Automated database-guided expert-supervised orientation for immunophenotypic diagnosis and classification of acute...

Area of Science:

• Hematology
• Immunophenotyping
• Computational Biology

Background:

• Accurate acute leukemia (AL) classification is vital for effective treatment strategies.
• The EuroFlow AL orientation tube (ALOT) was developed to guide classification towards T-cell acute lymphoblastic leukemia (T-ALL), B-cell precursor (BCP)-ALL, and acute myeloid leukemia (AML).

Purpose of the Study:

• To develop and validate a database-guided automated algorithm for classifying AL subtypes using the ALOT.
• To assess the algorithm's accuracy in selecting appropriate classification panels and providing diagnostic data comparable to WHO classifications.

Main Methods:

• A reference database was constructed using 656 typical AL samples (T-ALL, BCP-ALL, AML) analyzed via standardized protocols.
• Principal component analysis (PCA)-based plots and automated classification algorithms were employed to compare single cells from 783 new cases against the database.
• Patients were categorized based on database-guided results, including typical, transitional, atypical, and mixed-lineage classifications.

Main Results:

• The automated algorithm successfully selected the correct classification panel for 99.7% (781/783) of cases.
• Diagnostic data comparable to final WHO classifications were achieved in over 93% of cases, including T-ALL (85%), BCP-ALL (97%), AML (95%), and mixed-phenotype AL (87%).
• Accurate classification was possible even without data from full-characterization panels.

Conclusions:

• Database-guided analysis of ALOT results standardizes interpretation and enables accurate selection of classification panels.
• The developed algorithm provides a robust foundation for future World Health Organization (WHO) classifications of acute leukemia.
• This approach enhances diagnostic efficiency and accuracy in acute leukemia management.