Nonsurgical giant cell tumour of the tendon sheath or of the diffuse type: Are MRI or 18F-FDG PET/CT able to provide an accurate prediction of long-term outcome?

Laurent Dercle,Roland Chisin,Samy Ammari,Quentin Gillebert,Monia Ouali,Cyril Jaudet,Jean-Pierre Delord,Lawrence Dierickx,Slimane Zerdoud,Martin Schlumberger,Frédéric Courbon

DOI: https://doi.org/10.1007/s00259-014-2938-9

2014-11-01

Abstract:PurposeTo investigate whether MRI (RECIST 1.1, WHO criteria and the volumetric approach) or 18F-FDG PET/CT (PERCIST 1.0) are able to predict long-term outcome in nonsurgical patients with giant cell tumour of the tendon sheath or of the diffuse type (GCT-TS/DT).MethodsFifteen “nonsurgical” patients with a histological diagnosis of GCT-TS/DT were divided into two groups: symptomatic patients receiving targeted therapy and asymptomatic untreated patients. All 15 patients were evaluated by MRI of whom 10 were treated, and a subgroup of 7 patients were evaluated by PET/CT of whom 4 were treated. Early evolution was assessed according to MRI and PET/CT scans at baseline and during follow-up. Cohen’s kappa coefficient was used to evaluate the degree of agreement between PERCIST 1.0, RECIST 1.1, WHO criteria, volumetric approaches and the reference standard (long-term outcome, delay 505 ± 457 days). The response rate in symptomatic patients with GCT-TS/DT receiving targeted therapy was also assessed in a larger population that included additional patients obtained from a review of the literature.ResultsThe kappa coefficients for agreement between RECIST/WHO/volumetric criteria and outcome (15 patients) were respectively: 0.35 (p = 0.06), 0.26 (p = 0.17) and 0.26 (p = 0.17). In the PET/CT subgroup (7 patients), PERCIST was in perfect agreement with the late symptomatic evolution (kappa = 1, p < 0.05). In the treated symptomatic group including the additional patients from the literature the response rates to targeted therapies according to late symptomatic assessment, and PERCIST and RECIST criteria were: 65 % (22/34), 77 % (10/13) and 26 % (10/39).Conclusion18F-FDG PET/CT with PERCIST is a promising approach to the prediction of the long-term outcome in GCT-TS/DT and may avoid unnecessary treatments, toxicity and costs. On MRI, WHO and volumetric approaches are not more effective than RECIST using the current thresholds.