Postsplenectomy thrombocytosis with pseudohyperkalaemia

Abstract

A 52-year old man developed hyperkalaemia on the 11th postoperative day following an extensive open retroperitoneal liposarcoma resection that included splenectomy. Despite thorough investigations, no aetiology for the hyperkalaemia was identified and standard empirical treatment was ineffective. On reconsideration, in view of the patient's concurrent thrombocytosis, a pseudofactual or artefactual hyperkalaemia was suspected. This was confirmed by contemporaneous testing of serum and plasma potassium levels, with the latter value lying within the normal range. Treatment for hyperkalaemia was discontinued, thus averting an iatrogenic and potentially dangerous hypokalaemia. This case highlights pseudohyperkalaemia as an often-neglected cause of elevated serum potassium levels and discusses its association with thrombocytosis following splenectomy.

Background

Hyperkalaemia is a well-recognised and common electrolyte abnormality. Pseudohyperkalaemia, however, is incompletely understood and under-diagnosed despite the potentially serious consequences of its misidentification.

Pseudohyperkalaemia is the abnormal accumulation of potassium in blood as a result of its release from cells during in vitro clotting within the standard collecting specimen tubes.^{1 2} This gives the false impression of raised potassium on laboratory serum measurements. Failure to recognise these spurious laboratory results can lead to overtreatment of hyperkalaemia and can be detrimental to patients, and should thus be promptly averted. The case report that follows aims to raise clinical awareness of this condition, so that it may be anticipated and/or recognised early. To our knowledge, this is the first case, in the UK, of pseudohyperkalaemia following splenectomy and as part of cancer surgery.

Case presentation

An otherwise healthy 52-year-old man presented to his local hospital with non-specific abdominal pain in the absence of any other symptoms. He had previously undergone laparoscopic cholecystectomy for cholelithiasis, but his medical history was otherwise unremarkable. He was not on any regular medication, and had no family history of note and no known allergies. The patient had never smoked and reported a weekly alcohol consumption of 15 UK units. Physical examination of all major systems was unremarkable.

Despite the absence of clinical signs, general blood tests revealed an abnormal liver biochemistry profile, which prompted imaging studies. MR Cholangiopancreatography (MRCP) demonstrated left hydronephrosis, and CT of the abdomen and pelvis showed inflammatory changes within the pancreatic tail, in proximity to the left kidney and in association with localised lymphadenopathy. A subsequent CT-guided biopsy established the diagnosis of a retroperitoneal de-differentiated liposarcoma, for which the patient was referred to the regional Sarcoma Service.

The patient underwent an uneventful 5½ h operation under general anaesthesia for tumour resection via laparotomy. Owing to the tumour's anatomical relationships, the resection involved distal pancreatectomy, splenectomy, left nephrectomy, left hemicolectomy and, finally, local resection of the left hemidiaphragm. Furthermore, a right-iliac-fossa colostomy was formed and a left-sided chest drain was inserted, as well as a left-upper-quadrant abdominal drain.

Postoperative imaging revealed moderate bilateral pleural effusions with bibasal atelectasis, and a left diaphragmatic hernia with small bowel herniation and obstruction. On the seventh postoperative day, the patient was taken back to theatre, where the herniated bowel was reduced. Thereafter, he remained in the ward to recover. Despite an otherwise stable and satisfactory recovery, an incidental and asymptomatic hyperkalaemia at 5.7 mmol/L was noted on the 11th postoperative day.

Investigations

ECG was initially performed to exclude arrhythmia and conduction abnormalities. As this demonstrated normal sinus rhythm in the absence of any characteristic hyperkalaemic changes, blood tests were repeated for confirmation. Over the next 3 days, potassium levels fluctuated between 5.3 and 5.9 mmol/L, alongside serum sodium levels of 130–132 mmol/L, creatinine levels of 95–102 μmol/L, normal urea levels and high, yet steadily decreasing, inflammatory markers.

At a peak potassium value of 6.6 mmol/L on the 15th postoperative day, supplementary investigations were carried out to exclude various causes of persistent hyperkalaemia. These included paired urine-serum osmolalities (714 and 281 mOsm/L respectively), urinary sodium (203 mmol/L), fT4 (13.5 pmol/L), thyroid-stimulating hormone (TSH) (4.82 munit/L), early morning cortisol levels (569 nmol/L) and bicarbonate (21 mEq/L).

Differential diagnosis

Hyperkalaemia in patients after surgery has a variety of causes so our range of differential diagnoses was broad. It is crucial to discover the underlying cause, as the treatment required for each culprit is different. Treating the wrong cause can be detrimental for the patient.

