Neutrophilia is a nonspecific peripheral blood abnormality commonly encountered in clinical practice. The differential diagnosis of neutrophilia is broad, including but not limited to reactive states (i.e., infectious or inflammatory conditions), asplenia, tobacco use, iatrogenic from steroid-induced demargination, factitious from laboratory error, and primary clonal disorders.^{1, 2} The plurality of etiologies poses a diagnostic challenge to clinicians who strive to balance comprehensive evaluation with avoidance of excess diagnostic testing.

Chronic myeloid leukemia (CML) is a myeloid malignancy caused by the Philadelphia Chromosome, the fusion of the BCR-ABL genes, and a known cause of neutrophilic leukocytosis. Laboratory testing for BCR-ABL fusion is often performed to evaluate neutrophilia, especially when associated with concomitant basophilia.^{1, 3} However, data regarding the importance of basophilia and other clinical findings to inform testing decisions are lacking. This study aims to aid clinicians in selecting patients who are most likely to derive benefit from BCR-ABL testing in the evaluation of neutrophilia.

We reviewed medical records of patients who received BCR-ABL testing on peripheral blood and/or bone marrow aspirate in the evaluation of neutrophilia, either by qualitative polymerase chain reaction (PCR) or fluorescence in situ hybridization (FISH), between January 1st, 2014 and December 31st, 2018 (https://www.mayocliniclabs.com/test-catalog/overview/89006). Neutrophilia was defined as an absolute neutrophil count (ANC) > 6.45 × 10⁹/L and basophilia was defined as an absolute basophil count >0.08 × 10⁹/L per institutional parameters. Patients with a prior diagnosis of CML were excluded. Patient information including age, sex, complete blood count (CBC) and white blood cell (WBC) differential values, and the presence or absence of splenomegaly at time of BCR-ABL testing was assessed. The presence of splenomegaly was determined by documented physical exam and available imaging. CBC and WBC differential were assessed by automated methods with manual peripheral blood smear review when appropriate. CML diagnosis was defined by the World Health Organization 2016 criteria.⁴ Data were presented descriptively with medians, ranges, and percentages where appropriate. Two-tailed t-test and Pearson Chi-squared analysis were utilized for continuous and categorical variables respectively. Test characteristics (sensitivity, specificity, positive likelihood ratios [LR+], and negative likelihood ratios [LR−]) for the diagnosis of CML were calculated.

Seven-hundred and fifty patients with neutrophilia received BCR-ABL testing. Patient demographics and laboratory information are presented in Table 1. The median (range) age was 60 years (18–95 years) and 422 patients were female (51%). Documentation related to the absence or presence of splenomegaly was available for 678 patients (91%), of these 136 (20.1%) were noted to have splenomegaly. Median total WBC count (range) was 15.1 × 10⁹/L (9.7–746.5 × 10⁹/L) and the median ANC was 10.5 × 10⁹/L (6.46–408.8 × 10⁹/L). Common peripheral blood abnormalities noted to coincide with neutrophilia were absolute monocytosis (62.9%), absolute basophilia (45.8%), and thrombocytosis (41.5%). Two-hundred and twelve patients had circulating myeloid precursor cells (28.3%), 99 patients had circulating blasts (13.2%), and 117 patients had nucleated red blood cells on peripheral blood (15.6%).

One-hundred and thirty-two patients were diagnosed with CML (17.6%) (Table 1). The majority of patients without CML were noted to have reactive leukocytosis. Compared with the rest of patients, individuals diagnosed with CML had higher total WBC count, neutrophil count, and platelet count, lower hemoglobin, and were more likely to have splenomegaly, basophilia, circulating myeloid precursors, nucleated RBCs, and blasts. Sensitivity, specificity, and likelihood ratios are listed in Table 1. The sensitivity of basophilia for diagnosis of CML was 85.6%. Findings with greatest specificity for CML diagnosis included circulating blasts (94.3%), nucleated red blood cells (90.3%), ANC > 20.0 × 10⁹/L (88.8%), splenomegaly (85.0%), and circulating myeloid precursors (83.4%). All patients diagnosed with CML had either ANC > 20.0 × 10⁹/L, absolute basophilia, or circulating myeloid precursors. Over half of patients presenting with these findings were eventually diagnosed with a myeloid malignancy.

There are several notable observations from the present study. Compared with individuals with alternate etiologies for neutrophilia, patients with CML exhibited markedly different laboratory findings, including aberrancies of the WBC differential, platelet count, and hemoglobin. The degree of neutrophilia associated with CML was significantly higher than patients with reactive processes, with two-thirds of patients diagnosed with CML presenting with ANC > 20 × 10⁹/L. This is an important consideration for clinicians as they distinguish the probability of CML with additional causes of longstanding neutrophilia such as tobacco use and other chronic inflammatory states which are associated with lower ANC elevations.²

Basophilia was present in 86% of patients with CML. In contrast, a prior study by Ogunleye and colleagues reported the sensitivity of basophilia (defined as absolute basophil count >0.1 × 10⁹/L) for CML diagnosis to be 100%; a finding which prompted the authors to conclude that BCR-ABL testing should be restricted only to individuals presenting with elevated absolute basophil count.³ However, the present study notes that this limitation would exclude one in seven (14%) patients with CML. Careful evaluation of the entire clinical context is warranted in determining whether patients without basophilia may benefit from CML screening. Basophilia was present in over a third of patients with reactive neutrophilia, owing to the plurality of conditions which may result in this peripheral blood finding.⁵

The highest specificity for CML diagnosis was observed for peripheral blood blasts, nucleated red blood cells, ANC > 20 × 10⁹/L, and splenomegaly. Approximately 40% of patients with CML presented with splenomegaly, a finding consistent with prior reports and further encouraging the clinical utility of physical examination in patients presenting with neutrophilia.⁶ The highest positive likelihood ratio for CML diagnosis was noted for patients with circulating blasts and ANC > 20 × 10⁹/L, a group who appear to benefit the most from BCR-ABL testing. All patients diagnosed with CML either had ANC > 20.0, absolute basophilia, and/or circulating myeloid precursors. This finding suggests that BCR/ABL testing can be safely avoided in patients not exhibiting any of these peripheral blood findings; over 40% of the patients in the present study.

In conclusion, the vast array of etiologies for neutrophilia pose a challenge for clinicians aiming to optimize selection of patients who are most likely to benefit from BCR-ABL testing. The presence or absence of basophilia alone is not a sufficient discriminator in this process. Comprehensive clinical evaluation including spleen examination, assessing the degree of neutrophilia, and reviewing for other blood abnormalities remain key guiding considerations to avoid excess diagnostic testing. Patients with an ANC <20.0, who do not have basophilia or circulating myeloid precursors should not receive BCR/ABL testing as part of routine neutrophilia evaluation.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

PATIENT CONSENT

Study approved by Mayo Clinic Institutional Review Board. Patient consent waived due to retrospective nature of this minimal risk study. All analyzed clinical data are anonymous in presentation.