MANUSCRIPT CITATION

Kirpalani H, Bell EF, Hintz SR, Tan S, Schmidt B, Chaudhary AS, Johnson KJ, Crawford MM, Newman JE, Vohr BR, Carlo WA, D’Angio CT, Kennedy KA, Ohls RK, Poindexter BB, Schibler K, Whyte RK, Widness JA, Zupancic JAF, Wyckoff MH, Truog WE, Walsh MC, Chock VY, Laptook AR, Sokol GM, Yoder BA, Patel RM, Cotten CM, Carmen MF, Devaskar U, Chawla S, Seabrook R, Higgins RD, Das A; Eunice Kennedy Shriver NICHD Neonatal Research Network. Higher or Lower Hemoglobin Transfusion Thresholds for Preterm Infants. N Engl J Med. 2020 Dec 31;383(27):2639-2651. DOI: 10.1056/NEJMoa2020248. PMID: 33382931.

REVIEWED BY

Erin Grace
Neonatal Fellow
Department of Neonatal Medicine, Women’s and Children’s Hospital, North Adelaide, South Australia
SAHMRI Women and Kids, South Australian Health and Medical Institute, North Adelaide, South Australia
Robinson Research Institute and the Adelaide Medical School, the University of Adelaide, Adelaide, South Australia

Amy Keir
Consultant Neonatologist
Department of Neonatal Medicine, Women’s and Children’s Hospital, North Adelaide, South Australia
Healthy Mothers, Babies and Children Theme, South Australian Health and Medical Institute, North Adelaide, South Australia
Robinson Research Institute and the Adelaide Medical School, the University of Adelaide, Adelaide, South Australia

TYPE OF INVESTIGATION

Treatment

QUESTION

Among infants with a birth weight of less than 1000 grams (P), does restrictive (I) compared with liberal (C) red blood cell transfusion strategies, it affects death or neurodevelopmental impairment (O) at 22 to 26 months of corrected age?

METHODS

• Design: Multicentre open randomised control trial
  • Randomisation: 1:1 ratio to higher or lower threshold group
    • Stratified according to birth weight (<750 grams or 750-1000 grams)
    • Randomly selected block sizes of two or four infants
    • Multiple-birth siblings underwent randomisation individually
  • Caregivers not blinded to treatment, outcome assessors were blinded
• Allocation: Performed centrally by telephone with the data coordinating centre
• Follow-up period: 22-26 months corrected age
• Setting: 41 neonatal intensive care units (NICUs) from the Neonatal Research Network (NRN) in the USA
• Patients: 1824 infants
  • Inclusion criteria:
    • Infants weighing 1000 grams or less at birth with a gestational age between 22 weeks 0 days and 28 weeks and 6 days
    • Postnatal age of 48 hours or less (for initial participation)
  • Exclusion criteria:
    • Lack of viability
    • Cyanotic congenital heart disease
    • Family opposed to blood transfusion
    • Parent with hemoglobinopathy or congenital anaemia
    • History of transfusion in-utero, twin-to-twin transfusion syndrome or isoimmune hemolytic disease
    • Receipt of RBC transfusion after the first 6 hours of age
    • The low likelihood that the infant’s family would be able to return for follow-up assessment at 22-26 months corrected age
    • The receipt or planned receipt of erythropoietin
    • Congenital condition adversely affecting life expectancy or neurodevelopment
• Intervention/Comparison: Higher versus lower RBC transfusion thresholds. Infants received a transfusion when their haematocrit (or haemoglobin) fell below a predefined higher or lower threshold until 36 weeks of postmenstrual age or discharge from the NICU of the trial hospital, whichever occurred first.
  • Haemoglobin transfusion thresholds in both groups were determined according to postnatal age (highest in the first week) and use of respiratory support (a higher threshold when respiratory support in use). Respiratory support included mechanical ventilation, continuous positive airway pressure, FiO₂>0.35, or delivery of oxygen or room air by nasal cannula at a flow of 1 litre per minute or more.
  • Thresholds used in the study are provided in Table 1

Table 1: Haemoglobin RBC transfusion threshold used in the study in g/dL

*Respiratory support: mechanical ventilation, continuous positive airway pressure, FiO₂>0.35, or nasal cannula ≥1 litre per min (room air nasal cannula ≥1 litre per minute).