Renal dysfunction and renal tubular acidosis were excluded by the unremarkable urea, creatinine and electrolyte levels (urea range: 4.5–7.5 mmol/L, sodium: 132–139 mmol/L, creatinine: 85–115 µmol/L). Adrenal insufficiency was excluded by the normal serum cortisol and aldosterone levels as well as a normal urinary sodium value (cortisol: 692 nmol/L, urinary sodium: 203 mmol/L, aldosterone levels: 450 pmol/L (normal range 150–550 pmol/L). Normal thyroid function tests excluded thyrotoxicosis (fT4: 15.3 pmol/L, TSH: 4.1 mU/L).

Drug-induced hyperkalaemia is an important, often overlooked cause of high potassium levels. Drugs with significant hyperkalaemic properties include potassium-sparing diuretics, non-steroidal anti-inflammatory drugs (NSAIDs), ACE inhibitors, angiotensin receptor blockers and immunosuppressants. Our patient's drug regimen included none of the above, and consisted of prophylactic aspirin and phenoxymethylpenicillin (penicillin V), cyclizine, paracetamol and prophylactic tinzaparin. Although heparin can potentially lead to hyperkalaemia via hypoaldosteronism, normal serum sodium and aldosterone–renin ratio excluded this cause in our case. Finally, the absence of previous chemotherapy or trauma excluded tumour lysis syndrome and rhabdomyolysis, respectively.

Treatment

As the underlying cause of hyperkalaemia was initially unclear, empirical treatment was initiated. This consisted of calcium gluconate for cardioprotection, short-acting insulin with intravenous dextrose, nebulised salbutamol and furosemide as per hospital guidelines for asymptomatic moderate-to-severe hyperkalaemia. However, initial treatment was ineffective as demonstrated by the persistent hyperkalaemia. A concurrent tachycardia at between 100 and 120 bpm was noted, which was incongruous to the bradycardia one would expect in the setting of hyperkalaemia.

A repeat ECG confirmed sinus tachycardia without any hyperkalaemic conduction abnormalities. Empirical treatment continued as the patient (asymptomatic throughout) was placed on continuous cardiac monitoring.

At this stage, the aetiology was reconsidered by focusing on the patient's recent procedure, which could temporally be implicated in hyperkalaemia. Platelet levels had risen from a preoperative baseline of 243×10⁹/L to a postoperative peak of 1643×10⁹/L due to asplenia, for which the patient received daily prophylactic doses of aspirin and tinzaparin. Therefore, the possibility of thrombocytosis-induced pseudohyperkalaemia was entertained. To test this theory, contemporaneous blood samples were taken to measure serum and plasma potassium levels. Serum potassium levels were measured after blood collection in a gold-top BD Vacutainer Serum Separator Tube (BD SST II Advance), whereas plasma levels were measured from a green-top BD Vacutainer Lithium Heparin Plasma Tube. While serum potassium levels remained high at 6.5 mmol/L, plasma levels were normal (4.4 mmol/L). Interestingly, potassium levels on arterial blood gas (ABG) analysis were consistent with this value on two separate occasions (4.4 mmol/L and 4.3 mmol/L, respectively).

Based on this result, empirical treatment was finally discontinued and, in consultation with the local renal medicine, haematology and endocrinology teams, a diagnosis of pseudohyperkalaemia secondary to postsplenectomy thrombocytosis was established.

Outcome and follow-up

The patient was discharged home 4 weeks postoperatively without complications or treatment for hyperkalaemia. He has remained well since then and no further intervention has been required. His latest clinic appointment was 1 month postdischarge when his potassium levels were 5.2 mmol/L. The thrombocytosis was also resolving with a gradual drop from a peak value of 1640×10⁹/L to 859×10⁹/L measured on his latest clinic appointment.

In addition to preventing unnecessary treatment in our patient, managing the current case has influenced local clinical practice by raising awareness of this potentially dangerous and misleading condition, and thereby permitting swift diagnosis in subsequent cases.

Discussion

Hyperkalaemia is a common electrolyte disturbance that most often arises secondary to well-known clinical entities such as renal failure, adrenal insufficiency, thyrotoxicosis, trauma, tumour lysis syndrome or as a result of medications such as potassium-sparing diuretics, NSAIDs, ACE inhibitors and heparin. Normal serum potassium levels are between 3.5 and 5.3 mmol/L, as per the Pathology Harmony Agreed Adult Clinical Biochemistry Reference Intervals, although the normal reference range can still vary between laboratories.³ A potassium level above 6.5 mmol/L must be treated quickly to avoid arrhythmias and sudden cardiac death.⁴

