Transsphenoidal Surgery for Cushing’s Disease: Pitfalls, Results, and Long-Term Follow-up




Introduction


Cushing’s syndrome is an endocrinopathy, resulting in distinctive clinical features and associated systemic changes. Cushing’s disease is the name given to the disorder when it is caused by a pituitary source, usually an adrenocorticotrophic hormone (ACTH)-secreting pituitary microadenoma. The hypersecretion of ACTH by the pituitary adenoma, usually a microadenoma, results in bilateral adrenocortical hyperplasia and hypercortisolism. Cushing’s disease is a rare disorder with a reported incidence of between 0.7 and 2.4 cases per million inhabitants per year. If it is left untreated it is associated with a high morbidity and mortality, mainly from vascular disease associated with hypertension and impaired glucose tolerance. Cushing’s disease is associated with microadenomas in 90% of cases and macroadenomas in only 10%, and ACTH-secreting adenomas account for 15% of functional pituitary tumors.


Although the pituitary gland has been known from antiquity and the adrenal glands were first described in 1563, it was only in 1912 that Cushing’s described his famous patient with hypercortisolism, where he observed small basophilic adenomas in the pituitary gland, but he assumed it to be a polyglandular disorder. Bauer first postulated that hypercortisolism resulted from adrenal hypersecretion, either primary or secondary. The cause was disputed for many years and it was only in 1933 that the first patient had neurosurgery for Cushing’s disease. During the 1950s and 1960s the therapy was directed against the target organs, and bilateral adrenalectomy was the treatment of choice, but before the introduction of glucocorticoids this was an extremely perilous undertaking.


In 1969 Jules Hardy reported the first successful removal of a pituitary adenoma for the treatment of Cushing’s disease. Subsequently, with the introduction of microsurgery, intraoperative fluoroscopy, and advances in endocrine and radiological diagnostic techniques, transsphenoidal adenomectomy has become the treatment of choice for Cushing’s disease, with reported remission rates between 70% and 90%. It is the only treatment modality that allows for remission of Cushing’s disease without the need for replacement therapy or a high rate of serious complications. More recently endoscopic transsphenoidal surgery has been introduced with favorable results.




Pitfalls in the Treatment of Cushing’s Disease


The neurosurgeon treating Cushing’s disease should not only have a thorough anatomical understanding of the sella and surrounding structures, but also be competent in endocrinological interpretation. While the majority of published series on transsphenoidal surgery are from centers managing a high number of patients, the treatment of this condition is no longer only restricted to these centers. To perform transsphenoidal surgery for Cushing’s disease safely and effectively, it is important to have an understanding of the pitfalls in treating this condition. The difficulties encountered may be related to the incorrect interpretation of endocrinological and radiological tests and mistakes made in surgical technique.


Pitfalls in Endocrine Assessment


The diagnosis and differential diagnosis of Cushing’s disease remains a considerable challenge in clinical endocrinology. The diagnosis is usually a multistep process, involving first the diagnosis of Cushing’s syndrome, differentiating it from pseudo-Cushing’s states; then the diagnosis of ACTH-dependent Cushing’s syndrome; and finally determining if the source of ACTH overproduction is from the pituitary (Cushing’s disease) or an ectopic source.


The best screening for Cushing’s syndrome has been a 24-hour urine collection with analysis of urinary-free cortisol excretion. Other methods used are low-dose dexamethasone suppression testing, midnight cortisol, or repeated late night salivary cortisol measurements (lack of diurnal variation). It should, however, be remembered that urinary-free cortisol measurements have a relatively low specificity, and are frequently elevated in other conditions such as polycystic ovarian syndrome and depression. The dexamethasone suppressed corticotrophin-releasing hormone (CRH) test may differentiate Cushing’s syndrome from those with pseudo-Cushing’s states, who have decreased ACTH response to CRH because of negative feedback exerted by chronic hypercortisolism.


