Insufficiency of the zona glomerulosa of the adrenal cortex and progressive kidney insufficiency following unilateral adrenalectomy – case report and discussion

Abstract

Background: Primary aldosteronism (PA) is the most common cause of secondary hypertension and bilateral adrenal hyperplasia (BAH) and aldosterone-producing adenoma (APA) seem to be the most common causes of PA. Unilateral adrenalectomy (UA) is the preferred treatment for APA, although the benefits are still difficult to assess.

Case Report: We present a case report of a 69-year old man with a 30 year history of hypertension and probably long-standing PA due to APA, with typical organ complications. Since repeated abdominal CT scans were equivocal, not showing radiological changes characteristic for PA, the diagnosis of APA was delayed and was only finally confirmed by adrenal venous sampling which demonstrated unilateral aldosteronism. The patient underwent UA, complicated by mineralocorticoid deficiency syndrome and increased creatinine and potassium levels. At 12 months follow-up the patient still had hyperkalemia and was fludrocortisone dependent.

Conclusions: Older patients and patients with long-lasting PA who are treated with UA may demonstrate deterioration of renal function and develop transient or persistent insufficiency of the zona glomerulosa of the remaining adrenal gland necessitating fludrocortisone supplementation. Transient hyperkalemia may be observed following UA as a result of the prolonged effects of aldosterone antagonists and/or transient mineralocorticoid/glucocorticoid insufficiency. Additionally, the level of progression of chronic kidney disease after UA is difficult to predict. There likely exists a group of patients who might paradoxically have higher cardiovascular risk due to significant deterioration in kidney function not only resulting from the removal of the aldosterone induced glomerular hyperfiltration phenomenon. Identification of such a group requires further detailed investigation.

Keywords:

• Primary aldosteronism
• adrenal cortex insufficiency
• hyperkalemia
• chronic kidney failure

Introduction

Primary aldosteronism (PA) is one of the most common causes of secondary hypertension and represents up to 20% [1] of resistant hypertension (it is estimated for 15.7% in Poland [2]). Since the first case was described by Professor Jerome W. Conn in 1955, diagnostic methods in this disease have evolved and the negative impact of high aldosterone levels on the circulatory system and its complications have been well described. Patients with PA are characterised by resistant hypertension with rapidly developing organ damage. The main causes of PA are bilateral adrenal hyperplasia (BAH) and aldosterone-producing adenoma (APA). Aldosterone antagonists (AA) form the basis of treatment for BAH while unilateral adrenalectomy (UA) is preferred for APA. Surgery has been shown to be effective in preventing adverse consequences in PA, however the degree of benefit is still difficult to assess and prognostic factors continue to be investigated.

Case report

A 69 year-old man with a 30 year history of hypertension and type 2 diabetes, permanent atrial fibrillation, chronic obstructive pulmonary disease, chronic kidney failure G3 with proteinuria and a 10 year history of low blood potassium (lowest measured potassium level 2.5 mmol/l; 3.5-5.1 mmol/l) was referred to the Hypertension Outpatients Clinic, Medical University of Gdańsk by his family doctor. Although the patient was administered with five antihypertensive drugs including an aldosterone receptor-antagonist (that the patient has been receiving for 7 years), his mean 24-hours blood pressure was 146/87 mm Hg (SpaceLabs 90207) [3,4].

Transthoracic echocardiography revealed an enlarged left atrium (29 cm2), the remaining heart chambers were within normal ranges. The left ventricle did not show any signs of hypertrophy or contractile dysfunction; ejection fraction was 55%. The patient underwent thorough investigations for causes of secondary hypertension. A 24-hour urine collection found increased potassium excretion (149 mmol/24h; 25-125mmol/24h) with normal excretion of sodium, magnesium, calcium and phosphate ions. Other laboratory investigations included the following: diurnal cortisol rhythm, low-dose cortisol suppression test with 1 mg of dexamethasone (<28nmol/l), serum DHEA-S, androstenedione and ACTH levels, and a 24-hour urine collection for metoxycatecholamines; all were within normal ranges.

