Is Atelectasis Following Aortocoronary Bypass Related to Temperature?

Is Atelectasis Following Aortocoronary Bypass Related to Temperature?* Jackie A Thomas, MD; Robert James Cusimano, MD; and Victor Hoffstein, MD

Objective: To determine the frequency of acute postoperative atelectasis in patients undergoing aortocoronary bypass with either normothermic (warm) or hypothermic (cold) technique. Design: Prospective, randomized study comparing two groups. Setting: University-affiliate d hospital. Patients: Three hundred thirty-one patients (166 cold and 165 warm) undergoing isolated aortocoronary bypass. Measurements: Chest radiographs were obtained preoperatively, on the day of surgery, and subsequently as clinically indicated until discharge from the hospital. Radiologist (blinded to the patient allocation into warm or cold group) scored the atelectasis from 0 to 3 based on its severity. Regression analysis was used to determine if there was any difference in the atelectasis scores between the two groups. Results: Mean daily postoperative atelectasis scores were not different between the cold and warm groups. The number of patients requiring chest radiographs was similar in both groups. The percent of patients with abnormal chest radiographs was similar in both groups. Conclusion: The temperature of cardioplegia has no effect on the development of atelectasis following aortocoronary bypass, and therefore temperature-relate d cold injury is not a major cause of atelectasis following this type of surgery. (CHEST 1997; 111:1290-94) Key words: atelectasis; cold heart surgery; warm heart surgery

ulmonary abnormalities following cardiac surgery P are common. The most common of these is

atelectasis, especially of the left lower lobe. 1-4 Although many theories have been advanced regarding the etiology of the atelectasis, the one referred to the most has been related to cold-induced phrenic nerve injury.3 ·5 - 7 Various strategies to decrease the incidence of cold-related injury involved the elimination of an ice slush topical coolant,8 cooling jackets ,8 and insulating pads.9 However, all studies have utilized cold techniques as the primary form of myocardial and systemic protection. Warm heart surgery described recently 10 provides us with an opportunity to assess whether cold has any effect on the incidence of postoperative atelectasis. To answer this question, we studied patients operated on at normothermia using normothermic cardioplegia, and compared them with those undergoing the same operation under hypothermic conditions.

*From the Division of Cardiovascular Surgery (Drs. Thomas and Cusimano), The Toronto Hospital, and the Department of Medicine (Dr. Hoffstein ), St. Michael's Hospital, Toronto, Ontario, Canada. Manuscript received January 5, 1996; revision accepted November 8. Reprint requests: Dr. Victor Hoffstein, St. Michael's Hospital, 30 Bond Street, Toronto, Ontario, Canada MSB lWB 1290

MATERIALS AND METHODS Patients undergoing isolated aortocoronary bypass, recruited at St. Michael's Hospital between Nove mber 1990 and December 1992 participated in this study. All those scheduled for surge1y were deemed eligible, providing a n informed consent was obtained. Relative contraindications included cerebrovascular disease or renal insufficiency. Patients were randomized prospectively to receive either intermittent cold blood cardioplegia (5 to 8°C) as the form of myocardial protection or continuous normoth ermic (37°C) blood cardioplegia either in an antegrade or retrograde fashion. The nmdomization was by sealed envelope and occurred after the patient arrived in the operating room. These patients were a part of a larger, three-cente r randomi zed clinical trial compating warm and cold cardioplegia; full description of the human ethics committee approval, inclusion/exclusion criteria, enrollment tim e, number of surgeons, and random ization methods bas been published alreadyW Systemic temperatures were between 25 and 30°C for the cold group and between 33 and 37°C for the warm group. Insulating and cooling padswere not used in either group. Topical slush was used in 25 patients receiving cold cardioplegia. Its use was allowed as pe r the usual routine of the operating surgeon. The primary outcome that was measured in the present study was the development and amount of atelectasis. Since the clinical gold standard in the measurement of atelectasis is chest radiograph, this was used as the primmy outcome measure. Chest radiographs were perform ed as per th e usual routine of the hospital, ie, one preoperative radiograph , one the day of the surge1y, and however m any subsequent radiographs that were Clinical Investigations

