Outcomes in Patients With Severe Chronic Lung Disease Undergoing Index Cardiac Operations

Abstract

Background.

This study evaluated the impact of severe chronic lung disease on outcomes of index adult cardiac operations.

Methods.

A single-center, retrospective study of adult patients with severe chronic lung disease (as defined by The Society of Thoracic Surgeons) undergoing index cardiac operations between 2010 and 2018 was performed. Multivariable Cox regression and Kaplan-Meier analyses were used to evaluate survival.

Results.

Three hundred fifty-four patients (median age, 69 years; 32.77% women) were identified. Current smokers comprised 42.66% of the population, and 34.65% of patients required home oxygen. Median preoperative forced expiratory volume in 1 second was 48% of predicted (interquartile range, 41%−56%), and median diffusing capacity of the lungs for carbon monoxide was 78% of predicted (interquartile range, 55%−101%). Most patients underwent isolated coronary artery bypass (57.06%) or isolated aortic valve replacement (19.49%). Overall, 33 patients (9.07%) required a tracheostomy (median of 10 days from surgery) for a median of 49 days (interquartile range, 25–114) until decannulation. Preoperative home oxygen use was an independent predictor of 30-day (hazard ratio, 2.91; P = .030) and 1-year (hazard ratio, 2.12; P = .009) mortality. One-year and 5-year postoperative survival were 83.62% and 58.34%, respectively.

Conclusions.

Although severe chronic lung disease is a predictor of mortality and morbidity after index cardiac operations, only 9% of patients required a tracheostomy, and most were alive at 5 years after surgery. Home oxygen use may serve as a further stratification tool in this higher risk subset; however the presence of severe chronic lung disease alone should not deter from surgery in otherwise reasonable surgical candidates.

An estimated 33.2 million adults (13.4% of the US population) carry a diagnosis of chronic lung disease.¹ Because risk factors for chronic lung disease, specifically smoking history, overlap with cardiovascular disease risk factors, it is critically important to understand the interplay between both disease processes, particularly in patients undergoing cardiac surgery.² Prior studies focusing exclusively on patients with chronic obstructive pulmonary disease (COPD) have demonstrated increased 30-day mortality after cardiac operations,³ reduced long-term survival after coronary artery bypass grafting (CABG),⁴ and increased adjusted hazards for long-term mortality in patients undergoing surgical aortic valve replacement (AVR).⁵ Additionally baseline lung disease has been associated with the need for prolonged mechanical ventilation after open heart surgery,^6,7 which in turn increases the risk of requiring tracheostomy placement.

Taken together, lung disease is an important contributor to adverse outcomes in patients undergoing cardiac surgery, although scarce data correlate the severity of lung disease with survival and complication rates. Thus there is insufficient evidence with which to counsel patients, particularly those with more severe stages of lung disease, on anticipated postoperative outcomes. The aim of this study was to evaluate the impact of preexisting severe chronic lung disease on outcomes of index cardiac operations.

Patients and Methods

Study Cohort

We identified adult patients (≥18 years) with severe chronic lung disease who underwent index cardiac operations between 2010 and 2018. Chronic lung disease includes patients with COPD, chronic bronchitis, emphysema, occupational/environmental exposures, or prior lung radiation. As specified by The Society of Thoracic Surgeons (STS), severe chronic lung disease was defined as a forced expiratory volume in 1 second (FEV₁) of <50% predicted and/or a room air Pao₂ < 60 mm Hg or Paco₂ > 50 mm Hg.⁸ Index cardiac operations, as defined by the STS, included isolated CABG, isolated AVR, AVR + CABG, mitral valve replacement, mitral valve repair, mitral valve replacement + CABG, and mitral valve repair + CABG.⁹ Predicted risks of mortality and complications were obtained from the STS Risk Calculator.¹⁰

Pulmonary function test results for FEV₁ and the diffusion capacity of the lungs for carbon monoxide are represented as the percentage of the predicted value based on height, weight, and race. There were no strict pulmonary criteria that excluded operation; the decision to offer an operation was at the discretion of the surgeon. Additionally we investigated patients undergoing index cardiac operations during the study period with no history of chronic lung disease.

