Hematologic Effects of Placental Pathology on Very Low Birthweight Infants Born to Mothers with Preeclampsia

Home , Placental infarction

Journal of Perinatology (2009) 29,8–12 r 2009 Nature Publishing Group All rights reserved. 0743-8346/09 $32 www.nature.com/jp ORIGINAL ARTICLE Hematologic effects of placental pathology on very low birthweight infants born to mothers with preeclampsia

KJ Zook1,2, AB Mackley1, J Kern1 and DA Paul1,2 1Department of Pediatrics, Section of Neonatology, Christiana Care Health System, Newark, DE, USA and 2Department of Pediatrics, Thomas Jefferson University, Philadelphia, PA, USA

Introduction Objective: To investigate the effect of placental pathology on neonatal Preeclampsia is a common disorder of pregnancy affecting nearly 5 neutrophils, platelets, hematocrit and nucleated red blood cells in very low to 8% of pregnant women annually in the United States. The NIH birthweight (VLBW) infants born to mothers with preeclampsia. reported in 2000 that preeclampsia had risen by one-third over the Study Design: Retrospective cohort study of infants with birthweight previous decade.1 Preeclampsia is known to be a common cause of < 1500 g born to mothers with preeclampsia from july, 2002 to july, 2006 neonatal neutropenia and thrombocytopenia. It has previously at a single level III neonatal intensive care unit. Placental pathology been postulated that both neonatal neutropenia and was reviewed for the presence of placental infarction and vasculopathy. thrombocytopenia result from decreased production of WBCs and Hematologic parameters from day of life 0, 1 and 2 were obtained. platelets (Plt) from placental insufficiency.2–9 Statistical analysis included repeated-measures analysis of variance Preeclampsia is also known to have associated placental and multivariable analysis using logistic regression. pathology.10 Different placental histopathologies are associated with Result: The study sample included 203 infants with estimated specific clinical features of preterm preeclampsia. Severe maternal gestational age of 28±3 weeks; 45% had placental infarctions and 26% proteinuria is related to placental chronic inflammation, whereas placental vasculopathy. Infants with neutropenia and thrombocytopenia lower maternal antepartum Plt counts are related to placental did not have an increased occurrence of placental infarction or maternal abruption and infarction. Lower birthweight percentiles and lighter vasculopathy but were more likely to be of small gestational age (SGA) placentas are directly associated with uteroplacental vascular 11 and of lower gestational age compared with infants without neutropenia lesions. or thrombocytopenia. After multivariable analysis, gestational age and To date, the link between different placental pathological SGA remained associated with both neutropenia and thrombocytopenia conditions associated with preeclampsia, neonatal neutropenia and whereas placental infarction and vasculopathy did not remain in the thrombocytopenia have not been explored. In this study, we models. hypothesized that there would be specific placental findings associated with neonatal hematologic abnormalities in infants born Conclusion: In our population of VLBW infants born to mothers with to mothers with preeclampsia. preeclampsia, placental pathology was common. There was no association The purpose of this study was to investigate the association of placental infarction or vasculopathy with neonatal neutropenia and between placental pathology and neonatal absolute neutrophil thrombocytopenia. The data suggest that neonatal hematologic effects of count (ANC), Plt, hematocrit (Hct) and nucleated red blood cells maternal preeclampsia, if related to the placenta, are associated with (NRBC) in very low birthweight (VLBW) infants born to mothers factors other than placental histology. with preeclampsia. Journal of Perinatology (2009) 29, 8–12; doi:10.1038/jp.2008.104; published online 18 December 2008

