- Split View
-
Views
-
Cite
Cite
A. Bharadwaj, B. Dasgupta, K. Wolfe, C. Nordborg, E. Nordborg, Difficulties in the development of histological scoring of the inflamed temporal arteries in giant cell arteritis, Rheumatology, Volume 44, Issue 12, December 2005, Pages 1579–1580, https://doi.org/10.1093/rheumatology/kei180
- Share Icon Share
Sir, Giant cell arteritis (GCA) is a focal, granulomatous inflammation of medium and small-sized arteries in the elderly. Loss of vision occurs early and can be a presenting feature in 21.2% of patients [1]. The prognostic factors associated with neuro-ophthalmic complications are not entirely clear.
Temporal artery biopsy (TAB) is a clinically useful test with high predictive power [2]. A recent retrospective audit of 84 patients with TAB for suspected diagnosis of GCA at Southend General Hospital showed a high percentage of visual loss (29%) in biopsy-positive patients compared with biopsy-negative GCA.
The pathological hallmark of biopsy is granulomatous inflammation made of necrotic tissue, multinucleated giant cells, macrophages, lymphocytes and fibroblasts. There is some non-granulomatous infiltration of lymphocytes. The ischaemia causing the visual complication is mainly due to an occlusive vasculopathy rather than thrombosis. Usually the intima and media are avascular; hence neovascularization is an important pathological process, without which lumen obliteration is less likely [3]. Expression of vascular endothelial growth factor (VEGF) is associated with this process [4].
There is some evidence that the pathological findings in these biopsies are associated with the rate of ischaemic complications [5, 6]. There is also a suggestion of a difference between the cytokine profile in occlusive and non-occlusive disease. Those with significant intimal hyperplasia express higher levels of interferon-γ (secreted by T cells), interleukin-1β and the myofibroblast-stimulating platelet-derived growth factor. They also have a higher chance of having multinucleated giant cells, which secrete VEGF, allowing intimal growth.
It is therefore probable that tissue cytokine patterns in TAB correlate well with the clinical presentations and complications. Since the inflammatory cell infiltrate and cytokine production are closely associated and mutually interactive pathogenic processes we hypothesized that TAB may predict outcome in GCA. Our principal question therefore was whether histological examination with light microscopy could help differentiate patients at higher risk of ischaemic events, such as vision loss, from those with lower risk. We proposed a scoring system to quantify various aspects of the inflammatory response in TAB. This evolved after close discussion between histopathologists and rheumatologists.
The study was conducted in Southend General Hospital, UK in collaboration with Sahlgrenska University Hospital, Sweden. The TAB of 50 patients (25 from each centre) with biopsy-positive GCA was scored for these seven parameters: the general degree of inflammation; the extent of inflammatory invasion with regard to wall layers and the circumference; the presence and the extent of multinucleate giant cells; intimal thickening; fibrinous exudation and neovascularization (Table 1). Written consent was obtained from all the patients. The South Essex Local Research Ethics Committee approved the study. Each slide was scored twice (0–3) at an interval of 4 weeks. The slides were than exchanged between the two centres and scored again. The histopathologists were blinded to their own previous score, each other's scores and also to the clinical data. Weighted kappa statistics were used for intra- and interobserver reliability using the statistical software package SPSS.
