cushing es15

Treatment of Cushings Syndrome: An EndocrineSociety Clinical Practice Guideline

Lynnette K. Nieman (chair), Beverly M. K. Biller, James W. Findling,M. Hassan Murad, John Newell-Price, Martin O. Savage, and Antoine Tabarin

Program in Reproductive and Adult Endocrinology (L.K.N.), The Eunice Kennedy Shriver National Instituteof Child Health and Human Development, National Institutes of Health, Bethesda, Maryland 20892;Neuroendocrine Unit (B.M.K.B.), Massachusetts General Hospital, Boston, Massachusetts 02114;Medical College of Wisconsin (J.W.F.), Milwaukee, Wisconsin 53226; Mayo Clinic (M.H.M.), Division ofPreventive Medicine, Rochester, Minnesota 55905; Department of Human Metabolism (J.N.-P.), Schoolof Medicine and Biomedical Science, University of Sheffield, Sheffield S10 2RX, United Kingdom; WilliamHarvey Research Institute (M.O.S.), Barts and the London School of Medicine and Dentistry, LondonEC1M 6BQ, United Kingdom; and Department of Endocrinology (A.T.), Centre Hospitalier Universitairede Bordeaux and Inserm 862, University of Bordeaux, 33077 Bordeaux, France

Objective: The objective is to formulate clinical practice guidelines for treating Cushings syndrome.

Participants: Participants include an Endocrine Society-appointed Task Force of experts, a methodol-ogist, and a medical writer. The European Society for Endocrinology co-sponsored the guideline.

Evidence: The Task Force used the Grading of Recommendations, Assessment, Development, andEvaluation system to describe the strength of recommendations and the quality of evidence. TheTask Force commissioned three systematic reviews andused thebest available evidence fromotherpublished systematic reviews and individual studies.

Consensus Process: The Task Force achieved consensus through one group meeting, several con-ference calls, and numerous e-mail communications. Committees and members of The EndocrineSociety and the European Society of Endocrinology reviewed and commented on preliminarydrafts of these guidelines.

Conclusions: Treatment of Cushings syndrome is essential to reducemortality and associated comor-bidities. Effective treatment includes the normalization of cortisol levels or action. It also includes thenormalizationof comorbidities viadirectly treating thecauseofCushings syndromeandbyadjunctivetreatments (eg, antihypertensives). Surgical resection of the causal lesion(s) is generally the first-lineapproach. The choice of second-line treatments, including medication, bilateral adrenalectomy, andradiation therapy (for corticotrope tumors), must be individualized to each patient. (J Clin EndocrinolMetab 100: 28072831, 2015)

Summary of Recommendations

1. Treatment goals for Cushings syndrome1.1 In patients with overt Cushings syndrome (CS), we rec-ommendnormalizing cortisol levels or action at its receptorsto eliminate the signs and symptoms of CS and treating co-morbidities associated with hypercortisolism. (1QQQE)

1.2We recommendagainst treatment to reduce cortisollevels or action if there is not an established diagnosis ofCS. (1QEEE)

1.3We suggest against treatments designed to normal-ize cortisol or its action when there is only borderline bio-chemical abnormality of the hypothalamic-pituitary-ad-renal (HPA) axis without any specific signs of CS. The

ISSN Print 0021-972X ISSN Online 1945-7197Printed in USACopyright 2015 by the Endocrine SocietyReceived March 30, 2015. Accepted June 19, 2015.First Published Online May 14, 2015

Abbreviations:ACC,adrenocorticalcarcinoma;BMAH,bilateralmacronodularadrenalhyperplasia;BMD, bone mineral density; CD, Cushings disease; CS, Cushings syndrome; CT, computerizedtomography; EAS, ectopic ACTH secretion; HPA, hypothalamic-pituitary-adrenal; HRQOL, health-related QOL; MRI, magnetic resonance imaging; QOL, quality of life; RT, radiation therapy; SST,somatostatin receptor; TSS, transsphenoidal selective adenomectomy; UFC, urine free cortisol.

S P E C I A L F E A T U R E

C l i n i c a l P r a c t i c e G u i d e l i n e

doi: 10.1210/jc.2015-1818 J Clin Endocrinol Metab press.endocrine.org/journal/jcem 1

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 August 2015. at 06:18 For personal use only. No other uses without permission. . All rights reserved.

benefit of treating to normalize cortisol is not establishedin this setting. (2QEEE)

2. Optimal adjunctive management2.1 We recommend providing education to patients

and their family/caretaker(s) about their disease, treat-ment options, and what to expect after remission. (Un-graded best practice recommendation)

2.2We recommend that all patients receive monitor-ing and adjunctive treatment for cortisol-dependentcomorbidities (psychiatric disorders, diabetes, hyper-tension, hypokalemia, infections, dyslipidemia, osteo-porosis, and poor physical fitness). (Ungraded bestpractice recommendation)

2.3 We recommend that a multidisciplinary team, in-cluding an experienced endocrinologist, takes patient val-ues and preferences into consideration and provides edu-cation about the treatment options to the patient.(Ungraded best practice recommendation)

2.4We suggest evaluatingCSpatients for risk factors ofvenous thrombosis. (2QQEE)

2.5 In patients with CS undergoing surgery, we suggestperioperative prophylaxis for venous thromboembolism.(2QQEE)

2.6 We recommend that clinicians discuss and offerage-appropriate vaccinations toCSpatientsparticularlyinfluenza, Herpes zoster, and pneumococcal vaccina-tionsdue to an increased risk of infection. (Ungradedbest practice recommendation)

3. First-line treatment options3.1Werecommend initial resectionofprimary lesion(s)

underlying Cushings disease (CD), ectopic and adrenal(cancer, adenoma, and bilateral disease) etiologies, unlesssurgery is not possible or is unlikely to significantly reduceglucocorticoid excess (Figure 1). (1QQQQ)

3.1a We recommend unilateral resection by an experi-enced adrenal surgeon for all cases of benign unilateraldisease. (1QQQE)

3.1b We recommend localizing and resecting ectopicACTH-secreting tumors with node dissection as appro-priate. (1QQQQ)

3.1c We recommend transsphenoidal selective ad-enomectomy (TSS) by an experienced pituitary surgeon asthe optimal treatment for CD in pediatric and adult pa-tients. (1QQQQ)

3.1ci We recommend measuring serum sodium severaltimes during the first 514 days after transsphenoidal sur-gery. (1QQEE)

3.1cii We recommend assessing free T4 and prolactinwithin 12 weeks of surgery, to evaluate for overt hypop-ituitarism. (1QQEE)

3.1ciii We recommend obtaining a postoperative pitu-itary magnetic resonance imaging (MRI) scan within 13months of successful TSS. (Ungraded best practicestatement)

3.1d We recommend surgical resection of bilateral ad-renal disorders (1QQEE) and suggest medical therapy toblock aberrant hormone receptors for bilateral ma-cronodular adrenal hyperplasia (BMAH) (2QQEE).

4. Remission and recurrence after surgical tumorresection

4.1 We suggest an individualized management ap-proach based onwhether the postoperative serum cortisolvalues categorize the patients condition as hypocortiso-lism, hypercortisolism, or eucortisolism. (Ungraded bestpractice statement)

4.2 We recommend additional treatments in patientswith persistent overt hypercortisolism. (1QQQQ)

4.3 We recommend measuring late-night salivary orserum cortisol in patients with eucortisolism after TSS,including those cases where eucortisolismwas establishedby medical treatment before surgery. (1QQEE)

4.4 We recommend using tests to screen for hypercor-tisolism to assess for recurrence in patients with ACTH-dependent CS. (1QQQE)

5. Glucocorticoid replacement and discontinuation,and resolution of other hormonal deficiencies

5.1 We recommend that hypocortisolemic patients re-ceive glucocorticoid replacement and education about ad-renal insufficiency after surgical remission. (1QQQQ)

5.2We recommend follow-up morning cortisol and/orACTH stimulation tests or insulin-induced hypoglycemiato assess the recovery of the HPA axis in patients with atleast one intact adrenal gland, assuming there are no con-traindications. We also recommend discontinuing gluco-corticoid when the response to these test(s) is normal.(1QQQE)

5.3 We recommend re-evaluating the need for treat-ment of other pituitary hormone deficiencies in the post-operative period. (1QQQE)

6. Second-line therapeutic options6.1 In patients with ACTH-dependent CS who under-

went a noncurative surgery or for whom surgery was notpossible, we suggest a shared decision-making approachbecause there are several available second-line therapies(eg, repeat transsphenoidal surgery, radiotherapy, medi-cal therapy, and bilateral adrenalectomy). (2QQEE)

6.1a We suggest bilateral adrenalectomy for occult ormetastatic ectopic ACTH secretion (EAS) or as a life-pre-serving emergency treatment in patients with very severe

2 Nieman et al Guidelines on Cushings Syndrome Treatment J Clin Endocrinol Metab


ACTH-dependent disease who cannot be promptly con-trolled by medical therapy. (2QQQE)

6.1b We recommend regularly evaluating for cortico-trope tumor progression using pituitaryMRIs and ACTHlevels in patients with known CD who undergo bilateraladrenalectomy and in patients who undergo this proce-dure for presumed occult EAS (because some of the latterhave a pituitary and not ectopic tumor). (1QQQE)

6.2 Repeat transsphenoidal surgery6.2 We suggest repeat transsphenoidal surgery, partic-

ularly in patientswith evidence of incomplete resection, ora pituitary lesion on imaging. (2QQEE)

6.3 Radiation therapy/radiosurgery for CD6.3We recommend confirming that medical therapy is

effective in normalizing cortisol before administering ra-diation therapy (RT)/radiosurgery for this goal becausethis will be needed while awaiting the effect of radiation.(1QEEE)