Keywords: placenta; placental insufficiency; neonatal thrombocytopenia and neutropenia Methods The study was a retrospective nested-cohort study of VLBW infants (<1500 gm) born to mothers with preeclampsia at Christiana Care Correspondence: Dr KJ Zook, Christiana Care Health System, Department of Neonatology, Health System. Christiana Care Health System is the single level III 4755 Ogletown-Stanton Road, Newark, DE 19718, USA. regional neonatal intensive care unit serving the state of Delaware. E-mail: [email protected] Received 4 January 2008; revised 28 April 2008; accepted 15 May 2008; published online The study sample included inborn infants cared for between July 18 December 2008 2002 and July 2006. The study sample included only infants born Placental effects in preeclampsia KJ Zook et al 9 to mothers with preeclampsia. The Institutional Board Review at regression. A P-value < 0.05 was considered significant; all data Christiana Care Health Services approved this project. are expressed as mean±s.d. The diagnosis of preeclampsia was made by the attending obstetrician based on ACOG guidelines of elevated blood pressure (>140 systolic or >90 diastolic) and proteinuria (>0.3gm of Results protein in a 24 h urine collection) after 20 weeks of gestation.12 For During the study period, 211 inborn infants were born to mothers the purposes of this study, estimated gestational age was defined by with preeclampsia. The final study sample included 203 infants; best obstetric estimate. Small for gestational age (SGA) was defined 8 infants born to mothers with preeclampsia were excluded, as as birthweight of <10%. Data on prenatal steroid administration they did not have placental pathology results. In the study sample, were collected and mothers were classified as receiving antenatal 30 infants were diagnosed with neutropenia and 48 infants with steroids if they received a minimum of 1 dose of betamethasone or thrombocytopenia during the first 48 h of life. The distribution dexamethasone before the delivery. of the placental pathologies included infarction (45%), maternal Clinical pathology reports were obtained on all infants. The vasculopathy (26%), meconium staining (16%), histologic decision to send the placenta for pathologic evaluation was made chorioamnionitis (13%), calcification (9%) and funisitis (3%). by the attending obstetrician. The placentas were examined by Because of their prevalence, placental infarction, maternal one of nine attending pathologists at Christiana Care Health vasculopathy and placental calcifications were chosen to compare System who were blinded to the clinical and hematologic outcomes with the presence or absence of neutropenia and of the infants. Christiana Care Health System evaluates thrombocytopenia, total NRBCs and Hct values. Although more approximately 1500 placentas per year of the 7000 live births. The prevalent than placental calcifications, histologic chorioamnionitis placentas were systematically reviewed according to the Placental and meconium staining were not considered in relationship to Pathology Practice Guideline Development Task Force.13 These neonatal parameters as they are common comorbid factors for guidelines attempt to create uniformity for analysis of the placenta, many other conditions. which is a relatively subjective process. The reports were reviewed Infants with neutropenia were of lower gestation, lower for placental abnormalities using standard definitions for birthweight and more likely to be SGA compared with infants commonly described placental lesions.14 The guidelines encourage without neutropenia (Table 1). There were no differences in the the reporting of the most significant lesions such as presence occurrence of placental infarction, placental vasculopathy or of chorioamnionitis, villitis, central infarcts (rather than placental weight in infants with neutropenia compared with those marginal infarcts) and vasculopathy and discourage reporting without neutropenia. After multivariable analysis to control for minor findings. Those born to mothers with preeclampsia who potential confounding variables including gestational age, did not have a placental pathology report were excluded from birthweight, race, SGA, multiple gestation, use of prenatal steroids this study. and placental pathology, only gestational age (odds ratio of 1.3, Infants in the study sample had complete blood counts drawn 95% CI 1.1 to 1.6) and SGA (odds ratio of 4.2 (CI 1.3 to 13.5) as per neonatal intensive care unit protocol. This included a complete blood counts on admission and the first two days of life. Complete blood counts were obtained more frequently as clinically Table 1 Univariate analysis of neutropenia and placental pathologies, EGA, BW, indicated. Hematologic parameters investigated included ANC, Plt Race, SGA, multiple gestation, prenatal steroids count, Hct and total NRBC’s. All complete blood counts were Neutropenia Neutropenia P-value analyzed in standard fashion using a Sysmex Model Xe2100 present absent, Coulter Counter (Hialiah, FL, USA). Total NRBCs were calculated (n ¼ 30) (n ¼ 173) by multiplying the total WBC count by the number of NRBC ± ± counted per 100 WBC. EGA (weeks) 27.9 2.1 29.3 2.4 0.02 ± ± Thrombocytopenia was defined as Plt count less than 100 K/ BW (grams) 852 283 1091 267 <0.01 Placental infarction 33% 46% 0.2 mm3.15 Neutropenia was defined by the Mouzhino criteria: ANC Placental calcification 10% 9% 0.85 less than 500 on day of life 0, ANC less than 2000 on day of life 1, Maternal vasculopathy 33% 25% 0.41 16 and ANC less than 1100 on day of life 2. Placental weight (grams) 217±37 245±85 0.17 Race (Caucasian/African American) 31%/68% 50%/40% 0.02 Statistics SGA 37% 15% 0.01 Statistical analysis included both univariable and multivariable Multiple gestation 17% 16% 0.93 analysis. Univariable analysis included w2-test, MannFWhitney Antenatal steroids 83% 82% 0.9 test, U-test, ANOVA (analysis of variance), and repeated measures Abbreviations: BW, birth weight; EGA, estimated gestational age; SGA, Small for ANOVA as appropriate. Multivariable analysis includes logistic gestational age.