Statistical Analysis

Continuous data are presented as mean + SD for parametric data and median (interquartile range [IQR]) for nonparametric data. Categorical data are presented as number (percentage). One-way analysis of variance was used to compare categorical data for tracheostomy placement. A multivariable Cox regression model for mortality at 30 days and 1 year was constructed by using baseline characteristics displayed in Table 1. Backward selection was used to derive the final model that included variables with a P < .10; the tables display all variables with a P < .05. The supremum test was used to assess the proportional hazards assumption. Kaplan-Meier analysis was used to model survival up to 5 years and freedom from tracheostomy up to 1 year. The Nelson-Aalen estimation was used to model the cumulative hazards for readmission up to 5 years. Propensity score matching was used to compare patients with severe chronic lung disease and those with no chronic lung disease in a greedy nearest-neighbor matching approach with a caliper distance of 0.2. A standardized mean difference of <0.01 was considered balanced matching. This study was approved by the Institutional Review Board at the University of Pittsburgh.

Table 1.

Baseline Characteristics of Patients With Severe Chronic Lung Disease Undergoing Index Cardiac Operations (N = 354)

Characteristic	Value
Age, y	69 (61–74)
Female sex	116 (32.77)
Race
White	326 (92.09)
Black	23 (6.50)
Body mass index, kg/m2	28.9 (24.1–34.0)
Diabetes mellitus	179 (50.56)
Hypertension	314 (88.70)
Dyslipidemia	285 (80.51)
Dialysis	20 (5.65)
Current smoker	151 (42.66)
Immunosuppression	41 (11.58)
Peripheral arterial disease	134 (37.85)
Cerebrovascular disease	126 (35.59)
Family history of coronary artery disease	74 (20.90)
Previous heart failure	123 (34.75)
Previous myocardial infarction	201 (56.78)
Indication for cardiac presentation
Stable angina	22 (6.21)
Unstable angina	80 (22.60)
Non- ST elevation myocardial infarction	93 (26.27)
ST elevation myocardial infarction	16 (4.52)
Arrhythmia	111 (31.36)
No. of diseased vessels	56 (15.86)
0	37 (10.48)
1	66 (18.70)
2	194 (54.96)
3
Preoperative intraaortic balloon pump	14 (3.95)
Positive stress test	35 (9.89)
Status
Elective	136 (38.42)
Urgent	202 (57.06)
Emergent	15 (4.24)
Emergent salvage	1 (0.28)
Surgery type
Isolated CABG	202 (57.06)
Isolated AVR	69 (19.49)
Isolated mitral valve replacement	17 (4.80)
Isolated mitral valve repair	7 (1.98)
CABG and AVR	40 (11.30)
CABG and mitral valve replacement	9 (2.54)
CABG and mitral valve repair	10 (2.82)
Bilateral internal mammary artery utilization	21 (5.93)
Endocarditis	10 (2.82)
Serum creatinine, mg/dL	1.00 (0.80–1.30)
Albumin, mg/dL	3.50 (3.10–3.80)
Total bilirubin, mg/dL	0.60 (0.40–0.80)
Left ventricular ejection fraction, %	53 (40–58)
Previous valve procedure	10 (2.82)
Previous CABG	21 (5.93)
Previous percutaneous coronary intervention	90 (25.42)
Cardiopulmonary bypass time, min	101 (80–136)
Ischemic time, min	74 (56–98)
Pulmonary artery systolic pressure, mm Hg	42 (34–55)
Home oxygen	114 (34.65)
Forced expiratory volume in 1 second, %	48 (41–56)
Diffusion capacity of the lungs for carbon monoxide, % predicted	78 (55–101)
Society of Thoracic Surgeons risk prediction
Predicted risk of mortality, %	4.87 (2.64–9.31)
Predicted risk of deep sternal wound infection, %	0.74 (0.47–1.30)
Predicted risk of reoperation, %	9.39 (6.40–13.72)
Predicted risk of permanent stroke, %	1.65 (1.07–2.68)
Predicted risk of prolonged ventilation, %	20.95 (14.28–37.64)
Predicted risk of renal failure, %	5.83 (2.90–11.57)
Predicted risk of morbidity or mortality, %	27.99 (19.37–44.87)
Predicted risk of short length of stay, %	19.31 (9.45–31.29)
Predicted risk of long length of stay, %	14.46 (8.99–27.09)

Values are median (interquartile range) or n (%).

CABG, coronary artery bypass grafting.