After adequate preparation time, including modification of antihypertensive medication and correction of serum potassium, the patient underwent an intravenous saline suppression test (SST) with 2000 ml of 0.9% NaCl. Prior to the infusion, aldosterone was 42 ng/dl and renin was unmeasurable, plasma renin activity (PRA) was 0.01 ng/ml/h, at the end of the test aldosterone was 36.5 ng/dl (<10 ng/dl), indicating inadequate aldosterone suppression. An abdominal CT scan revealed a 15x16 mm thickening of the right adrenal gland possibly suggesting a low lipid content adenoma, however in this and in previous CT scans the radiological view was not typical for PA. Adrenal venous sampling was performed. An hour before the procedure, a continuous intravenous infusion with a synthetic corticotropin was administered. Following this, the suprarenal veins were catheterised via the right femoral vein. The localisation of the catheter was regularly validated by sampling cortisol from the veins. For further analysis, samples with the highest cortisol-based selective index were chosen from the right and left suprarenal vein (13.9 and 7.2, respectively). Aldosterone and cortisol levels from the venous samples were then used to calculate the lateralization index (26.8) and suppression index (0.17).

Taking into account the above findings, the patient underwent a laparoscopic right sided adrenalectomy without complications. Pathological examination of the excised adrenal gland revealed a benign cortical adenoma. The patient was discharged from hospital on a standard dose of hydrocortisone supplementation (20 mg +10 mg). During the following two weeks, the patient complained of progressive weakness and reduced exercise tolerance and was readmitted to hospital. Upon readmission, the patient had hyperkalemia (8.3 mmol/l), hyponatremia (128mmol/l) and metabolic acidosis. Serum glucose level was 188 mg/dl. Additionally, a significant reduction in kidney function was observed. Creatinine and eGFR values (calculated by the MDRD formula) before adrenalectomy were 1.32 mg/dl (0.73-1.17mg/dl) and 54 ml/min respectively, while post adrenalectomy values were 2.41 mg/dl and 27 ml/min respectively. The patient received hydrocortisone, calcium chloride, insulin with glucose, furosemide, 0.9% NaCl with 10% NaCl, calcium resonium and bicarbonates. As a result of the above treatment, the patient’s clinical state improved, his potassium and sodium levels normalized, and an improvement in kidney function, obviating the need for dialysis, was observed (at discharge creatinine was 1.59 mg/dl and eGFR was 44 ml/min).

Taking under consideration no previous abnormality in the ACTH-cortisol axis and the fact that the patient was receiving supplemental hydrocortisone after surgery, the most probable diagnosis was acute mineralocorticoid insufficiency syndrome.

Hydrocortisone supplementation was reduced and subsequently discontinued, however the patient required continued hormonal supplementation of 0.1 mg fludrocortisone daily. The patient’s antihypertensive medication was also modified due to the development of hyperkalemia; his angiotensin II receptor blocker (ARB) was discontinued and dose of carvedilol reduced. Due to lower BP values after UA also alpha blocker was discontinued. The patient also received lercanidipine, furosemide, atorvastatin, rivaroxaban, conventional insulin therapy and bronchodilators. ABPM on reduced hypotensive medication confirmed good BP control. Insufficiency of the ACTH-cortisol axis was excluded by means of a Synacthen stimulation test (cortisol results were 328 mmol/l, 460 mmol/l, 524 mmol/l and 629 mmol/l). On supplemental fludrocortisone the patient’s aldosterone was 5.3 ng/dl; (2.52-39.2ng/dl). Since chronic mineralocorticoid supplementation can cause aldosterone suppression, a resting renin level was measured (71.5 ulU/ml; 4.4-46.1ulU/ml). This result excluded suppression of the RAA axis by fludrocortisone and suggested instead insufficiency of the zona glomerulosa of the remaining adrenal gland.

Despite the above management, the patient continued to have a tendency towards high serum potassium levels after discharge (highest value 6.7 mmol/l). During the following months, fludrocortisone was reduced to 0.1 mg administered every other day. The patient underwent a follow-up visit at 6 months revealing a gradual worsening of kidney function. Creatinine had increased to 2.02 mg/dl with a reduction in eGFR to 33 ml/min. The urine albumin to creatinine ratio (UACR) was lower compared to preoperative results (92.5 mg/g vs. 126.5 mg/g). Mean BP (ABPM) was 142/83 mm Hg – hypertension treatment was intensified.

