Proteomics-based identification of haptoglobin as a favourable serum biomarker for predicting long-term response to splenectomy in patients with primary immune thrombocytopenia

BACKGROUND: Splenectomy is the most effective treatment for patients with primary immune thrombocytopenia (ITP) who fail to respond to steroid therapy. Thus far, there is no effective means to predict the long-term haematological response of the procedure. The purpose of this study was to identify serum biomarkers as predictors of long-term response based on a proteomics approach. METHODS: The serum samples of ITP patients were collected before splenectomy and seven days after surgery. After depletion of the abundant serum proteins, pooled preoperative serum samples from four responders to splenectomy, four nonresponders and four healthy controls were subjected to two-dimensional gel electrophoresis (2-DE). Nine protein spots with at least a five-fold alteration in expression between responders and nonresponders were all identified as haptoglobin (Hp) by matrix-assisted laser desorption/ionisation time-of-flight (MALDI-TOF) mass spectrometer (MS) analysis. The validation of serum Hp expression was performed using enzyme-linked immunosorbent assays (ELISA) in thirty-seven responders, thirteen nonresponders and twenty-one healthy controls. RESULTS: The preoperative serum levels of Hp in the nonresponders (925.9 ± 293.5 μg/ml) were significantly lower than those in the responders (1417.4 ± 315.0 μg/ml, p <0.001) and the healthy controls (1409.1 ± 354.2 μg/ml, p <0.001), while there was no significant difference between the latter two groups. The postoperative serum levels of Hp in responders and nonresponders were (1414.1 ± 225.0 μg/ml) and (952.9 ± 202.4 μg/ml), respectively. There were no significant differences between the serum Hp levels before and after surgery in both responders and nonresponders (p>0.05). The preoperative serum levels of Hp did not significantly correlate with preoperative platelet count of the same blood samples (r = 0.244, p = 0.087), while it positively correlated with postoperative peak platelet count (r = 0.622, p < 0.001). The optimal cutoff value of preoperative serum Hp levels (1173.80 μg/ml) derived from the receiver operating characteristic (ROC) curve led to 78.4% sensitivity and 84.6% specificity. CONCLUSIONS: These results suggest that serum Hp levels may serve as a favourable predictor for the long-term response to splenectomy in ITP and may help to understand the pathophysiological differences between responders and nonresponders.