- Email: [email protected]
Decreased vascularization and cell proliferation in placentas of intrauterine growth–restricted fetuses with abnormal umbilical artery flow velocity waveforms
Decreased vascularization and cell proliferation in placentas of intrauterine growth–restricted fetuses with abnormal umbilical artery flow velocity waveforms Chie-Pein Chen, MD, PhD,a Rekha Bajoria, PhD,b and John D. Aplin, PhDb Taipei, Taiwan, and Manchester, United Kingdom OBJECTIVE: The purpose of this study was to compare the morphologic features of placentas in severe intrauterine fetal growth restriction with abnormal umbilical artery blood flow velocity waveforms and normal gestation. STUDY DESIGN: Immunohistochemical methods were used to evaluate cell proliferation, vascular density, and α-smooth muscle actin expression by stromal cells in a group of 9 age-matched intrauterine growth–restricted and control placentas at 25 to 41 weeks of gestation. RESULTS: Fewer MIB1-positive nuclei were observed in both trophoblast and stromal cell populations in intrauterine growth restriction, which indicates fewer cells in cycle. Furthermore, a greatly reduced vascular density was observed, along with higher levels of α-smooth muscle actin expression in stromal cells. CONCLUSION: Intrauterine growth–restricted placentas show reduced cell proliferation in both trophoblast and stromal cell compartments. Peripheral villous vascularization is highly reduced. (Am J Obstet Gynecol 2002;187:764-9.)
Key words: Cell proliferation, fetal growth restriction, vascularization, umbilical artery, velocity waveform
Intrauterine fetal growth restriction (IUGR) is an important cause of perinatal death and morbidity. The precise definition of IUGR remains controversial and is likely to be a heterogeneous condition, but it has been shown in numerous studies that reduced umbilical artery blood flow velocity is associated with IUGR.1 Doppler waveform abnormalities have been shown to be useful in the prediction of poor perinatal outcome in IUGR fetuses.2 Elevated fetoplacental vascular impedance is proposed to be associated with impaired maternal-fetal respiratory gas exchange, which leads to an increased risk of chronic fetal hypoxia and acidosis. Abnormal (decreased or absent) end-diastolic flow velocity in the umbilical artery is associated with a perinatal mortality rate of 41%.3 Subsequent intensive fetal monitoring and timely elective delivery can reduce this risk up to 50%.4 Thus, Doppler ultrasound scanning is an important adjunct for the diagnosis and treatment of IUGR.2,4 The placental features that are associated with the delivery of an IUGR infant From the Division of High Risk Pregnancy, Department of Obstetrics and Gynecology, Mackay Memorial Hospital,a and the Academic Unit of Obstetrics and Gynecology, Medical School, University of Manchester.b Received for publication October 26, 2001; revised March 19, 2002; accepted March 28, 2002. Reprint requests: Chie-Pein Chen, MD, PhD, Flat 1 4F 116 Sec 3 ShinYi Road, Taipei 106, Taiwan. E-mail: [email protected] © 2002, Mosby, Inc. All rights reserved. 0002-9378/2002 $35.00 + 0 6/1/125243 doi:10.1067/mob.2002.125243
764
may give clues to the pathogenesis of the condition. However, studies of IUGR placenta have been hampered by imprecise or overly broad definitions of the study group and problems that involve the recruitment of wellmatched control subjects.5 In the present study, we have selected a well-defined group of severely growth-restricted preterm infants with abnormal end-diastolic blood flow in the umbilical artery and paired them with a normally grown control group of equal gestational age. We characterized cell proliferation, villous branching, villous vascularity, and stromal cell maturation in the peripheral villous placenta. We show dramatically altered placental characteristics that are suggestive of a developmental cause. Material and methods Patients. All pregnancies were singleton and resulted in normally formed infants. All women were healthy and normotensive in early pregnancy. In each case, the gestational age was assigned by an initial ultrasound examination that was performed before 16 weeks of gestation. All the clinical details of these patients were obtained retrospectively from the hospital chart records. Nine pregnancies were identified in which the diagnosis of IUGR was made according to the criteria of fetal birth weight <10th percentile6 with abnormal umbilical artery Doppler waveforms and reduced amniotic fluid volume (amniotic fluid index, ≤5). All the infants were delivered within 1 week
Chen, Bajoria, and Aplin 765
Volume 187, Number 3 Am J Obstet Gynecol
Table I. Primary antibodies Antibody
Clone
Supplier
Form
MIB-1
—
Dianova, Hamburg, Germany
Mouse immunoglobulin G1
CD34 α-Smooth muscle actin
QBEND10 1A4
Serotec, Ltd, Oxford, UK Sigma Chemical Company, St Louis, Mo
Mouse (culture supernatant) Mouse ascites
when abnormal umbilical artery Doppler waveforms were diagnosed. Nine deliveries from spontaneous preterm labor and term pregnancy without any signs and symptoms of chorioamnionitis and with grossly normal placentas that were matched for the closest gestational age were selected as control subjects. Pulsed Doppler ultrasonic recordings from the umbilical artery were made with a 100-Hz filter, and absent or abnormal end-diastolic blood flow velocity was documented from a minimum of five consecutive waveforms in two free-floating loops of cord. Abnormal end-diastolic blood flow was defined as pulsatility index (systolic-diastolic/mean) that was >2 SDs above the mean.7 Tissue. Placenta was obtained with local ethics committee approval. After delivery, the membranes and cord were removed, and the placentas weighed. One 2-cm3 tissue block, from the central area but excluding infarcted areas and peripheral margin of placenta, was randomly excised from each placenta and fixed in 10% buffered formaldehyde overnight at room temperature. Subsequently, the specimens were embedded in paraffin wax. Antibodies. The primary antibodies used for immunohistochemistry are shown in Table I. Immunohistochemistry. Sections 5 µm thick were deparaffinized. For microwave antigen recovery, the slides were put in a loosely covered thermoresistant plastic box (BDH Laboratory Supplies, Poole, UK) with 200 mL of 10 mmol/L citrate buffer (pH 6). They were exposed to microwaves at the medium setting (microwave: maximum power 800 W, with turntable), for 10 to 20 minutes (5 minutes per cycle); thereafter, the sections were allowed to cool for ≥15 minutes. Then, the slides were rinsed rapidly with distilled water at room temperature and washed with 0.05 mol/L TRIS-buffered saline solution (TBS; pH 7.6) three times for 5 minutes each. For immunostaining, the sections were treated with protein block (Dako, Carpinteria, Calif) for 20 minutes and then incubated for 1 hour at room temperature, with primary antibody diluted in phosphate-buffered saline solution. Next, the slides were washed three times for 5 minutes each in TBS and incubated with biotinylated rabbit anti-mouse secondary antibody in TBS (1:100; Dako) for 1 hour at room temperature. After the three 5-minute washes in TBS, the slides were incubated with an avidin-