6.3aWe suggest RT/radiosurgery in patients who havefailed TSS or have recurrent CD. (2QQEE)

6.3b We recommend using RT where there are con-cerns about the mass effects or invasion associated withcorticotroph adenomas. (1QQQE)

6.3cWe recommendmeasuring serum cortisol or urinefree cortisol (UFC) off-medication at 6- to 12-month in-tervals to assess the effect of RT and also if patients de-velop new adrenal insufficiency symptomswhile on stablemedical therapy. (1QQQE)

6.4 Medical treatment6.4 We recommend steroidogenesis inhibitors under

the following conditions: as second-line treatment afterTSS in patients with CD, either with or without RT/ra-diosurgery; as primary treatment of EAS in patients withoccult or metastatic EAS; and as adjunctive treatment toreduce cortisol levels in adrenocortical carcinoma (ACC).(1QQQE)

6.4a We suggest pituitary-directed medical treatmentsin patients with CD who are not surgical candidates orwho have persistent disease after TSS. (2QQQE)

6.4bWe suggest administering a glucocorticoid antag-onist in patients with diabetes or glucose intolerance whoare not surgical candidates or who have persistent diseaseafter TSS. (2QQQE)

6.4c We suggest targeted therapies to treat ectopicACTH syndrome. (2QEEE)

7. Approach for long-term follow-up7.1 We recommend treating the specific comorbidities

associated with CS (eg, cardiovascular risk factors, osteo-

porosis and psychiatric symptoms) in all patients with CSthroughout their lives until resolution (Figure 1). We alsorecommend testing for recurrence throughout life, exceptin patients who underwent resection of an adrenal ade-noma with a computerized tomography (CT) density of 10 Hounsfield units. (1QQQE)

7.2 We recommend educating patients and familiesabout the clinical features of remission. (Ungraded bestpractice statement)

7.3 In patients with adrenal adenoma, we suggest fol-low-up tests for the specific comorbidities associated withCS if the adenoma density on CT was 10 Hounsfieldunits. (2QQEE) For those with higher Hounsfield unitvalues or pathology consistent with possible carcinoma,we suggest evaluating for malignancy using imaging.(2QEEE)

7.4We recommend that patients with Carney complexhave lifelong follow-up tests for cardiac myxoma andother associated disease (testicular tumors, acromegaly,thyroid lesions). (1QQQQ)

8. Special populations/considerations8.1We recommend urgent treatment (within 2472 h)

of hypercortisolism if life-threatening complications ofCSsuch as infection, pulmonary thromboembolism, cardio-vascular complications, and acute psychosis are present.(1QQQE). The associated disorder(s) should be ad-dressed as well (eg, anticoagulation, antibiotics).

Developmental Method for Evidence-Based Clinical Practice Guidelines

The Clinical Guidelines Subcommittee of the EndocrineSociety deemed the treatment of CS a priority area in needof practice guidelines and appointed a Task Force to for-mulate evidence-based recommendations. TheTask Forcefollowed the approach recommended by the Grading ofRecommendations, Assessment, Development, and Eval-uation groupan international group with expertise inthe development and implementation of evidence-basedguidelines (1). A detailed description of the gradingscheme has been published elsewhere (2). The Task Forceused the best available research evidence to develop therecommendations. The Task Force also used consistentlanguage and graphical descriptions of both the strengthof a recommendation and the quality of evidence. In termsof the strength of the recommendation, strong recommen-dations use the phrase we recommend and the number1, and weak recommendations use the phrase we sug-gest and the number 2. Cross-filled circles indicate thequality of the evidenceQEEE denotes very low quality

doi: 10.1210/jc.2015-1818 press.endocrine.org/journal/jcem 3


evidence; QQEE, low quality; QQQE, moderate quality;and QQQQ, high quality. The Task Force has confidencethat persons who receive care according to its strong rec-ommendations will derive, on average, more good thanharm. Weak recommendations require more careful con-sideration of the persons circumstances, values, and pref-erences to determine the best course of action. Linked toeach recommendation is a description of the evidence andthe values that panelists considered when making the rec-ommendation; in some instances, panelists offer technicalsuggestions for testing conditions, dosing, and monitor-ing. These technical comments reflect the best availableevidence applied to a typical person being treated. Oftenthis evidence comes from the panelists values, prefer-ences, and unsystematic observations; therefore, these re-marks should be considered as suggestions. To make theguideline comprehensive, the task force included severaloverarching statements to emphasize some accepted clin-

ical principles that are not supported by clear direct evi-dence.These statements are explicitlymarkedasungradedbest practice statements.

The Endocrine Society maintains a rigorous conflict-of-interest review process for the development of clinicalpractice guidelines. All Task Force members must declareany potential conflicts of interest. These are reviewed be-foremembers are approved to serve on theTask Force andreviewed periodically during the development of theguideline. The Clinical Guidelines Subcommittee vets theconflict-of-interest forms before the members are ap-proved by the Societys Council to participate on theguideline Task Force.Most participantswho help developtheguidelinemusthavenoconflict of interest related to thematter under study.

Those participants who do have conflicts of interestmust disclose all conflicts. The Clinical Guidelines Sub-committee and the Task Force have reviewed all disclo-

Figure 1. An algorithm for the treatment of CS. Derived from Nieman LK, Biller BM, Finding, JW, et al. The diagnosis of Cushings syndrome: anEndocrine Society Clinical Practice Guideline. J Clin Endocrinol Metab. 2008;93:15261540. (17)



sures for this guideline and resolved or managed all iden-tified conflictsof interest.Mostmembersof theTaskForcehave relationships with companies that make pharmaceu-ticals for the treatment of CS. The Task Force explicitlydiscussed the possibility of a perception of conflict andagreed that the sections on individual pharmaceuticalswould bewritten by an individual without such a conflict;subsequent changes were approved by the entire group.

Conflicts of interest are defined as receiving any com-pensation from commercial interest(s) in the form ofgrants; research support; consulting fees; salary; owner-ship interest (eg, stocks, stock options, or ownership in-terest excluding diversified mutual funds); honoraria orother payments for participation in speakers bureaus, ad-visory boards, or boards of directors; or other financialbenefits. Completed forms are available through the En-docrine Society office.

The Endocrine Society was the only funding source forthis guideline; the Task Force received no funding or re-muneration from commercial or other entities.

Commissioned systematic reviewsThe three reviews summarized data from 29 case series

of radiotherapy in CD, 21 case series of radiosurgery inCD,and87case seriesof treatment-naiveCDpatientswhoreceived first-line transsphenoidal surgery. The outcomesof interest were biochemical remission and biochemicalrecurrence rates. In each review, we tested several patient,surgeon, and procedure variables to determine whetherthey were predictors of remission and recurrence rates.

Overall, analyses were underpowered to determineimportant independent predictors of the outcomes ofinterest. The quality of the evidence for recurrence andremission outcomes was low due to high risk of bias,heterogeneity, and imprecision.

1. Treatment goals for Cushings syndrome1.1 In patients with overt CS, we recommend normal-

izing cortisol levels or action at its receptors to eliminatethe signs and symptoms of CS and treating comorbiditiesassociated with hypercortisolism. (1QQQE)

1.2We recommendagainst treatment to reduce cortisollevels or action if there is not an established diagnosis ofCS. (1QEEE)

1.3We suggest against treatments designed to normal-ize cortisol or its action when there is only borderline bio-chemical abnormality of the hypothalamic-pituitary-ad-renal (HPA) axis without any specific signs of CS. Thebenefit of treating to normalize cortisol is not establishedin this setting. (2QEEE)

EvidenceCS is a condition of pathological hypercortisolism that

includes demonstrable clinical features. The goals of treat-ing CS are to eliminate its primary cause and achieve re-mission so as to eliminate the associated signs, symptoms,and comorbidities and to improve quality of life (QOL). In1952, before effective treatment was available, patientswith CS had a median survival of 4.6 years (3, 4). Sixtyyears later, despite available treatments for comorbidities,patientswith activeCS continue to have an increased stan-dardized mortality rate that is 1.7 to 4.8-fold greater thanthe general population (59). In contrast, when CS and itsassociated comorbidities are successfully treated, the stan-dardized mortality rate improves. Whether it is similar ornot to the general population or remains significantly in-creased remains debatable (59). As expected, patientswith persistent or recurrent hypercortisolism continue tohave higher than expected mortality (Hazard Ratio, 2.816) (6, 911).

Cardiovascular disease, venous thrombosis, and infec-tions are the primary causes of the excessmortality ratewesee in CS. Research indicates that the risk of infection islower in patients with mild to moderate vs severe hyper-cortisolism, which indirectly supports intervention (12).Altogether, these data suggest that prompt treatment ofCS and its comorbidities is important to reduce mortality.

Restoring eucortisolism leads to clinical and biochem-ical improvements regarding obesity, arterial hyperten-sion, insulin resistance, glucose tolerance, dyslipidemia,bone mineral density (BMD), linear growth in children,cognition, psychiatric disorders, and health-related QOL(HRQOL) (13). However, these complications persist inmany patients and should be addressed therapeuticallybefore and after remission from CS (14, 15) (see below).

Although there are no controlled studies in pediatricCS, growth andbody composition generally improve aftertreatment.Mostpatients reachanadult heightwithin theirpredicted parental target range, although they may needGH therapy to achieve this (16).

