Noncardiogenic pulmonary edema after cardiopulmonary bypass

Share Embed


Descrição do Produto

Noncardiogenic Pulmonary Edema After Cardiopulmonary Bypass An Anaphylactic Reaction to Fresh Frozen Plasma

Sabet W. Hashim, MD, New Haven, Connecticut Harold Ft. Kay, MD, New Haven, Connecticut Graeme L. Hammond, MD, New Haven, Connecticut Gary S. Kopf, MD, New Haven, Connecticut Alexander S. Geha, MD, New Haven, Connecticut

Noncardiogenic pulmonary edema after cardiopulmonary bypass is a rare clinical entity but one noted with increasing frequency. The syndrome is differentiated from pulmonary edema of cardiac origin by the occurrence of massive pulmonary edema in the presence of normal left ventricular filling pressure. Its cause is unknown, although anaphylactic reactions to protamine [I] or blood transfusion products [2] have been suggested. The purpose of this study is to report our experience with this syndrome at the Yale-New Haven Hospital over a 5 year period, to identify the etiologic factors and to discuss the therapeutic approach that has evolved from this experience. Clinical

Material

Between September 1977 and December 1982, we observed nine episodes of noncardiogenic pulmonary edema. They occurred in eight patients, one of whom had the syndrome twice. The patients’ ages ranged from 3 to 70 years; six were male and two were female. Noncardiogenic pulmonary edema occurred after corrective surgery for coronary, valvular, and congenital lesions. Table I summarizes the clinical features in the eight patients listed in chronologic order of occurrence of noncardiogenic pulmonary edema. The syndrome was characterized by massive and projectile outpouring of plasma-like edema fluid from the endotracheal tube that occurred from 30 minutes to 4 hours after the termination of cardiopulmonary bypass when fresh frozen plasma was being administered. Pulmonary artery diastolic pressure was elevated in the presence of low or normal left atria1 or pulmonary capillary wedge presFrom the Department of Surgery, Yale University School of Medicine, New Haven, Connecticut Requests for reprints should be addressed to Sabet W. Hashim, MD, Yale University School of Medicine. 333 Cedar Street, New Haven, Connecticut 06510 Presented at the 64th Annual Meeting of the New England Surgical Society, Bretton Woods, New Hampshire, September 30-October 2, 1983

580

sures. Arterial blood gases rapidly deteriorated, and the arterial partial pressure of oxygen decreased to less than 60 mm Hg in all cases despite a fractional inspired oxygen value of 1. Visual observation of the heart in two patients during development of the syndrome showed that the ventricles were not dilated and were contracting normally. In all but one patient, a progressive decrease in cardiac output and severe hypotension occurred. The syndrome resulted in the death of three patients despite treatment with vasopressors and an intraaortic balloon pump (Table

I). The case histories of Patients 2 and 5 represented in Table I (operated on in 1977 and 1980, respectively) and Patients 7 and 8 (operated on in 1982) are presented in detail to elucidate evolutionary patterns of the syndrome and its response to various treatment modalities, Patient 2. A 44 year old white woman underwent emergency coronary artery bypass grafting in September 1977 for left main coronary artery disease and unstable angina. The intraaortic balloon was placed preoperatively. Cardiopulmonary bypass was terminated without difficulty. Systemic systolic pressure was 120 mm Hg and the pulmonary artery diastolic pressure was 12 mm Hg. One hour later, after the sternotomy had been closed and while the patient was receiving fresh frozen plasma, copious amounts of plasma-like fluid exuded from the endotracheal tube. Pulmonary compliance decreased markedly, systemic pressure decreased, and the pulmonary artery diastolic pressure increased to 26 mm Hg. Pulmonary capillary wedge pressure was 7 mm Hg. The chest was reopened and the heart was found to be small and contractile, the grafts were patent, the lungs were distended and did not collapse during expiration. The patient became progressively hypotensive despite intraaortic balloon assist and administration of vasopressors. One gram of methylprednisolone, 50 mg of diphenhydramine and 100 mg of furosemide were given intravenously. The condition rapidly deteriorated, and the patient died in the operating room. Exploration of the pulmonary artery revealed no pulmonary embolus. Postmortem examination showed massive pulmonary edema. Each lung weighed 800 g.

