Macroscopic lesions of maternal and fetal vascular malperfusion in stillborn placentas: diagnosis in the absence of microscopic histopathological examination

Abstract

Introduction

Lesions of maternal vascular malperfusion (MVM) and fetal vascular malperfusion (FVM) are common in placentas associated with both stillbirth and live birth. The objective of this study was to identify lesions present more commonly in stillborn placentas and those most indicative of MVM and FVM without microscopic pathologic evaluation.

Methods

Data were derived from the Stillbirth Collaborative Research Network. Lesions were identified according to standard protocols published previously and categorized as either MVM or FVM according to the Amsterdam Placental Workshop Group Consensus Statement and macroscopic “umbilical cord at risk” findings. Multivariate logistic regression was used to determine the odds of stillbirth with macroscopic findings of MVM or FVM.

Results

595 stillbirths and 1,305 live births were analyzed. FVM lesions (85.2%) were marginally more common (though not statistically different) in stillbirths compared to MVM lesions (81.3%). Macroscopic findings of both MVM and FVM were more common in stillbirths versus livebirths (p <0.001). Odds ratios of macroscopic MVM and FVM lesions for stillbirth, adjusted for gestational age at delivery, maternal race (minority), ethnicity (Hispanic), age, and history of hypertension or diabetes, were 1.48 (95% CI 1.30 - 1.69) and 1.34 (95% CI 1.18 - 1.53), respectively.

Discussion

Macroscopic features of MVM and FVM are associated with higher odds of stillbirth versus live birth even when controlled for gestational age and maternal factors, which may be a useful clue in determining the pathophysiology of these events. This information is also useful for pathologists when microscopic examination is not available.

Introduction

The relationship between maternal and fetal placental lesions and adverse perinatal outcomes has been thoroughly studied. However, results are mixed regarding the association between such lesions and perinatal mortality [1-4]. While placental pathologies represent the largest category of cause of intrauterine death, maternal vascular malperfusion (MVM) and fetal vascular malperfusion (FVM) are variably linked to perinatal mortality, with lesions present at similar rates in fetal and neonatal deaths [3,5-7].

A recent prospective observational study by Kulkarni et al. (2021) sought to characterize these lesions in stillbirths, neonatal deaths, and livebirths [1]. These authors reported lesions associated with FVM at a rate of only 19.0% compared to MVM at 58.4% in placentas associated with fetal deaths. However, the rate of FVM among placentas associated with fetal death (19.0%) was higher than that of placentas associated with livebirth (8.3%). This study further found that MVM, but not FVM, was associated with maternal hypertension, small for gestational age fetuses, and antepartum hemorrhage.

This study did not assess the relative presence of macroscopic and microscopic lesions of MVM or FVM among placentas, an issue of import for hospitals without dedicated perinatal pathologists given the limited number of trained placental pathologists in the field [8]. The presence or absence of such lesions may have implications in assessing factors that contributed to an adverse outcome [9,10]. Therefore, it is important to be able to identify MVM or FVM lesions regardless of specialized perinatal pathology training and particularly without the use of placental microscopy.

Therefore, we sought to first compare the presence of FVM or MVM lesions among stillbirths and livebirths by presenting both macroscopic and microscopic findings associated with these pathologic phenomena. We then sought to understand how macroscopic placental features of vascular malperfusion may be able to classify placentas as related to stillbirth or livebirth. This builds upon previous work by both identifying which findings are present within stillbirths versus livebirths, and also which may be observed in the absence of microscopic pathologic evaluation.

Materials and Methods

Data were collected as part of the Stillbirth Collaborative Research Network (SCRN), a study of stillbirths and livebirths completed at 59 hospitals in five geographic regions from 2006 to 2009. All study procedures were approved by each center’s Institutional Review Board and by the Data Coordinating and Analysis Center. Patients gave written informed consent to participate. Inclusion and exclusion criteria for SCRN, as well as data collection procedures, have been described previously; all singleton stillbirths or livebirths of gestational age greater than 20 weeks who had undergone placental examination were included [11]. Participants completed an interview at which demographics, maternal health history, and current pregnancy history were collected. For stillbirths, cause of death was determined utilizing the INCODE system, as described previously [12]. For additional analyses, livebirths were stratified as preterm (<37 weeks) or term (≥37 weeks).