Because all treatments carry risk, clinicians should es-tablish a diagnosis of CS before administering them (17).However, in life-threatening situations, a clinical diagno-sis with minimal available biochemical data may justifyprompt treatment. Similarly, the consequences of mild orcyclic hypercortisolism are not clear, so that treatmentguidelines cannot be generalized to those patients. How-ever, because CS tends to progress to severe hypercorti-solism, it is possible that early recognition and treatmentofmild or cyclic disease (values1.5-fold upper referencerange) would reduce the risk of residual morbidity. Un-fortunately, few data address this assumption. If the cli-nician is uncertain of the clinical diagnosis (regardless of



themagnitude of biochemical perturbations), further test-ing over time is always the best approach. The treatmentof subclinical CS in the context of the evaluation of anadrenal incidentaloma isoutside the scopeof this guidelineon clinical CS.

2. Optimal adjunctive management2.1 We recommend providing education to patients

and their family/caretaker(s) about their disease, treat-ment options, and what to expect after remission. (Un-graded best practice recommendation)

2.2We recommend that all patients receive monitor-ing and adjunctive treatment for cortisol-dependentcomorbidities (psychiatric disorders, diabetes, hyper-tension, hypokalemia, infections, dyslipidemia, osteopo-rosis, and poor physical fitness). (Ungraded best practicerecommendation)

2.3 We recommend that a multidisciplinary team, in-cluding an experienced endocrinologist, takes patient val-ues and preferences into consideration and provides edu-cation about the treatment options to the patient(Ungraded best practice recommendation)

2.4We suggest evaluatingCSpatients for risk factors ofvenous thrombosis. (2QQEE)

2.5 In patients with CS undergoing surgery, we suggestperioperative prophylaxis for venous thromboembolism.(2QQEE)

2.6 We recommend that clinicians discuss and offerage-appropriate vaccinations toCSpatientsparticularlyinfluenza, Herpes zoster, and pneumococcal vaccina-tionsdue to an increased risk of infection. (Ungradedbest practice recommendation)

EvidenceSevere hypercortisolism impairs immunity and predis-

poses to severe, systemic infection and/or sepsis due tobacterial, fungal, and opportunistic pathogens (12, 1820). Based on this, we suggest immunization against in-fluenza, shingles, andpneumonia.Hypercortisolismalterscoagulation-factor profiles for up to 1 year after a surgicalcure (21) and carries an increased risk of venous throm-bosis, especially in the 4 weeks after surgery (2124). Cli-nicians caring for patients with CS should be aware of thisincreased risk, should evaluate their patients for throm-bosis and bleeding (25, 26), and should consider antico-agulation treatments. Clinicians should treat all othercomorbidities aggressively and seek appropriate consul-tations (including rehabilitation medicine). We suggestthat patients receive written information regarding theirdisorder.

3. First-line treatment options3.1Werecommend initial resectionofprimary lesion(s)

underlying CD, ectopic and adrenal (cancer, adenoma,and bilateral disease) etiologies, unless surgery is not pos-sible or unlikely to significantly reduce glucocorticoid ex-cess. (1QQQQ)

EvidenceComplete surgical resection of the causal tumor(s) (or

adrenal hyperplasia) is the optimal treatment of CS be-cause it alleviates hypercortisolism while potentially leav-ing the normal HPA axis intact (except for bilateral ad-renal disorders). The common causes of CS includeintrinsic adrenal glandabnormalities andACTHsecretionfrom a corticotrope tumor (CD) or from an ectopic tumor(EAS). Thus, differential diagnostic testing and tumor lo-calization studies are critical to a successful outcome (27).

3.1a We recommend unilateral resection by an experi-enced adrenal surgeon for all cases of benign unilateraldisease. (1QQQE)

EvidenceIn experienced hands, unilateral adrenalectomy is cu-

rative in nearly 100%of adults and childrenwith cortisol-producing adrenal adenomas; the complication rate ishigher when performed by surgeons with less experience(28). Adrenalectomy is generally performed via eithertrans or retroperitoneal laparoscopy unless other factorspreclude this. The role of adrenalectomy (vs intensivemedical treatment of comorbidities) in patients with ad-renal incidentaloma and subtle cortisol dysregulation (so-called subclinical CS) is not clear. However, retrospectivestudies suggest that patients withmetabolic abnormalitiesare likely to benefit (29).

Patients with ACC have a poor prognosis, and clini-cians should attempt definitive treatment via complete re-section. In children, complete resection was associatedwith survival rates of 80%, whereas the outlook forunresectable diseasewas very poor (30). Surgeons increas-ingly perform transperitoneal laparoscopic adrenalec-tomy, particularly for small tumors, but this is controver-sial becausemany recommendopen surgery to assess stageand achieve en bloc resection (31, 32). Because hypercor-tisolism is associated with increased mortality, medicaltreatment may be needed to achieve eucortisolism (33,34).

Follow-up, optimally by an adrenal cancer-specificmultidisciplinary team,may include cytotoxic chemother-apy and adjuvant radiotherapy or mitotane treatments(35). Further discussion is beyond the scope of theseguidelines.



3.1b We recommend localizing and resecting ectopicACTH-secreting tumors with node dissection as appro-priate. (1QQQQ)

EvidenceIn the absence of overt metastatic disease, ectopic

ACTH-producing tumor resection cured 76%of patients,as demonstrated by observational studies (36).

3.1c We recommend TSS by an experienced pituitarysurgeon as the optimal treatment for CD in pediatric andadult patients. (1QQQQ)

EvidenceMany centers have replaced the transnasal route with

an endoscopic endonasal approach (37), but both meth-ods canbe effective (38). Successful resection ismost likelywhenperformedbyanexperiencedneurosurgeonwhohasa high volume of these cases.

These are typically benign microadenomas (1 cm di-ameter) and are evident on pituitary MRI in approxi-mately 60% of adults (39) and approximately 55% ofchildren (40, 41). Some, but not all, reports demonstratea greater success when clinicians identify the tumor byMRI before surgery (4244). However, because 10% ofhealthy adults have pituitary lesions 6 mm that are vis-ible on MRI (45), the presence of a lesion does not defin-itively confirm the diagnosis of CD or identify the causaltumor. In addition, there is a 12% rate of abnormal pitu-itaryMRI scans in patients with EAS (18) and a 12% rateof false localization by MRI in patients with surgicallyproven CD (46). When feasible, all patients with ACTH-dependent CS and no obvious causal neoplasm (6 mm)should be promptly referred to an experienced center thatcan safely and reliablyperform inferiorpetrosal sinus sam-pling to distinguish a pituitary from a nonpituitary (ecto-pic) cause.

Inferior petrosal sinus sampling does not reliably iden-tify the tumor site because a side-to-side gradient of 1.4correctly predicted location in only 5669% of adults(4648) and 5981% of children (40, 49). Hemihy-pophysectomy, based on inferior petrosal sinus samplinglateralization, cured only 50%of patients in one study, anoutcome no better than chance (47).

3.1ci We recommend measuring serum sodium severaltimes during the first 514 days after transsphenoidal sur-gery. (1QQEE)

3.1cii We recommend assessing free T4 and prolactinwithin 12 weeks of surgery, to evaluate for overt hypop-ituitarism. (1QQEE)

3.1ciii We recommend obtaining a postoperative pitu-itaryMRI scan within 13months of successful TSS. (Un-graded best practice statement)

EvidenceAs with any TSS, potential complications include elec-

trolyte disturbances, hemorrhage, and meningitis (50).Hyponatremia occurs in 510% of patients, usually be-tweenpostoperativedays5and10 (51).This complicationismore commonafter extensive gland exploration inmen-struatingwomen. Diabetes insipidus is relatively commonin the first few postoperative days but is usually transient.We recommend measuring serum sodium several timesduring the first 514 days after surgery to address bothpossibilities, either daily or guided by the patients intakeandoutput. It is helpful to provide patientswith educationregarding post-transsphenoidal water balance disorders,including when to seek emergency care for the accompa-nying symptoms. It is advisable to provide the patientwithinformation about how to reach an endocrinologist in thiscircumstance because some emergency rooms are not fa-miliar with this cause of hyponatremia.

Because of the 5- to 7-day half-life of T4, a decrease offree or total T4 within 1 week of surgery may identifysignificant hypothyroidism (when compared to preoper-ative values). Acquired prolactin deficiency is amarker forhypopituitarism that may occur immediately after TSS(52). Hormonal deficits may be secondary to hypercorti-solism and also transient; therefore, we recommend re-evaluating the need for replacement therapy (see below).Permanent hypopituitarism ismore commonafter surgeryfor a microadenoma secreting ACTH than for those se-creting GH. This probably reflects a tendency to moreaggressive surgery but is not fully explained.

There are few data about the timing and need for post-operative imaging in patients with surgical remission, buttypical practice may include obtaining a postoperativescan 13 months after surgery to serve as a new baselinein case of future recurrence.

3.1d We recommend surgical resection of bilateral ad-renal disorders (1QQEE) and suggest medical therapy toblock aberrant hormone receptors for bilateral ma-cronodular adrenal hyperplasia (BMAH) (2QQEE).

Bilateral macronodular adrenal hyperplasiaThe terminology in this field is changing. In primary

BMAH, the disease eventually affects both adrenals, al-though itmay present initially as an asymmetric unilateralnodule (53).

Recent reports demonstrate an inherited mutation inthe armadillo repeat-containing 5 (ARMC5) gene in pa-



tients with BMAH and family members with unsuspectedBMAH (54, 55) (and in some with meningioma) (56).These findingsdonotprovidenew treatment(s).However,screening family members (with dexamethasone 1 mg) isindicated until genetic tests are available.

BMAH can be treated by an appropriate antagonist ifaberrant receptors are demonstrated that clearly couple tocortisol release, but the detailed workup of such patientsis a research undertaking beyond the scope of these guide-lines (57). Bilateral adrenalectomy, generally via the lapa-roscopic route, is the surgical treatment of choice, al-though some advocate for selective removal of the largeradrenal in older patients in whom nonfunctional tumorsare a possibility, especially if there is only a single noduleon each side (58).

