"Therapeutic" Carbon Monoxide May Be Toxic

To the Editor.

The report by Dolinay and colleagues (1) once again highlights the remarkable diversity of cellular effects mediated by carbon monoxide (CO). These workers and others have been working for some time toward establishing antiinflammatory effects of CO by mechanisms independent of CO-mediated hypoxia. Many of the "protective effects" of CO are intriguing, although sometimes not clearly distinguished from the cellular effects of lowering tissue oxygen tension, and are sometimes not reproducible in other laboratories (2). In the case of experimental ventilator-induced lung injury (VILI), these authors demonstrated differences in tumor necrosis factor-a elaboration and bronchoalveolar lavage cellularity but no convincing quantitative improvement in the extent of lung injury (for example, by observer-blinded lung injury scores). We were far more alarmed, however, when the authors suggested being "tempted" to move toward clinical trials using CO as a therapeutic agent to antagonize inflammatory processes, such as VILI.

Hypoxic effects of CO were described by Claude Bernard and John Haldane, and it is now clear that there are several additional pathways of cell stimulation mediated by CO. Activation of stress-dependent protein kinases, as shown by the authors and others, may have beneficial effects (1, 3). There are also effects related to perturbation of nitric oxide-dependent pathways that are injurious. Animals exposed for 1 hour to just 50 ppm CO exhibit protein tyrosine nitration in lung and large vessels, a macromolecular capillary leak, and leukocyte sequestration phenomena (4). Human beings exposed for hours to low CO concentrations also exhibit vascular leakage of macromolecules (5). Furthermore, patients with significant coronary or cerebrovascular disease tolerate even low carboxyhemoglobin (COHb) levels poorly.

Giving CO "therapeutically" as suggested by Dolinay and colleagues (1) could also result in blood COHb levels as high as 20% (6). Such CO exposures would be expected to cause brain injury similar to that caused by CO poisoning (7). Furthermore, some of the diseases that may require ventilatory support have been associated with brain injury themselves. One example is acute respiratory distress syndrome. Many patients with ARDS manifest brain injury 1 year after hospital discharge (8); an additional insult from iatrogenically administered CO could conceivably worsen brain-related outcomes.

Our main point is that there is a very real potential for unforeseen injury related to seemingly modest concentrations of CO. Additional information about CO pathophysiology is needed, and it is premature to suggest clinical trials purposefully administering this agent to injured patients.

Conflict of Interest Statement: S.R.T. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; L.K.W. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; N.B.H. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript.