Journal of Perinatology (2006) 26,89–92 r 2006 Nature Publishing Group All rights reserved. 0743-8346/06 $30 www.nature.com/jp ORIGINAL ARTICLE nucleated red blood cell counts: normal values and the effect of labor

JM McCarthy1, T Capullari2, Z Thompson3, Y Zhu3, WN Spellacy1 1Department of and Gynecology of the University of South Florida College of Medicine, Tampa, FL, USA; 2Department of Pathology of Tampa General Hospital, The University of South Florida, Tampa, FL, USA and 3Department of Epidemiology and Biostatistics, The University of South Florida, Tampa, FL, USA

The assumption has been that erythropoietin production, Objective: To determine the umbilical cord blood nucleated red blood cell secondary to hypoxia, causes the production of nRBC. It has been (UC-nRBC) count in uncomplicated delivered by elective shown that in sheep, acute hypoxemia results in erythopoietin cesarean section or delivered vaginally. increases after at least 3 h,10 and the resulting reticulocytosis does 11 Methods: A total of 57-term singleton pregnancies were studied: 33 with not peak until 4 days. However, in rats, acute hypoxia resulted in 12 elective cesarean sections and 24 with vaginal deliveries. UC-nRBC was elevated nRBC sometime after 4–12 h. It has also been suggested analyzed for its nucleated red blood cell counts. A logarithmic that acute hypoxia in humans can be associated with elevated 13–15 transformation of the data was used for statistical analysis. nucleated red blood cells. In humans, labor has been associated with elevated umbilical cord erythropoietin levels.16 Results: The mean±standard deviation (s.d.) for nucleated red blood Increased nRBC have also been found in placentas in association cell per 100 white blood cells (nRBC/100WBC) from the elective cesarean with cases of acute hypoxia.17 section group was 7.8±7.4. The vaginal delivery group had a mean value The literature has inconsistent data on the normal values for of 9.3±10.5, which was not significantly different. A value of 22 nRBC/ UC-nRBC. The purpose of this study was to determine the UC-nRBC 100WBC defined the upper 95% confidence limit. The correlation between values in elective repeat cesarean section deliveries of women with absolute nRBC and nRBC/100 WBC was 0.97. low-risk pregnancies. In addition, levels of UC-nRBC were Conclusion: Although chronic hypoxia is associated with elevated nRBC, measured after vaginal deliveries of women with uncomplicated the stress of uncomplicated labor does not change the level. This adds pregnancies and labors to determine the effects of labor on the credence to its use as a marker for hypoxia preceding labor and delivery. UC-nRBC levels. Journal of Perinatology (2006) 26, 89–92. doi:10.1038/sj.jp.7211437; published online 12 January 2006 Methods Keywords: nucleated red blood cells; umbilical cord blood; labor; fetal blood; hypoxia; markers of asphyxia This study was approved by the institutional review board. A prospective study was conducted at Tampa General Hospital, from October 2003 to May 2004. Subjects were chosen from women Introduction presenting to the labor floor for delivery when one of the authors The determination of whether a hypoxic insult to a newborn infant was available to collect data and specimens. Only patients with was sustained in the antepartum or intrapartum time period is of singleton pregnancies between 37 and 41 weeks gestation, crucial value to obstetrics research. It has been suggested that the confirmed by last menstrual period and early ultrasound, were presence of elevated nucleated red blood cells in umbilical cord blood included. Patients were excluded if they had any characteristics (UC-nRBC) is a sign of chronic fetal hypoxia.1,2 Studies have known to be associated with chronic or acute hypoxia (see established that in term infants, this index may be affected by factors Table 1). such as maternal diabetes mellitus,3–5 fetal growth restriction,6,7 RH Umbilical cord venous blood was obtained immediately after isoimmunization,8 maternal tobacco use,4 and .9 cord clamping by needle aspiration at the time of delivery from 57 women. The blood was transferred into a tube containing EDTA. Correspondence: Dr Joan McCarthy, Department of Obstetrics and Gynecology, University of The number of nRBC per 100 white blood cells (WBC) was South Florida, 4 Columbia Drive, Tampa, FL 33606 USA. measured using an automated Cell-DYN4000 instrument (Abbott E-mail: [email protected] Received 8 September 2005; revised 31 October 2005; accepted 12 November 2005; published Laboratories, Chicago, IL). This method was validated by online 12 January 2006 comparing aliquots using direct microscopic visualization of Umbilical cord nucleated red blood cell counts JM McCarthy et al 90

