CSF leak has been reported to occur with nCPAP treatment following trauma to the skull base, but its association with OSA and nCPAP treatment in the absence of trauma has been reported only in 2 cases in the literature.1
The frequent side effects of nCPAP treatment are related to the CPAP on the patient's face and the upper airway. They include irritation of the nasal bridge and the eyes, rhinitis, dryness of the oropharyngeal mucosa, ear pain and epistaxis.2
Rhinorrhoea is a frequent complaint that is commonly worsened by nCPAP treatment occurring in 30–50% of patients with OSA.3
CSF leak can have a severe presentation, such as meningitis, or a milder presentation, such as orthostatic hypotension or clear rhinorrhoea.4
It is most frequently caused by trauma or surgery involving the skull base.5
Despite that, ‘spontaneous’ CSF leak can occur without an identifiable aetiology and its frequency ranges from 2–40% of patients.5
The bulk of spontaneous CSF leaks are associated with raised intracranial pressure (ICP) indicated by the occurrence of an arachnoidocele into the sella turcica.5
OSA may prompt the evolution of CSF leak. Obesity, a significant risk factor for OSA, is likewise a risk factor for CSF leak.6
In addition, apnoea can lead to an increase in ICP to levels reaching 30 or 40 cm H2
O. This increase in ICP is due to the fact that apnoea leads to hypoxaemia and hypercapnia, both of which stimulate the dilation of cerebral vessels leading to an increase in central venous pressure.7
Nonetheless, the prevalence of elevated ICP among patients with OSA is unknown.
nCPAP may affect the ICP in patients with OSA through various possible mechanisms. Positive pressure ventilation (PPV) decreases venous return and increases central venous and longitudinal vertebral venous pressures, both of which may increase ICP and CSF pressure.8
These undesired effects of PPV may be counterbalanced by its favourable effects on apnoea-related changes in ICP. In addition, nCPAP may decrease the transdural pressure gradient and lower the heights of transdural pressure fluctuations. In the absence of prior nasal endoscopy or surgeries, the dura and bony structures of the skull base form a dense barrier between the subarachnoid space and the neighbouring air-containing cavities. However, the cribriform plate due to its thin structure and consequent susceptibility to positive pressure has been blamed in the occurrence of CSF leak. In fact, most spontaneous CSF leaks are thought to be due to congenital fistulas across the cribriform plate.9
It is also thought that previous nasal surgeries or nasal scopes, as well as erosive processes of the dura, may induce the formation of these fistulas.9 10
Meanwhile, nCPAP may lower the threshold for a subclinical defect/communication between the subarachnoid space and the aerated spaces and consequently leads to the clinical manifestation of CSF leak.
The temporal relation between the initiation of nCPAP and the onset of rhinorrhoea in our patient suggests such a causal relation. Such a temporal relation was observed in one of the two previously reported cases of CSF leak in nCPAP users without previous instrumentation of the nasal sinuses or mastoid.1
Interestingly, this case provides further evidence that despite the non-specificity of clear rhinorrhoea in nCPAP users, it can still be the presenting sign of a serious underlying cause such as CSF leak.
- CSF leak is a rare complication in OSA patients.
- Clear rhinorrhoea is a frequent complication of nCPAP in patients with OSA and can be the only presenting sign of CSF leak.
- In patients with OSA who are considered for nCPAP treatment, a history of prior mastoid or sinus surgery should be identified and physicians should consider the diagnosis of CSF leak when these patients present with clear rhinorrhoea.