Chronic Inflammation of the Placenta: Definition, Classification, Pathogenesis, and Clinical Significance

Abstract

Chronic inflammatory lesions of the placenta are characterized by the infiltration of the organ by lymphocytes, plasma cells and/or macrophages, and may result from infections (viral, bacterial, parasitic) or be of immune origin (maternal anti-fetal rejection). The three major lesions are villitis (when the inflammatory process affects the villous tree), chronic chorioamnionitis (which affects the chorioamniotic membranes), and chronic deciduitis (which involves the decidua basalis). Maternal cellular invasion is a common feature of the lesions.

Villitis of unknown etiology (VUE) is a destructive villous inflammatory lesion characterized by the infiltration of maternal T cells (CD8+ cytotoxic T cells) into chorionic villi. Migration of maternal T cells into the villi is driven by the production of T cell chemokines in the affected villi. Activation of macrophages in the villi has been implicated in the destruction of the villous architecture. VUE has been reported in association with preterm and term fetal growth restriction, preeclampsia, fetal death and preterm labor. Infants whose placentas have VUE are at risk for death and abnormal neurodevelopmental outcome at the age of 2.

Chronic chorioamnionitis is the most common lesion in late spontaneous preterm birth and is characterized by the infiltration of maternal CD8+ T cells into the chorioamniotic membranes. These cytotoxic T cells can induce trophoblast apoptosis and damage the fetal membranes. The lesion is frequently accompanied by villitis of unknown etiology and evidence of maternal anti-fetal antibodies and deposition of complement in the umbilical vein. Chronic deciduitis consists of the presence of lymphocytes or plasma cells in the basal plate of the placenta. This lesion is more common in pregnancies resulting from egg donation and has been reported in a subset of patients with premature labor. Chronic placental inflammatory lesions are now considered to represent maternal anti-fetal rejection and this process can be associated with the development of a novel form of fetal systemic inflammatory response syndrome characterized by an elevation of the fetal plasma T cell chemokine (CXCL10).

The evidence that maternal anti-fetal rejection may underlie the pathogenesis of many chronic inflammatory lesions of the placenta is reviewed. This article includes figures and histologic examples of all chronic inflammatory lesions of the placenta.

Introduction

The term “chronic inflammation” refers to an inflammatory process characterized by infiltration of lymphocytes, plasma cells, and histiocytes (i.e. tissue macrophages) ¹. Chronic placental inflammatory lesions can be present in the villous tree, extraplacental chorioamniotic membranes, chorionic plate, and basal plate of the placenta (Figure 1). The etiology of chronic inflammatory lesions of the placenta is an important challenge ^2–4. Infection ^5–7 caused by viruses ^8–12, bacteria (i.e. Treponema pallidum, Mycobacterium tuberculosis)^11,13, and parasites (i.e. Plasmodium spp., Toxoplasma gondii) ⁸ has been implicated; however, most chronic inflammatory lesions are of unknown etiology (i.e. an infectious agent cannot be identified) ¹⁴, and accumulating evidence suggests that an immune process caused by maternal anti-fetal rejection plays a role in the pathogenesis of these conditions ^4,15–24.

Figure 1.

The human placenta. Chronic inflammatory lesions can affect different parts of the placenta. Chronic villitis refers to the inflammation involving the villous tree. Chronic chorioamnionitis involves either the extraplacental chorioamniotic membranes or chorionic plate. Chronic deciduitis affects the basal plate. Modified from Benirschke K, Burton GJ, Baergen RN. Infectious Diseases. Pathology of the Human Placenta. Sixth ed. Berlin Heidelberg: Springer; 2012. p. 33.

The placenta and fetus are semi-allografts and a maternal (host) immune response against paternal antigens (expressed in the placenta or fetus) can be considered analogous to allograft rejection ^25–34. We will review the evidence in support of the idea that many cases of idiopathic chronic placental inflammation reflect maternal anti-fetal rejection, in which the main effector is the infiltration of maternal CD8+ T cells (cytotoxic lymphocytes) into fetal tissues (Figure 2) ^15,16,35. This state is associated with the presence of fetal HLA-specific antibodies in the maternal serum ^{17,18, 21}, and C4d deposition in the umbilical vein ^17,22 as well as syncytiotrophoblast²⁴. The presence of fetal HLA specific antibodies in maternal serum was determined by first performing HLA genotyping using fetal genomic DNA, and then assessing whether the maternal HLA antibodies were directed against fetal antigens using a Luminex assay ^18,21. Furthermore, the concentrations of T cell chemokine CXCL10 are elevated in different fetal compartments, such as amniotic fluid and fetal plasma ^15,16,20. These phenomena resemble those observed in allograft rejection in solid organ transplantation ^36–49.

Figure 2.

