Key words
hypertension - endocrine hypertension - potassium - hypopotassemia
ABPM Ambulatory blood pressure measurement
APA Aldosterone producing adenoma
ARR Aldosterone-to-renin ratio
BAH Bilateral adrenal hyperplasia
BMI Body mass index
CCT Captopril challenge test
ECG Electrocardiogram
HYRENE Hypertension Research Network
IQR Interquartile range
MR Mineralocorticoid receptor
NCCR Kidney.CH The National Centre of Competence in Research Kidney
Control of Homeostasis
PA Primary aldosteronism
REDCap Research Electronic Data Capture
RAAS Renin-angiotensin-aldosterone system
SIT Saline infusion test
USZ University Hospital Zurich
Introduction
Primary aldosteronism (PA) comprises a group of endocrine disorders with inadequate
oversecretion of aldosterone decoupled from the physiological regulatory circuits.
Bilateral adrenal hyperplasia (BAH) and aldosterone producing adenoma (APA)
represent the most frequent causes of PA with 70 and 30%, respectively,
while adrenal carcinoma and hereditary disorders are rare findings
(<1% each) [1]. In the setting
of low levels of renin and angiotensin II, aldosterone mediated activation of renal
mineralocorticoid receptors (MR) results in considerable sodium and secondary fluid
retention. The reabsorption of sodium in the aldosterone-sensitive distal portions
of the nephron further provides a sustained driving force for the luminal secretion
of potassium and hydrogen in equal quantities. Thus, PA typically results in
arterial hypertension and facultative hypokalemia and metabolic alkalosis. With a
prevalence of 5–12% in hypertensive patients, PA constitutes the
most common endocrine cause of secondary hypertension [2]
[3]
[4]
[5]
[6]. Despite this high prevalence, PA remains
underappreciated particularly in primary care, due to imprecise and therefore
difficult to apply screening indications as well as the necessity of complex,
multi-step, error-prone and cost-intensive diagnostics. An improvement and
simplification of disease recognition, however, would be of particular importance
from both, a patient and health economic perspective. Compared to essential
hypertension, PA is associated with worse cardiovascular morbidity and mortality due
to the systemic effects of aldosterone excess [7]. Since effective targeted drug therapies and potentially even curative
surgical treatments exist, an early diagnosis could prevent or at least delay the
development of PA-associated long-term complications in a large number of cases.
Against this background, the “Incidence of Primary Aldosteronism in Patients
with Hypokalemia” (IPAHK+) study was designed with the
long-term perspective to increase the detection rate of PA by improving the evidence
level of current screening recommendations, that are largely based on
cohort-specific prevalences. Specifically, screening for PA is indicated in patients
with moderate (prevalence 8%), severe (13%) and drug-resistant
hypertension (17–23%) [8].
Furthermore, patients with adrenal incidentaloma (2%) or sleep apnea
syndrome (34% among newly diagnosed hypertensive patients) should be
evaluated for the disease [8]. Last but not
least, also patients with spontaneous or diuretic-induced hypokalemia and
simultaneous hypertension are screening candidates [8]. The link between hypokalemia and PA is plausible from a
pathophysiological point of view and proven by numerous studies [8]
[9]
[10]
[11]. However, prevalence data of PA in a
hypokalemic population independent of other factors such as hypertension are still
missing. Hypokalemia is of particular importance in the context of the disease with
respect to pre-analytics, differential diagnosis, course of disease, treatment and
prognosis [8]
[9]. The IPAHK+trial, which
was launched in October 2019, will provide the first high-level evidence data on the
incidence of PA in an unselected hypokalemic outpatient cohort. This information
could ultimately allow more targeted screening recommendations to be made, ideally
based on specific potassium levels. In addition, we expect to gain insights into
whether different potassium levels correlate with specific features of the
disease.
Here, we describe the baseline characteristics of the first 100 patients eligible for
study participation.