Pseudohyperkalaemia is characterised by an elevated serum potassium level alongside a normal plasma potassium level. As such, it is diagnosed when the difference between serum and plasma potassium is greater than 0.4 mmol/L in the presence of raised serum potassium levels.^{5 6} We extend this definition to include the absence of ECG changes, absence of clinical manifestations of electrolyte imbalance and persistence of serum hyperkalaemia despite empirical treatment. Therefore, an artefactual result should be suspected when the hyperkalaemia persists without apparent aetiology, symptoms, ECG changes and response to treatment. It is a rarely reported phenomenon and is mainly attributed to the release of potassium from blood cells during the in vitro clotting process.² It can occur in the presence of leucocytosis, erythrocytosis, haemolysis and thrombocytosis, as intracellular potassium leaks from these cellular components of blood.^{7 8}

Causes of pseudohyperkalemia include an autosomal dominant familial syndrome in which red blood cells exhibit an abnormal potassium leak,^{9 10} hereditary spherocytosis,¹¹ leucocytosis of chronic inflammation such as in rheumatoid arthritis and infectious mononucleosis,^{12 13} chronic myeloid or lymphocytic leucaemia,^{14 15} essential thrombocythaemia^{16 17} and reactive postoperative thrombocytosis.¹⁸

Additionally, mechanical factors may disrupt cell membranes or promote potassium release in the bloodstream.⁷ These include prolonged tourniquet application, fist clenching, traumatic venepuncture, excessive force during syringe drawing and, occasionally, pneumatic tube transport of blood samples in cases of cell membrane fragility (eg, leucaemia). Storage of samples outside the range of 15–25°C and delays in processing can also promote cellular efflux of potassium via inhibition (in low temperatures) or exhaustion (in high temperatures or delayed processing) of the sodium-potassium pump. Furthermore, contamination of samples may occur by ethanol from antiseptic wipes, potassium from intravenous fluids or EDTA from collection tubes. The practical implication in the latter case is that blood samples for serum potassium measurement using serum tubes should be collected before samples for full-blood count (EDTA tubes). Finally, hyperventilation due to anxiety or crying can lead to hyperkalaemia via respiratory alkalosis and enhanced α-adrenergic activity.⁷ The factors described above should be considered in liaison with the Clinical Biochemistry Laboratory prior to assuming medical pathology, in order to avoid inadvertent treatment.

A review of the literature has identified only four reports of postsplenectomy thrombocytosis leading to pseudohyperkalaemia (table 1).^19–22 The clinical and biochemical features in all cases support our earlier suggestions on when to suspect and how to diagnose postsplenectomy thrombocythaemia-associated pseudohyperkalaemia. Additionally, all cases emphasise the fact that potassium-lowering treatment could cause life-threatening hypokalaemia and should be avoided, since spurious hyperkalaemia resolves naturally as platelet counts normalise. Of important note are cases where the misdiagnosis led to life-threatening overtreatment with haemodialysis.²¹

Table 1.

Summary of demographic, clinical and biochemical features of other cases of postsplenectomy pseudohyperkalaemia reported in the literature

Age/sex	Cause for splenectomy	Postoperative day of elevated potassium	Serum K+ (mmol/L)	Plasma K+ (mmol/L)	Thrombocytosis (×109/L)	Renal function	ECG changes	Symptoms related to electrolyte anomalies	Response to standard treatment	Reference
24/M	Grade IV splenic injury after accident	7	5.6	3.8	1634	Normal	Nil	Nil	Not initiated as diagnosis made immediately	18
49/M	Grade IV splenic injury after accident	16	5.9	3.9	NA	NA	NA	NA	NA	19
42/M	Hepatosplenic schistosomiasis with severe portal hypertension and bleeding oesophageal varices	365	7.0 (also high serum 8.9 mg/dL)	Normal (value not specified)	2 300 000		Nil	Nil	Nil	20
18/M	Grade V splenic injury after accident	9–10	5.7	NA	1737 (also white cells of 26.9)	Normal	Nil	Nil	Not initiated as K+ normalised with drop in platelet and WCC levels	21
22/M	Profuse bleeding from splenic hilum after accident	14	5.5	NA	1818 (also white cells of 29.2)	Normal	Nil	Nil	Not initiated as K+ normalised with drop in platelet and WCC levels	21

NA, not available; WCC, white cell count.

