MANUSCRIPT CITATION

Franz AR, Engel C, Bassler D, Rüdiger M, Thome UH, Maier RF, Krägeloh-Mann I, Kron M, Essers J, Bührer C, Rellensmann G, Rossi R, Bittrich HJ, Roll C, Höhn T, Ehrhardt H, Avenarius S, Körner HT, Stein A, Buxmann H, Vochem M, Poets CF; ETTNO Investigators.  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 Aug 11;324(6):560-570. PMID: 32780138.

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

Cathie Hilditch
Paediatric Registrar
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 have an effect on death or neurodevelopmental impairment at 24 months of corrected age (O)? As per the published ETTNO protocol, the study had one main objective: To test the hypothesis that ‘liberal’ red blood cell transfusion practices that aim to keep haematocrit levels >28% at all times during the initial hospitalisation will either improve or impair long-term outcome in extremely low birthweight infants when compared with ‘restrictive’ red blood cell transfusion guidelines that tolerate haematocrit levels as low as 21%.

METHODS

• Design: Multicentre randomised control trial
  • Randomisation: Within 72 hours of birth to 1 of 2 parallel treatment groups
    • Stratified by center and birth weight (birth weight stratum 400-749g; 750-999g).
    • Random sequence was computer generated with variable block size
  • Caregivers not blinded to treatment, outcome assessors were blinded (e.g., neurologists, paediatricians, psychologist, ophthalmologists)
• Allocation: Concealment performed using sealed, opaque envelopes
• Follow-up period: 24 months
• Setting: 36 level III/IV neonatal intensive care units (NICUs) in Europe
• Patients: 1013 infants
  • Inclusion criteria:
    • Infants weighing 400-999 grams at birth
  • Exclusion criteria:
    • Gestational age at birth greater than 29 weeks and 6 days
    • Major anomalies including chromosomal anomalies, cyanotic heart defects and syndromes that would affect long term outcomes
    • Malformations requiring surgical correct during the neonatal period
    • Participation in other studies precluding participation in this study
    • Lack of viability
    • Decision to provide comfort care only
    • For multiple pregnancies (twins or higher order), only the first eligible infant born first was included
• Intervention/Comparison: Restricted versus liberal red blood cell transfusion thresholds. Infants received a transfusion when their haematocrit (or haemoglobin) fell below a pre-defined restricted or liberal threshold. Four centres (total 36 infants) out of the 34 participating centres used haemoglobin trigger thresholds. Centres had to select either haematocrit or haemoglobin thresholds before commencing enrolment.
  • These thresholds were adjusted according to postmenstrual age and clinical condition e.g., whether an infant was deemed critical or non-critical, as per pre-specified criteria. These critical clinical criteria were as follows:
    • Requirement of mechanical ventilation (any mode, excluding continuous positive airway pressure (CPAP))
    • Requirement of CPAP with FiO₂>0.25 for greater than 12 hours per 24 hours
    • Patent ductus arteriosus (PDA) requiring therapy
    • Greater than 6 apnoea requiring stimulation per 24 hours or more than 4 desaturations to SpO₂ <60% per 24 hours despite methylxanthines (e.g., caffeine) and CPAP support
    • Acute sepsis or necrotising enterocolitis requiring inotropic/vasopressor support
  • Thresholds used in the study are provided in Tables 1 and 2.
• Outcomes:
  • Primary outcome: Incidence of death or neurodevelopmental impairment: any of the following:
    • Cognitive deficit (various measures)
    • Cerebral palsy
    • Severe visual impairment – best corrected visual acuity of less than 6/60
    • Severe hearing impairment – need for hearing aid or cochlear implant
  • Secondary outcomes:
    • Individual components of the primary outcome
    • Incidence of complications of prematurity
    • Growth at discharge and follow-up
    • Gross Motor Function Classification System (GMFCS) score
    • Time interval from birth to final discontinuation of:
      • Positive pressure respiratory support
      • Respiratory stimulant therapy
      • Gavage feeding
• Analysis: The primary outcome analysis (intention to treat) was performed according to the randomised treatment group assignment by logistic regression with the factors of treatment, centre, and birth weight stratum used to test the null hypothesis of equal proportions in the 2 treatment groups. A predefined per-protocol analysis was performed for all included infants who did not violate inclusion/exclusion criteria and underwent transfusion according to protocol from randomisation until discharge home.
• Sample size: The sample size was calculated based on rates of primary outcome the authors estimated from the previous PINT trial (i.e., 109/213 [51% in the liberal threshold group] compared to 126/208 [61% in restrictive threshold group]. This calculation was based on χ2 tests with 80% power, a 2-sided α = .05 significance level. For an absolute risk reduction of 10%, 390 patients per group were required and to allow for a 20% loss to follow up rate, the study needed to enrol 980 infants. Ultimately there were 1013 infants randomised (450 in liberal group and 478 in restrictive group).
• Patient follow-up: There was less than 10% loss to follow-up, with comparable numbers between each group.
• Additional notes: 25% of included infants had received at least one red blood cell transfusion prior to enrollment and there was significant cross over between groups (i.e., some infants received transfusion when their haematocrit was greater than assigned threshold, and some infants did not receive transfusion when hematocrit was less than assigned threshold). More details are provided below:
  • 14% of infants received a transfusion that was not as per study protocol
    • 47 infants in liberal threshold group
    • 97 in restrictive threshold group, predominantly in those with a birthweight of <750 grams
  • 7% of infants did not receive RBC transfusion within 48 hours of a haematocrit level below the assigned threshold
    • 65 infants in liberal threshold group
    • 5 infants in restrictive threshold group
    • Per protocol analysis confirmed findings of primary outcome