Placentas included in these analyses were associated with a livebirth or stillbirth of a singleton fetus (i.e. multiples were excluded) and were examined utilizing the SCRN placental examination protocol, as described previously [13]. Stillborn fetuses were examined according to the SCRN postmortem examination protocol [14]. All infants were weighed and examined for nuchal cord or cord entanglement. Placental weight was classified as small for gestational age (SGA) if below 10^th percentile for gestational age, according to established standards [15]. Lesions were classified as MVM or FVM according to previously published standards, including the Amsterdam Placental Workshop Group Consensus Statement [16-18]. Lesions were identified as macroscopic (i.e., observable on gross examination) or microscopic. We also included macroscopic “umbilical cord at risk” findings (according to Redline et al., 2018) not explicitly listed in the Amsterdam criteria for FVM, including marginal, velamentous, and furcate insertion, single umbilical artery, increased cord coiling, umbilical cord knots, and nuchal cord entanglement [17].

Analyses were conducted using R Studio software (R version 4.1.3). Demographic, placental, and lesion characteristics were compared between livebirths and stillbirths utilizing Mann-Whitney tests (continuous variables) and Chi-square or Fisher Exact tests (categorical variables). P-values for significance were set utilizing a Bonferroni correction for multiple comparisons for each analysis.

Multivariate logistic regression was conducted with the purpose of determining if macroscopic MVM or FVM placental findings were associated with stillbirths or livebirths. Macroscopic MVM or FVM findings were entered into regression alone, then entered controlling for several maternal factors (history of hypertension or diabetes, minority race, Hispanic ethnicity, and maternal age). Accuracy, sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) were collected for the model; unadjusted odds ratio (OR) or adjusted odds ratio (aOR) with 95% confidence intervals (95% CI) were collected for each covariate.

Results

Sample characteristics

Five-hundred and ninety-five (595) stillbirths and 1,305 livebirths (includes preterm and term, unless otherwise indicated) were analyzed (Table 1, Supplemental Figure 1). There were no significant differences in maternal age, ethnicity, or race between the two groups; stillbirth was associated with fewer years of education, previous stillbirth, and higher body mass index. Stillbirths were associated with significantly younger gestational age at birth than livebirths (28.66 ± 6.59 weeks versus 36.38 ± 5.46 weeks, p<0.001) and significantly lower fetal and placental disc weight. Among livebirths, preterm livebirths were less likely to have undergone induction of labor (15.5% versus 31.9%, p <0.001). Overall dimensions of the placental disc and umbilical cord were smaller in specimens associated with stillbirth versus livebirths (p<0.001); there was an increased rate of nuchal cord in livebirths versus stillbirths but more nuchal loops in stillbirths versus livebirths (not significant when adjusted for multiple comparisons).

Table 1:

