Objective
To determine the population-based sensitivities and specificities of various inflammatory markers for the diagnosis of GCA.
Design
Retrospective population-based cross-sectional study.
Participants
Patients who underwent temporal artery biopsy (TAB) from 01/01/1995 through 12/31/2019 in Olmsted County, Minnesota were identified using the Rochester Epidemiology Project (REP). Subjects were categorized as GCA or non-GCA using the 1990 American College of Rheumatology classification criteria for GCA.
Testing
Sensitivity and specificity of C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), and platelets for a positive GCA diagnosis were calculated.
Main Outcome Measures
GCA diagnosis.
Results
There were 553 patients who underwent TAB, 143 with GCA and 410 without GCA. Median age at time of TAB was 75 years (IQR 69-81) and 372 (67%) were female. Patients with GCA had higher CRP (91.1 vs 49.1 mg/L, P < .001), ESR (66.4 vs 52.0 mm/hr, P < .001), and platelets (370.5 vs 283.1 × 10 9 /L, P < .001) than patients in the non-GCA cohort. Sensitivity was 96% for CRP, 80% for ESR, and 49% for platelets. Specificity was 21% for CRP, 43% for ESR, and 79% for platelets. Combined ESR and CRP had a sensitivity of 77% and specificity of 54% while combined ESR, CRP, and platelets had a sensitivity of 41% and specificity of 84%. Normal inflammatory markers, including ESR, CRP, and platelets, were observed in 2 (3%) GCA patients.
Conclusions
This population-based study evaluated the performance of various inflammatory markers in the diagnosis of GCA. Inflammatory markers are helpful in the diagnosis of GCA, but normal values do not exclude the diagnosis.
INTRODUCTION
The gold standard diagnostic test for giant cell arteritis (GCA) with cranial manifestations is the temporal artery biopsy (TAB). TAB sensitivity ranges from 70% to 90%, therefore a negative result does not exclude a GCA diagnosis.[1] A TAB has some potential morbidity and therefore inflammatory markers, such as erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), and platelets, are widely used for initial screening. Few studies evaluate the performance of inflammatory markers in the diagnosis of GCA with existing literature based largely on single center and multicenter studies that are susceptible to referral bias. , This study utilizes the Rochester Epidemiology Project (REP), a collaborative medical records linkage system that captures nearly all healthcare utilization within Olmsted County, Minnesota. For over 5 decades, the REP has supported a wide range of longitudinal population-based studies, integrating healthcare data from the Mayo Clinic, Olmsted Medical Center, and their affiliated hospitals. By leveraging this dataset, this study aims to determine the population-based sensitivities and specificities of various inflammatory markers for the diagnosis of GCA, providing insights that may inform clinical decision-making in broader practice settings.
METHODS
Institutional review board approval was obtained from Mayo Clinic and Olmsted Medical Center, with adherence to the Declaration of Helsinki and Health Insurance Portability and Accountability Act. The REP was used to identify Olmsted County, Minnesota residents who underwent TAB from 1/1/1995 to 12/31/2019. The cohort of incident GCA patients was previously identified. Patients were included if they met the 1990 American College of Rheumatology classification criteria or the following criteria: age ≥50 years, elevated inflammatory markers, and radiographic evidence of large-vessel vasculitis involving the ascending aorta and/or its branches. Radiographic modalities used to detect large-vessel vasculitis included computed tomography angiography, magnetic resonance angiography, and positron emission tomography-computed tomography.
Using the GCA cohort, patients with TAB were subdivided into those with and without GCA. Medical records were reviewed for demographics, TAB results, systemic and ocular symptoms, glucocorticoid use, and inflammatory marker values including ESR, CRP, platelets, hemoglobin (Hgb), white blood cells (WBC), lymphocytes, and neutrophils. Inflammatory marker values from patients already on systemic glucocorticoids were excluded.
