Detection of stent restenosis in single vessel CAD: comparison of 201Tl and gated 99mTc-MIBI SPECT

BACKGROUND AND AIM: After successful percutaneous transluminal coronary angioplasty (PTCA), stent restenosis is observed in up to 40% of patients during the first year. The aim of this study was to determine the value of myocardial perfusion studies (MPSs) in the detection of stent restenosis in a symptomatic patient cohort. METHODS: A total of 80 patients without prior myocardial infarction (MI) and who underwent one-vessel PTCA with stent implantation were included to study. The diagnostic accuracy of two study protocols, Tl single photon emission computed tomography (SPECT) (38 patients) and rest-stress same day gated Tc sestamibi (MIBI) SPECT (42 patients), were compared. MPS data were visually evaluated by two experienced observers and stress induced perfusion defects with reversibility at rest was considered as restenosis. In gated MIBI data the wall motion (WM) and ejection fraction (EF) were also noted. The diagnostic value of a semiquantitative method based on 20 segment model and summed stress scores (SSSs) and summed difference scores (SDSs) were also tested. Results of MPSs were compared with control coronary angiography (CA) in all patients and agreement was defined as the kappa value (kappa). RESULTS: The average time between stent implantation and MPS was 8.9 +/- 2 months. Restenosis was detected in 58% of patients in CA. No significant differences were observed regarding age, gender, achieved exercise levels, vascular territorial distribution of lesions, imaging time interval after PTCA and degree of restenosis between Tl and gated MIBI groups. MPSs identified stent restenosis in 41 of 47 patients (sensitivity, 87%; specificity, 82%; accuracy, 85%; kappa=0.69). Four of six occluded stents that could not be detected in MPSs revealed intermediate degree stenosis (50-70%). Sensitivity, specificity and accuracy were not significantly different but better for gated MIBI group when compared to Tl (sensitivity, 90-83%; specificity, 85-80%; accuracy, 88-82%). Semiquantitative evaluation using SSS and SDS reached lower sensitivity than qualitative evaluation (MIBI, 90% vs 69%; Tl, 83% vs 72%) but higher specificity (MIBI, 85% vs 92%; Tl, 80% vs 100%) for both tracers and SSS were significantly correlated with occlusion degree (r=0.69). EF values calculated from the gated MIBI study were also inversely correlated with occlusion degree (r=0.55) and significantly different in patients with occluded stents (P<0.001). Agreement with CA for both tests were adequate (kappa=0.73, for MIBI; and kappa=0.63, for Tl). CONCLUSION: Semiquantitative evaluation of MPSs using SSS may enhance diagnostic specificity in the detection of stent restenosis. Both Tl and gated MIBI studies accurately detected stent restenosis. The gated MIBI method has advantages of WM analysis and evaluation of EF.