Methods

Patients and control subjects

Three patients with SSc referred to the Division of Rheumatology of the University of Florence were studied. Case 1, a 29-year-old woman with diffuse SSc12 and a modified Rodnan skin score of 2/51, presented with antinuclear and antitopoisomerase I antibodies, giant capillaries and capillary haemorrhages at nailfold videocapillaroscopy, a history of digital ulcers, normal pulmonary function, small areas of ground-glass opacity on the high-resolution CT scan, normal pulmonary artery pressure and no cardiac or renal involvement. During pregnancy she experienced a rapidly progressive increase in the skin score (14/51) and arthralgias and was treated with hydroxychloroquine and low-dose steroids. After 34 weeks of gestation the pregnancy was successfully ended. Case 2 was a 36-year-old woman with limited SSc,12 Raynaud’s phenomenon, antinuclear and antitopoisomerase I antibodies, giant capillaries and capillary haemorrhages at videocapillaroscopy with no internal organ involvement. Delivery was at 38 weeks. Case 3 was a 37-year-old woman with limited SSc, significant oesophageal involvement, antinuclear and antitopoisomerase I antibodies who delivered at 37 weeks.

Four healthy women with uncomplicated pregnancies and matched to the women with SSc for gestational age (trimester) were examined as a control group. All SSc and control women had a caesarean delivery for obstetric indications and gave written informed consent to use their placentas in the study as approved by the local ethics committee.

Histopathological analysis

At delivery the placentas were weighed and full-thickness biopsies were obtained. A stratified random sampling procedure was performed to obtain six full-thickness blocks per organ. The specimens were fixed in 10% buffered formalin, dehydrated in alcohol graded series and embedded in paraffin. Sections (3 μm) were stained with haematoxylin and eosin, Masson’s trichrome and phosphotungstic-acid-haematoxylin (PTAH; Bio-Optica, Milan, Italy) to evaluate pathological changes, tissue fibrosis and fibrinoid material deposition. The sections were carefully examined by an expert pathologist who was blind to sample classification.

Immunohistochemistry

Serial sections (3 μm) were heated in 10 mM citrate buffer (pH 6.0) for antigen retrieval. After blocking endogenous peroxidase activity and non-specific site binding, the sections were incubated with the following primary antihuman antibodies: mouse monoclonal anti-VEGF (VEGF-A) (1:50, BD PharMingen, Heidelberg, Germany), rabbit polyclonal anti-PlGF (1:100, Abcam, Cambridge, UK), rabbit polyclonal anti-VEGFR-1 (1:50, Santa Cruz Biotechnology, Santa Cruz, California, USA), rabbit polyclonal anti-VEGFR-2 (1:50, Abcam), mouse monoclonal anti-α-smooth muscle actin (α-SMA) (1:50, Abcam) and rabbit polyclonal anti-CTGF (1:400, Abcam). The sections were incubated with biotinylated secondary antibodies and avidin-biotin-peroxidase complex (UltraVision Detection System, LabVision, Fremont, California, USA). The immunoreactivity was developed using 3-amino-9-ethylcarbazole or 3,3′-diaminobenzidine tetrahydrochloride (Vector, Burlingame, California, USA) as chromogens. Some sections were counterstained with Mayer’s haematoxylin. Normal isotype-matched IgG (Sigma, St Louis, Missouri, USA) were used to replace the primary antibodies as negative controls. The sections were observed under a light microscope (Eclipse E400, Nikon, Tokyo, Japan) and photographed by digital camera (Coolpix 2500, Nikon).

Analysis of immunostaining intensity

The immunostaining for CTGF, VEGF, PlGF, VEGFR-1 and VEGFR-2 was assessed using a semiquantitative scoring system. Three fields (×20 magnification) from three random sections of each specimen were analysed by two blinded observers and scored for location of staining in trophoblast, stromal cells and vessels. Histological grading was based on the intensity of the staining (– negative, +/– weak, + moderate, ++ intense).

Results

Placental weights were consistent with the estimated gestational age in both patients with SSc and controls.

