key: cord-0842634-r3u9lv8c
authors: Thayyil, Sudhin; Pant, Stuti; Montaldo, Paolo; Shukla, Deepika; Oliveira, Vania; Ivain, Phoebe; Bassett, Paul; Swamy, Ravi; Mendoza, Josephine; Moreno-Morales, Maria; Lally, Peter J; Benakappa, Naveen; Bandiya, Prathik; Shivarudhrappa, Indramma; Somanna, Jagadish; Kantharajanna, Usha B; Rajvanshi, Ankur; Krishnappa, Sowmya; Joby, Poovathumkal K; Jayaraman, Kumutha; Chandramohan, Rema; Kamalarathnam, Chinnathambi N; Sebastian, Monica; Tamilselvam, Indumathi A; Rajendran, Usha D; Soundrarajan, Radhakrishnan; Kumar, Vignesh; Sudarsanan, Harish; Vadakepat, Padmesh; Gopalan, Kavitha; Sundaram, Mangalabharathi; Seeralar, Arasar; Vinayagam, Prakash; Sajjid, Mohamed; Baburaj, Mythili; Murugan, Kanchana D; Sathyanathan, Babu P; Kumaran, Elumalai S; Mondkar, Jayashree; Manerkar, Swati; Joshi, Anagha R; Dewang, Kapil; Bhisikar, Swapnil M; Kalamdani, Pavan; Bichkar, Vrushali; Patra, Saikat; Jiwnani, Kapil; Shahidullah, Mohammod; Moni, Sadeka C; Jahan, Ismat; Mannan, Mohammad A; Dey, Sanjoy K; Nahar, Mst N; Islam, Mohammad N; Shabuj, Kamrul H; Rodrigo, Ranmali; Sumanasena, Samanmali; Abayabandara-Herath, Thilini; Chathurangika, Gayani K; Wanigasinghe, Jithangi; Sujatha, Radhika; Saraswathy, Sobhakumar; Rahul, Aswathy; Radha, Saritha J; Sarojam, Manoj K; Krishnan, Vaisakh; Nair, Mohandas K; Devadas, Sahana; Chandriah, Savitha; Venkateswaran, Harini; Burgod, Constance; Chandrasekaran, Manigandan; Atreja, Gaurav; Muraleedharan, Pallavi; Herberg, Jethro A; Kling Chong, W K; Sebire, Neil J; Pressler, Ronit; Ramji, Siddarth; Shankaran, Seetha
title: Hypothermia for moderate or severe neonatal encephalopathy in low-income and middle-income countries (HELIX): a randomised controlled trial in India, Sri Lanka, and Bangladesh
date: 2021-08-03
journal: Lancet Glob Health
DOI: 10.1016/s2214-109x(21)00264-3
sha: a02ca4009fcfa5ad083a4e1d510a55555e1f508d
doc_id: 842634
cord_uid: r3u9lv8c

