Concordancia interobservador en el análisis visual y semicuantitativo de las imágenes SPECT 123I-FP-CIT en el diagnóstico del síndrome parkinsoniano

Abstract

Aims Using 123I-FP-CIT SPECT images makes it possible to identify presynaptic deterioration of the dopaminergic pathway by studying the dopamine transporter (DAT). A correct analysis of the SPECT images contributes to an adequate interpretation and diagnosis of movement disorders. Aims: 1. To compare visual and semiquantitative analysis of 123I-FP-CIT SPECT images in patients with movement disorders. 2. To evaluate interobserver agreement in visual and semiquantitative analysis. 3. To obtain a cut-off in the semiquantitative analysis to discriminate primary Parkinsonism Syndrome (PS) from non-primary PS. Methods A 123I-FP-CIT SPECT was performed in 32 patients with movement disorders suggestive of primary PS. Visual and semiquantitative images analyses were performed independently by two nuclear medicine physicians. Visual analysis was based on the visual interpretation. Semiquantitative analysis was calculated as specific uptake (caudate, putamen and striatum) versus non-specific uptake (occipital). Sensitivity, specificity, PPV, and NPV were calculated. Data were compared using ANOVA test followed by Bonferroni post-hoc test. Interobserver agreement of the visual and semiquantitative analysis was assessed by intraclass correlation coefficient and Kappa statistics, respectively. ROC curve was generated with semiquantitative data. Results Visual analysis showed 86% sensitivity and 100-88% specificity for the differential diagnosis of primary PS from non-primary PS. Semiquantitative analysis showed a gradual hypouptake proportional to the disease severity obtained in the visual analysis. Semiquantitative analysis did not provide any additional information to the visual analysis. Intraclass correlation coefficient and Kappa statistics showed 0.92 and 0.80 values, respectively. The Cut-off value to differentiate primary PS from non-primary PS was 1.9 on the putamen index.