Cardiac Dysrhythmia following Pneumonectomy

Cardiac Dysrhythmia following Pneumonectomy" Clinical Correlates and Prognostic Significance Michael]. Krowka, M.D.; Peter C. Pairolero, M.D., F.C.C.~; Victor F. Trastek, M.D., F.C.C.~; w Spencer Payne, M.D., F.C.C.~; and Philip E. Bernatz, M.D., F.C.C.~

Cardiac tachydysrhythmias occurred in 53 (22 percent) of

236 consecutive patients undergoing pneumonectomy. All

patients had preoperative electrocardiograms which showed normal sinus rhythm. Patients did not receive digitalis before surgery. Atrial6brillation was the most common dysrhythmia (64 percent; 34153~ followedby supraventricular tachycardia (23 percent; W53) and atrial flutter (13 percent; 7/53~ No episodes of ventricular tachycardia were documented. Elevated concentrations of cardiac enzymes were associated with 12 (28 percent) of 43 tachydysrhythmias. Recurrent or persistent dysrhythmias were documented in 29 (55 percent) of53 patients despite medical management or electrocardioversion (or both~ Thirty-one percent (9/29) of these patients subsequently died during their hospitalization. There was no correlation between standard preoperative pulmonary function tests and the incidence of postoperative dysrhythmia. In addition, there

The clinical course following pneumonectomy may involve certain complications which have poor correlation to selected preoperative pulmonary function tests, such as the forced vital capacity (FVC)and forced expired volume in one second (FEVJ. 1-3 Cardiac dysrhythmias are well-documented complications following pneumonectomy.t" and in selected patients, these dysrhythmias have been associated with significant mortality. e-s This retrospective analysis of patients undergoing pneumonectomy at the MayoClinic over a three-year period was undertaken to correlate the occurrence of various postoperative dysrhythmias to preoperative pulmonary function and to perioperative clinical data. The prognostic significance for patients who developed postoperative cardiac dysrhythmias was also evaluated. MATERIALS AND METHODS

The records of 244 consecutive patients who underwent pneumonectomyat the Mayo Clinic from 1982 through early 1985 were reviewed. Each patient had routine 12-lead electrocardiograms taken, both before and after pneumonectomy, and at any time that routine

was no correlation of dysrhythmia with postoperative diagnoses, surgical staging for lung cancer, postoperative arterial blood gas levels, or the fact that a completion pneumonectomy or chest wall resection was undertaken. An increased incidence of tachydysrhythmia was noted in patients undergoing intrapericardial dissections and those who developed postoperative interstitial or perihilar pulmonary edema. Twenty-6ve percent (13) of the patients experiencing tachydysrhythmias died within 30 days following their pneumonectomy. We conclude that taehydysrhythmias after pneumonectomy are associated with signi6cant mortality, have poor correlation to preoperative pulmonary function, and occur more frequently following intrapericardial dissection and in patients who develop postoperative interstitial pulmonary edema or perihilar pulmonary edema.

continuous electrocardiographic monitoring in intensive care demonstrated a change in rhythm. Each record was interpreted by a cardiologic consultant. While in intensive care, each patient had daily postoperative portable upright chest roentgenograms interpreted by radiologic consultants. The data reported include all preoperative and postoperative events during hospitalization up until the time ofthe patient's discharge or death. A description of the 236 patients studied is provided in the following tabulation, where the numbers within parentheses represent percentages:

Sex

65 (28) 171 (72)

Female Male

Age

Less than 50 yr Between 51 and 69 yr 70 yr or older (9 older than 80) Indications Primary lung cancer Metastatic disease Inflammatory disease Mesothelioma (malignant) Lymphoma!Av mal.lbroncholithiasis Operative notes RiJdtt pneumonectomy Left pneumonectomy Compretionpn~umon~romy

Intrapericardial pneumonectomy Previous thoracic irradiation Surgical staging

TI

*From the Department ofThoraclc Diseases and Internal Medicine, Section of Thoracic and Cardiovascular Surgery, Mayo Clinic and Mayo Foundation, Rochester, MN. Manuscript received September 11; revision accepted October 14. Reprint requests: Dr. Krowka, 4500 San Pablo lfoad, Jacksonville, FL32224 480