Design and Methods
IPAHK+was designed as an epidemiological cross-sectional study
including patients with defined hypokalemia (≤3.0 mmol/l) conducted
at the University Hospital Zurich (USZ), Switzerland. The recruitment of patients is
carried out continuously on a referral-basis by the central laboratory of the
hospital through an automated suitability testing and data delivery system.
Eligibility Criteria
Male and female outpatients of the University Hospital Zurich over 18 years of
age with a serum potassium level≤3.0 mmol/were defined eligible
to participate in the study if they had signed a general informed consent. The
cut-off value for hypokalemia was chosen with the intention to include only
patients with moderate to severe hypokalemia as mild hypokalemia often also
occurs in otherwise healthy individuals.
Key exclusion criteria include inpatients, pregnant or lactating women,
comorbidities resulting in a life expectancy of less than one year and the
inability to follow the procedures of the study.
Objectives
The primary objective of this study is to determine the incidence of PA in
outpatients from the USZ with a random serum potassium
level≤3 mmol/l in the presence or absence of hypertension.
Secondarily, the study aims to assess whether the incidence of PA correlates
with different serum potassium levels. Beyond that, evidence will be gathered
whether the level of hypokalemia correlates with the extent of clinical and
metabolic characteristics of PA. These include the stage of hypertension, the
history of cardiovascular events such as myocardial infarction, strokes and
atrial fibrillation, the occurrence of muscle weakness and evidence of
hyponatremia or hypomagnesemia.
Final diagnosis
The diagnosis of PA in this study follows the internationally recommended routine
procedure. Each step in this multistep process is performed as previously
described [8]. A prerequisite for
meaningful results of all tests is the prior adaption of the
participants’ drug regime according to the clinical standard to prevent
interference with the renin-angiotensin-aldosterone system (RAAS). If needed to
maintain hypertension control, the established therapy is therefore temporarily
switched to Verapamil slow-release 120–240 mg twice daily and/or
Doxazosin 4–8 mg once daily for one or four weeks, depending on
the original medication [8]. In addition,
the participants are given oral potassium supplements to provide normokalemic
conditions for testing, whenever necessary as hypokalemia can lead to false
negative sreening results. During blood collection, care is taken to minimize
factors that may induce hemolysis (e. g., fist clenching, late release
of the tourniquet, aspiration, etc.) as this can mask hypokalemia due to
incorrectly high potassium levels [12].
The calculation of the aldosterone-to-renin ratio (ARR) serves as the gold
standard for the screening for PA. Aldosterone and renin are measured with
chemiluminescence assays (CLIA) from DiaSorin with the Liasion XL analyzer. If
aldosterone is>50 ng/l and the ARR>11.5 ng/mU
(thresholds defined by the manufacturer DiaSorin) a subsequent confirmatory test
will either confirm or rule out the diagnosis. Many different test methods are
established for this latter purpose. For IPAHK+, the saline
infusion test (SIT) is primarily used. An aldosterone level above
60 ng/l after saline infusion in the sedentary patient confirms the
diagnosis of PA, provided that the renin concentration is simultaneously
suppressed and serum-cortisol did not increase [8]. Only if there are contraindications for the implementation of a
SIT or if the results from the SIT are ambiguous a captopril challenge test
(CCT) is used as an alternative. A decline in aldosterone to less than
30% of baseline after 120 minutes is consistent with the
diagnosis of PA [8].
Independently of the study, patients with newly diagnosed PA were offered further
diagnostics for subtype differentiation (adrenal imaging, adrenal vein catheter
sampling) and adequate treatment.
Data collection
Study related data are collected and managed using REDCap electronic data capture
tools hosted at the University Hospital Zurich. REDCap (Research Electronic Data
Capture) is a secure, web-based software platform designed to support data
capture for research studies [13].