12 months after his UA, the patient still had elevated (5-5.5mmol/l) serum potassium levels and fludrocortisone dependence. All attempts to discontinue fludrocortisone therapy resulted in hyperkaliemia and hyponatremia. Above parameters were compared in Table 1.

Table 1. Comparisons of blood pressure, antihypertensive treatment and biochemical variables prior and after the adrenalectomy.

	Before adrenalectomy	After adrenalectomy	6 month follow-up
24-hour ambulatory blood pressure (mm Hg)	146/87	97/54	142/83
Number of antihypertensive drugs	5	3	4
Serum creatinine (mg/dl)	1.32	2.41	2.02
GFR (ml/min/1.73m^2)	54	27	33
Potassium (mmol/l)	4.7	8.3	4.9
Sodium (mmol/l)	143	128	135
UACR (mg/g)	92.5	90.5	126.5

Discussion and problem review

Increased serum creatinine and potassium levels after UA are most commonly caused by insufficiency of the zona fasciculata of the contralateral adrenal cortex. This can be related to the presence of subclinical Cushing’s syndrome prior to surgery. In our patient, normal suppression of morning cortisol following 1 mg dexamethasone and a normal diurnal cortisol rhythm excluded Cushing’s syndrome. Further, hydrocortisone supplementation was implemented immediately following UA, and despite this, increased serum potassium and worsening of kidney function was observed. ACTH-cortisol axis dysfunction was again excluded 6 months after surgery by performing a synthetic corticotropin stimulation test (at that time the patient was no longer taking hydrocortisone supplementation).

A second cause of deteriorating kidney function and hyperkalemia following UA is zona glomerulosa insufficiency with corresponding mineralocorticoid deficiency. However, it is worth mentioning that normalisation of the RAAS after UA in patients with an APA is widely regarded as fast and sufficient. The majority of patients who have had UA do not require hormonal fludrocortisone supplementation and the time for RAAS recovery is about 1 month [5].

In our patient zona fasciculata function was checked before and after UA and additional examination did not reveal any abnormalities. On supplemental fludrocortisone the patient’s aldosterone was 5.3 ng/dl which is in normal range (2.52-39.2ng/dl), with strikingly high renin level 71.5 ulU/ml (despite the fludrocoritosone administration). Specimen for testing was sampled after 1 hour of upright position. Physical activity should stimulate renin and as a consequence aldosterone secretion. In this case we observed abnormal high renin level with poor aldosterone response. Obtained laboratory results suggest subclinical insufficiency of zona glomerulosa. Taking into consideration laboratory results and evident clinical symptoms such as hyperkaliemia, hyponatremia, metabolic acidosis together, proper function of zona fasciculata and no other hormonal causes of hyperkaliemia, insufficiency of the zona glomerulosa of the remaining adrenal gland with residual secretion is suspected. Moreover, in some studies aldosterone level lower than 3.5ng/dl was considered as undetectable [6]. Directly after UA hyperkaliemia coexisted with deterioration in renal function, but further observation did not reveal any relation between creatinine levels (or eGFR) and potassium levels. Moreover in the 12 months follow up any attempt to stop fludrocortisone administration ended up with hyperkaliemia and hyponatremia. It seems unlikely that postoperative hyperkalemia is due only (and directly) to a reduction in renal function.

Progression of chronic kidney disease

Worsening kidney function in patients treated with UA due to PA is widely reported. The scale of the problem, degree of worsening, predicting factors and mechanisms have not been adequately explained.

Iwakura et al. found a significant reduction in eGFR in patients with PA treated with both UA and an AA (a fall in eGFR from 81 to 68 ml/min/m2 and from eGFR 79 to 73 ml/min/m2 respectively), and also an increase in the occurrence of chronic kidney disease (CKD) (15.7% to 37.1% and from 8.1% to 28.3%). However, after one month there was no further significant decline in renal function. Of note, decline in eGFR was significantly higher in the UA treated patients. A possible explanation for this is that at one month the full therapeutic effect of the AA had yet to materialise. Backward stepwise regression analysis found that albuminuria, lower serum potassium and PRA before treatment had a significant impact on the degree to which kidney function may decline following treatment [7].