T2

1'3

1121835

19

9 16 4 4 2

112 124 53 7) 28 12 32 14

14

183

9 (4

II 61

28 178 32

NO Nl

N2

38jl6j

149 63 49 21

Stage 1 Stage 2

32 36 12

61)

Cardiac Dysrhythmia following Pneumonectomy (Krowkast aI)

(27)

Stage 3 Reoperation within 72 hr 3O-day mortality Intraoperative deaths Postoperative deaths

8 (3 26 (11

10 (38 16 (62

The operative notes and the intensive care unit record for each patient were reviewed to define medical or surgical problems which developed during surgery orin the immediate postoperative period. tJpon discharge from the intensive care unit, the patient's hospital record was reviewed to determine if complications occurred which required subsequent intervention prior to discharge. For those patients completing preoperative pulmonary function tests, an assessment was made of the FVC, FEV., maximum voluntary ventilation (MVV), and steady-state diffusing capacity (Dss). For any patient developing a cardiac dysrhythmia, the arterial blood gas levels (IL 1302 pH analyzer and IL 282 oximeter) obtained just prior to, or immediately fOllOwing, the dysrhythmia were reviewed for assessment of pH, arterial oxygen pressure (PaO.), and arterial carbon dioxide tension (PaCO.). hi addition, rhythms were considered as recurrent or persistent if they were again documented by 12-lead ECGs beyond 24 hours from initial documentation. RESULTS

All patients in this study had preoperative ECGs which showed normal sinus rhythm. Patients with preoperative atrial fibrillation (N = 8) were excluded; this resulted in records from 236 patients for review. Patients did not receive digitalis before surgery for the purpose of preventiIig postoperative dysrhythmias. Cardiac taehydysrhythmia occurred in 53 (22 percent) of the 236 patients who underwent pneumonectomy. In 12 (28 percent) of 43 patients, documented increases in levels of creatine phosphokinase were found and associated with elevations in the MBfraction (>2 percent). Three of those 12 patients had documented Q-wave patterns of infarction, and nine of the patients had non-Q-wave change abnormalities, primanly in the configuration of the T-waves,compared to preoperative ECGs. In sixof the subjects, elevations of concentrations ofcardiac enzymes preceded the development of the dysrhythmia, and in the remaining six occurred subsequent to the rhythm disturbance. In six of the 53 patients developing tachydysrhythmia, either dtgoxin (three), p-adrenergic blockers (two), or calcium-channel blockers (one) were preoperative medications that had been administered for reasons other than the proposed pulmonary resection and were continued through the perioperative period when the rhythm disturbances occurred. The initial

cardiac dysrhythmia occurring after surgery is shown in Table 1. Sixty-four percent (34) developed atrial fibrillation and no patient developed ventricular tachycardia following pneumonectomy. Hemodynamically, 45 percent (24) of those with dysrhythmias had ventricular responses greater than 150 beats per minute, and 38 percent (20) had relative hypotension associated with the dysrhythmia (defined by a drop in systolic blood pressure of greater than 25 mm Hg). Only two patients had an intraoperative onset of the dysrhythmia (both survived), and 95 percent (50) of the patients developed dysrhythmia within the first six days after surgery. Twenty-nine patients (55 percent) experienced persistent or recurrent 'tachydysrhythmias despite maximal therapy which included aggressive medical management with combinations of digitalis, quinidine, verapamil, and amiodarone, as well as cardioversion. In the four patients In whom cardioversion was attempted, two attempts were successful, and two patients remained in a persistent rhythm abnormality. Of the 53 patients, 5 percent (three) were dismissed from the hospital in atrial fibrillation (rate controlled with medication), 25 percent (13) died in the hospital, and 70 percent (37) were dismissed in normal sinus rhythm. Tachydysrhythmia occurred more frequently than expected (p
Table I-Initial Cardiac DyBrhythmia after Pneumoneciomy* Rhythm

Frequency

Totai Atrial fibrillation Supraventricular tachycardia Atrial flutter Ventricular tachycardia

53 (100) 34 (64) 12 (23) 7 (13) 0

Recurrent or Persistent 29 (55) 20 (69) 4 (14) 5 (17)

With Hypotensiont 20 14 4 2

(38) (70) (20) (10)

Resolved prior to Dismissal 37 23 7 7

(70) (62) (19) (19)