Table 1 Exclusion criteria used to define normal Table 2 Characteristics of the study populationa

Anemia-hemoglobin <10 mg/dla No labor (Group I) Labor (Group II) P value Maternal tobacco use during pregnancy Hypertension Maternal age (years) 25.8±6 22±4 0.005 Diabetes-gestational or pre-existing Parity 1.4±1.1 0.8±1.2 0.02 Fetal growth restriction (weight<10%/week gestation) Gestational age (weeks) 38.6±0.9 39.0±1.3 NS Abnormal fetal karyotype Birthweight (g) 3373±441 3396±404 NS Sex (% male/female) 68/32 51/49 NS RH Isoimmunization -stained amniotic fluid Apgar 1 min 8.7±0.5 8.5±1.2 NS Umbilical artery pH<7.10 (7–9) (7–9) Apgar<7 at 1–5 min Fetal heart rate repetitive late decelerations or severe variable decelerations 5 min 9.0±0.0 8.9±0.4 NS Cord knot or >1 nuchal cord (9–9) (7–9) Cord prolapse Cord pH 7.25±0.1 7.26±0.1 NS Abruption aData are presented as mean±s.d. (range). Chorioamnionitis NS, not significant Maternal drug or alcohol abuse

aAt any time during pregnancy. (ks ¼ 0.1742, P ¼ 0.72). A post hoc power analysis demonstrated that this study had a power of 0.99 to detect a difference in means prepared slides and read by experienced hematology technicians.18 of 5 nRBC (a ¼ 0.05). For absolute nRBC, the mean±s.d. (median) for groups I and II were 909±931 Â 106/l (715) and For comparisons, we also expressed the nRBC as an absolute 6 number by correcting the WBC for the presence of nRBC. 1180±1502 Â 10 /l (931) respectively; there was no significant Group I (n ¼ 33) were women with low-risk pregnancies who difference. NRBC/100 WBC and absolute nRBC were highly delivered by elective cesarean section, in the absence of labor. In correlated, r ¼ 0.97. all, 29 of these were elective repeat cesareans; three were carried Since there was no difference between the groups they were out for breech presentation, and one for prior myomectomy. Group combined to determine a mean UC-nRBC of 8.4±8.7 (median 6.1). The mean of the absolute nRBC was 1016±1184 Â 106/l. II (n ¼ 24) were women with low-risk pregnancies who delivered 6 vaginally. In both groups assessment of the infants included A value of 22 nRBC/100WBC, or 5600 Â 10 /l was the upper 95% weight, sex, gestational age, Apgar scores and umbilical artery confidence limit for term umbilical cord venous blood. blood gas analyses. In both samples the distribution of UC-nRBC was positively skewed so data was log transformed for analysis. The Discussion means and s.d. were determined for the variables and these data were evaluated using t-tests and Kolmogorov–Smirnov tests on the The two main purposes of this study were to determine (1) what log transformed data. For absolute nRBC, the F test for equality of term infant UC-nRBC levels are in pregnancies not complicated by variances showed that the groups did not share homogeneity of hypoxia; and (2) the effect of uncomplicated labor on UC-nRBC variance, therefore the Wilcoxon Rank Sum Test was used. levels. This information is critical in order to evaluate results Probability values <0.05 were considered significant. obtained from complicated pregnancies in which the might have become hypoxic and the UC-nRBC level is being used to attempt to time the hypoxic event. This study sought to develop a Results pure group of pregnant women with no acute or chronic hypoxic Comparison of the maternal and infant data for the two groups condition in order to determine a true ‘normal’ UC-nRBC value. is shown in Table 2. There was no difference in , Many studies have attempted to establish ‘normal’ UC-nRBC gestational age, acid base status or Apgar scores between groups. values (see Table 3), but most failed to eliminate some of the However, since Group I was largely composed of elective repeat conditions that could be associated with fetal hypoxic risk. cesarean sections, patients in that group were significantly older Saracoglu and associates simply excluded ‘obstetrical problems’ than the laboring patients, and had significantly greater parity. from their control group.19 Ghosh et al.20 did not eliminate The mean±s.d. for UC-nRBC of group I was 7.8±7.4 (range patients with hypertension, pre-, tobacco use, meconium, 0–28.6, median 6.0) and for group II it was 9.3±10.5 (range growth retardation or chorioamnionitis. Korst et al.2 did not 0–49.7, median 7.5). A Kovmogorov-Smirnov test confirmed that exclude patients with hypertension, pre-eclampsia, maternal there was no significant difference between the two distributions tobacco use, or chorioamnionitis. Axt et al.21 did not exclude

