterm birth compared to iatrogenic preterm birth.

Pathology variable	SPTB (141) n (% YES) of variable	IPTB (84) n (% YES) of variable	p- value
Features of maternal vascular malperfusion
Small trimmed placenta for GA	66 (47%)	49 (59%)	0.229
Macroscopic pathological infarction	9 (6%)	9 (11%)	0.729
Microscopic pathological infarction	20 (14%)	26 (31%)	0.009
Decidual arteriopathy	16 (11%)	22 (26%)	0.014
Accelerated villous maturation	30 (21%)	43 (51%)	0.000
Increased syncytial knots	42 (30%)	52 (62%)	0.000
Distal villous hypoplasia	12 (9%)	29 (35%)	0.000
Increased extravillous trophoblast/fibrinoid islands	26 (18%)	28 (33%)	0.037
Increased perivillous fibrin	42 (30%)	46 (55%)	0.001
Features of abruption
Macroscopic abruption	15 (11%)	16 (19%)	0.243
Microscopic abruption	28 (20%)	20 (24%)	1.000
Acute placental abruption (both macro and micro)	10 (7%)	11 (13%)	0.422
Features of fetal vascular malperfusion
Thrombosis of cord vessels (microscopic)	2 (1%)	6 (7%)	0.075
Avascular villi	24 (17%)	28 (33%)	0.017
Chorionic plate/stem villus thrombosis	11 (8%)	10 (12%)	0.937
Villous-stromal karyorrhexis	22 (16%)	25 (30%)	0.038
Fetal vascular thrombosis (microscopic only)	13 (9%)	24 (29%)	0.001
Vascular necrosis	0 (0%)	1 (1%)	0.469
Inflammatory response
Chorioamnionitis (MIR)	56 (40%)	25 (30%)	0.390
Villitis (MIR)	38 (27%)	19 (23%)	1.000
Funisitis of cord (FIR)	14 (10%)	1 (1%)	0.015
Vasculitis of cord vessels (FIR)	14 (10%)	4 (5%)	0.500
Outcome
Intrauterine growth restriction	12 (9%)	25 (33%)	0.000
Low birthweight	99 (72%)	69 (85%)	0.059
Stillbirths	27 (19%)	26 (31%)	0.137
Infant demise before 1 year	10 (7%)	4 (5%)	1.000

SPTB, spontaneous preterm birth; IPTB, iatrogenic preterm birth; Bold indicates significant p-value; MIR, maternal inflammatory response; FIR, fetal inflammatory response.

Gestational hypertension was observed in 42% of IPTBs in comparison to 8% in SPTBs (p = 0.000) and 8% in STBs (p = 0.000). Pre-eclampsia was observed in 23% of IPTBs in comparison to 3% in SPTBs (p = 0.000) and 2% in STBs (p = 0.000). Fetal growth restriction occurred in 33% of IPTBs in comparison to 9% in SPTBs (p = 0.000) and 19% in STBs (p = 0.012). IUGR also occurred significantly less in SPTBs compared to STBs (p = 0.016).

Significant differences were found when preterm stillbirths (n = 53) were compared to SPTB where the infants were alive at one year (n = 102). Amongst stillbirths 98% (n = 48) vs 61% (n = 62) had low birthweight (p = 0.000), 35% (n = 13) vs 5% (n = 5) were growth restricted (p = 0.000), 21% (n = 11) vs 1% (n = 1) had acute placental abruption (p = 0.000), 28% (n = 15) vs 6% (n = 6) had fetal vascular thrombosis (p = 0.003), 30% (n = 16) vs 9% (n = 9) had decidual arteriopathy (p = 0.012), 30% (n = 16) vs 11% (n = 11) had microscopic pathological infarction (p = 0.048), 18% (n = 10) vs 4% (n = 4) had macroscopic pathological infarction (p = 0.042), 23% (n = 12) vs 5% (n = 5) had distal villous hypoplasia (p = 0.016), and 8% (n = 4) vs 0% (n = 0) had thrombosis of the umbilical cord vessels (p = 0.042).

When preterm stillbirths (n = 53) were compared to IPTB infants who were alive at one year (n = 42), significant differences were found. Amongst stillbirths, 26% (n = 14) vs 81% (n = 34) had increased syncytial knots (p = 0.000), 57% (n = 30) vs 17% (n = 7) had chorioamnionitis (p = 0.000), and 98% (n = 48) vs 76% (n = 32) had low birthweight (p = 0.012).

Low birthweight was observed in 100% (n = 14) of preterm born infants who had died in contrast to 61% (n = 62) of SPTB infants who were alive at one year (p = 0.006). This was the only difference between the two groups.