deemed to be clinically necessary until th e day of hospital discharge. No additional radiographs were done as part of this study. To assess the atelectasis, the radiologists were given a standard method of reporting postoperative chest radiographs. The radiologists reporting the radiographs were blinded as to the allocation of the patients unde r study. They were asked to grade atelectasis as (l ) plate-like, (2) subsegmental or segmental, and/or (3) lobar atelectasis. This grading was believed to represent a continuum in severity of atelectasis and thus the mean atelectasis score was assigned for each group. All other radiologic findings, such as cardiomegaly, pleural effusions, pulmonary edema, etc, were reported separately. During preoperative assessment, the following information was recorded: age, sex, height, weight, smoking history (number of pack years of smoking), and body surface area. Intraoperative variables recorded included the body temperature of the patient, temperature of the cardioplegia, volume of cardioplegia used, time receiving cardiopulmonary bypass, aortic cross-clamp time, and the number of bypass grafts. Postoperatively, the mean atelectasis score for each group was calculated on a day-to-day basis for those patients having chest radiographs performed that day. Other postoperative variables assessed included postoperative duration of mechanical ventilation , postoperative myocardial infarction , death, low-output syndrome, and cerebrovascular accident. Statistical Analysis

Data are shown as mean::t:SD, and t tests and x2 tests were used when appropriate. An alpha level of 0.05 was assumed to be significant (tvvo-tailed). The main statistical method used to analyze these data was the Liang-Zeger statistics.ll· 12 This is a method for analyzing longitudinal data using regression models. It assumes that there is a correlation between the obseJvations on the same subject. Compared with the repeated measures analysis of variance, this method offers two advantages. First, it can be used with many different types of regression models-logistic regression and continuous regression. Second, it accommodates the uncertainty about the precise correlation between intrasubject variables. As such, this method provides robust information

about the relationship between the outcome variable (ie, atelectasis) and the associated explanatory variables (ie, type of cardioplegia). A power calculation to determine the sample size was perform ed. We found that a sample with more than 300 subjects with approximately equal allocation of treatm ent would have sufficient power (0.80 at a two-tailed alpha of 0.05) to detect treatm ent diffe rences if the tru e population odds ratio was at least 1.8.

RESULTS

Three hundred thirty-one patients were studied. One hundred sixty-five patients had the warm technique while 166 had the cold method. As expected, because of the randomization, the two groups were similar with respect to preoperative factors (Table 1). Comparison of the intraoperative variables (Table 1) shows that the warm group received a greater amount of cardioplegia solution. This was related to the use of a continuous vs intermittent method of delivery. Twenty-five patients who received cold cardioplegia had topical forms of hypothermia in addition to intracoronary administration of cardioplegia. They did not differ in terms of any preoperative, intraoperative, or postoperative variables. The result of the analysis leaving this "topical" group out was not different. They were thus included in the cold group. The mean body temperature for the warm group was 35.4±l.3°C compared with 29.5±2.9°C (p<0.0001 ) for the cold group. Postoperative variables (Table 1) indicate that there were no significant differences in clinical outcome variables for the two groups; postoperative ventilation time, number of chest tubes, and other complications were similar in the two groups.

Table !-Patient Variables* Preoperative

Group

No.