Journal of Perinatology Umbilical cord nucleated red blood cell counts JM McCarthy et al 91

Table 3 ‘Normal’ nucleated red blood cell count reported in literature

First author Sample size Source Normal value (s.d.) Range

Saracoglu19 2000 45 Mixed cord blood 7.56(3.85) nRBC/100WBC Ghosh20 2003 49 Umbilical venous blood 8.6 (7.01) nRBC/100WBC 1–26 Korst2 1996 83 Mixed cord blood 3.4 (3) nRBC/100WBC 0–12 Axt21 1999 261 Umbilical artery blood 3.7a nRBC/100WBC 0–14 Sheffer-Mimouni et al.23 2003 60–28 (vaginal) Neonatal CBC 356 Â 106/Lb (vaginal) 0–1260 32 (cesarean) 413 Â 106/Lb (cesarean) 0–1470

Results are mean (s.d.). aMedian, mean and s.d. not reported. bExpressed as absolute nRBC. patients with , tobacco use, or signs of in labor 5 Hanlon-Lundberg K, Kirby R, Gandhi S, Broekhuizen F. Nucleated red blood (meconium, low Apgars, low pH). The values obtained in this study cells in cord blood of singleton term neonates. Am J Obstet Gynecol 1997; establish a normal UC-nRBC in term pregnancies. 176: 1149–1156. From Table 3, it is apparent that the s.d. we report for nRBC is 6 Baschat A, Gembruch U, Reiss I, Gortner L, Harman CR, Weiner CP. larger than that previously reported in some other studies. We used Neonatal nucleated red blood cell counts in growth-restricted : automated readings of nRBC rather than manual counts, which relationship to arterial and venous Doppler studies. Am J Obstet Gynecol may account for higher counts overall and larger s.d. For purely 1999; 181: 190–195. 7 Ferber A, Grassi A, Akyol D, O’Reilly-Green C, Divon MY. The association of statistical reasons, a 100 cell differential manual count, which is fetal heart rate patterns with nucleated red blood cell counts at birth. Am J commonly used at most institutions, is inaccurate when the Obstet Gynecol 2003; 188: 1228–1230. 22 percentage of NRBC is <5%. When 100 cell differentials are used, 8 Fox H. The incidence and significance of nucleated erythrocytes in the fetal it is likely that the automated NRBC at the low values is more vessels of the mature human placenta. J Obstet Gynaecol Br Commonw 18 accurate than the manual smear. 1967; 74: 40–43. This study also determined that uncomplicated labor did not 9 Leikin E, Garry D, Visintainer P, Verma U, Tejani N. Correlation of neonatal affect nRBC levels in UC-nRBC. Sheffer-Mimouni et al.23 looked at nucleated red blood cell counts in preterm infants with histologic this question utilizing venous blood drawn from neonates up to chorioamnionitis. Am J Obstet Gynecol 1997; 177: 27–30. 12 h after birth and found that levels did not differ between laboring 10 Widness JA, Teramo KA, Clemons GK, Garcia JF, Cavalieri RL, Piasecki GJ and nonlaboring patients. However, the clearance time of nRBC et al. Temporal response of immunoreactive erythropoietin to acute varies,2,14 and nRBC levels can be affected by postdelivery hypoxemia in fetal sheep. Pediatr Res 1986; 20: 15–19. oxygenation and other conditions after birth.13 Thilaganathan 11 Georgieff MD, Schmidt RL, Mills MM, Radmer WJ, Widness JA. Fetal iron and cytochrome status after intrauterine hypoxemia and erythropoietin et al.24 did a similar study utilizing UC-nRBC, with the opposite administration. Am J