The American Journal of Surgery

Noncardiogenic Pulmonary Edema

TABLE I

Clhtksl Feature8 of NoncarcHogenk Pulnwnary Edema

Patient

Age (yr)

Operation’

Date of Operation

1 2 3

3 44 70

TOF CABG(3) CABG(2)

7177 9177 7179

w 24

4

49

CABG(3)

8180

26

5 6

59 21

CABG (3) AVR

10180 7181

24 19

... ..

8 12

7t

68

CABG (2)

10182

32

...

12

8

57

CABG(3)

10182

29

PAD (mm Hg)

;sj mm Hg) 8

...

...

7

5

... 10

1

.

.

12

Cardiac index

IABP

outcome

verylow

No

Very low 1 liier/m2/ min 1.8 liter/ m*/min very low 2.4 liter/ m*/min 1 liter/m*/ min 1.4 liter/

Yes No

Death Death Well

Yes

Well

Yes No

Death Well

No

Well

No

Well

Values in parenthesesindicatethe numberof grafts. t Noncardiogenicpulmonaryedema developedtwice in this patient.Immediatelyafter operationthe patienthad adultrespiratorydistress syndromewhich abated. ARDS = adultrespiratorydistresssyndrome;AVR = aotticvalvereplacement;CABG= coronaryattery bypassgaftlng; IABp = intraaor~ic balloonpump;LAP = low or normalatrial pressure;PAD = pulmonaryarterydiastolicpressure;TOF = tetralogyof Fallot;W = pulmonary capillary wedge pressure. l

Patient 5. A 59 year old white man underwent elective coronary artery bypass grafting in October 1980. He was hemodynamically stable for the first hour in the intensive care unit during which time he received nitroprusside for postoperative hypertension. Because of a large diuresis, he was given fresh frozen plasma for volume replacement. Suddenly, fulminant projectile episode of pulmonary edema occurred. The pulmonary artery diastolic pressure increased from 9 to 24 mm Hg. Mixed venous hemoglobin oxygen saturation decreased from 68 to 48 percent and arterial partial pressure of oxygen decreased from 265 to 45 mm Hg. The patient was immediately given 2 g of methylprednisolone, 50 mg of diphenhydramine, and 100 mg of furosemide. Endotracheal secretions were incessant, and hypoxia persisted despite administration of 15 cm of positive end-expiratory pressure (PEEP). The patient became progressively hypotensive, and catecholamines were administered with no resultant improvement. He was returned to the operating room and reexplored. The heart was small and contracted well, the grafts were patent, and the lungs were not compliant. The PEEP was increased to 20 cm of water which increased the arterial partial pressure of oxygen to 110 mm Hg. However, the patient continued to deteriorate hemodynamically. The intraaortic balloon was inserted but did not ameliorate peripheral perfusion. Respiratory failure decreased over the next few hours and secretions decreased, but the cardiac output remained very low. The patient was returned to the intensive care unit where he slowly but progressively showed hemodynamic and respiratory improvement. Hemodynamic and respiratory recovery occurred 40 hours after the operation at which time the arterial partial pressure of oxygen was 120 mm Hg with a fractional inspired oxygen concentration value of 0.4. Unfortunately, the patient became anuric and had a prolonged course of acute renal failure and sepsis. He died 3 weeks after surgery. Patient 7. A 68 year old white man underwent elective coronary artery bypass grafting in October 1982. Thirty

