Computed tomographic angiography for the diagnosis of blunt carotid/vertebral artery injury: a note of caution

Abstract

Objective: Computed tomographic angiography (CTA) by 16-channel multidetector scanner is increasingly replacing conventional digital subtraction angiography (DSA) for diagnosing or excluding blunt carotid/vertebral injuries (BCVI). To date there has been only 1 study in which all patients received both examinations. That study reported a high accuracy for 16-detector CTA. The current prospective parallel comparative study aims at validating this high accuracy and examining the rates of evaluability of CTA performed with a 16-detector scanner with image reconstruction by modern imaging software.

Methods: Patients at risk for BCVI (facial/cervical-spinal fractures; unexplained neurologic deficit; anisocoria; lateral neck soft tissue injury; clinical suspicion) underwent both CTA (16-channel multidetector scanner) and DSA. Results of the 2 studies and the clinical course were prospectively recorded.

Results: During the 40-month study period ending March 2007, approximately 7000 blunt trauma patients were evaluated and of these 119 (1.7%) consecutive patients meeting inclusion criteria were screened by CTA. Ninety-two patients underwent confirmatory DSA. Twenty-three (22%) DSA identified 26 BCVI (vertebral, 13; carotid, 13). Among these 23 CTAs, 17 identified 19 BCVIs (vertebral, 10; carotid, 9) (true positives), and 6 failed to identify 7 BCVIs (vertebral, 3; carotid, 4) (false negatives). Sixty-nine of the 92 DSA were normal. Of these 69 CTAs, 10 were falsely suspicious for 11 BCVIs (vertebral, 7; carotid, 4) (false positives), and 56 were normal (true negatives). The remaining 3 CTAs were nonevaluable (mistimed contrast, 1; streak artifact, 2). Sixteen of 89 (18%) evaluable CTAs, were suboptimal (mistimed contrast, 9; streak artifacts, 4; motion artifact, 2; body habitus, 1). Excluding the 3 nonevaluable CTAs, the sensitivity, specificity, positive and negative predictive values of CTA for diagnosing or excluding BCVI were 74%, 86%, 65%, and 90% respectively. One patient with grade II carotid artery injuries (by CTA and DSA) on antiplatelet agent developed stroke related to carotid artery injuries.

Conclusions: Current CTA technology cannot reliably diagnose or exclude BCVI. Twenty percent of CTAs are either nonevaluable or suboptimal. Until more data are available and the technique is standardized, the current trend towards using CTA to screen for and/or diagnose these rare but potentially devastating injuries is dangerous.