Results
Study recruitment started on 1 October 2019 and is ongoing. By 30 December 2019, a
total of 100 patients eligible for study participation were registered through the
electronic reporting system. Up to this point, 28 140 ambulatory potassium
measurements were performed of which 0.79% were found to
be≤3.0 mmol/l ([Fig. 1]). Of
222 hypokalemia reports, a total of 122 were excluded: 64 patients had not signed
the general consent form, in further 34 cases, multiple measurements were registered
in patients who had already been included, and 22 patients suffered from concomitant
diseases that did not allow study participation or resulted in a life expectancy of
less than one year.
Fig. 1 Overview on exclusion and inclusion of patients in the
IPAHK+trial.
The first 100 patients registered had a median age of 57 years with the youngest
being 22 years and the oldest 92 years ([Table
1]). Sex distribution was balanced with 53% men and 47%
women. Baseline potassium measurements were performed in the central laboratory on
patients of 22 different outpatient clinics of the USZ. Most patients were from the
Department of Medical Oncology and Hematology (20%), Gastroenterology
(12%) and Cardiology (11%). Nine percent of patients each came from
the interdisciplinary emergency unit and the Clinic for Nephrology. Eighteen other
departments accounted for the total 39 percent remaining patients. The exact
proportions of the respective medical departments are shown in [Fig. 2].
Fig. 2 Proportion of reported patients from the different medical
departments.
Table 1 Baseline characteristics of the first 100 patients in the
IPAHK+hypokalemia registry.
|
IPAHK+ Registry (n=100)
|
|
|
Age (years) median [range]
|
57 [22; 92]
|
|
Sex distribution male/female (%)
|
53/47
|
|
Hypertension (%)
|
53
|
|
Atrial fibrillation (%)
|
7
|
|
Primary aldosteronism (%)
|
4
|
|
Use of diuretics (%)
|
37
|
|
Use of laxatives (%)
|
12
|
|
Use of MR-agonists (%)
|
12
|
|
Potassium supplementation (%)
|
20
|
|
K+mean (mmol/l) [range]
|
2.9 [2.3, 3.0]
|
|
K+≤3.0 mmol/l (%)
|
100
|
|
K+≤2.9 mmol/l (%)
|
46
|
|
K+≤2.8 mmol/l (%)
|
29
|
|
K+≤2.7 mmol/l (%)
|
13
|
|
K+≤2.6 mmol/l (%)
|
8
|
|
K+≤2.5 mmol/l (%)
|
5
|
|
K+≤2.4 mmol/l (%)
|
3
|
|
K+≤2.3 mmol/l (%)
|
2
|
Serum potassium distribution
The mean serum potassium value was 2.89±0.02 mmol/l (range
2.3–3.0 mmol/l) independent of sex (2.92 vs. 2.87 mmol/l
in males and females, p=0.17). More than half of the included patients
(54%) had a potassium value of exactly 3.0 mmol/l. The number of
reported patients decreases steadily with the extent of hypokalemia ([Fig. 3a, b]). Overall, 92% of all
measurements were in a range between 3.0 and 2.7 mmol/l, while only
5% of patients had severe hypokalemia with values lower than
2.6 mmol/l.
Fig. 3 Percentage of the different potassium levels within the
total population studied. a: indicates the percentage of
hypertension (▀) and atrial fibrillation (▴) within
groups of different potassium levels. b: shows the percentage of
PA (●) and diuretic medication (x) within groups of different
potassium levels.
Twenty percent of patients were on oral potassium supplementation with an average
daily dosage of 52.5 mmol. Twelve percent of patients took MR
antagonists.