Age, yr

No. (%) of men

Warm Cold

165 166

60::'::8.6 60::'::9.0

130 (79) 132 (79)

1.97::'::0.45 1.95::':: 0.41

Smoking

LV Grade

35.0::':26.4 36. 1::':24.8

2.03::'::0.80 2.02::'::0.79

Intraoperative

Group

No. ofCABGs

Cross-Clamp Time, min

Bypass Time, min

Cardioplegia, L

Warm Cold

3.1::':0.8 3.1::': 1.0

71.2::':27.6 70.0::'::30.2

95.8::':34.2 101.5 ::':35. 1

5.0::':2.1 1 3.7::':1.6

Postoperative

Group

Hours Ventilated

No. of Chest Tubes

No. (%) Died

No. (%) With Low Output

CVA, No. (%)

MI, No. (%)

\Vann Cold

43::':79 42::':63

2.9::':0.7 2.9::':0.7

3 (1.8)

17 (10) 27 (16)

4 (2) 3 (2)

24 (15) 17 (11 )

6 (4)

*BSA=body surface area; LV = left ventricular; CABG=coronary arte1y bypass graft; CVA=cerebrovascular accident; MI = myocardial infarction. 1 p<0.05. CHEST/111 /5/MAY, 1997

1291

Table 2 shows the mean daily atelectasis score for the two groups for the first 5 days, including the first postoperative chest radiograph taken on return from the operating room. There was a change with time, particularly in the early period, but after that there was little daily change. The two groups did not differ with respect to each other. This lack of difference between the groups is also evident from Figure 1, which shows similar distribution of the atelectasis scores for the warm and cold groups. All patients exhibiting atelectasis showed changes confined to the left lower lobe. Table 2 shows the number (and percentage) of the radiographs that did not show any radiologic evidence of atelectasis. There was no difference between the two groups. The l analysis of atelectasis scores using 2 X 2 tables (warm vs cold groups, atelectasis vs no atelectasis) showed no difference between the groups. Using the methods of Zeger and Liang,n two models were utilized to determine if there were any treatment (ie, temperature) effects.l 2 Neither showed any treatment effects, while both showed a strong time effect as expected from the data shown in Table 1, ie, worsening scores with time. There were no differences between the two groups with respect to other radiographic findings, such as pleural effusion, cardiomegaly, pulmonary edema, etc.

DISCUSSION

This study demonstrates that temperature of cardioplegia has no effect on the frequency of atelectasis following aortocoronary bypass surgety. We defined atelectasis based on the classic features of the chest radiograph. It is possible, although unlikely, that in some patients, atelectasis was missed because of atypical appearance of the chest radiograph. Alternatively, there is a possibility that "typical" appearance may not represent volume loss at all, but fluid or inflammation. In the absence of CT scans, which could not be justified on clinical

OWarm

-

•Cold

c

~

G)

a..

Atelectasis score FIGURE

l. Distribution of atelectasis scores: warm vs cold groups.

grounds, we cannot be absolutely certain that chest radiographic diagnosis of atelectasis is correct. However, this issue is probably irrelevant, because our aim was not to determine the incidence of atelectasis after aortocoronary bypass surgety, but to compare the frequency of well-defined radiographic abnormalities (usually termed "atelectasis" by the radiologists) in two groups of patients-those undergoing warm and cold heart surgery. Our finding showing lack of any significant difference in the frequency of these radiologic abnormalities benveen the hvo groups remains valid regardless of whether there was actual volume loss. Pulmonary complications are common sequelae of cardiac surgery. One of the most frequently seen complications is atelectasis, particularly of the left lower lobe. The reasons for this are not well known. One possibility, first proposed by Scannell, 13 assumes impairment in the function of the left hemidiaphragm secondary to the use of cold cardioplegia. It is well known that peripheral nerves exposed to cold exhibit reduction in their conduction velocity and progressive pathologic changes. 14 ·15 Markland et al4 and Wilcox et al 16 found that the conduction

Table 2-Mean Atelectasis Scores Cold Cardioplegia Group

Warm Cardioplegia Group Postoperative Day 0 1 2 3 4 5

Atelectasis Score

No. of CXRs* (% Normal)

Atelectasis Score

No. of CXRs* (%Normal )

