Comparative analysis of third-generation dual-energy CT and IVUS for in-stent restenosis detection

Abstract

Purpose: Prior studies have assessed in-stent diameter restenosis (ISDR) in coronary arteries using 64-slice multidetector computed tomography coronary angiography (MDCT-CA) compared to invasive coronary angiography (ICA), which is the gold standard. This study aimed to compare the diagnostic accuracy of monoenergetic reconstruction using third-generation dual-source dual-energy CT (DSDECT) to that of ICA reconstruction via adjunctive intravascular ultrasonography (IVUS) for evaluating the ISDR.

Methods: A total of 95 patients with previously stented coronary arteries (involving 110 stents) underwent DSDECT followed by ICA and IVUS within a 24-h timeframe. The specificities, sensitivities, negative predictive values (NPVs), and positive predictive values (PPVs) of the DSDECT and ICA were compared for confirming or excluding the ISDR using in-stent area restenosis (ISAR) and a minimal luminal area (MLA) ≤ 4.0 mm² on IVUS as the reference standard.

Results: Compared with IVUS, the latest DSDECT demonstrated good sensitivity (100%), specificity (92.4%), and accuracy (96.1%) in detecting the ISDR. Our study highlights a limitation in assessability for stents with diameters < 3 mm, emphasizing the importance of careful patient selection. When employing an IVUS MLA of 4.0 mm² as a reference for identifying the ISDR, no significant difference was observed between DSDECT and ICA in the identification of the ISDR. However, it is important to note that the use of absolute cut-offs, such as < 6.0 mm² in the left main or < 4.0 mm², may not universally apply across varying ethnicities and between sexes. The interpretation of the minimal luminal area (MLA) should be considered in the context of individual patient characteristics, and caution is advised to avoid potential misleading conclusions based solely on absolute thresholds.

Conclusion: In summary, when assessing stent patency, the latest DSDECT exhibits similar performance to coronary angiography and IVUS. Moreover, it offers noninvasiveness, cost-effectiveness, and ease of operation, which are advantageous characteristics. However, it is essential to consider limitations in patient eligibility, including factors such as prior cardiac devices, arrhythmias, and any degree of chronic renal insufficiency, which may impact CT imaging analysis. The 100% negative predictive value (NPV) of third-generation DSDECT reliably excludes in-stent restenosis (ISDR), potentially obviating invasive angiography in stable patients with patent stents.

Trial registration: ZU-IRB#3915/13-8-2017 Registered 13 August 2017, email: IRB_123@medicine.zu.edu.eg.

Keywords: Computed tomography angiography; Coronary atherosclerosis; Coronary intervention planning; In-stent restenosis; Intravascular ultrasound; Noninvasive imaging.

Fig. 1

a Patient screening and enrollment flowchart. Of 163 consecutive patients with suspected in-stent restenosis, 68 were excluded (detailed in boxes). The final cohort included 95 patients (110 stents), with six stents excluded post-hoc due to non-diagnostic image quality. CKD = chronic kidney disease. b The 5-point grading scale for dual-source dual-energy CT image quality, with scores 1 through 3 (ranging from poor but technically assessable to good quality) considered diagnostic and included in analysis, whereas scores 4 (insufficient) and 5 (non-assessable) were excluded due to non-diagnostic quality. This classification ensured only images meeting minimum diagnostic thresholds were analyzed while maintaining transparency regarding borderline cases. c Distribution of stent diameter

Fig. 2

a Image quality by dual-source dual-energy CT. b Graphical Abstract: (Case 1)A 78-year-old male patient with a history of previous stenting to the proximal LCX 23 months prior via a DES 3.5×16 cobalt chromium stent presented with unstable angina. Panel A Coronal curved planar CT image showing suspected LCX in-stent stenosis (black shadow inside the stent), Panel B Invasive coronary angiography showing LCX in-stent restenosis, Panel C IVUS showing stent under expansion and Panel D ECG at presentation showing non-specific ST-T changes in inferolateral leads

Fig. 3

a Mechanisms of in-stent restenosis by IVUS. b Results for the 2.75 mm diameter group. c Results for the 3.0 mm diameter group. d Results for the 3.5 mm diameter group

Fig. 4

a Diagnostic accuracy of third-generation DSDECT for in-stent restenosis detection, comparing all patients (n=95) versus the stable subgroup (n=90; stable angina/NSTEMI). Error bars represent 95% confidence intervals. STEMI cases (n=5) were excluded as they proceeded directly to invasive coronary angiography