Table 1: Haematocrit thresholds (30 centres)

Table 2: Haemoglobin thresholds (4 centres)

MAIN RESULTS

Primary outcome: Amongst extremely low birthweight infants, at follow up at 24 months corrected age, liberal red cell transfusion strategy compared to restrictive strategy did not reduce the likelihood of the composite outcome of death or neurodevelopmental impairment or components of the outcome. Death or neurodevelopmental impairment by 24 months was present in 44.4% of the infants randomised to the liberal group compared to 42.9% randomised to the restrictive group, for a risk difference of 1.6% (95% CI, −4.8% to +7.9%) and an odds ratio of 1.05 (95% CI, 0.80-1.39; P = .72) adjusted for centre and birth weight stratum.

Secondary outcomes: There were no differences found between treatment groups in rates of the various components of the primary outcome or incidence of cognitive deficit (defined as Mental Development Index score <70), length of hospital stay, or time intervals from birth to final discontinuation of invasive respiratory support, positive pressure respiratory support and gavage feeding. Complications of prematurity including necrotising enterocolitis, bronchopulmonary dysplasia, retinopathy of prematurity or culture proven sepsis were not significantly different between treatment groups. Subgroup analyses for infants born at less than <750 grams did not find any differences in clinical outcomes between groups. The only difference in outcomes for treatment groups was weight at 36 weeks of postmenstrual age. This was higher in the liberal threshold group (mean, 2113 grams [standard deviation (SD), 356 grams] compared to 2068 grams [SD, 361 grams].

Transfusion-related data: Weekly mean haematocrit values were statistically significantly different between the treatment groups from week 1b (the week after randomisation) through to week 11. The mean difference in mean haematocrits ranged from 1.5 to 3.4% over this time period. Further summary results are in Table 3.

Main study outcomes: Amongst extremely low birthweight infants, at follow up at 24 months corrected age, a liberal red cell transfusion strategy compared to restrictive strategy did not reduce the likelihood of the following:

• • Death (absolute difference −0.7 [95%CI −4.3 to 2.9], odds ratio 0.91 [95% CI 0.58-1.45])
  • Cognitive impairment (absolute difference 3.1 [95%CI −3.3 to 9.6], odds ratio 1.12 [95% CI0.83-1.51])
  • Cerebral palsy (absolute difference −1.3 [95%−4.2 to 1.5], odds ratio 0.75 [0.40-1.40])
  • Composite outcome of death or neurodevelopmental impairment (absolute difference 1.6 [95%CI −4.8 to 7.9], odds ratio 1.05 [95% CI 0.80-1.390])