Sample characteristics

		Livebirth (n = 1305)			Stillbirth (n = 582)	p- Value1
Demographics	Maternal age (years)	27.8 +/− 6.1			27.7 +/− 6.7	0.5
	Paternal age (years)	29.8 +/− 7.2			29.4 +/− 7.4	0.3
	Ethnicity (Hispanic)	470 (36.0)			215 (36.9)	0.7
	Race (minority)	934 (71.6)			216 (37.1)	<0.001
	Black	795 (61.3)			147 (25.4)	<0.001
	American Indian	5 (0.4)			2 (0.3)	0.9
	Asian	46 (3.5)			35 (6.1)	0.013
	Pacific Islander	12 (0.9)			4 (0.7)	0.8
	Other	106 (8.2)			59 (10.2)	0.2
	Education (years)	132 +/− 3.1			12.8 +/− 3.0	0.003
	Previous pregnancy	1305 (100.0)			561 (96.4)	<0.001
	Previous stillbirth	0 (0.0)			42 (7.2)	<0.001
Maternal health history	Body mass index	26.4 +/− 7.0			27.5 +/− 7.2	0.001
	Any chronic condition	449 (34.4)			196 (33.7)	0.8
	Hypertension	192 (15.5)			81 (14.9)	0.7
	Diabetes	41 (3.3)			32 (5.9)	0.012
	Sexually transmitted infection	170 (13.0)			89 (15.3)	0.2
Current pregnancy	Gestational age at birth	36.4 +/− 5.5			28.7 +/− 6.6	<0.001
	Alcohol use	502 (40.6)			217 (40.0)	0.8
	Tobacco use	181 (14.6)			107 (19.7)	0.007
	Clinical abruption	51 (7.6)			70 (18.4)	<0.001
	Fetal sex (female)	647 (49.7)			233 (46.8)	0.3
	Birth weight (g)	2909.6 +/− 1046.1			1364.8 +/− 1157.6	<0.001
Delivery	Vaginal	839 (64.3)			492 (84.5)	<0.001
	Induction of labor	355 (28.3)			352 (63.0)	<0.001
		PTB	43 (15.5)	<0.001 2
		TB	312 (31.9)
	Preterm labor	184 (27.0)			102 (25.8)	0.7
		PTB	180 (68.4)	<0.001 2
		TB	4 (1.0)
	Pre-labor rupture of membranes	117 (17.1)			80 (20.1)	0.2
		PTB	104 (39.7)	<0.001 2
		TB	13 (3.1)
Placental disc examination	Disc weight (g)	403.7 +/− 135.3			240.4 +/− 152.0	<0.001
	Intact placental disc	1171 (92.2)			509 (88.7)	0.014
	Round disc	584 (45.7)			246 (43.9)	0.5
	Dimensions (cm)
	Longest diameter	19.1 +/− 3.5			15.1 +/− 3.9	<0.001
	Shortest diameter	15.8 +/− 2.9			12.1 +/− 3.3	<0.001
	Thickest portion	2.3 +/− 0.7			2.1 +/− 0.8	<0.001
	Thinnest portion	1.3 +/− 06			1.2 +/− 0.5	<0.001
	Irregular lobular pattern	71 (5.6)			46 (8.4)	0.025
Umbilical cord & membrane examination	Segments received	12 +/− 07			1.2 +/− 0.6	0.069
	Dimensions (cm)
	Length	36.0 +/− 16.7			33.3 +/− 17.1	<0.001
	Thinnest diameter	1.0 +/− 0.4			0.9 +/− 0.4	<0.001
	Thickest diameter	1.5 +/− 0.4			1.4 +/− 0.6	<0.001
	Mass present	4 (0.3)			5 (0.9)	0.14
	Decreased coiling (≤ 3 in 10cm)	910 (73.8)			358 (66.8)	0.003
	Increased coiling (>3 in 10 cm)	323 (26.2)			178 (33.2)	0.003
	Nuchal cord	287 (90.8)			83 (66.4)	<0.001
	Number of loops	1.2 +/− 0.5			1.4 +/− 0.7	0.004
	% Circumvallate membrane	1.1 +/− 9.1			1.8 +/− 11.3	0.12
	% Circummarginate membrane	5.7 +/− 17.6			11.4 +/− 25.8	<0.001

All values are represented as mean ± standard deviation or number (percent).

¹Mann-Whitney test, Chi-Square test, or Fisher Exact test.

²Chi-Square or Fisher Exact test comparing preterm births (PTB) versus term births (TB) among livebirths.

P-values considered significant for <0.0013 with Bonferroni correction for multiple comparisons.

Overall, FVM lesions (85.2%) were marginally more common in stillbirths compared to MVM lesions (81.3%), though this finding was not statistically significant (Tables 2-3). 402 (69.1%) stillborn and 543 (41.6%) liveborn placentas displayed features of both MVM and FVM, whereas 231 (39.7%) stillborn and 279 (21.4%) liveborn placentas displayed macroscopic features of both phenomena. Prevalence of such lesions were also significantly higher in stillbirths versus preterm livebirths (Figure 1). Furthermore, macroscopic and microscopic findings of MVM and FVM were more common among stillbirths compared to term and preterm livebirths; in post-hoc comparisons, all were significantly more common in stillbirths versus preterm livebirths except for microscopic MVM lesions.

