Adrenalectomy in ectopic Cushing’s syndrome: A retrospective cohort study from a tertiary care centre

First published: 13 August 2021


Neuroendocrine neoplasms (NENs) causing ectopic Cushing’s syndrome (ECS) are rare and challenging to treat. In this retrospective cohort study, we aimed to evaluate different approaches for bilateral adrenalectomy (BA) as a treatment option in ECS. Fifty-three patients with ECS caused by a NEN (35 females/18 men; mean ± SD age: 53 ± 15 years) were identified from medical records. Epidemiological and clinical parameters, survival, indications for surgery and timing, as well as duration of surgery, complications and surgical techniques, were collected and further analysed. The primary tumour location was thorax (n = 30), pancreas (n = 14) or unknown (n = 9). BA was performed in 37 patients. Median time from diagnosis of ECS to BA was 2 months (range 1–10 months). Thirty-two patients received different steroidogenesis inhibitors before BA to control hypercortisolaemia. ECS resolved completely after surgery in 33 patients and severe peri- or postoperative complications were detected in 12 patients. There were fewer severe complications in the endoscopic group compared to open surgery (p = .030). Posterior retroperitoneoscopic BA performed simultaneously by a two surgeon approach had the shortest operating time (p = .001). Despite the frequent use of adrenolytic treatment, BA was necessary in a majority of patients to gain control over ECS. Complication rate was high, probably as a result of the combination of metastatic disease and metabolic disorders caused by high cortisol levels. The two surgeon approach BA may be considered as the method of choice in ECS compared to other BA approaches as a result of fewer complications and a shorter operating time.


Endogenous Cushing’s syndrome (CS) has an estimated incidence of 0.2–5.0 per million people per year.1 In 5–10% of these, it is caused by ectopic secretion of adrenocorticotrophic hormone (ACTH) or, in extremely rare cases, corticotrophin-releasing hormone, from a non-pituitary tumour.12

The treatment of neuroendocrine neoplasms (NENs) with ectopic secretion of ACTH is challenging. Because of the rarity and heterogeneity of this condition, there is no established evidence-based recommendation.3 Most patients with ectopic Cushing’s syndrome (ECS) have severe hypercortisolaemia leading to disrupted electrolyte and glucose levels, metabolic alkalosis, thrombosis and life-threatening infections, amongst many other manifestations. Initiation of oncological treatment is often delayed as a result of the consequences of high cortisol levels. A reduction of the cortisol level is crucial for survival and hypercortisolaemia and hypokalaemia are negative prognostic factors.45 If radical surgery of the tumour is not possible because of metastatic disease, normo-cortisolaemia can be achieved either by medical treatment with steroidogenesis inhibitors (SI) or bilateral adrenalectomy (BA),6 and BA has also been considered a treatment option for patients with occult or cyclic ECS. In patients with metastatic neuroendocrine carcinomas, platinum-based chemotherapy may be applied as first-line action, combined by SI and/or followed by BA. Computed tomography-guided percutaneous adrenal ablation has been reported in several case reports as a possible therapeutic alternative for patients in whom medical treatment has failed and BA is not feasible,710 althhough more data is needed to recommend this method in daily practice.

In the 1930s, transabdominal open access BA was introduced as a treatment option for uncontrolled cortisol secretion.11 Sixty years later, in the 1990s, laparoscopic methods were established1213 and are now considered as the gold standard for BA (except for adrenal carcinomas) because they result in less postoperative pain, a shorter hospitalisation time and faster recovery.14 Laparoscopic transperitoneal adrenalectomy (LTA) is the most frequently applied surgical method. However, posterior retroperitoneoscopic adrenalectomy (PRA), introduced in 1995 by Walz et al,15 is gaining popularity.16 Using PRA compared to LTA offers a more direct approach to the adrenal glands, a shorter operating time (no need for reposition of the patient), less blood loss and faster recovery, and it aso has advantages for patients with obesity or a history of previous abdominal surgery.16 There are centres where PRA is performed by a two surgeon approach; thus, a simultaneous bilateral approach offers the possibility of decreasing the surgical time by up to 50% and reducing operative stress.1719

The present study aimed to evaluate BA as a treatment option for ECS, as well as the effects of different approaches on morbidity and mortality. We hypothesised that endoscopic surgery methods could be superior regarding complication rate, operating and hospitalisation time compared to open access surgery and also influence overall survival.