In another retrospective, observational study of patients treated for PA, a deterioration in kidney function was found 6 months after treatment. In this study, the degree of eGFR decline did not significantly differ between those treated with UA and those who received an AA (eGFR 15.4 ± 16.3 vs. 11.4 ± 11.2 ml/min/1.73m2 p = 0.32 and 72.4 ± 18.7 vs. 78.6 ± 17.9 ml/min/m2). However, they did find a significant correlation between post-treatment eGFR and degree of eGFR decline. Lower baseline eGFR, renin and potassium levels and higher aldosterone levels before treatment were risk factors for reduced renal function following treatment [8].

Sechi et al. in their prospective study assessed the influence of medical therapy on renal function in patients with essential hypertension (EH), PA treated with an AA, and PA treated with UA. At baseline, the PA group was characterized by a significantly higher eGFR and daily urinary albumin excretion than patients with EH, however the UA:UCr (log- transformed urine to albumin creatinine ratio) did not show any difference between groups. Increased urinary albumin excretion in the PA group was probably associated with hyperfiltration (i.e. patients with a higher serum aldosterone level had a higher eGFR). At six months, reduction in eGFR and UA:UCr was more marked in the PA group compared to the EH group. In subsequent blood tests, the degree of reduction in eGFR and UA:UCr did not vary significantly between groups. At the end of the follow-up period (mean time of observation 6.4 years), a significantly greater reduction in eGFR and UA:UCr was found in the PA group. Additionally, at six months, patients in the PA group whose baseline eGFR was above the median (105 ml/min/1.73 m2) showed a significantly greater reduction in eGFR and UA:UCr than in those whose baseline kidney parameters were lower. This relationship was not found in subsequent months. Analysis of the two PA groups found no significant renal function differences between patients who underwent UA and those who were treated with an AA. It is also worth noting that patients with PA more frequently returned to normal urinary albumin levels than patients with EH and this was independent of BP control. Therefore, increased urinary albumin excretion is suggestive of functional kidney dysfunction rather than of structural renal damage. Lack of complete remission of albuminuria after treatment in the majority of patients with PA suggests that alongside functional kidney dysfunction there is some irreversible damage to glomeruli, possibly resulting from concomitant elevated BP [9].

In another study comparing EH and PA, with the exception of increased urinary β-2-microglobulin and albumin excretion among PA patients, the two groups did not significantly differ with regard to kidney function and BP. In patients with PA treated with UA reduction in eGFR was inversely correlated with baseline eGFR and glomerular hydrostatic pressure and directly correlated with a reduction in effective renal plasma flow (eRPF), but not with BP values. Multiple factor analysis found that baseline eGFR, potassium, and the degree of postoperative reduction in eRPF all contributed to changes in eGFR. Through detailed analysis of renal blood flow, they found that patients with PA had, at baseline, lower resistance in their afferent glomerular vessels and higher glomerular hydrostatic pressures compared to patients with EH. After treatment, a significant reduction in afferent vessel resistance was found in both groups, as was a decrease in glomerular hydrostatic pressure following UA [10].

In summary, the probable factors contributing to declining renal function among patients treated for PA are: Microalbuminuria, increased aldosterone levels, lower renin and potassium levels prior to treatment and lower eGFR. Sechi et al. found the opposite with regard to eGFR. However, Sechi et al. applied other exclusion criteria to patients with CKD (eGFR <30 ml/min/1.73 m2). This differs significantly from criteria used in the other cited papers (eGFR <60ml/min/1.73 m2, eGFR <90ml/min/1.73 m2). Only Iwakura et al. included 10% of patients with an eGFR <60ml/min/1.73 m2, and they did not find a relationship between baseline eGFR and a reduction in eGFR following treatment. The exact prognostic value of baseline eGFR in the context of postoperative decline in renal function requires further research, including analysis of such in all stages of CKD.