In children, McCune-Albright syndrome may presentas BMAH, with a variable prognosis. It may be aggressiveand life threatening, with severe clinical features and hy-pertension (59), in which case bilateral adrenalectomy isindicated. Alternatively, research has reported spontane-ous remission after medical therapy to control hypercor-tisolemia (60). Therefore, in milder cases, clinicians mayconsider medical therapy.

Primary pigmented nodular adrenal diseaseLaparoscopic bilateral adrenalectomy is the definitive

treatment of choice and is curative in most cases of pri-marypigmentednodular adrenal disease, regardless of age(61, 62). However, in some pediatric patients, unilateraladrenalectomy leads to significant clinical and biochem-ical improvement (63), although studies have reported asubsequent relapse requiring the removal of the secondgland (58, 64). It is important to screen patients with sus-pected primary pigmented nodular adrenal disease at in-tervals for features of Carney complex, particularly foratrial myxoma (62). If clinicians detect Carney complex,they should also test family members.

EvidenceThe advantage of bilateral adrenalectomy is the swift

and definitive control of hypercortisolism; its disadvan-tages include the need for lifelong glucocorticoid andmin-eralocorticoid replacement therapy.

A review (65) of 23 studies reporting 739 patients un-dergoing bilateral adrenalectomy (about 70% laparo-scopic) showed a surgical morbidity of 18%and amedianmortality of 3%. In patients with CD, the surgical mor-tality was 1%. In 334% of cases, there was residualcortisol secretion due to accessory adrenal tissue or adre-nal remnants, but2%hada relapseofCS.Adrenal crisesoccurred in 9.3 cases per 100 patient-years. Overall mor-tality was 17% during a 41-month follow-up (range, 14

294mo).However, 46%of the deaths occurredwithin thefirst year after surgery, suggesting that careful manage-ment of cortisol-dependent comorbidities is needed. The10-year mortality after bilateral adrenalectomy for CD,bilateral adrenal hyperplasia, and EAS is 3, 10, and 44%,respectively (65).

4. Remission and recurrence after surgical tumorresection

4.1 We suggest an individualized management ap-proach based onwhether the postoperative serum cortisolvalues categorize the patients condition as hypocortiso-lism, hypercortisolism, or eucortisolism. (Ungraded bestpractice statement)

4.2 We recommend additional treatments in patientswith persistent overt hypercortisolism. (1QQQQ)

4.3 We recommend measuring late-night salivary orserum cortisol in patients with eucortisolism after TSS,including those cases where eucortisolismwas establishedby medical treatment before surgery. (1QQEE)

4.4 We recommend using tests to screen for hypercor-tisolism to assess for recurrence in patients with ACTH-dependent CS. (1QQQE)

Evidence

Postoperative initial remissionThere is no consensus on the criteria for remission after

resecting an ACTH-producing tumor. (In CD, because ofthe significant recurrence rate, the term remission ispreferable to cure.) Normal corticotropes are sup-pressed by sustained hypercortisolism; therefore, ACTHand cortisol levels are low after resecting the ACTH-pro-ducing tumor. Remission is generally defined as morningserum cortisol values 5g/dL (138 nmol/L) or UFC2856 nmol/d (1020 g/d) within 7 days of selectivetumor resection. Although clinicians have used glucocor-ticoid dependence to characterize surgical response, weadvocate measuring serum cortisol levels as a preferableand quantifiable alternative.

Adult series report a remission rate of 7376% for se-lectively resected microadenomas, but a lower remissionrate for macroadenomas (43%) (6668).

In two pediatric series using a strict criterion for remis-sion (post-TSS serum cortisol of 1 g/dL [28 nmol/L][41] or 1.8 g/dL [50 nmol/L] [40]), remission rateswere 100 and 69%, respectively. Follow-up data sug-gest that hypercortisolemia recurrence was uncommon(40, 70). Other series reported remission rates of 7098% (71, 72).

A recent report of 200 cases of pediatric CD (72) as-sociated initial postoperative remission with adenomaidentification at surgery. The study also associated long-



term remission with younger age, smaller adenoma, andmorning serum cortisol of 1g/dL after surgery (72). Inone study of adults, the initial remission rate for microad-enomas was greater than for macroadenomas (72.8 vs42.9%) (66). Technical aspects, surgeon experience, tu-mor size, and the lack of dural invasion likely contributemost to successful outcome in patients of any age (213).

Patients with mild or cyclic CS and those rendered eu-cortisolemic bymedical treatment before surgery may nothave suppressed corticotropes. Their postoperative UFCand morning cortisol may be normal. In these patients,clinicians must measure late-night serum or salivary cor-tisol. If there is a normal diurnal rhythm, ie, an appropri-ately low late-night serum or salivary cortisol level, then itis likely that the patient is in remission. Conversely, thelack of a diurnal rhythm suggests persistent disease. Pa-tients who are eucortisolemic after resection, especiallythose with moderate preoperative hypercortisolism, mayhave residual tumors and are more prone to recurrencethan patients with prolonged postoperative hypocortiso-lism (67).

After bilateral adrenalectomy, patients have undetect-able serum cortisol values, and morning plasma ACTHlevels are often between 200 and 500 pg/mL. Morningcortisol values are generally1.8g/dL (50 nmol/L) afterresecting an adrenal adenoma.

RecurrenceRecurrence rates of hypercortisolemia in pediatric CD

are very low (40, 70), especially when postoperativeplasma ACTH and cortisol levels are undetectable. Re-currence is a more significant problem in adults. In onestudy, after initial remission, 23%of adultswithmicroad-enomas and 33% with macroadenomas had recurrence.Recurrence rates vary from 1566%within 510 years ofinitially successful surgery (68, 73, 74).

Clinicians should evaluate patients for possible CD re-currence when the HPA axis recovers, and then annually,or sooner if they have clinical symptoms. Early recovery(within 6 mo) of HPA axis function may indicate an in-creased risk of recurrence (66). Two longitudinal studiesdemonstrated that elevated late-night serum/salivary cor-tisol is one of the earliest biochemically detectable signs ofrecurrence and almost always precedes elevated urine cor-tisol (75, 76). Although many such patients evolve to sig-nificant hypercortisolism, it is uncertain whether there isany benefit to treating patients with a mild early recur-rence if they are asymptomatic.

Although studies have reported that a number of tests(eg, CRH, desmopressin) predict recurrence risk, diagnos-tic accuracy is not high enough to provide certainty re-garding long-term outcomes. Thus, clinicians must mon-

itor every patient in remission from CD for the possibilityof recurrence (73, 77).

Rarely, residual adrenocortical tissue (usually in thesurgical bed, but sometimes in the gonads) regrows as aresult of prolonged ACTH hypersecretion. If patients be-come cushingoid after bilateral adrenalectomy, cliniciansshould test endogenous cortisol secretory capacity bywithholding glucocorticoid for 24 hours and checking se-rum/salivary cortisol levels.

Recurrence after ectopic ACTH-secreting tumor resec-tion generally reflects metastasis.

5. Glucocorticoid replacement and discontinuation,and resolution of other hormonal deficiencies

5.1 We recommend that hypocortisolemic patients re-ceive glucocorticoid replacement and education about ad-renal insufficiency after surgical remission. (1QQQQ)

5.2We recommend follow-up morning cortisol and/orACTH stimulation tests or insulin-induced hypoglycemiato assess the recovery of the HPA axis in patients with atleast one intact adrenal gland, assuming there are no con-traindications. We also recommend discontinuing gluco-corticoid when the response to these test(s) is normal.(1QQQE)

5.3 We recommend re-evaluating the need for treat-ment of other pituitary hormone deficiencies in the post-operative period. (1QQQE)

EvidenceAfter successful surgery, glucocorticoid replacement is re-

quired until the HPA axis recovers, which in adults occursabout 612 months after resecting ACTH-producing tu-mors and about 18 months after unilateral adrenalectomy(78, 79). In one large pediatric series, recovery occurred at amean of 12.6 3.3 months after surgery (80). (Obviously,bilateral adrenalectomy results in a need for lifelong gluco-corticoid and mineralocorticoid replacement.)

Despite the use of physiological glucocorticoid replace-ment, many patients suffer from glucocorticoid with-drawal. Patients should be warned that this is commonand expected (81). Symptoms include anorexia; nausea;weight loss; and other nonspecific symptoms such as fa-tigue, flu-like myalgias and arthralgias, lethargy, and skindesquamation. Accordingly, patients usually feel worsewithin a fewdays orweeks after successful surgery. Adultsmay experience persistent or new-onset atypical depres-sive disorders, anxiety, or panic symptoms (82). Recoveryfrom the glucocorticoid withdrawal syndromemay take 1year or longer.

The syndrome may persist even after the HPA axis hasrecovered and may even occur in patients who do notdevelop secondary adrenal insufficiency after TSS for CD



(83). The pathophysiology of the steroid withdrawal syn-drome is not known. Patients may improve with a tem-porary increase in the glucocorticoid dose, but it is im-portant to reduce the dose as soon as possible to avoidiatrogenic CS. Administering serotonin-specific reuptakeinhibitors may help, but this has not been systematicallystudied. Generally, themost important intervention is fre-quent support and reassurance by the medical team. Fam-ily, friends, and patient support groups also may behelpful.

We recommend glucocorticoid replacement with hy-drocortisone, 1012 mg/m2/d in divided doses, eithertwice or thrice daily, with the first dose taken as soon aspossible after waking (84). Although this dose is some-what higher than recently reported cortisol productionrates (85), it works well clinically, probably because ofinterindividual differences in hepatic and adipose metab-olism and glucocorticoid receptor polymorphisms. Hy-drocortisone is preferred because more potent syntheticglucocorticoids with a longer half-life may prolong HPAaxis suppression.

