Risk Factors for Febrile Neutropenia in Patients With Newly Diagnosed Diffuse Large B-Cell Lymphoma Undergoing...

Area of Science:

• Oncology
• Hematology
• Biostatistics

Background:

• Febrile neutropenia (FN) is a frequent complication in patients with diffuse large B-cell lymphoma (DLBCL) undergoing initial rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP) therapy.
• Identifying risk factors and developing predictive models for FN is crucial for managing this adverse event.

Purpose of the Study:

• To identify significant risk factors associated with FN in DLBCL patients treated with R-CHOP.
• To develop and validate a machine learning-based predictive model for FN occurrence.

Main Methods:

• Retrospective analysis of 238 newly diagnosed DLBCL patients treated with R-CHOP.
• Logistic regression for risk factor identification.
• Machine learning techniques for predictive model development and validation.

Main Results:

• The incidence of FN was 23.9%.
• Significant risk factors included bone marrow involvement, advanced disease stage (III-IV), ECOG Performance Status ≥2, elevated lactate dehydrogenase (≥240 U/L), and ≥2 extranodal sites.
• Machine learning models demonstrated moderate to strong predictive power (C-statistics ranging from 0.692 to 0.879).
• Patients with FN had significantly lower 5-year overall survival rates (57.6%) compared to those without FN (77.1%).

Conclusions:

• FN in DLBCL patients treated with R-CHOP is associated with significantly worse overall survival.
• Machine learning offers a viable approach for constructing reliable FN predictive models.
• Consideration of tailored prophylaxis, such as granulocyte-colony stimulating factor and antibiotics, may be beneficial for high-risk patients.