A Case of Acute Myeloid Leukaemia Presenting as Subacute Intracranial Hypertension in a Young Adult

ABSTRACT

We report a case of an 18-year-old woman presenting with headache, papilloedema, and cerebrospinal fluid (CSF) pleocytosis. She was subsequently diagnosed with acute myeloid leukaemia, which is to date the only reported case manifesting as central nervous system-localised disease in an adult. The intracranial hypertension was treated successfully with chemotherapy, acetazolamide, and CSF drainage, with no permanent visual impairment. The mechanism by which haematological malignancy causes intracranial hypertension is not fully elucidated, but we hypothesise that in our case, blast infiltration interfered with CSF reabsorption at the arachnoid granulations.

The purpose of this report is to demonstrate a case of acute myeloid leukaemia (AML) in an 18-year-old woman presenting with subacute intracranial hypertension and cerebrospinal fluid (CSF) pleocytosis. Ocular manifestations of acute leukaemia have been reported commonly¹; however, the majority of these cases are due to acute lymphoblastic leukaemia (ALL), not AML. We hope to broaden the differentials of a clinician who is confronted with a patient presenting with unexplained papilloedema and CSF pleocytosis and emphasise the need for prolonged follow-up until a formal diagnosis can be made.

Case report

A previously healthy 18-year-old woman presented to the emergency department with a 6-week history of bi-frontal headache, unresponsive to simple analgesia. She had some mild neck stiffness, but no photophobia, fevers, or altered cognitive state. Her physical examination was normal, including body mass index. Both optic nerve heads were noted to be prominent but considered to be clinically within normal limits using slit lamp biomicroscopy. The rest of the ocular exam was unremarkable. Visual acuity was 6/4 bilaterally. Her pupil exam was normal with no relative afferent pupillary defect. The intra-ocular pressures were 15 mmHg in each eye. She had normal computerised perimetry on Humphrey visual fields, and her motility examination was also normal. She had been investigated with a lumbar puncture two weeks prior in a different centre. This had shown a normal CSF opening pressure and CSF pleocytosis (248 mononuclears, 5 polymorphs, 2 RBCs, nil others). Peripheral blood count and film were normal. At that time, a presumptive diagnosis of viral meningitis was made, and she was observed with no active management.

Optical coherence tomography (OCT) (measured with Zeiss Cirrus, Cirrus HD-OCT Model 4000, Zeiss, Germany) demonstrated mild thickening of the mean retinal nerve fibre layer (RNFL) in both eyes (131 microns right eye, 133 microns left eye). Magnetic resonance imaging (MRI) of the brain and orbits revealed abnormal distension of the optic nerve sheaths, flattening of the posterior sclera bilaterally, and protrusion of the optic discs (Figure 1a and b). Magnetic resonance venogram ruled out venous obstruction. Peripheral blood vasculitic screen was normal except for mild toxic granulation on the peripheral blood smear. Repeat lumbar puncture demonstrated an opening pressure of 25 cmH₂O and a persistent, though improved, pleocytosis (141 mononuclears, 31 polymorphs, nil RBCs, nil others), with mildly raised protein (0.48 g/l). The extended culture was negative, including assays for Enterovirus, Herpes Simplex Virus, Varicella Zoster Virus, Epstein Barr Virus, Cytomegalovirus, Toxoplasma, Bartonella, Tuberculosis, and Cryptococcus. The CSF was sent for cytology which revealed a highly cellular fluid composed of numerous lymphoid cells together with a few monocytes, neutrophils, and eosinophils. Scant band, immature myeloid, and monocytic forms were detected but were thought to be non-specific and non-diagnostic. Flow cytometry did not detect an abnormal blast population.

Figure 1.

Axial (a) and coronal (b) T2 MRI of the brain demonstrates abnormal distension of the optic nerve sheaths, flattening of the posterior sclera bilaterally, and protrusion of the optic discs. Post-gadolinium sagittal T1 MRI of the brain and cervical spine (c) demonstrates leptomeningeal enhancement.

