Background Fluorine-18-fluorodeoxyglucose (FDG) PET/CT is increasingly used to diagnose large vessel GCA (LV-GCA) due to its excellent diagnostic accuracy[1]. However, PET/CT is not always readily available, which may compel the clinician to 1) either delay steroid treatment at the risk of GCA related complications, or 2) initiate treatment at the expense of diagnostic sensitivity of the FDG PET/CT study.

Objectives To evaluate if FDG PET/CT can accurately diagnose LV-GCA after 3 or 10 days of high-dose steroid treatment.

Methods Twenty-four treatment-naïve patients (16 women) with a mean age of 69 (range 57–84) years with FDG PET/CT (PET0) proven LV-GCA repeated FDG PET/CT after either 3 (PET3, n=10) or 10 days (PET10, n=14) of treatment with oral prednisolone 60 mg daily. Prior to treatment, clinical examination and laboratory tests were performed to confirm GCA and exclude differential diagnoses. A temporal artery biopsy (TAB) was performed in all patients.

Two experienced nuclear medicine physicians blinded to clinical data reviewed the FDG PET/CT images. LV-GCA was suspected if increased FDG uptake in the wall of the aorta and/or supra-aortic branches was observed. A semi-quantitative approach was applied (a.m. Meller) in which FDG uptake was graded on a 5-point scale (0; no uptake, 1; ≤ blood pool, 2; > blood pool, ≤ liver, 3; ≥ liver, 4; ≥2xliver). A score ≥3 was considered consistent with vasculitis[2]. Vascular composite scores (CS) was calculated summarizing grades from assessed vascular regions; Aortic: Aorta ascendens, aorta descendens and aortic arch; aortic branches: Vertebral, carotic and subclavian/axillary artery.

Results Mean CRP and ESR were 72 (95% CI: 55; 94) mg/l and 81 (95% CI: 72; 90) mm/h, respectively. ACR criteria for GCA was fulfilled by 18/24 patients and 17/21 had a positive TAB. Mean number of prednisolone doses before the post-treatment FDG PET/CT were 3.1 (SD 0.3) (PET3) and 10.3 (SD 0.7) (PET10).

Vascular CS in aorta did not decrease at PET3 (9 (IQR 9–9) vs. 9 (IQR 6–9)) whereas a significant decrease was observed in aortic branches at PET3 (6.5 (IQR 6–8) vs 5.5 (IQR 5–7), p<0.01) and both vascular domains at PET10 (Aortic; 9 (IQR 9–9) vs. 5.5 (IQR 3–6), aortic branches; 7 (IQR 7–8) vs 5 (IQR 4–6)). Although, FDG uptake decreased in aortic branches after 3 days, LV-GCA could still be accurately diagnosed in 10/10 patients. By contrast, LV-GCA could only be diagnosed in 5/14 patients after 10 days (PET0 vs. PET10, p<0.01).

At day 10, VAS global was significantly higher in patients with positive PET10 compared to patients with negative PET10 (5.2 (95% CI 3.6; 7.0) vs. 2.7 (95% CI 1.2; 4.2), p<0.05). No clinically significant differences in baseline phenotypical presentation, CRP or PET CS were found between patients with positive and negative PET10, respectively.

Interrater reliability of visual FDG-uptake-grading was substantial (agreement 90%, Cohens weighted kappa 0.67).

Conclusions In LV-GCA, high-dose steroid treatment for three or ten days differentially attenuates the regional uptake of FDG but diagnostic accuracy remains within the first three days.

References

1. Puppo et al. BioMed research international 2014.

2. Stellingwerff MD et al. Medicine 2015.

References

Acknowledgements Assesment of PET scans by Stine Kramer and Tronds Bogsrud is mostly appreciated.

Disclosure of Interest None declared