Older aged patients
Older age seems to correlate with the propensity to develop post-operative hyperkalemia in patients with primary hyperaldosteronism. Subjects aged more than 53 years old had greater odds of developing post-op hyperkalemia than those aged less than 53 years (OR = 15.6) as described in the Korean study [
5]. Fischer et al. noted that patients who developed hyperkalemia were significantly older as compared to normokalemic population (65 ± 9 vs. 50 ± 12 years.
P = 0.026) [
4]. This theme is consistent in the study by Chiang et al.: Patients who developed hyperkalemia were older (56.4 ± 4 vs. 46.5 ± 10.0 year.
P = 0.002). However, in the same study risk factors were analyzed further using multivariate regression and after backward elimination, older age was not one of the independent risk factors [
3]. It is possible that older patients had longer duration of hypertension with more advanced hypertensive co-morbidities including hypertensive nephropathy.
Duration of hypertension
Duration of pre-operative hypertension has been shown to be another potentially important risk factor for developing post adrenalectomy hyperkalemia. In the study by Chiang et al. [
3], post-op hyperkalemics had longer duration of hypertension (12.8 ± 9.3 vs. 6.7 ± 5.0 year.
P = 0.013). Park [
5] found that subjects with duration of hypertension greater than 9.5 years had 10 times higher risk of developing hyperkalemia than those with less than 9.5 years of hypertension in an age adjusted model. All of the three subjects presented in our report had a lengthy duration of hypertension (> 10 years) prior to diagnosis and surgery.
Impaired renal function
Impaired renal function is an important risk factor in predicting post-operative hyperkalemia. The Chiang paper demonstrated that CKD (Stages III to–V) was 30 % more prevalent in patients who ultimately developed hyperkalemia [
3]. Park took a more formal approach performing a ROC analysis to determine the most likely cut off values for pre-operative eGFR’s ability to predict post-operative hyperkalemia; an eGFR of < 58.2 ml/min was associated with increased risk of developing hyperkalemia (OR = 26.6
p < 0.05) [
5]. Fisher found that post-op hyperkalemics had significantly worse pre-operative renal function (creatinine: 91umol/l [76; 114 umol/l] vs. 61umol/l [53; 76 umol/l],
P = 0.001; GFR: 56 ± 17 vs. 84 ± 21 ml/min,
P = 0.024) [
4].
While overt pre-existing renal impairment may be a strong factor in predicting post-operative hyperkalemia, there is now evidence that PA itself may induce a hyper-filtration injury that may mask renal impairment until the operative reversal of aldosterone mediated hyper-filtration [
6]. In the first case presented above, the patient did not have any evidence of renal impairment pre operatively but after APA resection her eGFR decreased until it plateaued at 30–40 ml/min. Thus it may be that a “normal” eGFR (60–100 ml/min) in the PA pre-operative state may actually portend underlying nephropathy that will become evident and clinically relevant in the post-operative period.
Ribstein et al. [
7] studied 25 patients with tumoral PA. Renal studies (urinary excretion of proteins, GFR, and effective renal plasma flow [ERPF] were performed both before and 6 months after adrenalectomy. A control group consisting of patients with essential hypertension (EH) was studied before and after 6 months of antihypertensive therapy. At baseline, PA and EH patients were similar with respect to demographic data, duration and level of hypertension, and GFR and ERPF. Urinary excretion of albumin and β2 microglobulin (a marker of glomerular filtration) was higher in PA than EH. Adrenalectomy was followed by a decrease in arterial urinary excretion of albumin and β2 microglobulin, and GFR and ERPF. In EH, a similar decrease in pressure was associated with a decrease in albuminuria but no change in GFR or ERPF. In patients who received a 6 months treatment of spironolactone, both GFR and ERPF decreased in parallel with BP, similar to what was observed after surgery. This data suggests that PA was associated with relative hyperfiltration, unmasked after suppression, removal or blockade of aldosterone excess.
A few studies have attempted to identify predictors of decreasing eGFR after APA resection. Tanase-Nakao et al. [
8] studied patients with PA who underwent unilateral adrenalectomy and were followed for 1 month post operatively. Patients who underwent non-PA adrenalectomy were studied as control. Of PA patients, 37.8 % “developed CKD” (defined as eGFR < 60 ml/min/1.73 m2). None of non-PA group developed CKD post op. Of the pre op variables, logistic regression analysis showed that lower pre-surgical eGFR and higher aldosterone-to-renin ratios (ARR) were the independent predictors of post-op CKD in PA. Another Japanese study [
6] prospectively described 120 patients with APA who underwent adrenalectomy while 111 patients with bilateral adrenal hyperplasia (BAH) received MR antagonists. Urine albumin excretion (UAE), blood pressure and eGFR decreased significantly at 1 month after treatment in both arms but did not decrease further at 12 months. Multivariate regression analysis revealed that higher UAE and lower serum potassium at initial visit were independent predictors of decreasing eGFR after intervention.
Ribstein et al. [
7] showed that baseline serum potassium correlated with the post-operative change in GFR by univariate analysis. Rossi [
2] in the PAPY study provided evidence that low serum potassium at baseline predicts albuminuria. Renicke [
9] described that hypokalemic PA tended to be associated with renal impairment. Initial potassium concentration (odds ratio, 1.3; 95 % confidence interval, 1.0–5.2;
P < 0.01), plasma aldosterone concentrations (
P < 0.05), and presence or absence of the hypokalemic variant of PA (
P < 0.05) were independent predictors of higher creatinine.
