Critically Elevated Potassium in a 55-Year-Old Female With Chronic Lymphocytic Leukemia

Jing Cao, PhD; Amy B. Karger, MD, PhD


Lab Med. 2018;49(3):280-283. 

In This Article


Hyperkalemia can occur due to a variety of causes. Those causes include including renal failure, adrenal insufficiency, disorders associated with massive tissue breakdown (eg, trauma, rhabdomyolysis, marked hemolysis, or TLS), and pseudohyperkalemia.

TLS refers to a group of metabolic complications caused by the breakdown products of dying cancer cells. These complications include hyperkalemia, hyperphosphatemia, hypocalcemia, hyperuricemia, and hyperuricosuria, with the subsequent consequence of acute uric acid nephropathy and acute renal failure.

Hyperkalemia can alter cardiac excitability with significant changes in the ECG reading, including the induction of cardiac arrhythmias, which requires an aggressive therapeutic approach. Increased potassium levels are associated with a higher probability of ECG changes, and potassium levels higher than 8 mmol per L are almost invariably associated with ECG changes.[1] There has been an occasional report[2] of patients with extremely elevated potassium levels but unremarkable ECG findings, generally in patients with renal failure.

The term pseudohyperkalemia refers to an increased in vitro blood potassium level in the absence of in vivo hyperkalemia. The causes of pseudohyperkalemia include variables in blood specimen collection, such as fist clenching, mechanical trauma during or after phlebotomy, contamination from ethanol containing antiseptics, K+ carryover from K+–ethylenediaminetetraacetic acid (EDTA) or oxalate/fluoride tubes, refrigeration before centrifugation, delay in centrifugation, decreased transport or storage temperature, pneumatic tube transport, hemolysis, and clotting. Also, clinical conditions of leukocytosis, including CLL or infectious mononucleosis; thrombocytosis associated with myeloproliferative disorders; and erythrocyte disorders, such as in familial pseudohyperkalemia, renal disease, and rheumatoid arthritis may also present along with pseudohyperkalemia.

In CLL with extreme leukocytosis, cell lysis with release of intracellular K+ can result in pseudohyperkalemia. Clotting in serum collection was originally believed to contribute to the in vitro cell lysis; this theory is supported by a number of case reports on patients with CLL in whom hyperkalemia was noted in serum but not in plasma or in WB analyzed with a blood gas analyzer.[3] However, conflicting observations were also reported[4] in a patient with leukemia who had pseudohyperkalemia in plasma but not in serum or WB.

A few studies tried to address the discrepancy in K+ levels in different blood specimen types based on the cell sources of K+. Sevastos et al[5] studied 435 patients with thrombocytosis, erythrocytosis, leukocytosis, or mixed-type disorders and reported that different cell-type disorders contributed variably to the elevation of K+ levels. Patients with platelet, erythrocyte, or mixed-type disorders showed more increased levels of K+ in serum compared with plasma, whereas a subgroup of 20 patients with CLL or AML showed comparable increase in levels of K+ in serum and plasma. In another study of 37 paired specimens from patients with cancer,[6] K+ levels in serum were significantly greater than those in plasma, on average. Platelet counts were significantly associated with K+ levels in serum but not in plasma.

In this case, the laboratory measured K+ levels on the main chemistry analyzer (VITROS 5600) and the ABL90 FLEX blood gas instrument using heparin plasma. Both results demonstrated significant hyperkalemia which, in conjunction with the previous findings of hyperuricemia at the OSH, led to the diagnosis of TLS, despite that the null finding on ECG did not fully support the diagnosis of in vivo hyperkalemia.

Role of the Laboratory in Diagnosis

The patient was transferred to the medical intensive care unit (MICU). Intermittent hemodialysis (IHD) was performed for 2 hours, followed by left internal jugular catheter placement and overnight continuous renal replacement therapy (CRRT). The team started the patient on prednisone and planned to decide the next step once her K+ level became stabilized. Two hours after initiating intermittent hemodialysis, the plasma K+ level had dropped significantly to 5.6 mmol per L—such a decline would exceed the expected speed of K+ removal via hemodialysis (typically, 25 to 50 mmol of K+/hour). The potassium level, as measured the next morning after 6 hours of overnight CRRT, was 7.1 mmol/L, a value higher than the previous result. This finding did not make sense to us because the patient had undergone several hours of additional dialysis between measurements.

Given the unexpectedly large decrease in potassium after IHD and the increase in potassium level after CRRT, the health care providers suspected pseudohyperkalemia. It was noted that ionized calcium had been measured in WB at the same time as the 2 K+ measurements, so the laboratory measured a K+ level on both remnant WB specimens using the ABL90 FLEX blood gas analyzer. The WB K+ levels were 2.4 mmol per L and 2.5 mmol per L—these values were significantly discordant from the simultaneous plasma K+ measurements of 5.6 and 7.1 mmol per L (Table 1), and consistent with pseudohyperkalemia. After additional investigation within the laboratory, it was discovered that the initial confirmatory K+ test on the ABL90 FLEX blood gas instrument was run on a plasma specimen, rather than an uncentrifuged WB specimen. Therefore, given the discrepancy in K+ values between uncentrifuged WB and centrifuged plasma, it was concluded that centrifugation likely caused disruption of the CLL cells leading to pseudohyperkalemia.

Patient Follow-up

Although the patient underwent unnecessary dialysis with catheter placement, she did not experience any significant harm from this misdiagnosis. Given the documented susceptibility of patient specimens to centrifugation, the clinical team and laboratory put a plan in place to measure all future K+ levels in WB on the ABL90 FLEX blood gas analyzer to avoid breakage of cells via centrifugation. Also, according to this plan, all future specimens will be hand-delivered to the laboratory to avoid any potential for disruption of cells in the pneumatic tube system, although there was no evidence that disruption of cells was a contributing factor in this case. The patient was subsequently started on chemotherapy for CLL and died 3 months later.

The patient had no significant ECG findings despite that her K+ levels were higher than 9 mmol per L; in retrospect, this information could have been a clue that pseudohyperkalemia was the etiology. However, the similar results on the main chemistry analyzer (VITROS 5600) and ABL90 FLEX blood gas analyzer at presentation led the clinical team to discount the possibility of pseudohyperkalemia because the results were concordantly high. This finding highlights the important need for close communication between the laboratory and clinical care team—earlier discovery that both specimens were plasma could have avoided the unnecessary dialysis procedures. This case serves as a strong reminder that in patients with extremely elevated WBC counts or other cell counts, pseudohyperkalemia should be seriously considered as a potential cause for elevated K+ levels and actively ruled out to prevent unneeded K+-lowering therapies.