Pituitary Apoplexy: A Comprehensive Review
Keywords: Endocrinopathy, extra ocular palsy, headache, pituitary adenoma, pituitary apoplexy, trans-sphenoidal surgery, visual disturbancesKey Message: Pituitary apoplexy is a rare usually vision threatening and occasionally life threatening condition. A high index of suspicion is required for diagnosis. Prompt diagnosis and appropriate , timely intervention leads to good visual outcome in majority of the patients. Many patients, however, need long term hormone replacement therapy
Pituitary apoplexy (PA) is a clinical syndrome caused by a rapid expansion of the sellar contents due to infarction and/or hemorrhage. PA is often a misunderstood and misdiagnosed condition. This term is usually reserved for the typical clinical syndrome as in roughly 25% of cases of pituitary adenoma, there may radiological or histopathological evidence of infarction/hemorrhage. Usually, the typical syndrome occurs in a pre-existing, unrecognized pituitary tumor even though it can occur in patients with known tumors on observation/treatment. The term “subclinical PA” is used to refer to patients who are found to have radiological, intraoperative, or histopathological evidence of hemorrhage and/or infarction without the typical clinical presentation. However, in this chapter, PA refers to “classical, clinically diagnosed PA.”
The exact incidence of PA in the general population is likely to be underestimated as many cases remain subclinical. The incidence reported in the literature varies from 4% to 20%., This wide variation is often due to the definition of PA described in different studies and due to variations in the study populations in terms of gender, age, tumor size, tumor type, viz., functioning versus nonfunctioning, etc. In a population-based study in the United Kingdom, PA was found to occur in 7.9% of patients with pituitary adenomas. In one of the largest studies to be reported in recent times, Zhu et al. found an incidence of 4.8% among their 2,021 patients with pituitary adenoma. In general, many studies have reported a male predominance in the incidence of PA.,, In the vast majority of patients (~80%), PA is the first manifestation of pituitary adenoma. In patients with known pituitary adenomas, the incidence of PA varies from 3% to 25%.,,, There are studies which have found no gender predilection in PA, whereas there are other studies which have found a distinct male preponderance. PA can occur in all types of pituitary adenomas. However, it statistically is more common in macroadenomas and nonfunctioning adenomas followed by prolactinomas.[3.4] In their study of 2,021 patients with pituitary adenoma out of which 97 developed PA, Zhu et al. found that PA occurred 10 times more frequently in macroadenomas vis-à-vis microadenomas. Moller-Goede et al. in their study of 574 patients with pituitary adenomas of whom 42 suffered PA, the incidence of PA was three times greater in clinically nonfunctioning pituitary adenomas (NFPAs) than in secretory adenomas. The exact reason for the higher incidence of PA in patients with NFPA is not known. It is presumed that due to their non-secretory nature, these tumors become symptomatic at a later stage and therefore are larger at the time of diagnosis and therefore more prone for infarction/hemorrhage. While apoplexy is common in pituitary adenomas, it can be a presenting feature of other sellar pathologies like Rathke's cleft cysts. In a large single institution series by Jho et al., 10% of apoplectic episodes occurred in non-pituitary lesions including Rathke's cleft cysts. In another single institution study, 4.5% of apoplexy patients had histopathological evidence of Rathke's cleft cyst.