The diagnostic hallmark for Cushing’s disease is the diminished suppressibility of ACTH and cortisol in response to exogenous glucocorticoids (dexamethasone). In pituitary corticotroph adenomas, the set point for ACTH suppressibility by glucocorticoids is altered, as demonstrated by the need for larger doses of exogenous glucocorticoids to reduce serum cortisol levels. Cortisol nonsuppressibilty during administration of low-dose dexamethasone, but suppressibility during high-dose dexamethasone is the key diagnostic finding in patients with Cushing’s disease. However, this test is not definitive and sensitivity and specificity are in the 80% to 90% range, with reported false-positives for adrenal pathology and ectopic ACTH. Corticotrophin assays and the CRH stimulation test also have a high accuracy in proving pituitary dependency. It has been reported that the high-dose dexamethasone suppression test is less reliable in males. The diagnosis of Cushing’s disease in children may also be more difficult as some children with Cushing’s disease may suppress with low-dose dexamethasone or may show a waxing and waning in their cortisol concentration. Patients with macroadenomas have been reported as showing significantly less suppression of cortisol and its metabolites after a high-dose dexamethasone suppression test compared with microadenomas and the degree of cortisol suppression may not be a reliable test in such patients. In interpreting CRH responses it is important to know whether patients receive human or ovine sequence CRH as the responses to these peptides differ quantitatively and response criteria derived from one peptide may not be applicable to the other. A number of cases of pituitary apoplexy following the administration of CRH in the investigation of patients with Cushing’s disease have been reported.


In difficult cases, inferior petrosal sinus sampling (IPSS) is the most reliable diagnostic procedure for diagnosing pituitary-dependent Cushing’s disease and differentiating it from ectopic ACTH syndrome. The inferior petrosal sinus receives the drainage of the pituitary gland without mixture of blood from other sources and the concentration of ACTH will be much greater in the IPSS than at a peripheral location in patients with a pituitary adenoma. There will not be a significant difference between central and peripheral ACTH in ectopic Cushing’s syndrome. Results with cavernous sinus sampling are less accurate and it has a higher rate of complications, whereas jugular sampling is safer, but because it receives blood from multiple drainage areas is less accurate. Bilateral simultaneous sampling of the inferior petrosal sinuses with the catheters at the same level must be done because an adenoma may be localized to one side of the pituitary and the ACTH levels would be no different from the periphery if the incorrect petrosal sinus is sampled alone. *


* References , , , , , .

In sampling it is also important that the catheter tips are actually situated in the inferior petrosal sinus, otherwise there may be admixture from the internal jugular vein. Nonetheless, due to the varying venous anatomy, the exact positioning of the catheter tips within the IPSS may be difficult. The diagnostic accuracy of this test ranges from 80% to 90%, therefore even a positive result can occasionally result in an unsuccessful transsphenoidal procedure. Administration of CRH at the same time as IPSS may improve the accuracy of IPSS. IPSS has been questioned because it is a highly technical procedure requiring a considerable amount of experience, and severe adverse effects, although rare, do occur, and because of this should probably be reserved only for equivocal cases.




Radiological Pitfalls


Computed tomography (CT) scan is an unreliable diagnostic modality and has limited value in the diagnosis of microadenomas. In Cushing’s disease it is able to identify a pituitary microadenoma in 42% to 59% of cases. Magnetic resonance imaging (MRI) is the radiological investigation of choice for patients with Cushing’s disease ( Figure 19-1 ). The sensitivity of MRI in detecting microadenomas is reported to range from 44% to 80% and this can be increased to 90% with the use of intravenous gadolinium contrast medium. Visualization of the microadenoma on the MRI scan depends upon the size and composition of the pituitary adenoma. The smallest size adenoma that has been reported to be detected is 3 mm. T1-weighted coronal images after the administration of intravenous gadolinium are helpful in detecting microadenomas. The timing of the postcontrast images is critical ( Figure 19-2 ). The early and fast imaging after gadolinium injection improves the contrast difference between the microadenoma and normal pituitary tissue, and hence the detectability of the adenoma. The adenoma will typically appear as a low intensity area within the high intensity enhancing pituitary gland on T1 weighted images. However, in most instances the postgadolinium studies are performed 2 to 3 minutes after the injection of the gadolinium and by this time the contrast between adenoma and normal pituitary will already be reduced, which may increase the likelihood of a false-negative diagnosis. The use of dynamic MRI has been shown to improve the sensitivity of the MRI for visualizing ACTH secreting microadenomas. However, the specificity is reduced due to technical artefacts, lower clarity of images, and incidental pituitary lesions not detected on conventional MRI. Therefore the role of Dynamic MRI may be to detect microadenomas not visualized on conventional MRI in patients with a definite endocrinological diagnosis of Cushing’s disease. A MRI artefact of more than 1 mm mimicking an adenoma has been documented on 14% of scans, and this can easily lead to inaccurate exploration of the pituitary with failure of remission. With the rapidly improving technology and imaging techniques, it is likely that smaller microadenomas will be detected in the future.