0.58:t:0.81 0.92:t:0.87 l.09:t:0.86 l.l8:t:0.96 l.23:t:0.89 l.39:t:0.84

145 (61 ) 155 (40 ) 145 (32) 79 (33) 44 (27) 18 (22)

0.56:t:0.78 0.91:t:0.86 l.28:t:0.84 l.24:t:0.94 l.l8:t:0.83 l.l9:t:0.87

148 (61 ) 159 (40) 143 (27) 96 (28) 49 (24 ) 31 (8)

*CXR=chest radiograph. 1292

Clinical Investigations

latency of the phrenic nerve following bypass surgery was prolonged. Both groups reported approximately 10% incidence of diaphragmatic dysfunction following operation. However, radiographic evidence of atelectasis appears to be significantly more common-29% in the study by Thomas and Cusimano, 17 suggesting that although cold may play a part in damaging the phrenic nerve, other factors are also involved. Since the purpose of our study was not to investigate the mechanism of atelectasis, but to compare its frequency in the warm and cold groups of patients, no assessments of phrenic nerve function were carried out. Wheeler et al 9 and Esposito and Spencer6 used a prefabricated insulating pad in the pericardium to protect the phrenic nerve and maintain a cool myocardium. Both groups of authors found that with the pad, four to seven times fewer patients had elevated left hemidiaphragm than without it. Furthermore, this occurred only with the use of iced saline solution slush (below 0°C), and not with the use of saline solution cooled to 4 °C. It is not known to what extent a combination of cold cardioplegia with and without topical hypothermia alters the frequency of atelectasis. If it is truly the cold that damages the phrenic nerve, one would expect both phrenic nerves to be equally affected, and furthermore patients not exposed to cold would not develop atelectasis. This was not the case. Not only did atelectasis occur in the normothermic group, but in all cases it occurred in the left lower lobe. Our results show that the atelectasis is present in many patients already on the first postoperative chest radiograph. As time went on, the mean atelectasis score gradually rose but was not different behveen the groups. This gradual rise in atelectasis over the first few days may represent a fictitious worsening of atelectasis, since postoperative radiographs of the chest were taken only when required after the second day. Only those patients who warranted a radiograph for clinical reasons had one. Alternatively, it is possible that some patients had exhibited a progression of radiographic changes. The important point is that the same trend was seen in both groups, without regard for whether the patients were operated on under hypothermic or normothermic conditions. It is likely that the mean atelectasis scores for the two groups would have been lower had all patients received radiographs on subsequent days, but it is unlikely to introduce a statistically significant difference behveen the groups. If it is not the temperature that affects the incidence of atelectasis, what causes this common problem seen so frequently after cardiac surgery? One possibility is underventilation of the dependent parts

of the lung, which may produce atelectasis within 10 min of onset of anesthesia. 18 Predominance of the left lower lobe involvement indicates that local factors are involved. One such factor may be distention of the stomach (occurring during the bagging phase of induction of anesthesia), causing detrimental effect on the expansion of the left lower lobe. Another factor may be suspension of the pericardia! sac to the chest wall during the operation. This brings the heart and great vessels closer to the surface and stretches the pericardium so that it may be re-closed. This manipulation places tension on the left mainstem or lower lobe bronchus, favoring left lower lobe collapse. In addition, anatomic alterations may also stretch the phrenic nerve, further promoting the atelectasis. Since we did not measure the activity of the phrenic nerve or the diaphragm, we cannot rule out that it is the dysfunction of these organs that causes atelectasis. However, independently of the precise mechanism of atelectasis, temperature does not need to be implicated. Furthermore, since we found a similar degree of atelectasis in the hypothermic and normothermic groups, we believe it unlikely that cold-induced phrenic nerve injmy is the culprit of simple atelectasis seen after cardiopulmonary bypass. While we still do not know why atelectasis occurs, it does not appear to be a response to cold phrenic nerve temperatures. It may be that systemic temperature, which did not differ greatly between the two groups, has an important effect on the development of atelectasis. Alternatively, perhaps study of the conformational changes or other effects on the trachea and lungs during the operation "rill help to identify the cause(s) of postoperative atelectasis. ACKNOWLEDGMENT: The authors gratefully acknowledge the cont1ibution of Marko Katie at the Department of Research Design and Biostatistics at Sunnybrook Health Sciences Center, Toronto, for his assistance in the statistical analysis of this project.