Microscopic findings of chronic nonspecific villitis (villitis of unknown etiology; VUE). (A) Normal chorionic villi showing villous core with fetal vessels and stroma. The intervillous space is shown in white and contains maternal red blood cells. The rest of the image shows cross-sections of the villous tree of the placenta – each chorionic villus is lined with syncytiotrophoblast. Inside the villi, fetal capillaries are observed. (B) Destructive inflammation of the chorionic villus (asterisk). The inflammatory process is diagnosed by the presence of an infiltration of mononuclear cells. Obliteration of the villous capillaries is also seen in comparison with unaffected villi adjacent to the distorted villus (asterisk). Unaffected villi (black arrow). (C) Immunoperoxidase staining for CD8+ T cells. Cells stained in brown express CD8+ on their surface, and are, therefore, cytotoxic lymphocytes. These cells are of maternal origin, and are derived from the intervillous space. A-C, X200.

1.Chronic nonspecific villitis (Villitis of Unknown Etiology; VUE)

1.1 Definition and cellular composition

Villitis of unknown etiology (VUE) is a destructive inflammatory lesion characterized by the infiltration of maternal T cells into the chorionic villi (fetal tissue) ^50,51. The frequency of VUE varies among studies, and its prevalence ranges from 2% to 33.8% ^7,52–56. The wide range is thought to represent variations in the study population, sampling methods, and diagnostic criteria. The frequency of detection of VUE increases when the number of paraffin blocks from a given placenta increases to four ⁵².

The main maternal T cell subset infiltrating the chorionic villi is CD8+ cytotoxic T cells ³⁵. The maternal origin of these T cells was demonstrated by Redline et al. ⁵⁰ using in situ hybridization analysis with X and Y chromosome-specific probes, along with CD3 (a pan-T lymphocyte marker) and CD45 (a leukocyte marker) immunostaining of placental tissues obtained from four male neonates. T lymphocytes infiltrating the villous tree were consistently of maternal origin (as shown by the lack of Y chromosome signals) ⁵⁰.

Another important cell type in VUE is the macrophage, which is of fetal origin, as demonstrated by using chromogenic in situ hybridization with a Y chromosome-specific probe ³⁵. Placental macrophages, also known as Hofbauer cells, have an activated phenotype in VUE, as shown by expression of CD14 ³⁵. Increased CD14 immunoreactivity in VUE cases indicates that there pro-inflammatory activation of placental macrophages. Therefore, VUE is a unique inflammatory process involving T cells and macrophages originating from two different hosts (T cells from the mother and macrophages from the fetus) ^35,57. A key feature of VUE is fetal tissue damage by maternal cytotoxic T cells resembling allograft rejection. The relative numbers of T regulatory cells (CD4+ CD25+ FoxP3+) are also higher in placentas with VUE, compared to those without this lesion ⁵⁸. This is an interesting observation, given that T regulatory cells are implicated in the generation of a tolerogenic state in pregnancy ^34,59–63.

The destruction of the villous architecture by the inflammatory process appears to be related to apoptosis, which is more extensive in areas of the placenta with VUE, as demonstrated using transferase dUTP nick end labeling (TUNEL) staining ⁶⁴. Two mechanisms have been proposed to result in apoptosis: 1) the release of perforin and granzyme B by activated CD8+ T cells, which can induce caspase-dependent cell death in target cells; and 2) complement activation leading to pore formation. Immunohistochemistry studies show expression of perforin and granzyme B, and C5b-9 (membrane attack complex of the complement system) deposition in VUE ⁶⁴.

1.2 Grading

Histopathological grading of VUE is based on the number of chorionic villi affected, and whether the distribution of villous inflammation is patchy or diffuse. Redline has proposed a grading system using the number of villi affected per focus ⁵⁴. In low grade lesions less than 10 villi are affected and the lesions can either be focal (only one slide involved) or multifocal (more than one slide involved). High grade lesions are defined as those with more than 10 villi affected per focus, and are divided into patchy and diffuse subgroups ⁵⁴. The lesion is defined as diffuse when more than 5% of all distal villi are affected (Figure 3). Other grading systems have also been proposed ^52,65. The highergrade is the greater the risk of adverse pregnancy outcome ^66,67.

Figure 3.

The severity of villitis is assessed by a grading system. (A) Well-preserved chorionic villi in a normal placenta. (B) A low-grade lesion showing involvement of chorionic villi; less than 10 villi are affected. High-grade lesions are shown in 3C and 3D. 3C shows patchy involvement. The villi within the oval are affected, while those outside the oval are unaffected. Most of the villi are involved in 3D – this represents diffuse involvement. A-D, X100.

1.3 Pattern of involvement

Redline proposed that the patterns of VUE involvement be classified as distal, proximal and basal (Figure 4) ^54,66. The distal type is the most common (approximately 50% of cases), and is localized to distal villi (terminal and mature intermediate villi). Proximal occurs in 30% of the cases, and involves the proximal stem villi (sometimes the chorionic plate along with the distal villi) ⁵⁴. This type can be associated with obliterative vasculopathy and fetal vascular thrombo-occlusive disease, resulting in hyalinized avascular villi ⁵⁴. The third pattern is basal villitis, involving villi anchoring to the basal plate – this type is frequently associated with chronic deciduitis ⁵⁴.

Figure 4.

Histological description of villitis of unknown etiology (VUE). (A) An illustration showing the placental villous tree. Inflammation of the villous tree is the hallmark of villitis. (i) in Figure 4A is a stem villus. (ii) represents terminal and mature intermediate villi; (iii) represents anchoring villus. The infiltration of lymphocytes at this site represents proximal villitis. (B) The histological demonstration of villitis of the stem villi is shown (asterisk). (C) Distal villitis involving terminal and mature intermediate villi is the most common pattern of VUE (asterisks). (D) Basal villitis involves anchoring villus (asterisk). B-D, X100.

