LIVERPOOL — Cranial duplex ultrasound was more sensitive for confirming the diagnosis of giant cell arteritis (GCA) than temporal artery biopsy, which has been considered the gold standard, a researcher said here.
When compared with a clinical diagnosis of this large-vessel vasculitis at 3 months, the sensitivity of ultrasound was 81% compared with 53% for temporal artery biopsy, reported Adam P. Croft, PhD, of the University of Birmingham in England.
In addition, ultrasound had a specificity of 98%, a positive predictive value of 97%, and a negative predictive value of 88%.
In comparison, temporal artery biopsy had a specificity of 100%, a positive predictive value of 100%, and a negative predictive value of 47%, Croft reported at the annual meeting of the British Society for Rheumatology.
Giant cell arteritis is the most common primary systemic vasculitis. It has a predilection for the extracranial arteries and is characterized by granulomatous inflammation and a lymphocytic infiltrate.
“If left untreated, GCA progresses to an occlusive vasculopathy with downstream ischemic complications including stroke and acute, irreversible vision loss,” he said.
“However, the inflammation is segmental in nature, so if you stick a biopsy needle in randomly, as often happens in temporal artery biopsy, there’s a good chance you’ll end up with normal histology even in an affected artery,” he said.
Prompt initiation of steroids in patients with suspected GCA is sight-saving, but a challenge has been to determine which patients need to continue long-term steroid treatment and which can be safely discontinued at 3 months, avoiding unnecessary steroid exposure and toxicity if an alternate diagnosis is made.
Despite the fact that temporal artery biopsy has been considered the gold standard and is recommended by BSR guidelines for suspected GCA, the sensitivity has been quoted as being only about 40%.
In addition, temporal artery biopsy is a fairly straightforward procedure, but it’s not without risk, with possible morbidities such as permanent facial nerve injury and scalp necrosis, Croft pointed out.
“So we typically rely on the clinical diagnosis despite the guidelines,” and the decision on subsequent biopsy is clinician choice, he said.
Two decades ago the first reports of ultrasound for GCA suggested that typical findings included a hyperechoic halo sign, representing edema around the blood vessel that eventually can progress to arterial stenosis and even occlusion, he said.
Beginning in 2005, his group began performing ultrasound for patients with suspected GCA, and in this study they reviewed their findings to see how ultrasound affected real-life clinical decision making.
Clinical data were obtained from medical records, and a clinical diagnosis was made if three American College of Rheumatology (ACR) criteria were met. These criteria include age over 50, new-onset headache, abnormalities on palpation of the temporal artery, erythrocyte sedimentation rate over 50 mm/hour, and positive biopsy findings.
The investigators included 87 patients who had undergone the ultrasound, most of whom were male and whose ages were 71 to 75 years.
A total of 30 had the positive halo sign. About half of all the patients subsequently underwent biopsies, and about half of those were positive and half negative.
“What that means is that if you trust your biopsy result 100%, that would represent a 20% false-positive rate,” he explained.
Of the 57 patients with negative ultrasound findings, 11 went on to have biopsies. “Importantly, one did come back positive, which represented a true false-negative,” he noted.
The researchers also considered likelihood ratios for the ACR criteria and ultrasound for accurately confirming the GCA diagnosis at 3 months and found that a positive ACR result alone had a positive likelihood ratio of 2.18 (95% CI 1.32-3.61) and a negative likelihood ratio of 0.53 (95% CI 0.34-0.84).
For ultrasound, the positive likelihood ratio was 41 (95% CI 5.86-288) and the negative likelihood ratio was 0.2 (95% CI 0.1-0.39).
When both the ACR criteria and ultrasound were positive, the positive likelihood ratio was 25 (95% CI 3.56-182) and the negative likelihood ratio was 0.51 (95% CI 0.37-0.51).
“Therefore, if the pretest clinical probability is very high according to the ACR criteria, then a positive ultrasound is sufficiently high to rule in the diagnosis, and if you have a low pretest clinical probability and a negative test, it strongly rules out the diagnosis,” he said.
“So ultrasound is particularly useful in those patients at either end of the clinical spectrum, very high or very low, and perhaps we can avoid biopsies for those patients,” he said.
Limitations of ultrasound for GCA include the fact that it is investigator-dependent and usually requires someone with expertise in vascular ultrasound, and that the sensitivity falls rapidly after 4 to 5 days of steroid treatment, so it’s necessary to have the infrastructure to obtain scans early.