The current publication is certainly not the first suggestion that VZV or other infections might be important in the pathogenesis of GCA. In fact, as the authors acknowledge, at least seven prior studies have searched for VZV in arteries affected by GCA.5 Many of these studies were negative. For example, in 2014, Ami Bhatt, MD, and colleagues analyzed 12 formalin-fixed temporal artery biopsy specimens with GCA for microbes using comprehensive DNA sequencing compared to five controls.6 They did not find pathogenic viruses, including VZV.
In 2004, Rodriguez-Pla and colleagues used PCR to search for VZV DNA in 50 temporal artery biopsies showing histologic features of GCA.7 None were positive.
On the other hand, some smaller studies have suggested an association between VZV and GCA. For example, in a 2001 study of 35 specimens of GCA, 26% showed evidence of VZV using PCR compared with none in controls.8
Sampling Error?
The question remains whether the current study, which is larger than any of the previous investigations, definitively advances our understanding of GCA or is yet another inconclusive finding in the long-running debate about GCA and VZV.
One of the strengths of this study is how comprehensively the histological sections were analyzed. The authors not only studied sections showing GCA pathology but also analyzed the skip areas for the presence of VZV. As it turned out, VZV-positive regions tended to be adjacent to GCA pathology rather than being part of it. Secondly, the authors performed PCR selectively on regions that were positive for VZV using immunohistochemistry techniques. Because VZV-positive regions have intervening gaps, it’s possible the mixture of positive and negative VZV PCR publications reflects the substantial sampling error when trying to detect VZV. The one electron micrograph presented that shows enveloped viral particles is hard to ignore.
However, the study is cross-sectional, and we can say that there may be an association between VZV and GCA rather than causation. Also, the controls were from autopsy and obtained from different conditions than the temporal arteritis samples.
If we are to believe the authors’ argument that VZV is causative of GCA secondary to VZV reactivation in the cranial ganglia followed by axonal spread to the arterial adventitia, how do we reconcile the apparent brisk therapeutic response of GCA to high-dose corticosteroids? Although it’s possible the clinical symptoms are caused primarily by an inflammatory reaction to a self-limited viral reactivation, one would expect that treatment with steroids would ultimately lead to more disseminated infection. Examples of steroid-refractory GCA exist, but in the majority of cases, treatment with steroids over months leads to control of symptoms rather than deterioration.