A risk scoring system to predict coronary stent thrombosis

Abstract

Objective: Stent thrombosis (ST) is a potentially life-threatening complication of percutaneous coronary intervention (PCI). We aimed to develop a scoring system to predict the risk of ST following PCI.

Research design and methods: Odds ratios (ORs) for risk factors associated with ST were identified from a meta-analysis based on a systematic literature review, and through consensus expert opinion (Delphi–RAND method). The combined ORs were used to calculate risk scores for acute (within 24 hours), early (within 30 days) and late (31 days to 1 year) ST. Risk scores were validated against patient-level data from the TRITON-TIMI 38 study. Twenty risk factors were identified.

Results: The most highly predictive factor for early and late ST was “incomplete duration of dual antiplatelet therapy”. Derived total risk scores ranged from 0 to 22 for acute and early ST, and from 0 to 20 for late ST. Increasing scores were associated with an increasing risk of ST when applied to trial data. Model discrimination was 0.60 (p = .0028), 0.67 (p < .0001) and 0.66 (p < .0001) for acute, early and late ST respectively, indicating good discriminatory power for predicting ST. Key limitations included a lack of published data on acute ST, resulting in a risk score for this time point being based predominantly on expert opinion, and that it was not possible to map all risk factors to variables collected in the TRITON-TIMI 38 study.

Conclusion: Our weighted scoring system may help to stratify ST risk and individualize antiplatelet therapy in patients undergoing PCI.

Keywords:

• Delphi–RAND panel
• meta-analysis
• risk factors
• scoring system
• stent thrombosis
• systematic review

Transparency

Declaration of funding

Financial support for the Delphi panel, and analysis and writing of the manuscript was provided by Lilly UK.

Declaration of financial/other relationships

M.B. has disclosed that he is an employee and shareholder of Eli Lilly. W.F. and P.L.G.-C. have disclosed that they are employees of Eli Lilly. A.H.G. has disclosed that he has received meeting sponsorship from Boston Scientific, lecture fees, a travel bursary and meeting sponsorship from AstraZeneca, and advisory board/meeting travel assistance and meeting sponsorship from Abbott Vascular, Eli Lilly and Daiichi Sankyo. A.Bg. has disclosed that he has received personal fees from Eli Lilly and an institutional grant from Boston Scientific. A.Bh. has disclosed that he has received grants from Abbott Vascular, grants and personal fees from The Medicines Company and personal fees from AstraZeneca.

N.M., A.S.B., D.J.B., A.d.B., T.L., R.S. and A.Z. have disclosed that they have no significant relationships with or financial interests in any commercial companies related to this study or article.

CMRO peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Acknowledgments

The authors would like to acknowledge Fanni Natanegara (Eli Lilly), Ming-Dauh (Eli Lilly) and Aijun Gao (Inventiv Health Clinical) for developing the plan of analysis and the models, and for performing the simulation work to test the models and the data analyses. They would also like to acknowledge Dr Sue Chambers (Rx Communications, Mold, UK) for medical writing assistance with the preparation of this article, funded by Eli Lilly.

The interventional cardiologists on the Delphi–RAND expert panel included Prof Adrian Banning (John Radcliffe Hospital, Oxford, UK), Prof Andreas Baumbach (Bristol Heart Institute, Bristol, UK), Dr Daniel J. Blackman (Leeds Teaching Hospitals, Leeds, UK), Dr Adam De Belder (Brighton & Sussex University Hospitals, Brighton, UK), Dr Thierry Lèfevre (Institut Cardiovasculaire Paris Sud, Massy, France), Dr Rod Stables (Liverpool Heart and Chest Hospital, Liverpool, UK) and Prof Azfar Zaman (Freeman Hospital, Newcastle-upon-Tyne, UK).