Keywords
Cushing’s syndrome - older patients - aging - hypercortisolism
Introduction
Cushingʼs syndrome (CS) due to an adrenocorticotropic hormone (ACTH)-producing
pituitary adenoma [Cushing’s disease, (CD)] is most commonly diagnosed in women
(female to male ratio 2.3–4:1) in their fifth decade of life. Likewise, adrenal CS
occurs at a similar age or in a slightly older population with an even stronger
female preponderance (5.5:1) [1]
[2].
Although several features are typically associated with hypercortisolemia, including
central obesity, moon face, plethora, easy bruising, purple striae, muscle weakness,
and signs of hyperandrogenism, there is no single lead symptom, and the clinical
presentation can be extremely heterogeneous. In particular, CS may be misdiagnosed
with metabolic syndrome, especially in the first phases of the disease, leading to a
mean delay in diagnosis of about three years, also influenced by the rarity of the
condition (estimated incidence 0.2–5/1000 000 per year) [3]
[4].
Due to the increasing life expectancy and aging of the population in Western
countries, the percentage of older people presenting with CS may be higher than
previously reported, and this may represent a further diagnostic and therapeutic
challenge for modern endocrinology. Indeed, while the clinical presentation of CS
may vary according to age, some hypercortisolemia-related morbidities, such as
hypertension, osteoporosis, and cognitive dysfunction, may be falsely attributed to
aging, thus masking an underlying CS. The clinical risk scores that have been
recently developed to assist clinicians in discriminating CS in the context of
at-risk populations have not been validated in the elderly so far; however, they
might lack sensitivity even in the target group. Preselecting patients with
increased CS risk—which is crucial due to the rarity of the disease—remains
challenging. Studies suggest that osteoporosis at a young age, muscular atrophy,
skin changes, and dorsocervical fat pads are the most indicative symptoms suggesting
the need for CS screening, while an extended screening approach in at-risk
populations (patients with metabolic syndrome, diabetes or obesity) is not indicated
[4]. In older patients, the most
characteristic clinical symptoms of hypercortisolemia, such as skin changes or
central obesity, are often minimally present or lacking. Thus, the diagnosis of CS
in the elderly may be even more difficult than that in a younger population.
Moreover, standard therapy regimens may not be adequate in older patients, and the
response to the treatment may be suboptimal.
There is no clear age limit to define elderly CS patients; most authors use, as a
cutoff, an age of or above 65 years, as suggested by the World Health Organization
[5].
Age-related differences in the etiology of Cushing’s syndrome
Age-related differences in the etiology of Cushing’s syndrome
Recently, age-related differences in the etiology of CS have been shown. In a group
of 142 females with CS, Cushing’s disease (CD) was more often observed in those
younger than 45 years (70.6%), whereas a cortisol-producing adrenal adenoma was more
frequent in those older than 65 years (68.4%) [6]. This is in line with another study on adrenal tumors, showing their
highest prevalence among patients over 65 years (1900 per 100 000 inhabitants) with
a slight female preponderance; about 0.4% of them had overt CS [7]. Also, patients with mild autonomous
cortisol secretion are older than those with non-functioning adrenal incidentalomas
(mean age 65.5 vs. 60.1) [8].
Similarly, in the European Registry on Cushing’s Syndrome (ERCUSYN) cohort containing
1519 subjects, patients with adrenal-dependent CS were significantly older than
those with pituitary-dependent CS [1].
When analyzing only pituitary-dependent CS, compared to the younger group, older
patients more commonly had a pituitary macroadenoma [9]. On the contrary, in a study by Qiao et
al., including 45 patients with CD (age more than 60 years) and 90 patients (age
less than 60 years), no age-related differences in adenoma size were reported [10]. Aging also appears to attenuate the
typical female preponderance in the prevalence of pituitary CS, wherein slightly
more than a half the number of elderly patients were females in the ERCUSYN
population, and women represented almost three-quarters of the younger patients
[9]. Such reduction of the “gender
gap” in the prevalence of pituitary-dependent CS in the elderly may be due to the
lack of estrogen in postmenopausal women and resembles that reported in prepubertal
individuals [11]. Indeed, estrogens
enhance corticotrope proliferation in vitro, and the expression of the
estrogen receptor beta has been detected in the majority of the ACTH-secreting
adenomas, suggesting that estrogens may promote the development of CD in
premenopausal women [12].
Patients with ectopic ACTH secretion (EAS) are a very heterogeneous population,
typically diagnosed in adulthood [13].