Pretreatment
Working concentration
Microwave (4 times 1:20 at 5 minutes each) or wet autoclave (5-10 min) No 1:10 No 1:400
biotin-peroxidase complex that was prepared from the Vectorstain ABC Elite kit (Vector Laboratories, Peterborough, UK). Immunoreactivity was visualized with a diaminobenzidine (DAB)/nickel substrate (Vector Laboratories). The sections were then counterstained with 0.25% methyl green for 20 seconds, rinsed in tap water, dehydrated in ascending alcohol (70%-100%), rinsed in xylene, and mounted with XAM neutral medium (BDH). No staining was observed in the absence of primary antibody or in the presence of a variety of control primaries. Morphometric analysis. Each tissue section was covered with a transparent grid consisting of 4 8 6-mm squares, 15 of which were randomly selected and numbered. One microscopic field over the right lower angle of each square in each field was examined with a 10 objective. Therefore, 15 fields in each placenta were examined. The peripheral nonmuscularized intermediate and terminal villi constitute the largest proportion of the placental volume in the third trimester and are the primary sites of gas and nutrient transfer.8 These villi were evaluated together as peripheral villi. The total cross-sectional area of structurally intact villi and the number of peripheral villi in these squares were determined (imaging system KS 400; Kontron, Eching, Germany). The following features were also calculated: area of the villous stroma, the number of capillaries, the area of the capillaries, and MIB-1 proliferation antigen–positive cells per cross-section of villi. The outer trophoblast layer was visualized by methyl green counterstaining. Intravillous vascular elements were recognized by anti-CD34 antibody staining. The number of capillaries within the villous stroma was calculated after segmentation of captured images. Contours of the stroma of the placenta villi and all the included vascular elements were traced manually on the computer monitor with a mouse-controlled cursor at an on-screen magnification of 100 to calculate the areas of these components. MIB-1 recognized a nuclear proliferation-associated antigen, which is expressed in nearly all phases of the cell cycle except G0 and first G1 phase after a G0 stage.9 For each case, MIB-1 positive nuclei were counted separately for the trophoblast and stromal compartments. The sections were examined under a magnification of 250, and the numbers of MIB-1 positive nuclei
766 Chen, Bajoria, and Aplin
September 2002 Am J Obstet Gynecol
Table II. Clinical characteristics of pregnancies that were complicated by IUGR with abnormal umbilical artery Doppler waveforms IUGR
Gestational age (wk)
1 2 3 4 5 6 7 8 9
25 27 28 31 31 35 35 36 41
Birth weight (g)
Percentile
450 570 656 1150 580 1400 1880 1470 2040
<10 <10 <10 <10 <10 <10 <10 <10 <10
Umbilical artery Doppler waveform Absent end-diastolic volume Absent end-diastolic volume Reversed end-diastolic volume Absent end-diastolic volume Absent end-diastolic volume Absent end-diastolic volume Abnormal end-diastolic volume* Abnormal end-diastolic volume* Absent end-diastolic volume
*Pulsatility index, ≥2 SD.
Table III. Clinical characteristics of control pregnancies Control 1 2 3 4 5 6 7 8 9
Gestational age (wk)
Birth weight (g)
Percentile
26 27 28 30 32 32 33 34 38
613 980 1100 1560 1520 1654 2240 2510 3410
10-90 10-90 10-90 10-90 10-90 10-90 10-90 10-90 10-90
were related to the total villous cross-sectional area of peripheral villi. MIB-1–positive villous stromal cells per 10,000 square pixels total villous stromal area were assessed. The discrimination between relatively small numbers of cytotrophoblast nuclei and the much larger numbers of syncytiotrophoblast nuclei is difficult, but for the analysis of MIB-1 staining it was assumed that only the cytotrophoblast cells proliferate.8 For the α-smooth muscle actin staining, sections from different patients were processed in one run for the same monoclonal antibody, and the time incubated with DAB substrate was kept the same for all the slides. A scoring system for semiquantitative evaluation of the tissue sections was used in the following manner: 6, the strongest intensity with homogenous marked staining in all part of section; 5, >70% of the section showed marked staining, with the other part of the section revealing moderate staining; 4, the stain intensity is moderate and homogenous in all part of section; 3, >70% of the section showed moderate staining, the other part of the section revealing weak staining; 2, the stain intensity is weak and homogenous in all parts of section; and 1, >70% of the section showed weak staining, with the other parts of the section being revealed as very weakly stained or negative. All stainings were evaluated by two observers who were blind to the clinical diagnosis. Statistical analysis. All parameters of these two groups were described as mean (±SD) and were examined
Umbilical artery Doppler waveform Normal Normal Normal Normal Normal Normal Normal Normal Normal
whether the distribution of these parameters was parametric or nonparametric by Kolmogorov-Smirnov onesample test. The differences were assessed by Wilcoxon signed ranks test or the Student paired t test, when appropriate. Case patients and control subjects were examined independently. A probability value of <.05 was considered significant. Intrarater and interrater reliabilities were validated by the intraclass correlation coefficient. An intraclass correlation coefficient value of >0.75 was considered to be good agreement. A 95% CI for the intraclass correlation coefficient can also be estimated.10 The intraclass correlation coefficient for intrarater and interrater reliability of our measurements was 0.81 (95% CI, 0.65-0.97) and 0.77 (95% CI, 0.59-0.96), respectively. Results Patient characteristics. Gestational age at delivery was similar for the IUGR group (32.1 ± 5.1 weeks) and control group (31.1 ± 3.8 weeks). As expected, birth weight was significantly less in the IUGR group than in the control group (1132.9 ± 600.3 g vs 1731.9 ± 864.3 g; P = .002; Tables II and III). Villous morphologic features and vascularization. Intermediate and terminal villi were grouped together for the study of peripheral vascularization. There were significant differences between IUGR and control subjects in the number of capillaries per villous cross-section (P = .012). Similarly, the capillary area per villous stromal
Chen, Bajoria, and Aplin 767
Volume 187, Number 3 Am J Obstet Gynecol
Fig 1. Representative sections of control (A and C) and IUGR (B and D) placentas that were stained with antibody against CD34 to reveal vascular endothelium. IUGR villi showed hypovascularity and increased stroma (B). There was also a decreased number of terminal villi. (Original magnification: A and B, 100; C and D, 250).