Much has been written about the role of precipitating factors in PA. PA is known to be associated with a variety of precipitating factors which include head trauma, surgery, pregnancy, medication, pathological states like anticoagulation, endocrinological testing, etc.,,,, In a recent single institution study, 35% of patients with PA were found to have precipitating factors. The mechanism of precipitation of PA by these factors is variable as has been hypothesized. For example, in patients undergoing surgical procedures hypotension can reduce the blood supply to the already compromised tumor or the “surgical stress” can stimulate the pituitary gland to secrete more corticosteroids thereby precipitating apoplexy. Conditions like head trauma, lumbar puncture, spinal anesthesia can cause PA either by changes in blood pressure or changes in the intracranial pressure either of which can cause this event. Stimulation tests of the pituitary gland using GnRh or TRH or the use of GnRH analogues for the treatment of prostate cancer or endometriosis, can cause an acute increase in the size of the gland., Patients who are on anticoagulants and those susceptible to thrombocytopenia due conditions like dengue can also develop PA due to their increased tendency to bleed. Studies have shown that the risk of PA is three times more in patients on anticoagulants., In recent times, there has been an increase in the number of reports of gestational PA. The well-known changes in the anatomy and physiology of the pituitary gland during pregnancy including the increase in the volume of the gland, the increase in blood supply and enhanced hormonal activity tend to precipitate PA during pregnancy. As shown by Biousse et al., when co-existing factors are present, the chance of misdiagnosis of PA is more likely because of the confounding factors. For example, gestational PA is likely to be mistaken for other conditions like pre-eclampsia. Cavernous sinus invasion has also been cited as a risk factor for PA. This may be because cavernous sinus invasion is more common in macroadenomas and therefore, both may be inter-related.
Is the clinical presentation and outcome different in patients with precipitating factors?
Biousse et al. studied 30 patients with PA and found that 30% of their patients had associated conditions that can be considered as precipitating. The clinical presentation was not different between the groups with and without precipitating factors except for the slightly higher incidence of altered mental status in patients with precipitating factors. However, in their series, the visual outcome was worse in patients with associated conditions than in those without. They also noted that the presence of associated conditions confounded the recognition and treatment of PA despite a typical clinical presentation.
PA is often attributed to the following factors: a rapidly growing tumor that outgrows its blood supply resulting in ischemic infarction, or from compression of the superior hypophyseal artery against the diaphragmatic notch, inherent fragility of the tumor vessels, etc., However, even small adenomas may show evidence of hemorrhage. Mohr estimated the incidence of asymptomatic hemorrhages in pituitary adenomas to be 9.9% as opposed to 0.6% that presented with clinical findings. Onesti described five patients with extensive hemorrhage into the pituitary adenoma that was clinically silent. However, as small adenomas also show evidence of hemorrhage and PA occurs in the absence of precipitating factors, these raise the question whether intrinsic tumoral factors contribute to the occurrence of PA? This question was answered by a study by Oldfield et al.. Oldfield and Merrill showed that pituitary adenomas have a high metabolic demand, limited angiogenesis, reduced vessel density compared with the normal gland. In addition, they also showed that pituitary adenoma cells do not survive in the absence of reduced concentrations of glucose. Oldfield and Merrill showed that the expression of vascular endothelial growth factors (VEGFs) in pituitary adenomas is very less or negligible compared to normal brain, which explains the reduced vascularity of these tumors. Moreover, several studies have shown that pituitary adenomas have reduced vascular density when compared to the normal pituitary gland.,, The increased metabolic activity of these tumors has been documented by several authors.,, It has been shown by these studies that both micro- and macroadenomas as well as functioning and nonfunctioning pituitary adenomas have increased metabolic activity when compared to the normal pituitary gland. In addition, high intrasellar pressures have been documented in patients with pituitary macroadenomas., This increased intrasellar pressure may aggravate the already tenuous vascularity of pituitary adenomas. Thus, the combination of increased metabolic demand decreased vascularity, and increased intrasellar pressure may play a significant role in the pathogenesis of pituitary apoplexy.
Oldfield and Merrill proposed that these endogenous qualities of the tumor produce a tenuous balance between high metabolic demand and marginal tumor perfusion, thereby rendering the tumor susceptible to ischemia and infarction due to any factor that alters the tenuous balance between marginal tissue perfusion and increased metabolic demand. However, there is a caveat in the hypothesis proposed by Oldfield and Merrill. This hypothesis accounts for only patients with ischemic, coagulative necrosis, and does not account for the more common hemorrhagic variety. Thus, further studies are definitely warranted to find out the causation of PA. It is possible that more than one mechanism may be involved in the causation of PA.
