A 68-Year-Old Man With COPD Contemplating Colon Cancer Surgery

DR REYNOLDS: Mr A is a 68-year-old man with severe chronic obstructive pulmonary disease (COPD), coronary artery disease, peripheral vascular disease, hypertension, hypercholesterolemia, and ongoing tobacco use who needs surgery for presumed colon cancer.

In December 2005, Mr A underwent a colonoscopy for a single episode of melena. The colonoscopy revealed a malignant-appearing, friable infiltrative sigmoid mass. A previous routine colonoscopy performed in 2001 had revealed only 2 small polyps: a right-sided small adenoma and a left-sided small hyperplastic polyp. Since the recent colonoscopy, Mr A has seen 2 general surgeons who have both recommended that he proceed with resection of the mass.

Mr A smokes 1 to 2 packs per day, although he has been trying to cut down; he has a 75-pack-year history of smoking. He produces copious gray sputum in the morning; his chronic cough has not changed in the past several months. He has tried to quit smoking with the aid of a nicotine patch, but his periods of abstinence have been short. A trial of bupropion was limited by dysphoria. His recent attempts to quit have been limited in part by his increased stress level around the new diagnosis of near-certain cancer. He has numerous recent episodes of “bronchitis” and upper respiratory tract infection (URI) for which he has been treated with short courses of oral steroids and antibiotics. He has infrequent episodes of angina in a stable pattern. His exercise is limited by his peripheral vascular disease. He can walk 50 to 100 yards before his calves ache and he must stop to rest.

Mr A's past medical history is significant for COPD with numerous hospitalizations but no intubations. Magnetic resonance imaging has shown evidence of lacunar strokes. Mr A had a right femoral bypass graft placed in 1990, coronary artery bypass grafting in 1999, and a subsequent stent placement in 2001. He had no pulmonary complications after his cardiac surgery. A 2004 echocardiogram showed mild biventricular hypertrophy with preserved ventricular function and no wall motion abnormalities. Mr A has no history of congestive heart failure. A dipyridamole MIBI stress test performed in 2001 (after his stent placement) revealed mild, reversible defects in the inferior wall and inferior portion of the apex and anterior wall. Having been clinically stable since 2001 without change in his infrequent angina, Mr A visited his cardiologist before the planned surgery. Mr A declined the option of preoperative cardiac catheterization with possible stent placement. Stents would have delayed his surgery for at least 3 months because of the need for antiplatelet medications to prevent stent thrombosis.

Mr A's current medications include a fluticasone/salmeterol combination inhaler, an albuterol/ipratropium combination nebulizer, a tiotropium inhaler, amlodipine, lisinopril, atorvastatin, citalopram, isosorbide mononitrate, clonazepam, pantoprazole, and aspirin.

Mr A has commercial insurance. He lives with his wife, has a full-time job, and drinks several drinks daily.

On physical examination, Mr A has mild respiratory distress when moving from the waiting room to the examination room. At rest, he has no distress. He weighs 193 lb (87.7 kg); his body mass index is 27.5; blood pressure is 101/57 mm Hg, heart rate is 70/min, and oxygen saturation is 96% on room air. The lungs sound hyperinflated and have decreased breath sounds with scattered wheezes. There are no crackles. The neck is supple without jugular venous distension. The heart sounds are normal. The extremities show no clubbing, cyanosis, or edema.

Mr A's laboratory results are as follows: hematocrit, 33% with indices suggestive of anemia of chronic disease; platelet count, 157 000/μL; white blood cell count, 7900/μL; normal electrolyte levels; serum urea nitrogen, 15 mg/dL (5.35 mmol/L); serum creatinine, 1.1 mg/dL (97.24 μmol/L); serum albumin, 4.1 g/dL; and carcinoembryonic antigen, 11 ng/mL (normal values 0-4). A chest x-ray shows hyperinflation with flattening of the diaphragms. There are mediastinal clips and sternotomy sutures.

Mr A's spirometry results fluctuated minimally without any significant trends on the 4 intervening measurements among those shown in Table 1.

Mr A was very anxious about his upcoming surgery. Believing his anxiety would worsen if he was interviewed on camera, Mr A gave the editors of Clinical Crossroads permission to use his case history and asked us to speak to his doctor, whom he had known for years. Dr N had spoken extensively to Mr A and his wife before we interviewed him.

I met Mr A 8 or 9 years ago, when I was already caring for his wife and for his brother-in-law. Mr A's biggest medical problems have been his cardiac and pulmonary issues. In the early 1990s, Mr A had several admissions with chest pain and angina. He had angioplasties and then he had a 4-vessel coronary artery bypass graft. He had 1 subsequent PTCA [percutaneous transluminal coronary angioplasty] and has done well from a cardiac point of view for the last 4 or 5 years. So it took a few years for us to revascularize his coronary circulation. But we were finally able to do it.