Journal of Perinatology Placental effects in preeclampsia KJ Zook et al 10

Table 2 Univariate analysis of thrombocytopenia and placental pathologies, Table 3 NRBC and placental infarction, calcification, and vasculopathy EGA, BW, Race, SGA, multiple gestation, prenatal steroids Admission Day 1 Day 2 P-value Thrombocytopenia Thrombocytopenia P-value present absent Placental infarction 5.6 (4.3–6.9) 6.0 (3.6–8.3) 3.5 (1.3–4.7) 0.21 (n ¼ 48) (n ¼ 155) (n ¼ 92) No placental infarction 3.2 (2.0–4.3) 2.6 (0.5–4.7) 2.1 (1.1–3.8) EGA 28.3±3 29.2±2.3 0.04 Placental calcification 3.5 (0.7–6.4) 2.3 (0–7.4) 1.6 (0–4.2) 0.81 BW 882±327 1073±253 <.01 (n ¼ 18) Placental Infarction 50% 44% 0.47 No calcification 4.3 (3.3–5.2) 4.3 (2.6–5.9) 2.9 (2.0–3.7) Placental Calcification 10% 8% 0.8 Placental vasculopathy 4.3 (2.4–6.2) 4.5 (1.1–8.0) 3.2 (1.6–4.9) 0.94 Maternal Vasculopathy 25% 28% 0.73 (n ¼ 46) Placental Weight (grams) 231±130 223±70 0.87 No vasculopathy 4.1 (3.0–5.3) 4.0 (1.9–6.1) 2.6 (1.6–3.6) Race (Caucasian/African 37/50% 54/41% 0.06 Data shows total number of NRBC/mm3. All data are presented as mean with 95% CI. American) P-value represents repeated measures ANOVA. SGA 43% 17% <.01 Abbreviations: ANOVA, analysis of variance; NRBC, nucleated red blood cells. Multiple Gestation 14% 18% 0.53 Steroids 68% 69% 0.84