Results

Baseline Characteristics

Three hundred fifty-four patients were identified, of whom 116 (32.77%) were women, and the median age was 69 years (IQR, 61–74) (Table 1). Both hypertension (314, 88.70%) and dyslipidemia (285, 80.51%) were common in the patient population. One hundred ninety-four patients (54.96%) had 3-vessel coronary artery disease. Most cases (202, 57.06%) were classified as urgent, with 136 (38.42%) classified as elective. The most common case types included isolated CABG (57.06%), isolated AVR (19.49%), and CABG + AVR (11.30%). The predicted risk of mortality from the STS calculator was 4.87% and the predicted risk of prolonged ventilation was 20.95%.

Current smokers (n = 151) comprised 42.66% of the population. One hundred fourteen patients (34.65%) required home oxygen preoperatively. The median preoperative FEV₁ was 48% (IQR, 41%−56%), and the median diffusion capacity of the lungs for carbon monoxide was 78% (IQR, 55%−101%).

Survival Outcomes

Median follow-up time was 3.63 years (IQR, 1.82–5.88). Multivariable Cox regression analysis for 30-day mortality demonstrated that age (hazard ratio [HR], 1.12; P = .002), prior myocardial infarction (HR, 4.58; P = .019), and preoperative home oxygen use (HR, 2.91; P = .030) were independently associated with death (Table 2). Age (HR, 1.07; P < .001), diabetes (HR, 2.00; P = .022), dialysis (HR, 5.29; P < .001), cerebrovascular disease (HR, 2.44; P = .003), and preoperative home oxygen use (HR, 2.12; P = .009) were independent predictors of 1-year mortality. Survival at 30 days was 93.79%. Overall survival at 1 year and 5 years was 83.62% and 58.34%, respectively (Figure 1). The degree of FEV₁ or diffusion capacity of the lungs for carbon monoxide decrease was not significantly associated with reduced 1-year survival. Among 74 deaths 33.78% were attributable to cardiac causes, 9.46% to respiratory causes, 6.76% each to neurologic and infectious etiologies, and the remainder were classified as other/unknown.

Table 2.

Multivariable Cox Regression Analysis for 30-Day and 1-Year Mortality in Patients With Severe Chronic Lung Disease Undergoing Index Cardiac Operations

Variable	Hazard Ratio	95% Confidence Interval	P
30-day mortality			.002
Age (for each year)	1.12	1.04–1.20	.019
Prior myocardial infarction	4.58	1.29–16.23
Home oxygen	2.91	1.11–7.64	.030
1-year mortality
Age (for each year)	1.07	1.03–1.11	<.001
Diabetes	2.00	1.11–3.62	.022
Hypertension	0.39	0.19–0.82	.013
Dialysis dependence	5.29	2.12–13.23	<.001
Cerebrovascular disease	2.44	1.37–4.36	.003
Home oxygen	2.12	1.20–3.73	.009

Figure 1.

Kaplan-Meier analysis of 5-year survival in patients with severe chronic lung disease undergoing cardiac operations.

Respiratory Outcomes and Readmission

Thirty-three patients (9.07%) required a tracheostomy during the study period (Table 3). Median time from the index operation to the date of tracheostomy placement was 10 days (IQR, 7–13). One-year freedom from tracheostomy was 90.34% (Figure 2). Among the different procedure types, patients undergoing a CABG with a concomitant valve procedure (12/59, 20.34%) were significantly more likely to require a tracheostomy than those undergoing an isolated CABG (15/202, 7.43%) or an isolated valve procedure (6/93, 6.45%) (P = .006). The median duration of tracheostomy before decannulation was 49 days (IQR, 25–114).

Table 3.

Respiratory Outcomes Among Patients With Severe Chronic Lung Disease Undergoing Index Cardiac Operations (N = 354)

Outcome	Value
Need for tracheostomy	33 (9.07)
Time from index operation to tracheostomy, days	10 (7–13)
Duration of tracheotomy before decannulation, days	49 (25–114)
Discharge with tracheostomy	27 (7.63)

Values are median (interquartile range) or n (%).

Figure 2.

Kaplan-Meier analysis of freedom from tracheostomy up to 1 year in patients with severe chronic lung disease undergoing cardiac operations.