Histopathological findings

In all SSc placentas, signs of decidual vasculopathy, villous hypovascularity and severe perivascular and stromal fibrosis of decidua and chorionic villi were observed (fig 1). In addition, an increased syncytiotrophoblast knotting was found in all SSc cases.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Pathological features of placentas from patients with systemic sclerosis (SSc). (A, B) Masson’s trichrome staining of SSc and control placental tissue. In (A), placentas from women with SSc show signs of decidual vasculopathy such as foamy degeneration of endothelial cells (arrow) and thickening and delamination of the vessel wall (inset). Perivascular and stromal fibrosis of maternal decidua (A) and chorionic villi (B, arrows) is evident in SSc placentas compared with gestational age-matched healthy controls. Original magnification ×40 in A, ×10 in B. (C) Phosphotungstic-acid-haematoxylin staining of SSc and control placental tissue. SSc placentas show an extensive deposition of fibrinoid material in large infarcted areas (C) around decidual blood vessels and in the intervillous spaces (arrow in inset). Original magnification ×10.

Masson’s trichrome staining revealed an excessive perivascular and stromal extracellular matrix deposition in maternal decidua and chorionic villi of placentas from patients with SSc (fig 1A, B). In particular, cases 1 and 3 showed more severe fibrosis whereas, in case 2, areas of stromal fibrosis as well as areas of villous hyalinosis were observed. All SSc placentas had pathological features of vasculopathy and hypoperfusion of chorionic villi. Most decidual blood vessels showed a thickened and delaminated wall with reduction in the vessel lumen (fig 1A). Furthermore, foamy degeneration of decidual endothelial cells partially occluding the vessel lumen was found (fig 1A). In placentas from patients with SSc, extensive deposition of fibrinoid material was detected around decidual vessels and in the intervillous spaces (fig 1C). Placentas from women with SSc had increased and larger infarcts in chorionic villi (signs of hypermaturity) compared with gestational age-matched controls (fig 1C). In case 1 a greater number of placental infarcts as well as hypoplasic villi with small peripheral capillary vessels (signs of retarded growth and maturation) were observed.

Immunohistochemical findings

Immunohistochemical analyses for CTGF, the myofibroblast marker α-SMA and VEGF family angiogenic factors/receptors were performed. In placentas from women with SSc, strong CTGF immunopositivity was detected in decidual cells, in some areas of the trophoblast, in the vessel wall (pericytes, smooth muscle cells), endothelial cells lining the capillary lumen and in stromal fibroblasts of the chorionic villi (fig 2). In contrast, very weak immunostaining was observed in control placentas (fig 2).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Expression of connective tissue growth factor (CTGF) and α-smooth muscle actin (α-SMA) in placentas from patients with systemic sclerosis (SSc) and controls. CTGF immunostaining: in SSc placentas a strong immunopositivity for CTGF (brown staining) is seen in decidual cells (arrow), in some areas of the trophoblast (asterisk), in the vessel wall (arrowheads, inset) and in stromal cells of the chorionic villi. In contrast, very weak immunostaining is seen in control placentas. Blue haematoxylin counterstain. Original magnification ×20 (×40 in inset). α-SMA immunostaining: α-SMA immunopositivity (red staining) is evident in vascular smooth muscle cells and pericytes in both SSc and healthy placentas. Many stromal cells (arrows) show immunopositivity for α-SMA in SSc placentas only, suggesting a transdifferentiation of stromal fibroblasts to activated α-SMA+ myofibroblasts. Original magnification ×20.