Among 918 patients with thoracic and gastroenteropancreatic neuroendocrine tumors
(NET), patients with EAS were younger than those without hypercortisolemia (median
age 48.1 vs 58.7 years) [14]. However,
other studies suggested that patients with ectopic CS are older than those with CD
[15]
[16]. In the cohort gathered at the
National Institutes of Health over 20 years, the age of the patients with EAS ranged
from 8 to 72 years (mean 37.6), showing that EAS may present throughout the lifespan
with a variety of clinical symptoms [17].
In older women, a diagnosis of CS is often established within the evaluation of an
adrenal incidentaloma (36.8%) or workup of hypertension and/or diabetes (31.6%). On
the contrary, younger females more often looked for medical help due to weight gain,
virilization, and oligomenorrhoea [6].
This is also in line with the observation that menstrual irregularities and reduced
libido were more often seen in women with CD as compared with adrenal-dependent CS,
due to a concomitant ACTH-dependent rise in adrenal androgen production [1].
Age-related differences in the clinical presentation of Cushing’s
syndrome
Age-related differences in the clinical presentation of Cushing’s
syndrome
A clear age-dependent difference in the clinical presentation of CS has been recently
demonstrated in the ERCUSYN cohort. Patients older than 65 lacked the typical
features of hypercortisolism, including skin thinning, easy bruising, striae rubrae,
hirsutism, and central obesity, while showing lower BMI and waist circumference as
compared with the younger [9]. In another
study including 45 CD patients, those older than 60 less frequently had obesity as
compared with the younger and their BMI was also lower [10]. On the other hand, older age
correlated with the percentage of visceral fat and visceral to subcutaneous fat
ratio in patients with active CS [18].
Several mechanisms in hypercortisolism-related obesity could be altered in older
patients (e. g., diabetes-related catabolic effects or impaired AMPK signaling in
adipose tissue, hyporexia); however, evidence in humans is lacking [19]. Older females with CS rarely reported
hyperandrogenism-related signs, such as hirsutism, oligomenorrhoea, hair loss, and
reduced libido, which, as mentioned above, often lead younger female patients to
seek medical advice [6]
[9]
[10]. While the adrenal production of androgens may be less pronounced in
older patients due to aging, age-related differences in etiology may also play a
role. Indeed, older patients more often have adrenal CS, characterized by ACTH
suppression and consequent inhibition of adrenal androgen synthesis. Of note,
concentrations of 11-oxygenated C19 steroids were elevated in treatment-naïve
females with CD and correlated with salivary cortisol, suggesting that this hormone
excess is the main cause of hyperandrogenemia in CS [20]. However, it has been shown that
11-oxygenated C19 steroids were not lower in women aged>60 years in comparison to
females aged 20–40 years, in contrast to classic androgens concentration
(testosterone, androstenedione, and dehydroepiandrosterone-sulfate), which were
reduced after menopause [21].
The mechanisms whereby hypercortisolemia-related skin symptoms are lacking in the
elderly are not clear, although aging is associated with skin structural
derangements and decreased basal cell proliferation, which could make skin less
susceptible to the catabolic effects of cortisol [22].
Older patients more frequently present with muscle wasting and weakness, as compared
with their younger counterparts [9]
[10]. Such elevated prevalence of muscle
dysfunction in older CS patients is likely associated with age-related sarcopenia,
sedentary behavior, vitamin D deficiency, and concomitant disabilities, although a
specific effect of cortisol excess on aged muscle cannot be excluded [23]. Indeed, the administration of
dexamethasone to older mice reduced protein synthesis and promoted muscle atrophy
and hyperglycemia, partly by antagonizing the leucine stimulatory effect on muscle
[24]
[25]. Interestingly, in a longitudinal
study by Vogel et al., CS patients presented with sustained reduction of muscle
strength after successful treatment, which was associated with older age [26].
Moreover, type 2 diabetes was also more common in older CS patients as compared with
the younger, and this may contribute to the increased prevalence of muscle weakness
in the former [9]. An inverse association
between reduced handgrip strength and diabetes risk has been shown in older healthy
people [27]. In CS patients, Vogel et al.
also showed that metabolic derangements, such as greater waist-to-hip ratio and
higher HbA1C, predicted low muscle functionality in older subjects [26].
Growth hormone deficiency and hypogonadotropic hypogonadism, two common pituitary
hormone dysfunctions in both active CS and aging, maybe other important contributors
to the elevated prevalence of muscle weakness in aged patients with cortisol excess
[28]
[29]. Also, 11β-hydroxysteroid
dehydrogenase type 1 (11βHSD1), which converts inactive cortisone into active
cortisol, was up-regulated in healthy older women and associated with reduced grip
strength, insulin resistance, and adverse body composition profile [30].