On review two weeks later, her headache had worsened. Vision remained intact, including visual acuity and visual field, whereas the optic discs were now clearly abnormal. The swelling had progressed, with mean RNFL now at 250 microns right eye and 262 microns left eye. MRI of brain and spine with gadolinium demonstrated mild leptomeningeal signal changes and enhancement over the cerebral surface and spinal cord (Figure 1c). Lumbar puncture showed an opening pressure of >34 cmH₂0. Cytology of the fluid demonstrated numerous lymphocytes with scattered monocytes, erythrocytes, a few neutrophils, eosinophils, and small numbers of blast cells (<5%). Flow cytometry of the spinal fluid revealed an abnormal blast population expressing CD45dim/33/117, consistent with myeloblasts. The blood film now showed 0.9 × 10⁹/l blasts that had an identical immunophenotype as those observed in the CSF sample. The diagnosis of AML was confirmed with a bone marrow biopsy that demonstrated 32% myeloblasts, a normal karyotype, and the presence of NPM1 mutation on PCR testing.

She received intensive induction chemotherapy with cytarabine (1.5 g/m² BD days 1, 3, 5, and 7) and idarubicin (12 mg/m² days 1, 2, and 3) with intrathecal cytarabine and methotrexate. Acetazolamide, intravenous pulsed methylprednisolone, and weekly CSF drainage by lumbar puncture were administered for management of the increased intracranial pressure. The week following the diagnosis of AML, CSF pressure remained >34 cmH₂O and her visual acuity deteriorated to 6/9 right eye 6/6 left eye. Fundoscopy revealed increased nerve disc swelling with surrounding cotton wool spots and flame haemorrhages. OCT confirmed a thickened mean RNFL to 409 microns right eye and 401 microns left eye (Figure 2). The frequency of CSF drainage and dosage of acetazolamide were increased, and the patient was considered for optic nerve sheath fenestration. However, the optic disc swelling stabilised over the following week and then slowly resolved (Figure 3). Within three weeks of commencing induction chemotherapy, blast cells were no longer detected in the CSF, and the pressures had dropped to 22 cmH₂O with ongoing reduction in papilloedema and headache.

Figure 2.

Optical coherence tomography demonstrating bilateral RNFL swelling.

Figure 3.

Mean RNFL thickness following the diagnosis of AML. The timing of initiation of chemotherapy (A), blast-free CSF (B), resolution of meningeal gadolinium enhancement (C), and therapeutic CSF drainage (∎) are demonstrated.

Six months after initial diagnosis, the patient had completed three cycles of consolidation chemotherapy (cytarabine 3 g/m² BD days 1, 3, and 5). Visual acuity returned to 6/4 bilaterally, and the optic discs were no longer swollen (OCT mean RNFL: 109 microns right eye and 120 microns left eye). Lumbar puncture revealed an opening pressure of 15 cmH₂O and the CSF remained blast-free. Repeat MRI of brain demonstrated resolution of the meningeal gadolinium enhancement. The acetazolamide was weaned (from 1 g TDS to 500 mg BD) over a 6-month period.

Discussion

This case is an extremely rare presentation of AML, initially manifesting as purely central nervous system-localised disease with intracranial hypertension. Although ocular involvement is occasionally observed in ALL¹, the blast morphology, flow cytometry, and genetic mutation profile in this case were characteristic of AML. Papilloedema is recognised in the management of acute leukaemia but is more typically observed as a manifestation of disease relapse^2,3 or a side effect of drug therapy.⁴ To our knowledge, this is the first reported case of papilloedema as the initial presenting sign of previously undiagnosed AML in an adult. It highlights the importance of thorough analysis and follow-up of patients with a raised CSF pressure and pleocytosis with no identifiable cause.

Several case reports of leukaemia-induced intracranial hypertension suggest that the mechanism of the intracranial hypertension is possibly hyperviscosity secondary to leucocytosis, resulting in poor CSF reabsorption⁵, but our patient did not have a high systemic white cell count, indicating an alternative mechanism. Another case reported persistent increased intracranial pressure after the clearance of CSF blasts, suggesting the malignant cells themselves were not responsible and that perhaps an immune-mediated or inflammatory reaction altered CSF flow.²

We postulate that the blasts directly infiltrated and interfered with CSF reabsorption across the arachnoid granulations, resulting in increased CSF pressure. This is in keeping with a case report describing diffuse meningeal infiltrates discovered on autopsy of an adult suffering from intracranial hypertension and subsequently diagnosed with a terminal undefined acute leukaemia.⁶ Our case supports this mechanism with demonstration of meningeal gadolinium enhancement on MRI. Intracranial pressure remained elevated despite steroids, acetazolamide, and regular CSF drainage. Induction chemotherapy successfully eliminated blast cells and the CSF pressures reduced simultaneously, again suggesting malignancy as the underlying mechanism.

Funding Statement

The authors report no funding was received.

Declaration of interest

The authors report no conflict of interest.