BACKGROUND: Although therapeutic hypothermia reduces death or disability after neonatal encephalopathy in high-income countries, its safety and efficacy in low-income and middle-income countries is unclear. We aimed to examine whether therapeutic hypothermia alongside optimal supportive intensive care reduces death or moderate or severe disability after neonatal encephalopathy in south Asia. METHODS: We did a multicountry open-label, randomised controlled trial in seven tertiary neonatal intensive care units in India, Sri Lanka, and Bangladesh. We enrolled infants born at or after 36 weeks of gestation with moderate or severe neonatal encephalopathy and a need for continued resuscitation at 5 min of age or an Apgar score of less than 6 at 5 min of age (for babies born in a hospital), or both, or an absence of crying by 5 min of age (for babies born at home). Using a web-based randomisation system, we allocated infants into a group receiving whole body hypothermia (33·5°C) for 72 h using a servo-controlled cooling device, or to usual care (control group), within 6 h of birth. All recruiting sites had facilities for invasive ventilation, cardiovascular support, and access to 3 Tesla MRI scanners and spectroscopy. Masking of the intervention was not possible, but those involved in the magnetic resonance biomarker analysis and neurodevelopmental outcome assessments were masked to the allocation. The primary outcome was a combined endpoint of death or moderate or severe disability at 18–22 months, assessed by the Bayley Scales of Infant and Toddler Development (third edition) and a detailed neurological examination. Analysis was by intention to treat. This trial is registered with ClinicalTrials.gov, NCT02387385. FINDINGS: We screened 2296 infants between Aug 15, 2015, and Feb 15, 2019, of whom 576 infants were eligible for inclusion. After exclusions, we recruited 408 eligible infants and we assigned 202 to the hypothermia group and 206 to the control group. Primary outcome data were available for 195 (97%) of the 202 infants in the hypothermia group and 199 (97%) of the 206 control group infants. 98 (50%) infants in the hypothermia group and 94 (47%) infants in the control group died or had a moderate or severe disability (risk ratio 1·06; 95% CI 0·87–1·30; p=0·55). 84 infants (42%) in the hypothermia group and 63 (31%; p=0·022) infants in the control group died, of whom 72 (36%) and 49 (24%; p=0·0087) died during neonatal hospitalisation. Five serious adverse events were reported: three in the hypothermia group (one hospital readmission relating to pneumonia, one septic arthritis, and one suspected venous thrombosis), and two in the control group (one related to desaturations during MRI and other because of endotracheal tube displacement during transport for MRI). No adverse events were considered causally related to the study intervention. INTERPRETATION: Therapeutic hypothermia did not reduce the combined outcome of death or disability at 18 months after neonatal encephalopathy in low-income and middle-income countries, but significantly increased death alone. Therapeutic hypothermia should not be offered as treatment for neonatal encephalopathy in low-income and middle-income countries, even when tertiary neonatal intensive care facilities are available. FUNDING: National Institute for Health Research, Garfield Weston Foundation, and Bill & Melinda Gates Foundation. TRANSLATIONS: For the Hindi, Malayalam, Telugu, Kannada, Singhalese, Tamil, Marathi and Bangla translations of the abstract see Supplementary Materials section.

Index Table 1 Data are mean (standard deviation), median [inter-quartile range] plus median change (95% confidence intervals), or number (percentage) plus risk difference (95% confidence intervals). The laboratory data are based on the worst value between 24 to 48 hours after birth and were analysed at standard laboratory conditions without any temperature correction. Seizures are based on the number of babies who had clinical seizures between 24 to 48 hours after birth.

WBC: white blood cell count; CRP: C-reactive protein; pCO2: partial pressure of carbon dioxide; CPAP: continuous positive airway pressure; IVF: intravenous fluids; NG: Nasogastric feeds Data are number (%) or Odds ratio (OR) (95% Confidence interval: CI). *Odds ratio for MRI abnormalities score in cooled relative to non-cooled infants from ordinal logistic regression analysis. Basal ganglia and thalamic score: 0=normal, 1=mild (focal abnormal signal intensity), 2=moderate (multifocal abnormal signal intensity), 3=severe (widespread abnormal signal intensity). White matter score: 0=normal, 1=mild (exaggerated long T1 and long T2 in periventricular white matter only), 2=moderate (long T1 and long T2 extending out to subcortical white matter and /or focal punctate lesions or focal area of infarction), 3=severe widespread abnormalities including overt infarction, haemorrhage, and long T1 and long T2. Cortical involvement was scored as the presence of abnormal signal intensity, usually decreased T1 or cortical highlighting. 0=normal, 1=mild (1-2 sites involved), 2=moderate (3 sites involved), 3=severe (more than 3 sites involved). Error bars indicate 95% confidence intervals in the hypothermic (blue) and control (red) group. Quality assurance:

The inter-site variation in MRS measurements was quantified across all scanners using a spherical phantom containing a buffered solution of 10mM N-acetylaspartate (NAA) and 10mM lactate (Lac). In vivo quality assurance included the visual assessment of motion from imaging before and after MRS; protocol adherence; manual rejection of motion corrupted sub-spectra; automated spectral corrections for both frequency and phase; and the rejection of spectra with linewidths outside the normal distribution of the full dataset.

