Can cord blood sampling delay the first packed red blood cell transfusion?

Can cord blood sampling delay the first packed red blood cell transfusion?

Manuscript Citation: Balasubramanian H, Malpani P, Sindhur M, et al. Effect of Umbilical Cord Blood Sampling versus Admission Blood Sampling on Requirement of Blood Transfusion in Extremely Preterm Infants: A Randomized Controlled Trial. J Pediatr. 2019 Aug; 211

Rupin Kumar,¹ Brynne Sullivan¹

Affiliation:

• Division of Neonatology

Dept. of Pediatrics

University of Virginia Health System

Corresponding Author:

Rupin Kumar, MD

Contact Address:

749 Lords Ct, Charlottesville, VA 22901

Phone: (732) 614 2349

Email: rk9pg@hscmail.mcc.virginia.edu

Question:

In extremely preterm neonates born < 28 weeks’ gestational age and with birth weight < 1000g, does cord blood sampling from placental end compared to admission blood sampling from neonate affect the time from birth to first packed red blood cell (PRBC) transfusion?

Methods:

Design: Randomized Controlled Trial

Allocation: Mothers were approached to explain study interventions during the time of admission, but written informed consent was obtained after delivery if cord blood sampling (CBS) was successful and the infant(s) met eligibility criteria. Participants were allocated in random block sequences of varying sizes to CBS or admission blood sampling (ABS).

Blinding: Due to the nature of the study intervention, clinicians, nurses and parents of enrolled infants were not blinded to where blood was sampled from. Outcome assessors, which included primary investigators and laboratory technicians were blinded to the study group assigned to the patient.

Follow-up Period: From birth until death or NICU discharge.

Setting: Single center study conducted in a level III NICU of Surya Hospital, Mumbai, India.

Participants:

Inclusion Criteria: Gestational age < 28 weeks and birth weight < 1000 g

Exclusion Criteria: Chorioamnionitis, monochorionic twins or triplets, anomalies of the cord (true/false knots, strictures, funisitis), major congenital anomalies, Rh isoimmunization or placental anomalies.

Intervention (CBS): The placental end of the cord was used for cord blood sampling using aseptic technique. At least 5 ml blood was drawn from the umbilical and dispatched to the laboratory on ice within 10 minutes of collection. Further blood sampling was avoided for the next 12 hours, unless the samples were clotted, or clinical condition warranted earlier testing.

Control (ABS): Five ml of umbilical arterial or venous blood was collected from the neonate within the first hour after birth and the cord blood samples were discarded.

All infants, irrespective of randomization group, received standard anemia prevention strategies per unit policy.

Outcomes:

Primary Outcome: Time (in days) from birth to requirement of the first PRBC transfusion.

Secondary Outcomes:

1. Need for transfusion at 28 days and at discharge
2. Number of transfusions and mean hemoglobin levels at 4 and 6 weeks of postnatal age, and at the time of discharge
3. Mortality before discharge
4. Morbidities, including: duration of hospital stay, duration of respiratory support, incidence of IVH (Grade III- IV), NEC (Bell Stage 2 or higher), ROP requiring intervention, PDA needing medical/surgical treatment, PVL, BPD (defined as need for supplemental oxygen or respiratory support by 36 weeks’ postmenstrual age).

Analysis and Sample Size:

The study unit’s baseline time to first PRBC transfusion was a mean of 14 days postnatal for ELBW infants who underwent admission ABS. To demonstrate that the requirement of first PRBC transfusion can be postponed by 7 days in the CBS group, a sample size of 38 neonates in each group was estimated using a study power of 90% and 2-tailed alpha of 0.05 (using an assumed SD of 10 days in both groups). Continuous outcomes were compared using the 2-sample t-test (for parametric data) or Wilcoxon rank-sum test (for non-parametric data).  Fisher’s exact test was used to analyze categorical variables. All analyses were intention-to-treat, which was relevant for 3 infants randomized to CBS where samples were unable to be analyzed and blood was drawn on admission from the infant. The primary outcome was compared using survival analysis and plotting Kaplan Meier curves for time to first transfusion in each group. Neonates who did not require a transfusion until discharge or transfer, or those who died prior to first PRBC transfusion were censored. Log-rank tests were used to compare the survival curves between the two groups. To account for the simultaneous effect of several covariates on the primary outcome, unadjusted and adjusted hazard ratios were calculated using a Cox proportional hazards regression model. Weekly hemoglobin values were plotted for each group. Analysis of hemoglobin values used a mixed effects model with postnatal week considered to be a fixed effect and patients as random effect.

