Dr. Anderson, Dr. Pairolero, members and guests:It is estimated that as many as 150,000 cases of adult respiratory distress syndrome occur in the United States each year. A small subset of that group of patients is represented by those who develop it following pulmonary resection. There are very specific findings about this group of patients: the incidence has been reported as anywhere between 2 and 5 percent following pneumenectomy and 1 percent or less following lobectomy. It typically manifests itself between 24 and 72 hours following surgery, with the first few days being rather uneventful. Typically, the first finding is a very subtle interstitial infiltrate with an increase in respiratory rate followed by hypoxia and then, despite all measures, a very rapid progression to respiratory distress and death in many cases--as you see here, with reports ranging between 50 and 90 percent mortality. This is a group of patients that do not have evidence of pulmonary sepsis; no evidence of congestive heart failure; no evidence of bronchopleural fistula; aspiration; or any of the other causes that are so typically associated with the 50,000 ARDS cases that occur throughout the country. This is a phenomenon that is related solely to the performance of a pulmonary resection--or so it seems.
All of those in the audience who do pulmonary surgery will recognize this X-ray (that may or may not project to the audience), occurring 72 hours after surgery as a very subtle interstitial infiltrate. Aggressive supportive measures are instituted but, despite that, 2 or 3 days later, you can see a virtual whiteout of the remaining lung following right pneumenectomy and, in most cases, death soon follows.
There have been many proposed causes of this phenomenon: fluid overload; barotrauma; interruption of lymphatics at the time of surgery; the release of cytokines; or the activation of the complement system. It is fair to say that we aren't certain exactly what causes this phenomenon in patients following pulmonary resection. Physiologically, ARDS is characterized by intrapulmonary shunting, mild to moderate degrees of pulmonary artery hypertension, pulmonary edema, decreased pulmonary compliance, and varying degrees of right ventricular dysfunction.
In 1987, nitric oxide was identified as the factor responsible for the endothelium-derived relaxing factor. It is clearly produced in the pulmonary vascular endothelium. As we learn more about nitric oxide, we have found many properties that are attributed to it, many of which are beneficial in patients with adult respiratory distress syndrome. It is known to be a selective pulmonary vasodilator; it does not have the systemic vasodilating effects as many of the other vasodilating agents have because of its rapid binding with hemoglobin. It has been shown to relax bronchial smooth muscle and act as a free-radical scavenger and has been shown to inhibit platelet aggregation and neutrophile adhesion, which may be important components of the adult respiratory distress syndrome. Because it is inhaled, it has been proposed that it goes to the well-ventilated portions of the lung and by selectively vasodilating these segments improves ventilation perfusion matching and thereby improving oxygenation in those patients suffering from ARDS. It has been utilized in other series of patients where ARDS is attributed to sepsis, aspiration, burns, trauma, and many of the other things that we mentioned earlier. But to our knowledge, it has not been reported to have been used in a group of patients who have developed a very specific phenomenon of post-pulmonary resection ARDS. We were very keenly aware of this problem in the mid-1980s. We were able to identify 7 for comparison we had seen from 1986 to 1993, this represented about a 5 percent incidence following pneumenectomy.
The group of patients that we will report today are 10 patients that we have identified from 1993 to 1997. You can see the incidence of patients who had pneumenectomy during this interval of time and a 7.5% incidence for those patients who had pneumenectomy during this interval of time and 0.9 percent incidence for those patients undergoing lobectomies during the 1993 to 1997 interval. Post pulmonary resection ARDS is defined as the presence of a previous lung resection. We were certain there were no signs of sepsis, pulmonary emboli, heart failure, aspiration, fluid overload or a bronchopleural fistula. So it is just this specific group of patients who develop ARDS following pulmonary resection.
