The evaluation of minimal residual disease (MRD) in acute leukemia (AL) is currently recognized as a potential critical tool to assess the response and relapse rate of treatments. The present study investigated serum peptides from patients with AL to identify biomarkers that would be useful in providing clinical evaluations and independent prognostic information. The patterns of serum peptides from 123 patients with AL and 49 healthy controls were analyzed using matrix-assisted laser desorption/ionization-time of flight mass spectrometry. Furthermore, diagnostic models of differential peptides were established using the support vector machine (SVM) algorithm to discriminate between the AL patients and healthy controls or between the AL patients with various degrees of remission. Finally, the peptides were applied to evaluate the prognosis of the affected patients. The area under the receiver operating characteristic (ROC) curve (AUC), analyzed using the SVM algorithm to distinguish between the AL patients and healthy controls, was 0.921. The AUC of the models for distinguishing between the newly-diagnosed AL patients and those in AL-hematological complete remission (HCR) and between the AL-HCR patients from those in AL-molecular remission (MR), was 0.824 and 0.919, respectively. A short serum peptide of m/z 4625 was identified to decrease in density in parallel with an increase in the degree of remission, which was used to monitor the MRD level. The intensity of the m/z 4625 peptide was significantly correlated with a poor overall survival (OS). The m/z 4625 peptide was identified to be a partial fragment of SERPINA3. The serum peptide pattern is high in sensitivity and specificity and may be used to discriminate between AL patients with various degrees of remission. The m/z 4625 peptide may be used to monitor the MRD levels and provide independent prognostic information in patients with AL.
Keywords: acute leukemia; mass spectrum; minimal residual disease; peptide pattern.