Thrombocythaemia/thrombocytosis logically arises when the rate of platelet production exceeds the rate of platelet destruction or sequestration. Excessive platelet production occurs in myeloproliferative haematological disorders such as polycythaemia rubra vera and essential thrombocythaemia, as well as in states of inflammation (eg, chronic infections, rheumatoid arthritis). Inadequate sequestration or destruction, on the other hand, can occur in states of either functional or actual asplenia/hyposplenism.¹⁹ Persistent thrombocytosis is expected following splenectomy, particularly in the first three postoperative months.^{23 24} Samples for serum potassium measurement are normally collected and centrifuged in serum separator tubes (SST; ‘gold-top’ tubes), which contain silica and an acrylic-based gel polymer. The silica particles encourage coagulation, whereas the gel forms a barrier between serum and blood cells, thus stabilising the sample and enabling analysis following centrifugation. As platelets aggregate and degranulate in vitro prior to sample centrifugation and separation, intracellular potassium is released.⁵ Thus, in the setting of thrombocytosis, potassium levels tend towards artefactual elevation. If pseudohyperkalaemia is suspected, a concurrent blood sample should be collected in a heparinised tube (such as the green-top BD Vacutainer Lithium Heparin Plasma Tube), which avoids the above process and allows for a more representative measurement.⁵

The link between platelet and serum potassium counts dates back to 1955 when Hartmann and Mellinkoff¹ noticed a ‘spurious hyperkalaemia’ in the presence of thrombocytosis and hypothesised that potassium was released from platelets during the in vitro clotting process. Using platelet-free plasma from the same patient yielded normal potassium levels.² Given that thrombocytosis affects serum but not plasma potassium ion concentration, the latter is a more reliable measurement and its use automatically excludes pseudohyperkalaemia.⁵ For the same reason, we believe that arterial or venous blood gas analysis (which uses heparinised syringes) could be a useful, quick test in cases of high serum potassium in which pseudohyperkalaemia is suspected and plasma potassium levels are pending. As in our case, when retrospectively examined, potassium values on ABG were, in fact, consistent with plasma potassium levels on two separate occasions. Further studies would be necessary to evaluate the clinical utility and reliability of ABG in such circumstances.

Despite the correlation between thrombocytosis and serum potassium levels reported in the literature and in our case (figure 1), the proportional rise in potassium tends to be variable rather than to show a direct relation to the number of platelets, thus it has been difficult to uniformly correct potassium measurements based on platelet counts.^{25 26} Furthermore, pseudohyperkalaemia has been shown to occur only at platelet and leucocyte counts of more than 600×10⁹/L and 70×10⁹/L, respectively.^{14 18 19 27} Furthermore, the reason why pseudohyperkalaemia only occurs in certain cases of asplenia remains to be elucidated.

Figure 1.

Graph showing the rise in serum potassium levels as measured in Serum Separator (‘gold-top’) tubes (y axis) and blood platelet count (x axis) for the patient postsplenectomy. Each point represents a unique day on which measurements were carried out. The trend illustrates a strong and statistically significant positive correlation (R=0.91, p<0.0001****) between the two variables highlighting the link between thrombocytosis and pseudohyperkalaemia.

This case demonstrates that pseudohyperkalaemia can occur in postsplenectomy patients and, as such, requires a high degree of clinical suspicion in order to avoid inadvertent hypokalaemia through empirical treatment. Furthermore, the anticipation and prompt diagnosis of pseudohyperkalaemia is relevant to clinicians in surgical as well as medical specialties, and should always be considered in the differential diagnosis of hyperkalaemia, especially in cases of concomitant thrombocytosis.

Patient's perspective.

• During my stay at the hospital, being treated for a liposarcoma, the medical team members were concerned because I had raised potassium levels in my blood. I found this very distressing because of potential problems that could occur.

• As I understand it, with raised potassium levels, there would normally be clinical symptoms and signs on the ECG, however, this was not the case. The treatment for high potassium (hyperkalaemia) was initiated but did not improve the potassium levels.

• The doctors looking after me took blood samples from me a number of times and asked the pathology department to test each sample before it had clotted. When this was performed, the potassium levels actually proved normal. Therefore, I was diagnosed with pseudohyperkalaemia, rather than true hyperkalaemia, which the cardiac, haematology and endocrinology teams also agreed on and confirmed.

• Had this diagnosis not been made, I would have been treated for a condition I did not have, and subjected to unnecessary medications, which again could have led to even more complications.

Learning points.

• Pseudohyperkalaemia should be considered in all hyperkalaemic patients in the absence of symptoms and aetiology, and should be excluded in cases of increased blood cellularity (erythrocytosis, leucocytosis or thrombocytosis) by measuring plasma potassium levels in heparinised plasma tubes or blood gas syringes and comparing to serum potassium levels from serum separator tubes.

• A difference >0.4 mmol/L between serum and plasma potassium alongside a normal ECG without hyperkalaemic changes, absence of clinical features and lack of biochemical response to potassium-lowering treatment suggest a diagnosis of pseudohyperkalaemia.

• Failure to recognise pseudohyperkalaemia can lead to patient harm from iatrogenic hypokalaemia.

• Arterial blood gas analysis may be a reasonable quick test in cases of high serum potassium levels where pseudohyperkalaemia is suspected and plasma potassium levels are pending, in order to avoid unnecessary treatment.