Table 2:

Lesions associated with Fetal Vascular Malperfusion (FVM)

	Livebirth	Stillbirth	p-Value1
Any FVM lesion	774 (59.3)	496 (85.2)	<0.001
Macroscopic features – umbilical cord	630 (48.3)	357 (61.3)	<0.001
Marginal insertion	158 (12.1)	117 (20.1)	<0.001
Velamentous insertion	19 (1.5)	25 (4.5)	<0.001
Length of unprotected cord (cm)	0.13 +/− 1.09	0.47 +/− 2.13	<0.001
Furcate insertion	36 (2.9)	9 (1.6)	0.12
Length of unprotected cord (cm)	0.08 +/− 0.46	0.15 +/− 1.28	0.8
Knot
True knot	13 (1.0)	6 (1.0)	0.9
False knot	184 (17.3)	68 (23.3)	0.020
Single umbilical artery	28 (2.2)	41 (7.2)	<0.001
Increased coiling (>3 coils/10 cm)	323 (26.2)	178 (33.2)	0.003
Nuchal cord entanglement	38 (20.0)	35 (50.7)	<0.001
Microscopic features	291 (22.3)	360 (61.9)	<0.001
Distal villous immaturity	84 (6.5)	99 (17.1)	<0.001
Diffuse distribution	43 (51.2)	55 (55.6)	<0.001
Focal distribution	41 (48.8)	44 (44.4)
Avascular villi	91 (7.1)	118 (20.3)	<0.001
Diffuse distribution	1 (1.1)	24 (20.3)	<0.001
Focal distribution	90 (98.9)	94 (79.7)
Stromal-vascular karyorrhexis	21 (1.6)	180 (31.0)	<0.001
Diffuse distribution	1 (4.8)	29 (16.1)	0.14
Focal distribution	20 (95.2)	151 (83.9)
Fetal vascular thrombi present	131 (10.0)	155 (26.6)	<0.001
Thrombi with complete occlusion	15 (1.1)	38 (6.5)	<0.001
Thrombi with degenerative changes	23 (1.8)	7 (1.3)	0.4

All values are represented as mean ± standard deviation or number (percent).

¹Mann-Whitney test, Chi-Square test, or Fisher Exact test. P-values considered significant for < 0.002 with Bonferroni correction for multiple comparisons.

Table 3:

Lesions of the placental disc associated with Maternal Vascular Malperfusion (MVM)

	Livebirth	Stillbirth	p-Value1
Any MVM lesion	856 (65.6)	473 (81.3)	<0.001
Macroscopic features	570 (43.7)	370 (63.6)	<0.001
Placenta small for gestational age	511 (39.8)	313 (54.8)	<0.001
Retroplacental hematoma	78 (6.1)	126 (22.7)	<0.001
Microscopic features	560 (42.9)	324 (55.7)	<0.001
Accelerated villous maturity	82 (6.3)	101 (18.3)	<0.001
Distal villous hypoplasia	117 (9.1)	68 (11.8)	0.071
Diffuse distribution	26 (22.2)	15 (22.1)	0.9
Focal distribution	91 (77.8)	53 (77.9)
Increased syncytial knots	325 (25.2)	126 (21.8)	0.11
Diffuse distribution	0 (0.0)	0 (0.0)	1.00
Focal distribution	325 (100.0)	126 (100.0)
Decidual arteriopathy	48 (3.7)	74 (12.8)	<0.001
Parenchymal infarction	213 (16.5)	164 (28.4)	<0.001
Diffuse distribution	1 (0.5)	16 (9.8)	<0.001
Focal distribution	212 (99.5)	148 (90.2)

All values are represented as mean ± standard deviation or number (percent).

¹Mann-Whitney test, Chi-Square test, or Fisher Exact test. P-values considered significant for <0.004 with Bonferroni correction for multiple comparisons.

Figure 1: Maternal and fetal vascular malperfusion, microscopic and macroscopic lesions, according to stillbirth versus preterm livebirth versus term livebirth status.

Percent of each lesion among term livebirths, preterm livebirths, and stillbirths. P-values are shown for among group analyses and between group comparisons of preterm livebirths versus stillbirths (both Chi-Square); bolded p-values indicate statistical significance (p < 0.05).