Results:

Of the 127 neonates that met inclusion criteria, 80 ELBW infants (40 in each group) were randomized and all were included in the survival analysis for the primary outcome. Maternal and neonatal characteristics were similar in both groups.

Primary Outcome: Median time to first PRBC transfusion was 30 (IQR 21-41) days in the CBS group, compared to 14 (IQR 7-26) days in the ABS group. The probability of needing a transfusion in the CBS group was 56% lower than the ABS group (hazard ratio = 0.44, 95% CI 0.27-0.72, log-rank p value < 0.001). Repeat analysis using a cox proportional hazards model adjusted for covariates did not change this result.

Secondary Outcomes: While fewer PRBC transfusion were needed in the first 4 weeks for infants in the CBS group (30 % in CBS vs 75% in ABS, p < 0.001), the percent receiving PRBC transfusion was equal by the time of NICU discharge (80% in CBS vs 87% in ABS, p = 0.54). Baseline hemoglobin levels were similar between groups (15.2 ± 2.2 g/dL in CBS, 15.7 ± 1.7 g/dL in ABS, p = 0.24), but separated slightly at 1-2 weeks (higher by at most 3 g/dL in the CBS group with overlapping confidence intervals) and then fell to similar levels again from 4-8 weeks postnatal age. In the mixed effects model, hemoglobin values in the ABS group were lower compared to the CBS group in the first 2-3 weeks after birth (p <0.001). There were no significant differences in other outcomes, including mortality.

Study Conclusion:

Blood sampling using cord blood, when combined with anemia prevention strategies, significantly prolonged time to first PRBC transfusion and reduced need for transfusions in the first 4 weeks after birth in infants <28 weeks’ gestation and <1000 g birthweight.

Commentary:

In this study, cord blood sampling prolonged the time to first PRBC transfusion with fewer transfusions in the first 4 weeks after birth, but nearly all infants required transfusion by 8 weeks. This randomized controlled trial analyzing the effect of CBS on need for transfusion compared to admission blood sampling contributes important information on this practice and adds to the body of literature that shows CBS is a feasible and usually successful procedure.^1,2 Only 8 infants were excluded due to failure to collect a cord blood sample and 3 samples from infants randomized to the CBS group were not able to be analyzed. The authors acknowledge that the study was limited by its small sample size and was not adequately powered to study the impact of this practice on relevant outcomes. A larger study might answer the question of whether reducing the need for PRBC transfusion in the first 4 weeks has an effect on morbidity and mortality in ELBW infants. While the evidence for the association of blood transfusions and some outcomes, such as NEC, is controversial, ^3,4 several studies demonstrate reason to believe that reducing early transfusions might impact short- and long-term outcomes, such as ROP, BPD and neurodevelopment, due to alteration of oxygen delivery, blood volume, and other physiologic factors during a critical phase of development.^5,6

Considering the pathophysiology of retinopathy, fewer transfusions in the first month after birth might show a decrease in severe ROP in a large enough trial. Hyperoxia in the first few weeks after birth disrupts vasculogenesis of the developing retina⁵. The transfusion of adult hemoglobin A leads to a rightward shift in the oxygen dissociation curve, which can cause increased oxygen delivery to retinal tissues and promote retinopathy.⁷ Observational studies have shown that the age at transfusion or number of transfusions in the first 30 days^7,8 might increase risk of severe ROP. In the current study, the authors have highlighted a trend toward reduced incidence of ROP needing treatment in the CBS group (32% vs 57%, p = 0.057).  While interpreting results, it is also important to consider the generalizability of this study. The standard anemia prevention practice used in the study center includes routine umbilical cord milking and use of recombinant EPO for all ELBW infants, two practices that are not routinely used in most centers.⁹

In conclusion, this randomized trial of cord blood sampling shows that this procedure is usually successful and reduces the need for PRBC transfusion in the first month after birth when adhering to strict transfusion guidelines in ELBW infants. A larger trial might show a difference in important outcomes such as severe ROP.