When hydrocortisone is not available, or patientsrequest once-daily dosing, clinicians can administerother glucocorticoids at the lowest possible replace-ment dose. Written instructions about stress dosing forintercurrent illnesses, injectable emergency steroids,and the need to obtain and wear a medical alert tagindicating adrenal insufficiency/glucocorticoid replace-ment are essential (84).

Although some practitioners prescribe supraphysi-ological doses (eg, hydrocortisone 20 mg two to threetimes daily) in the immediate postoperative period, thereare no controlled studies that address whether this (or aslower taper) minimizes the glucocorticoid withdrawalsyndrome.Other clinicians use only physiological replace-ment doses to avoid continued excessive glucocorticoidexposure.

There are a variety of tapering and discontinuationstrategies, none of which has been systematically studied;the following are general comments. Some centers reducethe hydrocortisone dose as weight decreases and discon-tinue abruptly when the HPA axis is recovered; otherstaper the dose at fixed intervals. Clinicians can assessHPAaxis recovery with a morning cortisol level obtained (be-fore that days glucocorticoid dose) every 3 months, fol-lowedbyanACTHstimulation test startingwhen the levelis 7.4g/dL (200 nmol/L) ormore. The axis has recoveredif the baseline or stimulated level is approximately 18g/dL (500nmol/L)or greater. Patientswith cortisol levelsbelow 5 g/dL (138 nmol/L) should remain on glucocor-ticoids until retested in 36 months. Stimulation testingmay be helpful with intermediate values. However, vari-

ous studies propose different cutoffs, and assays differ, soclinical judgment should be used. It is rare that the HPAaxis does not eventually recover.

Any etiology of hypercortisolism can cause preoper-ative functional central hypothyroidism and central hy-pogonadism. Although these may resolve after 6 post-operative months (86), patients may need continuedreplacement therapy. Clinicians should repeat testing toestablish when and if the patient has recovered. GH defi-ciency, frequently present during hypercortisolism, per-sisted in over 50% of children during the first 12 monthsafter cure (87) and to a lesser extent into adult life (88). Itmight not be possible for patients to attain normal lineargrowth after surgically cured CD, with patients often notreaching their genetic target (89). For this reason, somecenters test GH stimulation at 3 months after surgery andinitiate human GH replacement, combined with GnRHagonist therapy, in pubertal subjects to maximize growthpotential in children with abnormal responses (16). Theuse of GH therapy in adults after remission should followpreviously published guidelines (90). Because hypercorti-solemia can affect the GH axis, some experts advise wait-ing at least 12months after remission to test for deficiencyof this hormone in adults (91).

6. Second-line therapeutic options6.1 In patients with ACTH-dependent Cushings syn-

dromewhounderwent a noncurative surgery or forwhomsurgery was not possible, we suggest a shared decision-making approach because there are several available sec-ond-line therapies (eg, repeat transsphenoidal surgery, ra-diotherapy, medical therapy, or bilateral adrenalectomy).(2QQEE)

EvidenceWhen surgery is not possible or is noncurative, the

choice of second-line therapy must take into account pa-tient preferences; treatment goals; biochemical control;the size and location of residual tumors; the urgency totreat; other medications (drug-drug interactions); the pa-tients personal history; the method of delivery, side ef-fects, and cost of medication; gender; age; and the avail-ability of medical therapies.

After unsuccessful transsphenoidal surgery, if cortisollevels remain consistently elevated, the prompt titration ofmedical therapies (or bilateral adrenalectomy) is needed.Even if cortisol levels are normal, careful observation isneeded because cortisol levels may fall over subsequentweeks (67).

6.1a We suggest bilateral adrenalectomy for occult ormetastatic EAS or as a life-preserving emergency treat-



ment in patients with very severe ACTH-dependent dis-ease who cannot be promptly controlled by medical ther-apy. (2QQQE)

6.1b We recommend regularly evaluating for cortico-trope tumor progression using pituitaryMRIs and ACTHlevels in patients with known CD who undergo bilateraladrenalectomy and in patients who undergo this proce-dure for presumed occult EAS (because some of the latterhave a pituitary and not ectopic tumor). (1QQQE)

EvidenceMedical therapy may be an initial approach in any pa-

tient with ACTH-dependent CS who has failed surgery,who has persistent or metastatic disease, or who has anoccult tumor. It is also useful in severely ill patients (92).

InpatientswithCD, theprimary corticotrophadenomaremains in situ after adrenalectomy. Therefore, both chil-dren and adults have a significant risk of developing mac-roscopic (1 cm) enlargement of the tumor and hyper-pigmentationthe so-called Nelsons syndrome.

After bilateral adrenalectomy for CD, 21% of adultsdeveloped Nelsons syndrome (65), whereas corticotrophtumor progressionwithout all of the features described byNelson was observed on MRI in 50% of cases (93). Nel-sons syndrome was less likely in patients without visibletumor at the time of adrenalectomy (93). It is not knownconclusively whether RT mitigates this risk.

The risk of corticotroph tumor progression is associ-ated with increased plasma ACTH, but other predictorsare not understood. Thus, lifelong follow-up is importantand should include clinical examinations for hyperpig-mentation, ACTH measurements, and MRI scans. Theseshould be performed initially and at annual intervals witha decrease in frequency based on previous findings.

Patientswith presumedoccult EASmay in fact haveCDand thus should be regularly examined for the emergenceof a pituitary tumor (94).

6.2 Repeat transsphenoidal surgery6.2 We suggest repeat transsphenoidal surgery, partic-

ularly in patientswith evidence of incomplete resection, ora pituitary lesion on imaging. (2QQEE)

In 12 of 17 patients with persistent hypercortisolismafter TSS, early repeat transsphenoidal surgery inducedhypocortisolism. All patients had tumors that were foundduring the initial procedure (95). Repeat surgery for re-current hypercortisolism led to hypocortisolism in 22 of31 patients with previous surgical remission. Evidence ofprevious incomplete resection, the presence of a pituitarylesionon imaging, and intraoperative tumordetectionpre-dicted remission (96). Repeat surgery carries an increasedrisk of hypopituitarism compared to initial surgery. Al-

though remission is less likely than after the first surgery,it can be achieved rapidly compared to someother second-line treatments and is important to consider, particularlywhen there is access to an expert pituitary surgeon.

6.3 Radiation therapy/radiosurgery for Cushingsdisease

6.3We recommend confirming that medical therapy iseffective in normalizing cortisol before administering RT/radiosurgery for this goal because thiswill be neededwhileawaiting the effect of radiation. (1QEEE)

6.3aWe suggest RT/radiosurgery in patients who havefailed TSS or have recurrent CD. (2QQEE)

6.3b We recommend using RT where there are con-cerns about the mass effects or invasion associated withcorticotroph adenomas. (1QQQE)

6.3c We recommend measuring serum cortisol or UFCoff-medication at 6- to 12-month intervals to assess theeffect of RT and also if patients develop new adrenal in-sufficiency symptoms while on stable medical therapy.(1QQQE)

EvidenceAlthough pituitary radiation may serve as a primary

treatment for CD in individuals who cannot undergo sur-gery or when the tumor is invasive or unresectable, it ismost often used as a second-line treatment when surgeryfails. The effects of radiation occur over months to years,so it is important tomedicallymanage cortisol excess untilcortisol is controlled off-medication. If pituitary RT isplanned, one should first demonstrate that adequate bio-chemical control is possible by medical means to ensurecontrol while waiting for radiation effects to occur. RT ofsome type may be indicated for an invasive, expansive,atypical corticotroph tumor, in which case it is possiblethat cortisol may be controlled in a different way (eg, ad-renalectomy) or may not be elevated.

Several different forms of radiation are available. Thetermconventional radiation typically refers to fraction-ated photon beamRT.Many centers traditionally admin-istered RT from a linear accelerator, via a three-field tech-nique (two lateral, one frontal) to deliver a total dose of 45Gy over 6 weeks (97). However, intensity-modulated RTis increasingly used tomodulate the dose to accommodatetumor contours and spare nearby critical structures.

Conventional RT results in remission in up to 83% ofadult patients, from 660 months after treatment, butoftenwithin 2 years. In children, conventionalRTwithoutadjunctive medical treatment was effective in less timethan in adults (98). In pediatric series, 18 of 23 (78%)children were cured in 918 months after RT alone (99,100).



Stereotactic radiation uses a frame to position the pa-tient accurately and includes computer-assisted planningcombined with MRI to deliver radiation to the tumorthrough many ports so as to minimize radiation to sur-rounding structures. Using single-dose stereotactic radia-tion is often termed radiosurgery; severalmillimeters ofclearance between the tumor and the neurovisual appa-ratus are required to avoid damaging the optic nerves.Gamma knife, linear accelerator, and proton beam areadministered in this way; there are no direct comparisonsof effectiveness and safety, but forCD, results appear to besimilar. There are no reported series of stereotactic radio-surgery in pediatric CD.

Radiosurgery increases patient convenience by allow-ing a single treatment day, rather than 6 weeks of therapy(101). It may provide more rapid biochemical control ofcortisol excess than conventional radiation and confer lessrisk of radiation damage to surrounding brain structures,but these anecdotal findings have not been definitivelyestablished.

Some experts recommend radiosurgery only whenthere is a clear target on MRI, bearing in mind that thismay not be an ACTH-secreting tumor, whereas othersadvise using radiosurgery in all patients with adequateclearance of the neurovisual apparatus, targeting thewhole sella and a few millimeters beyond if no lesion isvisible.