Abbreviations: BW, birth weight; EGA, estimated gestational age; SGA, Small for gestational age. neutropenia or thrombocytopenia. After controlling for potential confounding variables such as lower birthweight, lower gestational age and/or SGA remained associated with increased odds of remained associated with neutropenia. Neither placental weight nor neonatal thrombocytopenia and neutropenia. the presence of placental infarction, calcification or vasculopathy Preeclampsia continues to be an increasing problem associated was associated with neonatal neutropenia. with preterm birth and is known to be a common cause of Similarly, infants with thrombocytopenia were also found to be neonatal neutropenia and thrombocytopenia.1 The incidence of of lower gestational age, birthweight and more likely to be SGA. neutropenia is reported as high as 49% in pregnancy-induced Again, there were no differences found in the occurrence of hypertension exposed infants.2 Consistent with our findings, placental infarction, placental vasculopathy, calcification or neutropenia severity has been correlated with the severity of growth placental weight when compared with infants without restriction and prematurity.4 Kinetic investigations have suggested thrombocytopenia (Table 2). After multivariable analysis to control that neutropenia in infants born to mothers with preeclampsia is for potential confounding variables including gestational age, secondary to decreased production of WBCs as evidenced by birthweight, SGA, multiple gestation, use of prenatal steroids and decreased circulating, marginated, storage, progenitor and placental pathology, only gestational age (OR 1.2, 95% CI 1.0 to proliferative neutrophils.2 Our data suggest that the decreased 1.4) and SGA (OR 4.2, CI 1.6 to 11) were associated with an production of WBCs is not associated with histologic changes in the increased odds of thrombocytopenia. Neither placental weight nor placenta including placental infarction or vasculopathy. the presence of placental infarction, calcification, or vasculopathy Early onset of neonatal thrombocytopenia is most commonly was associated with neonatal neutropenia. observed secondary to placental insufficiency with maternal Thirty-eight percent of the infants were both neutropenic and hypertension being the largest risk factor.17 Burrows and colleagues thrombocytopenic. Of the infants without neutropenia, 20% were found that in hypertensive mothers, preterm birth was the major thrombocytopenic (P ¼ 0.3). In the infants with both neutropenia risk factor for neonatal thrombocytopenia. Term infants exposed to and thrombocytopenia 27% had placental infarctions, 18% had maternal hypertension had the same incidence of placental calcifications and 33% had placental vasculopathies. thrombocytopenia as the control population.5 This finding suggests No association was found between the presence of placental that hypertension effects may be of a different pathophysiologic infarction, placental calcification or placental vasculopathies and pathway in term versus preterm infants. This observation is Hct or NRBCs in the first 48 h of life (Tables 3 and 4). consistent with our finding of thrombocytopenia associated with lower gestational age in infants born to mothers with preeclampsia. Preelcampsia is known to have associated placental pathology. Discussion Redline et al.18 evaluated 609 placentas and found most placental In our population of VLBW infants born to mothers with lesions before 37-week gestation, specifically maternal preeclampsia, placental pathology was common. Despite the vasculopathy, villitis and increased intervillous fibrin were common occurrence of placental infarction and maternal increased in pregnancies complicated by hypertension. Soma et al. vasculopathy, these findings were not associated with neonatal described that there are many microscopic findings common in the

Journal of Perinatology Placental effects in preeclampsia KJ Zook et al 11

Table 4 Hematocrit and placental infarction, calciﬁcation, and vasculopathy

Admission Day 1 Day 2 P-value

Placental infarction (n ¼ 92) 47.6 (45.8–49.3) 48.5 (46.5–50.4) 46.6 (44.7–48.5) 0.22 No placental infarction 45.8 (44.3–47.3) 45.3 (43.6–47.0) 43.6 (41.9–45.2) Placental calciﬁcation (n ¼ 18) 44.9 (45.8–49.3) 45.6 (41.2–50.0) 44.0 (39.7–48.0) 0.89 No calciﬁcation 46.8 (45.6–47.9) 46.8 (45.4–48.2) 45.0 (43.7–46.4) Placental vasculopathy (n ¼ 46) 48.7 (46.4–51.0) 49.4 (46.8- 52.0) 47.0 (44.6–47.7) 0.67 No vasculopathy 46.4 (44.7–47.5) 46.2 (44.6–47.7) 44.7 (43.1–46.2)

Data shows % Hematocrit. All data are presented as mean with 95% CI. P-value represents repeated measures ANOVA. Abbreviation: ANOVA, analysis of variance.