Post-treatment eGFR decline may be connected with the disappearance of the so-called “escape” phenomenon which is present in hyperaldosteronism and consists of natriuresis stimulated via increased ANP secretion. Suppression of this mechanism may lead to a decrease in eGFR and eRPF of up to 20% [11]. In vitro studies have shown that aldosterone causes vasoconstriction of both glomerular arterioles, however it causes a greater degree of efferent arteriole vasoconstriction [12] which favours glomerular hyperfiltration. Studies on rats with deoxycorticosterone-induced hypertension found that glomerular pressure and flow were increased, and discontinuance of deoxycorticosterone resulted in improved renal function and reduced albuminuria [13]. Functional changes in glomeruli that occur following treatment of PA can reveal structural changes which are previously masked by hyperfiltration. Molecular examination of renal biopsy material from patients with proteinuria and kidney failure of various origin have shown thathighserum aldosterone negatively correlates with kidney function, and positively correlates with the degree of kidney fibrosis [14]. Some studies have also found that hypokalemia itself may cause structural kidney damage due to tubular-vascularization, cyst formation and interstitial inflammation [15,16].

The patient we have described, despite PA, hypokalemia and arterial hypertension also had a 30-year history of type 2 diabetes. Thus, in this case, reduced GFR and albuminuria may also represent diabetic nephropathy, though this does not explain the rapid postoperative progression of CKD which we observed. Data from referenced publications were summarized in Table 2.

Table 2. Assessment of factors affecting renal function after primary aldosteronism (PA) treatment.

Authors	Study protocol	No. of patients	Age (years)	Baseline eGFR (ml/min/1.73m^2)	Aim	Determinants for reduced renal function following PA treatment	Comments
Iwakura et al. [7]	12-month prospective study	213	54 ± 0.8	80 ± 1	Assessment of chronic kidney disease prevalence in PA following treatment Identification of the determinants of eGFR decline after PA-treatment	Microalbuminuria before treatment Lower potassium serum levels before intervention Lower plasma renin activity before treatment.	Consistent eGFR in the cohort at baseline Patients with diabetes not excluded.
Kramers et al. [8]	Retrospective study	134	52 ± 11	89.5 ± 17	Renal function determinants following PA treatment	Lower eGFR Lower renin levels before treatment Lower potassium levels before treatment Higher aldosterone levels before treatment.	Inconsistent methods of renin measurements.
Sechi et al. [9]	9-year prospective study (mean FU 6.4 years)	50	25-70	104 ± 32	Assessment of long-term renal function after PA (GFR and albuminuria).	Baseline eGFR above the median (105 ml/min/1.73 m2) associated with greater reduction in GFR and UA:UCr.	Relatively long-term follow-up (mean of 6.4 years) Wide range of GFR at baseline, incl. <30ml/min.
Ribstein et al. [10]	6-month prospective study	25	33-66	102 ± 3	PA surgical treatment on renal hemodynamics and function.	Lower baseline eGFR before treatment Lower serum potassium concentration before intervention The degree of postoperative reduction in ERPF	Comparable eGFR changes for PA and EH treatment Additionally to eGFR, the ERPF was assessed Surgical treatment only

eGFR: estimated glomerular filtration rate (Cocroft-Gault formula; except for Iwakura et al.: equation established for Japanese population by Japanese Society of Nephrology and Kramer et al.: Chronic Kidney Disease Epidemiology Collaboration).; EH: essential hypertension; ERPF: effective renal plasma flow; PA: primary aldosteronism.

Hyperkalemia

The first reports of postoperative hyperkalemia in patients with PA who underwent UA arose in the 1960's. Over the years few reports have been published, and in the majority of these hyperkalemia was temporary and only rarely required hormonal substitution.