MR spectroscopy data was first analysed in LCModel (Provencher, 2001) with the basis set, optimised for the timings of the Philips implementation and with ideal RF pulses. This was then repeated in LCModel with the STEAM protocol, generating a basis set with the same 1Hz Gaussian lineshape in VeSPA-Simulate (Soher et al., 2011) , and including the same list of metabolites (with NAA, NAAG, Cho, Cr and PCr methyl peaks). From the STEAM series, the same relaxivity corrections were applied to the metabolite signals, and water signals.

All water suppressed spectra were analysed using LCModel (v6·3-1J), with basis sets simulated using VeSPA (v0·9·11) with ideal RF pulses according to the PRESS (TE = 288 ms) and STEAM (TE = 20 ms) sequence timings employed by each vendor for each acquisition (personal communication). The following metabolites were included in the simulations and analyses: acetate (Act), alanine (Ala), ascorbate (Asc), betaine (Bet), aspartate (Asp), choline (Cho), phosphocholine (PCh), glycerophosphocholine (GPC), creatine (Cr), phosphocreatine (PCr), gamma-aminobutyric acid (GABA), glucose (Glc), glutamate (Glu), glutamine (Gln), glutathione (GSH), glycine (Glyc), lactate (Lac), myo-inositol (mIns), N-acetylaspartate (NAA), N-acetylaspartyl glutamate (NAAG), phosphoethanolamine (PE), propylene glycol (PGC), scyllo-inositol (Scyllo), taurine (Tau), and threonine (Thr).

Specific LCModel control parameters were, for PRESS (TE=288 ms):

And for STEAM (TE=20 ms):

The methyl peaks of NAA, NAAG, choline (Cho), phosphocreatine (PCr), and creatine (Cr) were separated from other groups in the basis spectra to allow quantification of individual relaxation rates. NAA+NAAG methyl peaks at ~2·0ppm were combined and referred to as 'NAA', and PCr+Cr methyl peaks at ~3·0ppm were combined and referred to as 'Cr' due to strong covariance. Lac+Thr were combined and referred to as 'Lac' due to strong covariance. A single peak was used to fit the choline signal at ~3·2ppm and referred to as 'Cho'. Water unsuppressed signals were quantified using HLSVD.

NAA/Cho, NAA/Cr and Lac/NAA were all derived from the LCModel fitted results of first water suppressed PRESS acquisition only (TR/TE=2290ms/288ms) ( Figure S1a ).

[tNAA] was calculated from the fitted NAA methyl singlets through two STEAM experiments (TE=20 ms; TM= 20ms, TR=1500; 5000 (or 3500 ms) acquired for T1 and T2 corrections) (Figure S1 b and c) , comparing the relaxation corrected NAA signal to the relaxation corrected unsuppressed water signal. This process is outlined in the following Figure 

Two-Point Saturation Recovery STEAM: To accurately measure changes in the parenchyma water concentration, !2 relaxometry of the water signal was performed using saturation recovery experiment. In-vivo experiments, were optimised with two-point saturation recovery by choosing an appropriate pair of TRs for accurate and precise measurement of metabolite concentrations. A pair of TRs sufficiently long to ensure near fully recovered Lip/MM signals (i.e. TR≥1500ms) were used.

Choice of acquisition scheme for tNAA concentration: This protocol comprises 10 separate STEAM acquisitions in the same 15×15×15mm 3 thalamic voxel. In this study scan duration (approximately 13 minutes) for NAA concentration estimation is significantly shorter than in the PRESS protocol (~25 min) for the previous MARBLE study.

In the present two-point saturation recovery STEAM protocol (Table S2) , steps 1 and 3 comprise a two-point !1 relaxometry experiment with TR 1500ms and 5000ms (or 3500ms) for each metabolite, allowing the calculation of fully !1 relaxed signal intensities. The short TE of 20ms ensures that differential !2 decay effects between metabolites and water are minimised. Steps 2 and 4 allow eddy current correction of steps 1 and 3, and can be combined to determine the fully !1 relaxed brain water signal under the same phase cycling scheme as steps 1 and 3.

Although STEAM acquisitions provide a lower inherent signal-to-noise ratio than PRESS, they are less susceptible to chemical shift displacement artefacts, and can achieve a shorter TE to minimise variation due to the variance !2 relaxation effects. Though, the use of a short TE complicates the fitting of key metabolites due to overlapping with broader resonances from lipids and macromolecules, their short !1 ensures that they are fully relaxed throughout the present experiment. The estimates of the NAA concentration and !1 of acetate were in agreement between PRESS (used in our earlier MARBLE study) and present STEAM protocols. Within normal variation and mild pathology, the PRESS and STEAM protocols are comparable in their measurements of [tNAA] in the neonatal brain, although STEAM tends to overestimate this.

Optimising Two-Point Saturation Recovery: Initial few STEAM acquisitions (total 07 included in analysis) were taken with TR =1500 ms and 3500ms. The protocol was further optimised by increasing TR from 3500ms to 5000ms. By increasing the TR of the second acquisition in the two-point saturation recovery experiment, the estimates of fully relaxed signals and !1s are less biased and more precise, which is a desirable property for robust estimation of metabolite concentrations. With the longer TR at 3500ms, estimates from the two acquisitions were sensitive to the signal to noise ratio, however, with 5000ms, fully !1 relaxed signal intensities are largely independent of the error at the shorter TR. In the proposed STEAM protocol, the Lip/MM resonances are fully relaxed due to !1 effects throughout the experiment. A TR of 5000ms was also observed as a reasonable balance of signal-to-noise ratio efficiency and overall fitting accuracy, while also allowing effective post-hoc motion correction on the collected sub-spectra. DTI acquisition in 30 diffusion gradient directions were carried out using 2D spin-echo echo-planar imaging with 31 total images per slice (30 b = 750 s/mm 2 and 1 b = 0 s/mm 2 ), TE = 79 ms, 1·95×1·95×2 mm 3 voxel size and parallel imaging acceleration factor of 2·0.

Data was assessed for excessive motion and acquisition artefacts, both before and after isolating the brain signal and undertaking corrections for eddy currents and motion using FSL.

Diffusion tensors were then calculated in FSL, and datasets of sufficient quality were spatially normalised to skull stripped JHU_neonate_SS_fass FA map. Major white matter tracts were selected by thresholding according to fractional anisotropy (FA > 0·15) and skeletonised using FSL.

To select a region of interest in the posterior limbs of the internal capsule, a binarized mask was extracted using JHU neonatal brain atlas and placed over mean FA map, as shown below:

The mean FA and standard deviation within this region of interest was then used for assessing prognostic accuracy.

Tract-based spatial statistics (FSL) was then used to examine relationships between FA values in Hypothermia vs control study groups, estimated across all of the major white matter tracts included in the mean skeleton. These analyses were corrected for multiple comparisons using threshold-free cluster enhancement. 

Small for gestational age was defined as birth weight less than 2 standard deviations on the World Health Organisation growth chart Figure 11 . Effect of hypothermia on mortality at hospital discharge: Subgroup analysis based on co-existent sepsis. The centres are arranged in the order of increasing mortality in the left panel. The centers with higher mortality appears to have lower rates of disability.

-vivo spectra: (a) PRESS: TE 288ms

STEAM: TE 20ms

STEAM: TE 20ms TR 5000ms

(a) The PRESS TE 288ms, TR 2290ms acquisition protocol for peak area ration (Lac/NAA, NAA/Cr, NAA/Cho) incorporated the T1 or T2 relaxation effects in the observed metabolites. (b) STEAM TE 20ms TR 1500ms (c) STEAM TE 20ms TR