Figure 19-1


A, T1-weighted sagittal MRI scan precontrast. B, T1-weighted coronal MRI scan precontrast. C, Gadolinium- enhanced T1-weighted sagittal MRI scan showing enhancement of the pituitary, and a hypointense adenoma anteriorly. D, Gadolinium enhanced T1-weighted coronal MRI scan demonstrating the microadenoma in the coronal plane.



Figure 19-2


A, Standard gadolinium-enhanced T1-weighted sagittal MRI scan showing diffuse enhancement within the sella. B, Standard gadolinium-enhanced T1-weighted sagittal MRI. C, Delayed gadolinium-enhanced T1-weighted sagittal MRI scan, done 10 minutes after contrast administration. The adenoma is now clearly seen as a hypointense lesion anteriorly. D, Delayed gadolinium-enhanced T1-weighted sagittal MRI scan. Poorly defined hypointense region demonstrated in right lateral aspect of gland in keeping with a microadenoma.


In up to 20% of cases of patients with endocrinologically proven Cushing’s disease, the MRI scan will be normal with no evidence of a microadenoma. This may be due to an incorrect diagnosis with the source of the ACTH being ectopic; it may be due to a microadenoma being present but too small to see on a MRI scan; or rarely it may be due to diffuse hyperplasia of the pituitary being the source of the ACTH excess. Cases of Cushing’s disease with a normal MRI in which a pituitary adenoma was found in the cavernous sinus or pituitary stalk have been described. A patient has also been reported as having an elevated ACTH from pituitary metastasis in the liver following surgical, radiotherapeutic, and medical treatment for pituitary dependut Cushing’s syndrome in which there was no longer any evidence of disease within the pituitary itself. When the neurosurgeon is faced with this dilemma of biochemical-proven Cushing’s disease and a normal MRI, he or she is faced with the choice of exploring the entire gland in the hope of identifying a small adenoma not identified on the MRI, or of performing a hypophysectomy or hemihypophysectomy in the hope of removing the tumor at the same time. Other treatment options include bilateral adrenalectomy, suppressive or ablative medical treatment, or conventional radiotherapy or radiosurgery. The identification and removal of the microadenoma clearly not only offers the best chance of remission from the Cushing’s disease, but also the least chance of unwanted side effects. Semple and Laws have demonstrated that the remission rate in patients with biochemically proven Cushing’s disease and a normal MRI is the same as when a microadenoma is identified on MRI.


IPSS is a useful method for intrapituitary localization of a microadenoma when the MRI is normal in a patient with Cushing’s disease. It has been suggested that IPSS may be more accurate than MRI for lateralization of a microadenoma. However, because ACTH secretion may be intermittent, there is a risk of sampling in the interval between secretory phases and thus obtaining a false-negative result. CRH stimulation is a successful method to improve the sensitivity of IPSS. The correct lateralization of the microadenoma is reported to be 66% to 99%. A ratio of 1.4 or greater of the level of ACTH in one inferior petrosal sinus to the other reliably localizes the tumor to that side of the gland. An important factor in the success of IPSS is that the majority (73% to 80%) of ACTH-producing microadenomas are located in the lateral wings of the pituitary gland. In midline tumors there is a lack of side to side gradient. A dominant petrosal sinus, draining the majority of the pituitary, could explain why even when there is a large R:L ratio, this ratio does not always correctly predict the side on which a small microadenoma may be found.