placentas of hypertensive patients. These lesions included increased Practice Guideline Development Task Force.13 Findings such as syncytial knots, villi hypovascularity, cytotrophoblastic proliferation, increased amounts of syncytial knots and necrosis, thickened with thickened basement membranes, atherosis of the spiral trophoblastic basement membranes and narrowed fetal capillaries arteries and narrowed fetal capillaries.10 Different placental as previously discussed findings in preeclamptic placentas tend to histopathologies have been associated with specific clinical features be subtle variations that are not typically commented upon in a of preterm preeclampsia. Severe maternal proteinuria has been routine placental pathology analysis. In addition, some frequent related to placental chronic inflammation, whereas lower maternal findings in preeclampsia, such as meconium staining may be too antepartum Plt counts were related to placental abruption and nonspecific to draw reliable conclusions from. It may be of benefit infarction. In addition, lower birthweight percentiles and lighter to form a prospective study designed to have evaluations for specific placentas were directly associated with uteroplacental vascular placental lesions and then analyze whether these lesions have lesions.19 Ghidini’s 11prospective study used a scoring system based direct hematologic correlation. on severity and extent of placental vascular and villous lesions and It is plausible that the placental pathologies we investigated found that the clinical diagnosis of preeclampsia increased with were not necessarily associated with physiologic placental progressive uteroplacental impairment. The lesions defined in insufficiency. Perhaps, other placental factors may be the causative Ghidini’s study were analyzed for more specific lesions than is agents of neonatal hematologic aberrations. Histologic typically found in a standard placental pathology report and could chorioamnionitis, common in preterm birth, has been associated not be applied to this study. Kush et al.20 evaluated the relationship with increased levels of cord blood cytokines and increase between umbilical artery end diastolic velocity in growth-restricted intracranial hemorrhage in neonates.22,23 Inflammation may have fetuses and associated hematologic alterations that were a contributing role and warrant further investigation with regard to proportional to the degree of placental dysfunction, which was able preeclampsia and neonatal neutropenia and thrombocytopenia. to be predicted by fetal Doppler measurements. Absent end diastolic The number of infants born to mothers with histologic umbilical artery velocity was also associated with increased NRBC chorioamnionitis in our study cohort was too small to make a counts.21 meaningful association. Furthermore, histologic chorioamnionitis We were unable to show any association between placental may have been associated with other comorbidities such as pathology and neonatal hematologic abnormalities. There are a premature labor. Other limitations of our study may include the number of potential explanations for this finding and our data possibility of inter-observer variance and lack of consensus on the should be interpreted with caution. There may be no causal effect diagnosis of placental lesions. However, adhering to methodical of placental pathology on neonatal neutropenia or practice guidelines at a single institution and standard definitions thrombocytopenia. Alternatively, placental pathology may still be likely limited this variablity. associated with the hematologic effects of maternal preeclampsia; In conclusion, in our population of VLBW infants born to however, the placental pathologies leading to neutropenia or mothers with preeclampsia, placental pathology was common. thrombocytopenia may be different than those selected in this There was, however, no association of placental infarction or study. As our investigation was retrospective, we may have missed vasculopathy with neonatal neutropenia, thrombocytopenia some frequent subtle placental pathologies, which may not have or other hematologic parameters. Our data suggest that the been routinely reported on clinical pathology reports. Placental neonatal hematologic effects of maternal preeclampsia, if related analysis is relatively subjective with less than 5% of the tissue to the placenta, are associated with factors other than placental examined histologically according to the Placental Pathology histology.