Physiol 1992; 262: R485–R491. result, however, they did not exclude patients with conditions 12 Blackwell SC, Hallak M, Hotra JW, Refuerzo J, Hassan SS, Sokol RJ, associated with chronic hypoxia. The present results show that labor Sorokin Y. Timing of fetal nucleated red blood cell count elevation in in low-risk pregnant women, in the absence of abnormal fetal heart response to acute hypoxia. Biol Neonate 2004; 85: 217–220. tracings or other signs of fetal stress, has no significant effect on the 13 Naeye R, Localio R. Determining the time before birth when ischemia and UC-nRBC level. These data can now be used in studies of high-risk hypoxemia initiated . Obstet Gynecol 1995; 86: 713–719. pregnancies to determine their effects on UC-nRBC levels. 14 Phelan J, Ock Ahn M, Korst L, Martin G. Nucleated red blood cells: a marker for fetal asphyxia? Am J Obstet Gynecol 1995; 173: 1380–1384. 15 Hermansen MC. Nucleated red blood cells in the fetus and newborn. Arch Dis References Child Fetal Neonatal Ed 2001; 84: F211–F2115. 1 Phelan J, Korst L, Ock Ahn M, Martin G. Neonatal nucleated red blood cell 16 Widness JA, Clemons GK, Garcia JF, Oh W, Schwartz R. Increased and lymphocyte counts in fetal brain injury. Obstet Gynecol 1998; 91: immunoreactive erythropoietin in cord serum after labor. Am J Obstet 485–489. Gynecol 1984; 148: 194–197. 2 Korst L, Phelan J, Ock Ahn M, Martin G. Nucleated red blood cells: an 17 Fox H. The incidence and significance of nucleated erythrocytes in the fetal update on the marker for fetal asphyxia. Am J Obstet Gynecol 1996; 175: vessels of the mature human placenta. J Obstet Gynaecol Br Commonw 843–848. 1967; 74: 40–43. 3 Green DW, Mimouni F. Nucleated erythrocytes in healthy infants and in 18 McCarthy J, Capullari T, Spellacy W. The correlation between automated infants of diabetic mothers. J Pediatr 2001; 116: 129–131. hematology and manually read smears for the determination of nucleated 4 Yeruchimovich M, Dollberg S, Green D, Mimouni F. Nucleated red blood red blood cells in umbilical cord blood. J Matern Fetal Neonatal Med 2005; cells in infants of smoking mothers. Obstet Gynecol 1999; 93: 403–406. 17: 199–201.

Journal of Perinatology Umbilical cord nucleated red blood cell counts JM McCarthy et al 92

19 Saracoglu F, Sahin I, Eser E et al. Nucleated red blood cells as a marker in 22 Rumke CL. Statistical reflections on finding atypical cells. Blood Cells 1985; acute and chronic fetal asphyxia. In J Gynecology and Obstetrics 2000; 71: 11: 141–144. 113–118. 23 Sheffer-Mimouni G, Mimouni F, Lubetzky R et al. Labor does not affect the 20 Ghosh B, Mittal S, Kumar S, Dadhwal V. Prediction of neonatal absolute nucleated red blood cell count. Am J Perinatology 2003; with nucleated red blood cells in cord blood of newborns. In J Gynecology 20: 367–371. and Obstetrics 2003; 81: 267–271. 24 Thilaganathan B, Athanasiou S, Ozmen S, Creighton S, Watson NR, 21 Axt R, Ertan K, Hendrik J, Wrobel M et al. Nucleated red blood cells in cord Nicolaides KH. Umbilical cord blood erythroblasts counts as an index blood of singleton term and post-term neonates. J Perinat Med 1999; 27: of . Arch Dis Child Fetal Neonal Ed 1994; 70: 376–381. F192–F194.

Journal of Perinatology