cross-section in the IUGR group was significantly lower than that of the control group (P = .008). This contrasted with the two cases of IUGR in which end-diastolic blood flow was preserved; in these cases, highly capillarized terminal villi with sinusoidal dilation were seen in some areas. In cases with reduced capillarization of the peripheral villi, other features included a thin trophoblast covering of the villi, stromal fibrosis, and unusually straight sections of peripheral villus that suggested reduced branching. Indeed, the overall mean number of peripheral villi per microscopic field (at a magnification 100) was significantly lower in the IUGR than in the control subjects. Representative histologic sections are shown in Fig 1, and data are summarized in Fig 2. Cell proliferation. The number of villous profiles that contained proliferating cells was lower in the IUGR group. MIB-1–positive nuclei were scored separately in the trophoblast and stromal compartments. The number of MIB-1–positive cytotrophoblasts was not significantly different between the IUGR and control groups (P = .115), although there were a lower number of MIB-1–positive cytotrophoblasts in the IUGR group (14.7 ± 5.3 vs 20.6 ± 9.3; P = .115). A reduction in the number of stained stromal cells was noted in the IUGR group compared with the control group (4.7 ± 2.4 vs 7.4 ± 2.5; P = .02; Fig 3, A and B, and Fig 4). Smooth muscle α-actin expression. Smooth muscle αactin expression was significantly elevated in the villous stromal cells of IUGR placenta (Fig 3, C and D; Fig 5). No smooth muscle α-actin was detected in trophoblasts.
Fig 2. Villous vascularization (based on CD34 immunohistochemistry) and number of peripheral villi in the placentas of control (C) and IUGR (I) fetuses. a, The number of capillaries per 10,000 square pixels of placental stroma; b, the capillary area as a percentage of placental stroma; c, the number of villi per microscopic field. (Original magnification 100).
Comment A decreased number of peripheral villi with reduced vascularization and cell proliferation in the placentas of the IUGR group was found in this study. Cell proliferation continues actively in both trophoblast and stromal cell compartments of the placenta to term. Placentas from IUGR infants are usually smaller than those from normally grown counterparts,11 and it is apparent from the differences in MIB-1–positive cells that the rates of proliferation are reduced in IUGR in the present cohort, both in stromal and tropohoblastic compartments. Thus, the increased cytotrophoblast proliferation that were reported in some other studies12 and that were associated with perivillous hypoxia5 was not observed. The increased expression of α-smooth muscle actin in villous stroma of IUGR placenta is consistent with Macara et al13 and suggests an increase in myofibroblasts. The cytoskeletal phenotype of this cell population in vivo has been defined by Kohnen et al14 and varies in a continuum from undifferentiated fibroblasts to α-smooth muscle actin–positive and myosin-positive vascular smooth muscle. These α-smooth muscle actin–positive cells are thought to be one kind of specialized extravascular stromal cell,15 with the ability of contraction or relaxation that could compress or widen the extravascular space and enable the placenta to gain some control over maternal intervillous blood flow. Because stromal cell proliferation is decreased, this may suggest an increased differentiation of undifferentiated precursor cells to myofibroblasts or smooth muscle cells.14 Reduced villous branching is suggested by the finding of fewer villous cross-sections per unit placental cross-sectional area. This is consistent with several previous find-
768 Chen, Bajoria, and Aplin
September 2002 Am J Obstet Gynecol
Fig 4. Proliferating cell numbers in trophoblast and stroma of placental villi from control (C) and IUGR (I) fetuses. a, The number of MIB-1-positive nuclei per 10,000 square pixels of placental stroma; b, the number of MIB-1–positive trophoblast nuclei per 10,000 square pixels of placental villous area; c, percentage of villous cross-sections that contained MIB-1 positive nuclei. Fig 3. Representative sections of control (A and C) and IUGR (B and D) placentas stained with antibody against MIB-1 (original magnification: A and B, 250) to localize nuclei of proliferating cells (A, arrow) or against α-smooth muscle actin to localize stromal myofibroblasts. There was a trend to decreased MIB-1 positive cells (B, arrow) in the trophoblast and stromal compartments of peripheral villi of IUGR placentas. There was higher α-smooth muscle actin staining in the villous stroma of IUGR placentas (D) compared with control placentas (C) (original magnification, 100). The inset in (D) shows an area at increased magnitude to demonstrate intracellular actin staining in a proportion of fibroblasts (original magnification, 250).
ings; proportional volume occupancy by intermediate and terminal villi was found to be reduced in placentas from preterm pregnancies that were complicated by IUGR, with abnormal umbilical artery Doppler waveforms, compared with gestational age-matched control subjects.16,17 Smaller mean placental terminal villous profile cross-sectional area and diameter were observed in cases of absent or reversed end-diastolic blood flow than in control subjects at 30 weeks of gestation.18 This type of alteration in terminal villus architecture suggests a developmental cause and carries the implication of a reduced surface area for exchange.8 Placental development, including maturation of intermediate villi and development of new terminal villi, continues during the third trimester.8 However, the design of the present study, which used paired, age-matched placentas in the range of 28 to 41 weeks of gestation, by definition will not reveal developmental compromise that occurs in the time window under examination; in general, the connection between villous branching morphogenesis and cell proliferation remains to be explored. An independent impairment in amino acid transfer efficiency of trophoblast has been reported in IUGR,19 which may be compounded by increased glucose consumption by the
Fig 5. α-Smooth muscle actin in the villous stroma of IUGR (I) placentas compared with control placentas (C). Staining was evaluated with the use of a semiquantitative scoring system.
