Development of imaging-based risk scores for prediction of intracranial haemorrhage and ischaemic stroke in patients taking antithrombotic therapy after ischaemic stroke or transient ischaemic attack: a pooled analysis of individual patient data from cohort studies

Jonathan G. Best, Gareth Ambler, Duncan Wilson, Keon-Joo Lee, Jae-Sung Lim, Masayuki Shiozawa, Masatoshi Koga, Linxin Li, Caroline Lovelock, Hugues Chabriat, Michael Hennerici, Yuen Kwun Wong, Henry Ka Fung Mak, Luis Prats-Sanchez, Alejandro Martinez-Domeno, Shigeru Inamura, Kazuhisa Yoshifuji, Ethem Murat Arsava, Solveig Horstmann, Jan PurruckerBonnie Yin Ka Lam, Adrian Wong, Young Dae Kim, Tae-Jin Song, Robin Lemmens, Sebastian Eppinger, Thomas Gattringer, Ender Uysal, Zeynep Tanriverdi, Natan M. Bornstein, Einor Ben Assayag, Hen Hallevi, Jeremy Molad, Masashi Nishihara, Jun Tanaka, Shelagh B. Coutts, Alexandros Polymeris, Benjamin Wagner, David J. Seiffge, Philippe Lyrer, Ale Algra, L. Jaap Kappelle, Rustam Al-Shahi Salman, Hans R. Jager, Gregory Y. H. Lip, Urs Fischer, Marwan El-Koussy, Jean-Louis Mas, Laurence Legrand, Christopher Karayiannis, Thanh Phan, Sarah Gunkel, Nicolas Christ, Jill Abrigo, Thomas Leung, Winnie Chu, Francesca M Chappell, Stephen Makin, Derek Hayden, David J Williams, Werner H Mess, Paul J Nederkoorn, Carmen Barbato, Simone Browning, Kim Wiegertjes, Anil M. Tuladhar, Noortje Maaijwee, Anne Cristine Guevarra, Chathuri Yatawara, Anne-Marie Mendyk, Christine Delmaire, Sebastian Köhler, Robert van Oostenbrugge, Ying Zhou, Chao Xu, Saima Hilal, Bibek Gyanwali, Christopher Chen, Min Lou, Julie Staals, Régis Bordet, Nagaendran Kandiah, Frank-Erik de Leeuw, Robert Simister, Jeroen Hendrikse, David Calvet, Simon Jung, Vincent I. H. Kwa, Stefan T. Engelter, Nils Peters, Franz Fazekas, Vincent Thijs, Ji Hoe Heo, Vincent Mok, Roland Veltkamp, Hakan Ay, Toshio Imaizumi, Beatriz Gomez-Anson, Kui Kai Lau, Eric Jouvent, Peter M Rothwell, Kazunori Toyoda, Hee-Joon Bae, Joan Marti-Fabregas, David J. Werring*, Microbleeds Int Collaborative

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

4 Citations (Scopus)


Background Balancing the risks of recurrent ischaemic stroke and intracranial haemorrhage is important for patients treated with antithrombotic therapy after ischaemic stroke or transient ischaemic attack. However, existing predictive models offer insufficient performance, particularly for assessing the risk of intracranial haemorrhage. We aimed to develop new risk scores incorporating clinical variables and cerebral microbleeds, an MRI biomarker of intracranial haemorrhage and ischaemic stroke risk.

Methods We did a pooled analysis of individual-patient data from the Microbleeds International Collaborative Network (MICON), which includes 38 hospital-based prospective cohort studies from 18 countries. All studies recruited participants with previous ischaemic stroke or transient ischaemic attack, acquired baseline MRI allowing quantification of cerebral microbleeds, and followed-up participants for ischaemic stroke and intracranial haemorrhage. Participants not taking antithrombotic drugs were excluded. We developed Cox regression models to predict the 5-year risks of intracranial haemorrhage and ischaemic stroke, selecting candidate predictors on biological relevance and simplifying models using backward elimination. We derived integer risk scores for clinical use. We assessed model performance in internal validation, adjusted for optimism using bootstrapping.

Findings The included studies recruited participants between Aug 28, 2001, and Feb 4, 2018. 15 766 participants had follow-up for intracranial haemorrhage, and 15 784 for ischaemic stroke. Over a median follow-up of 2 years, 184 intracranial haemorrhages and 1048 ischaemic strokes were reported. The risk models we developed included cerebral microbleed burden and simple clinical variables. Optimism-adjusted c indices were 0.73 (95% CI 0.69-0.77) with a calibration slope of 0.94 (0.81-1.06) for the intracranial haemorrhage model and 0.63 (0.62-0.65) with a calibration slope of 0.97 (0.87-1.07) for the ischaemic stroke model. There was good agreement between predicted and observed risk for both models.

Interpretation The MICON risk scores, incorporating clinical variables and cerebral microbleeds, offer predictive value for the long-term risks of intracranial haemorrhage and ischaemic stroke in patients prescribed antithrombotic therapy for secondary stroke prevention; external validation is warranted. 

Original languageEnglish
Pages (from-to)294-303
Number of pages10
JournalThe Lancet neurology
Issue number4
Early online date17 Mar 2021
Publication statusPublished - 1 Apr 2021


  • neurology

Cite this