Of the 33 patients who underwent tracheostomy, 27 patients were discharged with the tracheostomy (ie, were not decannulated before discharge). The 30-day readmission rate for all patients was 20.34%. One-year and 5-year freedom from readmission was 51.83% and 31.82%, respectively (Figure 3). Severe chronic lung disease was not independently associated with readmission (HR, 1.20; P = .079) when compared with no chronic lung disease (Supplemental Table 1).

Figure 3.

Nelson-Aalen cumulative hazards for readmission in patients with severe chronic lung disease undergoing cardiac operations.

Comparison of Patients Without Chronic Lung Disease With Those With Severe Chronic Lung Disease

In a comparison of patients with no chronic lung disease and those with severe chronic lung disease, those without lung disease had fewer comorbidities, although there were no differences in age (Supplemental Table 2). Propensity score matching resulted in a well-matched cohort (Supplemental Table 3). One-year, 5-year, and overall mortality were significantly higher in matched patients with severe chronic lung disease (Supplemental Table 4). At 1 year the overall survival was 91.04% in patients with no chronic lung disease and 83.43% in those with severe chronic lung disease (Supplemental Figure 1). Readmission rates were also higher in those with severe chronic lung disease. Rates of major adverse cardiac and cerebrovascular events were significantly higher in patients with severe chronic lung disease (58.00%) as compared with those without chronic lung disease (39.86%, P < .001). Patients with severe chronic lung disease were also more likely to require a tracheostomy (9.43% vs 2.66%, P < .001).

Comment

Chronic lung disease is the fourth leading cause of death in the United States.¹¹ Because smoking contributes to the risk of both lung disease and atherosclerotic heart disease, chronic lung disease is an important consideration among patients undergoing open heart surgery. In this retrospective, single-institution study of 354 patients with severe chronic lung disease undergoing index cardiac operations, the major findings were that most patients (58%) were alive at 5 years, 9% required a tracheostomy in the postoperative period, and preoperative home oxygen use was a predictor of increased 30-day and 1-year mortality in this higher-risk patient subset. Collectively these data suggest that although severe chronic lung disease is a known risk factor for mortality and morbidity after index cardiac operations, its presence alone should not deter surgery in otherwise reasonable surgical candidates.

Prior studies have suggested reduced postoperative survival in patients with underlying respiratory diseases who undergo cardiac operations. Our 1-year survival of 83.62% is lower than the 88.7% 1-year survival reported by Angouras and colleagues⁴ in patients with COPD undergoing CABG and contrasts strikingly with their reported 94.0% survival in patients without COPD. This distinction in mortality rates is likely related to our inclusion of patients undergoing valve and combined operations as well as our focus on only those patients classified as having severe lung disease. Few studies have specifically examined the severity of lung disease in relation to postoperative outcomes; however even mild or moderate COPD has been linked to reduced 30-day survival after cardiac surgery.³ COPD has also been associated with all-cause mortality and cardiac death in all patients with ischemic heart disease, and thus surgical revascularization may potentially mitigate this excess mortality.¹² Improved understanding of specific contributors to death in patients with both chronic lung disease and indications for cardiac surgery will inform individualized risk assessment and help guide surgeons and patients in decision-making when considering the relative risks and benefits of surgery.

The relative role of surgical (including open and minimally invasive approaches) as opposed to percutaneous interventions has received some attention in patients with lung disease. Cardiopulmonary bypass and associated proinflammatory factors may induce lung injury that may be poorly tolerated in patients with preexisting pulmonary dysfunction.¹³ With this in mind several groups have explored outcomes in on-pump versus off-pump CABG. A subgroup analysis of patients with COPD in the Randomized On-Off Bypass (ROOBY) trial, however, demonstrated no significant differences in 30-day or 1-year composite outcomes in patients undergoing on- or off-pump CABG.¹⁴ In contrast in a randomized study of 58 patients with COPD undergoing CABG, off-pump CABG was associated with a reduced intensive care unit length of stay and a reduced time to extubation (4.5 vs 8.65 hours).¹⁵ Interestingly patients undergoing minimally invasive CABG displayed the lowest time to extubation (2.13 hours) and spent the fewest days in the intensive care unit (1.1 days) as compared with those undergoing off-pump CABG (1.4 days) or on-pump CABG (2.6 days).¹⁵ This study was limited to patients with single-vessel disease and thus results may not apply in patients with multivessel disease; however this work highlights the importance of both cardiopulmonary bypass and incision type on postoperative respiratory function. Further study in a larger patient population is warranted to explore the relative efficacy of these different approaches.