2.1 Patients and data

A cohort of 59 patients with ECS was identified retrospectively from medical records of 894 patients with NENs, referred to the Department of Endocrine Oncology, Uppsala University Hospital between 1984 and 2019. None of the patients had a small-cell lung cancer (SCLC) because these tumours are not treated at our centre and possibly have a different mechanism behind ACTH production compared to that of NENs. Furthermore, SCLCs have a much more severe course of disease compared to well differentiated NENs and including them in the present study could mask any results important for NEN clinical management. Six patients were from outside Sweden and were excluded from further analysis because of a lack of follow-up data; thus, in total, 53 patients were available for analysis. Diagnosis of ECS was confirmed by histopathological examination of tumour specimen (n = 48) together with the clinical and biochemical picture of ACTH-dependent Cushing’s syndrome (elevated serum and urinary cortisol, high ACTH and pathological functional tests). In five patients where neither primary tumor, nor metastatic disease was found despite several PET examinations, including 68 Ga- DOTATOC-PET, 11C-5HTP-PET and 18FDG-PET in four of the five patients, ECS was confirmed on the basis of the clinical/biochemical picture and exclusion of pituitary origin by magnetic resonance imaging, as well as inferior sinus petrosus sampling.

Epidemiological data, data on clinical parameters, survival, indication and duration of surgery, complications and surgical technique were extracted and further analysed.

2.2 Surgery

BA was performed either by an open access approach, LTA or PRA. PRA was performed either by one surgeon (PRA-1S) or by two surgeons operating on both sides simultaneously (PRA-2S). Some patients were operated twice (one adrenal at the time) and, for those patients, operating time was pooled from both surgeries, if both sessions were performed within 1 week. Cases where conversion from an endoscopic to an open access approach was made peroperatively were grouped as open access surgery in further analysis. Patients who died during the postoperative stage (within 30 days) were excluded from calculation of hospitalisation time.

Postoperative complications were graded using the Clavien–Dindo classification where complications of Grade 1 are defined as “any deviation from the normal postoperative course without the need for pharmacological treatment or surgical, endoscopic and radiological interventions. Allowed therapeutic regimens are drugs as antiemetics, antipyretics, analgesics, diuretics and electrolytes and physiotherapy”.20 Because almost all patients had mild, Grade 1 postoperative complications (metabolic disturbances caused by hypercortisolaemia), this variable is not described. We defined complications up to Grade 2 as mild and Grade 3–5 as severe.

2.3 Statistical analysis

All parameters were analysed by descriptive statistics: normally distributed data as the mean ± SD, and data with skewed distribution and/or outliers were described as medians, accompanied by the 25th to 75th percentile ranges (Q1-Q3) or minimum-maximum (min-max). The defined event was death from any cause. Overall survival (OS) was defined as time from diagnosis of ECS or time of BA until date of death or, if the event was not found, censored at date of last observation, 31 December 2019. Kaplan-Meier plots were used for survival analysis and the log-rank test was used for comparison. Chi-squared was used for testing relationships between categorical variables. p < .05 was considered statistically significant. All statistical analyses were performed using IBM, version 27 (IBM Corp., Armonk, USA).


3.1 Studied patients

ECS represented six% (n = 59) of NENs in our cohort. Six patients were excluded from further analysis, resulting in 53 ECS patients who were analysed; there were 35 females and 18 males with a mean ± SD age of 53 ± 15 years. The localisation of the primary NEN was thorax (n = 30), pancreas (n = 14) or unknown (n = 9). Histopathological staining for Ki-67 was available in 38 patients and Ki-67 was < 2% in five patients, 3–20% in 22 patients and > 20% in 11 patients. Patient characteristics are shown in Tables 1 and 2. Twenty-two patients (41.5%) in this cohort had concomitant hypersecretion of hormones other than ACTH from their tumour (5-HIAA, n = 10; calcitonin, n = 3; 5-HIAA + calcitonin, n = 2; glucagon, n = 3, gastrin, n = 2; growth hormone, n = 1; insulin + gastrin + vasointestinal peptide, n = 1).

3.2 Surgery

Adrenalectomy was performed in 37 patients (70%); 24 patients were operated at Uppsala University Hospital, nine at Karolinska University Hospital in Stockholm and four at Umeå University Hospital. Median time from diagnosis of ECS to BA was 2 months (range 1–10 months). Median Ki-67 in patients who were operated within 2 months after ECS diagnosis was higher (Ki-67 18.5%) compared to those with BA performed later in the course of disease (Ki-67 9.5%), although the difference was not statistically significant (p = .085).

