A longitudinal analysis of pseudohyperkalemia in patients with chronic lymphocytic leukemia (CLL).
Lauren Katkish1, Thomas Rector2, 1, Areef Ishani3, 4, Pankaj Gupta5, 6
1
Department of Medicine, University of Minnesota, Minneapolis, MN
Center for Chronic Disease Outcomes Research, Minneapolis VA Health Care System, Minneapolis, MN
3
Nephrology and 5Hematology/Oncology Sections, Minneapolis VA Health Care System, Minneapolis, MN
4
Nephrology and 6Hematology/Oncology/Transplantation Divisions, Department of Medicine, University of
Minnesota, Minneapolis, MN
2
BACKGROUND: Pseudohyperkalemia in patients with leukocytosis due to chronic lymphocytic leukemia
(CLL) is reported in case studies, and has led to unnecessary or inappropriate potassium-lowering interventions.
It is important to be able to estimate the likelihood of this phenomenon when interpreting measured potassium
levels in patients with CLL. However, there are no studies that have systematically related elevated potassium
levels with elevated WBC counts in a large cohort of patients with CLL over time, that can provide guidance
for clinical decision-making.
METHODS: We identified 310 patients diagnosed with CLL between 1997 and 2014 from the Tumor Registry
of the Minneapolis VAHCS. Patients who had alternative causes for hyperkalemia were excluded. A total of 57
eligible patients with a white blood cell (WBC) count that reached 50.0 x 109 cells/L during the course of their
disease were included in the analysis. All WBC counts and plasma potassium levels measured within 12 hours of
each other were recorded. A total of 1,119 data points were extracted for 57 male patients, age 49-95 years at
diagnosis, with a WBC count between 5.0 x 109 cells/L and 282.6 x 109 cells/L. Longitudinal fixed effects linear
regression was employed to test for a relationship between WBC counts and differences between the measured
plasma potassium concentrations and the upper limit of normal (ULN) for the assay.
RESULTS: Overall, 19% of potassium values were > ULN, and 7.3% exceeded the ULN by at least 0.5 mmol/L.
For every increase of 100.0 x 109 WBC/L, the potassium value increased by 0.5 mmol/L on average. The adjusted
odds of a patient’s potassium level being above the ULN increased by 1.4 (95% confidence interval, 1.2-1.5; p <
0.0001) with every 10.0 x 109 cells/L increase in WBC counts. Below a WBC count of 50.0 x 109 cells/L, the
median estimated percentage of a patient’s potassium values being above the ULN was low (1.7%; IQR, 0.93.45), whereas the estimated percentage above the ULN was 8.1% (IQR 3.9-19) when WBC count was ≥ 100.0 x
109 cells/L. However, within patients, variation in WBC counts explained only part of the variation in their
potassium values.
CONCLUSIONS: A considerable proportion of measured plasma potassium values are elevated in patients with
CLL and high WBC counts. It is likely that the majority of these values represent pseudohyperkalemia. However,
a “correction factor” cannot be created to account for pseudohyperkalemia because it is not possible to predict the
potassium value based on the WBC count alone. This is likely a consequence of the erratic effect of diverse
artifactual phenomena that influence potassium measurement in patients with CLL. Clinical judgment therefore
needs to be used when interpreting potassium values in such patients. We recommend that a warning in the
electronic medical record system be placed to alert clinicians to the potential for pseudohyperkalemia in patients
with CLL and WBC count ≥ 50.0 x 109 cells/L.