Reappraisal of haemorrhagic suprasellar pilocytic astrocytoma during adulthood
- 1 Department of Neurosurgery, Lyerly Neurosurgery, Jacksonville, Florida, USA
- 2 Research Department, Jacksonville University, Jacksonville, Florida, USA
- Correspondence to Dr Ricardo Hanel; rhanel@lyerlyneuro.com
Abstract
Pilocytic astrocytomas comprise the most common central nervous system tumour during childhood and have an excellent response to surgical treatment in this population. The tumour incidence decreases with age, whereas more aggressive behaviour tends to increase. Haemorrhage as a presenting feature of pilocytic astrocytomas is a rare phenomenon, especially in the adult population. We present a case of a 55-year-old patient with progressive headaches and dizziness. MRI confirmed a sellar and predominantly retrochiasmal suprasellar lesion with heterogeneous signal, enhancement and blood products. Management via transsphenoidal approach was performed, and histopathology revealed the unexpected diagnosis of haemorrhagic pilocytic astrocytoma. Haemorrhagic pilocytic astrocytoma is an infrequent entity in the adult population and it is essential to recognise the peculiarities regarding diagnostic evaluation and management, which differ from the paediatric population. During adulthood, this tumour carries an overall unfavourable prognosis, with higher rates of progression and recurrence.
Background
Pilocytic astrocytomas (PAs) are the most common primary central nervous system tumour during childhood, with a mean age at diagnosis varying between 6.8 and 7.7 years old. They tend to have an indolent progression and a thriving treatment response to surgical resection.1 2 The presentation of this subtype of astrocytoma, however, is not common later in life, with a decrease in incidence with age.3 Adding tumour-associated bleeding to the equation restricts the number of cases described in the literature.4
We describe a distinct case of a man on his fifties who presented with a suprasellar mass and underwent transsphenoidal resection. Histopathology revealed the unexpected diagnosis of a suprasellar haemorrhagic PA. An overview of the literature is discussed.
Case presentation
A 55-year-old man presented to an outside hospital complaining of progressively worsening headaches and dizziness for the past 6 weeks. The headaches would worsen when standing or looking up to the ceiling. He also had occasional episodes of blurry vision with spontaneous resolution, but denied issues with his peripheral vision. On review of the systems, there were no complaints of weakness, paresthesia, excessive thirst or polyuria. The remainder of the neurological examination was normal. Medical history was relevant for uncontrolled hypertension due to long-term non-compliance with prescribed medication.
Investigations
Initial head non-contrast CT, performed in the emergency department, demonstrated a heterogeneously attenuating, partially haemorrhagic suprasellar lesion. It measured approximately 2.6×1.7×1.3 cm and concomitant hydrocephalus was noted, most likely related to mass effect (figure 1). No expansion or remodelling of the bony sella was identified. CT angiography of the head revealed mild local mass effect on the circle of Willis structures, but no associated vascular abnormality. Chest X-ray and initial basic laboratory evaluation were unremarkable.
Axial (A), sagittal (B),and coronal (C) non-contrast CT head images demonstrate a heterogeneously attenuating sellar and suprasellar lesion with hyperdensity centrally, presumed to reflect intratumoural haemorrhage. Note dilatation of the lateral ventricular temporal horns bilaterally, consistent with hydrocephalus.
The patient was transferred to our hospital for further clinical evaluation by neurosurgery and endocrinology. MRI demonstrated a sellar and predominantly retrochiasmal suprasellar lesion with heterogeneous signal, enhancement and blood products (figure 2). The possibility of normal pituitary tissue along the sellar floor was raised. Findings favoured craniopharyngioma rather than pituitary adenoma as possible diagnoses. Neurological examination remained unchanged, with fluctuating headache from mild to none. He had also started on corticosteroids, secondary to optic chiasm impingement.
Non-contrast coronal T2 sequence demonstrates the sellar and suprasellar lesion with central T2 hypointense signal, correlating with blood products and peripheral T2 hyperintense signal (A). Precontrast sagittal T1 sequence demonstrates a lobulated sellar and retrochiasmal suprasellar lesion with heterogeneous T1 signal (B). Postcontrast depicts enhancing pituitary tissue displaced along the floor of the sella turcica, with heterogeneous enhancement throughout the lesion (C).