Pitfalls in Transsphenoidal Surgery


Some of the potential pitfalls that one may encounter in transsphenoidal surgery for Cushing’s disease apply to transsphenoidal surgery in general. In Cushing’s disease the pituitary fossa is normal in size in the majority of cases, but it is important to assess the type of sphenoid sinus the patient has on the MRI scan, lateral skull x-rays, or CT scan. If one operates on a patient using the transsphenoidal approach with a conchal sphenoid sinus, it may in fact become very difficult to reach the pituitary fossa even with extensive drilling. This may be a problem that is more common in children who have poorly developed sphenoid sinuses and may sometimes require a two-stage procedure. The carotid arteries should always be visualized on the coronal cuts of the MRI to ascertain how close they come to each other. Occasionally, “kissing carotids,” where the internal carotid arteries almost touch each other across the midline, may be encountered, and this may lead to vascular disaster in the transsphenoidal approach. Maintaining vertical and horizontal orientation is essential during the approach to the sphenoid sinus and pituitary fossa. This may be more difficult if the patient has undergone previous transsphenoidal surgery. Vertical deviation can result in penetration of ethmoidal air cells and anterior fossa, and lateral deviation increases the likelihood of carotid artery and cavernous sinus injury. Seven percent of pituitary adenomas may be accompanied by an aneurysm. It is possible for a suprasellar or even intrasellar aneurysm to be mistaken for a pituitary macroadenoma.


There appears to be three main reasons for failure of transsphenoidal surgery for the treatment of Cushing’s disease: (1) Patients in whom extensive pituitary exploration did not reveal an adenoma; (2) patients in whom the pituitary adenoma was not completely removed either because of poor surgical technique or where an invasive tumor has penetrated surrounding tissue, making selective removal technically impossible; and (3) finally a small number of patients with a seemingly discrete lesion that was in fact, at microscopic examination, surrounded by hyperplastic corticotrophs. Most failed explorations occur because of the inability to identify an adenoma, and this is most commonly in patients in whom the MRI of the pituitary demonstrated normal findings. There have also been cases reported where the adenoma is in an ectopic, extrasellar position, such as along the pituitary stalk, in the sphenoid sinus, or in the cavernous sinus. In these cases the intrasellar exploration will be negative and the patient will not go into remission until the ectopic site of the pituitary adenoma is found and treated. Rarely there may be multiple microadenomas in patients with Cushing’s disease, not all of which secrete ACTH. Reports have described how remission was not obtained following the removal of an initial microadenoma, but only after a second procedure where the ACTH producing adenoma was removed.


Negative exploration of the pituitary gland has been reported in 6% to 30% of cases, and remains the most common reason for surgical failure. The procedure of choice, undoubtedly, is to identify the microadenoma intraoperatively and macroscopically remove it ( Figure 19-3 ). The majority of intratumoral locations are lateral and paramedian (91%) and medial in only 9% of cases. On opening the sella, Knappe and Ludicke found that the tumor was visible in 13.5% of patients, after horizontal incision in 42.5%, and after pituitary exploration in 44.2%. If the tumor is visible on opening or surgical exploration, adenomectomy may be performed. The varied locations, the small average size, the occasional semiliquid consistency of Cushing’s adenomas and their frequent resemblance to the neurohypophysis emphasize the need for meticulous surgical exploration. In addition to removal of all visible tumor, a wide rim of pituitary tissue surrounding the tumor site should also be removed in premenopausal women; in postmenopausal women an even wider rim should be removed. This reduces the chance of microscopic invasion of the surrounding gland from the microadenoma being left behind, resulting in recurrence or failure of remission. If no tumor is visible on surgical exploration, especially with a normal MRI, and when IPSS suggests a unilateral location of the tumor, then a hemihypophysectomy may be performed in the hope that a very small microadenoma can be removed with the pituitary tissue. Total hypophysectomy can be considered if the patient is beyond childbearing age and has agreed to the procedure in advance. The procedure is normally reserved for those cases with severe Cushing’s disease and typical preoperative testing, in which a definite adenoma cannot be found during surgery. Sometimes it may be pertinent to await postoperative endocrine and pathology reports, even if no obvious tumor is found, and then consider hypophysectomy if there is no remission and pathology reports are negative. If no adenoma tissue is identified in the adult patient, the virtual certainty of cure with a relatively low morbidity when compared with the severe complications of persistent hypercortisolism provide a compelling argument for hypophysectomy. Because the microadenoma may occupy the posterior lobe, total hypophysectomy requires total clearance of the sella. However, in general the success rate with total hypophysectomy is no better than with more selective surgery. Some authors encourage an aggressive approach as they believe the cure rate is higher with hypophysectomy, whereas others prefer a more conservative approach as they feel many of these cases eventually are shown to come from an ectopic source. In general it has been reported that the more aggressive the surgical approach, the higher the rate of remission, but at the cost of a higher rate of postoperative endocrine replacement.




