CONFERENCE UPDATE : EHA25
Novel prognostic index for pediatric acute lymphoblastic leukemia risk stratification
Risk stratification is important in planning the suitable treatment options for pediatric acute lymphoblastic leukemia (ALL) patients and hence improving the prognostic outcomes.1 While the traditional binary risk stratification algorithms might be insufficient in predicting ALL, the spectrum expansion in genetic abnormalities and the understanding of minimal residual disease (MRD) have enabled more accurate prognosis prediction of the disease.2 To better stratify ALL risk, a novel prognostic index (PIUKALL) was proposed at the virtual 25th European Hematology Association Annual Congress (EHA25).
Traditional risk stratification categorizes patients into risk groups based on clinical factors and genetic alternations. “Because you are using binary choices (in typical risk stratification algorithm) where you have continuous variables, you are losing predictive power, and it’s more difficult to integrate risk factors in order to improve your decision-making process. This ultimately disregards the biological heterogeneity that exists within patients,” said Dr. Anthony V Moorman, professor of genetic epidemiology at Newcastle University.
One specific approach to address the limitations of traditional risk stratification is to intergrate MRD and genetic factors.2 Since the association between absolute relapse risk and specific MRD level varied by genetic subtype, genotype-specific MRD thresholds can help enhance the modified risk stratification algorithm.2 Although this refinement improved the accuracy of prognosis prediction, Dr. Moorman added, “this intervention, the idea of combining and integrating MRD and genetics, has overcome some but not all of the issues. The key thing is that it was still not integrating all of the risk factors.”
To develop a suitable prognostic index incorporating the relevant risk factors, researchers selected three variables relevant to the relapse prediction: White cell count at diagnosis, cytogenetics prior to treatment, and MRD values at the end of induction. A linear equation was derived from coefficients in the multivariable model.1
Three thresholds were identified with reference to the percentage of patients included, relapse rate and overall survival rate. Patients were then categorized into four different risk groups based on their prognostic index score. The proposed prognostic index was validated by event-free survival rate, overall survival rate and relapse rate comparison between the validation and discovery cohorts (Figure 1).1
Analysis of the area under the curve confirmed that PIUKALL was significantly better at predicting outcome when compared to traditional risk stratification algorithms. Investigators argued that the new prognostic index could provide a more accurate and flexible relapse risk stratification for allocating low-risk patients to treatment deintensification and high-risk patients to more experimental therapies. The improvement in risk stratification could address the treatment-related mortality in low-risk group and poor prognosis in high-risk group.1
“Crucially, we have demonstrated that the integration of risk factors is absolutely key to refining the risk prediction in pediatric ALL,” stated Dr. Moorman. He continued, “Optimal precision or personalized prediction in the future possibly would come from integrating multiple risk factors into a single prognostic index, and I believe that would be the future of risk stratification in pediatric ALL.”
- Enshaei A, et al. A validated novel continuous prognostic index to deliver stratified medicine in pediatric acute lymphoblastic leukemia. Blood. 2020 Apr 23;135(17):1438-46.
- O'Connor D, et al. Genotype-specific minimal residual disease interpretation improves stratification in pediatric acute lymphoblastic leukemia. J Clin Oncol. 2018 Jan 1;36(1):34-43.
A case review on cardiac risk stratification in the secondary prevention of cardiovascular diseases
Advances in treating chronic lymphocytic leukemia patients
Although chronic lymphocytic leukemia (CLL) patients now have a broader selection of treatment options available, therapeutic efficacies of most treatments still largely depend on the patients’ fitness, cytogenetics and previous treatment responses. Some clinically relevant mutations have now been identified to predict