. | Inflammation (general) . | Layers . | Circumference . | Giant cells . | Intimal proliferation . | Fibrous exudate . | Neovascularization . |
---|---|---|---|---|---|---|---|
Obs. Ia | 68.6 | 84.2 | 88.1 | 35.3 | 84.3 | 31.3 | 25.4 |
Obs. Ib | 66.6 | 86.2 | 90.1 | 49 | 84.2 | 37.2 | 41.2 |
Obs. IIa | 92.1 | 100 | 98 | 21.4 | 84.2 | 21.5 | 50.9 |
Obs. IIb | 92.1 | 100 | 99.9 | 19.6 | 84.2 | 23.4 | 41.1 |
. | Inflammation (general) . | Layers . | Circumference . | Giant cells . | Intimal proliferation . | Fibrous exudate . | Neovascularization . |
---|---|---|---|---|---|---|---|
Obs. Ia | 68.6 | 84.2 | 88.1 | 35.3 | 84.3 | 31.3 | 25.4 |
Obs. Ib | 66.6 | 86.2 | 90.1 | 49 | 84.2 | 37.2 | 41.2 |
Obs. IIa | 92.1 | 100 | 98 | 21.4 | 84.2 | 21.5 | 50.9 |
Obs. IIb | 92.1 | 100 | 99.9 | 19.6 | 84.2 | 23.4 | 41.1 |
. | Inflammation (general) . | Layers . | Circumference . | Giant cells . | Intimal proliferation . | Fibrous exudate . | Neovascularization . |
---|---|---|---|---|---|---|---|
Obs. Ia | 68.6 | 84.2 | 88.1 | 35.3 | 84.3 | 31.3 | 25.4 |
Obs. Ib | 66.6 | 86.2 | 90.1 | 49 | 84.2 | 37.2 | 41.2 |
Obs. IIa | 92.1 | 100 | 98 | 21.4 | 84.2 | 21.5 | 50.9 |
Obs. IIb | 92.1 | 100 | 99.9 | 19.6 | 84.2 | 23.4 | 41.1 |
. | Inflammation (general) . | Layers . | Circumference . | Giant cells . | Intimal proliferation . | Fibrous exudate . | Neovascularization . |
---|---|---|---|---|---|---|---|
Obs. Ia | 68.6 | 84.2 | 88.1 | 35.3 | 84.3 | 31.3 | 25.4 |
Obs. Ib | 66.6 | 86.2 | 90.1 | 49 | 84.2 | 37.2 | 41.2 |
Obs. IIa | 92.1 | 100 | 98 | 21.4 | 84.2 | 21.5 | 50.9 |
Obs. IIb | 92.1 | 100 | 99.9 | 19.6 | 84.2 | 23.4 | 41.1 |
Intraobserver reliability showed moderate correlation for the first observer [k = 0.59, 95% confidence interval (CI) 0.51–0.66] and good correlation (k = 0.77, 95% CI 0.71–0.84) for the second observer. Interobserver reliability showed poor agreement (k = 0.16, 95% CI 0.12–0.20). The intimal proliferation produced fair interobserver agreement (k = 0.34, 95% CI 0.19–0.48) with poor reliability for giant cells (k = −0.02, 95% CI −0.11 to 0.07) and neovascularization.
This study was carried out to see whether any distinct patterns emerged from a retrospective review of TABs. Any significant relationship between histopathological scoring and clinical outcome would have helped target aggressive therapy for patients with poor prognosis.
This study highlighted the difficulties in developing valid, reliable scoring for various diagnostic parameters in TAB. The involvement is segmental and non-uniform, making it difficult to grade the number of giant cells. Neovascularization is not always visible with haematoxylin and eosin stain in light microscopy and may need special stains for better characterization. It is felt that vague parameters like ‘general degree of inflammation’ are difficult to grade and should not be used for quantification. Different methods can be applied to reduce interobserver variability. More stringent definitions of criteria should be used with a prior consensus training phase to remove subjectivity. Other methods like pre-agreed photographs showing mild, moderate and severe involvement may be used for reporting, although their acceptance varies between pathologists. We feel that it is only with a valid, reliable scoring system that the correlation between histopathological findings and clinical features can be commented upon.
The authors have declared no conflicts of interest.
References
Heyreh SS, Podhajsky PA, Zimmerman B. Occult giant cell arteritis: ocular manifestations.
Hall S, Persellin S, Lie JT, O’Brien PC, Kurland LT, Hunder GG. Therapeutic impact of temporal artery biopsy.
Kaiser M. Formation of new vasa vasorum in vasculitis. Production of angiogenic cytokines by multinucleated giant cells.
Weyand CM, Hicok KC, Hunder GG, Goronzy JJ. Tissue cytokine patterns in patients with polymyalgia rheumatica and giant cell arteritis.
Author notes
Southend Hospital NHS Trust, Essex, UK and 1Sahlgrenska University Hospital, Göteborg, Sweden
Comments