minutes after decannulation, pink frothy secretions came from the endotracheal tube. One unit of fresh frozen plasma was being administered. The heart, which was still exposed, was small and contracted well. The pulmonary artery diastolic pressure increased from 12 mm Hg to 32 mm Hg while the pulmonary capillary wedge pressure remained at 12 mm Hg. Arterial partial pressure of oxygen decreased to 60 mm Hg, and PEEP was instituted. Corticosteroids were administered. The patient became hypotensive and dopamine was started. Pulmonary capillary wedge pressure decreased to 8 mm Hg. Two liters of normal saline solution were infused over 30 minutes with immediate improvement in blood pressure and cardiac output. Secretions from the endotracheal tube diminished, and over the next 2 hours 4 more liters of crystalloid solutions were given. Arterial blood gases improved, and 2 hours later the arterial partial pressure of oxygen was 240 mm Hg with a fractional inspired oxygen concentration value of 0.6. The patient continued to be stable in the intensive care unit, but continuous blood drainage from the chest tubes was noted, and abnormal clotting parameters were found. Accordingly, another unit of fresh frozen plasma was ordered. After 50 ml of fresh frozen plasma was administered, a second episode of noncardiogenic pulmonary edema occurred. The patient was again treated with infusion of normal saline solution to restore volume and improve hemodynamics. Administration of fresh frozen plasma was discontinued. The secretins progressively decreased, and a satisfactory cardiac output was maintained. Arterial blood gases again improved, but the patient required ventilatory assistance for 5 days. He survived and is doing well 11 months after surgery. Patient 8. A 5’7year old white man underwent elective coronary bypass in October 1982. Two hours postoperatively, he became stable in the intensive care unit when he was given fresh frozen plasma for volume replacement. Massive frothy secretions suddenly exuded from the endotracheal tube. The pulmonary artery diastolic pressure

561

Hashim et al

increased from 14 mm Hg to 29 mm Hg while the wedge pressure decreased from 12 mm Hg to 8 mm Hg. Systemic hypotension ensued and dopamine was started. Over the next hour, 4 liters of normal saline solution were administered. The hemodynamics quickly stabilized, and hypoxia, which was initially severe, progressively decreased with the use of PEEP. Pulmonary secretions decreased despite rapid saline infusion. The patient continued to be stable hemodynamically, and his oxygenation improved over an 8 hour period. He was extubated 48 hours later with satisfactory blood gases. He was discharged 10 days postoperatively and continues to do well 11 months after surgery.

Comments Pulmonary edema formation follows Starling’s law of capillary exchange which describes net flow of water across the capillary wall in terms of gradients

of hydrostatic and oncotic pressures: Net rate of filtration = K [hydrostatic pressure difference - osmotic pressure difference]. K represents the permeability coefficient. The hydrostatic and osmotic pressures are measured in the plasma and interstitial fluid. In hemodynamic pulmonary edema, the gradient between intravascular hydrostatic pressure and oncotic pressure is increased which leads to a high transudation of fluid across the endothelium to the instertitium and alveoli. This situation is most frequently seen in patients with cardiac valvular or left ventricular disease. In permeability pulmonary edema, the coefficient of permeability is increased secondary to an alteration of the alveolar capillary membrane. Accordingly, the edema fluid is rich in plasma protein. Permeability edema is seen in toxic and anaphylactic reactions to drugs or blood products. The syndrome has been described after the administration of fresh frozen plasma in noncardiac surgical patients [3]. A consistent finding in all of our eight patients in whom noncardiogenic pulmonary edema developed was that it occurred while they were receiving fresh frozen plasma. Protein content of the edema fluid measured approximately 3.5 g/ liter. In all eight patients, blood incompatibility was ruled out. In three patients, leukocyte agglutination tests were performed; the results were unequivocal. These observations imply, but by no means prove, that an anaphylactic reaction produces permeability edema. The temporal association between noncardiogenic pulmonary edema and fresh frozen plasma administration was most impressive in Patient 7. After an initial complete physiologic recovery, the initiation of a second infusion of fresh frozen plasma caused the reoccurrence of noncardiogenic pulmonary edema. Because noncardiogenic pulmonary edema occurred in patients with heart disease and was associated with an elevated pulmonary artery diastolic pressure, we initially resorted to the use of cardiotonics, diuretics, and the intraaortic balloon pump for therapy. Of six patients so treated, three