Thirty-two (86%) patients received different SI prior to BA to control hypercortisolaemia. Eight of those were treated with chemotherapy as well in an attempt to reduce cortisol levels. The majority of patients was treated with ketoconazole, often in combination with other drugs (Table 3). Indications for BA in our cohort included (1) persistent hypercortisolaemia despite use of SI (n = 30); (2) BA as first choice of treatment to reduce cortisol levels (n = 5); and (3) no effect combined with severe side effects from SI including liver toxicity and severe leukopenia (n = 2). In 16 patients, BA was not performed as a result of (1) good control of ECS with SI (n = 4); (2) radical surgery of the primary tumour (n = 3); (3) good control of ECS with SI followed by radical surgery of the primary tumour (n = 5) and (4) the bad condition of the patient (n = 4).

3.3 Survival analysis

There was no operative mortality in this cohort. Four patients died within 1 month after adrenalectomy (on day 5, 16, 22 and 30, respectively) as a result of multiple organ dysfunction syndrome and progression of NEN. At the end of the follow-up period, 14 patients were still alive and 39 had died.

Median survival after BA was 24 months (95% confidence interval [CI] = 7–41, min-max: 0–428) with a 5-year survival of 22%. Median follow-up time for all patients from time of ECS diagnosis was 26 (range 6–62) months and after BA was 19 (range 3–50) months. OS was longer in patients where ECS was treated by radical surgery of the primary tumour or where good biochemical control was achieved by SI compared to patients who underwent BA, 96 months (95% CI 0–206) vs 29 months (95% CI 7–51), respectively. However, this difference was not statistically significant (p = .086), most likely as a result of the small sample size. Multiple hormone secretion correlated with shorter OS after BA (p = .009; hazard ratio = 2.9; 95% CI= 1.3–6.7). There was no significant difference in OS after BA depending on localisation of primary tumour (thoracic NENs 24 months [95% CI = 8–40, min-max: 0–428], pancreatic NENs 19 months [95% CI = 0–43, min-max: 0–60], p = .319) or surgical approach (open access approach 24 months [95% CI = 1–47], endoscopic approach 19 months [95% CI = 1–37], p = .720).

Median time from ECS diagnosis to BA was 2 months (range 1–10). Patients who underwent BA within 2 months after ECS diagnosis had shorter OS compared to those who were operated at a later stage: 6 months (95% CI = 0–18) and 45 months (95% CI = 3–86) respectively (p = .007). The former group had a slightly higher median Ki-67 level (18% vs 9%), lower potassium (2.7 mmol L-1 vs 3.0 mmol L-1) and higher hormone levels (ACTH 217 vs 120 ng mL-1, morning cortisol 1448 vs 1181 nmol L-1 and UFC 5716 vs 4234 nmol per 24 h) at diagnosis compared to those who were operated later in the course of disease.


The present study highlights new aspects of the advantages of an endoscopic approach of BA compared to open access surgery, regarding the incidence of severe complications graded using the Clavien-Dindo classification, as well as operation- and hospitalisation time. Our results indicate that PRA performed by two surgeons simultaneously is the method of choice for patients with ECS. However, despite these advantages, the endoscopic approach did not appear to improve overall survival.

Recent Endocrine Society guidelines recommend SI as primary treatment for ECS in patients with occult or metastatic ECS followed by BA.6 Although the toxicity of SI in our cohort was low (n = 2; 6%), 32 patients (73%) had persistent hypercortisolaemia despite medical treatment and proceeded to BA. BA, especially with an endoscopic approach, with a short operating time and low complication risk, appears to play a major role in the appropriate management of hypercortisolaemia in ECS, where rapid reduction of cortisol levels is very important.

Prolonged exposure to high cortisol levels, in combination with high risk for hepatotoxic and nephrotoxic SI side effects, increases morbidity and risk for severe complications, and often delays the start of oncological treatment. However, the trauma caused by surgery can also postpone initiation of chemotherapy.21 Therefore, a fast and minimally invasive surgical procedure appears to be a crucial factor for the better survival in ECS. The endoscopic approach is now considered as the gold standard for BA. Our study presents fewer severe complications, as well as shorter operating and hospitalisation times, when the endoscopic approach is compared with open surgery. In line with previous studies,1922 we observed a significantly shorter operating time when applying PRA compared to LTA because there is no need for repositioning of the patient during PRA. PRA-2S had the shortest operating time and should be considered as the best choice of surgical approach in ECS. This result ties well with previous studies reporting the median operating time to be between 43 and 157 min in PRA-2S, which is significantly shorter compared to LTA and PRA-1S.1719

The median time from diagnosis to BA was 2 months, which is consistent with a previous study.23 However, OS was significantly shorter in patients who were operated within 2 months after diagnosis of ECS in our cohort compared to those operated at a later stage. These early operated patients probably had a more aggressive clinical course of disease, as indicated by slightly higher median Ki-67, lower potassium and higher hormone levels at diagnosis, and they were operated as a result of more acute indications (without time to proper pre-treatment with SI) than the other group.