1.4 Pathogenesis

What drives maternal T cells in the intervillous space to infiltrate the chorionic villi? Gene expression profiling of placentas affected by VUE demonstrates overexpression of genes involved in the immune response. For example, there is overexpression of major T cell chemokines and their receptor (i.e. CXCR3). The T cell chemokines CXCL9, CXCL10, and CXCL11 are overexpressed in Hofbauer cells (placental macrophages of fetal origin), stromal and endothelial cells ¹⁵. Therefore, a chemotactic gradient could be established between villi and the intervillous space, where maternal lymphocytes circulate. Importantly, maternal CD8+ T cells within the villi express the receptor for the T cell chemokines CXCR3+, indicating that they are able to receive the chemotactic signal ¹⁵. Based on this evidence, we propose that activated Hofbauer cells attract maternal T cells into the villous compartment through the production of chemokines. Another potential mechanism facilitating maternal T cell infiltration into the chorionic villi includes increased expression of intercellular adhesion molecule-1 (ICAM-1) both in syncytiotrophoblast and maternal immunocytes ^68–71. It is also possible that small breaks in the syncytiotrophoblast contribute to the disease process ⁷².

In support of an immunological origin of this lesion, we reported that using transcriptomic analysis, placentas with VUE have an enrichment of genes involved in antigen presentation, such as class II major histocompatibility antigens (HLA-DM, -DO, -DP, -DQ, -DR) ¹⁵ and overexpression of class I molecules (HLA-B, -C, -G) ¹⁵. VUE is characterized by the proliferation of Hofbauer cells within the villi (demonstrated by nuclear immunoreactivity for Ki-67, a protein associated with cell proliferation). In summary, the gene expression profile of VUE resembles that observed in transplant rejection and graft-versus-host disease ^73–78.

The immunological changes in VUE are not restricted to the villi, since maternal and fetal plasma concentrations of CXCL9, −10, and −11 are also significantly higher in cases with VUE than in controls ¹⁵. Moreover, VUE is histologic evidence of maternal cell trafficking into the fetus, which could lead to the development and establishment of fetal-maternal microchimerism ⁷⁹. It remains to be established if maternal T cells are grafted into the visceral organs of the fetus. The precise nature of the fetal antigens that induce the maternal immune response has not been established. However, we have evidence that maternal sensitization to fetal specific anti-HLA antigens occurs in patients with VUE (see below).

Placentation provides a unique anatomical context for allograft rejection. The fetus (the placental villi are of fetal origin) differs from typical allografts, such as the kidney and the liver, in that it is a distinct immunologically-competent host (Figure 5).

Figure 5.

The unique immunological nature of villitis of unknown etiology (VUE). (A, B) Circulating maternal T cells entering the intervillous space can infiltrate the villus. (C) Following maternal T cell infiltration, Hofbauer cells (fetal placental macrophages) are activated. This would be considered semi-allograft rejection, because maternal T cells (recipient of the semi-allograft) infiltrate the placenta (semi-allograft).

1.5 Clinical significance

VUE is generally considered to be a lesion of term pregnancies, and is often subclinical in nature (normal pregnancy outcome) ⁸⁰. However, extensive involvement of the placenta by VUE has been associated with preterm and term fetal growth restriction ^{5,6,15,53,65,81–87}, spontaneous preterm delivery ^15,85,88, small for gestational age (SGA) associated with preeclampsia ^15,82,89,90, and fetal death ^{6,15,52,65,91}. The frequency of VUE in the “great obstetrical syndromes” ^92–94 is displayed in Figure 6. This lesion is also associated with the occurrence of chronic chorioamnionitis (Figure 6B, and the next section in this review).

Figure 6.

VUE is frequently associated with chronic chorioamnionitis; (A) The frequency of VUE in women at term not in labor, preterm labor with intact membranes, preterm PROM, preeclampsia at term, preterm preeclampsia, and small for gestational age. Modified from Kim CJ et al. Mod Pathol. 2010 Jul;23(7):1000-11; (B) The frequency of VUE is higher in the placenta of patients who had preterm labor with intact membranes and in patients with preterm PROM who had chronic chorioamnionitis than in those who did not. Modified from KIM CJ et al. Mod Pathol. 2010 Jul;23(7):1000-11.

VUE: villitis of unknown etiology; PROM: prelabor rupture of membranes

In a cohort study of 180 neonates with a birthweight below the 10[san]^th[san] centile delivered <34 weeks of gestation with abnormal umbilical artery Doppler velocimetry, the presence of villitis was an independent risk factor for infant death (odds ratio 5.7; 95% CI, 1.16–28.1) ⁹⁵ and VUE was a predictor of necrotizing enterocolitis ⁹⁵. Moreover, infants from pregnancies with placental villitis were at risk for an abnormal developmental outcome at the age of 2 (odds ratio 3.19; 95% CI 1.26–8.09) ⁹⁵. The mechanisms whereby VUE may lead to SGA with an abnormal Doppler of the umbilical artery are unknown. Becroft et al. argued that the extent of the placental parenchymal damage by VUE is not sufficient to restrict fetal growth, as only 0.01%–5% of the villous tree would be affected ⁶. We propose that the most likely mechanism whereby VUE compromises fetal growth is a systemic fetal inflammatory response (a fetal inflammatory response syndrome, type II – see below for details).