Journal of Perinatology Placental effects in preeclampsia KJ Zook et al 12

References 13 Langston C. Practice guideline for examination of the placenta. Arch Pathol Lab Med 1997; 121: 449–476. 1 http://www.nhlbi.nih.gov/health/prof/heart/hbp_preg.htm. 14 Kurman R. Blaustein’s Pathology of the Female Genital Tract 5th edn. Springer: New 2 Koenig JM, Christensen RD. Incidence, neutrophil kinetics, and natural history of York, 2002, pp 1103–1192. neonatal neutropenia associated with maternal hypertension. N Engl J Med 1989; 321: 15 Sola M, Del Veechio A, Rimsza L. Evaluation and treatment of thrombocytopenia in the 557–562. neonatal intensive care unit. Clin Perinatol 2000; 27: 655–679. 3 Koenig JM, Christensen RD. The mechanism responsible for diminished neutrophil 16 Mouzinho A, Rosenfeld CR, Sanchez PJ, Risser R. Revised reference ranges for production in neonates delivered of women with pregnancy-induced hypertension. Am circulating neutrophils in very-low-birthweight neonates. Pediatrics 1994; : 76–82. J Obstet Gynecol 1991; 165: 467–473. 94 Arch Dis 4 Christensen RD, Henry E, Wiedmeier SE, Stoddard RA, Lambert DK. Low blood 17 Roberts I, Murray NA. Neonatal Thrombocytopenia: causes and management. neutrophil concentrations among extremely low birthweight neonates: data from a Child Fetal Neonatal Ed 2003; 88: F359–F364. multihospital health-care system. Journ of Perinatol 2006; 26: 682–687. 18 Redline RW, Patterson P. Patterns of placental injury. Correlations with gestational 5 Burrows RF, Andrew M. Neonatal thrombocytopenia in the hypertensive disorders of age, placental weight and clinical diagnoses. Arch Pathol Lab Med 1994; 118: pregnancy. Obstet Gynecol 1990; 76(2): 234–238. 698–701. 6 Murray NA, Roberts I. Circulating Megakaryocytes and Their Progenitors in Early 19 Salafia CM, Pezullo JC, Ghidini A, Lopez-Zeno JA. Clinical correlations of patterns of Thrombocytopenia in Preterm Neonates. Pediatric Research 1996; 40: 112–119. placental pathology in preterm preeclampsia. Placenta 1998; 19: 67–72. 7 Beiner M, Simchen M, Sivan E, Chetrit AM, Kuint J, Schiff E. Risk factors for neonatal 20 Kush ML, Gortner L, Harman CR, Baschat AA. Sustained hematological consequences thrombocytopenia. Am Journ of Perinat 2003; 20: 49–54. in the first week of neonatal life secondary to placental dysfunction. Early Hum Devel 8 Chakravorty S, Murray N, Roberts I. Neonatal Thrombocytopenia. Early Hum Dev 2006; 82: 67–72. 2005; 81: 35–41. 21 Axt-Fleidner R, Hendrik HJ, Schmidt W. Nucleated red blood cell counts in growth- 9 Fraser SH, Tudehope DI. Neonatal neutropenia and thrombocytopenia following restricted neonates with absent or reversed-end-diastolic umbilical artery velocity. maternal hypertension. Journ of Ped and Child Health 1996; 32: 31–34. Clin Exp Obst and Gyn 2002; 29: 242–246. 10 Soma H, Yoshida K, Mukaida T, Tabuchi Y. Morphologic changes in the hypertensive 22 Tauscher MK, Berg D, Brockmann M, Seidenspinner S, Speer CP, Groneck P. placenta. Contrib Gynecol Obstet 1982; 9: 58–75. Association of histologic chorioamnionitis, increased levels of cord blood cytokines, 11 Ghidini A, Salafia CM, Pezullo JC. Placental vascular lerions and likelihood of and intracerebral hemorrhage in preterm neonates. Biology of the Neonate 2003; 83: diagnosis of preeclampsia. Obstet Gynecol 1997; 50: 542–545. 166–170. 12 ACOG Committee on Obstetric Practice. ACOG practice bulletin. Diagnosis and 23 Kim CJ, Yoon BH, Park SS, Kim MH, Chi JG. Acute funisitis of preterm but not term management of preeclampsia and eclampsia. No. 33, January 2002. American College placentas in associated with severe fetal inflammatory response. Hum Pathol 2001; 32: of Obstetricians and Gynecologists. Obstet Gynecol 2002; 99: 159–167. 623–629.

Journal of Perinatology