Identification of Herpes Zoster–Associated Temporal Arteritis Among Cases of Giant Cell Arteritis

Erin M. Buckingham; Maria A. Foley; Charles Grose; Nasreen A. Syed; Morton E. Smith; Todd P. Margolis; Matthew J. Thurtell; Randy Kardon

doi:10.1016/j.ajo.2017.12.017

Identification of Herpes Zoster–Associated Temporal Arteritis Among Cases of Giant Cell Arteritis

Erin M. Buckingham, Maria A. Foley, Charles Grose, Nasreen A. Syed, Morton E. Smith, Todd P. Margolis, Matthew J. Thurtell, Randy Kardon

Research output: Contribution to journal › Article › peer-review

33 Scopus citations

Abstract

Purpose: To examine whether herpes zoster antigen (also called varicella-zoster virus antigen) was detectable in temporal artery biopsies taken from individuals with giant cell arteritis (GCA). Design: Retrospective comparative case series. Methods: Sections of formalin-fixed paraffin-embedded temporal arteries were examined first by hematoxylin-eosin (H&E) staining to establish the diagnosis of GCA. Adjacent sections of the same biopsy were then examined by immunohistochemistry, using 2 different monoclonal antibodies against a major antigen of varicella-zoster virus called gE. Pathologic specimens were obtained from patients cared for at the University of Iowa and Washington University in St. Louis ophthalmology clinics. Results: The study included biopsies from 25 patients with symptoms of GCA as well as positive H&E pathology and 25 patients with symptoms compatible with GCA but negative H&E pathology. Among the GCA-positive group, 3 patients had positive staining for herpes zoster antigen. Among the GCA-negative group, herpes zoster antigen was not detected in any biopsy. In both groups of patients, false-positive staining for herpes zoster antigen was detected in the presence of calcifications in the arteries. False-positive staining was also detected on some extra-arterial skeletal muscle and erythrocytes. Conclusion: Herpes zoster antigen was detected in 3 of 25 temporal arteries from patients with biopsy-proven GCA. One of the 3 positive cases was noteworthy because the patient had had herpes zoster ophthalmicus diagnosed 3 weeks before the onset of GCA symptoms. False-positive staining for herpes zoster antigen was detected on several temporal artery biopsies.

Original language	English
Pages (from-to)	51-60
Number of pages	10
Journal	American journal of ophthalmology
Volume	187
DOIs	https://doi.org/10.1016/j.ajo.2017.12.017
State	Published - Mar 2018

Access to Document

10.1016/j.ajo.2017.12.017

Cite this

@article{4a54fd73e8154ae8a2ea5d21464016e3,

title = "Identification of Herpes Zoster–Associated Temporal Arteritis Among Cases of Giant Cell Arteritis",

abstract = "Purpose: To examine whether herpes zoster antigen (also called varicella-zoster virus antigen) was detectable in temporal artery biopsies taken from individuals with giant cell arteritis (GCA). Design: Retrospective comparative case series. Methods: Sections of formalin-fixed paraffin-embedded temporal arteries were examined first by hematoxylin-eosin (H&E) staining to establish the diagnosis of GCA. Adjacent sections of the same biopsy were then examined by immunohistochemistry, using 2 different monoclonal antibodies against a major antigen of varicella-zoster virus called gE. Pathologic specimens were obtained from patients cared for at the University of Iowa and Washington University in St. Louis ophthalmology clinics. Results: The study included biopsies from 25 patients with symptoms of GCA as well as positive H&E pathology and 25 patients with symptoms compatible with GCA but negative H&E pathology. Among the GCA-positive group, 3 patients had positive staining for herpes zoster antigen. Among the GCA-negative group, herpes zoster antigen was not detected in any biopsy. In both groups of patients, false-positive staining for herpes zoster antigen was detected in the presence of calcifications in the arteries. False-positive staining was also detected on some extra-arterial skeletal muscle and erythrocytes. Conclusion: Herpes zoster antigen was detected in 3 of 25 temporal arteries from patients with biopsy-proven GCA. One of the 3 positive cases was noteworthy because the patient had had herpes zoster ophthalmicus diagnosed 3 weeks before the onset of GCA symptoms. False-positive staining for herpes zoster antigen was detected on several temporal artery biopsies.",

author = "Buckingham, {Erin M.} and Foley, {Maria A.} and Charles Grose and Syed, {Nasreen A.} and Smith, {Morton E.} and Margolis, {Todd P.} and Thurtell, {Matthew J.} and Randy Kardon",