Detailed preoperative laboratory evaluation showed mild low testosterone (233.4 ng/dL), low Follicle-Stimulating Hormone (3.9 mIU/mL), and low Luteinizing Hormone (2.7 mIU/mL) suggestive of secondary hypogonadism. Decreased cortisol (3.7 μg/dL) and Adrenocorticotropic Hormone (<5 pg/mL) levels were present, which was assumed to result from dexamethasone use before labs were drawn. Prolactin (12.9 ng/mL), testosterone (233.4 ng/dL), Thyroid-Stimulating Hormone (1.57 μIU/ml), free T4 (0.9 ng/dL), Human Growth Hormone (0.52 ng/dL) and Insulin-like Growth Factor-1 (233 ng/dL) were unremarkable. A non-functional mass was, therefore, favoured.
Treatment
Given the history of progressive headaches, endocrine and radiological findings, surgery was considered and discussed with the patient. Because the lesion was sited at midline and there was no major involvement of adjacent structures, we opted for a transsphenoidal approach of the lesion. A preoperative lumbar spinal drain was placed. Endoscopic endonasal access was performed after extensive drilling of the sella, to gain access to suprasellar and retrosellar compartments. After dural incision and access to the tumour, the pathology sample was undertaken and sent to analysis, while we proceeded with tumour excision. Preliminary frozen intraoperative report favoured astrocytoma rather than craniopharyngioma. In either case, we decided that total resection was the best aim. After excision of the tumour, a cerebrospinal fluid (CSF) leakage became evident and was managed by multilayer closure involving a nasoseptal flap. No major intraprocedural complications were identified.
Outcome and follow-up
Pathology demonstrated glial neoplasm with Rosenthal fibres and eosinophilic granular bodies (figure 3). The tumour was composed of monotonous spindle cells and round oligodendrocyte-like cells within microcystic areas. Hyalinised blood vessels and fragments of organising thrombus were present. There was no evidence of necrosis. Immunohistochemical stains for Glial Fibrillary Acidic Protein (GFAP), synaptophysin and S100 were positive. Neurofilament demonstrated staining at the periphery of two fragments. Isocitrate dehydrogenase-1 (IDH-1) R132H was negative and ATRX showed intact nuclear staining by immunohistochemistry. Mitotic figures were inconspicuous, and the Ki-67 proliferation index was low. The findings were consistent with a suprasellar PA, WHO grade I. Molecular studies for BRAF gene rearrangement, and BRAF mutations were negative.
H&E-stained sections show a biphasic tumour with mildly cellular areas comprised of piloid cells and hyalinised blood vessels (A) and areas with oligodendrocyte-like cells containing microcysts and eosinophilic granular bodies (B). Focally, Rosenthal fibres are frequent (C). The tumour is diffusely GFAP positive (D) and has only focal neurofilament staining (E). The Ki-67 proliferation index is low (F).
Postoperative course was complicated by right hemiparesis, with subsequent dysphagia and urinary retention. The patient also experienced acute encephalopathy, persistent hydrocephalus and adrenal insufficiency, which were successfully managed. A postoperative MRI revealed anterior left thalamic acute infarction and debulking of the previously identified suprasellar mass. There was a thin peripheral rim of enhancement along the anterior margin of the suprasellar resection cavity, which was also present on the preoperative MRI and could reflect a small volume of residual tumour (figure 4). The patient was discharged for rehabilitation with a remnant right hemiparesis, but an improvement of his dysphagia was noted.
Immediate postoperative axial MR images demonstrate thin peripheral rim of enhancement along the anterior margin of the suprasellar resection cavity on T1-weighted (A) and acute anterior infarct of let thalamus on DWI (B). Three weeks follow-up, confirming large debulking on sagittal flair T1-weight images (C).
Discussion
We report a case of a patient with progressive headache and essentially unremarkable medical history. Imaging studies revealed a haemorrhagic suprasellar mass, and initial differential diagnosis was craniopharyngioma. The patient underwent a transsphenoidal approach, and histopathology returned the unexpected diagnosis of PA.