Categorical variables were summarized as raw counts and proportions, and continuous variables as mean ± standard deviation. Rank sum tests compared continuous variables and chi-square tests compared categorical variables between GCA and control cohorts. Sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of inflammatory markers, individually or in combination, for a positive GCA diagnosis were calculated based on proportions of elevated or decreased markers with exact 95% confidence intervals from the binomial distribution. Normal ranges of inflammatory markers are in Supplemental Table 1. All statistical tests were 2-tailed with an alpha of p < 0.05. Analyses were performed using SAS version 9.4 (SAS Institute, Cary, NC, USA).
RESULTS
The study included 553 patients who underwent TAB, 143 with GCA and 410 without GCA. Median age at time of TAB was 75 years (IQR 69-81), 372 (67%) were female and 508 (92%) were White. Among the 143 GCA patients, 113 (79%) were TAB positive, 6 (4%) met radiologic criteria for GCA, and 24 (17%) were clinically diagnosed with GCA. Common systemic diagnosis and symptoms in the GCA cohort are shown in Table 1 . There were 16 (12%) patients with ocular manifestations of GCA without systemic symptoms. Ocular symptoms included diplopia in 8 (6%), permanent vision loss in 10 (7%), and transient vision loss in 9 (6%) ( Table 1 ). A sub-analysis of patients with permanent vision loss found that 2 (20%) had ocular manifestations of GCA without systemic symptoms.
| Characteristics | Patients with GCA (N = 143) n (%) |
|---|---|
| Polymyalgia rheumatica | 41 (29) |
| Systemic symptoms | |
| Headache | 103 (72) |
| Scalp tenderness | 74 (53) |
| Jaw claudication | 70 (50) |
| Fatigue | 51 (37) |
| Subjective weight loss | 38 (27) |
| Anorexia | 29 (21) |
| Subjective fever | 18 (13) |
| Ocular symptoms | |
| Diplopia | 8 (6) |
| Permanent vision loss | 10 (7) |
| Transient vision loss | 9 (6) |
| No systemic symptoms | 16 (12) |
Comparisons between patients with and without GCA showed that GCA patients had higher mean age (77.6 vs 73.5 years, P < .001), CRP (91.1 vs 49.1 mg/L, P < .001), ESR (66.4 vs 52.0 mm/hr, P < .001), and platelets (370.5 vs 283.1 × 10 9 /L, P < .001) ( Table 2 ). Other biomarkers of inflammation including WBC, neutrophil to lymphocyte ratio (NLR), and platelet-lymphocyte ratio (PLR) were also significantly higher in GCA patients ( Table 2 ). The proportion of elevated or decreased inflammatory markers for patients in the GCA and control cohorts is noted in Supplemental Table 2.
| Characteristic | Patients with GCA (N = 143) mean (SD) | Patients without GCA (N = 410) mean (SD) | P-value |
|---|---|---|---|
| Sex, female, n (%) | 108 (76) | 264 (64) | .015 |
| Age in years | 77.6 (8.0) | 73.5 (10.3) | <.001 |
| CRP, mg/L | 91.1 (82.5) | 49.1 (61.0) | <.001 |
| ESR, mm/hr | 66.4 (29.9) | 52.0 (33.1) | <.001 |
| Hgb, g/dL | 12.0 (1.3) | 12.1 (1.7) | .495 |
| WBC, x10 9 /L | 9.5 (2.9) | 9.0 (9.3) | <.001 |
| Lymphocytes, x10 9 /L | 1.6 (0.6) | 2.3 (9.5) | .881 |
| Neutrophils, x10 9 /L | 6.6 (2.4) | 6.0 (3.3) | .001 |
| Platelets, x10 9 /L | 370.5 (118.5) | 283.1 (112.2) | <.001 |
| NLR | 4.9 (2.7) | 4.9 (4.8) | .019 |
| PLR | 263.1 (127.5) | 235.4 (232.1) | <.001 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