Fetal vascular malperfusion

Significantly more stillbirths had any FVM lesion compared to livebirths (85.2% versus 65.6%, p <0.001; Table 2). Both macroscopic and microscopic findings of FVM were more likely to be identified in stillbirths versus livebirths (p <0.001).

Among macroscopic lesions, placentas associated with stillbirth were more likely to have marginal or velamentous umbilical cord insertion, single umbilical artery, and nuchal cord entanglement than those associated with livebirth. Increased cord coiling was observed at a higher frequency in stillbirths versus livebirths but was not statistically significant when controlled for multiple comparisons.

Among microscopic lesions, placentas associated with stillbirth were more likely to show distal villous immaturity, avascular villi, stromal-vascular karyorrhexis, and thrombi of the fetal vessels compared to those associated with livebirths. Diffuse lesions were more common than focal lesions in stillbirths versus livebirths in the categories of distal villous immaturity and avascular villi. Fetal vascular thrombi with complete occlusion were also more common among placentas associated with stillbirths versus livebirths.

Maternal vascular malperfusion

Significantly more stillbirths had any MVM lesion compared to livebirths (81.3% versus 65.6%, p<0.001; Table 3). Both macroscopic and microscopic findings of MVM were more likely to be identified in stillbirths versus livebirths (p<0.001).

Among macroscopic features, placentas associated with stillbirth were more likely to be SGA and have retroplacental hematoma compared to livebirths. Among microscopic features, placentas associated with stillbirth were more likely to have accelerated villous maturity, decidual arteriopathy, and parenchymal infarction. Only diffuse parenchymal infarction (versus focal) was more common in stillbirths versus livebirths. There were no differences in the prevalence of distal villous hypoplasia or increased syncytial knots (Tenney-Parker changes) between stillbirths and livebirths.

Classifying stillbirth based on macroscopic lesions

Logistic regression was able to classify data points as stillbirth versus livebirth based on the presence of macroscopic MVM and FVM features and several maternal factors with an accuracy of 74.4% (95% CI 70.1 – 78.4%), sensitivity of 88.6%, specificity of 42.2%, PPV of 77.7%, and NPV of 62.0%. Unadjusted OR (with 95% CI) for MVM and FVM were 1.48 (1.31 – 1.66) and 1.32 (1.17 – 1.48), respectively (p<0.001); when adjusted for race, ethnicity, maternal age, and history of hypertension or diabetes, these ORs did not significantly change (aOR for MVM = 1.48 [1.30 – 1.69], FVM = 1.34 [1.18 – 1.53], p<0.001).

Discussion

In this study we investigated the association between placental vascular lesions and stillbirth; both MVM and FVM were significantly associated with stillbirth, with a lesion rate of 85.2% for FVM and 81.3% for MVM. This rate is significantly higher than reported in previous literature and emphasizes the role of vascular malperfusion in the etiology of stillbirth. Furthermore, we were able to identify the odds ratios for stillbirth versus livebirth based on the presence of macroscopic MVM and FVM features, which may be useful for pathologists when microscopic examination is not available.

MVM and FVM lesions were both incredibly common in stillbirths, though certain macroscopic features were present more often across the sample. Over half (50.7%) of umbilical cords associated with stillbirth had nuchal entanglement; indeed, umbilical cord accidents account for cause of death in 10% of stillbirths and are thought to be related to fetal hyperactivity and uterine stimulation [19]. However, the presence of nuchal cord entanglement alone is not necessarily associated with increased risk of stillbirth, implying that other factors may be contributing to these antenatal deaths that should be investigated thoroughly when nuchal entanglement is present [20]. A large portion (33.2%) of these stillbirths also were associated with increased cord coiling, which, similarly, does not alone predict stillbirth; rather, umbilical coiling index > 90^th percentile for gestational age is associated with a multitude of adverse outcomes in addition to fetal death, including preterm birth, intrapartum fetal distress with need for interventional delivery, meconium-stained amniotic fluid, fetal anomalies, and acidosis [21,22].