Infants who had died (n = 14), differed from IPTB infants who were alive at one year (n = 42) regarding increased syncytial knots 29% (n = 4) vs 81% (n = 34) (p = 0.006) and avascular villi, 0% (n = 0) vs 36% (n = 15) (p = 0.020). These were the only variables that differed significantly.

4. Discussion

Placental pathology has strong associations with PTB, both spontaneous and iatrogenic, and provides important insights into these subtypes. We found many significant differences between SPTB and IPTB which will be discussed separately.

4.1. SPTB

The pathologic conditions associated with disturbed maternal blood flow are currently known as MVM (see Fig. 2), as defined in the Amsterdam consensus [13], with villous maldevelopment being the most important histologic finding [14]. Accelerated villous maturation i. e. premature formation of terminal villi is diagnosed in placentas prior to mid third trimester and can be seen as an adaptation of the placenta to decreased perfusion or placental insufficiency [15,16]. In MVM, global or partial obstruction of blood flow from abnormal spiral arteries leads to accelerated villous maturation while segmental, complete occlusion causes a villous infarct [17]. The normal villous maturation pattern is characterized by decreasing villous size and diameter, increasing syncytial knots for gestational age accompanied by increased intervillous fibrin deposition alternating with areas of villous paucity (distal villous hypoplasia) in order to maximize oxygen diffusion exchange surface [14,16,18]. In our SPTB group we found more cases showing accelerated maturation of villi (21%) but combined with less proliferation of syncytial knots, perivillous fibrin, and extravillous trophoblast. There was no association with distal villous hypoplasia. These findings were significantly associated with SPTB, suggesting that placentas not adapting by the accelerated villous maturation pattern, are more likely to suffer from hypoxia, stillbirth, infant demises or SPTB.

Fig. 2.

Microscopic lesions found in preterm births. A: Accelerated villous maturation at 31 weeks gestation. Haematoxylin and Eosin stain, 10x. B: Increased syncytial knots at 33 weeks gestation (indicated by ‘→‘). Haematoxylin and Eosin stain, 10x. C: Villous stromal-vascular karyorrhexis involving the villi with endothelial cell karyorrhexis and red blood cells extravasation and fragmentation (as indicated by the ‘*‘). Haematoxylin and Eosin stain, 20x. D: Distal villous hypoplasia at 36 weeks gestation. Note the large empty intervillous spaces in which very small villi are lying. Haematoxylin and Eosin stain, 4x. E: Increased perivillous fibrin encasing the villi. The villous lining trophoblast, in some of the villi, has already degenerated (as highlighted by ‘→‘). Other villi still have retained their syncytiotrophoblastic lining. Haematoxylin and Eosin stain, 40x.

A previous study found that MVM and placental ischemia were more prevalent than infectious findings [19], while others have shown that the incidence of MVM in PTB is higher than inflammation or infection [14], and among all PTBs, evidence of MVM was identified more often than infection [7]. These studies included both SPTB and IPTB, but we highlight the importance of separating PTBs as our research demonstrates different associations with placental pathology suggesting different mechanisms for PTB.

MVM and infection/inflammation are the most common placental findings associated with PTB [20,21]. We found MVM to be central for IPTB and funisitis (FIR) significant for SPTB. However, other studies found that a significant proportion of SPTBs were associated with MVM and accelerated villous maturation [2,7,14,22-24].

Funisitis (10% of SPTBs vs 1% of IPTBs or vs 5% of STBs) was the only significant inflammatory response finding. Weiner et al. [24] reported a high rate of FIR in SPTB, supporting our finding. Although they and Kovo et al. [25] also documented a high rate of MIR in SPTB, we found no significant association. Kim et al., found preterm infants to be more vulnerable and respond more vigorously to intrauterine infection than term fetuses constituting a trigger for preterm labour [26]. As funisitis was the only variable that was significantly more prevalent in SPTB than in IPTB, this might be linked to spontaneous initiation of preterm labour.

Our results did not support chorioamnionitis as a predominant cause of PTB [7,26-29]. Others, however, submit that chorioamnionitis is a consequence rather than a cause of PTB, as 10–15% of placentas at term have histological acute chorioamnionitis [6]. Given that 29% of our placentas at term had acute chorioamnionitis, we surmise that chorioamnionitis may not be a cause of PTB. In the literature, placentas from SPTB more commonly show acute chorioamnionitis with funisitis, vasculitis and chronic nonspecific villitis [6,29]. Van Vliet et al. and Goldstein et al. found chronic villitis associated with PTB [20,29], but we found no association with SPTB or IPTB.