Associated with Mortality 13 (25) 9 (69) 4 (31) 0

*N = 53 (22 percent), 18ble data are numbers of patients; numbers within parentheses are percents. tDrop in systolic blood pressure of at least 25 mm Hg. CHEST I 91 I 4 I APRIL. 1987

481

Table I-Clinical C()fWlata with Cardiac Dy".hythmituJ following Pneumonectomy· Parameter Preoperative correlates All patients Previous thoracic irradiation FEV.<2.0 L FEV.>2.0 L MVV<80 percent of predicted MVV>80 percent of predicted Mean MV\': percent of predicted Mean Dss Age<50 yr Age 51-59 yr Age 70+ yr Postoperative correlates Diagnosis Primary lung cancer Metastatic lung cancer In8ammatory disease Mesothelioma (malignant) Tl surgical stage T2 surgical stage T3 surgical stage NO surgical stage Nl surgical stage N2 surgical stage Stage 1 Stage 2 Stage 3 Pa02<60 mm Hg pH<7.35 pH>7.45 Completion pneumonectomy Chest wall resection Intrapericardial dissection Interstitial or perihilar edema' 3O-day mortality

N 236 9 58 90 46 104

ISO 94

38 149 49 197 14 12 9 21 112

SO

58 57 68 20 112 51

53t 53t 53t

28 13 32 58 26

With DysWithout rhythmia Dysrhythmia 53 (22) 3 17 22 12 25 94 17.7 6 31 16

41 3 5 3

5

22 12 11 12 16 5

22

12 8 (15) 6 (11) 13 (25) 8 4 17*

3011

13'

183 (78) 6 41 68

34 79 88 18.8 32 118 33

156 11 7 6 16 90

38 47 45 52 15 90 39

20

9

15*

2811

13'

*N =236.

Table data are numbers of patients; numbers within parentheses are percentages. t Arterial blood gas levels not assessed unless dysrhythmia occurred. *X2 = 19.80; p
need to complete chest wall resections. The 3O-day mortality in the series was 11 percent (261236). There were ten intraoperative deaths and 16 postoperative deaths. Table 3 summarizes relevant data. Intraoperative deaths were not caused by disturbances in cardiac rhythm; hemorrhage was the major problem. Of the postoperative deaths, 81 percent (13116) were associated with or preceded by tachydysrhythmias, and 69 percent (9/13) of those patients had persistent or recurrent dysrhythmias. Nine (31 percent) of 29 patients with recurrent or persistent tachydysrhythmia died during hospitalization. Twenty-five percent (13) of the 53 patients developing tachydysrhythmias died, compared to 7 percent (131183) who died without dysrhythmias (p
cent (7/13) of dysrhythmias occurred within 96 hours of death. If dysrhythmia occurred prior to 72 to 96 hours of death, prolonged hypoxemia, hypotension, and recurrence or persistence of dysrhythmia were present prior to death. DISCUSSION

This retrospective analysis from a tertiary referral center showed that patients who underwent pneumonectomy who did not receive digitalis before surgery but went on to develop persistent or recurrent tachydysrhythmias experienced a greater degree of mortality than those who maintained normal sinus rhythms after surgery. The difference in mortality was significant. Tachydysrhythmia, regardless of ventricular response, appeared to have no apparent association with preoperative pulmonary function, surgical indication, or TMN staging of lung cancer. 'Iachydysrhythmia did occur more frequently than expected in those with interpericardial dissections and postoperative interstitial or perihilar pulmonary edema. Previous studies concerning cardiac dysrhythmias following pneumonectomy are summarized in Table 4. Our data represent the largest series assessing the relationship between dysrhythmia and mortality in patients who underwent pneumonectomy. Of those series which correlated the incidence of dysrhythmia with subsequent mortality, 23 to 33 percent ofpatients with significant rhythm disturbances subsequently died, a higher percentage than the accepted mortality for pneumonectomy. Statistics on 30-day mortality were not available in any other study except that of Keagy et aI,I in which they reported a similar 30-day mortality; however, they did not correlate the onset of rhythm disturbances to subsequent mortality. The incidence of dysrhythmia in patients summarized in Table 4 ranged from 9 to 29 percent, with the major rhythm disturbances being atrial fibrillation (most commonly), supraventricular tachycardia, and atrial Buttet: The data in these studies were reviewed to eliminate the inclusion of patients with premature ventricular complexes or premature atrial complexes as part of the definition of cardiac dysrhythmia following pneumonectomy. Ventricular tachycardias were extremely uncommon in the studies reviewed, and none occurred in our study, even with the onset of perioperative Q-wave or non-Q-wave electrocardiographic changes associated with elevated concentrations of cardiac enzymes. The only study that significantly addressed prophylactic digitalization, that of Shields and Ujiki," showed a decreased incidence of dysrhythmia following prophylactic digitalis but did not discuss the incidence of rhythm disturbances as they related to subsequent mortality. Our study is different from the previous studies in that most of our patients had continuous C8rdtac Dysrhythmia following PneumotMK:tomy (Krowka et aJ)