The classical clinical presentation of PA includes acute onset of headache with associated visual disturbances. However, in general, clinical practice, PA can present with varying symptomatology ranging from the only headache to blindness, coma, and even death., In one of the largest series of PA to be reported in the literature, Zhu et al. noted the following symptoms occurring in the descending order of frequency: headache (86%), visual disturbances (62%), vomiting (40%), extraocular palsies (25%). Except for visual disturbances, there was a higher incidence of other three symptoms in patients with PA than in those without PA. Headache is often due to the rapid increase in intrasellar pressure due to hemorrhagic infarction of the pituitary gland. Visual deterioration has been reported to occur in 52–90% of patients with PA., It is due to compression of the optic chiasm by the rapidly expanding mass and can vary from varying degrees of visual field defects to complete blindness.,,,, In a large series of patients reported by Semple et al., visual acuity was impaired in 56%, visual field defects were found in 34%, and 10% of patients presented with blindness. Extraocular nerve palsies are common because of extensions of these tumors into the cavernous sinus,, the incidence of which varies from 25% to 39%. PA can also lead to ischemic infarction of the brain due to vasospasm caused by extension of the blood into the basal cisterns or by direct compression of the vasculature by the enlarged and hemorrhagic mass.,, Altered sensorium can occur in up to 22% of patients. However, the presence of associated conditions can confound the diagnosis and often leads to a delay in the diagnosis.
Recently, Giritharan et al. proposed a clinical grading system known as Pituitary Apoplexy Score (PAS). The PAS ranges from 0 to 10 and includes the following parameters: level of consciousness, visual acuity, visual fields, and extraocular palsies. In their series, they found that patients with lower scores could be managed conservatively, whereas those with higher scores required surgical intervention. However, the authors of the PAS themselves conceded the limitations of PAS and suggested that PAS should not be used as the sole guiding system for management. For example, in PAS visual symptoms are graded from 0 to 2; a score of 1 is attributed to mono-ocular visual deficit, whereas a score of 2 is attributed to visual deficits in both eyes. However, even in mono-ocular visual deficits with a score of 1, the deficits may range from a visual acuity of 6/12 to complete blindness in one eye. As both are given a score of 1, it is imperative to understand that this PAS has limitations in practical applicability.
Another grading system was proposed more recently by Jho et al. where patients are stratified into five grades: Grade 1: patients are asymptomatic with typical radiological findings of PA (subclinical); Grade 2: patients have endocrinopathy with appropriate MRI findings with no other clinical symptoms; Grade 3: patients with headache; Grade 4: patients with ocular palsies; and Grade 5: patients with visual deficits or altered sensorium. As reported in the previous grading system, there was a trend toward surgical management in higher grade patients than lower grade patients. However, this grading system also has potential pitfalls as it includes subclinical PA (Grade 1) and it is uncommon to diagnose PA in patients presenting only with endocrinopathy—Grade 2 in Jho et al.'s grading system, as attested to by the very low numbers of patients in both these groups in Jho et al.'s series itself.
Endocrine status at the time of presentation
In view of the emergent nature of this condition, it is sometimes not possible to perform a complete endocrinological evaluation prior to intervention. However, one study showed that 70% of patients presenting with PA were hypopituitaric, 16% were eupituitaric and in the rest, the status could not be evaluated.
MRI is the imaging modality of choice in patients with PA [Figure 1] and [Figure 2]. In patients in whom it is contraindicated or could not be done, a dedicated pituitary CT scan can be done. MRI detects confirms the presence of a sellar mass with hemorrhagic/ischemic components within the tumor. Within the first week (often referred as “acute phase”) the lesion is isointense or slightly hypointense in T1 weighted sequences and hypointense in T 2 weighted sequences. In the subacute phase (7 days to 21 days) the lesion appears hypointense in both T1 and T 2 images due to the presence of methemoglobin. Gadolinium enhancement may be present in the peripheral rim and this is often referred to as “pituitary ring sign” though it is not specific for PA. In the chronic phase (>21 days) the lesion appears hypointense in T1 and T2 due to the presence of hemosiderin and ferritin. Thickening of the sphenoid sinus mucosa in patients with PA, especially, beneath the sella turcica was first described by Arita et al. This is seen only during the acute phase and is due to swelling of the subepithelial layer of the sinus mucosa. It is important to note that this thickening does not indicate infectious sinusitis and is not a contraindication for transsphenoidal surgery. Fluid-debris levels with an upper layer of fluid hyperintensity and a lower layer of hypointensity on T 1 images during the subacute stages is commonly found in PA. However, it is important to remember that there is no pathognomonic sign of PA in imaging.