Diffuse chronic villitis has been associated with neurologic impairment in other studies ^{67, 96}. VUE is more frequent in the placenta of the smaller twin compared to the larger twin in dichorionic pregnancies ^97,98. Interestingly, the incidence of VUE is higher in placentas from patients who had ovum donation ^99–101. Pregnancy in cases of ovum donation represents a total allograft rather than a semi-allograft. These results strengthen the case for an immunological origin of VUE.

2. Chronic chorioamnionitis

2.1 Definition and cellular composition

Chronic chorioamnionitis is defined by the infiltration of mononuclear cells into the chorioamniotic membranes or the chorionic plate. The typical lesion of chronic chorioamnionitis shows patchy or diffuse infiltration of maternal CD8+ T cells (Figure 7). The extent and severity of T cell infiltration is less than neutrophilic infiltration in cases of acute chorioamnionitis ^102–104. The primary locus of interaction is between maternal CD8+ T cells, and the fetal cells are the choriodecidual border in the chorion laeve ¹⁶. Trophoblast damage by CD8+ T cells in the form of apoptosis can be demonstrated using double immunofluorescence staining with antibodies against CD8+ lymphocytes and M30 (a specific antibody against an epitope associated with cleavage of cytokeratin 18 during apoptosis) ¹⁰⁵. Figure 8 shows direct contact between CD8+ T cells and trophoblasts expressing M30. Trophoblast apoptosis often leads to thinning of the chorionic trophoblast layer with a “moth-eaten” appearance of the choriodecidual border ¹⁶.

Figure 7.

The mechanisms and progression of chronic chorioamnionitis. (A) Increased concentration of the intra-amniotic T cell chemokine CXCL10 (and increased expression of CXCL9, −10, and −11 in the chorioamnionitic membranes) induces migration of CXCR3+ T cells from the decidua (maternal tissue) into the chorioamniotic membranes (fetal tissue). (B) T cells (depicted in with a blue nucleus, originally located in the decidua on the left side of the figure) migrate into the chorioamniotic membranes (“amniotropism”; right side of the figure). The term “amniotropic” implies that the T cells are migrating towards the amnion.

Figure 8.

Chorionic trophoblast apoptosis by cytotoxic T cells. (A) Confocal microscopy with immunofluorescence of the chorioamniotic membranes in a case with chronic chorioamnionitis. Nuclei of all cells are stained blue. Green represents staining with an antibody against CD8. These cells are CD8+ cytotoxic cells of maternal origin which are infiltrating the chorion (fetal). M30 (antibody against an epitope associated with cleavage of cytokeratin 18 during apoptosis) detects trophoblast that has undergone apoptosis; apoptotic cells are depicted in red. Direct contact between cytotoxic CD8+ cells (green) and apoptotic trophoblast (red) is indicated with white arrows – this represents the cytotoxic effect of maternal CD8+ cells in inducing apoptosis of fetal cells. (B) Light microscopic image showing chorionic trophoblast apoptosis with increased cytoplasmic eosinophilia and pyknotic nuclei (arrows). Loss of trophoblasts results in shaggy and irregular choriodecidual border (moth-eaten appearance). Hematoxylin & Eosin, X200.

2.2 Grading

We proposed a grading system to assess the severity of chronic chorioamnionitis ¹⁶. Grade 0 indicates the absence of inflammation; grade 1 when there are more than two foci of inflammation or patchy inflammation; and grade 2 when diffuse inflammation is present ¹⁶. The stage of inflammation was scored as 1 if amniotropic lymphocytic infiltration was limited to the chorionic trophoblast layer sparing the chorioamniotic connective tissue, and stage 2 if lymphocytic infiltration into the chorioamniotic connective tissue was noted (Figure 9) ¹⁶. Importantly, cases with mild inflammatory lesions have significantly higher amniotic fluid CXCL10 concentrations than patients without these lesions¹⁶ The substantial increase in the amniotic fluid CXCL10 concentration in cases with low grade, low stage inflammation underscores that the lesion can be associated with a unique form of intra-amniotic inflammation, characterized by an increase in T cell chemokines, but not necessarily amniotic fluid IL-6, which is a marker of acute inflammation ¹⁶.

Figure 9.