EBM Lesson:

Mixed-Effects Models for analysis of differences in repeated measures

When collecting longitudinal data, repeated measurements from study subjects are taken over a period of time. Two features of this type of data impact their analysis: first, the measurements are correlated within a subject (random effect). For example, an infant with low hemoglobin levels one week is more likely to be low the next week than an infant who had a high hemoglobin level the week before. Second, hemoglobin levels vary with postnatal age, which would be true for all subjects (fixed effect).  These features will impact the distribution of hemoglobin along with the study intervention. The goal of a mixed effects model is to take into account the correlated nature of repeatedly measured data points that might have a significant impact on the results and conclusions.¹⁰

The distribution of hemoglobin levels across a study cohort can be plotted showing mean hemoglobin level of all subjects as a function of time. Each individual subject’s hemoglobin curve can then be compared to the distribution of curves among all subjects and subjects within each study arm. If one assumes that each subject is drawn at random from the sample of interest, an individual measurement can be theoretically separated into that due to the subject (the random effect), that due to postnatal age or time (the fixed effect), the study intervention, and residual variance around the mean.

Acknowledgment:  

The Journal club is a collaboration between the American Academy of Pediatrics- Section of Neonatal Perinatal medicine and the International Society of Evidence- based neonatology (EBNEO.org)

Conflict of Interest: None declared

References:

1. Carroll PD, Nankervis CA, Iams J, Kelleher K. Umbilical cord blood as a replacement source for admission complete blood count in premature infants. Journal of Perinatology. 2012 Feb;32(2):97.
2. Baer VL, Lambert DK, Carroll PD, Gerday E, Christensen RD. Using umbilical cord blood for the initial blood tests of VLBW neonates results in higher hemoglobin and fewer RBC transfusions. Journal of Perinatology. 2013 May;33(5):363.
3. Patel, R. M., Knezevic, A., Shenvi, N., Hinkes, M., Keene, S., Roback, J. D., Josephson, C. D. Association of Red Blood Cell Transfusion, Anemia, and Necrotizing Enterocolitis in Very Low-Birth-Weight Infants. JAMA, 2016; 315(9), 889–897
4. Kirpalani H, Zupancic JA. Do transfusions cause necrotizing enterocolitis? Semin Perinatol. 2012;36(4):269–276
5. Lust C, Vesoulis Z, Jackups R Jr, Liao S, Rao R, Mathur AM. Early red cell transfusion is associated with development of severe retinopathy of prematurity. J Perinatol. 2019;39(3):393–400.
6. Wang YC, Chan OW, Chiang MC, Yang PH, Chu SM, Hsu JF, Fu RH, Lien R: Red blood cell transfusion and clinical outcomes in extremely low birth weight preterm infants. Pediatr Neonatol 2017; 58: 216–222
7. De Halleux V, Truttmann A, Gagnon C, Bard H. The effect of blood transfusion on the hemoglobin oxygen dissociation curve of very early preterm infants during the first week of life. Semin Perinatol. 2002;26(6):411–5.
8. dos Santos AM, Guinsburg R, de Almeida MF, Procianoy RS, Leone CR,Marba ST, et al. Red blood cell transfusions are independently associated with intra-hospital mortality in very low birth weight preterm infants. JPediatr 2011; 159:371-6.e1-e3
9. Ohlsson A, Aher SM. Early erythropoiesis‐stimulating agents in preterm or low birth weight infants. Cochrane Database of Systematic Reviews. 2017(11)
10. Vittinghoff E, Glidden DV, Shiboski SC, McCulloch CE. Regression methods in biostatistics: linear, logistic, survival, and repeated measures models. Chapter 8.3 (Springer Science & Business Media, 2^nd Edition, 2012)