Discussion

In mammals, automatic respiratory drive originates from and is modulated by neuronal networks located in the pons and medulla oblongata.1 Central chemoreceptors, which respond to variations in blood CO₂ levels, are spread at the surface of the ventral medulla and allow minute ventilation adjustments to maintain PaCO₂ and pH homoeostasis.2 In addition to specific respiratory centres, the neighbouring solitary tract nucleus is crucial for the control of breathing, as it integrates information from peripheral and central receptors.2 Vigilance states have a strong effect on respiratory centre control.2 Thus, it is not surprising that cases of CHS secondary to brainstem damage have been reported following infarction, compression, brainstem tethering, tumour or infection. Bilateral injury of these areas results in rapid death. However, less extensive damage may be compatible with survival but causes respiratory centre dysfunction and central hypoventilation.

LM is a gram-positive bacillus that can cause opportunistic foodborne infections and is associated with a high mortality rate.3 Because of its ability to cross the blood-brain barrier, LM can infect the central nervous system (CNS) and lead to encephalitis, meningitis, cerebellar abscess or rhombencephalitis (or brainstem encephalitis).3 Although its exact prevalence remains unknown, LM rhombencephalitis is an uncommon infection in humans.3 Unlike other forms of listeriosis, predisposing factors are usually absent in patients developing rhombencephalitis, as was the case for our patient. After a non-specific prodromal phase, evidence of brain stem dysfunction appears, often leading to a fatal outcome, particularly when treatment is delayed.3 With proper early antibiotic therapy and supportive care, the mortality rate decreases to approximately 30%, with patients often displaying neurological sequelae. Although initial blood and cerebrospinal fluid samples may be negative for LM, cultures are eventually positive thereby delaying timely diagnosis. Brain CT may be normal or inconclusive but MRI typically reveals multiple abscesses in the brain stem and the cerebellum. Autopsy may show non-specific inflammation but usually confirms the presence of unilateral or bilateral multiple abscesses located in the brainstem and cultures are positive for LM.3 In our patient, neither cerebral CT scan nor MRI scans had been performed at the time of the initial infection but cerebral MRI scans obtained 34 years after the acute episode revealed two hypodense brainstem lesions, which argue against a congenital cause which would not exhibit these types of MRI abnormalities.4 Diagnosis and treatment may be delayed because symptoms are non-specific, blood and cerebrospinal fluid direct examination are often negative and CT or MRI may be inconclusive. This delay may lead to excess mortality, particularly in immunocompromised individuals. Fortunately, this was not the case in our patient.

Ataxic respiration during wakefulness and sleep, marked daytime alveolar hypoventilation without PHOX-2B mutation (known to cause congenital CHS1), a normal pulmonary function with extreme impairment of ventilatory response to CO₂, and MRI findings in our patient were consistent with a diagnosis of acquired central hypoventilation. One aspect of our case, however, was quite unusual compared to congenital CHS: the presence of clearly defined apnoeas and hypopneas. Given the history of severe rhombencephalitis, MRI brainstem lesions were considered to be a sequel of the LM abscess. The patient's brainstem lesions were unilateral, likely accounting for her extraordinary recovery. To our knowledge, this case is only the third report of LM brainstem lesions leading to acquired secondary central hypoventilation.5 ,6 The course and outcome of the acquired central hypoventilation was quite different for these three patients. The two previous patients were diagnosed with the syndrome during the acute infection,5 ,6 one died approximately 7 months later in spite of nocturnal ventilation,5 while the other, like our patient, survived but continued to require ventilator support during sleep.6 Our case is remarkable not only because of the positive outcome in spite of severe initial sepsis and CNS infection but also because of the extreme delay between the rhombencephalitis and the diagnosis and treatment of acquired central hypoventilation. Our patient presented for the first time to the Sleep Clinic with a suspicion of sleep apnoea approximately 15 years after the acute episode. Because of the patient's reluctance to pursue investigations and treatment, it took an additional 10 years to initiate non-invasive bi-level nocturnal ventilation and correct alveolar hypoventilation.

With the incidence of listeriosis rising, more cases of LM rhombencephalitis are likely to occur. Respiratory abnormalities are common during the acute phase of LM CNS infection and may be a cause of death.3 A subset of patients may apparently recover but display subclinical ventilatory control impairment due to brainstem lesions. As is the case with late-onset central hypoventilation, such patients are at risk for acute respiratory failure during anaesthesia or with concurrent pulmonary infection, opioid or sedative use.1 Thus, in the recovery phase of LM rhombencephalitis, respiratory status should be evaluated during sleep with overnight polysomnography, arterial blood gases and chemo-sensitivity testing.