CASE REPORT

A 54 year old man presented with fever, altered mental status, and seizure. He had a history of alcohol and tobacco abuse but no other previous illnesses, and he was not taking any medications. After emergency intubation, examination revealed normal breath sounds and right upper extremity paralysis. The chest radiograph and brain CT scan were normal. The cerebrospinal fluid (CSF) contained 180 white blood cells/mm³ with 39% neutrophils, 55% lymphocytes, a glucose level of 103 mg/dl, and a protein level of 81 mg/dl. The CSF immunoglobulin M (IgM) for WNV was positive. Other laboratory studies were unrevealing.

On day 7 the patient showed significant clinical improvement and his mental status returned to normal, but he still remained ventilator dependent. On days 14 and 75 electrodiagnostic studies were performed (table 1). Initially they revealed reduced motor responses and preserved sensory responses with denervation potentials (positive sharp waves, fibrillation potentials, and fasciculation potentials) and decreased recruitment of motor units, consistent with the involvement of either motor axons or anterior horn cells. The sensory abnormalities that developed over the course of his illness may have reflected an additional component of critical illness neuropathy.

By day 37 the patient was clinically stable but he had persistent thick respiratory secretions and the chest radiograph showed left lower lobe atelectasis. Despite bronchoscopy, atelectasis persisted and nebulised 3% saline was administered every 4 hours with a resulting subjective reduction in the viscosity of the secretions. Theophylline was started on day 51 potentially to improve diaphragmatic strength.

Throughout the course of the illness the patient experienced occasional fevers with associated purulent respiratory secretions. Cultures of tracheal aspirates consistently grew Pseudomonas aeruginosa despite multiple courses of appropriate intravenous antibiotics. On day 67 nebulised colistimethate sodium (Coly-Mycin M Parenteral; Monarch Pharmaceuticals, Bristol, TN, USA) 150 mg twice daily was started. The fevers subsided soon thereafter and subsequent respiratory cultures failed to grow Pseudomonas.

During his stay in hospital the patient remained ventilator dependent and a tracheostomy was performed on day 16. Objective measures of respiratory muscle strength were consistently reduced with a negative inspiratory pressure of –10 cm H₂O on day 5 and −8 cm H₂O on day 80. Phrenic nerve conduction studies were performed on day 109. Bilateral transcutaneous electrical stimulation of the phrenic nerve with surface recordings of diaphragmatic activity failed to elicit a motor action potential on either side, suggesting bilateral diaphragmatic paralysis. On day 112 the patient still required full mechanical ventilatory support and was discharged to a long term facility.

DISCUSSION

Approximately one of 100 WNV infections in humans results in clinical disease.² In the New York outbreak of 1999, 27% of patients showed signs of muscle weakness, 10% had flaccid paresis, 32% had respiratory symptoms, and 17% required mechanical ventilation.² Previously published case series do not expand on the reason for mechanical ventilation. In our review of the literature this is the first report to identify bilateral diaphragmatic paralysis in a patient with WNV infection and ventilator dependent respiratory failure.

When presented with neuromuscular disease in a critically ill patient a number of contributing factors must be considered including electrolyte disturbances, poor nutrition, or medications such as neuromuscular blockers, corticosteroids, and aminoglycosides. We were able to maintain adequate nutrition and electrolyte balance in our patient, he did not receive neuromuscular blocking agents, but he did receive corticosteroids and aminoglycoside antibiotics. Although we cannot rule out a contribution from these agents, the phrenic nerve conduction study was performed more than 30 days after the last dose of corticosteroid or aminoglycoside. The persistence of inspiratory muscle weakness after discontinuation of these agents makes their contribution unlikely.

In general, treatment of patients with WNV infection is mainly supportive care, although immunoglobulin containing anti-WNV antibodies have been used successfully in immunocompromised patients with WNV encephalitis.³ Theophylline can potentially increase diaphragmatic strength, and recent animal studies have shown improvement in diaphragmatic strength with theophylline after spinal cord hemisection.⁴ Bronchial mucus hypersecretion has been described in patients with spinal cord injury due to inhibition of peripheral sympathetic stimulation,⁵ and the copious thick sputum production in our patient may have been produced via a similar mechanism.

Previous reports have neglected to discuss the respiratory complications associated with WNV infection. Although mucus hypersecretion and Pseudomonas infection may occur simply with prolonged mechanical ventilation, they contributed to morbidity in this patient and should be considered as potential complications of WNV infection. If the prevalence of WNV continues to increase in the future, recognition of diaphragmatic paralysis and related respiratory complications will become increasingly important to the pulmonary/critical care physician.