TOPLINE:

Ultrahigh-resolution (UHR) photon-counting detector (PCD)-CT demonstrated more than 90% accuracy in detecting coronary in-stent restenosis (ISR), outperforming standard-resolution reconstructions, with improvements in assessing small-calibre stents (below 3 mm in diameter), according to a retrospective study.

METHODOLOGY:

• Researchers conducted a retrospective analysis including 50 consecutive patients with 106 coronary stents (mean age, 70.9 years; 16% women) who underwent UHR coronary CT angiography (CCTA) at two academic centres between May 2023 and 2025.
• All scans were reconstructed with UHR imaging parameters at a slice thickness of 0.2 mm and an increment of 0.1 mm, and standard-resolution reconstructions were generated by downsampling to a slice thickness of 0.4 mm with an increment of 0.2 mm to simulate energy-integrating detector CT.
• Stents were visually evaluated for the presence and severity of ISR in both reconstruction types, with the severity of ISR categorised as negative (0%), non-obstructive (< 50%), or obstructive (≥ 50%) on the basis of the visual assessment of narrowing within the stent or within 5 mm of stent edges.
• Invasive coronary angiography performed within 8 weeks of CCTA served as the reference standard, with quantitative coronary angiography yielding continuous measurements including reference vessel diameter, minimal luminal diameter, and percentage diameter stenosis.
• Among the 106 stents evaluated, non-obstructive ISR was detected in 10 stents (9.4%) using the 0.2 mm UHR reconstruction and 14 stents (13.2%) using the 0.4 mm reconstruction, and obstructive ISR was found in 26 stents (24.5%) using 0.2 mm and 32 stents (30.2%) using 0.4 mm. Stents were evaluable in 99.1% of cases for the 0.2 mm reconstruction vs 91.5% for the 0.4 mm reconstruction.

TAKEAWAY:

• For the detection of non-obstructive ISR, the 0.2 mm UHR reconstruction achieved higher sensitivity (90.5% vs 82.9%), specificity (100.0% vs 78.5%), and overall diagnostic accuracy (96.2% vs 80.2%) than the 0.4 mm standard-resolution reconstruction.
• For the detection of obstructive ISR, the UHR vs standard-resolution reconstruction demonstrated improved specificity (98.7% vs 82.3%), positive predictive value (96.3% vs 64.1%), and diagnostic accuracy (98.1% vs 84.9%), whereas sensitivity remained high in both cases (96.3% vs 92.6%).
• In small-calibre stents (below 3 mm in diameter), UHR PCD-CT showed superior overall diagnostic accuracy than 0.4 mm reconstructions for both non-obstructive ISR (94.1% vs 73.5%) and obstructive ISR (94.1% vs 79.4%).
• Interreader agreement was higher with the 0.2 mm UHR reconstruction (Cohen's kappa, 0.71; 95% CI, 0.45-0.86) than with the 0.4 mm reconstruction (Cohen's kappa, 0.40; 95% CI, 0.15-0.61).

IN PRACTICE:

"[The study] findings support the broader clinical utility of PCD-CT in ISR assessment and provide a rationale for larger prospective studies to confirm clinical impact," the authors wrote.

SOURCE:

This study was led by Bálint Szilveszter, Semmelweis University Heart and Vascular Center, Budapest, Hungary. It was published online on March 05, 2026, in the European Journal of Radiology.

LIMITATIONS:

Recent updates on the first commercially available PCD-CT scanner have allowed for combined spectral and UHR image acquisition that could further improve diagnostic performance for certain clinical tasks. The study employed a reconstruction simulating energy-integrating detector CT rather than performing a repeat scan in the same patients to avoid additional radiation exposure. Partial verification bias may have influenced diagnostic performance.

DISCLOSURES:

This study was supported by the János Bolyai Research Scholarship of the Hungarian Academy of Sciences; a grant from the Hungarian National Research, Development and Innovation Office; the Semmelweis University Excellence Scholarship Program; and the Ministry of Culture and Innovation of Hungary from the National Research, Development and Innovation Fund. Four authors reported receiving travel support, speaker fees, and research support or having other ties with Siemens Healthineers and various other sources. Full disclosures are noted in the original article.

This article was created using several editorial tools, including AI, as part of the process. Human editors reviewed this content before publication.