Perfusion computed tomography as a screening tool for pending delayed cerebral ischemia in comatose patients after aneurysmal subarachnoid hemorrhage: a retrospective cohort study

Abstract

Abstract Background/Objective Delayed cerebral ischemia (DCI) after aneurysmal subarachnoid hemorrhage (aSAH) is a severe complication often attributed to vasospasm. Endovascular rescue therapies (ERT) are available; however, pending DCI is difficult to diagnose in comatose patients. Therefore, symptomatic vasospasms (sVS) on perfusion computed tomography (P-CT) is often used as a surrogate. The overarching aim of this retrospective cohort study was to evaluate P-CT as a screening modality for pending DCI on days four- and eight after aSAH in comatose patients. The specific objectives were (i) to explore the modality's validity by comparing the proportion of patients who received ERT due to sVS to patients developing DCI, and (ii) to explore the modality’s ability to risk-stratify patients by calculating the negative predictive value (NPV) of a patient with normal screening not developing DCI and by comparing the negative post-test probability (NPP) of developing DCI to the presumed incidence of DCI. Methods Records from 56 comatose aSAH patients Jan. 2019 to Dec. 2021 receiving 98 P-CT screening scans were retrospectively reviewed. With DCI, sVS and ERT as primary outcome measures, event rate analysis was undertaken, and NPP and NPV was calculated. sVS was defined as arterial narrowing with corresponding hypoperfusion on P-CT. DCI was defined as new cerebral infarct, not attributable to other causes, occurring >48 hours <6 weeks post aSAH. Results The incidence of DCI was 40%. sVS was detected in 9% of P-CTs, involving six patients. Positive P-CTs were followed by digital subtraction angiography (DSA) in five patients with ERT eligible in three patients. The combined NPV of screening on days four- and eight was 0,65. Days four and eight had an isolated NPV of 0,63 and 0,61, respectively. The NPP was 0,35 for negative scans combined on days four- and eight. Days four and eight had a NPP of 0,37 and 0,39, respectively. Seven (13%) patients had manifest DCI on day 4 P-CT, and 12 (21%) on day 8 P-CT. Conclusions Screening for pending DCI with P-CT on days four- and eight often resulted in late detection and had a poor prognostic certainty. The negative post-test probability of developing DCI was approximately equal to the incidence. Based on our analysis, we cannot recommend using P-CT as a screening modality on days four and eight alone for screening purposes in this group of patients. There is a need for continuous monitoring modalities to timely predict sVS and prevent DCI