Platelet inhibitors linked to lower risk for digital ulcers in systemic sclerosis
Click Here to Manage Email Alerts
A prediction model presented at ACR Convergence 2021 demonstrated that the use of platelet inhibitors could predict a lower probability of digital ulcer occurrence among patients with systemic sclerosis.
“We all know that digital ulcers are common in patients with systemic sclerosis; up to 50% of the patients may develop these lesions, and if we leave these lesions untreated, they may turn gangrenous and the patients may require amputation,” Alexandru Garaiman, a doctoral student in clinical science at the department of rheumatology at the University Hospital Zurich in Switzerland said during a press conference. “Although, the use of iloprost, PDE5 inhibitors or bosentan is beneficial in the management of digital ulcers, the prevention and healing of digital ulcers is still a challenge in clinical practice.”
Garaiman and colleagues used prospective data from the European Scleroderma Trials and Research group (EUSTAR) registry to identify patients who met the 2013 ACR/EULAR SSc classification criteria with comprehensive longitudinal data on the presence of digital ulcers and platelet inhibitors.
The dataset — which included information from each patient’s (n = 3,710) last follow-up visit — was split into two cohorts: patients recorded in the EUSTAR registry before Jan. 1, 2017 (derivation cohort), and those recorded after (validation cohort).
More than half of the patients (67.8%) had limited cutaneous systemic sclerosis, 14.6% were male, median disease duration was 8.94 years and median age was 56.97 years.
At baseline, 486 patients had digital ulcers and 150 had taken platelet inhibitors. Median follow-up was 1.03 years, at which time 487 patients had digital ulcers and 90 had taken platelet inhibitors.
According to the researchers, their model confirmed previously reported risk factors for the presence of digital ulcers at follow-up, including smoking, prior digital ulcers and SCL-70 positive status, among others.
Additionally, their model revealed that exposure to platelet therapy was associated with a lower odds ratio of digital ulcer occurrence at follow-up (OR = .25; 95% CI, .07-.77).
“We developed and validated a performance prediction model, which is ready to assist the rheumatologist in decision making regarding the management of digital ulcers,” Garaiman said. “... Our model revealed that platelet inhibitors are associated with a lower probability of digital ulcer occurrence. But, of course, all these findings should be confirmed, ideally in a randomized, clinical, controlled trial before treatment recommendations or guidelines can be made.”
He added that “if the therapeutic effect of platelet inhibitors can be also demonstrated in this study design, this would mean we’d have another treatment option for systemic sclerosis patients with digital ulcers, and one that is readily available in most countries at a low price.”
Perspective
Back to Top
David A. McLain, MD, MACR, FACP, FRCP
Clinical prediction modelling is a complex area of study that plays an increasingly important role in clinical care by informing of outcome risks. These, of course, help us and the patient with shared medical decision-making. The FRAX score is an example of a clinical prediction model.
Steyerberg’s classic text, Clinical Prediction Models: A Practical Approach to Development, Validation, and Updating, totals 558 pages. Further study requires Harrell’s Regression Modeling Strategies: With Applications to Linear Models, Logistic Regression, and Survival Analysis, which is 582 pages. Mastery of this area is not practical for most busy clinicians who wish to use a clinical prediction model to answer a clinical question. They certainly would need a biostatistician to collaborate.
When you dig into clinical prediction modelling, you find an area of complex mathematics and statistics. It is like asking what time it is and having to understand how a clock works for the answer. Clinical prediction models can be either diagnostic or prognostic. Diagnostic prediction models aim to calculate an individual’s risk that a disease is already present, while prognostic prediction models aim to calculate the risk of particular heath states occurring in the future. The need to validate the model is part of the process and this article discusses derivation and validation for their model for digital ulcer prediction.
Collapse
Garaiman A. Abstract #1383. Presented at: ACR Convergence 2021; November 5-9, 2021 (virtual meeting).
Read more about
Click Here to Manage Email Alerts