562

died from hemodynamic difficulties. Patients 1 and 2 died in the operating room from intractable hypotension despite catecholamine administration in both patients and the use of an intraaortic balloon pump in Patient 2. Patient 5 died from acute renal failure and sepsis secondary to prolonged hypotension. The cause of the hemodynamic deterioration can best be explained by hypovolemia. The pulmonary capillary wedge pressures or left atria1 pressures were within normal limits at the onset of the syndrome and decreased as it developed. The hypovolemia resulted from the loss of fluid into the lungs and possibly across other capillary beds. In the last two patients in our series, consistent improvement in cardiac output, without a worsening of the respiratory distress was observed when adequate left-sided filling pressures were restored. The reluctance to give volume replacement in the presence of pulmonary edema in our earlier patients undoubtedly contributed to the high mortality. It is, therefore, essential in this syndrome to give fluids rapidly at the onset of hypotension until a satisfactory cardiac output is obtained or until the pulmonary capillary wedge pressure reaches 15 mm Hg. This did not exacerbate the pulmonary edema. Olinger et al [1] described four patients in whom noncardiogenic pulmonary edema developed. He suspected that the syndrome was an allergic reaction to protamine administration. Two patients died from profound hypoxia and ventricular fibrillation. Both had no tolerance for the administration of exogenous catecholamines which caused ventricular fibrillation at very low rates of infusion. The hypovolemic nature of the hemodynamic difficulties was recognized, and the patients were treated with massive amounts of fresh frozen plasma, 5 percent albumin solution and blood. It is possible that the large amount of colloid administered leaked through the altered alveolar capillary membrane exaggerating pulmonary edema and prolonging its course. Also, the administration of fresh frozen plasma may have triggered recurrent episodes of noncardiogenic pulmonary edema, as observed in one of our patients. It is important to recognize that in all of the survivors, the edema was temporary, and oxygenation compatible with survival could always be maintained with PEEP. As one would expect, the hematocrit value increased as the edema developed. Therefore, we were reluctant to give blood as primary fluid replacement and have found that normal saline solution is a very satisfactory alternative. Before October 1982, we were giving fresh frozen plasma to many patients after cardiopulmonary bypass to prevent clotting abnormalities and for volume replacement. Since recognizing the link between fresh frozen plasma and noncardiogenic pulmonary edema, we have restricted the use of fresh frozen plasma to the correction of demonstrable coagulation abnormalities in patients with postop-

The American Journal of Surgery

Noncardiogenic Pulmonary Edema

erative bleeding. We are able to report that there have been no additional cases of noncardiogenic pulmonary edema in the 12 month period between our restriction of the use of fresh frozen plasma and the completion of this study. We have observed a temporal association between noncardiogenic pulmonary edema and fresh frozen plasma infusion after cardiopulmonary bypass. The resultant hypoxia is transient, and mortality is a consequence of hemodynamic deterioration. If hemodynamics are maintained at acceptable levels with rapid infusion of normal saline solution at the onset of hypotension a better outcome can be expected. Summary Nine episodes of fulminant noncardiogenic pulmonary edema after cardiopulmonary bypass were observed in eight patients between September 1977 and December 1982. All these catastrophic reactions occurred during administration of fresh frozen plasma 30 minutes to 6 hours after discontinuation of cardiopulmonary bypass. In one patient, two episodes of noncardiogenic pulmonary edema occurred 4 hours apart. In each instance, fresh frozen plasma was being administered. In all patients, pulmonary artery diastolic pressure became elevated during the administration of fresh frozen plasma while left atrial pressure or pulmonary capillary wedge pressure progressively decreased, and cardiac output deteriorated markedly in all but one patient. Corticosteroids, postive end-expiratory pressure, and catecholamines were administered to all patients. All deaths were due to a decrease in cardiac output. Cardiac output did not increase substantially with the use of an intraaortic balloon pump or the administration of catecholamines. The last two patients in the series showed a steady and remarkable improvement in cardiac output when the wedge pressure was increased to a level above 15mm Hg with the administration of normal saline solution. Our data suggest the following: (1) noncardiogenic pulmonary edema after cardiopulmonary bypass is most probably an anaphylactic reaction to fresh frozen plasma. (2) The syndrome is reversible within hours; in only one patient (who suffered noncardiogenic pulmonary edema twice) did adult respiratory distress syndrome develop. (3) The three deaths were not related to hypoxia but to the deleterious effects of low cardiac output associated with hypovolemia secondary to fluid loss through the lungs and possibly across other capillary beds. Therefore, treatment should include restoration of adequate left-sided filling pressures to achieve satisfactory cardiac output. References 1. Olinger, GN, Wer

