Practical RECIP: Visual assessment of response evaluation criteria In PSMA-PET 1.0 - 04/03/23
, I. Rauscher 3, W.P. Fendler 4, B. Hadaschik 5, K. Herrmann 4, J. Calais 1, J. Czernin 1, M. Rettig 6, M. Eiber 3, M. Weber 4, M.R. Benz 1, 7, A. Farolfi 1, 8Résumé |
Introduction |
To evaluate the accuracy of Response Evaluation Criteria In PSMA-Imaging (RECIP) 1.0 determined by visual reads of nuclear medicine physicians for response evaluation using PSMA-PET/CT in men with mCRPC treated with 177Lu-PSMA radioligand therapy.
Methods |
In total, 124 patients were included in this multicenter retrospective study. All patients received 177Lu-PSMA and underwent PSMA-PET/CT scans at baseline (bPET) and at 12 weeks (iPET). Five nuclear medicine specialists individually interpreted each pair of PSMA-PET/CT scans for responses in PSMA-positive total tumor volume (PSMA-VOL) assessed visually and for occurrence of new lesions. Responses in PSMA-VOL were also assessed quantitatively using semi-automatic segmentation software. Occurrence of new lesions was combined: 1) with visual responses in PSMA-VOL to determine visual RECIP; and 2) with quantitative responses in PSMA-VOL to determine quantitative RECIP. RECIP response classes were dichotomized for differentiation of clinically relevant progressive (PD) vs. non-progressive disease (nPD): RECIP-PD vs. RECIP-nPD. Primary outcome measure was the inter-reader reliability of visual RECIP and agreement between visual vs. quantitative RECIP (by Cohen's κ coefficient). Secondary outcome measure was the prognostic value of visual RECIP (by Cox regression) for overall survival (OS).
Results |
Based on readers’ majority rule, 41/124 (33%) and 83/124 (67%) patients had visual RECIP-PD and RECIP-nPD, respectively. Agreement among readers in classifying visual RECIP-PD vs. RECIP-nPD was observed in 103 (83%) patients (excellent agreement: κ = 0.81). Agreement between visual vs. quantitative RECIP was observed in 118 (95%) patients (excellent agreement: κ = 0.89). OS was significantly shorter in men with visual RECIP-PD vs. RECIP-nPD (HR 2.55; 95%CI, 1.71–3.81; P<0.001). The median OS was 8.4 months (95%CI, 7.0–9.8) for patients with RECIP-PD and 16.1 months (95%CI, 13.6–18.5) for patients with RECIP-nPD.
Conclusion |
Visual RECIP demonstrated excellent inter-reader reliability and excellent agreement with quantitative RECIP calculated using semi-automatic segmentation software. Visual RECIP was prognostic for overall survival of men undergoing 177Lu-PSMA radioligand therapy. RECIP 1.0 evaluated visually by nuclear medicine physicians can be used in practice and clinical trials for response evaluation using PSMA-PET/CT in men with mCRPC.
Le texte complet de cet article est disponible en PDF.Keywords : PSMA, Lu-PSMA, PET response assessment
Plan
Vol 47 - N° 2
P. 41-42 - mars 2023 Retour au numéroBienvenue sur EM-consulte, la référence des professionnels de santé.
L’accès au texte intégral de cet article nécessite un abonnement.
Déjà abonné à cette revue ?