By utilizing this lung-injury score, we can identify the severity of lung injury in this group of patients. The average in our 10 patients was 3.1 and it is generally accepted that a lung-injury score over 2.5 is indicative of severe pulmonary injury. The treatment for this group of patients is seen here: aggressive supportive measures including intubation; bronchoscopy to clear secretions and identify or eliminate the possibility of a bronchopleural fistula; diuresis when appropriate to dry the patients out; broadening antibiotic coverage in case there was underlying sepsis. A point which I think is worth emphasizing and not often talked about: in our intensive care units, it is the standard practice to utilize changes in the posture of patients undergoing mechanical ventilation. Not only does this improve the management of their secretions but improves the ventilation and perfusion to both dependent and non-dependent portions of the lung. It is routine for our patients to spend at least 6 hours prone during a 24-hour period when undergoing mechanical ventilation, as was the case with this series. All of these patients were started immediately, at the first sign of this problem, on inhaled nitric oxide--between 5 and 20 parts per million and, because of the belief that this may represent an inflammatory problem (as evidenced by the earlier mentioning of the cytokines), some of the patients were treated with steroids in the hopes of reducing this inflammatory component. Nitric oxide was administered according to the guidelines established by our department of anesthesia and critical care medicine and, most importantly, are the last three points that nitric oxide levels were monitored very carefully, as were the levels of methemoglobin and nitrogen dioxide which have the potential to cause problems. If these levels were elevated, the concentration of nitric oxide was adjusted accordingly.
Patient demographics are seen here: 5 of these patients had pre-operative irradiation and represent a disproportionate share of the patients--obviously there were a smaller percent of patients during this time interval who had had pre-operative irradiation compared to those who didn't and yet, t of the patients in this series developed ARDS following pre-operative irradiation. I think this is an area that requires further investigation to quantify the risk that pre-operative irradiation poses in the development of post-operative ARDS. There are 8 patients who had pneumonectomy and, interestingly enough, 5 of them had left pneumonectomies; most series report this much more commonly following right pneumonectomy. Two had lobectomy, one of which had had a previous lobectomy, about 9 months prior to the subsequent lobectomy. All patients, based on their pre-operative pulmonary function (DLCO's), were thought to be good operative candidates, and could tolerate the planned pulmonary resection. That underscores the fact that it is very difficult to predict which patients are going to develop this problem and other than right pneumonectomy, marginal pulmonary function, and preoperative irradiation; there is very little that has been identified to allow us to predict which patients are at risk for developing this problem.
The median number of days on nitric oxide was 8; the patients were kept on nitric oxide and monitored very carefully for the levels of methemoglobin and nitrogen dioxide and generally kept on it until the requirements for mechanical ventilation were minimal, with their FIO2 less than 40 percent, low levels of PEEP and low inspiratory pressures.
The results: all of these patients had dramatic improvement in their chest X-ray: 8 of the 10 normalized their chest X-ray within 48 hours. All experienced improvements in their level of oxygenation, diminished requirements for the concentration of oxygen, peak inspiratory pressures,thereby having a secondary benefit of avoiding the complications of prolonged high concentrations of oxygen and the baro trauma from excessive levels of peak inspiratory pressure. Most importantly, by carefully monitoring these patients and keeping the NO levels between 5 and 20 parts per million, there were no demonstrable findings of toxicity of this treatment that went on anywhere from 8 to over 30 days.
If we look at this ratio, the PAO2 to FIO2, it gives you a sense of the severity of oxygenation faced by these patients. An example here of a patient that has a PAO2 of 100, on a 100 percent, their ratio is 100,--in this particular group of patients, their initial value ratio was 95 and you can see that immediately after the institution of inhaled nitric oxide, this rose to about 128 and then in the first 24 hours, another marked improvement that was steady improvement over the ensuing days with the maximal improvement as these patients recovered at about 6 to 7 days. If you look at the percent improvement (again, another way of looking at that ratio of improved oxygenation), the greatest percent improvement occurred in the first 24-hour period but, again, on each subsequent day, one could notice this improvment. The use of steroids did not show statistically significant improvement compared to those patients who did not receive steroids. There seemed to be some slight benefit initially but did not persist. The role of steroids in this group of patients is unclear. There were no deaths directly attributable to ARDS; 3 deaths related to late sepsis. For what it's worth, you compare the deaths of this small group of 10 patients to the ones we had seen from 1986 to 1993--a 30-percent mortality versus nearly a 90-percent mortality.
So, in conclusion, we still are uncertain exactly what triggers ARDS following pulmonary resection, what mediates this syndrome we are so familiar with. We aren't certain of which of the many possibilities which nitric oxide might work to improve these patients but it seems clear, based on our experience, that early institution of nitric oxide and aggressive supportive measures does change the course of these patients who develop ARDS. And then, just to reemphasize, we think that when the syndrome is recognized, aggressive measures are warranted, with the institution of inhaled nitric oxide at the levels described here, with careful monitoring for increased levels of methemoglobin and nitrogen dioxide.
I would like to thank the Society for the privilege of presenting this paper.