RM, Bonchek LI. Noncardiogenic pulmonary

edema and peripheral vascular collapse following cardio-

Volume 147, April 1984

pulmonary bypass. rare protaminereaction?AnnThoracSwg 1980; 29:20-5. 2. CullifordAT, ThomasS, SpencerF. Fulminatingnoncardiogenic pulmonary edema. Thorac CardiovascSurg 1980; 80:8683

75. O’Connor PC, Erskine JG, Pringle TH. Pulmonary edema after transfusionwith fresh frozen plasma. Br Med J 1981;282: 379-80.

Discussion Lawrence Cohen (Boston, MA): This truly is a very important paper to those of us who do cardiac surgery because it points out once again that even a therapy that seems to have tremendous theoretic advantage and is seemingly very benign has, like any other therapy, its serious and even fatal complications. The use of fresh frozen plasma has become extremely fashionable in open heart surgery because of the hemodilution that is used in patients, since plasma clotting proteins are diluted, and there is of course platelet damage with the variety of different oxygenators in use. So, in theory it seems like a very good idea to use fresh frozen plasma. I agree 100 percent with the authors that this, in fact, is the reason that this allergic phenomenon developed in these patients. We did not, and do not, use this therapy routinely. When one thinks that he had done a fairly good job on a very routine, very lowrisk operation, to find a profound hypotension and then ultimately hypoxia is perhaps one of the most disheartening feelings imaginable. We have had two patients, one of whom had a bypass, in whom this severe volume problem resulted, and we had another patient in whom it occurred postoperatively and in whom an anaphylactic reaction with severe bronchospasm and terrible hypoxia developed despite the fact that the grafts were open and the heart was functioning very well. It appears now that the principal reason to give fresh frozen plasma is if the operation is expected to be very long, if the patient comes to the operating room very sick; if the patient is going to have reoperation; and, of course, if the patient has preexistent liver disease and might need open heart surgery. One other important reason to consider not using fresh frozen plasma routinely is the cost-effectiveness issue. I think as surgeons, particularly in New England and especially in Massachusetts, we have to be very aware that any therapy we do had better be justified on very solid grounds. The cost of one unit of fresh frozen plasma is $75. If you take a very busy cardiac surgical unit, such as Yale’s that probably does at least 750 to 800 open heart operations per year, multiply that by 3 units per patient (roughly) and then $75 per unit, it’s small wonder that we’re constantly fighting the accelerating cost of care, particularly after open heart surgical therapy. I would ask you only one question: Of the patients who did die, what did the pulmonary and cardiac tissue look like at the time of autopsy? Again, I want to compliment the authors for their courage and the timeliness of pointing out the effect of a relatively commonly used product, which I think should be thought about more critically before it’s used in every patient undergoing open heart surgery.

Graeme Hammond (New Haven, CT): Several years ago in the state of Connecticut, we went to component therapy for volume replacement; every unit of blood that is harvested in the state or that comes into it is sent to Farmington. At Farmington, the blood is broken down into