Co-morbidities
The percentage of patients with a history of arterial hypertension was
53%. Seven percent of the first 100 study patients had a history of
atrial fibrillation. The proportion of hypertensive patients and patients with
atrial fibrillation related to the different potassium levels is provided in
[Fig. 3a]. Notably, the percentage of
hypertensive patients, but also of atrial fibrillation decreased with decreasing
potassium levels. However, the intake of oral potassium supplements and
potassium-sparing diuretics was higher in the group of hypertensives than in
non-hypertensives (24.5 vs. 14.9% and 15.1 vs. 8.5%,
respectively). The same was seen when comparing the two subgroups of patients
with and without atrial fibrillation (28.6 vs. 19.4% and 14.3 vs.
10.8%, respectively).
Causes of hypokalemia
The current evaluation suggests that hypokalemia-inducing drugs were the most
common cause of hypokalemia in the study population: 55% of all patients
had been prescribed at least one potentially hypokalemia-inducing drug with a
leading usage of diuretics at 37%. Twelve percent of the patients were
on laxatives. These two groups of drugs therefore accounted for the majority of
potentially hypopotassemia-promoting drugs. Glucocorticoids and various
immunomodulary drugs accounted for a significantly smaller share.
After drug-related causes, acute or chronic gastrointestinal losses account for
the second most common cause of hypokalemia in the study population, with a
proportion of 9%. In third place among the most common causes of
hypokalemia in the population studied was PA with a prevalence of 4%.
Within the subgroup of hypertensive patients, the prevalence was as high as
7.5%. [Fig. 3b] shows the
percentage of patients with diuretic use and those diagnoses with PA related to
the different potassium levels. More rare causes were renal losses due to
Gitelman disease (2%) or renal tubular acidosis type 1 (1%) and
a decreased intake of potassium due to eating disorders (2%).
Discussion
Herein, we introduce the IPAHK+hypokalemia cohort and report
baseline characteristics of the first 100 patients eligible for
IPAHK+study enrollment. Numerous previous studies have
provided information on the prevalence of hypokalemia in cases of confirmed
diagnosis of primary hyperaldosteronism [11]
[14]
[15]
[16]
[17]
[18]. Based on these evaluations, a
hypokalemic course of the disease is expected in 9–37% of cases
[11]. In contrast, high-level evidence
about the prevalence of PA in a hypokalemic population is missing. To our knowledge,
the IPHAK+trial is the first prospective study investigating the
incidence of PA in a hypokalemic population. Although hypokalemic PA only accounts
for the smaller proportion of the disease, it is of major importance because of its
impact on pre-analytics, the diagnostic pathway, treatment and prognosis [8]
[9].
Along with hyponatremia, hypokalemia is one of the most common electrolyte disorders.
The prevalence of mild hypokalemia (3.0–3.5 mmol/l) in outpatients
is reported to be 14% [19]. Lower
values have a significantly lower prevalence. Indeed, in the total of 28 140
outpatient potassium measurements performed at the USZ between October and December
2019, the rate of hypokalemia less than or equal to 3.0 mmol/l was
0.79%. As expected, case numbers decreased sharply with further decreasing
potassium levels. The mean potassium level was 2.9 mmol/l irrespective of
sex. With a share of 20%, most of the reports were attributable to the
Oncology and Hematology department, followed by the Department for gastroenterology
(12%), cardiology (11%), nephrology and the interdisciplinary
emergency unit (11% each). Obviously, this ranking is potentially biased by
the total number of potassium measurements in each clinical setting.
Ophthalmologists, for example, are likely to check potassium levels less frequent
than cardiologists, for example, due to lack of indication. In addition, however,
the patient population of the departments are characterized by various
dyselectrolytemia predisposing factors such as polymorbidity, malnutrition and
polypharmacy. The aforementioned properties favor an inadequate supply of potassium,
impaired potassium distribution between intra- and extracellular space and in
particular, increased gastrointestinal or renal excretion of potassium, which
finally results in hypokalemia [20]
[21].