There is no consensus in the literature about the ideal management for PA. In general, patients who present with a headache with minimal or no visual symptoms, without altered sensorium or those with severe associated comorbidities are chosen for conservative management, while those who present with significant visual symptoms, with altered sensorium, rapidly deteriorating neurological status are chosen for surgical management., Even though several studies have shown the advantages and disadvantages of conservative and surgical management, there is no clear consensus in the literature as the available evidence is based on retrospective and often small case series with dissimilar clinical presentations; patients in worse clinical condition being taken up for surgery while those who are well preserved being chosen for conservative treatment thus there is an inherent selection bias in many of these studies.
Conservative management generally consists of initiation of supportive measures which includes hormonal replacement therapy in the form of glucocorticoids, electrolyte management, close monitoring of visual symptoms. If conservative management is chosen, it is imperative that the patient be kept under close clinical monitoring to observe for clinical deterioration. Surgical treatment consists of transsphenoidal resection of the lesion after stabilization of the general condition of the patient. There are several studies which have shown differences in outcome between conservative versus surgical management and many of these studies, as pointed out in the following discussion, have shown better outcomes for surgical vis-à-vis conservative treatment.
Most patients with PA end up with a hormonal deficiency, especially, deficiency of anterior pituitary hormones. One single centre study found that overall 84% of patients had anterior pituitary hormone deficiency which was in the descending order of frequency: growth hormone, glucocorticoids, thyroxine, and gonadotrophic hormones in that order. Despite the less severe clinical presentation of patients undergoing conservative treatment, counterintuitively, there was no significant difference if the endocrine outcomes between conservative and surgical groups in certain studies. In another large single institutional study, anterior pituitary hormone deficiency was found in approximately two-thirds of patients while 23% required replacement therapy with desmopressin. The poor endocrine outcome of patients with PA has been noted by several authors., This poor outcome is attributed to the increased intrasellar pressure that leads to necrosis of the anterior pituitary cells as shown by the elegant study by Zayour et al. who measured intrasellar pressure and used serum prolactin levels as markers for anterior pituitary necrosis. Hence, most patients with PA would require long-term hormone replacement therapy which should be explained to the family while counselling. In contrast to the above mentioned studies, Verees et al. found that patients who underwent surgery within three days had a better endocrine outcome as 11 of their 15 patients the return of endocrine function. They postulated that the pituitary gland is capable of secreting sufficient amounts of hormones even if as little as 10% glandular tissue is preserved and at least in some cases, the endocrine deficits are due to compression of the pituitary gland rather than destruction. Randeva et al. also noted better endocrine outcome with early surgical intervention. Chuang et al. noted three times increase in the need for hormone replacement therapy in patients undergoing delayed surgical intervention than those undergoing early surgery.
Most patients have significant improvements in their visual status which may vary from partial improvement to complete recovery of their visual deficits and often there is no difference between the conservative and surgically treated group., Dubuisson et al. found that among their 24 patients with PA with impaired visual function, all but two improved. Woo et al. noted improvement in visual acuity and visual fields in 96% and 55% at three months following surgery. Bill et al. found that visual acuity and visual field defects improved in 88% and 96% of their patients with preoperative visual dysfunction. Visual function has been noted to improve even in patients with blind eyes.,,, In another large recent study, 94% of patients had significant improvements in visual function.
Does the timing of surgery influence the visual outcome and is there a difference in visual outcome between conservative and surgical treatment cohorts?