placenta,20 thus contributing multiply to fetal growth impairment. A further key factor that must contribute to restriction of fetal growth is the very marked reduction of capillary density and area in peripheral villi. Several vascular characteristics have been reported to be altered in peripheral villi from IUGR placentas with abnormal Doppler velocity waveforms.8,21 Capillary loops are sparse in number, significantly longer, less branched, and less coiled than those in control subjects, as observed by scanning electron microscopy of vascular casts. A progressively reduced branching of the stem arteries was seen in placentas with
Chen, Bajoria, and Aplin 769
Volume 187, Number 3 Am J Obstet Gynecol
abnormal umbilical artery end-diastolic blood flow. This phenotype may arise by a failure of branching angiogenesis in development,8,12 and MIB-1 staining was seen rarely in the current study in endothelial or pericapillary cells. However, reduced vascular density in the terminal villi may also arise as a result of vascular regression that occurs as a result of occlusive lesions.22 Our observations suggest that increased fetoplacental vascular impedance is developed, at least in part, at the capillary level. This is consistent with several previous studies13,16; however, others have shown alterations in the number or wall thickness of stem villus vessels.11,23,24 It is possible that both lesions occur in association. In contrast to this picture, in two cases with retained end-diastolic blood flow, highly capillarized terminal villi with sinusoidal dilation were observed. Similarly, Lee and Yeh25 reported many capillary bud projections and numerous anastomoses in the capillary network of the placental villi of small-for-gestational age placentas that were delivered at term. Todros et al21 found peripheral villi in IUGR with positive end-diastolic blood flow to contain numerous capillary cross-sections with multiple branching. They suggested the numerous capillary buds and anastomoses are characteristics of neovascularization and may be a compensatory phenomenon of the capillary network. Thus, IUGR is likely to have heterogeneous causes. REFERENCES
1. Francis ST, Duncan KR, Moore RJ, Baker PN, Johnson IR, Gowland PA. Non-invasive mapping of placental perfusion. Lancet 1998;351:1397-9. 2. Fong KW, Ohlsson A, Hannah ME, Grisaru S, Kingdom J, Cohen H, et al. Prediction of perinatal outcome in fetuses suspected to have intrauterine growth restriction: Doppler US study of fetal cerebral, renal, and umbilical arteries. Radiology 1999;213: 681-9. 3. Karsdorp VHM, van Vugt JMG, van Geijn HP, Kostense PJ, Arduini D, Montenegro N, et al. Clinical significance of absent or reversed end diastolic velocity waveforms in umbilical artery. Lancet 1994;344:1664-8. 4. Newnham JP, O’Dea MRA, Reid KP, Diepeveen DA. Doppler flow velocity waveform analysis in high-risk pregnancies: a randomized controlled trial. Br J Obstet Gynecol 1991;94:929-34. 5. Fox, H. Pathology of the placenta. 2nd ed. London: Saunders; 1997. 6. Forbes JF, Smalls MJ. A comparative analysis of birth weight for gestational age standards. Br J Obstet Gynaecol 1983;99:297-303. 7. Soothill PW, Campbell S, Nicolaides KH. Prediction of morbidity in small and normally grown fetuses by fetal heart rate vari-
8. 9. 10. 11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
ability, biophysical profile score and umbilical artery Doppler studies. Br J Obstet Gynaecol 1993;100:742-5. Benirschke K, Kaufmann P. Pathology of the human placenta. 4th ed. New York: Springer; 2000. p. 42-170. Kaufmann P, Castellucci M. Extravillous trophoblast in the human placenta. Trophoblast Res 1997;10:21-65. Chien PFW, Owen P, Khan KS. Validity of ultrasound estimation of fetal weight. Obstet Gynecol 2000;95:856-60. Kreczy A, Fusi L, Wigglesworth JS. Correlation between umbilical arterial flow and placental morphology. Int J Gynecol Pathol 1995;14:306-9. Ahmed A, Dunk C, Ahmad S, Khaliq A. Regulation of placental vascular endothelial growth factor (VEGF) and placenta growth factor (PIGF) and soluble Flt-1 by oxygen: a review. Placenta 2000;21(Suppl):S16-24. Macara L, Kingdom JCP, Kohnen G, Bowman AW, Greer IA, Kaufmann P. Elaboration of stem villous vessels in growth restricted pregnancies with abnormal umbilical artery Doppler waveforms. Br J Obstet Gynaecol 1995;102:807-12. Kohnen G, Kertschanska S, Demir R, Kaufmann P. Placental villous stroma as a model system for myofibroblast differentiation. Histochem Cell Biol 1996;105:415-29. Kingdom J, Poston L. Stem villous arteries as regulators of fetal placental blood flow: a workshop report. Trophoblast Res 1998;12:403-8. Hitschold T, Weiss E, Beck T, Hunterfering H, Berle P. Low target birth weight or growth retardation? Umbilical Doppler flow velocity waveforms and histometric analysis of fetoplacental vascular tree. Am J Obstet Gynecol 1993;168:1260-4. Hitschold TP. Doppler flow velocity waveforms of the umbilical arteries correlate with intravillous blood volume. Am J Obstet Gynecol 1998;179:540-3. Karsdorp VHM, Dirks BK, van der Linden JC, van Vugt JMG, Baak JPA, van Geijn HP. Placenta morphology and absent or reversed end diastolic flow velocities in the umbilical artery: a clinical and morphometrical study. Placenta 1996;17:393-9. Mahendran D, Donnai P, Glazier JD, D’Souza SW, Boyd RD, Sibley CP. Amino acid (system A) transporter activity in microvillous membrane vesicles from the placentas of appropriate and small for gestational age babies. Pediatric Res 1993;34:661-5. Challis DE, Pfarrer CD, Ritchie JW, Koren G, Adamson SL. Glucose metabolism is elevated and vascular resistance and maternofetal transfer is normal in perfused placental cotyledons from severely growth-restricted fetuses. Pediatr Res 2000;47:309-15. Todros T, Sciarrone A, Piccoli E, Guiot C, Kaufmann P, Kingdom J. Umbilical Doppler waveforms and placental villous angiogenesis in pregnancies complicated by fetal growth restriction. Obstet Gynecol 1999;93:499-503. Salafia C, Pezzullo JC, Minior VK, Divon M. Placental pathology of absent and reversed end-diastolic flow in growth-restricted fetuses. Obstet Gynecol 1997;90:830-6. Giles WB, Trudinger BJ, Baird PJ. Fetal umbilical artery flow velocity waveforms and placental resistance: pathological correlation. Br J Obstet Gynaecol 1985;92:31-8. Bracero LA, Beneck D, Kirshenbaum N, Peiffer M, Stalter P, Schulman H. Doppler velocimetry and placental disease. Am J Obstet Gynecol 1989;161:388-93. Lee MML, Yeh M. Fetal microcirculation of abnormal placenta. Am J Obstet Gynecol 1986;154:1133-9.