Older CS patients more often experience bone fractures than their younger
counterparts, partly due to age-related frailty and muscle weakness, which increase
the risk of falls [9]. Older age and
prior hypercortisolism predicted low trabecular volumetric bone mineral density and
impaired measures of bone quality at the proximal femur in patients with CS in
remission, suggesting that fracture risk may be especially significant in the
elderly who had been exposed to elevated cortisol levels in the past [31].
As stated above, an age-related decline in sex steroids and growth hormones may also
contribute to bone loss [32]
[33]. An age-related increase in the
expression of 11βHSD1 has been described in animal models at the vertebral bone,
although there was no clear association with bone loss. Therefore, further studies
are needed to clarify the association between age-related changes in tissue
expression of 11ΒHSD1 and musculoskeletal impairment in CS [34].
Older age was also described as an important risk factor for venous thromboembolism
(VTE). In a study comprising 176 patients with active CS, the age over 69 years,
reduced mobility, infections, previous cardiovascular events, increased cortisol
level, and shortened APTT were independent VTE risk factors, with age and mobility
being the strongest among them [35].
However, other data showed that VTE risk was the highest at 3–6 postoperative
months, and a number of operations, male sex, and urine-free cortisol level (UFC)
were associated with VTE on multivariate analysis in patients with treated CS [36].
Cognitive impairment is another important issue in CS patients, which improves only
partially after successful treatment [37]. Studies in animals suggest that hypercortisolemia contributes to
age-related cognitive dysfunction [38].
It has been shown that aging, together with a lower level of education and a longer
duration of hypercortisolism, were related to both visual and vertebral memory
impairment and smaller hippocampal volume [39]. However, studies using functional magnetic resonance imaging (MRI),
especially longitudinal, are needed to evaluate the deleterious effect of
hypercortisolemia on cognitive function in different age groups, as well as its
reversibility after treatment [40].
Finally, whereas older CS patients typically presented with a significantly higher
burden of comorbidities, including hypertension, diabetes, cardiovascular disease,
and VTE, as compared with younger individuals, the prevalence of depression was
lower in the former [9]. This surprising
finding may be due to underdiagnosing. Indeed, late-life depression is often
characterized by less mood and motivation symptoms and more somatic features in
comparison to that diagnosed in younger patients, which may make the diagnosis of
affective disorders particularly challenging in aged people [41].
Although the available data do not show any age-related differences in terms of delay
in diagnosis, recognition of hypercortisolism in older patients may be difficult due
to lack of the typical stigmata of CS in them and attribution of comorbidities to
aging rather than hypercortisolemia [3]
[9].
Clinical presentation of ECS is frequently different from that in pituitary and
adrenal CS, in that symptoms are more severe, and their onset is more rapid.
Therefore, catabolic symptoms and comorbidities, such as hypertension and diabetes,
may predominate in the clinical picture, similar to what is reported in the elderly.
However, the profound hypokalemia and rapid progression are clues that often evoke
the diagnosis of EAS [13]. In a study
using machine learning, the most important features to distinguish EAS from CD were
serum potassium, ACTH, and MRI, while other clinical variables, including age, were
less important [42].
Potential mechanisms underlying the age-related differences in clinical presentation
of CS are summarized in [Table 1].
Table 1 Clinical presentation in the older patients with CS
compared with younger counterparts and potential mechanisms
associated.
|
Older patients
|
Potential mechanisms
|
|
Etiology
|
Mostly ACTH-independent
|
|
|
Gender distribution
|
Weaker female preponderance
|
|
|
Proteocatabolic symptoms
|
|
|
|
|
|
|
|
Hyperandrogenism-related symptoms
|
|
|
|
|
|
|
|
|
|
|
|
Cardiometabolic comorbidities
|
|
-
Physiological age-related increase in cardiometabolic
risk factors (endothelial dysfunction, atrial
fibrillation, dyslipidemia, insulin resistance) and
alteration of coagulation and fibrinolytic factors
|
|
|
|
|
|
|
|
|
CS, Cushing’s syndrome; ACTH, adrenocorticotropic hormone; IGF-I,
Insulin-like growth factor 1.