Figure 19-3


Intraoperative photograph of pituitary fossa. Arrowheads show margin of microadenoma within the pituitary gland.


Intraoperative ultrasound has been successfully used as the primary means of detecting and localizing very small microadenomas that were not visualized on MRI or at surgical inspection of the surface of the gland. The adenoma is hyperechoic in relation to the rest of the pituitary gland. Intraoperative measurement of ACTH in adenoma and in anterior lobe micro samples has been used to improve localization of minute microadenomas. This method permits a clear differentiation between adenoma and pituitary tissue, and may be valuable in difficult cases with unclear intraoperative findings. However, in another study on intraoperative ACTH levels, it was found that the decline of ACTH during surgery did not accurately predict complete tumor resection.


In general identification rate for cushing’s adenoma on histologic examination is lower than that achieved macroscopically at surgery. The possible reasons for this are the tumor may be lost during sampling or histological examination, the tumor may be too small to be detected on routine histology, or that although the tumor was never removed it was somehow effectively treated by manipulation of the pituitary gland or alteration of its blood supply. The relatively high remission rate in these patients suggests that they do not represent a different etiological group. Sheehan et al in a report from a center treating a high volume of Cushing’s adenomas found no difference in outcome between this group and those in which an adenoma was identified at histology.


Hemorrhage at the operative site has been reported as the major complication in transsphenoidal surgery for Cushing’s disease. Problems with venous bleeding appear to be more common in patients with Cushing’s disease than in patients with other pituitary tumors. The venous bleeding may arise from the anterior intercavernous sinus or the cavernous sinus. The venous bleeding from the anterior intercavernous sinus is more common in Cushing’s disease because the majority of adenomas are microadenomas. Macroadenomas are more likely to compress the sinus, eventually obliterating it, whereas in microadenomas these sinuses remain patent. In general the bleeding can be controlled by bipolar cautery or Surgicel. Occasionally, if the bleeding is profuse. it may be worthwhile to pack off the bleeding area and then return a few days later to complete the operation because the sinus will have probably thrombosed, making surgery easier. Arterial bleeding from injury to the carotid artery can also occur. It is particularly important to study the carotid arteries on the MRI before surgery to ensure they do not loop into the sella. For the same reason one should not do a primary deep cut into the gland with a sharp instrument because it may injure the carotid artery that is partially in the sella. The carotid artery injury may not bleed initially if the injury to its wall is partial, but may develop a false aneurysm that can result in delayed hemorrhage.




Complications


Transsphenoidal surgery in the hands of an experienced surgeon is a safe procedure with mortality rate ranging from 0.4% to 2%. The overall complication rate for transsphenoidal surgery is 3.3% to 9.3%. Although the complication rate is low, when they do occur they can be life-threatening. An in-depth analysis of transsphenoidal complications is beyond the scope of this chapter but includes diabetes insipidus; syndrome of inappropriate secretion of antidiuretic hormone; vascular injury; visual complications; cranial nerve injuries; cerebrospinal fluid leaks and meningitis; nasal complications including sinusitis, septal perforations, and epistaxis; hypopituitarism; and medical problems such as deep vein thrombosis and pneumonia. *


* References , , , , , .