Key words: Cell proliferation, fetal growth restriction, vascularization, umbilical artery, velocity waveform
Intrauterine fetal growth restriction (IUGR) is an important cause of perinatal death and morbidity. The precise definition of IUGR remains controversial and is likely to be a heterogeneous condition, but it has been shown in numerous studies that reduced umbilical artery blood flow velocity is associated with IUGR.1 Doppler waveform abnormalities have been shown to be useful in the prediction of poor perinatal outcome in IUGR fetuses.2 Elevated fetoplacental vascular impedance is proposed to be associated with impaired maternal-fetal respiratory gas exchange, which leads to an increased risk of chronic fetal hypoxia and acidosis. Abnormal (decreased or absent) end-diastolic flow velocity in the umbilical artery is associated with a perinatal mortality rate of 41%.3 Subsequent intensive fetal monitoring and timely elective delivery can reduce this risk up to 50%.4 Thus, Doppler ultrasound scanning is an important adjunct for the diagnosis and treatment of IUGR.2,4 The placental features that are associated with the delivery of an IUGR infant From the Division of High Risk Pregnancy, Department of Obstetrics and Gynecology, Mackay Memorial Hospital,a and the Academic Unit of Obstetrics and Gynecology, Medical School, University of Manchester.b Received for publication October 26, 2001; revised March 19, 2002; accepted March 28, 2002. Reprint requests: Chie-Pein Chen, MD, PhD, Flat 1 4F 116 Sec 3 ShinYi Road, Taipei 106, Taiwan. E-mail: [email protected] © 2002, Mosby, Inc. All rights reserved. 0002-9378/2002 $35.00 + 0 6/1/125243 doi:10.1067/mob.2002.125243
764
may give clues to the pathogenesis of the condition. However, studies of IUGR placenta have been hampered by imprecise or overly broad definitions of the study group and problems that involve the recruitment of wellmatched control subjects.5 In the present study, we have selected a well-defined group of severely growth-restricted preterm infants with abnormal end-diastolic blood flow in the umbilical artery and paired them with a normally grown control group of equal gestational age. We characterized cell proliferation, villous branching, villous vascularity, and stromal cell maturation in the peripheral villous placenta. We show dramatically altered placental characteristics that are suggestive of a developmental cause. Material and methods Patients. All pregnancies were singleton and resulted in normally formed infants. All women were healthy and normotensive in early pregnancy. In each case, the gestational age was assigned by an initial ultrasound examination that was performed before 16 weeks of gestation. All the clinical details of these patients were obtained retrospectively from the hospital chart records. Nine pregnancies were identified in which the diagnosis of IUGR was made according to the criteria of fetal birth weight <10th percentile6 with abnormal umbilical artery Doppler waveforms and reduced amniotic fluid volume (amniotic fluid index, ≤5). All the infants were delivered within 1 week
Chen, Bajoria, and Aplin 765
Volume 187, Number 3 Am J Obstet Gynecol
Table I. Primary antibodies Antibody
Clone
Supplier
Form
MIB-1
—
Dianova, Hamburg, Germany
Mouse immunoglobulin G1
CD34 α-Smooth muscle actin
QBEND10 1A4
Serotec, Ltd, Oxford, UK Sigma Chemical Company, St Louis, Mo
Mouse (culture supernatant) Mouse ascites
when abnormal umbilical artery Doppler waveforms were diagnosed. Nine deliveries from spontaneous preterm labor and term pregnancy without any signs and symptoms of chorioamnionitis and with grossly normal placentas that were matched for the closest gestational age were selected as control subjects. Pulsed Doppler ultrasonic recordings from the umbilical artery were made with a 100-Hz filter, and absent or abnormal end-diastolic blood flow velocity was documented from a minimum of five consecutive waveforms in two free-floating loops of cord. Abnormal end-diastolic blood flow was defined as pulsatility index (systolic-diastolic/mean) that was >2 SDs above the mean.7 Tissue. Placenta was obtained with local ethics committee approval. After delivery, the membranes and cord were removed, and the placentas weighed. One 2-cm3 tissue block, from the central area but excluding infarcted areas and peripheral margin of placenta, was randomly excised from each placenta and fixed in 10% buffered formaldehyde overnight at room temperature. Subsequently, the specimens were embedded in paraffin wax. Antibodies. The primary antibodies used for immunohistochemistry are shown in Table I. Immunohistochemistry. Sections 5 µm thick were deparaffinized. For microwave antigen recovery, the slides were put in a loosely covered thermoresistant plastic box (BDH Laboratory Supplies, Poole, UK) with 200 mL of 10 mmol/L citrate buffer (pH 6). They were exposed to microwaves at the medium setting (microwave: maximum power 800 W, with turntable), for 10 to 20 minutes (5 minutes per cycle); thereafter, the sections were allowed to cool for ≥15 minutes. Then, the slides were rinsed rapidly with distilled water at room temperature and washed with 0.05 mol/L TRIS-buffered saline solution (TBS; pH 7.6) three times for 5 minutes each. For immunostaining, the sections were treated with protein block (Dako, Carpinteria, Calif) for 20 minutes and then incubated for 1 hour at room temperature, with primary antibody diluted in phosphate-buffered saline solution. Next, the slides were washed three times for 5 minutes each in TBS and incubated with biotinylated rabbit anti-mouse secondary antibody in TBS (1:100; Dako) for 1 hour at room temperature. After the three 5-minute washes in TBS, the slides were incubated with an avidin-
Pretreatment
Working concentration
Microwave (4 times 1:20 at 5 minutes each) or wet autoclave (5-10 min) No 1:10 No 1:400