Microscopic staging of chronic chorioamnionitis. (A) Normal chorioamniotic membranes with amnion, chorion and decidua. (B) Stage 1 chronic chorioamnionitis showing T cell infiltration confined to the chorionic trophoblast layer. The choriodecidual border is studded with lymphocytes, while the chorioamniotic connective tissue is spared. The arrows indicate infiltration of T cells along the choriodecidual border. (C) In stage 2 chronic chorioamnionitis, lymphocytic infiltration into the chorioamniotic connective tissue is clearly seen. The arrows indicate infiltration of T cells into the amniotic connective tissue below the epithelium of the amnion at the top of the picture. A-C, X200

2.3 Pathogenesis

The choriodecidual junction is a large interface between the mother and the fetus (see Figure 1), and maternal immune cells in the decidua can recognize fetal cells, specifically chorionic trophoblasts (i.e. chorion laeve). Under normal circumstances, an inflammatory response in the choriodecidual junction does not occur. The mechanisms responsible for the tolerogenic state in pregnancy include: 1) expression of non-polymorphic HLA-G expression in the trophoblasts ^106–109; and 2) T cell chemokine gene silencing in the decidual cells, as shown by Nancy et al. ¹¹⁰. Other mechanisms have been implicated, such as a role for T-regulatory cells ^{59, 60,111,112}, tryptophan catabolism by indolemine 2,3-dioxygenase (IDO) ¹¹³, T-cell apoptosis ¹¹⁴, complement ^115,116, and co-stimulatory molecules such as the programmed death ligand 1 (PDL 1) ¹¹⁷. The mechanisms implicated in fetomaternal tolerance are shown in Table 1. A full discussion of this subject is outside the scope of this review.

Table 1.

Proposed mechanisms implicated in the tolerogenic state of pregnancy

Expression of non-classical major histocompatibility complex molecules by trophoblast cells106–109,202

Chemokine silencing by methylation in decidual stromal cells limits T cell access to the placenta 110

Regulatory T cells 59,60,111,112

Maternal T cell apoptosis 114,203

Increased expression of tryptophan catabolism by the enzyme indoleamine 2,3-dioxygenase (IDO) 113,204–207

Increased expression of programmed death ligand (PDL) 1 117,208,209

Expression of high levels of complement regulatory proteins [Decay Accelerating factor (DAF), Membrane co-factor protein , CD59] 115,116,210–214

Progesterone 215–219

Activation of placental exosomes 220,221 and placenta endogenous retroviral envelope proteins 222,223 which have the following properties: Shift of the immune response from Th1/Th17 toward a Th2/Treg cell response 224,225 Inhibit natural cell killer cytotoxicity against trophoblast 220,226 Immunosuppressive properties including FasL (Fas Ligand) 203,227–229, programmed death ligand 1 (PDL1), MHC (major histocompatibility complex) molecules etc.

Chronic chorioamnionitis was first described by Gersell et al. in a case series of 17 placentas in which the inflammatory infiltrate in the chorioamniotic membranes consisted predominantly of lymphocytes ¹⁰². Importantly, an extensive study of amniotic fluid and placental microbiology, as well as immunohistochemical staining and maternal serology, did not show evidence of infection with either bacteria or viruses ¹⁰². Patients had no clinical evidence of infection, such as fever, a flu-like syndrome, or a rash, however, preterm birth occurred in 76% (13/17) of patients. Prior to this seminal observation, few reports had documented a chronic inflammatory infiltrate of the fetal membranes in patients with maternal rubella infection ¹¹⁸ and toxoplasmosis ¹¹⁹. Subsequently, Qureshi and Jacques reported 31 cases of chronic chorioamnionitis and emphasized that 71% (22/31) of the cases had VUE, suggesting a common immunological etiology for the two lesions ¹⁰⁴. Immunohistochemistry showed that the predominant cells were CD8+ lymphocytes, 39% (12/31) of the cases were associated with a preterm birth, and 16% (5/31) were SGA ¹⁰⁴.

The chemotactic gradient favoring migration of T cells from the decidua into the chorioamniotic membranes has been proposed to represent increased concentrations of the chemokine CXCL10 in amniotic fluid, and overexpression of CXCL9, CXCL10 and CXCL11 in the chorioamniotic membranes ¹⁶. Therefore, the gradient of T cell chemokine concentrations leads to CXCR3+ T cell chemotaxis into the chorioamniotic membranes and the chorionic plate. This is analogous to the situation of acute chorioamnionitis in that elevated amniotic fluid neutrophil chemokines, such as IL-8 ^120–129, promote chemotaxis of neutrophils. Chronic chorioamnionitis is also characterized by distinct changes in the amniotic fluid proteome compared to acute chorioamnionitis ¹³⁰. Among patients with preterm labor, 31 differentially expressed proteins have been identified in chronic chorioamnionitis compared to cases with either acute chorioamnionitis or controls without inflammatory lesions. Of interest is that the amniotic fluid concentration of glycodelin-A is decreased in chronic chorioamnionitis ¹³⁰. This molecule has been implicated in the maintenance of maternal tolerance against the semi-allogeneic placenta/fetus ¹³¹.

2.4 Clinical significance

The prevalence of chronic chorioamnionitis in the “great obstetrical syndromes” ^92–94 is displayed in Figure 10. The frequency in patients with spontaneous preterm labor and delivery is 34%, and 39% in patients with preterm PROM ¹⁶. In a separate study of consecutive cases of 1,206 preterm deliveries, chronic chorioamnionitis was the most common lesion (20.8%), and was particularly common among late preterm birth, which represents 70% of all preterm deliveries ²³. Importantly, we found that 60% of cases of fetal death had evidence of chronic chorioamnionitis, and high amniotic fluid concentrations of CXCL10. This suggests that some cases of unexplained fetal death may represent an extreme form of maternal anti-fetal rejection ^19,132.