REFERENCES 1 Turnbull KW, Miyagishima RT, Gerein AN. Pulmonary complications and cardiopulmonaty bypass: a clinical study in adults. Can Anaesth Soc J 1974; 21:181-93 2 Gale CD, Teasdale SJ, Sanders DC, et al. Pulmona1y atelectasis and other respiratory complications after cardiopulmonary bypass and investigations of aetiological factors. Can Anaesth Soc J 1979; 26:15-21 3 Benjamin JT, Cascade PN , Rubenfire M, et al. Left lower lobe atelectasis and consolidation follov.ing cardiac surgery: the effect of topical cooling on the phrenic nerve. Hadiology 1982; 142:11-4 4 Markland ON, Mom·thy SS, Mahomed Y, e t al. Postoperative phrenic nerve palsy in patients with open-heart surgery. Ann Thorac Surg 1985; 39:68-73 5 Marco JD, Hahn JW, Barne r H. Topical cardiac hypothermia and phrenic nerve injury. Ann Thorac Surg 1977; 23: 235-37 CHEST/111 /5/MAY, 1997

1293

6 Esposito RA, Spencer FC. The effect of pericardia! insulation on hypothermic phrenic nerve injury during open-heart surgery. Ann Thorac Surg 1987; 43:303-08 7 Diehl JL, Lofaso F, Deleuze P, et a!. Clinically relevant ruaphragmatic dysfunction after cardiac operations. J Thorac Cardiovasc Surg 1994; 107:487-98 8 Roussou JA, Parker T, Engelman RM, eta!. Phrenic nerve paresis associated with the use of iced slush and the cooling jacket for topical hypothermia. J Thorac Cardiovasc Surg 1985; 89:921-25 9 Wheeler WE, Rubis LJ, Jones CW, et a!. Etiology and prevention of topical cardiac hypothermia-induced phrenic nerve injury and left lower lobe atelectasis during cardiac surgery. Chest 1985; 88:680-83 10 Warm Heart Investigators. Randomized trial of normothermic versus hypothermic coronary bypass surgery. Lancet 1994; 343:559-63 ll Zeger SL, Liang KY. Longitudinal data analysis for ruscrete and continuous outcomes. Biometrics 1986; 42:121-30

1294

12 Matthews DE, Farewell Vf. Using and understanding medical statistics. 3rd ed. New York: Karger 1996; 162-65 13 Scannell SC. Discussion of McGoon DC, Mankin HT, Kirklin JW: Results of open heart operation for acquired aortic valve disease. J Thorac Cardiovasc Surg 1963; 45:47-66 14 Franz DN, Iggo A. Conduction failure in myelinated and nonmyelinated axons at low temperatures. J Physiol 1980; 199:319-28 15 Nukuda H, Pollock M, Allpress S. Experimental cold injury to peripheral nerve. Brain 1981; 104:779-811 16 Wilcox P, Baile EM, Hards J, et a!. Phrenic nerve function and its relationship to atelectasis after coronary artery bypass surgery. Chest 1988; 93:693-98 17 Thomas JT, Cusimano RJ. Left lower lobe atelectasis following aortocoronary bypass surgery [abstract]. Clin Invest Med 1990; 13:bll5 18 Strandberg C, Tokics L, Brismar B, et a!. Atelectasis during anaesthesia and in the post-operative period. Acta Anaesthesia! Scand 1986; 30:154-58

Clinical Investigations