An artificial intelligence tool can identify potential complications after bone marrow or stem cell transplantation before symptoms arise, which could allow for more accurate monitoring of patients and pre-emptive treatment.
The BIOPREVENT algorithm combines machine learning with immune biomarkers and clinical data to predict chronic graft-versus-host disease (GVHD) or death after hematopoietic cell transplantation (HCT).
The tool, described in the Journal of Clinical Investigation, is currently designed for risk assessment and clinical research.
But ultimately, it could allow doctors to receive real-time, personalized risk estimates of post-transplant outcomes based on clinical and biomarker data.
“By the time chronic GVHD is diagnosed, the disease process has often been unfolding for months, quietly hurting the body,” explained researcher Sophie Paczesny, PhD, from the Medical University of South Carolina.
“We wanted to know whether we could detect warning signs earlier, before patients feel sick, and soon enough for clinicians to intervene, before the damage becomes irreversible.”
GVHD arises when donor cells infused to treat blood diseases react against the recipient and commonly affects the skin, gut, or liver. It is one of the leading causes of debilitating illness and death following HCT transplantation.
In an attempt to predict its occurrence, Paczesny and team developed BIOPREVENT using data from 1310 stem cell and bone marrow transplant recipients across four well-characterized, multicenter studies.
The data incorporated seven previously validated plasma biomarkers measured from blood samples collected between 90 and 100 days after transplantation that are linked with inflammation, immune activation and regulation, and tissue injury and remodeling.
This was combined with nine key clinical factors that included the patient’s age, transplant type, primary disease, and prior complications identified from transplant registries.
Patients were divided into training and validation datasets, and several machine-learning and deep-learning models were assessed for their ability to predict outcomes at various times over the year and a half following transplantation.
The researchers found that deep learning performed, at best, similarly to the other machine-learning approaches considered. This, they suggested, may have been due to an inability of the complex neural network structure in deep learning models to effectively understand relationships between biomarkers and chronic GVHD risk without a much larger sample size, in the tens of thousands.
Bayesian Additive Regression Trees (BART) produced consistently high results and was eventually chosen for the final model.
BIOPREVENT was the best-performing of several machine-learning models and illustrated its real-world applicability as a biomarker-based predictive tool in two power calculations for a hypothetical trial under two patient scenarios.
The researchers believe that theirs is the largest biomarker study of chronic GVHD to date and that making their web-based tool freely available will allow it to be tested further.
“It was important to us that this not remain a theoretical model or a tool limited to a single institution,” Paczesny said. “Making BIOPREVENT freely available helps ensure that researchers and clinicians can test it, learn from it, and, ultimately, improve care for transplant patients.”