Macroadenomas in Cushing’s disease, although relatively uncommon, may present similarly to nonfunctioning tumors. Their complications are similar to those one would expect to find in nonfunctioning tumors. The fact that the majority of adenomas in Cushing’s disease are small and occur in relatively young people makes them less likely to suffer from complications related to the surgical procedure itself. Rees et al published a complication rate of 15% in their series of 54 patients with Cushing’s disease who underwent transsphenoidal surgery. A series of complications in patients undergoing transsphenoidal surgery for Cushing’s disease is summarized in Table 19-1 . The overall rate of complications for patients in Semple and Laws’ series was 13.5%, which is higher than other published series for transsphenoidal surgery for pituitary adenomas in general. However, if only the permanent complications are included, then the rate is 1.8%. There was an increased rate of complications in those patients with macroadenomas and patients with normal MRI findings. The elevation of the complication rate in this series was due to the increased medical complications, namely deep vein thrombosis, pneumonia, and poor wound healing. Cushing’s disease patients are frequently debilitated because of associated medical conditions such as obesity, hypertension, and diabetes mellitus, which make them more susceptible to medical problems in the perioperative period. It is to be expected that patients undergoing hemihypophysectomy and hypophysectomy for Cushing’s disease will have a higher rate of hypopituitarism than those patients who have undergone adenomectomy. Cushing’s disease patients often have a diuresis due to the excess body water that is lost in the early postoperative period and this may be mistaken for diabetes insipidus. It has been reported that children have a higher rate of anterior pituitary dysfunction and diabetes insipidus compared with adults.



Table 19-1

Complications in a Contemporary Series of Transsphenoidal Surgery for Cushing’s Disease *




























































































































Case No. Age (yr) Sex Previous Treatment MR Findings Invasion Complication
1 24 F micro no Permanent DI
2 48 M trans macro no Nasal septal perforation
3 52 F macro yes Deep vein thrombosis
4 32 F micro yes CSF rhinorrhea (postop)
5 52 M normal no Deep vein thrombosis
6 37 F normal no Ocular neuropathy
7 53 M micro yes Deep vein thrombosis
8 68 F GK normal no SIADH, hyponatremia (fatal)
9 71 F micro ? Pneumonia and UTI
10 36 F macro yes Deep vein thrombosis
11 28 F micro no Epistaxis
12 45 F macro yes 3rd cranial nerve palsy (transient)
13 32 F normal no Nasal septal perforation
14 47 M normal no Wound dehiscence

Table reproduced from Semple PL, Laws ER: Complications in a contemporary series of patients who underwent transsphenoidal surgery for Cushing’s disease. J Neurosurg 1999; 91:175-179 with permission of the publishers.

* All adenomas presumed from MR findings were confirmed on pathological examination, except in case 9. Abbreviations: GK, gamma knife; invasion, invasion of dura seen on pathological examination; macro, macroadenoma; micro, microadenoma; trans, transsphenoidal; UTI, urinary tract infection; – , none; ?, uncertain (no dural specimen was obtained).


These procedures were performed at another institution.


Normal pituitary gland according to pathological examination.





Results and Long-Term Follow-up


Definitions, Methods, and Timing of Follow-up


Although the terms “cure” and “remission” have been widely used in publication of results in patients treated for Cushing’s disease, there has been very little uniformity in their definition, making it extremely difficult to compare and analyze results from different centers. *


* References , , , , , , .

The majority of authors use biochemical and clinical factors to define cure, some use only biochemical and others only clinical factors. Cure suggests a permanent outcome. Recurrence has occurred in patients many years following “cure”; indeed the longer the period of follow-up, the greater the recurrence rate. Therefore, until longer periods of follow-up have been published, and some determination made as to when the chance of recurrence is so minimal that “cure” can be used, remission is probably a more accurate way to describe both short-term and long-term results. The majority of published series have follow-up periods that are relatively short and this makes it difficult to determine with accuracy the recurrence rates and actual prevalence of long-term control. An additional problem because of the varying methods used to determine “cure” is the difficulty in determining whether patients had a recurrence of the disease or regrowth of a residual tumor that was never cured.


Bochicchio et al relied on clinical assessment alone to determine remission postoperatively but the majority of series have used biochemical methods in addition to clinical assessment. Biochemical tests used to assess remission are:



  • 1.

    Unmeasurable early serum cortisol has been advocated by the majority of authors of large series as a definition of cure. Measurements are usually performed 5 to 14 days after surgery and at least 12 hours following withdrawal of oral hydrocortisone replacement therapy. The rationale surrounding this method is presumed inhibition of ACTH secretion from the normal pituitary in a patient harboring an ACTH-secreting adenoma. The lack of ACTH release from the suppressed pituitary gland leads to lack of stimulation of the adrenal gland and therefore the serum cortisol should become undetectable. The inhibition of the pituitary gland may resolve over time. However, long-term clinical remission has been reported with measurable serum cortisol levels postoperatively, and recurrence of Cushing’s disease can occur in patients who postoperatively have subnormal cortisol.