Mr A has had some severe exacerbations of COPD. So he’ll occasionally get a URI. He hasn't needed to be intubated, but he’ll be hospitalized and need intensive treatment, either in or out of the ICU [intensive care unit]. One of the disappointing things is that he has not been able to stop smoking cigarettes. He would like to, but he still smokes a couple of packs a day.

Mr A has a sedentary lifestyle, so his lung disease doesn't limit him. He walks. He works as a security guard and he doesn't have to restrict his work activities or his usual household activities. He paces himself and avoids exercise.

I think anxiety is a big issue for him. He doesn't come across as an overly anxious person. But he reports the experience of anxiety despite antianxiety medications and various antidepressants over the years. I think he finds the cigarettes very comforting, and to some extent, he's turned to alcohol from time to time at periods of increased stress.

It's been very stressful for Mr A to have this diagnosis of presumed colon cancer. He has expressed to his wife many times that he doesn't think he's going to make it. It's reminiscent of other times when he has had a bad feeling about his health. He is concerned about not coming off the ventilator. But unless he had an unforeseen complication, his underlying lung disease wouldn't make us think that's a realistic fear.

He saw his cardiologist about a month or two ago and he had a repeat stress test. That was good. He was reassured by that. He saw his pulmonologist and had repeat pulmonary function tests. Those were okay. He has his long-standing abnormalities. But they were no worse. He was asked at that time to stop smoking and advised that if he stopped smoking for as short as 1 month, his perioperative complications would likely be less. He apparently did stop for a week or two, but subsequently he started smoking again.

I have spoken to Mr A's wife a few times in the last month and she's been concerned about how he's been dealing with anxiety, both that he tells her how worried he is and that when he comes home from work, his first inclination is to have a couple of drinks and then take a nap.

How do you assess preoperative pulmonary risk in a patient with COPD and what are the important patient- and procedure-related risk factors? What preoperative evaluation should be performed and by whom? Do spirometry or pulmonary risk indexes help to stratify risk? When should an alternative low-risk procedure be considered and is pulmonary status ever an absolute contraindication to surgery? What strategies have been shown to reduce the risk of postoperative pulmonary complications? What is the role for stopping or reducing smoking? What do you recommend for Mr A?

DR SMETANA: Mr A has severe COPD that causes symptoms most days and restricts his lifestyle to sedentary activities. He is contemplating surgery for a potentially life-threatening condition but has concerns about the impact of his COPD on the risk of the procedure. In particular, he is concerned about the potential that he may require long-term mechanical ventilation after surgery and may fail to wean from the ventilator. He is a long-term smoker and worries that this will increase the risk of his proposed surgery.

Important postoperative pulmonary complications (PPCs) include pneumonia, lobar atelectasis, mechanical ventilation longer than 48 hours (respiratory failure), and bronchospasm.¹ Most studies of PPCs estimate risk for a composite outcome that includes several or all of these possible complications. With few exceptions, it is not possible to derive a risk estimate for individual PPCs. In moderate- to high-risk surgeries, pulmonary complications are equally morbid as cardiac complications and contribute equally to prolonged length of stay and mortality.² For example, in a prospective cohort study of 3970 unselected patients undergoing major nonurgent, noncardiac surgery, postoperative cardiac and pulmonary complications occurred with similar frequency; rates were 2.5% and 2.7%, respectively.³

A recent guideline from the American College of Physicians (ACP) provides summary estimates of the risk conferred by particular patient- and procedure-related factors and helps to estimate Mr A's risk of PPCs as he considers possible colon cancer surgery.^{1 ,4 - 5} The following discussion focuses primarily on preoperative pulmonary evaluation and strategies to reduce the likelihood of postoperative pulmonary complications including ventilator dependence. Venous thromboembolic risk is not discussed, as this is a separate topic with different risk factors.

The American Society of Anesthesiologists (ASA) physical status classification, which estimates perioperative mortality by use of a general clinical impression of burden of illness, also is one of the most important patient-related risks for PPCs^{1 ,6} (Table 2). According to this classification, Mr A would be ASA class 3 (a patient with severe systemic disease); this finding alone increases his risk of PPCs (odds ratio [OR], 4.87; 95% confidence interval [CI], 3.34-7.10) when compared with risk for ASA class 1 patients).¹

The impact of age on PPC risk has been controversial. Advancing age increases the prevalence of comorbidities that may themselves increase risk, but age remains a risk factor even after multivariate adjustment for such comorbidities.¹ No single deflection point exists; PPC risk increases with each decade beyond 50 years of age. Mr A is 68 years old; this confers an adjusted odds ratio (AOR) of 2.09 (95% CI, 1.65-2.64) compared with patients younger than 50 years. Considering Mr A's absolute risk, the unadjusted PPC rates for unselected patients age 65 years and older vs younger than 65 years are 11.3% and 1.5%, respectively.¹