biotin-peroxidase complex that was prepared from the Vectorstain ABC Elite kit (Vector Laboratories, Peterborough, UK). Immunoreactivity was visualized with a diaminobenzidine (DAB)/nickel substrate (Vector Laboratories). The sections were then counterstained with 0.25% methyl green for 20 seconds, rinsed in tap water, dehydrated in ascending alcohol (70%-100%), rinsed in xylene, and mounted with XAM neutral medium (BDH). No staining was observed in the absence of primary antibody or in the presence of a variety of control primaries. Morphometric analysis. Each tissue section was covered with a transparent grid consisting of 4 8 6-mm squares, 15 of which were randomly selected and numbered. One microscopic field over the right lower angle of each square in each field was examined with a 10 objective. Therefore, 15 fields in each placenta were examined. The peripheral nonmuscularized intermediate and terminal villi constitute the largest proportion of the placental volume in the third trimester and are the primary sites of gas and nutrient transfer.8 These villi were evaluated together as peripheral villi. The total cross-sectional area of structurally intact villi and the number of peripheral villi in these squares were determined (imaging system KS 400; Kontron, Eching, Germany). The following features were also calculated: area of the villous stroma, the number of capillaries, the area of the capillaries, and MIB-1 proliferation antigen–positive cells per cross-section of villi. The outer trophoblast layer was visualized by methyl green counterstaining. Intravillous vascular elements were recognized by anti-CD34 antibody staining. The number of capillaries within the villous stroma was calculated after segmentation of captured images. Contours of the stroma of the placenta villi and all the included vascular elements were traced manually on the computer monitor with a mouse-controlled cursor at an on-screen magnification of 100 to calculate the areas of these components. MIB-1 recognized a nuclear proliferation-associated antigen, which is expressed in nearly all phases of the cell cycle except G0 and first G1 phase after a G0 stage.9 For each case, MIB-1 positive nuclei were counted separately for the trophoblast and stromal compartments. The sections were examined under a magnification of 250, and the numbers of MIB-1 positive nuclei
766 Chen, Bajoria, and Aplin
September 2002 Am J Obstet Gynecol
Table II. Clinical characteristics of pregnancies that were complicated by IUGR with abnormal umbilical artery Doppler waveforms IUGR
Gestational age (wk)
1 2 3 4 5 6 7 8 9
25 27 28 31 31 35 35 36 41
Birth weight (g)
Percentile
450 570 656 1150 580 1400 1880 1470 2040
<10 <10 <10 <10 <10 <10 <10 <10 <10
Umbilical artery Doppler waveform Absent end-diastolic volume Absent end-diastolic volume Reversed end-diastolic volume Absent end-diastolic volume Absent end-diastolic volume Absent end-diastolic volume Abnormal end-diastolic volume* Abnormal end-diastolic volume* Absent end-diastolic volume
*Pulsatility index, ≥2 SD.
Table III. Clinical characteristics of control pregnancies Control 1 2 3 4 5 6 7 8 9
Gestational age (wk)
Birth weight (g)
Percentile
26 27 28 30 32 32 33 34 38
613 980 1100 1560 1520 1654 2240 2510 3410
10-90 10-90 10-90 10-90 10-90 10-90 10-90 10-90 10-90
were related to the total villous cross-sectional area of peripheral villi. MIB-1–positive villous stromal cells per 10,000 square pixels total villous stromal area were assessed. The discrimination between relatively small numbers of cytotrophoblast nuclei and the much larger numbers of syncytiotrophoblast nuclei is difficult, but for the analysis of MIB-1 staining it was assumed that only the cytotrophoblast cells proliferate.8 For the α-smooth muscle actin staining, sections from different patients were processed in one run for the same monoclonal antibody, and the time incubated with DAB substrate was kept the same for all the slides. A scoring system for semiquantitative evaluation of the tissue sections was used in the following manner: 6, the strongest intensity with homogenous marked staining in all part of section; 5, >70% of the section showed marked staining, with the other part of the section revealing moderate staining; 4, the stain intensity is moderate and homogenous in all part of section; 3, >70% of the section showed moderate staining, the other part of the section revealing weak staining; 2, the stain intensity is weak and homogenous in all parts of section; and 1, >70% of the section showed weak staining, with the other parts of the section being revealed as very weakly stained or negative. All stainings were evaluated by two observers who were blind to the clinical diagnosis. Statistical analysis. All parameters of these two groups were described as mean (±SD) and were examined
Umbilical artery Doppler waveform Normal Normal Normal Normal Normal Normal Normal Normal Normal
whether the distribution of these parameters was parametric or nonparametric by Kolmogorov-Smirnov onesample test. The differences were assessed by Wilcoxon signed ranks test or the Student paired t test, when appropriate. Case patients and control subjects were examined independently. A probability value of <.05 was considered significant. Intrarater and interrater reliabilities were validated by the intraclass correlation coefficient. An intraclass correlation coefficient value of >0.75 was considered to be good agreement. A 95% CI for the intraclass correlation coefficient can also be estimated.10 The intraclass correlation coefficient for intrarater and interrater reliability of our measurements was 0.81 (95% CI, 0.65-0.97) and 0.77 (95% CI, 0.59-0.96), respectively. Results Patient characteristics. Gestational age at delivery was similar for the IUGR group (32.1 ± 5.1 weeks) and control group (31.1 ± 3.8 weeks). As expected, birth weight was significantly less in the IUGR group than in the control group (1132.9 ± 600.3 g vs 1731.9 ± 864.3 g; P = .002; Tables II and III). Villous morphologic features and vascularization. Intermediate and terminal villi were grouped together for the study of peripheral vascularization. There were significant differences between IUGR and control subjects in the number of capillaries per villous cross-section (P = .012). Similarly, the capillary area per villous stromal
Chen, Bajoria, and Aplin 767
Volume 187, Number 3 Am J Obstet Gynecol
Fig 1. Representative sections of control (A and C) and IUGR (B and D) placentas that were stained with antibody against CD34 to reveal vascular endothelium. IUGR villi showed hypovascularity and increased stroma (B). There was also a decreased number of terminal villi. (Original magnification: A and B, 100; C and D, 250).