Figure 10.

The frequency of chronic chorioamnionitis in “Great Obstetrical Syndromes”. The frequency is notably higher in the order of fetal death, preterm prelabor rupture of membranes (PROM), and preterm labor/delivery. Modified from Kim CJ et al. Mod Pathol. 2010 Jul;23(7):1000-11 and Lee et al. Histopathology 2011:59:928–938.

3. Chronic deciduitis

Chronic deciduitis is diagnosed by the presence of lymphocytes and plasma cells in the basal plate of the placenta (Figure 1 and Figure 11) ^133,134. Chronic microbial infection and immune mechanisms have been implicated in the etiology of chronic deciduitis. Originally recognized by Naeye when reviewing the placentas from the Collaborative Perinatal Study of the National Institute of Neurological and Communicative Disorders and Stroke (study of nearly 40,000 pregnancies in which clinical outcome and placentas were collected prospectively), chronic deciduitis was thought to be present in 1–2% of all pregnancies, and to be associated with fetal growth restriction and fetal death ¹³⁵. However, the original description involved infiltration of chronic inflammatory cells in both the decidua basalis and capsularis; therefore, chronic chorioamnionitis as well as chronic deciduitis of the basal plate may have been included in Naeye’s original report ¹³⁵.

Figure 11.

Chronic deciduitis. (A) Normal basal plate along with anchoring villus (asterisk). This image has cross-sections of villi, and the intervillous space is in white. The basal plate of the placenta is the horizontal tissue at the bottom of the picture. (B) Dense infiltration of the basal plate of the placenta and anchoring villi with mononuclear and plasma cells. Basal villitis (asterisk) is present. (C) Immunostaining confirms the presence of CD138+ positive plasma cells (brown color) in the basal plate. A-C, × 200.

Bendon and Miller focused attention on a lesion in the basal plate of the placenta whose prominent feature was the infiltration by plasma cells ¹³⁶. The frequency of the lesion was 4% (25/600) in placentas examined because of complications of pregnancy requiring admission to a neonatal intensive care unit. Bendon and Miller emphasized that plasma cells were not normally seen in the decidua parietalis of the same placentas, and proposed an immune origin for the lesion based upon the proximity of the plasma cells to trophoblast cells ¹³⁶. Adverse pregnancy outcome was common in these 25 cases, as there were 5 infants with intrauterine growth restriction, 5 fetal deaths, and 3 mothers had systemic lupus erythematosus. Of interest is that recurrent lesions were observed in 3 patients ¹³⁶.

The precise definition of chronic deciduitis in the basal plate was established by an international collaborative effort reported by Khong et al., in which 30 slides of placental sections were distributed to 6 experienced perinatal/placental pathologists to determine the degree of concordance in the identification of chronic deciduitis ¹³³. Pathologists scored the presence or absence of a lesion, its extent (focal, multi-focal, or diffuse), severity (mild, moderate, or severe), and the presence or absence of plasma cells; however, no prespecified definitions were provided to the pathologists, and the first round of examinations was used to generate a diagnostic tool for defining the disease ¹³³. The conclusion of the study was that the diagnosis should be made on the basis of severe and extensive lymphocyte infiltration and the presence of plasma cells ¹³³. Recently, the scoring system for chronic deciduitis has been proposed ¹³⁷. The criterion for the diagnosis of chronic deciduitis is the presence of ≥ 50 lymphocytes/per high-power field (Grade 1). Grade 2 and Grade 3 chronic deciduitis are characterized by the presence of lymphocyte in multiple and diffuse foci in at least one slide, respectively ¹³⁷. This classification correlates well with that of Khong et al ¹³³.

The association between chronic deciduitis and preterm labor without clinical chorioamnionitis was first reported by Edmonson et al. after examining the placentas of 39 patients with idiopathic preterm labor and 39 age-matched control placentas of singleton pregnancies (induced because of fetal congenital anomalies, excluding aneuploidy) ¹³⁸. The frequency of chronic deciduitis was significantly higher in patients with preterm labor than in a control group (41% vs 15%; p=0.02) ¹³⁸.

Evidence that chronic deciduitis may have an immune origin has been suggested, as this lesion is significantly associated with basal villitis and has been reported more frequently in the basal plate of the placenta of pregnancies resulting from egg donation vs. non-donor in vitro fertilization (IVF) pregnancies. In one study, the frequency of chronic deciduitis was 42% (14/33) in egg donation pregnancies, and 1.6% (1/60) in the controls (p=0.001) ¹⁰⁰. However, in a subsequent study, the frequency was 2.8% after oocyte donation vs. 1.8% in non-oocyte donor IVF (p=0.03) ¹⁰¹.

4. Chronic placental inflammation as maternal anti-fetal rejection

The placenta and fetus express both paternal and maternal antigens; they are semi-allografts ^34,139,140. The syncytiotrophoblast is in direct contact with maternal blood, and the chorion laeve is with decidua; thus, the maternal immune system is exposed to paternal antigens expressed by the fetus ^141,142. Immune tolerance is a requirement for successful pregnancy ^{25,26, 31,32,59,143,144}. Immune effector cells with fetal specificity are selectively “silenced” during pregnancy by complex mechanisms ^31,32. A full discussion of the mechanisms responsible for tolerance during pregnancy is outside the scope of this article; however, the main mechanisms are listed in Table 1.