In 2012, the results of a retrospective, observational study of 110 patients from Germany with PA due to APA treated with UA were published. 18 patients (14%) had hyperkalemia (serum potassium level >5 mmol/l) postoperatively. During 3 months of observation and on only one occasion did 12 of these patients have hyperkalemia, with subsequent spontaneous normalisation. In the remaining 6 patients (5%) hyperkalemia persisted and required fludrocortisone therapy. All those who required postoperative hormonal supplementation had undetectable serum aldosterone levels (<35 ng/l). In contrast, patients with transient hyperkalemia had either undetectable or low (<50 ng/l) levels of aldosterone. Detailed comparative analysis found that patients who required postoperative mineralocorticoid supplementation were significantly older than patients with postoperative normokalemia (65 ± 9 vs. 50 ± 12 years, p = 0.024), and their kidney parameters were worse before treatment (creatinine 1.2mg/dl vs. 0.8 mg/dl, p = 0.001, eGFR 56 ± 17 ml/min vs. 84 ± 21 ml/min). Worsened renal function was found in both groups, significantly poorer stillin patients with hyperkalemia (these patients also had higher levels of postoperative microalbuminuria). Multiple factor analysis found the following predictors of postoperative hyperkalemia: Lower eGFR, higher creatinine levels before UA, degree of increase in creatinine levels following treatment, creatinine concentration and microalbuminuria after UA. Treatment with an AA prior to UA did not protect from postoperative hyperkalemia. It is also worth mentioning that during the follow-up period (11-46 months), patients taking fludrocortisone underwent unsuccessful attempts at stopping mineralocorticoid supplementation, resulting in recurrence of hyperkalemia [6].

In contrast, in a retrospective analysis of 55 patients with PA who underwent UA due to APA, the frequency of postoperative hyperkalemia (potassium >5 mmol/l) was 29%, of which only 3 patients (5.5%) had chronic electrolyte imbalance requiring fludrocortisone supplementation, while among the remaining patients increased potassium lasted from 7 days to 2 months. Patients who had postoperative hyperkalemia (both periodic and chronic) tended to be older, male, had a longer hypertension history, and had lower preoperative potassium and eGFR levels in comparison to patients with normokalemia after UA. Multivariate regression analysis found duration of hypertension (OR 1.12; 95%PU 1.01-1.24; p < 0.05) and eGFR (OR 0.97; 95% PU 0.94-1.0; p < 0.05) to be independent risk factors for postoperative hyperkalemia [17].

In 2015 the European Journal of Endocrinology published results of a retrospective analysis of 124 patients who underwent UA due to PA between the years 2000-2012 in South Korea. The incidence of postoperative hyperkalemia (potassium level >5.5mmol/l) was 10.5%, while 7.3% (9 patients) had persistently elevated potassium levels (shortest follow-up period was 3 months). Transient hyperkalemia was diagnosed if it appeared within 3 months of treatment regardless of the time of electrolyte imbalance, and disappeared spontaneously. Patients with persistent hyperkalemia were significantly older than patients who had transient hyperkalemia and postoperative normokalemia. Furthermore, they had a longer hypertension history, and significantly lower eGFR and higher creatinine levels both pre and postoperatively. Finally, adenoma size (estimated postoperatively) was significantly larger in the group with persistent hyperkalemia. ROC analysis found the following values correlate with postoperative chronic hyperkalemia: Age ≥53 years, hypertension history ≥9.5 years, size of adrenal lesion ≥1.95cm, preoperative eGFR <58.2ml/min; albuminuria was not recorded. Similarly to the German study they did not find a protective effect of preoperative treatment with an AA [18].

The articles cited above differ in methodology; some patients with transient and persistent postoperative hyperkalemia were analyzed together while others were analyzed separately. Further, different cut-off points for high serum potassium levels were used. In all the above publications, a lower preoperative eGFR was a risk factor for postoperative hyperkalemia. Nonetheless postoperative reduction in renal function was associated with increased potassium levels in only one paper.

So far, only two mechanisms have been proposed to explain inadequate aldosterone production following UA: Lack of proper stimulation of cells in the zona glomerulosa under normal renin levels and more rarely, secondary hypoaldosteronism connected with low postoperative renin levels secondary to renal damage [6,19]. It seems unlikely that postoperative hyperkalemia is due only (and directly) to a reduction in renal function.

Undoubtedly, an important factor in the development of postoperative hyperkalemia is inadequate aldosterone production by the remaining healthy adrenal gland. In Cushing's syndrome, for example, excessive production of the glucocorticoid hormone cortisol leads to atrophy of previously healthy adrenal glands. Paradoxically, in patients with APAs pathological examination reveals hyperplasia of the zona glomerulosa cells surrounding the tumor [20]. This could be connected with zona glomerulosa cell stimulation by ACTH and high serum potassium levels in the low renin activity environment which is secondary to high aldosterone levels [21].