note = "Funding Information: The virology laboratory at the University of Iowa Children's Hospital has had extensive experience over 3 decades with detection of viral antigens in biopsies of human tissues from patients with serious varicella and herpes zoster infections. 18,25,26 The current arterial analysis was undertaken in response to the recent publication of several articles from another university center, describing the detection of herpes zoster antigen in more than 70% of TA biopsies from patients with GCA. 14,27 In their first report, the group examined formalin-fixed paraffin-embedded temporal artery biopsies from 82 patients with diagnosed GCA as defined by positive histopathology. 14 They examined serial sections of these biopsies and found herpes zoster antigen by IHC in slides from 74% of the patients (61/82). When they looked at adjacent sections, they found GCA histopathology in 89% of the slides. In other words, there was a statistically significant correlation between sites of positive H&E histopathology for GCA and positive IHC for zoster antigen. The herpes zoster antigen was most common in the adventitia (49%), followed by media (32%) and intima (19%). Among the set of 61 zoster-positive biopsy samples, 32 also included adjacent skeletal muscle. Of these 32 cases, zoster antigen was detected by IHC in 38% of the skeletal muscles. In subsequent reports, the same group expanded their GCA and control populations. 27–29 Altogether, they reassessed a total of 100 temporal artery biopsies from patients with symptoms of GCA but with previously negative H&E histopathology for GCA. Among these 100 biopsies, they found 58% to be positive for zoster antigen. In contrast to the findings of this neurology group, we found herpes zoster antigen in only 3 of 25 TA biopsies from patients with GCA. When we performed a similar analysis of TA biopsies from 25 patients with symptoms of GCA but negative histopathology, we found no sample positive for VZV antigen, using the same MAb reagent as the neurology group. In contrast, we did find false-positive IHC tests for herpes zoster antigen in the presence of calcifications in temporal arteries from both the GCA-positive and GCA-negative patients, a point that was not mentioned in prior publications. 14,27 This false-positive result obviously would increase the apparent number of VZV-positive temporal artery biopsies. Calcifications have been observed in TA biopsies by several ocular pathology laboratories, but calcifications are often not mentioned in recent publications about GCA because they are not thought to play a major role in pathogenesis of GCA. 30 In one informative analysis of GCA histologic features, calcifications were found in 37% of temporal arteries when no ischemic signs were present and in 53% of temporal arteries when ischemic manifestations were documented. 31 In a second, similar analysis, calcifications were found in 26% of temporal artery biopsies from patients with the clinical features of GCA but negative histopathology. 32 As noted, we observed easily detectable false-positive IHC reactivity at sites of calcification; these false-positive results occurred even in IHC tests with no primary antibody. Without observing adjacent H&E-stained sections to identify sites of calcifications, we would not have been able to discriminate false-positive staining of calcifications from true-positive staining of herpes zoster antigen in areas lacking calcifications. We also observed false-positive IHC reactivity on skeletal muscle. We are confident that attachment of VZV antibody to skeletal muscle is a false-positive result for several reasons. First of all, the prototypic alpha herpesvirus HSV-1 cannot spread through human skeletal muscle tissue. Many experiments have been performed with HSV-1, as part of protocols to determine if recombinant HSV-1 vectors could be delivered into mature adult muscle to treat human muscular dystrophies. Numerous experiments demonstrated that mature basal lamina acts as a physical barrier to HSV-1 infection of adult myofibers. 33,34 HSV-1 has a much wider tropism for different tissues that VZV. Further, adenovirus does not infect and spread in adult human skeletal muscle tissue. 35 The fact that HSV-1 and adenovirus cannot infect adult muscle fibers leads to a conclusion that VZV is unlikely to infect adult skeletal muscle fibers. In a prior analysis of TA biopsies, skeletal muscle was found in 32 of 36 specimens; VZV antigen was detected in skeletal muscle of 12 of 32 (38%) samples. 14 In our investigation, we observed that our control antibody against reovirus antigen also attached to skeletal muscle. Based on this reanalysis of antibody binding to skeletal muscle, we conclude that VZV positivity in skeletal muscle is more likely to be a false-positive than a true-positive. We made no attempt to detect VZV DNA in our artery biopsies for 4 reasons. First, in prior studies 14 the percent of arteries positive for VZV DNA was always lower than the percent positive by IHC; and secondly, in earlier studies from other centers, 2 different virology laboratories found artery biopsy samples from 65 GCA patients to be uniformly negative for VZV DNA. 36,37 As a third reason, the temporal arteries are known to be innervated by nerve fibers from ganglia that harbor latent VZV. 12 Therefore, at any time a small number of temporal arteries may be positive for VZV DNA. Yet, in the absence of any VZV antigen in the artery, we do not think that VZV infection could elicit the immunopathology for which giant cell arteritis is known. As a fourth reason, we point out that the published studies have been retrospective. Therefore, all laboratories used formalin-fixed paraffin-embedded biopsies of temporal arteries. These tissues may be suboptimal for performance of polymerase chain reaction (PCR) testing. 38 To perform optimal PCR testing, we would need to organize a prospective study in which the temporal artery biopsies were divided immediately after surgery. A piece that was not formalin fixed would be used instead for DNA extraction with subsequent PCR analysis. The concept that VZV reactivation, either as herpes zoster ophthalmicus or as zoster sine herpete, can cause occasional ocular disease is not a new concept. 4 There are several papers in the literature describing VZV-related corneal disease, iridocyclitis, and retinitis in the absence of a clinically apparent skin eruption. 39–41 In a report of 13 cases of VZV-related acute retinal necrosis, all occurred in the absence of a vesicular skin disease. 42 The cases were diagnosed by PCR amplification of VZV DNA in vitreous and aqueous specimens. If we remove the patients with an underlying immunodeficiency, the ages of the adult patients with VZV-related retinal necrosis were 39, 48, 58, 67, and 72 years, a clear overlap with the age group most commonly associated with GCA. Thus, VZV appears able to enter any tissue in the head and neck that is innervated by nerves originating from ganglia containing latent VZV. Our Case #3, of a 78-year-old woman with clinically apparent herpes zoster ophthalmicus and GCA 3 weeks later, may provide an important confirmatory link to this association. However, in the current survey we were not able to document an association between GCA-positive histopathology and herpes zoster antigen positivity in the majority of TA biopsies from GCA patients. Therefore, we do not recommend that all patients with GCA be treated with antivirals, unless there is an immediate past history of documented herpes zoster. Funding/Support: Research on GCA at the University of Iowa was supported by grants from the Iowa City VA Center for the Prevention and Treatment of Visual Loss C-9251-C RR&D and the National Shingles Foundation, New York, New York. Varicella reagents were produced during research supported by NIH grant AI89716, Bethesda, Maryland. Research at Washington University in St. Louis was supported in part by an NIH Core Grant (EY002687) and an unrestricted grant from Research to Prevent Blindness to the Department of Ophthalmology and Visual Sciences. Financial Disclosures: None of the authors has any financial disclosures. The authors attest that they meet the current ICMJE criteria for authorship. Publisher Copyright: {\textcopyright} 2017 The Author(s)",