In solid organ transplantation, breakdown of tolerance leads to rejection of the graft, and ultimately, injury ^145–152. Allograft rejection results from a cellular and/or humoral (antibody-mediated) immune response by the recipient of a graft ^145–152. The major histocompatibility complex (MHC) class I and II molecules include human leukocyte antigens (HLA), and this system is implicated in the rejection of solid organs ^153–157 as well as bone marrow ¹⁵⁸. We have proposed that disruption of the tolerogenic state of normal pregnancy leads to maternal anti-fetal rejection, placental damage, and complications of pregnancy (i.e. fetal growth restriction and spontaneous preterm labor). The extreme form of graft failure in organ transplantation is extensive damage of the transplanted organ – the equivalent to this in pregnancy would be fetal death caused by maternal anti-fetal rejection ^19,132.

The two mechanisms of allograft rejection are cell- (T-cell) and antibody-mediated ^145–152. In the context of pregnancy, the histopathologic manifestations of cell-mediated rejection are VUE and chronic chorioamnionitis. In the former, the battleground for rejection is the chorionic villi, where maternal T-cells invade the villous tree; in the latter, the battleground is the extraplacental choriodecidual interface, where maternal T-cells infiltrate the chorioamniotic membranes. A cellular mediated feature of maternal anti-fetal rejection can be detected in maternal systemic circulation in patients with chronic chorioamnionitis ¹⁵⁹. The proportion of CD300a+, cytotoxic T lymphocytes, is significantly increased in patients with chronic chorioamnionitis than in those without this lesion ¹⁵⁹. Moreover, we have found an increased number of CD CD300a+CD8+ T cells, which overexpress mRNA of granzyme genes (GZMA, GZMB, and GZMK), granulysin (GNLY), and perforin (PRF1) indicated that CD300a+CD8+ T lymphocytes have more cytotoxic phenotype ¹⁵⁹.

Antibody-mediated rejection has been reported in both VUE ^24,64 and chronic chorioamnionitis ^17,18,21. The effector mechanism for rejection in antibody-mediated allograft damage involves complement activation. Deposition of complement C4d in the umbilical vein endothelium in mothers with complications of pregnancy and detectable maternal anti-fetal antibodies has been demonstrated in patients with both chronic chorioamnionitis ¹⁷ and VUE ^{22, 24}.The most frequent alloantigens involved in transplant rejection are encoded by the major histocompatibility complex (MHC) genes, or HLA in humans ^160–162. The degree of HLA mismatch between the donor and the recipient is a risk factor for rejection, as is the presence of existing donor-specific antibodies in the recipient ^163–171. Sensitization of mothers to fetal-specific HLA antigens occurs frequently during pregnancy, increases with gestational age and parity, and has traditionally been considered benign ^172–175. However, we have recently demonstrated that maternal HLA sensitization diagnosed in the midtrimester is a risk factor for spontaneous preterm delivery (OR 2.8; p=0.01) ²¹, and the strength of the association increases (OR 5.9; 95% CI 1.6–21.83, p=0.008) when patients present with spontaneous preterm labor ¹⁷.

Maternal HLA sensitization is necessary, but not sufficient, for antibody-mediated maternal anti-fetal rejection. Antibodies specific to fetal antigens must also cross the placenta, activate complement, and damage the semi-allograft (placenta and/or fetus) – we have reported that this occurs in cases of chronic chorioamnionitis and massive perivillous fibrin deposition, also known as maternal floor infarction ¹⁷⁶. In both conditions, the following evidence supports maternal anti-fetal rejection: 1) maternal HLA sensitization; 2) fetal HLA-specific antibodies; 3) complement deposition in the umbilical vein; and 4) chronic chorioamnionitis or VUE ²⁰. We consider preterm labor/delivery ²¹, fetal growth restriction, and fetal death ^19,132 as the clinical manifestations of graft injury. Similar findings have been reported in cases of maternal floor infarction with fetal death ¹⁷⁷. Why some sensitized mothers develop maternal anti-fetal rejection and others do not requires further investigation.