PAs are benign WHO grade 1 neoplasm, with an annual incidence during childhood varying from 4.2 to 8.4 for every hundred thousand patients.2 In this population, this tumour accounts for approximately 18% of all primary brain tumours and are most commonly located in the cerebellum. They have an overall good prognosis, with a 10-year survival rate over 90%.1 2 It is distinctly uncommon for PA to present in the late decades of life.5 The average age of presentation in adults is 32.2 (SD 9.4), with a progressive decrease in incidence with age.3 4 In contrast to the paediatric population, the 5-year survival rate goes from 89.9% in patients>20 years old to 59.6% in patients older than 60, with age significantly associated with hazard of death in univariate analysis (HR 1.5, 95% CI 1.4 to 1.6, for each 10 years of age).3 During adulthood, PAs are predominantly extracerebellar, with the most substantial proportion located supratentorially. They carry an overall poor prognosis, with associated increased rates of recurrence, progression, malignant transformation and leptomeningeal dissemination.6 In particular, leptomeningeal seeding is commonly associated with a primary lesion located near a CSF space, mainly when situated in the sellar region. Additionally, surgical manipulation, haemorrhage and features of anaplasia are also considered risk factors for leptomeningeal dissemination.7 8
Tumour-associated haemorrhage is scarcely seen in low-grade astrocytomas, rendering it challenging to determine how it might influence patient outcomes. It was postulated that bleeding risk in this subset of low-grade tumours might be related to an abnormal vasculature within the tumour and local metabolic factors.9 The exact bleeding risk cannot be inferred with the available literature. However, the risk for a haemorrhagic transformation seems to be equal between the paediatric and adult populations (ratio 1.1:1).10 In a review of radiologic aspects, Kikuchi et al described a 10% prevalence of intratumoural haemorrhage in PAs and suggested perfusion dynamic susceptibility MRI cut-off values that would corroborate in the differentiation of PA from high-grade glial tumours.11 Using an ROC-AUC analysis, they found that optimal diagnostic cut-off values of ≤1.08 for regional cerebral blood flow (rCBFmax), ≤4.08 for relative cerebral blood volume (rCBVmax), and ≥1.38 x 10–3 mm2/s for apparent diffusion coefficient (ADCmin) would favour a PA diagnosis. Rather haemorrhagic or not, PA may occupy the suprasellar region when they arise from the pituitary gland or when they arise from adjacent structures and invade it by extension.12 The overall prevalence in this region range from 7% to 17% in the adult population.4 13 Not many cases have been reported in which the origin of the tumour was proven to be from the pituitary gland itself. Parish et al reported two of these cases, which were managed with a transsphenoidal approach.12 They were complicated by a CSF leaking, a recognised possible complication of this approach, but that can be avoided with the suitable use of a pedicled nasoseptal flap, as in our case.
Radiological features are paramount for diagnostic evaluation and treatment planning. From this standpoint, different factors have to be considered, such as the epicentre of the lesion, detailed imaging characteristics (including solid, cystic or mixed components), presence of sellar remodelling, calcification, haemorrhage, vascular features and enhancement characteristics. In an adult with a heterogeneous sellar or suprasellar multilobulated lesion felt to arise separately from the pituitary gland, the primary differential diagnosis favours craniopharyngioma.14 The papillary subtype presents with solid and enhancing components, commonly seen in the fifth and sixth decade of life; differing from adamantinomatous subtype, which is seen earlier in life. Given our patient demographics, imaging features and difficulty in determining the exact origin of the lesion, additional differential diagnosis was a macroadenoma, as well as less common entities like pituicytoma and granular cell tumour.15 16 PAs commonly present radiologically as a cystic mass with characteristic postcontrast enhancement.5 In low-grade astrocytomas, local oedema is not a common finding but may become prominent in the occurrence of haemorrhagic transformation.17 Imaging findings require from upfront a high level of suspicion given their most non-specific findings and variable baseline characteristics of these patients. Radiological features are reviewed in table 1.