year = "2018",

month = mar,

doi = "10.1016/j.ajo.2017.12.017",

language = "English",

volume = "187",

pages = "51--60",

journal = "American journal of ophthalmology",

issn = "0002-9394",

}

TY - JOUR

T1 - Identification of Herpes Zoster–Associated Temporal Arteritis Among Cases of Giant Cell Arteritis

AU - Buckingham, Erin M.

AU - Foley, Maria A.

AU - Grose, Charles

AU - Syed, Nasreen A.

AU - Smith, Morton E.

AU - Margolis, Todd P.

AU - Thurtell, Matthew J.

AU - Kardon, Randy

N1 - Funding Information: The virology laboratory at the University of Iowa Children's Hospital has had extensive experience over 3 decades with detection of viral antigens in biopsies of human tissues from patients with serious varicella and herpes zoster infections. 18,25,26 The current arterial analysis was undertaken in response to the recent publication of several articles from another university center, describing the detection of herpes zoster antigen in more than 70% of TA biopsies from patients with GCA. 14,27 In their first report, the group examined formalin-fixed paraffin-embedded temporal artery biopsies from 82 patients with diagnosed GCA as defined by positive histopathology. 14 They examined serial sections of these biopsies and found herpes zoster antigen by IHC in slides from 74% of the patients (61/82). When they looked at adjacent sections, they found GCA histopathology in 89% of the slides. In other words, there was a statistically significant correlation between sites of positive H&E histopathology for GCA and positive IHC for zoster antigen. The herpes zoster antigen was most common in the adventitia (49%), followed by media (32%) and intima (19%). Among the set of 61 zoster-positive biopsy samples, 32 also included adjacent skeletal muscle. Of these 32 cases, zoster antigen was detected by IHC in 38% of the skeletal muscles. In subsequent reports, the same group expanded their GCA and control populations. 27–29 Altogether, they reassessed a total of 100 temporal artery biopsies from patients with symptoms of GCA but with previously negative H&E histopathology for GCA. Among these 100 biopsies, they found 58% to be positive for zoster antigen. In contrast to the findings of this neurology group, we found herpes zoster antigen in only 3 of 25 TA biopsies from patients with GCA. When we performed a similar analysis of TA biopsies from 25 patients with symptoms of GCA but negative histopathology, we found no sample positive for VZV antigen, using the same MAb reagent as the neurology group. In contrast, we did find false-positive IHC tests for herpes zoster antigen in the presence of calcifications in temporal arteries from both the GCA-positive and GCA-negative patients, a point that was not mentioned in prior publications. 14,27 This false-positive result obviously would increase the apparent number of VZV-positive temporal artery biopsies. Calcifications have been observed in TA biopsies by several ocular pathology laboratories, but calcifications are often not mentioned in recent publications about GCA because they are not thought to play a major role in pathogenesis of GCA. 30 In one informative analysis of GCA histologic features, calcifications were found in 37% of temporal arteries when no ischemic signs were present and in 53% of temporal arteries when ischemic manifestations were documented. 31 In a second, similar analysis, calcifications were found in 26% of temporal artery biopsies from patients with the clinical features of GCA but negative histopathology. 32 As noted, we observed easily detectable false-positive IHC reactivity at sites of calcification; these false-positive results occurred even in IHC tests with no primary antibody. Without observing adjacent H&E-stained sections to identify sites of calcifications, we would not have been able to discriminate false-positive staining of calcifications from true-positive staining of herpes zoster antigen in areas lacking calcifications. We also observed false-positive IHC reactivity on skeletal muscle. We are confident that attachment of VZV antibody to skeletal muscle is a false-positive result for several reasons. First of all, the prototypic alpha herpesvirus HSV-1 cannot spread through human skeletal muscle tissue. Many experiments have been performed with HSV-1, as part of protocols to determine if recombinant HSV-1 vectors could be delivered into mature adult muscle to treat human muscular dystrophies. Numerous experiments demonstrated that mature basal lamina acts as a physical barrier to HSV-1 infection of adult myofibers. 