Acute chorioamnionitis is more common than chronic chorioamnionitis. We found 34.2% and 29.2% of acute chorioamnionitis in our preterm and term birth groups respectively. Kim et al. reports that acute chorioamnionitis is present in 3–5% of placentas [30], while Roberts et al. reports a presence in 25% of term placentas [31], making our incidence higher. However, 64,3% of our chorioamnionitis overall was subchorionic, and only 18,8% chorionic and 16,6% chorioamniotic, respectively. The subchorionitis may be clinically insignificant and would now be diagnosed as maternal stage 0 (subchorionitis only) [31, 32].

Kim et al. reported chronic chorioamnionitis in preterm and term birth groups as 34–39% and 8–19%, respectively [33]. This is significantly higher than the 1.8% and 2.7% chronic chorioamnionitis found in our preterm and term groups. Infectious villitis, acute or chronic, is rare. Our prevalence of 0.2% and 0.6% respectively corresponds with the literature [29,34-37]. The overall frequency of VUE, has varied from 1% to 34%, depending on placental sampling, diagnostic criteria and the population group examined [34,35,38-45]. Our overall frequency was 22%, being higher (29%) in growth restricted cases.

Funisitis was found in 5.1% of our whole cohort and 6.7% of our PTB group, corresponding well with Jessop et al.‘s frequency of 5.7% [46]. The prevalence of 4.7% in our STB group was less than 6.7% reported by Lee et al. in term births [47], while higher than Liu et al. who reported an overall prevalence of 3.25%, 5.8% in the preterm and 2.8% in term group [48], but much less than the incidence of 44.4% in a preterm labour group reported by Pacora et al. [49].

Inflammation of umbilical vessels, vasculitis, can be seen in the umbilical vein and/or one or both umbilical arteries. Kim et al. found vasculitis in 17% of PTBs; 45% of vein and 55% of arteries [50]. We found vasculitis in only 8% of our PTB group. However, our findings of 44% with inflammation of vein and 56% of arteries, were similar.

4.2. IPTB

In IPTB we found features of MVM more commonly than in SPTB including microscopic pathological infarction (31%), decidual arteriopathy (26%), accelerated villous maturation (51%), increased syncytial knots (62%), distal villous hypoplasia (35%), increased extravillous trophoblast (33%) and increased perivillous fibrin (55%) (Table 4). Furthermore, we found high rates of obstetric complications such as gestational hypertension (42%), pre-eclampsia (23%) and fetal growth restriction (33%). This is not unexpected as other investigators have also found distal villous hypoplasia, increased syncytial knots and accelerated villous maturation were associated with MVM, fetal growth restriction, and pre-eclampsia [16,51]. MVM, is frequently cited as causing fetal growth restriction [28,52,53]. According to Visser et al., IUGR is associated with MVM and SPTB [2], but we found no evidence of this. Researchers have found pre-eclampsia [54], gestational hypertension, IUGR [55], abruption [2], and chronic villitis [23], to be commonly associated with placental insufficiency. Pregnancies complicated by MVM in our IPTB group, had a 4.5 times higher risk of developing pre-eclampsia or delivering a growth restricted baby according to Wright et al. [56].

Pre-eclampsia was observed in 23% (19) of IPTB women with only four having decidual arteriopathy (21%). The literature finds a higher prevalence of decidual arteriopathy at from 34 to 53% of women with pre-eclampsia [57,58], much more than our 21%. Possibly our pre-eclampsia was less severe, warranting other indications (see results) for induction.

In IPTB compared to STB we found no significant difference for chorioamnionitis, chronic or acute villitis, vasculitis of cord vessels or funisitis. Although Kovo et al. [25] discloses a significantly higher rate of inflammatory lesions for PPROM, we found no association in our IPTB group where nearly 16% were induced for PPROM.

FVM is a recently introduced term [13] to characterize a group of lesions previously described under the headings of fetal vascular obstructive lesions or thrombi, fetal thrombotic vasculopathy, and extensive avascular villi. FVM can be separated into partial or complete obstruction of fetoplacental blood flow [17]. Van Vliet et al. found fetal vascular thrombosis to be associated with PTB [20]. Our study revealed that fetal vascular thrombosis was associated with IPTB only, present in 29% vs 9% (SPTB) and 15% (STB). Those with fetal vascular thrombosis are more likely to have IUGR or oligohydramnios and necessitate delivery. This underscores the importance of differentiating between SPTB and IPTB rather than using the umbrella term: PTB. Our results only confirm a significant association between FVM and IPTB, but not with villitis as reported by Faye-Petersen [6].