A history of congestive heart failure, as established by clinical diagnosis, is a risk factor for PPC and is particularly important among elderly patients. For example, in a multivariable analysis of 544 patients older than 70 years undergoing noncardiac surgery, a history of heart failure was the strongest preoperative predictor of PPCs (OR, 5.7; 95% CI, 2.1-15.5).⁸ Two studies derived from the National Veterans Administration Surgical Quality Improvement Program each found functional dependence (inability to perform some or all activities of daily living) to be an independent risk factor.^{9 - 10} Cigarette use increases PPC rates, even after adjustment for the presence of COPD. The AOR is, however, modest at 1.26 (95% CI, 1.01-1.56). Among these 3 risk factors, only current cigarette smoking applies to Mr A.

Of greatest importance to Mr A is the impact of COPD on PPC rates. In a pooled analysis of 8 studies, the AOR for COPD was 1.79 (95% CI, 1.44-2.22), which is lower than previously estimated.¹ The unadjusted PPC rate for patients with COPD across all types of surgery is 18.2%.¹

How accurate is clinical evaluation in diagnosing and estimating the severity of COPD? Table 3 summarizes composite operating characteristics obtained from a 1995 review¹¹ and a more recent multicenter study.¹² Clinicians can correctly diagnose obstructive airways disease with a high degree of accuracy. The finding of all 4 of the values cited in the study by Straus and colleagues¹² confers a positive likelihood ratio of 220.5, and for the absence of all factors, a negative likelihood ratio of 0.13. Limited literature suggests that the greater the number of abnormal physical findings present, the more severe is the obstructive lung disease.¹³ Relevant to Mr A is the observation in 1 study that the presence of any of 5 abnormal physical findings, including decreased breath sounds, prolonged expiration, rales, wheezes, or rhonchi, increased PPC rates after elective abdominal surgery (OR, 5.8; 95% CI, 1.04-32.1).¹⁴ When he performed his preoperative physical examination, Dr N found that Mr A had 3 of the 5 physical findings. The literature does not allow a composite estimate of the OR for PPCs for patients with multiple risk factors.

Does the severity of COPD by clinical evaluation influence PPC rates? Kroenke and colleagues¹⁵ studied outcomes of 78 patients with COPD after thoracic or abdominal surgeries. The authors classified severity of COPD by spirometric criteria (severe = forced expiratory volume in the first second [FEV₁] <50% predicted and ratio of FEV₁ to forced vital capacity [FEV₁/FVC] <70%), but clinicians would likely have been able to identify these patients based on clinical grounds. Compared with patients with mild to moderate COPD, those with severe COPD were more likely to be taking long-term bronchodilators (54% vs 25%), to have abnormal results of lung examination (62% vs 37%), and to have an ASA class of 3 or higher (96% vs 85%). Mr A meets all 3 of these clinical criteria. Mortality (19% vs 4%) and severe PPCs (23% vs 10%) were more common among the severe COPD cohort. The number of pack-years of cigarette use also appears to correlate with PPC rates, but this is not a consistent finding.^{16 - 19}

In contrast to COPD, well-controlled asthma is not a risk factor for PPCs. In the largest study to address this question, only 12 of 706 patients with asthma had a PPC; all of the complications were minor and consisted of bronchospasm and laryngospasm.²⁰

Pulmonary hypertension, regardless of etiology, is an important risk factor for PPCs. In 1 study, the rate of respiratory failure for such patients undergoing noncardiac surgery with general anesthesia was 28%.²¹ Small incremental PPC risk also exists for patients with recent weight loss (AOR, 1.62; 95% CI, 1.17-2.26), impaired sensorium (AOR, 1.39; 95% CI, 1.08-1.79), and alcohol use (AOR, 1.21; 95% CI, 1.11-1.32).^{9 - 10}

In contrast to cardiac risk stratification, procedure-related risk factors contribute more to overall PPC risk than do patient-related risk factors.¹ The single most important risk factor for PPCs is surgical site.¹ The traditional teaching still applies: PPC risk is greatest for incisions near the diaphragm. The highest-risk procedures are aortic (AOR, 6.90; 95% CI, 2.74-17.36), thoracic (including esophageal surgery) (AOR, 4.24; 95% CI, 2.89-6.23), and upper abdominal (AOR, 2.91; 95% CI, 2.35-3.60). Neurosurgery (AOR, 2.53; 95% CI, 1.84-3.47) and head and neck procedures (AOR, 2.21; 95% CI, 1.82-2.86) are also intrinsically high-risk procedures.¹ Mr A is planning abdominal surgery, which confers an OR of 3.01 (95% CI, 2.43-3.72) for PPC rates among unselected patients.¹ The unadjusted PPC rate for Mr A, based on whether he undergoes upper or lower abdominal surgery, would be 19.7% or 7.7%.¹ Postoperative pulmonary complication rates specific to colorectal surgery range from 5% to 14%.^{22 - 24}