Similarly poor respiratory reserve may potentially advocate for the use of transcatheter valve interventions, including transcatheter AVR, in select eligible patients.¹⁶ A retrospective analysis of data from patients with chronic lung disease in the PARTNER (Placement of Aortic Transcatheter Valves) trial, however, demonstrated no significant differences in 30-day, 1-year, or 2-year mortality between high-risk patients undergoing transcatheter AVR versus surgical AVR.¹⁷ Preoperative oxygen dependence was similarly found to be an independent prognosticator of poor postprocedural outcomes.¹⁷ It remains to be seen whether a subset of chronic lung disease patients may derive benefit from a transcatheter over a surgical approach.

In addition to intraoperative factors, modifiable baseline characteristics may represent a potential avenue to improve outcomes in patients with severe chronic lung disease. As an example we found that over 40% of our population were current smokers. In a retrospective review of nonsmokers, former smokers, and current smokers undergoing CABG, current smoking was significantly associated with pulmonary complications.¹⁸ These authors suggested that delay of surgery to allow for a minimum of 4 weeks of abstinence from tobacco may improve outcomes.¹⁸ A prospective study exploring the relative postoperative morbidity and mortality in patients who delay elective surgery to allow for a short interval of smoking cessation would be particularly informative. Preoperative adjuncts including bronchodilator therapy, a short course of steroids, and pulmonary rehabilitation have also been suggested to reduce the rates of pulmonary complications.^19,20 Ultimately a combination of preoperative and perioperative strategies may allow for additive benefits to reduce postoperative pulmonary dysfunction and potentially improve long-term survival.

We believe it is critically important to include a discussion of the need for tracheostomy in this patient population as part of the informed consent discussion. In our population of patients with severe chronic lung disease, 9% required a tracheostomy, and most patients were discharged with the tracheostomy. Tracheostomy has been linked to poor quality of life and may alter the discharge disposition, both of which are key patient-centered outcomes that may assist patients and families in anticipating the postoperative course.^21,22 It is also important to consider the association between tracheostomy and the rate of sternal wound infection, which was quoted as 7% in a meta-analysis.²³ Thus interventions to avoid the need for prolonged mechanical ventilation and tracheostomy placement may result in improved patient satisfaction in addition to potential decreases in infectious complications, prolonged length of stay, and readmission.²⁴ Conversations with patients and their families regarding the potential for tracheostomy and what the procedure entails are especially vital in the preoperative setting because they avoid the potential issue of families refusing tracheostomy because they may feel overwhelmed by their loved one’s suboptimal recovery after surgery. These expectations and potential realities should therefore be discussed upfront before surgery.

This study has several limitations. Because we explored only patients undergoing cardiac operations, the study population represents only those patients deemed fit for surgery, and thus the results may not be generalizable to potential outcomes among all patients with severe chronic lung disease who present with indications for surgery. Because of the retrospective nature of the study we were unable to account for provider decision-making to proceed with surgical intervention, including the influence of multidisciplinary discussions. Additionally we were unable to determine the cause of death; thus it is unclear whether poor baseline respiratory function is the primary contributor to long-term survival. We also did not have granular data in our database to calculate rates of atelectasis, pulmonary edema, pneumothorax, or acute respiratory distress syndrome. Finally we do not have granular data to assess respiratory status immediately before the operation. Thus we were not able to account for the influence of preoperative respiratory decompensation or infection in these patients with underlying lung disease.

In conclusion we report long-term outcomes among patients with severe chronic lung disease undergoing index cardiac operations. We found that patients who required home oxygen displayed more than double the short-term mortality compared with those not requiring oxygen. Although severe chronic lung disease is a known risk factor for operative mortality and morbidity, the findings that only 9% required tracheostomy and most patients were alive at 5 years suggest that severe chronic lung disease alone should not deter index cardiac operations in otherwise surgically fit patients. Finally given the high prevalence of chronic lung disease among adults in the United States, this work calls for a directed inquiry into specific preoperative and perioperative strategies to improve outcomes among patients with compromised pulmonary function who are undergoing open heart surgery.