Table 3-ClinictJl DaIG ofTIaoN Who Died within 30 Daya following Pneumonectomy* Group and Case/Age (yr) Intraoperative 1/66 2/68 3/39 4180 5163

Surgical Indication, cell typet

Procedure or Stage

L squamous R adenocarcinoma R metastatic Wilms' tumor L metastatic squamous R squamous L large cell R Lymphoma/BP fistula R adenocarcinoma R squamous R squamous

T3N2 Completion

In3 2/82

L alveolar

3/52

R mesothelioma L squamous

T2Nl T3N 1; chest wall resection Pericardial window T2N2

6165

5152

7n4

R mesothelioma R squamous R squamous

T2NO T3N2

8136

R AV malfOrmation

9/73

L adenocarcinoma

10180

13/84

R squamous R metastatic osteosarcoma R metastatic adenocarcinoma R squamous

14/65

R squamous

15164

R metastatic breast

16168

R squamous

6154 7n4

8153

9/75

10164

Postoperative

4159

11/15 12/62

R squamous

Postoperative Day of DysrhythmialDeath

Hypotension

Recurrence or Persistence

Commentsi

Arterial hemorrhage Hemorrhage Hemorrhage (diffuse) Arterial hemorrhage Hemorrhage SVC hemorrhage Arterial hemorrhage

Completion Completion Completion

Hemorrhage (diffuse) Arterial hemorrhage Hemorrhage

Completion T3NO T3N2

2/4 3/6

Yes No

No Yes

23/27 2/23

Yes No

Yes Yes

None/6 24/28 619

No No

No No

6/6

Yes

Yes

T3N2 atrial resection

3/20

Yes

Yes

Completion

None/15 Nonell0

Yes

7/16

Yes

Yes

TIN2

619

No

No

Completion; intrapericardial

4/16

Yes

No

1/16

Yes

Yes

Intrapericardial

7/21

Yes

Yes

Perioperative MI Hypoxemia/pulmonary hypertension Cardiac compression; METS Hypoxemia; pulmonary hypertension; BP 6stulat COl retention; aspiration Pneumonitis; sepsis; lobar collapse Effusion; tension pneumothorax; ARDS Staphylococcal pneumonitis; subendocardial MI Stump dehiscence Pulmonary emboli Hypoxemia; pulmonary emboli Hypoxemia; ARDS; pneumonitis Hypoxemia; sepsis Stump dehiscence; sepsis; pulmonary hypertension Hemothorax; reoperation: hypoxemia; stump dehiscence

*N=26. bronchopleural; and AV, arteriovenous malfOrmation; METS, metastatic disease; none, no dysrhythmia occurred, postop day of death added. iSVC, Superior vena cava; MI, myocardial infarction; and ARDS, adult respiratory distress syndrome. tB~

electrocardiographic monitoring in the intensive care unit after surgery, which certainly improved the detection of cardiac dysrhythmias, especially those which may have been asymptomatic. In addition, the types of patients who were referred to our institution for subsequent pulmonary resection had had previous thoracic surgery, as well as other types of therapy such as chemotherapy and irradiation. Speeifcally, completion pneumonectomy was not uncommon, and although it did not correlate with the incidence of dysrhythmia, it appeared to be associated with a higher