Even though intuitively early surgery should provide better visual ouctomes in PA, there is considerable controversy in the literature regarding the influence of early surgical intervention on the visual outcomes. One of the reasons for this controversy is due to the definition of the time frame of “early surgery” in different reports. In the series by Woo et al., complete recovery of visual acuity occurred in all patients who underwent surgery within three days of symptom onset versus 83% in patients who underwent surgery after three days and complete or partial recovery of visual field defects occurred in 66% in those who underwent early versus 40% in those who underwent late surgery. Bill et al. noted complete recovery of visual function in patients operated within the first week of symptom onset. Randeva et al. noted that all patients operated within eight days had complete recovery of visual acuity versus 46% operated after eight days and visual field defects improved in 75% versus 23% in those operated before and after eight days, respectively. Muthukumar et al. studied the visual outcomes in patients presenting with unilateral or bilateral blindness and noted that recovery occurred even in patients with blindness when surgery was done within seven days. Similar results were noted by Agrawal and Mahapatra. However, Giritharan et al. found no difference in visual outcomes between early and delayed surgical interventions. Giritharan et al. also found no difference in visual outcomes between conservatively and surgically treated cohorts even though surgical cohorts tended to have more severe visual impairment preoperatively.
Outcome of extraocular palsies
Most studies have shown good to excellent outcomes for recovery of extraocular palsies both following surgical and conservative treatments.,, Singh et al. found that 100% of patients with extraocular palsies improved. Woo et al. found that sixth cranial nerve palsy recovered much earlier than other extra ocular palsies. Dubuisson et al. found that 85% of their patients with extra ocular palsies improved. Semple et al. found that 91% of patients with extra ocular palsies improved. Bill et al. and Muthukumar et al. found that extraocular palsies improved irrespective of the timing of surgery, that is, even patients who underwent delayed surgery improved., In the series by Semple et al., 91% of patients with extraocular palsies improved; however, there was no mention about the influence of timing of surgery on the outcome.
Is the outcome different in patients with infarction alone and hemorrhage alone?
The evidence in this regard is controversial. Semple et al. noted that patients who had infarction alone on histopathological examination had a more benign clinical course including less severe clinical signs, less incidence of visual symptoms, ophthalmoplegia, and the visual and extraocular palsy outcomes were much better in the infarction alone group. Interestingly, in their study, there was no difference in the endocrine outcome between the two groups. Contrary to the above study, a more recent study by Ogawa et al. found that ischemic PA had a more progressive course with worse outcomes in visual, extraocular palsy and endocrine. They attributed this due to the secondary edema of the necrotizing tissue with compression of the neighboring structures.
In a series of 66 patients reported by Semple et al., a mortality rate of 5% was noted. Singh et al., in a recent study of 87 patients, found a mortality rate of 4.5%. A recent study by Zhu et al. found a mortality rate of 1%. Decrease in mortality rates in recent studies may be due to increased awareness, lower thresholds for imaging and more aggressive management protocols instituted in recent times.
Recurrence of PA, though rare, is a known event and has been reported by several authors., Recurrent PA is well documented after conservative treatment of PA.,,, Recently, Hosman et al. studied 76 patients with PA and noted recurrence of PA in 5.6% of patients surgically treated. They found that greater Knosp grading was associated with higher frequency of recurrent hemorrhages probably because of the higher incidence of residual tumors in the cavernous sinus which were not resected. They also noted that among patients with residual tumors within the cavernous sinus recurrent hemorrhage occurred in 23.5% of patients and these recurrent PAs were associated with higher incidence of ophthalmoplegia which was often permanent and all patients with recurrent PA required hormonal replacement therapy during follow-up. Therefore, they concluded that cavernous sinus invasion and residual tumors are higher risks for recurrent PA. Interestingly, in their series of PA, 8% of cases occurred in patients with residual tumors after initial resection which indicates that tumor remnant is a risk factor for PA.