• Outcomes:
  • Primary outcome: Incidence of death or neurodevelopmental impairment at 22-26 months of age, corrected for prematurity.
  • Neurodevelopmental impairment was defined as one or more of the following components – cognitive delay, moderate or severe cerebral palsy or severe vision or hearing loss.
    1. Cognitive delay – composite cognitive score of <85 on the Bayley Scales of Infant and Toddler Development (3^rd edition)
    2. Moderate cerebral palsy – level II or III in the Gross Motor Function Classification (GMFCS) and severe cerebral palsy as levels IV or V.
    3. Severe visual impairment – best-corrected visual acuity of less than 20/200
    4. Severe hearing impairment – bilateral hearing loss requiring hearing aids or cochlear implants
  • Secondary outcomes measured at two-time points:
    • • During initial hospitalisation:
        1. Ultrasound grade 3 or 4 IVH, cystic PVL, or ventriculomegaly
        2. Bell stage 2 or 3 NEC
        3. Apnoea mandating either caffeine or respiratory support
        4. Age at final tracheal extubation
        5. Number of donor exposures by RBC donors or other blood product
        6. Number of transfusions
        7. ROP stage 3 or greater or requiring treatment
        8. Time to regain birthweight
        9. Time to full feedings
        10. Death before discharge
        11. BPD at 36 weeks postmenstrual age based on the need for oxygen supplementation following an oxygen reduction test
        12. Decreased height, weight, or head circumference at 36 weeks postmenstrual age
        13. length of hospital stay at Level 3 care
        14. Costs of hospitalisation.
      • At 22-26months corrected age, each of the composite outcomes individually
        1. Moderate or severe cerebral palsy
        2. Severe vision impairment
        3. Severe hearing impairment
        4. BSID-III Cognitive score < 85
        5. BSID-III cognitive, language, or motor score <70
        6. Gross Motor Function level ≥ II
        7. BITSEA score <15 for girls and <13 for boys (changed to the Child Behavior Checklist after 2013)
        8. Death before 22-26month follow-up
        9. Decreased height, weight, or head circumference
        10. Cost-benefit analysis of primary outcome at 22-26months
  • Analysis: The primary outcome analysis (intention to treat) was performed. All the analyses were adjusted for stratification according to birthweight group and centre. The primary outcome was analysed using Poisson regression to obtain adjusted relative risks and 95% confidence intervals.
  • Sample size: Based on data from the Neonatal Research Network (NRN) centres and previous results from the TOP trial, a conservative estimate based on assumed for the primary outcome (combined death and/or neurodevelopmental impairment) event of 53.5% (lower threshold) and 46.5% (higher threshold) for the two treatment groups. With a two-tailed alpha of 0.05 and 80% power, the initial sample size estimate was 1658 infants. The sample size was then increased by 10% to account for loss to follow up at 22-26 months. The final sample size calculation was 1824 infants (as per published protocol) detect an absolute difference of 7 percentage points in the incidence of the primary outcome between the trial group.
  • Patient follow-up: There was less than 10% loss to follow-up, with comparable numbers in each group.
  • Additional notes: The number of protocol compliant transfusions was 5342/5624 (95%) in the high haemoglobin threshold group, 3115/4055 (77%) in the low haemoglobin group. There was a total of 8457/9679 (87%) RBC transfusions compliant with the study protocol.

MAIN RESULTS

Main study outcomes: Amongst infants born at 1000 grams or less with gestational age 22 weeks and 0 days to 28 and 6 weeks, at follow up at 22-26 months corrected age, a high threshold RBC transfusion strategy compared to a liberal RBC transfusion strategy did not reduce the likelihood of the following:

• • Death or neurodevelopmental impairment – higher threshold, 423/845 (50.1%) and lower threshold 422/847 (49.8%) (adjusted relative risk 95% CI 1.00 (0.92-1.10) p=0.93
  • Death – higher threshold, 146/903 (16.2%) and lower threshold 135/901 (15%) (adjusted relative risk 95% CI 1.07 (0.87-1.32)
  • Neurodevelopmental impairment – higher threshold, 277/699 (39.6%) and lower threshold 287/712 (40.3%) (adjusted relative risk 95% CI 1.00 (0.88-1.13)
  • Cognitive delay – higher threshold, 269/695 (38.7%) and lower threshold 270/712 (37.9%) (adjusted relative risk 95% CI 1.04 (0.91–1.18)
  • Moderate or severe cerebral palsy – higher threshold, 48/711 (6.8%) and lower threshold 55/720 (7.6%) (adjusted relative risk 95% CI 0.87 (0.60–1.26)
  • Severe vision impairment – higher threshold, 5/713 (0.7%) and lower threshold 6/720 (0.8%) (adjusted relative risk 95% CI 0.83 (0.25–2.76
  • Severe hearing impairment – higher threshold, 14/710 (2.0%) and lower threshold 25/715 (3.5%) (adjusted relative risk 95% CI 0.56 (0.29–1.07)
• Primary outcome: The outcome of death or neurodevelopmental impairment at 22-26 months was present in 50.1% of the infants randomised to the higher group compared to 49.8% randomised to the restrictive group, for an adjusted relative risk of 1.0 (95% CI 0.92-1.10) adjusted for centre and birth weight stratum.
• Secondary outcomes: There were no differences between treatment groups in rates of the various components of the primary outcome or for any short-term or longer-term secondary outcomes.
• Transfusion-related data: The mean (±SD) number of transfusions was 6.2±4.3 and 4.4±4.0, respectively (mean difference, 1.71; 95% CI, 1.37 to 2.05). At randomisation (48 hours of age or less), mean (±SD) haemoglobin levels were similar in the two treatment groups; (13.8±2.6 grams per decilitre (g/dL) in the higher-threshold group and 13.7±2.6 g/dL in the lower-threshold group). Across the study period, the pretransfusion mean haemoglobin levels differed between the groups by 1.9g/dL (p<0.001).