563

red blood cells, the platelet buffy coat, and the plasma. It is then shipped out to the various centers as they need it. We can still get whole blood, but it’s difficult to do, and they’re trying to phase that out. As a result, giving open heart surgery patients packed cells for volume replacement which obviously did not have the clotting proteins in them and so on, was the rationale for our using fresh frozen plasma routinely on cardiac surgical patients. This went on over a 5 year period. Now this in fact does not make any sense, because the blood would go to Farmington where it was broken down into components; we would get the components back to New Haven, and we’d mix them all back together again and give this to the patient. Of course, our blood bank pointed out innumerable times that what we were doing didn’t make any sense, yet our response was that we wanted the whole blood. Dr. Joe Bove who is the director of our blood bank very carefully analyzed our blood usage before we went to component therapy, and after we had gone into component therapy and were giving the fresh frozen plasma. It became clear that there was really very little difference in the use of our blood whether we were giving components or whether we were giving whole blood. So really, there is very little necessity to give fresh frozen plasma. Then these problems started developing which we could not put our finger on, and I must say I was very much fooled. As Dr. Cohen said, you perform an operation that goes very nicely without any problems and then all of a sudden this devastating pulmonary edema develops. As a cardiac surgeon dealing with cardiac surgical patients, naturally the first thing you think is that there has been some horrendous cardiac poblem, like a big infarct, a clotted graft, a valve that’s not working, and so on. So we thought this was a hypertensive type of edema with accompanying elevated pressures rather than a permeability edema. To Dr. Hashim’s tremendous credit, he was the one who observed in the operating room that, in fact, when it developed in the operating room the hearts were very small and contracting normally with low filling pressures. Our understanding of the problem and corrective measures evolved from there. The lesson is that fresh frozen plasma absolutely should not be used routinely on cardiac surgical patients or any other patients, for that matter. It should be reserved only for treating serious coagulation defects. George Clowes (Boston, MA): I can’t resist speaking about this paper because it has implications far beyond cardiac surgery. What these gentlemen have described is essentially the evolution of septic shock in one of its parts, namely, the leaking capillary syndrome and the development of adult respiratory distress syndrome. Do the au-

564

thors have any explanation for it, because an explanation would lead to other forms of treatment perhaps. It occurs to me that plasma that’s been partly denatured by the separation process has the ability to activate complement, either in itself or in the remaining portion of the plasma in the patient. This means that the macrophages, in particular, are activated, not only to produce the agents that establish the immunologic responses, but more importantly, to produce lysosomal products, namely, proteases, superoxides, and things that can really injure bystander tissue; Jacobs has emphasized this phenomenon in adult respiratory distress syndrome. Furthermore, it may well activate the platelet system and other producers of prostaglandins. If either of these phenomena was the answer, and they could show by complement consumption or prostaglandin analysis that they were different from normal, then of course the treatment would be the use of steroids. It was not mentioned in the paper, but I presume that steroids were used. Rationally that would be correct because the steroids would shut off this activity of the macrophages. Sabet Hashim (closing): In answer to Dr. Cohen, autopsies were performed on the two adult patients; the baby did not have an autopsy. One of these patients died in the operating room, and the lungs at autopsy weighed 300 g each and showed massive pulmonary edema. The grafts were patent, and there was no myocardial infarction even though the time that elapsed between death and the operation was too short to show gross criteria for myocardial infarction. The other patient who was autopsied died 3 weeks after operation from sepsis and renal failure. The lungs were normal in that patient, which shows that the recovery of the lung is rapid. As a matter of fact, the physiologic recovery, as we said before, occurs within a few hours. The heart in that patient was normal (there were no infarctions). The grafts were also patent. In answer to the last discussant, the cortical steroids were given to all of our patients. Every one of them received 1 g methylprednisolone, and they were also given 50 mg of diphenhydramine because the fit thing we thought of was an allergic reaction, even though at that time we were not sure what had caused it. We did not study prostaglandins in these patients, but we did try to find out if the plasma was causing leukocyte agglutination in the recipient. The three tests we performed were unfortunately equivocal and, therefore, we could not assert that this was a problem due to administration of fresh frozen plasma. However, we do believe that this is an allergic phenomenon; it occurs rapidly, as soon as a few milliliters of fresh frozen plasma are given to the patient.

The American Journal ol Surgery

Lihat lebih banyak...

Comentários

Copyright © 2017 DADOSPDF Inc.