Subgroup comparison revealed a marginally but significantly lower mean potassium
level in non-hypertensive compared with hypertensive patients. In addition, the
percentage of hypertensive patients decreased successively with decreasing potassium
levels. This is initially surprising, since hypokalemia per se promotes the
development of hypertension at different levels [22]
[23]
[24]
[25]
[26]. Based on these findings, one would have
expected lower potassium values in patients suffering from hypertension and a higher
proportion of hypertensive patients with decreasing potassium. However, the
comparatively higher intake of oral potassium supplements and potassium-sparing
diuretics in the group of hypertensive patients may represent a relevant confounder
and plausible explanation. Interestingly, no significant differences in mean
potassium levels were found in patients with and without atrial fibrillation or
PA.
The current evaluation suggests that hypokalemia-inducing drugs were the most common
cause of hypokalemia in the study population, followed by acute or chronic
gastrointestinal disorders and PA. The prevalence of PA in the study population was
at least 4%. In relation to all patients with hypertension, PA even accounts
for at least 7.5%. In a current retrospective observational study from
Italy, the prevalence of PA in outpatients with hypokalemia and hypertension was
reported to be as high as 28.1% [27].
The lower rate in our study is most likely due to the low proportion of screening
examinations. It is likely that very few study candidates ever received an
appropriate screening examination regarding PA. In order to obtain meaningful data
and allow for subgroup analysis we therefore aim to recruit a minimum of 1000
candidates undergoing a standardized screening assessment. To achieve this goal we
began to enroll patients continuously in the IPAHK+study protocol
([Fig. 4]). Patients are invited to a
baseline visit in which all relevant basic characteristics are assessed. This
includes a targeted medical history and physical examination as well as a 24-hour
ambulatory blood pressure measurement (ABPM) and an electrocardiogram (ECG). In
addition, we perform urine examinations in spot and 24-hour collection urine and
blood tests. The diagnosis of PA follows internationally recommended routine
procedure including adaptation of antihypertensive medication (slow-release
verapamil 120–240 mg twice daily and/or Doxazosin 4 to 8 mg
once daily for one or four weeks, depending on the initial medication) for the
determination of the aldosterone-to-renin ratio (ARR) [8]. In addition, participants are provided with
oral potassium supplements to achieve normokalemic conditions for testing. Saline
infusion test (SIT) is used as a subsequent confirmatory test [8]. Independently of the study, patients with
newly diagnosed PA are offered further diagnostics for subtype differentiation
(adrenal imaging, adrenal vein catheter sampling) and adequate treatment.
Irrespective of whether PA was found or not, a final follow-up one year after study
enrollment is scheduled that includes re-assessment of the baseline characteristics
and adverse events.
Fig. 4 IPAHK+study design and patient flow.
The study in its current stage has several limitations. The main weakness of the
study is the patient selection. Patients with hypertension or chronic or unclear
hypokalemia are likely to be motivated to participate in the study. Conversely,
patients who do not have hypertension or those with known plausible cause of their
hypokalemia, or only one-time low potassium are more likely to decline participation
in the study because the personal benefit is considered to be low. The risk of
selection bias may thus arises from the predominant inclusion of patients with
already elevated pretest probability for PA into the trial. This in turn may
ultimately lead to an overestimation of the disease in a hypokalemic population.
Another limitation is the monocentric design of the study whereby the involvement of
further international study centers is planned in order to enhance the power of the
study by greater diversity and cohort size.
In conclusion, we have reported the baseline characteristics of the first 100
suitable candidates for inclusion in the IPAHK+trial. The
prevalence of PA was 4% in the total study population and 7.5% in
the subgroup of hypertensives. However, guideline-compliant screening of all
patients with spontaneous or diuretic-induced hypokalemia≤3.0 mmol/l
would likely result in a significantly higher rate of PA. In the coming years, the
IPAHK+study will therefore fill a long-standing knowledge gap
by providing evidence on the incidence of PA in a hypokalemic population. It will
also investigate whether the degree of hypokalemia correlates with various features
of the disease such as symptomatology, disease severity, and sub-group
differentiation.