CONCLUSION

Amongst infants born at 1000 grams or less with gestational age 22 weeks and 0 days to 28 and 6 weeks, at follow up at 22-26 months corrected age, a higher RBC transfusion threshold compared to a liberal RBC transfusion threshold did not reduce the likelihood of death or neurodevelopmental impairment. There were no significant differences in complications of prematurity or adverse outcomes between groups.

COMMENTARY

The Transfusion of Prematures (TOP) trial explored the effects of red cell transfusion thresholds on neurodevelopmental outcomes in extremely low birthweight (ELBW) infants. The study was proposed due to a post hoc analysis demonstrating a slightly higher mild-moderate cognitive delay in the restrictive transfusion group in a previous transfusion threshold trial.(2)

The TOP trial was a multicentre, unmasked randomised control trial, involving 1824 infants, born 1000 grams or less with a gestational age between 22 weeks 0 days and 28 weeks and 6 days. It was conducted in 19 centres (41 NICUs) in the United States of America between 2012 to 2017, with follow-up through to 2020. The primary outcome, a composite of death or neurodevelopmental impairment at 22-26 months corrected age, was not significantly different between the two groups; liberal transfusion 423/845 (50.1%) and restrictive transfusion (422/847 (49.8%) (adjusted relative risk 95% CI 1.00 (0.92-1.10) p=0.93. The rates of common complications of prematurity, including IVH, NEC, ROP, BPD, growth parameters and length of hospital stay, were similar. In terms of haemoglobin (Hb) data, there was a difference of 1.9g/dL in pretransfusion mean Hb levels through the study period.

Randomisation occurred within the first 48 hours with thresholds for transfusion applied until infants were 36 weeks postmenstrual age and differed depending on whether infants received respiratory support (defined as mechanical ventilation, continuous positive airway pressure, FiO2 >0.35, or nasal cannula ≥1 litre per min). Delayed cord clamping (DCC) or cord milking was documented in 27% in the high threshold and 25% of the low threshold analysis cohort (the cohort of which had complete data for the primary outcome at follow up). The practice of DCC or milking is relatively new in neonatal practice and was described as occurring in the original PINT trial.(2)

We used the Cochrane Risk of Bias (ROB) tool(3) to assess ROB in the study, and deemed it low risk overall. The risk of post randomisation confounding exists with variation in per-protocol transfusions. There were non-protocol compliant transfusions; 23% of all transfusions administered in the restrictive transfusion and 5% in the liberal transfusion groups. Despite this, separation of haemoglobin level was maintained between treatment groups throughout the study period. Transfusions that should have taken place per protocol but were not performed were deemed missed transfusions. 185 (3.3%) missed transfusions in the high threshold group and 33 (1%) in the low threshold group. Additional potential for confounding included centre variation in blood banking practises, i.e., only transfusing fresh red blood cells. The authors argued that this variability in blood bank practices increases the generalisability of findings.

With the recent ETTNO trial(4) results consistent with the TOP findings, a low (restrictive) transfusion threshold appears safe and appropriate to use. Combining low (restrictive) transfusion threshold practices with measures that reduce anaemia in preterm infants, such as enhanced placental transfusion, restrictive phlebotomy sampling practices,(5) and cord blood sampling for admission blood tests,(6) should result in less transfusion. This may have flow-on effects of reducing the risk for potential transfusion side effects, improved allocation of a limited resource product and reduced health care costs.

REFERENCES

1. Kirpalani H, Bell EF, Hintz SR, et al. Higher or Lower Hemoglobin Transfusion Thresholds for Preterm Infants. N Engl J Med 2020;383(27):2639-51.
2. Kirpalani H, Whyte RK, Andersen C, et al. The Premature Infants in Need of Transfusion (PINT) study: a randomized, controlled trial of a restrictive (low) versus liberal (high) transfusion threshold for extremely low birth weight infants. J Pediatr 2006;149(3):301-07.
3. Higgins JPTC, R.; Chandler, J.; Cumpston, M.S. Cochrane Handbook for Systematic Reviews of Interventions Version 5.2.0 Cochrane, Updated June 2017.
4. Franz AR, Engel C, Bassler D, et al. Effects of Liberal vs Restrictive Transfusion Thresholds on Survival and Neurocognitive Outcomes in Extremely Low-Birth-Weight Infants: The ETTNO Randomized Clinical Trial. JAMA 2020;324(6):560-70.
5. Alan S, Arsan S. Prevention of the anaemia of prematurity. Int J Pediatr Adolesc Med 2015;2(3-4):99-106.
6. 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. The Journal of pediatrics 2019;211:39-45 e2.