cross-section in the IUGR group was significantly lower than that of the control group (P = .008). This contrasted with the two cases of IUGR in which end-diastolic blood flow was preserved; in these cases, highly capillarized terminal villi with sinusoidal dilation were seen in some areas. In cases with reduced capillarization of the peripheral villi, other features included a thin trophoblast covering of the villi, stromal fibrosis, and unusually straight sections of peripheral villus that suggested reduced branching. Indeed, the overall mean number of peripheral villi per microscopic field (at a magnification 100) was significantly lower in the IUGR than in the control subjects. Representative histologic sections are shown in Fig 1, and data are summarized in Fig 2. Cell proliferation. The number of villous profiles that contained proliferating cells was lower in the IUGR group. MIB-1–positive nuclei were scored separately in the trophoblast and stromal compartments. The number of MIB-1–positive cytotrophoblasts was not significantly different between the IUGR and control groups (P = .115), although there were a lower number of MIB-1–positive cytotrophoblasts in the IUGR group (14.7 ± 5.3 vs 20.6 ± 9.3; P = .115). A reduction in the number of stained stromal cells was noted in the IUGR group compared with the control group (4.7 ± 2.4 vs 7.4 ± 2.5; P = .02; Fig 3, A and B, and Fig 4). Smooth muscle α-actin expression. Smooth muscle αactin expression was significantly elevated in the villous stromal cells of IUGR placenta (Fig 3, C and D; Fig 5). No smooth muscle α-actin was detected in trophoblasts.
Fig 2. Villous vascularization (based on CD34 immunohistochemistry) and number of peripheral villi in the placentas of control (C) and IUGR (I) fetuses. a, The number of capillaries per 10,000 square pixels of placental stroma; b, the capillary area as a percentage of placental stroma; c, the number of villi per microscopic field. (Original magnification 100).
Comment A decreased number of peripheral villi with reduced vascularization and cell proliferation in the placentas of the IUGR group was found in this study. Cell proliferation continues actively in both trophoblast and stromal cell compartments of the placenta to term. Placentas from IUGR infants are usually smaller than those from normally grown counterparts,11 and it is apparent from the differences in MIB-1–positive cells that the rates of proliferation are reduced in IUGR in the present cohort, both in stromal and tropohoblastic compartments. Thus, the increased cytotrophoblast proliferation that were reported in some other studies12 and that were associated with perivillous hypoxia5 was not observed. The increased expression of α-smooth muscle actin in villous stroma of IUGR placenta is consistent with Macara et al13 and suggests an increase in myofibroblasts. The cytoskeletal phenotype of this cell population in vivo has been defined by Kohnen et al14 and varies in a continuum from undifferentiated fibroblasts to α-smooth muscle actin–positive and myosin-positive vascular smooth muscle. These α-smooth muscle actin–positive cells are thought to be one kind of specialized extravascular stromal cell,15 with the ability of contraction or relaxation that could compress or widen the extravascular space and enable the placenta to gain some control over maternal intervillous blood flow. Because stromal cell proliferation is decreased, this may suggest an increased differentiation of undifferentiated precursor cells to myofibroblasts or smooth muscle cells.14 Reduced villous branching is suggested by the finding of fewer villous cross-sections per unit placental cross-sectional area. This is consistent with several previous find-
768 Chen, Bajoria, and Aplin
September 2002 Am J Obstet Gynecol
Fig 4. Proliferating cell numbers in trophoblast and stroma of placental villi from control (C) and IUGR (I) fetuses. a, The number of MIB-1-positive nuclei per 10,000 square pixels of placental stroma; b, the number of MIB-1–positive trophoblast nuclei per 10,000 square pixels of placental villous area; c, percentage of villous cross-sections that contained MIB-1 positive nuclei. Fig 3. Representative sections of control (A and C) and IUGR (B and D) placentas stained with antibody against MIB-1 (original magnification: A and B, 250) to localize nuclei of proliferating cells (A, arrow) or against α-smooth muscle actin to localize stromal myofibroblasts. There was a trend to decreased MIB-1 positive cells (B, arrow) in the trophoblast and stromal compartments of peripheral villi of IUGR placentas. There was higher α-smooth muscle actin staining in the villous stroma of IUGR placentas (D) compared with control placentas (C) (original magnification, 100). The inset in (D) shows an area at increased magnitude to demonstrate intracellular actin staining in a proportion of fibroblasts (original magnification, 250).
ings; proportional volume occupancy by intermediate and terminal villi was found to be reduced in placentas from preterm pregnancies that were complicated by IUGR, with abnormal umbilical artery Doppler waveforms, compared with gestational age-matched control subjects.16,17 Smaller mean placental terminal villous profile cross-sectional area and diameter were observed in cases of absent or reversed end-diastolic blood flow than in control subjects at 30 weeks of gestation.18 This type of alteration in terminal villus architecture suggests a developmental cause and carries the implication of a reduced surface area for exchange.8 Placental development, including maturation of intermediate villi and development of new terminal villi, continues during the third trimester.8 However, the design of the present study, which used paired, age-matched placentas in the range of 28 to 41 weeks of gestation, by definition will not reveal developmental compromise that occurs in the time window under examination; in general, the connection between villous branching morphogenesis and cell proliferation remains to be explored. An independent impairment in amino acid transfer efficiency of trophoblast has been reported in IUGR,19 which may be compounded by increased glucose consumption by the
Fig 5. α-Smooth muscle actin in the villous stroma of IUGR (I) placentas compared with control placentas (C). Staining was evaluated with the use of a semiquantitative scoring system.