3O-day mortality (see Table 3). Our 30-day mortality of II percent is similar to that reported in the series by Keagy et al," Compared to the recent Lung Cancer Study Group analysis from five participating centers analyzing 569 pneumonectomies, a 3O-day mortality of 6.2 percent was reported, but those data were not correlated with cardiac dysrhythmia. 10 In reviewing those studies presented in Table 4, proposed causes of dysrhythmia after pneumonectomy have included hypoxemia, vagal irritation, atrial inflammation, preexisting cardiac disease, pulmonary CHEST I 91 I 4 I APRIL, 1887

413

Table 4-CardiGc DysrhfitlatnitJa following Pneumonectomy·

Study Baily and Betts" Currens et al6 Massie and Valle8

Year

Mowry and Beynolds" Shields and Ujiki9

1943 1943 1947 1957 1964 1968

Keagy et all Krowka et al

1983 1987

Cerney'

No. of Pneumonectomies

78

43

120

34 72

9t

18;

58 236

Incidence of Dysrhythmia, percent 10 21 9 29 19 11 6 28 22

Dysrhythmias Associated with Mortality, percent (cases) No deaths reported

3O-Day Mortality, percent (cases)

33 (319)

30 (3110) Not discussed

Unavailabie Unavailable Unavailabie Uhavailable Unavailable

Not discussed Not discussed 25 (13153)§

Unavailable 12 (11/90) 11 (26/236)

23 (3113)

*Includes only atrial fibrillation ("auricular fibrillation"), atrial8utter, nodal or supraventricular tachycardia, and ventricular tachycardias.

tNot prophylactically digitalized.

*Prophylactica1ly digitalized. §3O-day mortality.

hypertension, and right cardiac dilation. In most of these studies, the onset of dysrhythmia was during the first postoperative week. Apparently, most patients responded well to combinations of digitalis, quinidine, or no medication. There appeared to he an increased frequency with age and, in at least one report, a ques.. tion of anesthesia-related cause for dysrhythmia." We could not demonstrate an age-related association. A previous study also addressed the association between previous thoracic irradiation and the frequency of cardiac rhythm disturbances following pneumonectomy,'' Although a relationship was suggested, it was not a statistically significant finding in that study or in our data. In no study could hypoxemia or preoperative pulmonary function be related to the rhythm disturbances. None of the studies addressed hemodynamic consequences (specifically, hypotension) in terms. of the effect of dysrhythmia. Our data did not suggest that hypoxemia is a major contributing factor to the development of dysrhythmia, especially acute hypoxemia. The association of cardiac dysrhythmia with pulmonary edema or increased interstitial infiltrates on the chest x-ray film after pneumonectomy is intriguing. The entity of pulmonary edema after pneumonectomy has recently been discussed by Zeldin et al," who reported the findings in ten patients. These investigators indicated that this phenomenon usually occurs within 48 hours after surgery and can occur in a relatively hemodynamically stable patient. Risk factots appear to be right pneumonectomy and interoperative administration of fluid in an amount greater than 2,000 ml, Pulmonary capillary wedge pressures are not necessarily elevated in these individuals. In fact, pulmonary arterial pressures may be falsely low following pneumonectomy due to balloon occlusion causing increased right ventricular afterload, reduced cardiac output, and decreased left atrial pressure. 13 It appears, based on other studies," that pulmonary capillary hydrostatic pressure can be raised in the setting ofiow

or unaltered precapillary vascular tone and, hence, result in acute fluid overload of the remaining lung. Based on our chest roentgenographic results, it would appear that certainly this entity is more common than realized and may indeed be responsible for an acute subclinical fluid overload in some cases resulting in the dysrhythmias noted. This needs further prospective study. The relationship of intrapericardial dissection and cardiac dysrhythmia following pulmonary resection has not been addressed in the literature. The relationship between pericardial disease and dysrhythmia in nonsurgical patients has been documented," but the mechanism of intrapericardial dissection as a cause for dysrhythmia may not be comparable. This finding needs additional confirmation. Finally, as previously documented," We could find no relationship between cardiac rhythm disturbances following pneumonectomy and preoperative pulmonary function. This finding Was not unexpected. and reinforces the fact that preoperative pulmonary function tests are not useful parameters on which to base the postoperative clinical course with reference to cardiac and hemodynamic function. 2.15 The implications of this study focus on the increased mortality which occurred in patients that had nonventricular tachycardias after pneumonectomy. The role of digitalis or other rhythm-stabilizing prophylactic medications is unclean Whether or not prophylaxis would result in fewer dysrhythmias and subsequent mortality and morbidity has not been well addressed in the literature. Our retrospective study, although documenting the effects of postoperative tachydysrhythmias, does not imply that prophylactic cardiac medications will favorably affect morbidity and mortality. Future studies should address the hypothesis that cardiac dysrhythmias after pneumonectomy are, in part, volume-induced problems which may not correlate well to standard hemodynamic monitoring C8rdiac Dysrhythmiafollowing Pneumoi1ectomy (KtrNlka et aI)