  • 2.

    Various levels of urinary or serum cortisol have been cited below which surgery is considered successful. However, other studies have cited measurable cortisol levels as diagnostic of a residual tumor and failed remission.


  • 3.

    Normal suppression of serum cortisol with low-dose dexamethasone (2 to 3 mg) overnight has been used to define remission. There is a good, but not complete (78%) concordance between the cortisol response to dexamethasone and postoperative remission.


  • 4.

    Serum ACTH levels have been evaluated, both basally and after dynamic stimulation or suppression of the pituitary axis. In a study by Fahlbusch and Buchfelder, all the patients with reduction of serum ACTH to low levels had clinical remission.


  • 5.

    A subnormal response of cortisol to ACTH or CRF stimulation in the early postoperative period has been considered a criterion for remission. An early response to ACTH or CRH would indicate the presence of a residual tumor.



The majority of endocrine testing occurs within 2 weeks of surgery. This has the advantage of providing valuable early prognostic information and expediting definitive treatment if hypercortisolism persists after surgery. Other authors, however, reported that serum cortisol levels at 12 weeks were lower than levels obtained at 2 weeks and better at predicting remission. However, the majority of authors seem to agree that immediate postoperative cortisol values provide a good guide to long-term remission. When doing postoperative testing, the cyclical secretion of cortisol must be considered. Finally, as experience with transsphenoidal surgery for Cushing’s disease is still relatively short, all patients need lifelong endocrine follow-up.


The definition of recurrence is also varied and imprecise. The progressive increase in recurrences the longer the follow-up begs the question whether they are in fact recurrences or represent residual disease. Lamberts et al propose that true recurrences should be defined as “the return of clinical and biochemical characteristics of Cushing’s disease after a period of adrenal insufficiency following surgery, followed by a period during which the hypothalamo-pituitary-adrenal axis is shown to be completely normalized, including normal diurnal rhythm of cortisol, a normal response of cortisol to hypoglycemia, and a normal response to dexamethasone.”


Remission and Recurrence Rates


The remission rates from large published series vary from 64% to 93% ( Table 19-2 ). *


* References , , , , , , , .

However, the rates of remission also differ between short-term and long-term follow-up. Postoperative remission rates vary from 70% to 93%.

References .

Studies with longer-term follow-up, in excess of 3 years, have lower rates of remission, and recurrence develops in up to 20% of patients. Chee et al reported an initial remission rate of 78%, but on longer follow-up in excess of 2 years, the remission rate had dropped to 67%. Rees et al in a report on 54 consecutive patients reported an initial remission rate of 77%. The median follow-up was 6 years but only a 5% recurrence rate was documented. However, in another study by Pieters et al, 25% of the patients who were in remission had relapsed after 4.5 years, and Tahir and Sheeler reported a recurrence rate of 20%. In a study of 162 patients, Sonino and co-workers showed that the cumulative percentage of patients remaining in remission after 2 years was 93%, but after 10 years this had decreased to 70%. It can therefore be expected that the longer the period of follow-up the greater the number of relapses there will be, and this reinforces the notion that patients who are in remission should have lifelong endocrine follow-up. Patients who undergo a second procedure have a lower remission rate. Shimon et al reported a remission rate of 78% for patients after one operation, but only 62% in patients who had undergone a second procedure. Blevins et al noticed that the initial remission rate was lower for macroadenomas than microadenomas (67% vs. 91%) and the recurrence rate was higher in macroadenomas (36% vs. 12%). In children treated for Cushing’s disease, the rate of remission in patients with extended follow-up was 50% to 73%, which is much lower than in adults and seemed to indicate that Cushing’s disease is more aggressive in children. However, other authors have published results in children with remission rates no different than adults ( Table 19-3 ).
Jun 10, 2019 | Posted by in NEUROLOGY | Comments Off on Transsphenoidal Surgery for Cushing’s Disease: Pitfalls, Results, and Long-Term Follow-up

Full access? Get Clinical Tree

Get Clinical Tree app for offline access