andre ramon1, helene Greigert2, adrien guilloteau3, claudie cladière4, marion ciudad5, Sylvain Audia6, paul ornetti7, jean-Francis Maillefert7, Bernard Bonnotte2 and Maxime Samson2, 1Department of Rheumatology, Dijon University Hospital, Dijon, France, 2Department of Internal Medicine and Clinical Immunology, Dijon University Hospital, Dijon, France, 3Methodological Support Unit (USMR), Dijon-Bourgogne University Hospital, Dijon, France, 4INSERM, EFS BFC, UMR 1098, RIGHT Graft-Host-Tumor Interactions/Cellular and Genetic Engineering, Bourgogne Franche-Comté University, Dijon, France, 5INSERM, EFS BFC, UMR 1098, RIGHT Graft-Host-Tumor Interactions/Cellular and Genetic Engineering, Dijon, France, 6Department of Internal Medicine and Clinical Immunology, Dijon-Bourgogne University Hospital, Dijon, France, 7Department of Rheumatology , Dijon-Bourgogne University Hospital, Dijon, France
Background/Purpose: Polymyalgia rheumatica (PMR) and giant cell arteritis (GCA) are frequently overlapping conditions. It is estimated that 16 to 21% of PMR patients have GCA. In PMR, the presence of cranial features of GCA easily lead the clinician to diagnose GCA. However, the diagnosis of isolated large vessel (LV) GCA is more challenging and requires the use of vascular imaging techniques that are costly and/or irradiating. It is currently not recommended to perform vascular imaging or temporal artery biopsy to rule out GCA in PMR patients without features of GCA. Therefore, the identification of biomarkers able to rule out GCA in PMR patients could be helpful to select PMR patients in whom these exams should be performed. The aim of this study was to assess the performance of several biomarkers to identify GCA in PMR patients.
Methods: Patients were prospectively enrolled at the Dijon University Hospital (France). At inclusion, all patients were free of glucocorticoids or immunosuppressive drugs. All isolated PMR patients fulfilled the 2012 ACR/EULAR classification criteria and GCA was ruled out by at least one exam evaluating the temporal arteries (biopsy and/or doppler US scan) and an evaluation of large vessels by angio-CT or 18FDG-PET-CT. All PMR/GCA overlap patients fulfilled the 2022 ACR/EULAR classification criteria for GCA. For all patients, vascular remodeling markers (MMP2, MMP3, MMP9, ANGPTL4, ANGPTL6), endothelial markers (CD141, CD31, CD146, vWFA2, ICAM-1, VCAM-1), cytokines (IL-6, CXCL-9, IL-27, IL-11) and CD163 were determined by Luminex assay in serum samples. Mann-Whitney tests were performed to compare PMR and PMR/GCA patients. Receiving operator characteristic (ROC) with area under the curve (AUC) were used to assess the performance of the tested biomarkers. Optimal cut off were identified according to Youden Index.
Results: Fifty isolated PMR and 29 GCA/PMR patients were included. GCA/PMR patients had higher serum levels of CD141 (p = 0.004) and CXCL9 (p = 0.004) than isolated PMR patients. By contrast, serum levels of MMP3 (p = 0.02) and IL-6 (p = 0.003) were lower in GCA/PMR patients compared to isolated PMR patients. There was no difference between groups for the other markers. AUC were calculated for CD141, CXCL9, IL-6 and MMP3. Separately, none of them reached an AUC > 0.7. However, when combined, high accuracy performance was revealed for the CXCL9/IL-6 ratio, which was significantly increased in GCA/PMR patients (p < 0.0001) with an AUC of 0.75 (cutoff > 35.06) while the MMP3/CD141 ratio was significantly lower in GCA/PMR patients (p < 0.0001; AUC = 0.77, cutoff < 5.1).
Conclusion: This study demonstrated that combining serum markers such as MMP3/CD141 and CXCL9/IL-6 could help to discriminate between individuals with isolated PMR and those with GCA/PMR. These markers warrant prospective evaluation in larger cohorts of PMR and PMR/GCA patients.
a. ramon: None; h. Greigert: None; a. guilloteau: None; c. cladière: None; m. ciudad: None; S. Audia: None; p. ornetti: None; j. Maillefert: None; B. Bonnotte: None; M. Samson: ARGENX, 2, Boehringer-Ingelheim, 2, CHUGAI, 2, CSL Vifor, 2, GlaxoSmithKlein(GSK), 2, NOVARTIS, 2, 5.