33,34 HSV-1 has a much wider tropism for different tissues that VZV. Further, adenovirus does not infect and spread in adult human skeletal muscle tissue. 35 The fact that HSV-1 and adenovirus cannot infect adult muscle fibers leads to a conclusion that VZV is unlikely to infect adult skeletal muscle fibers. In a prior analysis of TA biopsies, skeletal muscle was found in 32 of 36 specimens; VZV antigen was detected in skeletal muscle of 12 of 32 (38%) samples. 14 In our investigation, we observed that our control antibody against reovirus antigen also attached to skeletal muscle. Based on this reanalysis of antibody binding to skeletal muscle, we conclude that VZV positivity in skeletal muscle is more likely to be a false-positive than a true-positive. We made no attempt to detect VZV DNA in our artery biopsies for 4 reasons. First, in prior studies 14 the percent of arteries positive for VZV DNA was always lower than the percent positive by IHC; and secondly, in earlier studies from other centers, 2 different virology laboratories found artery biopsy samples from 65 GCA patients to be uniformly negative for VZV DNA. 36,37 As a third reason, the temporal arteries are known to be innervated by nerve fibers from ganglia that harbor latent VZV. 12 Therefore, at any time a small number of temporal arteries may be positive for VZV DNA. Yet, in the absence of any VZV antigen in the artery, we do not think that VZV infection could elicit the immunopathology for which giant cell arteritis is known. As a fourth reason, we point out that the published studies have been retrospective. Therefore, all laboratories used formalin-fixed paraffin-embedded biopsies of temporal arteries. These tissues may be suboptimal for performance of polymerase chain reaction (PCR) testing. 38 To perform optimal PCR testing, we would need to organize a prospective study in which the temporal artery biopsies were divided immediately after surgery. A piece that was not formalin fixed would be used instead for DNA extraction with subsequent PCR analysis. The concept that VZV reactivation, either as herpes zoster ophthalmicus or as zoster sine herpete, can cause occasional ocular disease is not a new concept. 4 There are several papers in the literature describing VZV-related corneal disease, iridocyclitis, and retinitis in the absence of a clinically apparent skin eruption. 39–41 In a report of 13 cases of VZV-related acute retinal necrosis, all occurred in the absence of a vesicular skin disease. 42 The cases were diagnosed by PCR amplification of VZV DNA in vitreous and aqueous specimens. If we remove the patients with an underlying immunodeficiency, the ages of the adult patients with VZV-related retinal necrosis were 39, 48, 58, 67, and 72 years, a clear overlap with the age group most commonly associated with GCA. Thus, VZV appears able to enter any tissue in the head and neck that is innervated by nerves originating from ganglia containing latent VZV. Our Case #3, of a 78-year-old woman with clinically apparent herpes zoster ophthalmicus and GCA 3 weeks later, may provide an important confirmatory link to this association. However, in the current survey we were not able to document an association between GCA-positive histopathology and herpes zoster antigen positivity in the majority of TA biopsies from GCA patients. Therefore, we do not recommend that all patients with GCA be treated with antivirals, unless there is an immediate past history of documented herpes zoster. Funding/Support: Research on GCA at the University of Iowa was supported by grants from the Iowa City VA Center for the Prevention and Treatment of Visual Loss C-9251-C RR&D and the National Shingles Foundation, New York, New York. Varicella reagents were produced during research supported by NIH grant AI89716, Bethesda, Maryland. Research at Washington University in St. Louis was supported in part by an NIH Core Grant (EY002687) and an unrestricted grant from Research to Prevent Blindness to the Department of Ophthalmology and Visual Sciences. Financial Disclosures: None of the authors has any financial disclosures. The authors attest that they meet the current ICMJE criteria for authorship. Publisher Copyright: © 2017 The Author(s)

