Treatment of Pulmonary Embolism Is All About Relieving Obstruction? N.O., It Is Not!*

Abstract

Pulmonary embolism (PE) is a common clinical entity seen by clinicians in a myriad of treatment arenas (1). Fortunately, PE infrequently leads to death in low risk, hemodynamically stable patients with normal right ventricular (RV) function (2). However, when PE is the driver of acute RV dysfunction or shock, mortality sharply increases (3). It is this patient that is most concerning to intensivists. A PubMed search of “pulmonary embolism” or “pulmonary embolism AND treatment” will reveal a nondigestible number of publications. Despite this, the management principles in general remain the same: judicious manipulation of preload, maintenance of pressure and perfusion to the right ventricle, augmentation of cardiac output with inotropes and inodilators, avoidance of further RV afterload increase (e.g. hypoxia and acidosis), RV afterload reduction, and treatment of the clot itself. Despite thrombolytic therapy, a small but significant portion of patients will not improve (4). This highlights the fact that the increase in RV afterload seen in acute PE is much more complex than mechanical obstruction from clot, and active pulmonary vasoconstriction plays an integral role in the hemodynamics of acute PE (4, 5). Well known in this interplay is the role of nitric oxide (NO) in maintaining low pulmonary tone for the RV to work against (5, 6). The authors hypothesize that it is perhaps hemolysis, from PE itself as well as treatment of PE, that leads to the release of cell-free hemoglobin, which can scavenge NO quite efficiently (7). It is under this backdrop that, in this issue of Critical Care Medicine, Sertorio et al conducted their physiologic study revolving around the role that cell-free hemoglobin and NO consumption play in the sometimes, anomalous hemodynamic findings of acute PE (8). In the first part of the study, the authors used a large animal model (n = 14 lambs) to assess the role of cell-free hemoglobin in NO consumption. The animals were randomized to three groups: sham, embolism with autologous blood clot, and embolism with microspheres. Each animal had continuous hemodynamic monitoring with arterial and pulmonary artery catheters, and sedation and paralysis maintained with ketamine, midazolam, and pancuronium. Blood samples were obtained at baseline, 30, and 180 minutes to measure plasma levels of cell-free hemoglobin and NO consumption. As hypothesized, lung embolization increased both cell-free hemoglobin concentrations and NO consumption. Further, there was a significant correlation between cell-free hemoglobin concentration and NO consumption. Expected hemodynamic changes paralleled these findings. With similar objectives, the authors went to the bedside and compared cell-free hemoglobin and NO consumption in healthy historical controls (n = 28) with two groups of PE patients treated with thrombolytic therapy (n = 14). The main inclusion criteria included documented PE with systolic blood pressure less than 90 or “stable” hemodynamics with RV dysfunction (or a surrogate thereof). Similar results were obtained in these human data: in patients with embolism, both cell-free hemoglobin concentration and NO consumption were greater than in control, with a significant correlation between the two. This study has some attractive strengths. The authors have taken preclinical data and sound physiologic rationale and taken it two steps further. The similar results across the animal model and small human trial seem to show that the authors are on to something. Their results are also consistent with other disease states associated with the decompartmentalization of hemoglobin as well (9). Consistency across methodologies lends strength to the argument for causality. As does biologic plausibility and strength of association. These results seem to show a strong correlation between cell-free hemoglobin and NO consumption, supporting the authors’ hypothesis in acute PE. The relevance here was also enhanced by including patients most concerning to intensivists: PE patients in shock and/or with RV dysfunction. Most of the weaknesses lie in the human arm of the study. The small numbers here make establishing cause–effect impossible. It should be noted that this was not necessarily the aim of this physiologic study. All patients were Caucasian; therefore, extrapolation of these data to other races, which may have genetic differences in NO kinetics, is impossible. There is also a lack of detail concerning very significant comorbid conditions for a study such as this: heart failure, chronic obstructive pulmonary disease, or any hematologic conditions that may predispose to hemolysis (e.g., sickle cell). Although the animal arm of the study seems to show a temporal relationship to NO kinetics and expected hemodynamic changes, we know nothing about the hemodynamics in the humans or when they were studied in their disease course. Is it possible that the increase in cell-free hemoglobin and NO consumption was associated with absolutely no change in pulmonary pressures *See also p. e118.