Prognostic accuracy of head computed tomography for prediction of functional outcome after out-of-hospital cardiac arrest: Rationale and design of the prospective TTM2-CT-substudy

Abstract

Background: Head computed tomography (CT) is a guideline recommended method to predict functional outcome after cardiac arrest (CA), but standardized criteria for evaluation are lacking. To date, no prospective trial has systematically validated methods for diagnosing hypoxic-ischaemic encephalopathy (HIE) on CT after CA. We present a protocol for validation of pre-specified radiological criteria for assessment of HIE on CT for neuroprognostication after CA.

Methods/design: This is a prospective observational international multicentre substudy of the Targeted Hypothermia versus Targeted Normothermia after out-of-hospital cardiac arrest (TTM2) trial. Patients still unconscious 48 hours post-arrest at 13 participating hospitals were routinely examined with CT. Original images will be evaluated by examiners blinded to clinical data using a standardized protocol. Qualitative assessment will include evaluation of absence/presence of "severe HIE". Radiodensities will be quantified in pre-specified regions of interest for calculation of grey-white matter ratios (GWR) at the basal ganglia level. Functional outcome will be dichotomized into good (modified Rankin Scale 0-3) and poor (modified Rankin Scale 4-6) at six months post-arrest. Prognostic accuracies for good and poor outcome will be presented as sensitivities and specificities with 95% confidence intervals (using pre-specified cut-offs for quantitative analysis), descriptive statistics (Area Under the Receiver Operating Characteristics Curve), inter- and intra-rater reliabilities according to STARD guidelines.

Conclusions: The results from this prospective trial will validate a standardized approach to radiological evaluations of HIE on CT for prediction of functional outcome in comatose CA patients.The TTM2 trial and the TTM2 CT substudy are registered at ClinicalTrials.gov NCT02908308 and NCT03913065.

Keywords: Cardiac arrest; Computed tomography; GWR grey-white matter ratio; Hypoxic-ischaemic encephalopathy (HIE); Neuroprognostication; Outcome; Targeted temperature management.

© 2022 The Authors.

Figures

Fig. 1

Flowchart patient inclusion for CT substudy CT, head computed tomography; N, number of patients; h, hours.

Fig. 2

SOP checklist for qualitative analysis Standardised operating procedure checklist for qualitative radiological evaluation. CSF; cerebrospinal fluid, SAH; subarachnoid haemorrhage, HIE; hypoxic-ischaemic encephalopathy.

Fig. 3

SOP for the qualitative measurement of grey-white matter ratio Placement of Regions of Interest (ROI) for determination of the grey-white matter ratio (GWR) at the basal ganglia level including 8 or 4 ROIs., Yellow indicates white matter ROIs and blue indicates ROIs in the grey matter. All axial slices containing basal ganglia structures should be evaluated and ROIs placed bilaterally in the slice best representative of that target region. Thus, these 8 ROIs may be placed in different slices: 1) Putamen (PU); 2) Head of the caudate nucleus (CN); 3) Posterior limb of the internal capsule (PIC), and 4) Genu of the corpus callosum (CC). In case of complete loss of grey-white distinction, use ventricles and midline as landmarks. In some patients with severe HIE radiodensity is similar in grey and white matter, exact location of target regions cannot always be determined. Nonetheless, ROIs should be placed and patients should not be excluded from GWR determination. Crosses indicate the ROIs included in each grey-white matter ratio method: 8 BG (basal ganglia model), 4SI (simple model) and auto GWR (automated GWR determination).