Among macroscopic MVM lesions, SGA placenta was found commonly in stillbirths (54.8%); many factors appear to contribute to SGA placentas, including low levels of placental growth factor, epigenetic changes affecting growth, and clinical complications such as hypertension, therefore making it difficult to pinpoint a specific cause of stillbirth in these cases [23-25]. As stillbirths and preterm births due to hypertensive disorders often share placental findings, it is significant that macroscopic MVM findings were more common in stillbirths versus preterm livebirths, though microscopic findings were similar in prevalence. This may point to the high rates of obstetric complications among our preterm livebirth sample. Retroplacental hematoma was less common but still prevalent among stillbirths (22.7%); of note, clinically-identified abruption was not necessarily associated with macroscopic placental hematoma. Failure to promptly diagnose placental abruption is often cited in malpractice cases, making this distinction particularly important in medicolegal settings [26,27]. Macroscopic examination of the placenta at the time of delivery, by midwives or obstetricians, may reveal true attached retroplacental blood clots and suggest chronic MVM, which also has significant legal (and clinical) implications [28,29]. These findings can help adjudicate these cases, as well as provide an explanation for a patient’s or plaintiff’s distress.

There exists overlap between antemortem and postmortem findings of FVM, which may be delineated by clinical findings (e.g., clinical identification of fetal demise), the extent of maceration, and/or histopathological changes (autolysis within the fetus and/or temporal heterogeneity of findings within the placenta) [30]. On a macroscopic level, this overlap is less significant, given that no macroscopic lesions are shared by antepartum and postmortem FVM findings. Indeed, “umbilical cord at risk” findings are more specific for antemortem FVM than postmortem changes [31]. However, it is important to note for microscopic placental examination that uniform avascular villi or stromal-vascular karyorrhexis may be postmortem, requiring pathologists to distinguish these findings accordingly.

There are several strengths to the SCRN study that add to the validity of our findings. Firstly, all placental examinations were standardized according to SCRN protocols, which mitigates variability among different pathologists. Furthermore, the SCRN data contain a diverse group of participants (approximately 36% Hispanic and 36% minority race), which captures a wide breadth of individuals experiencing stillbirth. Our study should also be interpreted in the context of several limitations. We included an additional set of findings not explicitly stated within the Amsterdam criteria but consistent with other sources in the literature, including those lesions that suggest an “umbilical cord at risk.” This may contribute to the increased rate of FVM lesions observed within the sample but did not impact the rate of MVM lesions we reported [17]. Furthermore, the SCRN data does not include findings that, while traditionally identified microscopically, may be observed on careful macroscopic examination, including parenchymal infarcts, chorionic plate or umbilical thromboses, or large areas of avascular villi. Our data were collected from 2006 to 2009, which may underestimate the impact of evolving environmental, medical, and social factors on etiologies of stillbirth. Though we assessed differences between stillbirths and livebirths in this study, we did not control for factors such as neonatal morbidity or mortality; future analyses of this cohort of placentas may seek to determine differences in placental pathologies among all adverse pregnancy outcomes. Finally, though macroscopic lesions of MVM and FVM were associated with higher odds of stillbirth versus livebirth, NPV and specificity were low (62.0% and 42.2%, respectively), likely due to the high prevalence of these lesions within the livebirth sample; the predictive capacity of these lesions therefore needs to be assessed in an external sample.

Maternal vascular malperfusion and fetal vascular malperfusion are common abnormalities that may be identified through macroscopic placental examination. Further studies will continue to elucidate the macroscopic correlates of these findings to further assist general pathologists, obstetricians, and midwives in identifying at-risk placental features such as these that may have implications for further childbearing.

Highlights.

• The Stillbirth Collaborative Research Network data contains comprehensive placental examination data for stillbirths.

• Maternal and fetal vascular malperfusion (MVM, FVM) are associated with stillbirth, with macroscopic lesions more common in stillbirths.

• These lesions can identify placentas associated with stillbirths versus livebirths.

Abbreviations:

GA

gestational age

MVM

maternal vascular malperfusion

FVM

fetal vascular malperfusion

SCRN

Stillbirth Collaborative Research Network

SGA

small for gestational age