© 2005 American Thoracic Society doi: 10.1164/rccm.2410003
Therapeutic HypercapniaCareful Science, Better TrialsHospital for Sick Children Toronto, Ontario, Canada Reducing tidal volumes in patients with injured lungs—a strategy associated with elevated carbon dioxide—improves patient survival (1, 2). Laboratory studies have documented benefits of hypercapnia, as well as some mechanisms of action (3–5). Moreover, buffering hypercapnic acidosis attenuates its benefit, and hypocapnia can be harmful (6). All of the above suggests that hypercapnia might soon evolve into a testable clinical therapy (7). In the current issue of the Journal (pp. 147–157), a paper from Lang and coworkers injects a dose of disquiet into this evolving carbon dioxide story (8). The authors have investigated an animal model of sepsis and found that their particular strategy of permissive hypercapnia increased—not decreased—lung inflammation (8). This article follows an earlier report of adverse effects of hypercapnia from the same group (9), multiple studies of its free radical biochemistry (10), and extensive experience at the bedside—particularly with pulmonary or intracranial hypertension. So there are two sides to the hypercapnia story and a spectrum of interpretations. The current study is superficially at odds with several previous studies (5, 11–13), including some from our laboratory. The worst possible approach would be to ignore or condemn the new data because they do not fit with one's prior ideas. This would be anti-science. The appropriate approach is to review all the information carefully to understand it fully. How then, do we learn from the current data? I see three questions and three lessons. First, are the findings of the current study attributable to hypercapnia, or to the means of achieving it? Lang and colleagues allocated animals to a strategy of "permissive hypercapnia" by lowering the respiratory rate, not the tidal volume (8). Of course, achieving "permissive hypercapnia" by reducing respiratory rate is unusual at the bedside. In fact, clinicians usually reduce tidal volume, and to compensate, increase the respiratory rate. This may be important, because low values of tidal volume, positive end-expiratory pressure, and respiratory rate may cause atelectasis. Indeed, it has been suggested that benefits associated with one low tidal volume strategy may actually be due to intrinsic positive end-expiratory pressure resulting from the increased respiratory rate (14). As well, the authors remind us that in contrast to adding inspired carbon dioxide, hypoventilation may result in uneven distribution of carbon dioxide throughout the lung (15). Indeed, "therapeutic hypercapnia," through raising the arterial carbon dioxide by increasing inspired concentration, is not frequently practiced in the clinical setting but has proved effective in the laboratory (5). Finally, the authors point out that the use of 100% inspired oxygen (8), a level avoided by most clinicians, might have exacerbated inflammatory events (16). Therefore, in answer to the first question, all hypercapnia may not be equal, and the differences may in part be due to other aspects of the ventilatory management. Second, what do we learn about the complexity of hypercapnia? The current study expands our horizons regarding the pathogenic mechanisms associated with elevated levels of carbon dioxide which can alter the formation of peroxynitrite, and result in either "relatively protective" or "relatively injurious" intermediates. The authors discuss the recent study by Laffey and coworkers that demonstrated protection against endotoxin-induced lung injury with added carbon dioxide (11). Whereas that study suggested that the nitrotyrosine formation might represent a reservoir that could "mop up" the more toxic nitrate/nitrite intermediates in the lung (11), the current study demonstrated that inhaled nitric oxide decreased tissue injury (8), but without altering the formation of nitrotyrosine. Thus, in answer to the second question, inhaled nitric oxide may lessen lung injury associated with endotoxin and hypercapnia, and illustrates the lesson that nitrotyrosine formation might be more a marker of injury than a pathogen. Third, what are the implications for rapid implementation at the bedside? There is an intense personal and professional desire within all of us to get results for patients—and quickly. This was advocated courageously by the actor Christopher Reeve, who died recently. Quadriplegic following spinal trauma, Reeve exhorted researchers to rapidly implement promising therapies, and cut through obstructionist red tape. But many "novel" therapies do not prove successful, and some cause harm. Among many possible explanations as to why a trial may fail, we don't often hear that the trials were premature, or that clinical trials methodology was applied onto a scientifically unsound concept. A classic case in point was the high-frequency ventilation study which, reporting a worse outcome in infants treated with high-frequency oscillatory ventilation, jeopardized the subsequent use of this important therapy (17). Later studies, combining surfactant use and improved understanding of the physiology, resulted in high-frequency oscillation being incorporated into standard care for neonates. A more recent example involves inhaled nitric oxide for acute lung injury. This is an example of an idea that underwent rapid translation into randomized controlled trials, where survival was the key outcome of interest. The last of these negative trials was published in 2003, but had completed enrolling patients by late 1999 (18). However, since 1999, additional clinical and laboratory studies have provided potentially important insight into where nitric oxide does (and doesn't) work, improved means of administration, and better adjuvant therapy; in short, many of the issues that might impact positively on trial design. The current article certainly complicates the carbon dioxide scene (8). Although challenging, we should be as reassured by this complexity as we would be skeptical of undue simplicity. Studies such as this one will move us closer to eventual safe application, and dissuade us from undertaking premature clinical trials. Careful reflection on the sort of data produced by Lang and colleagues (8), coupled with careful preliminary human studies that do not rely on mortality, will optimize trial design. Why expect the road from bench to bedside to be straight or short? FOOTNOTES Conflict of Interest Statement: B.P.K. does not have a financial relationship with a commercial entity that has an interest in the subject of the manuscript. REFERENCES
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
B activation. Am J Respir Cell Mol Biol 2003;29:124–132.