Noninvasive Ventilation in Myasthenic Crisis

Abstract & Commentary

By David J. Pierson, MD, Editor, Professor, Pulmonary and Critical Care Medicine, Harborview Medical Center, University of Washington, Seattle, is Editor for Critical Care Alert.

Synopsis: In a retrospective study of acute respiratory failure complicating myasthenic crisis, 14 of 24 patients who were managed initially with noninvasive ventilation successfully avoided intubation, but pneumonia occurred in 80% of the instances in which this approach was unsuccessful.

Source: Seneviratne J, et al. Noninvasive ventilation in myasthenic crisis. Arch Neurol 2008;65:54-58.

Investigators in the Department of Neurology at the Mayo Clinic in Rochester, MN, reviewed the charts of patients admitted with myasthenic crisis (MC) between 1987 and 2006 who received either invasive or noninvasive mechanical ventilation. Only patients with de novo MC were included, and postoperative cases plus those associated with underlying cardiopulmonary disease were excluded. Fifty-two patients qualified, with a total of 60 episodes of MC. In 24 (40%) of these episodes, noninvasive ventilation (NIV) was attempted; NIV was the only form of ventilator support used in 14 of the 24, and in 10 instances endotracheal intubation and invasive mechanical ventilation (ET-MV) was subsequently used. In the other 36 episodes (60%), ET-MV was the initial procedure. The authors examined the clinical, physiologic, and outcome information available in the patients' charts for these 3 groups (i.e., NIV alone, NIV followed by ET-MV, and ET-MV alone).

Statistical analysis revealed no significant differences among the 3 groups with respect to age, precipitating event, initial arterial blood gases, initial vital capacity, or initial maximal inspiratory or expiratory pressures, although the numbers of patients who had these different variables measured prior to initiation of ventilatory support are not provided. Among patients initially treated with NIV, those who subsequently were intubated were more likely to have initial arterial PCO2 values exceeding 45 mm Hg. Patients who were intubated spent longer on the ventilator than those in whom NIV was successful. Pneumonia occurred in 3 of the 14 patients successfully treated with NIV, and in 25 of the 46 patients who required intubation—including 8 of the 10 in whom NIV was tried unsuccessfully.


Admission of a patient in MC is a very uncommon event in most ICUs. However, such patients share a number of clinical features and management challenges in common with a larger population of acutely ill patients cared for by intensivists, nurses, and respiratory therapists. The most important of these features are respiratory muscle weakness and impaired airway protection, which are encountered in patients with Guillain-Barré syndrome, spinal cord injury, amyotrophic lateral sclerosis, multiple sclerosis, stroke, and a variety of other disorders affecting neuromuscular function. The most important life-threatening complications in patients with these problems are acute ventilatory failure and pneumonia. Thus, this study of patients with MC, which found very high rates of both of these complications, may be more broadly applicable to ICU practice than just in the context of myasthenia gravis.

The study by Seneviratne and colleagues comes from an acknowledged center of excellence in managing patients with myasthenia gravis, and has the advantage of access to the records of a larger number of such patients than would be the case at most institutions. This large cohort permitted comparisons of patient characteristics and outcomes in patients receiving NIV with those in patients who were intubated during an episode of MC. However, a retrospective chart review has serious inherent limitations when it comes to detecting differences between patients with the same diagnosis, and I think the fact that the patients appeared similar in most aspects examined by the authors illustrates some of these limitations. In this center of excellence, where patients were managed by clinicians experienced in the management of MC, some patients received trials of NIV and some were intubated as the initial form of ventilator support. Why? There were no statistically significant differences in initial arterial blood gases or maximal respiratory muscle pressures. In fact, the mean arterial pH and PCO2 values in the patients receiving NIV were within normal limits. Because this was not a prospective, randomized trial, and instead describes actual practice based on individual patient presentation and clinician judgment, there must have been other findings that led to the decision to try NIV (in 40% of the patients) or to intubate (in 60%). Initial respiratory rates and other vital signs are not provided, nor are descriptions of the patients' levels of alertness or cooperation.

Noninvasive ventilation did not come into wide use in treating acute respiratory failure until the second half of the 19-year period of this review. Thus, it is likely that most of the NIV patients were managed relatively recently, and possible that other important temporal trends in patient assessment and management could have occurred. These things illustrate some of the inherent limitations of a retrospective chart review for telling us why patients were managed in a particular way. Because of such limitations and other factors, studies like this one are useful in generating hypotheses as to the reasons for differences observed among the patients—hypotheses appropriate for testing in prospective clinical trials—but much less reliable when it comes to drawing conclusions about how patients should be managed.

A very high proportion of the patients in this study developed pneumonia—including 80% of those initially treated with NIV who subsequently were intubated. This finding supports the notion that many of these patients were unable to protect their lower airways, and raises concern about the advisability of ventilator support without intubation. Patients managed with NIV who did not have initial hypercapnia tended to do better in this series, but the numbers are fairly small. In fact, bulbar weakness and other circumstances of impaired airway protection are generally listed among the contraindications to NIV in acute respiratory failure. Thus, although this study's findings may be used to support the authors' conclusion that NIV "is effective in the treatment of acute respiratory failure in patients with myasthenia gravis," they could also be cited in support of the opposite conclusion. Prospective studies of NIV vs ET-MV in patients with acute respiratory failure and neuromuscular weakness could tell us with much greater confidence whether the former is an advisable approach. In the meantime, I think great caution should be exercised whenever the use of NIV is contemplated in such patients—particularly if they have hypercapnia or obvious bulbar weakness on initial presentation.