5. Fetal Inflammatory Response Syndrome, type 2

When maternal anti-fetal antibodies cross the placenta and induce an alloimmune reaction (Figure 12), this can lead to a systemic fetal inflammatory response; however, this process is different from fetal systemic inflammation observed in the context of intrauterine infection ^20,178–182. In microbial invasion of the amniotic cavity, bacteria lead to activation of the innate immune system and the production of neutrophil chemokines (e.g. IL-8), which generate a chemotactic gradient attracting neutrophils from the decidua into the chorioamniotic membranes, leading to acute histologic chorioamnionitis ^120–129. Fetal systemic inflammation in the context of intra-amniotic infection leads to an elevation of the cytokine IL-6 in fetal plasma – we refer to this condition as the fetal inflammatory response syndrome (FIRS) type 1, and the histologic hallmark is funisitis and chorionic vasculitis ¹⁸³. A full description of FIRS type 1 and details about its biology and long-term consequences are described in a companion article in this issue of the Journal ¹⁸⁴. Fetuses born to mothers whose placentas have evidence of VUE and/or chronic chorioamnionitis frequently have elevated concentrations of the chemokine CXCL10 ^{15, 20}. This is also the case in amniotic fluid; however, fetal plasma concentrations of IL-6 are generally not elevated in these fetuses, suggesting a different form of systemic inflammatory process ²⁰. CXCL10, a ligand for CXCR3, is chemotactic for activated T-cells, macrophages, and NK cells ^185–187. Notably, CXCL10 is one of the most commonly expressed chemokines during allograft rejection and graft-versus-host disease ^{47,49,188–191}. An elevated intra-graft CXCL10 is associated with renal ^{36,38–42,46}, lung ^44,192, and cardiac allograft rejection ^37,43,193 (akin to the overexpression of CXCL10 in the chorionic membranes in cases with chorioamnionitis, or in the placenta in cases with VUE). An elevated CXCL10 concentration before organ transplantation is predictive of poor allograft outcome ^{38,41,43,44,191}.

Figure 12.

The mechanism whereby an antibody-mediated alloimmune reaction leads to a fetal inflammatory response. (A) Maternal anti-fetal HLA antibodies (depicted as “Y” shape) are present in the maternal circulation, can cross the placenta, and enter the fetal circulation. This process is represented by the blue arrow. (B) Antibodies in the fetal circulation (in the lumen of the fetal vessels in the villus) gain access to the fetus and illicit a systemic inflammatory response, illustrated by the different color of the fetus. Antibodies can lead to complement deposition in the fetal endothelium.

Studies of the white blood cell transcriptome and plasma proteome of fetuses with evidence of maternal anti-fetal rejection identified stereotypic changes in both ¹⁵. The maternal anti-fetal rejection phenotype is defined when the patients/fetuses met two or more of the following criteria: 1) presence of chronic placental inflammation; 2) ≥80% of maternal HLA class I panel reactive antibodies (PRA) positivity; and 3) fetal serum CXCL10 concentration >75th percentile ²⁰. Gene ontology analysis showed enrichment of 24 biological processes, such as “response to other organism” and “killing by host of symbiont cells”. A set of differentially-expressed genes discovered by microarray experiments were subsequently confirmed using qRT-PCR ²⁰. We found universal down-regulation of mRNA expression of genes for neutrophil granule proteins and the polymorphonuclear leukocyte surface marker, CD66b ²⁰. All the evidence indicates that there is another type of systemic fetal inflammatory response associated with maternal anti-fetal rejection and an elevated fetal plasma concentration of CXCL10. We propose this as a second type of FIRS (FIRS type 2) (Figure 13).

Figure 13.

Fetal inflammatory response syndrome (FIRS): type 1 vs. type 2. (A) Systemic fetal inflammatory response associated with intra-amniotic infection is FIRS type 1, while the one associated with maternal anti-fetal rejection is FIRS type 2. In FIRS type 1 and type 2, there is elevation of amniotic fluid IL-6 and CXCL10, respectively. (B) Fetal serum CXCL10 concentration is significantly elevated in cases with chronic placental inflammation, while there is no such change in those with acute chorioamnionitis. (C) In acute chorioamnionitis cases, there is significant elevation of fetal serum IL-6 concentration, while CXCL10 concentration does not change. ACA: acute chorioamnionitis, CCA/VUE/CDP: chronic chorioamnionitis/villitis of unknown etiology/chronic deciduitis with plasma cells. Modified from Lee J Am J Reprod Immunol 2013 70:265-80

6. Summary and conclusions

A large body of evidence now supports that idiopathic chronic placental inflammation (villitis of unknown etiology/chronic chorioamnionitis/chronic deciduitis) represents the histological manifestations of maternal anti-fetal cellular rejection, and is common in human pregnancies. It is possible to screen for anti-fetal antibody-mediated rejection by identification of maternal serum antibodies against fetal HLA, and determine whether they are specific for the fetus in the index pregnancy (fetal HLA profile). Further investigation is required to develop and implement tools for the diagnosis of maternal anti-fetal rejection during pregnancy.”

Maternal anti-fetal rejection is significantly associated with late preterm birth and other obstetrical complications ²³. Epidemiologic studies show an association between late preterm birth and clinical adverse outcomes in childhood, adolescence, and adulthood ^194–200. A recent report of an association between late preterm birth and impaired neurocognitive performance in late adulthood in a Finnish birth cohort underscores the potential importance of this mechanism of disease²⁰¹. Therefore, further clinical and molecular phenotyping of maternal anti-fetal rejection is needed so that effective monitoring and management of pathological maternal anti-fetal rejection during pregnancy be feasible. Once methods for the identification and diagnosis of maternal anti-fetal rejection become available, clinical trials would be required to determine if interventions such as steroids or immunosuppressive agents could improve pregnancy outcome. The major questions to be answered include the mechanistic explanation of preterm labor associated with maternal anti-fetal rejection, and the solution of the riddle for seemingly normal pregnancy outcome with substantial evidence of maternal anti-fetal rejection.