Suprasellar pilocytic astrocytoma—radiological features and differential diagnosis
Suprasellar pilocytic astrocytoma—radiological features and differential diagnosis
Until a few years ago, every astrocytic tumour that presumably arises from the posterior pituitary was considered a pituicytoma, and the terms were used interchangeably.18–20 Notwithstanding, more recent pathological studies showed that, although they share similarities, pituicytomas lack Rosenthal fibres, eosinophilic granular bodies, and other distinctive characteristics commonly seen on astrocytomas.15 Attempts have been made to distinguish the two entities via imaging, with a tendency for a faster enhancement after injection of contrast and a tail of enhancement to the pituitary stalk described with pituicytomas.18 19 Practically, the lesions are very difficult to differentiate through imaging. Other common differential diagnoses, such as craniopharyngioma and suprasellar PA, may also be indistinguishable on imaging.14
Both craniopharyngioma and PA have surgical resection as first-line treatment, with surgery frequently performed with diagnostic and therapeutic intent. Gross total resection is the only treatment with evidence of improving survival in newly diagnosed adults’ patients.6 In this age group, adjuvant radiotherapy may be considered when the lesion is not amenable to complete resection.21 Radiation therapy, however, was not found to impact overall survival, and there may be an association with an increased risk for malignant transformation.3 22 23 The exact radiation technique also interferes with the potential benefits and risks.24 25
To the best of our knowledge, seven cases of haemorrhagic PA after the fifth decade of life had been described in the literature (table 2), but only one of them was in the suprasellar region.9 17 26–30 The patient similarly presented with a history of headache and visual disturbances, arriving obtunded on the emergency department.28 CT was performed addressing a hyperdense suprasellar mass and hydrocephalus, as in our patient. He underwent craniotomy with subtotal resection of the lesion. We, conversely, opted for a transsphenoidal approach in our patient once the suprasellar lesion was sitting at midline and there was no major involvement of adjacent structures. Injury to thalamic perforator vessels, a complex and variable irrigation system, may explain the postoperative right-sided hemiparesis.31 Overall, though, transsphenoidal and transcranial approaches demonstrated similar safety profiles and rates of total gross resection, which may vary according to the exact type of tumour.32 33 Although some may argue that an initial craniotomy approach, instead of transsphenoidal, could have allowed total resection and avoided the ischaemic complication, there is no robust data that supports these statements. According to the National Surgical Quality Improvement Program, postoperative complications occurred in one-third of patients undergoing craniotomy for craniopharyngioma resection.34 Jeswani et al found similar surgical outcomes between the groups, except that more CSF leak was seen in the transsphenoidal group, whereas more neurological injuries occurred in the craniotomy group.35 Also analysing patients with craniopharyngiomas, a meta-analyse by Komotar et al elucidated a higher rate of total resection in patients who underwent a transsphenoidal approach (p=<0.003), with no differences in the rate of stroke between the groups.36 Wannemuehler et al also described similar surgical outcomes between the two approaches.37 Additionally, rates of recurrence tend to be lower with tumours managed via transsphenoidal approach.32 33 There is a paucity of data regarding surgical approach to suprasellar PA, therefore, the decision for the approach must aim for lower morbidity risk and must be evaluated on a case-by-case basis.
Haemorrhagic pilocytic astrocytoma in patients older than 50 years old
| Author | Year | Age/sex | Clinical presentation | Location | Treatment | Comments | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| F, female; GTR, gross total resection; ICH, intracranial haemorrhage; IVH, intraventricular haemorrhage; M, male; N/A, not applicable; OB, open biopsy; STR, subtotal resection. | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Lones30 | 1991 | 69/F | Left side weakness | Cerebral hemisphere | N/A | Acute IVH and death before any intervention | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Sorenson26 | 1995 | 58/F | Transient amnesia | Hypothalamic | OB | Memory impairment following biopsy | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Lyons27 | 2007 | 75/M | Aphasia | Temporal region | GTR | Two previous episodes of ICH associated with warfarin |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| White9 | 2008 | 58/F | Headache and ataxia | Pons | GTR | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Sekula28 | 2008 | 65/M | Headache and visual changes | Suprasellar | STR | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Kim17 | 2011 | 53/M | Headache | Cerebellum | GTR | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Sun29 | 2018 | 62/M | Headache and ataxia | Cerebellum | GTR | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Footnotes
-
Contributors GC: substantial contributions to the conceptualisation of the work, data acquisition and interpretation; drafting and revision of the work. AM: substantial contributions to the conceptualisation of the work, data acquisition and interpretation. BL: substantial contributions to the conceptualisation and design of the work. RH: original idea, data supervision and interpretation; revision of the work, approval of the final version; agreement to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.
-
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
-
Competing interests RH, senior author of the above-mentioned manuscript, is consultant for Covidien, Stryker, Codman and MicroVention.
-
Patient consent for publication Obtained.
-
Provenance and peer review Not commissioned; externally peer reviewed.
- © BMJ Publishing Group Limited 2020. No commercial re-use. See rights and permissions. Published by BMJ.
References
Use of this content is subject to our disclaimer