Intuitively, PA should lead to destruction of the tumor and/or pituitary gland and should not be associated with tumor recurrence. However, the converse is true. Acikogz et al. reported a patient who had two episodes of PA following which she developed an adenoma that warranted surgery. Randeva et al. found 6% incidence of tumor recurrence in patients with PA. Sibal et al. noted a 4% and 22% incidence of recurrence in surgically and conservatively treated groups, respectively, thereby implying that conservative treatment is associated with a higher risk of tumor recurrence. Pal et al. found a 11% risk of tumor recurrence during a mean follow-up of 81 months. In view of the above mentioned evidence, UK guidelines for PA suggested MRI follow up at 3–6 months after PA, annually for five years and once in two years thereafter.
Pituitary incidentalomas are known to occur in 1.6% of individuals undergoing MRI. In a large, recent series of 2,021 patients with pituitary adenomas, incidentalomas occurred in 4.6% of patients. Arita et al. studied 42 patients with incidental nonfunctioning adenomas and found that 9.5% developed PA during a mean follow-up of 5 years. A more recent study of all types of pituitary incidentalomas by Ishii et al. where 65 patients with incidentalomas were studied, 50% were nonfunctioning adenomas, and the overall incidence of PA in the study was 7.7%. Ishi et al. also concluded that patients with incidentally discovered nonfunctioning adenomas had a higher rate of PA than other types of tumors.
While majority of the PA occurs in patients with macroadenoma, there have been reports of PA in patients with microadenomas., Randall and Couldwell reported three patients with microadenomas and PA in whom the diagnosis was initially mistaken. In Zhu et al. series of 97 patients with PA, 4.1% occurred in microadenomas with a distinct male preponderance.
Controversy continues regarding the optimal management of patients with PA, viz., conservative versus surgical management. Such controversies can be laid to rest only by randomized control trials with matching cohorts in both the conservative and surgical arms. However, such RCTs are unlikely to be done in the foreseeable future in view of the risks involved to the patients and the medicolegal issues. Therefore, it becomes imperative for medical professionals who manage this condition to take a balanced view but keeping in mind the overall safety and outcome of the patient. We and others, have shown that even those who present with unilateral or bilateral blindness can recover serviceable vision in a significant percentage of patients provided the surgical intervention is done early, especially, within 1 week of the ictus. The improved outcome following surgical treatment has been confirmed by a recent meta-analysis of the literature on this subject by Tu et al. Moreover, as pointed out in the previous discussion, the incidence of tumor recurrence after PA is nearly 6 times in conservatively treated group when compared to the surgical group and the recurrence of PA is more frequent in patients treated conservatively than surgically.,,, In view of the above as well as the recent advances that have taken place in the surgical management of pituitary adenoma with low complication rates, this author is of the firm opinion that patients who present with visual compromise/extra ocular palsies or those with altered sensorium should be subjected to early surgical decompression as it has been shown to improve the visual outcome and to a lesser extent the endocrine outcome vis-à-vis conservative treatment.
PA often presents with acute onset of headache with or without visual, extraocular palsy, altered sensorium. MRI is the investigation of choice. It is important to rule out the presence of precipitating factors for PA. Most patients with PA are hypopituitaric at presentation and therefore will require emergent hormone replacement, management of electrolyte disturbances, and associated co-morbid conditions. The decision to proceed with conservative or surgical treatment should be tailored to each patient depending on the severity of clinical symptoms and signs, general condition of the patient, presence of co-morbid conditions, etc. If conservative management is chosen, close monitoring of the patient is warranted to identify worsening clinical status. Surgical management consists of transsphenoidal resection. Generally, visual acuity, fields and extraocular palsies improve in most patients. However, most patients require long term hormone replacement therapy. Recurrence of PA is rare but possible, especially, if there is a residual tumor. Recurrence of pituitary adenoma after PA is also rare but is still a possibility therefore surveillance imaging is required for several years., In recent years, the mortality from PA has decreased significantly. A high index of suspicion and prompt multidisciplinary management will often lead to an overall good outcome in patients with PA.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Figure 1], [Figure 2]