In our previous report on patients with ACTH-producing NENs, multiple hormone secretion was identified as the strongest indicator of a worse prognosis.4 A similar pattern of results was observed in this cohort, showing that patients with NENs, with concomitant hypersecretion of other hormones than ACTH from their tumour, had a shorter OS after BA compared to those with ACTH hypersecretion only.

As a result of the extremely high preoperative cortisol levels in ECS, the substitution therapy needed after successful BA may be challenging.21 Over-replacement of glucocorticoids may lead to higher morbidity24 and mortality, especially in patients with metastatic NENs, who often have impaired immune function because of oncological treatment. Many patients suffer from glucocorticoid withdrawal syndrome, despite adequate replacement therapy, and it can take ≥ 1 year to gain control over these symptoms.6 This frequently leads to high dosage of glucocorticoids. The Endocrine Society guidelines recommend glucocorticoid replacement with hydrocortisone, 10–12 mg m-2 day-1 in divided doses.6 If we assume that most of our patients have body surface area around 2 m2 or less, the daily hydrocortisone dose should not exceed 25 mg. However, 1 year after BA, only one patient received 25 mg of hydrocortisone daily, with the majority receiving 30 mg or more. One-third of the patients had residual arterial hypertension and diabetes 3 months after BA, probably partially depending on too high a dose of glucocorticoids.

There was a higher complication rate in our cohort compared to other studies192526 and five patients needed conversion from an endoscopic approach to open surgery. In particular, the outcome of BA in ECS has not previously been systematically evaluated27 because most of the reports include patients with various aetiologies of CS.1922232829 In a systematic review of the literature published between 1980 and 2012 on BA in CS, Reincke et al23 identified 37 studies and ECS was present in 13% of the patients. There are only few papers focused on BA in ECS solely2125263031 and only one has a cohort with > 50 patients (n = 54).26 Patients with ECS have almost always a more aggressive course and more severe metabolic disturbance than patients with other types of Cushing’s syndrome, which probably leads to higher risk for postoperative complications. Furthermore, multiple liver metastases, fibrotic processes in the abdomen as a result of previous surgery or large primary tumour in pancreas could be some of the factors influencing surgical outcome in ECS.

The present study has several limitations. First, all data were collected retrospectively from patient records and not all the preferred parameters were available for all patients. Second, even if our cohort is one of the largest regarding studies on BA in ECS, the number of patients is too low for reliable statistical analysis. Finally, our study covered more than three decades, BAs were performed at different clinics and by different surgeons. Therefore, the data should be interpreted carefully.

In conclusion, the present study is one of few reports focusing on BA in specifically NEN patients with ECS and includes one of the largest patient cohorts analysed in the field. PRA-2S can be considered as method of choice in ECS compared to other BA approaches. The aim is to avoid administrating too high a hydrocortisone replacement dosage postoperatively because this can worsen the metabolic disturbance. As a result of the rarity of the condition, multicentre studies are needed with large, prospective cohorts and standardised inclusion criteria, aiming to further improve our knowledge about the management of ECS.


This study was funded by the Swedish Cancer Society (grant number CAN 18 0576), the Lions Foundation for Cancer Research at Uppsala University Hospital, Selanders Foundation and Söderbergs foundation at Uppsala University.


The authors declare that they have no conflicts of interest.


Ieva Lase: Conceptualisation; Data curation; Formal analysis; Investigation; Methodology; Visualisation; Writing – original draft; Writing – review & editing. Malin Grönberg: Formal analysis; Supervision; Visualisation; Writing – review & editing. Olov Norlén: Conceptualisation; Writing – review & editing. Peter Stålberg: Conceptualisation; Writing – review & editing. Staffan Welin: Conceptualisation; Supervision; Writing – review & editing. Eva Tiensuu Janson: Conceptualisation; Funding acquisition; Methodology; Supervision; Writing – review & editing.


The need for informed consent was waived by the local ethics committee. All procedures performed in this study were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. The study was approved by the local ethics committee, Regionala etikprövningsnämnden (EPN), in Uppsala, Sweden.


The peer review history for this article is available at

The entire article, PDF, supporting tables and more can be found at

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: