Sodium Nitroprusside on AF Occurrence After Cardiac Surgery
Sodium Nitroprusside on AF Occurrence After Cardiac Surgery
Although a lower incidence of POAF was found in patients who received sodium nitroprusside compared to controls, the difference between groups was not statistically significant. Also, despite multivariate analysis showing a possible 17% decrease in the odds of POAF, and patient matching based on propensity score showing a possible 23% decrease, the results were still not significant. For the secondary outcomes a significant decrease in ICU length of stay was found with sodium nitroprusside, which could have potential cost benefits. However, a higher incidence of TIA was also found in the sodium nitroprusside group.
One unexpected finding in the results was that patients taking preoperative β-blockers were more than twice as likely to experience POAF. This result was consistent within both the entire study cohort and the sodium nitroprusside group. It is interesting, considering β-blockers are the primary therapy recommended for the prevention of POAF. Although this result cannot be fully rationalized, one possible explanation is that patients on preoperative β- blocker therapy may have had a greater propensity for being prescribed therapy. In other words, patients with a higher risk of POAF were more likely to be receiving β- blockers than were those with a lower risk. Another explanation may be late use or underuse of β-blockers postoperatively. Failure to reinitiate β-blocker therapy postoperatively has been associated with almost twice the risk of POAF. However, data indicating when in the postoperative period β-blockers were reinitiated were not available.
The pharmacologic mechanisms by which sodium nitroprusside is theorized to reduce POAF are thought to be its antiinflammatory effects in the vasculature and its ability to serve as an NO donor. Research has recently been conducted on the role of NO as a potential cause of atrial fibrillation. Lin and colleagues demonstrated that loss of NO in pulmonary veins, an important focus of electrical signaling, increases the rate of uncontrolled conduction. Their study also demonstrated that use of sodium nitroprusside in canine pulmonary vessel cells decreased irregular signaling. These findings establish a link between irregular cardiac myocyte conduction and NO. The research conducted by Lin and colleagues was based on studies that examined the influence of pulmonary veins on the occurrence of atrial fibrillation.
The study conducted by Cavolli and colleagues was predicated on the findings that decreased serum NO plays a role in the development of atrial fibrillation. It is perhaps the only other clinical trial published that has evaluated the effect of sodium nitroprusside on postoperative atrial fibrillation. It enrolled 100 patients undergoing first elective CABG surgery and randomly assigned them to receive either placebo or sodium nitroprusside intraoperatively. Study medications were administered continuously for 60 minutes beginning on release of the aortic cross clamp before rewarming had begun. A significant reduction in POAF was seen between the control group (36%) and sodium nitroprusside group (12%) in the first 5 postoperative days (p = 0.005). On multivariate analysis, sodium nitroprusside remained a significant predictor of POAF (OR 4.28; 95% CI 1.50 to 12.27).
Considering the results seen by Cavolli and colleagues we expected to find a positive association between POAF and sodium nitroprusside in our study population. However, there are differences between the 2 studies that may explain why the results of our trial were not significant. First, based on the results observed by Cavolli and colleagues we anticipated that the rate of POAF in the sodium nitroprusside group would be half that in the control group. Although the incidence was lower in the sodium nitroprusside group in our study, the relative difference was only 9%. Therefore, the trial lacked sufficient power to show a statistically significant difference in the incidence of POAF between groups. Second, the populations in both studies were different. The study by Cavolli and colleagues excluded patients undergoing surgery other than elective CABG. Our study enrolled patients undergoing elective, urgent, or emergent isolated CABG or valve surgery, or a combination of the 2. Thus, our patient population was at a higher risk of POAF. The effect of nitroprusside as an NO donor may not be as beneficial in patients undergoing these various procedures. Finally, the time frame assessed, in which the development of POAF occurred, differed. Cavolli and colleagues assessed only the incidence of POAF for 5 days postoperatively, while our study included the entire postoperative period, regardless of length of stay. This longer time frame may have negated the significant benefit of sodium nitroprusside observed by Cavolli and colleagues in the early postoperative period.
Despite these differences, our study builds on the limitations of those in the study conducted by Cavolli and colleagues. For instance, in Cavolli and colleagues' study, patients with a history of atrial fibrillation were not excluded and, although the preoperative incidences of arrhythmias in both groups were reported, the authors did not define this variable. We eliminated this confounder by excluding patients with a history of atrial fibrillation. Another improved aspect of our study was the inclusion of variables associated with a higher incidence of POAF in the analysis. These included body mass index; mitral valve disease; cerebrovascular disease; peripheral artery disease; preoperative use of statins, steroids, and inotropes; use of an intraaortic balloon pump; readmission to the ICU; and postoperative pneumonia.
Despite these design improvements our study has several limitations. Chief among these is its retrospective design. A prospective study including methodologic improvements similar to those in this trial would help eliminate some of the confounding variables observed in our population. Our trial was also limited by the methods used for data collection, which could not confirm dose, timing, or administration of sodium nitroprusside. This adds bias to the results, as patients included in the sodium nitroprusside group may never have received the medication. In addition, the dose and timing of sodium nitroprusside administration may play a crucial role in the prevention of POAF. Research has found that oxidative stress occurs during ischemia-reperfusion. Also, administration of sodium nitroprusside after release of the aortic cross clamp has been found to decrease systemic inflammatory markers. The protocol used by Cavolli and colleagues called for the administration of sodium nitroprusside in the setting of ischemia-reperfusion injury after aortic cross clamp removal. It is unlikely that all patients in our study who received sodium nitroprusside had it administered during this crucial time. Another limitation of our trial was that a standard definition of POAF was not used. However, no standardized definition for POAF exists and the version of the STS Adult Cardiac Database that was used in this trial only defines POAF as atrial fibrillation or flutter that was not present preoperatively. Finally, patients receiving preoperative digoxin or other antiarrhythmic medications were not excluded. Digoxin has been associated with an increase in POAF and antiarrhythmic medications such as amiodarone decrease the risk of POAF.
Use of sodium nitroprusside during cardiothoracic surgery was not associated with a significant difference in the incidence of POAF. These results do not rule out a possible association; however, the limitations of this trial prevent definitive conclusions. There is both scientific and clinical evidence supporting a role for NO donors such as sodium nitroprusside in the prevention of POAF. Additional research in this area should seek to further elucidate the role that NO plays in the development of POAF and build on the limitations of previous clinical investigations using NO donors.
Discussion
Although a lower incidence of POAF was found in patients who received sodium nitroprusside compared to controls, the difference between groups was not statistically significant. Also, despite multivariate analysis showing a possible 17% decrease in the odds of POAF, and patient matching based on propensity score showing a possible 23% decrease, the results were still not significant. For the secondary outcomes a significant decrease in ICU length of stay was found with sodium nitroprusside, which could have potential cost benefits. However, a higher incidence of TIA was also found in the sodium nitroprusside group.
One unexpected finding in the results was that patients taking preoperative β-blockers were more than twice as likely to experience POAF. This result was consistent within both the entire study cohort and the sodium nitroprusside group. It is interesting, considering β-blockers are the primary therapy recommended for the prevention of POAF. Although this result cannot be fully rationalized, one possible explanation is that patients on preoperative β- blocker therapy may have had a greater propensity for being prescribed therapy. In other words, patients with a higher risk of POAF were more likely to be receiving β- blockers than were those with a lower risk. Another explanation may be late use or underuse of β-blockers postoperatively. Failure to reinitiate β-blocker therapy postoperatively has been associated with almost twice the risk of POAF. However, data indicating when in the postoperative period β-blockers were reinitiated were not available.
The pharmacologic mechanisms by which sodium nitroprusside is theorized to reduce POAF are thought to be its antiinflammatory effects in the vasculature and its ability to serve as an NO donor. Research has recently been conducted on the role of NO as a potential cause of atrial fibrillation. Lin and colleagues demonstrated that loss of NO in pulmonary veins, an important focus of electrical signaling, increases the rate of uncontrolled conduction. Their study also demonstrated that use of sodium nitroprusside in canine pulmonary vessel cells decreased irregular signaling. These findings establish a link between irregular cardiac myocyte conduction and NO. The research conducted by Lin and colleagues was based on studies that examined the influence of pulmonary veins on the occurrence of atrial fibrillation.
The study conducted by Cavolli and colleagues was predicated on the findings that decreased serum NO plays a role in the development of atrial fibrillation. It is perhaps the only other clinical trial published that has evaluated the effect of sodium nitroprusside on postoperative atrial fibrillation. It enrolled 100 patients undergoing first elective CABG surgery and randomly assigned them to receive either placebo or sodium nitroprusside intraoperatively. Study medications were administered continuously for 60 minutes beginning on release of the aortic cross clamp before rewarming had begun. A significant reduction in POAF was seen between the control group (36%) and sodium nitroprusside group (12%) in the first 5 postoperative days (p = 0.005). On multivariate analysis, sodium nitroprusside remained a significant predictor of POAF (OR 4.28; 95% CI 1.50 to 12.27).
Considering the results seen by Cavolli and colleagues we expected to find a positive association between POAF and sodium nitroprusside in our study population. However, there are differences between the 2 studies that may explain why the results of our trial were not significant. First, based on the results observed by Cavolli and colleagues we anticipated that the rate of POAF in the sodium nitroprusside group would be half that in the control group. Although the incidence was lower in the sodium nitroprusside group in our study, the relative difference was only 9%. Therefore, the trial lacked sufficient power to show a statistically significant difference in the incidence of POAF between groups. Second, the populations in both studies were different. The study by Cavolli and colleagues excluded patients undergoing surgery other than elective CABG. Our study enrolled patients undergoing elective, urgent, or emergent isolated CABG or valve surgery, or a combination of the 2. Thus, our patient population was at a higher risk of POAF. The effect of nitroprusside as an NO donor may not be as beneficial in patients undergoing these various procedures. Finally, the time frame assessed, in which the development of POAF occurred, differed. Cavolli and colleagues assessed only the incidence of POAF for 5 days postoperatively, while our study included the entire postoperative period, regardless of length of stay. This longer time frame may have negated the significant benefit of sodium nitroprusside observed by Cavolli and colleagues in the early postoperative period.
Despite these differences, our study builds on the limitations of those in the study conducted by Cavolli and colleagues. For instance, in Cavolli and colleagues' study, patients with a history of atrial fibrillation were not excluded and, although the preoperative incidences of arrhythmias in both groups were reported, the authors did not define this variable. We eliminated this confounder by excluding patients with a history of atrial fibrillation. Another improved aspect of our study was the inclusion of variables associated with a higher incidence of POAF in the analysis. These included body mass index; mitral valve disease; cerebrovascular disease; peripheral artery disease; preoperative use of statins, steroids, and inotropes; use of an intraaortic balloon pump; readmission to the ICU; and postoperative pneumonia.
Despite these design improvements our study has several limitations. Chief among these is its retrospective design. A prospective study including methodologic improvements similar to those in this trial would help eliminate some of the confounding variables observed in our population. Our trial was also limited by the methods used for data collection, which could not confirm dose, timing, or administration of sodium nitroprusside. This adds bias to the results, as patients included in the sodium nitroprusside group may never have received the medication. In addition, the dose and timing of sodium nitroprusside administration may play a crucial role in the prevention of POAF. Research has found that oxidative stress occurs during ischemia-reperfusion. Also, administration of sodium nitroprusside after release of the aortic cross clamp has been found to decrease systemic inflammatory markers. The protocol used by Cavolli and colleagues called for the administration of sodium nitroprusside in the setting of ischemia-reperfusion injury after aortic cross clamp removal. It is unlikely that all patients in our study who received sodium nitroprusside had it administered during this crucial time. Another limitation of our trial was that a standard definition of POAF was not used. However, no standardized definition for POAF exists and the version of the STS Adult Cardiac Database that was used in this trial only defines POAF as atrial fibrillation or flutter that was not present preoperatively. Finally, patients receiving preoperative digoxin or other antiarrhythmic medications were not excluded. Digoxin has been associated with an increase in POAF and antiarrhythmic medications such as amiodarone decrease the risk of POAF.
Use of sodium nitroprusside during cardiothoracic surgery was not associated with a significant difference in the incidence of POAF. These results do not rule out a possible association; however, the limitations of this trial prevent definitive conclusions. There is both scientific and clinical evidence supporting a role for NO donors such as sodium nitroprusside in the prevention of POAF. Additional research in this area should seek to further elucidate the role that NO plays in the development of POAF and build on the limitations of previous clinical investigations using NO donors.