Thrombosis of umbilical cord vessels is rare, reported at <0.1% of perinatal autopsies or high risk gestations [59,60]. We found umbilical cord vascular thrombosis present in 7.3% of our IPTB cases. Venous thrombosis occurred in 70%, arterial thrombosis (one or both umbilical arteries) in 10%, and both venous and arterial thrombosis in 20% of cases, according to literature reports [18,60]. Our venous thrombosis in 68% of overall cases, is consistent with the literature. Poor fetal outcome is more likely with arterial thrombosis [60]. We identified 27% of cases with arterial thrombosis in our cohort, (compared to 10% reported), which highlights the severity, especially in the IPTB group (33%). Furthermore Klaritsch et al. postulates that arterial thrombosis may cause severe IUGR [59], supporting the 33% growth restriction found in the IPTB group. Thrombosis of large fetal vessels in the placenta leads to regions of downstream avascular villi and have been associated with adverse outcomes and increased rates of IUGR [61]. Fox et al. who studied avascular villi, reported an overall incidence of 4.5% and a higher frequency of 14% in stillbirths [62], which is lower than the avascular villi incidence of 33% found in our IPTB group. Villous-stromal karyorrhexis (30% of our IPTB placentas), is deemed as a precursor of avascular villi [18]. Chorionic plate/stem villus thrombosis, found in 12% of our IPTB group, was mostly indeterminate. These features of FVM were all significantly increased and mostly associated with our IPTB group.

4.3. Demises

Preterm stillbirths, in comparison to SPTB infants alive at one year, had lower birthweights and more were more often growth restricted. They had more findings of placental abruption, fetal vascular thrombosis, and features of MVM (pathological infarction, decidual arteriopathy and/or distal villous hypoplasia). However, preterm stillbirths, compared to IPTB infants alive at one year, had significantly less proliferation of syncytial knots, increased chorioamnionitis and lower birthweights. Other researchers found infants delivered to mothers with acute chorioamnionitis were more preterm [5,63], which could explain our higher prevalence of low birthweight in stillbirths. Chorioamnionitis is frequently associated with perinatal death [26], which is consistent with our findings.

Preterm infants dying within one year of life, all had low birthweight compared to SPTB infants alive at one year, and, presented with less increased syncytial knots and no (0%) avascular villi in comparison to IPTB infants alive at one year.

The finding of less increased syncytial knots in preterm stillbirths, preterm infant demises and SPTBs differed significantly from increased syncytial knots found in IPTBs and STBs and is contrary to literature reports of extensive syncytial knotting with villous maturation. During pregnancy, syncytial knots are consistently present, increasing with gestational age, and utilised to evaluate villous maturity. Loukeris et al. found increased syncytial knots to be associated with conditions of MVM [12], supporting our findings regarding the IPTB group (significant MVM and increased syncytial knots) and the SPTB group (lack of both MVM and increased syncytial knots). Morgan et al. concluded that accelerated villous maturation with extensive syncytial knotting is a histological marker for placental insufficiency, and a common finding in idiopathic PTB [64]. Again, this was only true for our IPTB and not the SPTB group. According to Fogarty et al., the formation of increased syncytial knots with richly capillarised and highly branched terminal villi, in placentas from complicated pregnancies, can be attributed to premature aging of the placenta, providing a mechanism to enhance the placenta’s ability to transfer oxygen to the fetus [65]. But in cases where syncytial knots are not increased despite the accelerated villous maturation we propose that a “blocked” accelerated maturation pattern might be a trigger for SPTB and or result in a stillbirth or an infant demise. The lack of increased knotting in preterm stillbirths, in preterm born infants demised within one year of life, and in SPTB combined with a predominance of terminal villi, therefore demonstrating only partial accelerated maturation or abnormal/“blocked” maturation, needs to be investigated further.

5. Conclusion

SPTB and IPTB are pathologically different entities and should be investigated separately.

SPTB was not associated with features of maternal or fetal vascular malperfusion. Only funisitis occurred more frequently in this group.

We found significantly less increased syncytial knotting combined with a predominance of terminal villi in preterm demises and in SPTB, raising the question of whether abnormal villous maturation reflects a process that could trigger SPTB or result in a preterm demise?

IPTB was not associated with inflammation of the umbilical cord, membranes and chorionic plate, or parenchyma.

Abbreviations:

ANOVA

analysis of variance

CI

confidence interval

FIR

fetal inflammatory response

FVM

fetal vascular malperfusion

GA

gestational age

IPTB

iatrogenic preterm birth

IUD

intrauterine demise

IUGR

intrauterine growth restriction

MIR

maternal inflammatory response

MOU

midwife obstetric unit

MVM

maternal vascular malperfusion

OR

odds ratio

PPROM

preterm premature/prelabour rupture of membranes

PTB

preterm birth

SA

South Africa

SPS

Safe Passage Study

SPTB

spontaneous preterm birth

STB

spontaneous term birth

TBH

Tygerberg Academic Hospital

VUE

villitis of unknown etiology

ZAR

South African Rand