Reports of radiation effectiveness in CD include vari-ous methods, definitions of biochemical and tumor con-trol, and lengths of follow-up, making comparison acrossstudies difficult.Amongpublicationsover the last 15yearswith at least 20 subjects, cortisol excesswas biochemicallycontrolled in2886%ofpatients (102104).Recurrencesmay develop, so patients need long-term monitoring. Ithas been suggested that recurrence may be more frequentafter radiosurgery than after conventional RT, but thispoint is not firmly established. Tumor size was better con-trolled (83100% success rate) than cortisol (102104);this treatment option is particularly valuable for largetumors.

A normal diurnal rhythm is not necessarily achievedafter RT, so increased late-night cortisol levels should notbe a criterion for remission. The increased nocturnal valuemay represent the persistence of mild residual hypercor-tisolism in patients with normalized UFC, but this possi-bility has not been investigated.

Clinicians use medications to normalize cortisol untilradiation takes effect. We recommend assessing serumcortisol orUFCoff-medication at 6- to 12-month intervalsand if patients develop new adrenal insufficiency symp-toms while on stable medical therapy.

Hypopituitarism is a riskwithall formsof radiation.Upto two-thirds of patients develop anterior pituitary hor-mone deficiency after RT (101, 105). All patients shouldundergoacareful assessmentof anteriorpituitary functionpost-therapy annually (at least) or sooner if hormone de-ficiency symptoms develop. Additional radiation risks in-clude optic neuropathy (12%) and other cranial neurop-athies (24%) as well as a small risk of secondaryneoplasia within the radiation field (most commonly me-ningiomas) (101).

In children, GH deficiency and hypogonadism werecommonwithin 1 year of RT; thyroid hormone deficiencywas not reported (106). Administering humanGHforGHdeficiency is indicated if linear growth potential exists. Areassessment of GH secretion at growth completion is ad-vised because only one of six patients in one study showedunequivocally normal GH secretion (88). Changes in cog-nitive functionhavenotbeen studiedafterRT forpediatricor adult CD.

ValuesPatients should consider the cost, accessibility, andcon-

venience of radiosurgery vs RT when choosing betweenthe two.

6.4 Medical treatment6.4 We recommend steroidogenesis inhibitors under

the following conditions: as second-line treatment afterTSS in patients with CD, either with or without RT/ra-diosurgery; as primary treatment of EAS in patients withoccult or metastatic EAS; and as adjunctive treatment toreduce cortisol levels in ACC. (1QQQE)

6.4a We suggest pituitary-directed medical treatmentsin patients with CD who are not surgical candidates orwho have persistent disease after TSS. (2QQQE)

6.4bWe suggest administering a glucocorticoid antag-onist in patients with diabetes or glucose intolerance whoare not surgical candidates or who have persistent diseaseafter TSS. (2QQQE)

6.4c We suggest targeted therapies to treat ectopicACTH syndrome. (2QEEE)

ValuesThe choice of medical therapy should be guided by ef-

ficacy, individual patient factors, and cost. The goal isclinical normalization using cortisol levels as a proxy end-point (except for mifepristone, see below). This can beachieved either with a block and replace strategy inwhich circulating cortisol is reduced to minimally detect-able levels and glucocorticoid replacement is added (avoid-ing supraphysiological doses) (107, 108) orwith a normal-ization strategy aimed to achieve eucortisolism (109, 110).



If there is evidence of significant cyclicity, block and replacemaybepreferable,but itcarriesadditional risk ifhigherdosesand multiple medications are needed.

RemarksHypoadrenalism may occur when treating with mife-

pristone or any of the steroidogenesis inhibitors, due toovertreatment, the inability to mount a cortisol responseto intercurrent infection, or cyclical or variable hypercor-tisolism. The gastrointestinal symptoms of hypoadrenal-ism overlap in many cases with side effects of the drugs.Thus, the possibility of adrenal insufficiency must be ad-dressed (see below) (Table 1).

Monitoring includes assessing clinical response and thebiochemical evaluation (24-h UFC, morning serum corti-

sol, or serum cortisol day curves [except for mifepristone;see below]) to evaluate for hypercortisolism control. As-sessing adrenal insufficiency in patients on medical treat-ment is mainly done clinically; dose interruption or reduc-tion should be considered when adrenal insufficiency issuspected.

Except as noted, none of the medical treatments dis-cussed have U.S. Food and Drug Administration (FDA)approval for the treatment of hypercortisolism.

Many of these agents stimulate or inhibit CYP3A4,which may lead to significant drug-drug interactions(111).Thus, reviewof all othermedications is importantwhen initiating therapy and when adding other medi-cations. These agents also may increase the QT interval(Table 1).

Table 1. Medical Treatment of CS

Drug Pros Cons Dosea

Steroidogenesisinhibitors

Ketoconazoleb Quick onset of action Adverse effects: GI, hepatic dyscrasia (death),male hypogonadism; requires acid forbiological activity; DDIs

4001600 mg/d;every 68 hdosing

Metyraponeb Quick onset of action Adverse effects: GI, hirsutism, HT,hypokalemia; accessibility variable acrosscountries

500 mg/d to 6 g/d;every 68 hdosing

Mitotanec Adrenolytic, approved for adrenalcancer

Slow onset of action; lipophilic/long half-life,teratogenic; adverse effects: GI, CNS,gynecomastia, low WBC and T4,1 LFTs;1 CBG, DDIs

Starting dose, 250mg; 500 mg/dto 8 g/d

Etomidate Intravenous, quick onset of action Requires monitoring in ICU Bolus and titratePituitary-directedCabergoline Adverse effects: asthenia, GI, dizziness 17 mg/wkPasireotided Most successful when UFC 2-fold normal;

sc administration; adverse effects: diarrhea,nausea, cholelithiasis, hyperglycemia,transient1 LFTs;1QTc

600900 g twicedaily

Glucocorticoidreceptor-directed

Mifepristonee Difficult to titrate (no biomarker);abortifacient; adverse effects: fatigue,nausea, vomiting, arthralgias, headache,hypertension, hypokalemia, edema,endometrial thickening

3001200 mg/d

Abbreviations: GI, gastrointestinal; DDI, drug-drug interactions; HT, hypertension; CNS, central nervous system; WBC, white blood cell count; LFTs,liver function tests; CBG, corticosteroid binding globulin; ICU, intensive care unit; QTc, corrected QT interval.a Except as noted, the lowest dose may be used initially, unless the patient has severe hypercortisolism (UFC more than five times normal), in whichcase the starting dose may be doubled.b Ketoconazole and metyrapone are approved by the European Medicines Agency for the treatment of CS.c Mitotane has FDA approval for treatment of adrenal cancer.d Pasireotide has FDA approval for treatment of patients with CD who are not surgical candidates or have failed surgery. The agent is alsoapproved in Europe.e Mifepristone has FDA approval for treatment of patients with CS and diabetes or glucose intolerance who are not surgical candidates or havefailed surgery.

The use of other agents listed here represents an off-label use for the treatment of CS.



Evidence

KetoconazoleKetoconazole, an imidazole derivative with antifungal

activity, impairs adrenal and gonadal steroidogenesis byinhibiting side-chain cleavage, 17,20-lyase, and 11- hy-droxylase enzymes (112).

Taking all case series together, ketoconazole mono-therapy (at daily doses of 4001200mg) normalized UFCin 57 of 82 patients with presumed CD (2593% rate inindividual studies); normalization was not dependent onthe dose or duration of treatment (113120). Recent dataconfirm these findings with a greater than 50% drop inUFC in 75% of 200 patients, and with clinical improve-ments in diabetes, hypertension, and hypokalemia (121).Efficacy in the ectopic ACTH syndrome is lower: of ninepatients, only four (44%) achieved eucortisolism (113115, 122).

Ketoconazoles side-effect profile (Table 1) is relativelybenign, except for idiosyncratic severe hepatic dyscrasia,which is estimated to occur in one in 15 000 exposed in-dividuals (115, 122124). The FDA issued a black boxwarning for this in 2013, and the European MedicinesAgency has restricted access to the agent to physiciansspecialized in treating CS (125). Among 33 cases of po-tential ketoconazole-induced liver injury submitted to theFDA from the time of initial marketing in 1980, 18 pa-tients had an 8-fold elevation in transaminases, five hadcholestatic injury, andninehadamixtureof the two (126).Thus, monitoring liver function is necessary.Mild asymp-tomatic elevation in serum transaminases occurs in ap-proximately 1015% of cases (121, 126), usually whentherapy starts or when doses increase, so closemonitoringis needed at these times. Values typically return to normalwithin 24 weeks after stopping therapy or reducingdoses. If liver enzyme elevations remain less than threetimes the upper limit of normal, most clinicians will con-tinue therapy. However, when enzyme levels are higher,discontinuation or dose reduction is advised.

MetyraponeMetyrapone inhibits 11- hydroxylase, which cata-

lyzes the conversion of 11-deoxycortisol to cortisol. Its2-hour half-life necessitates three to four doses daily. Phe-nytoin and phenobarbital accelerate metyrapone metab-olism, and estrogens reduce it; metyrapone is excreted inbreast milk.

Metyrapone controls hypercortisolemia in 5075% ofpatients with CS. In the largest published single centerseries (91 patients), chronic therapy controlled hypercor-tisolemia inCDdespite a rise in serumACTH(127). Thesefindings were confirmed recently in 195 patients (128).Although nomedication for CS is approved for use during

pregnancy, metyrapone has been given occasionally inpregnantwomenwithCSwith no apparent adverse effectsto mother or offspring (129132).