While hypokalemia may simply be a marker of more severe aldosteronism, it must also be considered that hypokalemia could actually play a pathophysiologic role in renal impairment. Low potassium itself can cause functional and structural defects in kidneys, such as cyst formation, tubular vacuolization, interstitial inflammation and gradual loss of renal function [
10].
Post-operative hypoaldosteronism
Connell et al. [
11] described a case of a 50 year old woman who developed hyperkalemia and acute kidney injury 1 month after resection of APA. Her aldosterone level was inappropriately low, given hypovolaemia and hyperkalaemia, suggesting hypoaldosteronism (plasma aldosterone 100 pmol/l; plasma renin activity 4.4 pg/ml/h). Gadallah [
12] reported a case in which unilateral adrenalectomy for adrenal adenoma was followed by severe hyperkalemia, marked volume depletion and undetectable plasma renin activity and serum aldosterone suggesting chronic suppression of the renin-aldosterone axis. One year later, a gradual return to normokalemia, normotension, and normal plasma renin activity and aldosterone levels was achieved, indicating resolution of the suppression of the renin-aldosterone axis.
In the German Conn’s registry post-operative hyperkalemia caused by suppression of the zona glomerulosa (ZG) insufficiency was examined. They measured serum potassium, post-operative aldosterone and trans-tubular potassium gradient (TTKG) to determine ZG insufficiency which was defined as was defined as undetectable plasma aldosterone level (< 97pmol/l) in the presence of hyperkalemia (K > 5.0). Of the 18 patients who developed hyperkalemia 14 had undetectable and four had low aldosterone levels (aldosterone level < 138 pmol/l). In all six patients who developed persistent hyperkalemia, serum aldosterone level was less than 97pmol/l. The calculation of the TTKG showed decreased values of less than 6, in three of four evaluated hyperkalemic patients, showing inappropriate renal response to hyperkalemia. The normal response to hyperkalemia is an increase in aldosterone secretion leading to an increased urinary potassium excretion, with an increase of TTKG to greater than 10 [
13]. A result of less than 6 in a hyperkalemic patient indicates inadequate aldosterone concentration or effect [
14]. They also assessed the contralateral gland suppression index in those patients who had selective adrenal vein sampling (AVS) prior to surgery (
n = 77). Five of five investigated patients with persistent hyperkalemia demonstrated a contralateral aldosterone/cortisol ratio suppression index of aldosterone in the non-dominant gland less than 1 versus inferior vena cava [
4].
Taniguchi et al. [
15] reported a case of a 46-year-old woman who underwent unilateral adrenalectomy and developed episodes of severe hyperkalemia for 4 months. Plasma aldosterone concentration (PAC) was low (102.5 pmol/l) and undetectable renin activity. Serum potassium levels gradually decreased with concomitant increase in PAC (162pmol/l).
It is plausible that aldosterone synthesis of adjacent and contralateral adrenal glands is severely impaired in some cases with primary hyperaldosteronism, as is glucocorticoid synthesis in Cushing syndrome. Since many instances of hyperkalemia may go undetected post operatively, a larger series with prospective, standardized, sequential potassium measures may be needed in order to better determine if contralateral suppression might predict hyperkalemia.
Pre-operative mineralocorticoid receptor (MR) antagonists
MR antagonists are often administered in APA patients to control hypokalemia. Reversal of hyperaldosteronism after long term treatment with spironolactone (SP) has been reported in the literature, and it appears that it has anti-steroidogenic effect. Yoneda et al. [
16] reported a case of a 41 year old man who had a left sided APA. The plasma renin activity (PRA) was 0.6 ng/ml/h and the plasma aldosterone concentration (PAC) was 362 pg/ml in the supine position. CT showed a 5 mm left sided nodule and AVS indices showed left lateralization. The patient received spironolactone 75 mg OD daily for 5 years and 50 mg for another 4–5 years. At 10 year follow up after discontinuation of spironolactone, the patient’s blood pressure, serum potassium level, and plasma aldosterone concentration remained in the normal range. Although the left adrenal gland tumor was still present on computed tomography after treatment, a furosemide and upright posture test, a captopril challenge test, and a saline loading test produced no evidence of PA. Adrenal vein sampling demonstrated no sign of lateralization.
Demura et al. [
17] have shown that the aldosterone synthetic enzyme, CYP11B2, is influenced by epigenetic factors. Long-term treatment with spironolactone may influence the epigenetic modulation of CYP11B2. SP (spironolactone) not only is an MR antagonist, but also forms SP body inclusions in cells after long-term treatment. Histological examination of SP body-containing cells demonstrates enhanced but abortive steroidogenic activity [
18]. These observations suggest that MR antagonists, at long exposure times, may actually impair aldosterone production, something that may explain or aggravate the post adrenalectomy hypoaldosterone state.
Park et al. [
5] found that out of 124 patients, 93 were treated with spironolactone at a mean dose of 88 mg for mean duration of 45 days. The proportion of MR antagonist users was not significantly different between subjects with persistent hyperkalemia, transient hyperkalemia and normokalemia. Preoperative treatment with MR antagonists did not influence the incidence of hypoaldosteronism and hyperkalemia.
These clinical reports do not yet confirm the role of pre-operative MR antagonists in the development of post-operative hyperkalemia but at the very least should emphasize the importance of immediate post-operative discontinuation of such medications.