PY - 2018/3

Y1 - 2018/3

N2 - Purpose: To examine whether herpes zoster antigen (also called varicella-zoster virus antigen) was detectable in temporal artery biopsies taken from individuals with giant cell arteritis (GCA). Design: Retrospective comparative case series. Methods: Sections of formalin-fixed paraffin-embedded temporal arteries were examined first by hematoxylin-eosin (H&E) staining to establish the diagnosis of GCA. Adjacent sections of the same biopsy were then examined by immunohistochemistry, using 2 different monoclonal antibodies against a major antigen of varicella-zoster virus called gE. Pathologic specimens were obtained from patients cared for at the University of Iowa and Washington University in St. Louis ophthalmology clinics. Results: The study included biopsies from 25 patients with symptoms of GCA as well as positive H&E pathology and 25 patients with symptoms compatible with GCA but negative H&E pathology. Among the GCA-positive group, 3 patients had positive staining for herpes zoster antigen. Among the GCA-negative group, herpes zoster antigen was not detected in any biopsy. In both groups of patients, false-positive staining for herpes zoster antigen was detected in the presence of calcifications in the arteries. False-positive staining was also detected on some extra-arterial skeletal muscle and erythrocytes. Conclusion: Herpes zoster antigen was detected in 3 of 25 temporal arteries from patients with biopsy-proven GCA. One of the 3 positive cases was noteworthy because the patient had had herpes zoster ophthalmicus diagnosed 3 weeks before the onset of GCA symptoms. False-positive staining for herpes zoster antigen was detected on several temporal artery biopsies.

AB - Purpose: To examine whether herpes zoster antigen (also called varicella-zoster virus antigen) was detectable in temporal artery biopsies taken from individuals with giant cell arteritis (GCA). Design: Retrospective comparative case series. Methods: Sections of formalin-fixed paraffin-embedded temporal arteries were examined first by hematoxylin-eosin (H&E) staining to establish the diagnosis of GCA. Adjacent sections of the same biopsy were then examined by immunohistochemistry, using 2 different monoclonal antibodies against a major antigen of varicella-zoster virus called gE. Pathologic specimens were obtained from patients cared for at the University of Iowa and Washington University in St. Louis ophthalmology clinics. Results: The study included biopsies from 25 patients with symptoms of GCA as well as positive H&E pathology and 25 patients with symptoms compatible with GCA but negative H&E pathology. Among the GCA-positive group, 3 patients had positive staining for herpes zoster antigen. Among the GCA-negative group, herpes zoster antigen was not detected in any biopsy. In both groups of patients, false-positive staining for herpes zoster antigen was detected in the presence of calcifications in the arteries. False-positive staining was also detected on some extra-arterial skeletal muscle and erythrocytes. Conclusion: Herpes zoster antigen was detected in 3 of 25 temporal arteries from patients with biopsy-proven GCA. One of the 3 positive cases was noteworthy because the patient had had herpes zoster ophthalmicus diagnosed 3 weeks before the onset of GCA symptoms. False-positive staining for herpes zoster antigen was detected on several temporal artery biopsies.

UR - http://www.scopus.com/inward/record.url?scp=85041421723&partnerID=8YFLogxK

U2 - 10.1016/j.ajo.2017.12.017

DO - 10.1016/j.ajo.2017.12.017

M3 - Article

C2 - 29294312

AN - SCOPUS:85041421723

SN - 0002-9394

VL - 187

SP - 51

EP - 60

JO - American journal of ophthalmology

JF - American journal of ophthalmology

ER -

Identification of Herpes Zoster–Associated Temporal Arteritis Among Cases of Giant Cell Arteritis

Abstract

Access to Document

Other files and links

Fingerprint

Cite this