Adverse effects of metyrapone are most commonwhentherapy starts or doses increase, and theymainly consist ofgastrointestinal disturbances (in the absence of hypo-adrenalism). However, this reaction is uncommon whenthe medication is taken with food or milk. With chronictherapy, hirsutism and acne may worsen due to the ac-cumulation of androgenic precursors secondary to theblockade of cortisol synthesis; the accumulation of min-eralocorticoid precursors requires monitoring for hypo-kalemia, edema, and hypertension.

RemarksMetyrapone causes a gross elevation of circulating lev-

els of the cortisol precursor 11-deoxycortisol, which cancross-react in many immunoassays for serum and urinarycortisol. This artificially elevates apparent cortisol values,potentially masking biochemical hypoadrenalism. Liquidchromatography-tandem mass spectrometry assays forcortisol circumvent this issue (133). With no cross-reac-tivity issues, the treatment target is either UFC in the nor-mal range or mean serum cortisol levels that are between5.4 and10.8g/dL (150300nmol/L) throughout the day(134).

Combination therapyMetyrapone and ketoconazolemay be combined to en-

hance the control of severe hypercortisolemia (92, 135).

MitotaneMitotane is primarily used to treat adrenal carcinoma.

It inhibits CYP11A1 (P450 side-chain cleavage) and has adirect cytotoxic action on the adrenal cortex (136). Mi-totane has a long half-life and sustained effects because ofits storage in adipose tissue. Hence, the dose may be in-creased at weekly intervals, and UFC normalization takesalmost 6 months (107). The dose and mitotane plasmalevels required to control hypercortisolism in CD werelower than those needed to achieve ACC antineoplasticactivity,withmedians around2.7 gofmitotane equivalentper day and 8.5 mg/L, respectively (107).

Mitotane as monotherapy does not cure CD. It is aneffective adjunctive therapy in patients with CD as a first-or second-line treatment (after unsuccessful TSS) whileawaiting the effects of pituitary RT or when surgery is notpossible. In three studies of patients with CD receivingmitotane for these indications, 7282%had sustained hy-percortisolism remission (107, 108, 110).

Because mitotane increases the production of cortisol-binding globulin, plasma total cortisol levels increase pro-



portionally (137). Biochemicalmonitoring therefore reliesonUFCor salivary cortisolmeasurements. Inpatientswhodevelop adrenal insufficiency, the usual hydrocortisonereplacement dose should be increased because mitotanestrongly activates CYP3A4 and increases hydrocortisoneclearance (138). A safe approach is to increase the initialhydrocortisone daily dose by one-third. Dexamethasone,a strongCYP3A4 inducer, should be avoided (139). At thedoses used forCD,mitotane has less adrenolytic effects onthe zona glomerulosa; mineralocorticoid replacementmay not be required (110).

Side effects lead to discontinuation of the drug in up to28%of patients (107).Mitotane is a teratogen; pregnancyshould be avoided for years after stopping the drug be-cause measurable plasma levels may persist for months.Measuring plasma levels may help guide this decision. Po-tential drug interactions are numerous (140) (Table 1).

Glucocorticoid receptor antagonistIn one study, Mifepristone, a glucocorticoid receptor

antagonist and antiprogestin, led to an improvement inhypertension and/or diabetes in 40 and 60%, respectively,of 34 patients (141). At least one other clinical parameter(weight, depression, cognition, clinical appearance, orQOL) improved in 87%. Mifepristone is approved in theUnited States for the control of diabetes or glucose intol-erance secondary to hypercortisolism in patients whofailed surgery or are not surgical candidates.MeanACTHlevels increased by more than 2-fold in 31 of 43 patientswith CD followed for a median of 11.3 months. Threemacroadenomas increased in size, whereas one regressed(142).

Cortisol levels remain unchanged or may increase dur-ing mifepristone treatment, and therefore practitionerscannot use hormonalmeasurements to guide efficacy or todiagnoseadrenal insufficiency.Becausepractitionersmustuse clinical cortisol-dependent variables for these pur-poses, it is difficult to estimate the correct dose. For thisreason, clinicians should start mifepristone at 300 mg/d,titrate it slowly, and base dose adjustment on clinical pa-rameters, primarily glucose, and weight reduction. Ad-verse events include symptoms of cortisol insufficiency(fatigue, nausea, vomiting, arthralgias, and headache), ev-idence of increased mineralocorticoid action (hyperten-sion, hypokalemia, edema), and antiprogestin effects (en-dometrial thickening) (Table 1). One study treatedsuspected adrenal insufficiencywith drug discontinuationor 28 mg of dexamethasone daily (141).

EtomidateEtomidate is the only medical treatment available for

severe hypercortisolism in seriously ill patients of any age

who are not immediate surgical candidates and who can-not take oralmedications. It is also useful in an emergencysetting with acute unmanageable symptoms such as respi-ratory failure or severe psychosis (143, 144) and can be aneffective bridge to other medical or surgical therapies(145).Etomidate is an imidazolederivative (likeketocona-zole) that is often used for anesthesia induction. Subhyp-notic doses rapidly decrease steroidogenesis within 1224hours by inhibiting 11-hydroxylase and cholesterol side-chain cleavage (146). Due to the need for iv infusion, thesepatients should be managed in an intensive care unit. Itmaybeprudent to administer etomidatepreparations con-taining propylene glycol (which may cause thrombophle-bitis and pain on injection) through a central venous line.Measuring cortisol levels every 46 hours is required, andclinicians can titrate the infusion rate to achieve a stableserum cortisol level between 10 and 20 g/dL (280560nmol/L) or they can use a block and replace strategy. Aloading dose of 35 mg is followed by a continuous infu-sion of 0.030.10 mg/kg/h (2.53.0 mg/h). Studies havenot reported sedation at these doses; however, patientsmay need a dose reduction if renal failure occurs, due to aresulting increase in free etomidate concentrations.

Medical pituitary-directed treatmentsCabergoline and pasireotide act directly on cortico-

troph tumors to inhibit ACTH production. They are gen-erally not effective in adrenal causes of CS, and their rolein the treatment of ectopic ACTH production remains tobe determined.

CabergolineCabergoline is a dopamine agonist with high affinity

for the dopamine receptor subtype 2, which is expressedby most corticotroph adenomas (147, 148).

In small studies, 3040% of patients responded andcontinued to have normal UFC levels after 23 years ofcabergoline treatment (149, 150). However, the cortisol-lowering effect did not last in 29% of initial responders.The serum prolactin concentration did not predict long-termUFCresponse.Thedoses patients receivedwereup to7mg orally per week, with a median dose of 3.5 mg/wk inone study and a mean dose of 2.1 mg/wk in the other,which is higher than typical doses for hyperprolactinemia(149, 150).

Systolic and diastolic blood pressure, fasting glucose,and insulin improved. Tumor volume decreased or re-mained stable in the small number of reported patientswith visible adenomas (149, 150).

Side effects were typical of dopamine agonist use, suchas nausea, dizziness, and asthenia (Table 1), which werenot reported as adrenal insufficiency (150). In one study,



a single patient progressed from mild to moderate tricus-pid regurgitationon echocardiogramafter 2 years of treat-ment; in the other, no patients who had echocardiogramsshowed any significant valvulopathy (149, 150).

Cabergoline has been combined with other medica-tions to treat CD (151, 152). In 12 patients with persistenthypercortisolism after TSS, cabergoline monotherapy (upto3mg/wk)normalizedUFC in threepatients at 6months.Adding ketoconazole (up to 200 mg twice daily) normal-ized UFC in six of the nine nonresponders (152). In an-other study, adding cabergoline (up to 6 mg/wk) to pasir-eotide normalized UFC in four patients who were notcontrolled on pasireotide alone (151). We need larger tri-als to establish the role of combination therapies.

PasireotidePasireotide is a somatostatin receptor (SST) agonist

that binds to four of the five SST subtypes with substan-tially higher affinity for SST1 and SST5 than octreotide orlanreotide (153, 154). Corticotroph tumors have a highexpression of SST5, and pasireotide decreased ACTH se-cretion and cell proliferation in cultured human cortico-troph tumors (49, 155).

A phase 3 trial administered pasireotide 600 or 900 gsc twice daily in 162 CD patients who had failed (or werenot candidates for) surgery and had a mean baseline UFClevel at least 1.5-fold above normal (156). After 6months,20% of the subjects attained a normal UFC. Systolic anddiastolic blood pressure, triglycerides, low-density lipo-protein cholesterol, weight, and HRQOL improved.About 90%of patientswhowere not controlled bymonth1 or 2 remained uncontrolled at 6 and/or 12months; thus,a brief trialmay predict the chance of biochemical control.

Mean tumor volume decreased by 44% in 75 patientswith a lesion onMRI at the 900-g dose (156). Most sideeffects were similar to those of other somatostatin analogs(predominantly gastrointestinal, including biliary sludgeand gallstones) except for the important finding of hyper-glycemia (73% of patients) (Table 1). Glucose and gly-cated hemoglobin increased soon after drug initiation inmost patients, regardless of whether UFC was controlled;no patient developed diabetic ketoacidosis or hyperosmo-lar coma (156). Pasireotide was approved in 2012 in theEuropeanUnion and theUnited States for the treatment ofCDwhen surgery is not successful or cannotbeperformed.

Clinicians should correct hypokalemia and hypomag-nesemia before initiating pasireotide. And they should ad-minister tests forbaseline liver function tests, thyroid func-tion (including free thyroid hormone), IGF-1, fastingglucose/glycated hemoglobin, as well as gallbladder ultra-sounds and electrocardiograms for corrected QT intervalprolongation or bradycardia. Clinicians should evaluate

changes in these parameters on pasireotide (hyperglyce-mia, prolonged quality corrected QT interval thyroid ab-normalities, gallstones, and GH deficiency) based on clin-ical symptoms and signs. At a minimum, they shouldmonitor these at 34months after initiating treatment andafter any dose increase. Because of hyperglycemia, clini-cians should monitor postprandial glucose and also rec-ommend that patients take the drug after (not before)meals and follow dietary recommendations for diabetes.