techniques. Finally, the introduction of newer cardiac medications (ie, calcium-channel blockers and amiodarone), noninvasive cardiac techniques (ie, Doppler echocardiography and exercise stress testing), and a combined cardiopulmonary approach to the patient with a potential pulmonary resection may be of clinical importance. This type of consideration, in conjunction with a prospective trial of rhythm-stabilizing medications, should help remove the mystique surrounding preoperative digitalization in the candidate for pulmoresection.

nary

REFERENCES 1 Kirsh MM, Botman H, Behrenst DM, Orringer MB, Sloan H. Co~plications of pulmonary resection. Ann Thorac Surg 1975; 2Q:215-36 2 KeagyBA, Sborlemeyer GR, Murray GF, Starek PIC, Wl1cox BR. Correlation of preoperative pulmonary function testing with clinical course in patients after pneumonectomy. Ann Thorac Surg 1983; 36:253-57 3 Kohlman LJ, Meyer }A, Ikins PM, Oates BE Random versus predictable risks of mortality after thoracotomy for lung cancer. I Thorac Cardiovasc Surg 1986; 91:551-54 4 Bailey CC,"Betts RH. Cardiac arrhythmias following pneumoneetomy, N Eng} I Med 1943; 229:~ 5 Mowry F~, Reynolds EW Jr. Cardiac rhythm disturbances com-

plicating resectional surgery of the lung. Ann Intern Med 1964;

61:688-95

6 Currens JH, White PO, Churchill ED. Cardiac arrhythmias following thoracic surgery. N Eng) J Med 1943; 229:360-64

7 Cerney CI. The prophyluis ofcardiac arrhythmias complicating

pulmonary surgery. I Thorac Surg 1957; 34:105-10 8 Massie E, Valle AB. Cardiac arrhythmias complicating total pneumonectomy. Ann Intern Med 1947;26:231-39 9 Shields lW, Ujiki GT. Digitalization for prevention of arrhythmias following pulmonary surgery. Surg Gynecol Obstet 1968; 126:743-46 10 Ginsburg RI, Hill LD, Eagan Thomas ~ Mountain CF, Deslauras I, et ale Modem thirty day operative mortality for surgical resections in lung cancer. I Thorac Cardiovasc Surg 1983; 86: 654-58 11 MarkJBD, Call E~ von Essen Cit: Preoperative irradiation in patients undergoing pneumonectomy for carcinoma of the lung. J Thorac Cardiovasc Surg 1966;51:30-5 12 Zeldin RA, Normandin D, Landtwing D, Peters RM. Postpneumonectomy pulmonary edema] Thorac Cardiovasc Surg 1984;

m:

87:~9-69

13 Wittnich C, 'Irudel ], Zidul1caA, Cbu-Jeng R. Misleading "pulmonary wedge pressure tt after pneumonectomy: its importance in postoperative fluid therapy. Ann Thorac Surg 1986; 42:192-96 14 Hutchin ~ Terzi ac, Hollandsworth LA, Johnson G Jr, Peters RM. Pulmonary congestion following infusion oflarge fluid loads in thoracic surgery patients. Ann Thorac Surg 1~; 8:339-46 Sanders CA. Clinical cardiology. New York: Grune 15 Willerson and Stratton, Inc, 1977:382-94

n:

Congres de Pneumologle de Langue Prancalse This congress on pneumology (in the French language) will be held June ll-13 in Grenoble at the faculty of medicine. For information, contact Dr. Cb. rambilla, PavilIon D2, ~B. 217 X, 38043 Grenoble Cedex, France.

a

CHEST I 81 I 4 I APRIL, 1987

411