Factors distinguishing leukemoid reaction from hematological malignancy in children.

BACKGROUND

Information on the etiology and prognosis for leukemoid reaction (LR) in children is still limited and little is known about the factors that distinguish LR from hematological malignancy (HM).

METHODS

This was a single-center, case-control study. Pediatric patients (<18 years) with a white blood cell (WBC) count of 50,000/μL or more were included in the study. Clinical information and laboratory test results were extracted from the electronic medical records. Patients were divided into the LR and HM groups. Logistic regression analysis was performed to investigate the factors that discriminated LR from HM.

RESULTS

We found 214 cases (115 cases in the LR group and 99 cases in the HM group) eligible for analysis. Approximately half of the LR cases were due to infectious diseases (n = 58, 50%); bacteremia and respiratory infections were the most common infections. Age younger than 2 years (odds ratio [95% confidence interval]) (2.154 [0.690-6.727]), presence of known underlying diseases (10.006 [3.119-32.102]), WBC count <60,555/μL (20.676 [6.357-67.251]), platelet count 118,000/μL or higher (15.059 [3.876-58.504]), lactate dehydrogenase (LDH) below 781 U/L (4.219 [1.378-12.915]), and C-reactive protein (CRP) ≥0.91 mg/dL (10.568 [2.736-40.825]) were identified as the predictive factors for LR by logistic regression analysis. Thirty-day mortality was higher in the LR group than in the HM group but the difference was not statistically significant (13% vs. 6%, p = 0.087).

CONCLUSIONS

Approximately half of the etiology of LR was infectious diseases and prognosis was poorer for LR than for HM. Age, presence of known underlying conditions, and laboratory tests, including WBC count, platelet count, LDH, and CRP, may be useful in distinguishing LR from HM.