A new model accurately distinguished between indolent systemic mastocytosis (SM) and advanced SM using blood sample parameters, according to a recent abstract of an analysis presented at the American Society of Hematology’s Annual Meeting and Exposition.
“To help guide care across practice settings and to minimize procedural diagnostic interventions that can have potential side effects, we developed a predictive model aiming to accurately classify ISM vs AdvSM based on objective, easily assessed, and broadly available laboratory and clinical parameters,” the researchers said.
What is the difference between advanced and nonadvanced SM?
Nonadvanced SM comprises the indolent SM and smoldering SM subtypes. Advanced SM includes aggressive SM, SM with an associated hematologic neoplasm and mast cell leukemia.
SM is a complex disease with several clinical subtypes. Both indolent and advanced subtypes can present with symptoms associated with excessive mast cell activation, which can significantly limit daily activities and affect quality of life.
Read more about SM testing and diagnosis
The key difference between the two subtypes: Patients with indolent SM have a normal life expectancy, while advanced SM is associated with organ infiltration and damage, with a significant effect on survival.
The authors collected data from more than 400 patients with SM participating in SM clinical trials such as PATHFINDER and EXPLORER. Of them, 261 patients had indolent SM, and 176 patients had various forms of advanced SM.
Researchers fed a forest algorithm various demographic parameters, including age, sex and country of origin, as well as clinical parameters such as the presence of intra-abdominal fluid (ascites), enlarged spleen (splenomegaly) and pleural effusion.
They also introduced laboratory data such as liver enzymes, albumin, alkaline phosphatase, basophil count, eosinophil count, hemoglobin, white blood cell (WBC) count and platelet count.
The algorithm produced various models to distinguish patients with indolent SM and advanced SM. Model 1 showed an excellent discriminative ability with an area under the curve (AUC) of 0.97. Statistically speaking, the closer the AUC is to 1, the greater the predictive value of a clinical tool. Model 2 also had an excellent predictive ability, with an AUC of 0.95.
Model 1 relied on age, platelets, absolute monocyte count, hemoglobin, alkaline phosphatase, tryptase and total bilirubin. Model 2 used pleural effusion, ascites, hemoglobin, absolute neutrophil count and platelets.
Model 1 misclassified 31 patients; model 2 misclassified 33 patients. The authors pointed out that misclassified patients had several high-risk characteristics. “Patients aberrantly classified by these models may identify current limitations in our understanding of SM biology,” the authors said.
