Authors
Lukas K van Vugt, Dennis A J van den Broek, Pedro Miranda Afonso, Danny van der Helm, Alvaro Assis de Souza, Brenda C M de Winter, Marlies E J Reinders, Marian C Clahsen-van Groningen, Jesper Kers, Aiko P J de Vries, Dennis A Hesselink
Published in
Immunobiology. Volume 231. Issue 6. Pages 153261. Oct 27, 2026. Epub Oct 27, 2026.
Abstract
Alemtuzumab is an effective therapy for severe kidney transplant rejection. Yet, not all patients benefit from this treatment, while they are at risk for its toxic effects. A prediction model could help to identify these patients and potentially guide treatment decisions.
We developed a prediction model using logistic regression and four prespecified, clinical predictor variables on a clinical dataset of two hundred patients who were treated for glucocorticoid-resistant and/or severe kidney transplant rejection with alemtuzumab. The outcome was a composite of graft loss or having a marginal graft function over the entire six months post-treatment. The model was internally validated using bootstrapping and a pilot external validation was performed with an independent validation cohort (n = 96). To make optimal use of the available data, the model was refitted to the combined datasets. Its potential for clinical utility was subsequently assessed with a decision curve analysis.
The model included the predictors rejection subtype, timing, donor type, and dialysis dependence. It showed an optimism-corrected c-statistic and calibration slope of 0.75 and 0.81. The pilot external validation demonstrated a c-statistic of 0.71 and calibration slope of 0.86. There was a potential benefit of model-informed treatment decisions over a 'treat-all' strategy.
The model showed moderate discrimination and reasonable calibration for estimating the risk of an unfavorable graft outcome after alemtuzumab treatment and warrants further external validation.
PMID:
42956456
Bibliographic data and abstract were imported from PubMed on 30 Oct 2026.
Read full publication at:
Please sign in
to see all details.
Advertisement
Stats
- Recommendations n/a n/a positive of 0 vote(s)
- Views 9
- Comments 0