Search tips
Search criteria 


Logo of jrsocmedLink to Publisher's site
J R Soc Med. 2002 October; 95(10): 501–502.
PMCID: PMC1279176

Meningitis four years after treatment of macroprolactinoma

Rachel Wood, BM MRCP, Derek D Sandeman, MB FRCP, Mary L Gawne-Cain, MD FRCR,1 and Richard I G Holt, PhD MRCP2

After treatment of macroprolactinoma, patients are advised to look out for cerebrospinal fluid (CSF) rhinorrhoea. But meningitis can develop without this warning sign.


A man aged 33 was admitted with pneumococcal meningitis. Four years previously the occurrence of three generalized convulsions had led to a diagnosis of macroprolactinoma. An MRI of his brain at that time showed hydrocephalus and a large heterogeneously enhancing pituitary mass with suprasellar expansion and extension into the sphenoid sinus (Figure 1). His serum prolactin was 100 000 U/L. Further pituitary testing indicated that he had secondary hypothyroidism and hypogonadism, for which thyroxine and testosterone were prescribed.

Figure 1
MRI coronal view of brain showing hydrocephalus and a large heterogeneously enhancing pituitary mass with suprasellar expansion and extension into sphenoid sinus

For the prolactinoma he was treated for one month with bromocriptine 7.5 mg daily and subsequently with cabergoline 0.5 mg twice weekly. After three months on this regimen, serum prolactin was 1800 U/L (normal < 450), so the dose of cabergoline was increased to 0.5 mg three times weekly. The dose of cabergoline was finally increased to 1 mg twice weekly twelve months after diagnosis, and serum prolactin then became normal. A repeat MRI scan at thirty months showed a large reduction in tumour size. It was noted that the sphenoid sinus was destroyed and occupied by CSF. He never required neurosurgical intervention.

During the two weeks before his current admission he developed flu-like symptoms and occasional mild headache. On the day of admission, he woke with a severe frontal headache, which was not relieved by simple analgesia and was associated with nausea, dizziness and mild neck stiffness. Glasgow coma score was 15 and there were no signs of meningism. His temperature was 37.1°C. No abnormalities in his cranial or peripheral nerve examination were found. Initial blood testing revealed a neutrophilia—white cells 15.39/L, neutrophils 13.79/L. After treatment with intravenous morphine in the accident and emergency department the symptoms completely resolved. He was admitted for observation with a working diagnosis of postviral headache, but in view of his pituitary tumour the possibilities of a pituitary haemorrhage or infarction were considered. Later that evening he had a generalized convulsion, after which he was drowsy and sweaty. He became hypotensive and tachycardic. A CT scan showed a 3.5 × 7 cm mass in the pituitary fossa, with calcification of the superior aspect and no haemorrhage. Cerebrospinal fluid protein was 5.3 g/L, white cells 147 / μL, glucose < 0.5 mmol/L; Streptococcus pneumoniae was subsequently cultured from cerebrospinal fluid and blood. It was a penicillin-insensitive strain with penicillin minimum inhibitory concentration 0.75 mg/L and cefotaxime 0.19 mg/L. He was treated with 3 g cefotaxime 4-hourly and rifampicin 600 mg once daily for 14 days and made a full recovery. Review of his CT scans showed extensive bony erosion of the sphenoid sinus and it was felt that the source of his infection was a sinus connecting his pituitary fossa with the sphenoid sinus (Figure 2). There had never been any rhinorrhoea.

Figure 2
CT transverse view showing destruction of the sphenoid sinus


The pneumococcal meningitis in this patient was probably caused by a defect in the pituitary sella floor. The case highlights several issues that are important in the management of patients with macroprolactinomas. First, the complication of meningitis occurred several years after the start of treatment of his prolactinoma. This, together with the insidious nature of the presentation followed by a rapid decline, meant that the cause of his illness was not immediately apparent. Secondly, although CSF rhinorrhoea is a recognized complication after neurosurgery or dopamine agonist therapy, the absence of a history of rhinorrhoea did not exclude a defect in the skull base. Finally, it is unclear what measures should be taken to prevent meningitis in patients with macroprolactinomas.

The dopamine agonist cabergoline is an effective and well-tolerated treatment for macroprolactinoma. It is now regarded as first-line therapy but may also be useful in bromocriptine-resistant tumours. CSF rhinorrhoea is a rare but potentially lethal complication following primary dopamine agonist treatment with bromocriptine1 or cabergoline2. Dopamine agonist therapy causes a rapid shrinkage of the tumour and may produce a defect in the sella floor, particularly when dopamine agonists are used to treat invasive prolactinomas with skull-base destruction. These defects predispose to bacterial meningitis. In a review of 15 patients with bromocriptine-induced CSF leaks3, 8 patients had developed leaks within one month of the start of treatment and the remainder occurred by seventeen months. 4 individuals developed meningitis. The most likely organisms to cause meningitis in this setting are upper respiratory tract pathogens—typically S. pneumoniae, Neisseria meningitidis and Haemophilus influenzae. Consequently, patients with large macroprolactinomas under treatment with dopamine agonists should be made aware of the symptom of CSF rhinorrhoea and the risk of meningitis.

The absence of a known CSF leak does not exclude the diagnosis of a skull-base defect. This has been noted before4. Although our patient had never reported rhinorrhoea, a skull-base defect was presumed to be the portal of bacterial entry. The interval of 4 years between the start of treatment and the development of meningitis is the longest yet reported. What preventive measures can be adopted? Where a CSF leak is proved and surgical repair is not feasible, vaccination against S. pneumoniae, H. influenzae and meningococci has been recommended. We suggest that this strategy might be extended to all patients treated for invasive macroprolactinoma. As regards treatment for meningitis in patients with a known CSF leak, the recommendation is to start with a third-generation cephalosporin and vancomycin pending the results of culture, because of the increasing prevalence of penicillin-resistant pneumococci, as seen in this case.


1. Kok JG, Bartelink AK, Schulte BP, et al. Cerebrospinal fluid rhinorrhea during treatment with bromocriptine for prolactinoma. Neurology 1985;35: 1193-5 [PubMed]
2. Leong KS, Foy PM, Swift AC, Atkin SL, Hadden DR, MacFarlane IA. CSF rhinorrhoea following treatment with dopamine agonists for massive invasive prolactinomas. 2000;52: 43-9 [PubMed]
3. Barlas O, Bayindir C, Hepgul K, et al. Bromocriptine-induced cerebrospinal fluid fistula in patients with macroprolactinomas: report of three cases and a review of the literature. Surg Neurol 1994;41: 486-9 [PubMed]
4. Davis JR, Sheppard MC, Heath DA. Giant invasive prolactinoma: a case report and review of nine further cases. Quart J Med 1990;74: 227-38 [PubMed]
5. Moss PJ, Graham JC, McKendrick MW. Cerebrospinal fluid rhinorrhea and group B streptococcal meningitis. Clin Infect Dis 1998;27: 411-12 [PubMed]

Articles from Journal of the Royal Society of Medicine are provided here courtesy of Royal Society of Medicine Press