placenta,20 thus contributing multiply to fetal growth impairment. A further key factor that must contribute to restriction of fetal growth is the very marked reduction of capillary density and area in peripheral villi. Several vascular characteristics have been reported to be altered in peripheral villi from IUGR placentas with abnormal Doppler velocity waveforms.8,21 Capillary loops are sparse in number, significantly longer, less branched, and less coiled than those in control subjects, as observed by scanning electron microscopy of vascular casts. A progressively reduced branching of the stem arteries was seen in placentas with
Chen, Bajoria, and Aplin 769
Volume 187, Number 3 Am J Obstet Gynecol
abnormal umbilical artery end-diastolic blood flow. This phenotype may arise by a failure of branching angiogenesis in development,8,12 and MIB-1 staining was seen rarely in the current study in endothelial or pericapillary cells. However, reduced vascular density in the terminal villi may also arise as a result of vascular regression that occurs as a result of occlusive lesions.22 Our observations suggest that increased fetoplacental vascular impedance is developed, at least in part, at the capillary level. This is consistent with several previous studies13,16; however, others have shown alterations in the number or wall thickness of stem villus vessels.11,23,24 It is possible that both lesions occur in association. In contrast to this picture, in two cases with retained end-diastolic blood flow, highly capillarized terminal villi with sinusoidal dilation were observed. Similarly, Lee and Yeh25 reported many capillary bud projections and numerous anastomoses in the capillary network of the placental villi of small-for-gestational age placentas that were delivered at term. Todros et al21 found peripheral villi in IUGR with positive end-diastolic blood flow to contain numerous capillary cross-sections with multiple branching. They suggested the numerous capillary buds and anastomoses are characteristics of neovascularization and may be a compensatory phenomenon of the capillary network. Thus, IUGR is likely to have heterogeneous causes. REFERENCES
1. Francis ST, Duncan KR, Moore RJ, Baker PN, Johnson IR, Gowland PA. Non-invasive mapping of placental perfusion. Lancet 1998;351:1397-9. 2. Fong KW, Ohlsson A, Hannah ME, Grisaru S, Kingdom J, Cohen H, et al. Prediction of perinatal outcome in fetuses suspected to have intrauterine growth restriction: Doppler US study of fetal cerebral, renal, and umbilical arteries. Radiology 1999;213: 681-9. 3. Karsdorp VHM, van Vugt JMG, van Geijn HP, Kostense PJ, Arduini D, Montenegro N, et al. Clinical significance of absent or reversed end diastolic velocity waveforms in umbilical artery. Lancet 1994;344:1664-8. 4. Newnham JP, O’Dea MRA, Reid KP, Diepeveen DA. Doppler flow velocity waveform analysis in high-risk pregnancies: a randomized controlled trial. Br J Obstet Gynecol 1991;94:929-34. 5. Fox, H. Pathology of the placenta. 2nd ed. London: Saunders; 1997. 6. Forbes JF, Smalls MJ. A comparative analysis of birth weight for gestational age standards. Br J Obstet Gynaecol 1983;99:297-303. 7. Soothill PW, Campbell S, Nicolaides KH. Prediction of morbidity in small and normally grown fetuses by fetal heart rate vari-
8. 9. 10. 11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
ability, biophysical profile score and umbilical artery Doppler studies. Br J Obstet Gynaecol 1993;100:742-5. Benirschke K, Kaufmann P. Pathology of the human placenta. 4th ed. New York: Springer; 2000. p. 42-170. Kaufmann P, Castellucci M. Extravillous trophoblast in the human placenta. Trophoblast Res 1997;10:21-65. Chien PFW, Owen P, Khan KS. Validity of ultrasound estimation of fetal weight. Obstet Gynecol 2000;95:856-60. Kreczy A, Fusi L, Wigglesworth JS. Correlation between umbilical arterial flow and placental morphology. Int J Gynecol Pathol 1995;14:306-9. Ahmed A, Dunk C, Ahmad S, Khaliq A. Regulation of placental vascular endothelial growth factor (VEGF) and placenta growth factor (PIGF) and soluble Flt-1 by oxygen: a review. Placenta 2000;21(Suppl):S16-24. Macara L, Kingdom JCP, Kohnen G, Bowman AW, Greer IA, Kaufmann P. Elaboration of stem villous vessels in growth restricted pregnancies with abnormal umbilical artery Doppler waveforms. Br J Obstet Gynaecol 1995;102:807-12. Kohnen G, Kertschanska S, Demir R, Kaufmann P. Placental villous stroma as a model system for myofibroblast differentiation. Histochem Cell Biol 1996;105:415-29. Kingdom J, Poston L. Stem villous arteries as regulators of fetal placental blood flow: a workshop report. Trophoblast Res 1998;12:403-8. Hitschold T, Weiss E, Beck T, Hunterfering H, Berle P. Low target birth weight or growth retardation? Umbilical Doppler flow velocity waveforms and histometric analysis of fetoplacental vascular tree. Am J Obstet Gynecol 1993;168:1260-4. Hitschold TP. Doppler flow velocity waveforms of the umbilical arteries correlate with intravillous blood volume. Am J Obstet Gynecol 1998;179:540-3. Karsdorp VHM, Dirks BK, van der Linden JC, van Vugt JMG, Baak JPA, van Geijn HP. Placenta morphology and absent or reversed end diastolic flow velocities in the umbilical artery: a clinical and morphometrical study. Placenta 1996;17:393-9. Mahendran D, Donnai P, Glazier JD, D’Souza SW, Boyd RD, Sibley CP. Amino acid (system A) transporter activity in microvillous membrane vesicles from the placentas of appropriate and small for gestational age babies. Pediatric Res 1993;34:661-5. Challis DE, Pfarrer CD, Ritchie JW, Koren G, Adamson SL. Glucose metabolism is elevated and vascular resistance and maternofetal transfer is normal in perfused placental cotyledons from severely growth-restricted fetuses. Pediatr Res 2000;47:309-15. Todros T, Sciarrone A, Piccoli E, Guiot C, Kaufmann P, Kingdom J. Umbilical Doppler waveforms and placental villous angiogenesis in pregnancies complicated by fetal growth restriction. Obstet Gynecol 1999;93:499-503. Salafia C, Pezzullo JC, Minior VK, Divon M. Placental pathology of absent and reversed end-diastolic flow in growth-restricted fetuses. Obstet Gynecol 1997;90:830-6. Giles WB, Trudinger BJ, Baird PJ. Fetal umbilical artery flow velocity waveforms and placental resistance: pathological correlation. Br J Obstet Gynaecol 1985;92:31-8. Bracero LA, Beneck D, Kirshenbaum N, Peiffer M, Stalter P, Schulman H. Doppler velocimetry and placental disease. Am J Obstet Gynecol 1989;161:388-93. Lee MML, Yeh M. Fetal microcirculation of abnormal placenta. Am J Obstet Gynecol 1986;154:1133-9.