Biochemical tests in elderly patients with Cushing’s syndrome
Biochemical tests in elderly patients with Cushing’s syndrome
Biochemical diagnosis may also be more challenging for older patients
Daily cortisol secretion, assessed as the summation of the major glucocorticoid
metabolites of 24-hour urine samples (C21) when corrected for body surface area,
decreases slightly from 3–4 years to 7–8 years and, subsequently, starts to rise
again. After the age of 11–12 years, it becomes divergent between the sexes,
with significantly higher levels in males [43]. In the cohort of elderly included in the Baltimore Longitudinal
Study of Aging, the UFC/creatinine ratio presented a U-shaped pattern across the
lifespan, with the lowest results in the 20 s and 30 s, relative stability until
the 50 s, and increase thereafter [44]. Another study showed that 24-hour plasma-free cortisol
concentrations and cortisol production rate increased with age, independent of
body size [45]. Moreover, a
meta-analysis evaluating if healthy elderly had an abnormal cortisol response to
challenge showed a higher responsivity to stimulation and less inhibition after
a suppression test compared to that in younger people [46]. However, other authors reported
that, in acute stress conditions, older adults showed overall lower cortisol and
a lower stress-induced cortisol increase than younger adults [47]. In the ERCUSYN cohort, UFC
concentrations less frequently supported the diagnosis of hypercortisolism in
the older patients with pituitary CS in comparison to the younger, whereas other
hormonal tests did not differ according to age [9].
In the study by Qiao et al., no differences in any hormone measurements,
including late-night salivary cortisol (LNSC) and UFC, were documented (10). In
137 CD patients treated with long-acting pasireotide, Newell-Price et al. showed
greater LNSC levels in patients older than 60 years as compared with their
younger counterparts [48]. A linear
relationship between increasing age and LNSC levels has also been described in
healthy individuals [49].
Future studies are needed to evaluate the hypothalamus-pituitary-adrenal axis in
the elderly and assess the potential interference of several factors, including
suboptimal renal function, malnutrition, chronic illness, and concomitant
medications, on the reliability of the tests commonly used to diagnose
hypercortisolism in them.
Treatment and outcome according to age
Treatment and outcome according to age
Transsphenoidal surgery (TSS) remains the first-choice treatment for CD, but older
patients should be carefully selected, preferably based on complication avoidance
protocols in dedicated excellence centers. Transsphenoidal surgery was proven to be
a safe and effective procedure in 123 elderly patients with a pituitary adenoma
operated on in a tertiary neurosurgery center, with delayed hyponatremia being the
main reason for readmissions, which highlights the need for particular attention to
postoperative fluid restriction and sodium control protocols in this vulnerable
population [50]. Qiao et al. reported
that, despite a higher number of comorbidities and worse scores in preoperative risk
evaluation, older patients had similar complication and surgical remission rates as
those in the younger patients [10]. They
also found that recurrent disease after pituitary surgery was more frequent in the
younger patients than the older, which was confirmed in another study on 273
patients with CD followed up during 3.4 years [51]. Younger age was also one of three predictors of relapse, together
with postoperative cortisol and ACTH concentrations, in a study aimed at developing
machine-learning-based predictive models of disease recurrence [52]. These findings suggest that the
biological behavior of corticotropinomas in younger patients may have a more
aggressive pattern compared with that in the older population [52]. In line with this,
estrogen-sufficient CD women in the ERCUSYN showed a greater recurrence rate when
compared with postmenopausal female patients [9].
On the other hand, older European CD patients of both sexes more commonly had a
macroadenoma and less frequently experienced remission after TSS when compared with
the younger population [9].
Conservative approaches, such as radiotherapy and or medical therapy, were the
preferred therapeutic options in the older patients included in the ERCUSYN, whereas
surgery was the first-choice treatment in the younger [9].
The response to medical treatment might potentially differ according to age;
nonetheless, the data are scarce. Osilodrostat, a recently introduced 11βhydroxylase
inhibitor, was effective in normalizing hypercortisolemia also with a favorable
safety profile; however, the trials included only patients below 70 years of age
[53]
[54]. Ketoconazole was also used in
patients over 75 years of age with no increase in the percentage of adverse effects,
compared with younger patients, while age was not a significant predictor of the
drug efficacy [55]
[56]. Metyrapone was also administered to
patients older than 70, but detailed information about effectiveness and side
effects in this group is lacking [57]. In
case of severe hypercortisolemia, etomidate infusion may be used in older patients,
but they may require lower than standard doses [58]. In a study concerning mitotane use in CS (patients aged 14–71
years), age was not a predictor of drug efficacy [59]. In a longitudinal study on
twice-daily use of pasireotide, only one of five patients older than 65 reached 60
months of therapy, therefore no relevant conclusions could be drawn in this group of
patients [60].