α-SMA immunoreactivity was evident in vascular smooth muscle cells and pericytes in both healthy and SSc placentas. Interestingly, many α-SMA+ stromal cells were found in SSc placentas only, suggesting a transdifferentiation of resting fibroblasts to activated α-SMA+ myofibroblasts (fig 2). In SSc specimens, endothelial cells of most vessels, stromal cells and the trophoblast showed a stronger VEGF immunopositivity than in control placentas (fig 3). In SSc and controls, endothelial cells, fibroblasts, decidual cells and the trophoblast were immunopositive for PlGF, but SSc samples showed weaker expression (fig 3). Furthermore, in SSc placentas, VEGFR-1 immunostaining was similar to controls in all placental components (fig 3). VEGFR-2 was strongly expressed by endothelial cells in many capillary vessels within the chorionic villi of SSc placentas but was weakly or not expressed in controls (fig 3). The results of immunostaining intensity analysis are summarised in table 1 in the online supplement.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Expression of vascular endothelial growth factor (VEGF), placenta growth factor (PlGF) and the receptors VEGFR-1 and VEGFR-2 in placentas from patients with systemic sclerosis (SSc) and controls. VEGF immunostaining: in SSc placentas, endothelial cells (arrows in the inset), stromal cells and the trophoblast (arrowheads in the inset) display a stronger immunopositivity for VEGF than in control placentas. PlGF immunostaining: SSc samples show weaker PlGF immunopositivity than controls in all placental components. VEGFR-1 immunostaining: in SSc the expression of VEGFR-1 is similar to controls in the trophoblast (arrowhead in the inset), endothelial cells (arrows in the inset) and stromal cells within chorionic villi. VEGFR-2 immunostaining: VEGFR-2 is strongly expressed by endothelial cells of many capillary vessels (arrows in the inset) within chorionic villi of SSc placentas, whereas it is weakly or not detected in control placentas. Original magnification ×10 (×20 in insets).

Discussion

In this study we provide evidence to show that SSc placentas are affected by fibrosis and abnormal vascular remodelling. These modifications may affect the placental function and favour the intrauterine growth restriction and premature delivery seen in clinics.3 5 13

The main pathological features observed in SSc placentas were a thickened and delaminated wall of decidual vessels and foamy degeneration of endothelial cells, as well as chorionic villous infarcts, increased syncytiotrophoblast knotting and severe villous hypovascularity and fibrosis. These vascular findings are similar to those observed in other organs in SSc14 and are consistent with the definition of “decidual vasculopathy” proposed by Langston et al15 and the “uteroplacental vascular pathological process” described by Salafia et al.16 Moreover, SSc placentas showed signs of hypermaturity, as indicated by the increased and larger infarcted areas.

We also analysed the expression of specific markers of fibrosis and the vascular remodelling process. Severe perivascular and stromal fibrosis of decidua and chorionic villi, as well as marked overexpression of CTGF and many α-SMA+ myofibroblasts, were detected. Thus, SSc placentas display fibrotic features similar to those observed in the skin and visceral organs of patients with SSc.1 17 Furthermore, an extensive deposition of fibrinoid material was found around decidual vessels and in the intervillous spaces leading to vessel collapse. Taken together, such pathological features may lead to uteroplacental hypoperfusion and tissue hypoxia, which is a major stimulus for angiogenesis.

Increasing evidence suggests that the final geometry of the villous vascular bed is defined, to some degree, by the balance of VEGF/PlGF and their receptors.8 A predominance of VEGF/VEGFR-2 promotes establishment of richly branched low-resistance capillary beds within immature placental villi during the first two trimesters of pregnancy. By contrast, the presence of poorly branched terminal capillary loops in the last trimester may be controlled by the predominance of PlGF/VEGFR-1.8 The balance between VEGF and PlGF expression is altered in SSc placentas. In fact, we found increased immunostaining for VEGF and VEGFR-2, while PlGF was markedly decreased. Our data suggest that an impairment in the physiological switch from VEGF-driven branching to PlGF-driven non-branching angiogenesis may be responsible for the altered vascular remodelling and hypoperfusion in SSc placentas. Interestingly, these results mirror those previously reported in the skin of patients with SSc where an uncontrolled expression of VEGF and its receptors has been proposed to lead to altered vessel morphology/function such as giant and bushy capillaries.11

In conclusion, this study shows an active vascular and fibrotic pathological process in the placenta, suggesting that the pathogenetic mechanism in SSc is an ongoing one that may affect newly formed tissues. Such pathological changes may favour the hypoperfusion of the maternal-fetal interface causing fetal suffering and the risk of premature delivery.