It was also postulated that the lack of proper cell response to stimulation following surgery may be attributed to the antagonistic action of spironolactone on mineralocorticoid receptors preceding surgery. This effect may persist for 4-6 weeks even after drug discontinuation. The above mechanism does not, however, explain long-lasting postoperative hyperkalemia. Detailed descriptions of cellular mechanisms responsible for the absence of normal functioning of the cells comprising the zona glomerulosa of the remaining adrenal gland requires further study. Data from referenced publications were summarized in Table 3.

Table 3. Assessment of factors predicting hyperkalemia after unilateral adrenalectomy (UA).

Authors	No. of patients with PA	No. of patients with postoperative hyperkaliemia (transient/ persistent)	Baseline age of patient with hyperkalemia (years.)	Baseline eGFR (ml/min/1.73m^2)	hyperkalemia definition (mmol/l)	additional criteria	Primary outcome	Predictors of postoperative hyperkalemia	Comments
Fischer et al. [6]	110	18 (12/6)	Transient 54 ± 3 Persistent 65 ± 9	Transient 79 ± 23 Persistent 56 ± 17	>5.0	Transient hyperkaliemia defined as one time detected, restored spontaneously	Hyperkalemia following UA in large series of APA patients	Low eGFR before treatment High creatinine level before treatment Degree of increase in creatinine levels following treatment Creatinine concentration after UA Microalbuminuria after UA	Predictors of postoperative hyperkalemia assessed in multivariate analysis Limited number of patients with persistent hyperkalemia
Chiang et al. [17]	55	16 (13/3)	56.4 ± 11.8	73 ± 30	>5.0	Prolonged hyperkalemia: timeframe from 7 days to 2 months	Incidence and risk factors for hyperkalemia after UA	Duration of hypertension Low eGFR after UA	High percentage of hyperkalemia after UA Transient and persistent hyperkalemia pooled analysis
Park et al. [18]	124	13 (4/9)	Transient 48.5 ± 12.4 Persistent 57.2 ± 6.6	Transient 96.0 ± 37.1 Persistent 53.1 ± 23.0	>5.5	Transient hyperkalemia detected within 3 months following treatment/resolved spontaneously	Predictors of postoperative hyperkalemia in APA patients undergoing UA	Age ≥53 years, Hypertension history ≥9.5 years, Size of adrenal lesion ≥1.95cm, Preoperative eGFR <58.2ml/min	Risk factors for persistent hyperkalemia analyzed separately

eGFR: estimated glomerular filtration rate (abbreviated Modification of Diet in Renal Disease Study Group equation, except for Fischer et al. and Park et al.- formula specified); APA: aldosterone producing adenoma; PA: primary aldosteronism; UA: unilateral adrenalectomy.

Summary

We presented a very rare complication of persistent insufficiency of the zona glomerulosa of the adrenal cortex after unilateral adrenalectomy in a patient who probably had long-lasting PA due to an APA with signs of typical organ complications (left ventricular hypertrophy, atrial fibrillation, nephropathy). The diagnosis of APA was delayed since detailed ambulatory laboratory investigations were not possible in the GP’s office, and there were no characteristic changes in the adrenal glands on repeat abdominal CT scans. The patient underwent UA complicated with mineralocorticoid deficiency syndrome. It is important to remember that older patients, with impaired kidney function and long-lasting PA treated with UA might develop transient or persistent insufficiency of the zona glomerulosa of the remaining adrenal gland necessitating fludrocortisone supplementation. This is most likely the result of atrophy of the “healthy” adrenal gland cortex resulting from a hyperfunctioning APA in the contralateral adrenal gland. Transient hyperkalemia is frequently observed after UA resulting from the prolonged effects of AAs and/or transient mineralocorticoid insufficiency. In patients following UA, serum potassium levels should be monitored since higher levels may also represent glucocorticoid insufficiency associated with subclinical Cushing’s syndrome or, as in this case, total and irreversible insufficiency of the zona glomerulosa of the adrenal cortex. Additionally, there likely exists a group of patients who, due to significant progression of CKD following UA resulting not only from the removal of the aldosterone induced glomerular hyperfiltration phenomenon, might paradoxically have increased cardiovascular risk. Identification of such a group requires further detailed investigation.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

Statutory work ST-02-0017/07/125, Medical University of Gdańsk.