Table 3: Transfusion information after randomisation until 36 weeks post-menstrual age

CONCLUSION

Amongst extremely low birthweight infants, at follow up at 24 months corrected age, a liberal red blood cell transfusion strategy compared to restrictive red blood cell transfusion strategy 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

In this European multicentre randomised controlled trial, the Effects of Transfusion Thresholds on Neurocognitive Outcome of Extremely Low birth Weight Infants (ETTNO), 1013 preterm infants with a birthweight less than 1000 grams were included. The study was undertaken across 2011-2014 with longer term follow-up beyond this period.(1) At 24 months’ corrected age, rates of the primary outcome of death or neurodevelopmental impairment were similar in the liberal and restrictive threshold groups (44.4% and 42.9%; odds ratio, 1.05). The rates of mortality, cognitive impairment, and complications including bronchopulmonary dysplasia and necrotising enterocolitis was similar between groups.

To examine the study more closely, randomisation occurred with the first 72 hours of age with the majority of infants randomised at 2.5 days of age. Twenty-five percent of infants in each group received at least one red blood cell transfusion prior to randomisation. Deferred/delayed cord clamping (DCC) for 30-45 seconds was implemented at each study site with 308/492 (63%) in the liberal group and 319/519 (61%) in the restrictive group receiving DCC. The mean difference in mean haematocrits ranged from 1.5 to 3.4% from the week of randomisation to week 11. Whether this difference is clinically meaningful is worth considering. The authors note that the separation in these haematocrit values, and consequently the potential oxygen-carrying capacity achieved, may be too small to cause a difference in study outcomes.(1) The lowest mean haematocrit in the restrictive group was in week 9 at 31.7% (SD 4.6). A comprehensive list of further study limitations is well laid out by the authors. These included potential limited generalisability due to the infants being primarily from a white German population group, the changes in clinical practices that may have occurred since 2011-2014, and the proportion of infants who received a red blood cell transfusion not justified by the study protocol. Risk of bias (ROB) in the study, assessed using the Cochrane ROB tool(2), was deemed low risk overall. The potential for post randomisation confounding exists, with 14% of infants receiving a transfusion not justified by the protocol and 7% did not receive a transfusion when dictated by the protocol. Analysis of the per-protocol population was consistent with the primary analysis findings. Other post randomisation confounding risks were deemed low as there was minimal loss to follow up with similar numbers between groups.

The results of the recently published Transfusion of Prematures (TOP)  trial (3) are consistent with this trial. Combining the results will hopefully enable a much anticipated individual patient data meta-analysis.(1) Anaemia of prematurity is a term used broadly across the highly variable and complex course of prematurity in relation to red blood cell transfusion practice. It is possible that a difference approach to transfusion in the first days of age requires a transfusion trigger based on optimisation of oxygen kinetics, rather than a threshold per se. (4). At this point, however, lower red cell transfusion thresholds compared to more liberal thresholds seem an appropriate clinical approach, especially for those preterm infants 48-72 hours of age or older.

REFERENCES

1. Franz AR, Engel C, Bassler D, Rüdiger M, Thome UH, Maier RF, 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.
2. Higgins JPTC, R.; Chandler, J.; Cumpston, M.S. Cochrane Handbook for Systematic Reviews of Interventions Version 5.2.0 Cochrane; Updated June 2017
3. Kirpalani H BE, D’Angio C, Hintz S, Kennedy K, Ohls R, Poindexter B, Schibler K, Schmidt B, Vohr B, Widness J, Das A, Higgins R, Zupancic J, Roberts R, Whyte R, Chaudhary A, Johnson K. Protocol: Transfusion of Prematures (TOP) trial – does a liberal red blood cell transfusion strategy improve neurologically-intact survival of extremely-low-birth-weight infants as compared to a restrictive strategy? : ClinicalTrials.gov 2012; NCT01702805; 2012
4. Andersen CC, Keir AK, Kirpalani HM, Stark MJ. Anaemia in the Premature Infant and Red Blood Cell Transfusion: New Approaches to an Age-Old Problem. Current Treatment Options in Pediatrics 2015; 1 3:191-201.