Reversible cerebral vasoconstriction syndrome secondary to caffeine withdrawal

Introduction

Reversible cerebral vasoconstriction syndrome (RCVS) is a clinical–radiologic syndrome that is characterised by spontaneous or provoked severe recurrent thunderclap headaches with or without seizures, focal neurological deficits, and diffuse segmental constriction of cerebral arteries. The syndrome resolves spontaneously within three months. ¹ Several vasoactive drugs have been implicated as causative, and the syndrome can be precipitated with first ever exposure, after long-term use or acute withdrawal at normal or excessive doses.^2–5

Case

A 39-year-old man, a commercial heavy goods vehicle driver, was admitted with accidental superficial burns to his back which was managed conservatively with morphine analgesia and wound dressings. No vasoconstrictive medications were given. Two days after admission, he developed severe recurrent occipital ‘pressure like’ headache associated with vomiting and visual disturbance with a shimmering effect in both visual fields during the headaches. Time-to-peak headache was ≤1 min and symptoms initially persisted for few hours (subsequently 30 min) before gradually improving. He developed multiple similar thunderclap headaches over the following few days with no other focal neurological disturbance. There was no significant past medical history. His first cousin had a brain tumour, and his mother has pernicious anaemia. He smoked 10–20 cigarettes/day, drank alcohol 2–5 units/week and smoked an ounce of cannabis in a week. He was not taking any prescribed or over-the-counter medications.

On examination, he was alert and orientated. Pupils were equal and reactive to light. Fundoscopy, eye movements and visual fields were normal. The rest of neurological examination was normal. He was treated with oral morphine and Non-steroidal anti-inflammatory drugs (NSAIDS). Five doses of sumitriptan were given over the first three days after onset of headache and then stopped. Non-contrast computed tomography (CT) brain scan was normal. Blood investigations were unremarkable, and he refused lumbar puncture on multiple occasions. A magnetic resonance imaging (MRI) scan of head, eight days after onset of headache, showed no brain parenchymal abnormality.

On further questioning, it emerged that he had a significant daily caffeine intake. He drank four to five cans of high-energy drinks which contained high doses (250–270 mg/can) and around six large cups of coffee a day, amounting to a total caffeine intake of 1900 to 2000 mg/day. His MR cerebral angiogram showed diffuse segmental vasoconstriction (Figure 1(a)). He was advised to restart his usual caffeine immediately and to gradually reduce intake over the following six to eight weeks. His symptom resolved within a day after restarting oral caffeine and was discharged a day later. At three-month follow-up, he was ‘caffeine-free’. He had initially stayed off cannabis (after discussions re possible causative drugs) but had restarted his usual 1 ounce/week two weeks after discharge. His headaches and visual symptoms had resolved completely, and he remained asymptomatic at three months when a follow-up MR angiogram showed complete resolution of vasoconstriction (Figure 1(b)) satisfying the diagnostic criteria ⁶ for reversible cerebral vasoconstriction. The RCVS seemed most likely due to caffeine withdrawal as no other causative factor was identified. OPEN IN VIEWER

Discussion

Although RCVS can occur spontaneously, at least 50% of the cases have been known to be secondary to exposure to vasoactive medications (e.g. Selective serotonin re-uptake inhibitor (SSRI), Serotonin-norepinephrine re-uptake inhibitor (SNRI), pseudoephedrine, triptans) or postpartum. ¹ Several drugs have been implicated, and despite evidence of vasodilatory effects of caffeine on cerebral ⁷ and renal vasculature, this is the first reported case of RCVS secondary to caffeine withdrawal. In our patient, there was an overlap with cannabis, another potential vasoactive drug. In a prospective study ³ of 48 stroke patients, 10 out of 13 cannabis users were observed to have multifocal cerebral vasoconstriction; however, the dose variance of cannabis use was not documented. The temporal correlation between radiological features and changes to caffeine intake but not cannabis, with improved clinical features, is the basis for our conclusion. It is conceivable that an RCVS caused by cannabis withdrawal was here ‘treated’ by caffeine. However, that would represent a major coincidence – and might suggest caffeine as a very effective antidote to any episode of RCVS. The pathophysiologic mechanisms of RCVS are thought to be transient failure of regulation of cerebral arterial tone with sympathetic over-activity. No histologic changes pertaining to arterial wall inflammation have been found on biopsy specimens. ⁸ The deregulation is presumed to result from endothelial dysfunction, and/or oxidative stress, based on the significant overlap seen with posterior reversible encephalopathy syndrome. ¹ Treatment includes gradual withdrawal of inciting agents and supportive care. Calcium channel blockers improve symptoms but do not influence the cerebral vasoconstriction or complications.

Practical implications

Rigorous review of all prescribed medications is standard.