Targeted therapies for ectopic ACTH syndromeACTH-secreting tumors may express functional SST2

and Dopamine 2 receptors (157, 158). Drugs targetingthese receptors may reduce ACTH secretion and, conse-quently, control hypercortisolism. Several case reportsshow that octreotide, a potent SST2 agonist, may controlACTH and cortisol secretion for a short- to midterm pe-riod in patients with recurrent or unresectable ectopicACTH-secreting tumors (159, 160). However, octreotidetreatment usually had little or no effect on tumor growth.

Studies have occasionally reported hormonal controlwith the dopamine-2 agonist receptor cabergoline givenalone or in combination with an SST2 agonist (150, 161).

In three reports, the tyrosine kinase inhibitors vandet-anib and sorafenib rapidly and fully controlled hypercor-tisolism caused by ACTH secretion from metastatic med-ullary thyroid carcinomas (162164). The dissociationbetween the decrease in ACTH secretion and the lack oftumor reduction suggests a direct antisecretory effect.

7. Approach for long-term follow-up7.1 We recommend treating the specific comorbidities

associated with CS (eg, cardiovascular risk factors, osteo-porosis and psychiatric symptoms) in all patients with CSthroughout their lives until resolution. We also recom-mend testing for recurrence throughout life, except in pa-tients who underwent resection of an adrenal adenomawith a computerized tomography (CT) density of 10Hounsfield units. (1QQQE)

7.2 We recommend educating patients and familiesabout the clinical features of remission. (Ungraded bestpractice statement)

7.3 In patients with adrenal adenoma, we suggest fol-low-up tests for the specific comorbidities associated withCS if the adenoma density on CT was 10 Hounsfieldunits. (2QQEE) For those with higher Hounsfield unitvalues or pathology consistent with possible carcinoma,we suggest evaluating for malignancy using imaging.(2QEEE)

7.4We recommend that patients with Carney complexhave lifelong follow-up tests for cardiac myxoma and



other associated disease (testicular tumors, acromegaly,thyroid lesions). (1QQQQ)

Evidence

Monitoring and treating Cushings features andcomorbidities

With CS, biochemical remission or a cure is usuallyassociated with significant clinical improvement, butsome comorbidities may not completely normalize. Eval-uating and treating the long-term negative effects ofchronic hypercortisolism may therefore be important toreduce morbidity, improve QOL, and reduce the long-term excess mortality associated with CS.

After surgical or medical remission of CS, a significantproportion of patients will either develop or have a recur-rence of autoimmune or inflammatory diseases, such ashypothyroidism, psoriasis, celiac disease, ulcerative coli-tis, Crohns disease, asthma, lupus, and rheumatoid ar-thritis after surgical or medical remission (165). There isno evidence that surgical or medical remission of CS in-creases the risk of the development recurrence of autoim-mune or inflammatory diseases. It is likely that previoushypercortisolism suppressed the conditions.

Patients with Carney complex and primary pigmentednodular adrenocortical disease should be followed annu-ally for the development of new or recurrent cardiac myx-omas and other features of the syndrome (166).

Clinicians should tell patients that theymay feel unwellfor 69 months, that their mood will gradually improve,and that improvementmay continue formore than 1 year.Generally, weight, bruising, and the physical appearanceof the face change first. Patients should receive recom-mendations for physical therapy and nutrition to optimizethe recovery of muscle strength and the normalization ofweight. The recovery of comorbidities is variable, and cli-nicians should continue treating these conditions, taperingoffordiscontinuing treatmentasgoals aremet (seebelow).

Metabolic syndrome and cardiovascular riskMetabolic syndrome, including central obesity, arterial

hypertension, insulin resistance, impaired glucose toler-ance, and dyslipidemia, is present in at least two-thirds ofpatients with CS (167), and it is believed to contribute toincreased cardiovascular morbidity.

After successful treatment, decreased body weight andfat mass, improved blood pressure, a reduction in antihy-pertensive treatment, or a reversal of hypertension usuallyoccurs within weeks to a year. However, excess weightand hypertension may persist in up to 25% of patients (8,14, 15, 168176). Patients with a history of CD haveincreased blood pressure levels compared to age-, sex-,and body mass index (BMI)-matched patients (177). Per-

sistent hypertension has been attributed to microvesselremodeling, abdominal obesity, and insulin resistance (15,177). A longer duration of hypercortisolism is associatedwith persistent hypertension (15).

Similarly, glucose and lipid metabolism improve, andmedications may be reduced and/or discontinued inmanypatients, but the prevalence of diabetes and dyslipidemiaremains increased compared to BMI-matched controlsseveral years after remission (8, 168, 171, 175, 177). Fast-ing blood glucose may underdiagnose diabetes mellituscompared to an oral glucose tolerance test (178). Clini-cians should carefully monitor adult patients treated withGH replacement for diabetes mellitus (179). The persis-tence of central obesity plays a key role in continued in-sulin resistance (14, 171, 175, 177).

There is increased long-term risk of myocardial infarc-tion (7, 8). Because hypertension anddiabetes appear tobethe main controllable determinants of cardiovascularevents and mortality (8, 11, 174, 180), rigorous and re-peated long-term follow-up evaluation and treatment ofthese conditions are mandatory. One study found thatdepression was associated with an increased risk of car-diovascular disease (180).

Mood and cognitive functionPsychiatric referral may benefit adult patients with CS.

Many patients experience improvement in psychiatricsymptoms, particularly depression and emotional lability.However, several years after CD is cured, patients have anincreased prevalence of psychopathology, including anx-iety, depression symptoms, and maladaptive personalitycompared to patients treated for other types of pituitaryadenoma (173, 181184). Although a few short-termstudies indicate a significant improvement in psychopa-thology within the first year after successful surgical ormedical remission (82, 182), little is known regardinglong-term reversibility.

Cognitive impairment, particularly declarative short-termmemory deficits, is a prominent feature of hypercor-tisolism.Deficits also occur in attention, visuospatial abil-ities, and working memory (185). These are likely relatedto the abnormal function of key central nervous systemstructures such as the hippocampus, amygdala, and fron-tal cortex that are peculiarly vulnerable to glucocorticoidexcess. Several imaging studies identified atrophy in thesespecific brain areas including decreased hippocampal vol-ume during active CS; this improves after remission insome but not all patients. Some, but not all, studies founda parallel improvement in cognitive function (186190).

Pediatric patients show personality changes such asmoodiness, irritability, compulsive behavior, and over-achievement in school (191).



There are scarce data regarding the long-term recoveryof cognitive impairments. Compared tomatched controls,patients in long-term remission from CD or adrenal ade-nomas performed significantly worse in tests of executivefunction (185), memory (185, 190), and other cognitivetests (192). When evaluated, hypopituitarism and hydro-cortisone dependency were associated with worse perfor-mance, but cognitive impairment was independent of co-existent chronic fatigue or affective disorders (192). Thepatients with memory scores below normative cutoff val-ues showed decreased hippocampal volume on MRI.

Quality of lifePatients with active CS of any cause have impaired

HRQOL compared to patients with other pituitary tu-mors (181, 193, 194). Because CS is associated with long-term physical, psychological, emotional, and cognitive af-ter-effects despite cure, HRQOL is likely to be affected inthe long term and is an important outcome measure toassess in patients with a previous history of CS.

There are fewpublished cross-sectional studies ofQOLduring long-termremission.Most usedgenericQOLques-tionnaires. QOL improves in patients in remission com-pared to those with hypercortisolism, regardless of thecause of CS or treatment used (13, 194196). Neverthe-less, two large cross-sectional studies report long-term re-sidual impairment in physical and social functioning aswell as limitations in the ability to perform usual activitiesof daily living due to physical and emotional problems,pain, fatigue, sleep problems, reduced vitality, and feelingof health impairment (13, 197).Togetherwith the residualpsychological and cognitive impairments, chronic de-creased perceived QOL has significant social and eco-nomic consequences, such as the inability to return towork (198).

The duration of remission did not affectQOL results inone study (197) but was associated with lower physical(but not psychological) component scores in another (13).Postoperative hypopituitarism may decrease QOL (197).However, it is unlikely to play a large role because GH-deficient females in remission from CD have a greater im-pairment in QOL than those with other causes of GHdeficiency (199).

OsteoporosisOsteoporosis results from direct effects of cortisol on

bone cells and indirect events such as glucocorticoid-in-duced hypogonadism, secondary hyperparathyroidism,GH deficiency, and reduced bone strain due to myopathy(200). However, the prevalence of fractures is not known(201). Studies of adolescents and adults document majorimprovement and sometimes complete normalization of

BMD after curing CS, although 45 years may be neededto achieve full recovery (202205). Importantly, supra-physiological postoperative hydrocortisone doses mayhamper bone recovery. The scant fracture data suggestthat the increased symptomatic fracture risk disappearsafter curing CS, similar to that seen in exogenous gluco-corticoid-induced osteoporosis (204, 206). It is not clearwhether complete normalization occurs because the ex-pected BMD for a given person cannot be predicted (8,173, 200).We recommend a detailed fracture assessment,evaluating BMD at the spine and hip, obtaining lateralmorphometric imaging of the spine, adequate calciumandvitamin D intake, avoidance of excessive glucocorticoidsupplementation, and personalized long-term follow-upbasedon the resultsofboneevaluation.Arecent consensusp

cushing es15

Documents