Researchers have developed a new tool for helping clinicians tailor personalized treatment plans for patients with a rare blood cancer called chronic myelomonocytic leukemia (CMML). The tool, called the international CMML Prognostic Scoring System (iCPSS), may help improve a patient’s chance of survival.
CMML is caused by genetic mutations in bone marrow stem cells that are acquired over one’s lifetime. The only curative treatment is a stem cell transplant, which replaces diseased cells with healthy donor cells. But it is a risky procedure that can cause severe complications such as infections or graft-versus-host disease. It is not a suitable option for many patients, and oncologists have limited data to help them choose which patients are best suited for transplantation.
To address that challenge, the research team created iCPSS using machine learning techniques to analyze clinical and genetic data from more than 3,000 patients. The tool helps clinicians identify which patients will have the best outcomes from stem cell transplantation and determine the optimal timing for the procedure. The researchers published their findings in the Journal of Clinical Oncology.
“Our tool can help physicians decide when is the best time to discuss transplantation with their patients.”
Luca Lanino, MD, a Postdoctoral Associate at Yale School of Medicine and the Study’s First Author
New tool identifies best stem cell transplant candidates
CMML presents in different ways, and oncologists classify the cancer into two main subtypes. Low white blood cell counts are associated with the myelodysplastic subtype, in which blood cells fail to mature properly. High white blood cell counts are linked to the myeloproliferative subtype, where the bone marrow produces too many blood cells.
The disease presentation can also change over time. “You might start as a dysplastic subtype, but then evolve into a more proliferative phenotype as the disease progresses,” Lanino says.
Genetic mutations underlying the cancer partly drive these differences in disease presentation. In the latest study, researchers examined clinical and genetic data from an international cohort of more than 3,000 patients who had already received care, mainly in North America or Europe. “This genetic information is routinely collected in the clinical workup of patients,” Lanino explains.
The international team of researchers combined genomic sequencing data with clinical features to build iCPSS, a new prognostic scoring system that sorts patients with CMML into five risk categories based on their chance of survival. The researchers also developed a decision support system that models expected survival under different transplantation strategies, enabling clinicians to determine whether stem cell transplantation is likely to be beneficial and when it should be performed.
Compared with existing prognostic tools, the iCPSS-guided approach changed the recommended transplantation strategy in nearly one in three patients, resulting in a significant gain in life expectancy.
“We tried to stratify patients to help us decide which subsets of patients we should try to more aggressively treat,” Lanino says. “And what other subsets we can be slower in treating—reserving aggressive treatment for later when the disease is worse.”
Patients who underwent a transplant within the iCPSS-recommended time frame had a significantly higher chance of survival, they found. The researchers then validated iCPSS in a separate, prospectively enrolled international cohort of around 500 patients.
iCPSS gives oncologists an evidence-based way to create personalized treatment plans for people with CMML. Understanding the different genetic mutations that cause CMML can also help guide future research on the disease, Lanino adds. “If we study the subsets of patients that share a homogenous genetic background, we might be able to find treatments that work for that specific subset.”
Amer Zeidan, MBBS, MD, professor of internal medicine (hematology) and member of Yale Cancer Center, was a co-author of the study.
(Newswise/HG)