Radiotherapy is an important treatment option in elderly CD patients with
contraindications to surgery due to an estimated general efficiency in two-thirds of
patients [61]. Indeed, older CD patients
included in the ERCUSYN more commonly underwent radiotherapy as a first-line
treatment as compared with the younger, in whom surgery was the preferred approach
[9]. Among risks potentially
associated with this treatment, cerebrovascular complications, cognitive
deterioration, and hypopituitarism seem especially important in the elderly
population [61]. Age, however, was not
associated with initial and durable control of hypercortisolism after stereotactic
radiosurgery in a study comprising 278 patients with CD [62]. Longitudinal head-to-head data
comparing the efficacy and safety of each treatment option according to age are
needed.
Mortality
In longitudinal studies with longer periods of observation, older age was associated
with an increased mortality, mainly due vascular events, regardless of remission
status [63]
[64]. Moreover, older age, depression,
diabetes, and male sex predicted the occurrence of cardiovascular events in a group
of 346 patients with CD [65]. In the
ERCUSYN cohort, mortality was 3% during a median follow-up of 2.7 years and was
independently associated with older age. Moreover, age was the only factor that
remained significant in a subanalysis of mortality within 90 days from the start of
treatment [66].
In ectopic CS, the most important predictors of overall survival are tumor histology
and non-lymph node metastases, while age-related differences in survival have not
been described [67].
Quality of life (QoL)
Older age negatively impacts the QoL in CS patients [68]. Younger CD patients more frequently
experienced an improvement in QoL compared to their older counterparts after
successful treatment. Elderly patients more often present with other factors that
may contribute to persistent QoL impairment, such as comorbidities, residual
morbidity, and multiple treatments [69].
In a questionnaire addressing many factors related to QoL in CS patients, some
age-related differences were observed. Younger patients more often reported concerns
about hair loss, loneliness, pain, and the impact of the disease on their marriage
or children than their older counterparts. The commonest complaints were related to
anxiety/depression, loss of physical strength, memory loss, and fatigue, without any
differences based on age [70].
Telomere theory
Telomeres are repetitive DNA sequences located at the end of linear chromosomes,
which prevent the loss of genetic material during cell division. Their critical
shortening induces apoptosis. Aging is characterized by a progressive shortening of
telomeres, which compromises their structure and function [71]. Telomerases are enzymatic complexes
maintaining telomere length, and their function might be regulated by genetic,
epigenetic, environmental, and hormonal factors [72]. Some studies have shown that cortisol
levels are correlated negatively with telomerase activity or telomere length [73]
[74]
[75]. Therefore,
hypercortisolemia might predispose to the acceleration of telomeres shortening in
CS, similar to what occurs in aging. After successful treatment of CD, telomeres
length increased compared to that observed during the active phase of the disease
and correlated with the time of remission. However, during the baseline case-control
study, no difference in telomere length was observed between patients and controls.
Of note, age was one of the factors related to telomeres shortening in patients with
CS [76]. However, because there is no
evidence that older CS patients have shorter telomeres than their younger
counterparts, future studies are needed to confirm the speculation that there is a
potential association between CS and telomere length over the course of the
disease.
Conclusions
Management of CS in the elderly may be more challenging than in younger individuals
due to many potential pitfalls in terms of symptom recognition and reliability of
hormone testing, greater burden of comorbidities, and lack of information about the
safety and efficacy of treatment options ([Table 2]). Thus, future studies are needed to establish the most
effective approach aimed at shortening the exposure time to excessive cortisol and
optimizing the diagnostic and therapeutic strategies specifically addressed to older
patients, with the ultimate goal to normalize survival and promote a satisfactory,
age-adjusted QoL.
Table 2 Potential pitfalls in CS management in the elderly
population.
|
DIAGNOSIS
|
Lack of some typical CS symptoms (skin changes, weight gain – see
Table 1)
|
|
Overlap between some symptoms/comorbidities that might be related
to CS and aging (hypertension, diabetes, muscle weakness,
cognitive impairment)
|
|
Many factors influencing laboratory results in diagnostics of
hypercortisolism (age-related changes in
hypothalamus-pituitary-adrenal axis, suboptimal renal function,
malnutrition, comorbidities, drugs)
|
|
TREATMENT
|
High preoperative risk
|
|
Lack of data concerning safety and efficacy of new drugs in older
population
|
|
OUTCOME
|
High mortality rate
|
|
Persistent QoL impairment after successful treatment
|
CS, Cushing’s syndrome; QoL, quality of life.