Other important procedure-related risks are emergency surgery (AOR, 2.52; 95% CI, 1.69-3.75) and prolonged surgery (usually defined as >3 hours) (AOR, 2.26; 95% CI, 1.47-3.47). The effect of anesthetic type on PPC rates remains controversial. In the largest meta-analysis of randomized trials to date, authors identified 141 eligible trials of intraoperative neuraxial blockade (spinal or epidural anesthesia) vs general anesthesia that included 9559 patients.²⁵ Pneumonia (OR, 0.61; 95% CI, 0.48-0.76) and respiratory failure (OR, 0.41; 95% CI, 0.23-0.73) were both significantly less common among patients randomly assigned to receive neuraxial blockade (with or without concomitant general anesthesia) than among those assigned to receive general anesthesia alone. Potential sources of bias in this meta-analysis are the inclusion of studies with small sample sizes, study heterogeneity, and older trials. Several recent randomized trials have not confirmed lower PPC rates among patients receiving neuraxial blockade.^{26 - 27} While the data are mixed, clinicians should consider general anesthesia to be a probable risk factor for PPCs, pending future study.

In addition to Mr A's risk for PPCs, he and his physicians must consider his risk for postoperative cardiac complications including myocardial infarction, pulmonary edema, and cardiac death. The revised cardiac risk index is a validated tool that uses 6 clinical factors to estimate risk.²⁸ In this index, each of the following factors is significant after adjustment for potential confounders: (1) high-risk surgery, (2) ischemic heart disease, (3) history of congestive heart failure, (4) history of cerebrovascular disease, (5) insulin therapy for diabetes, and (6) serum creatinine level higher than 2.0 mg/dL (176.8 μmol/L). Mr A has 3 of the 6 factors (1, 2, and 4), which puts him into class IV; this class carries an estimated risk of postoperative cardiac complications of 9.1% to 11.0%. This is comparable to his risk of PPCs.

Based on the American College of Cardiology/American Heart Association guideline,²⁹ Mr A has the following: (1) coronary revascularization within 5 years with recurrent symptoms; (2) a recent coronary evaluation that showed ischemia; (3) intermediate clinical predictors (mild angina); and (4) poor functional capacity (50-100 yards). The guideline thus recommends noninvasive testing for Mr A. However, the value of this approach has been challenged by the recent finding that coronary revascularization does not reduce the risk of postoperative cardiac complications among patients with stable coronary artery disease.³⁰ An updated version of this guideline is expected this year and will likely differ substantially from the existing guideline regarding the role of preoperative cardiac testing. Further discussion of preoperative cardiac evaluation is beyond the scope of this review.

Spirometry. Spirometry has a well-established role in the evaluation of patients with lung cancer who are potential candidates for resection,^{31 - 32} but its role in other types of high-risk surgery has been debated. Four studies have used multivariable analysis to adjust for clinical features known to increase PPC risk.^{14 ,16 ,33 - 34} In 3 of these studies, abnormal spirometry predicted higher PPC rates; 2 of these studies reported that clinical predictors conferred higher ORs than abnormal spirometry. In the fourth study, results of spirometry did not differ between patients with and without PPC, whereas clinical predictors in each significantly correlated with PPC rates. The value of spirometry before surgeries other than lung resection remains unproven, and the test is unlikely to add incremental information beyond the clinical evaluation.

While spirometric values are more likely to be abnormal among patients who sustain a PPC than those who have an uncomplicated postoperative course, clinicians can likely identify these patients based on history and physical examination. Mr A's example supports this premise. While he has an FEV₁ of less than 50% predicted, suggesting severe COPD, the severity of his COPD is apparent from his history and physical examination; his spirometry results do not modify the clinical impression of severity.

Chest Radiography. Potential rationales for preoperative chest radiographs include providing a baseline in the event of need for a postoperative film, detecting clinically inapparent cardiopulmonary disease that may modify anesthetic or surgical technique or influence PPC risk estimates, and stratifying risk among patients with known COPD. Limited evidence from a single study supports the value of obtaining a preoperative study solely to serve as a baseline.³⁵ Most of the literature regarding screening chest radiographs has evaluated the likelihood that a study would change preoperative management rather than stratify PPC rates. In a review of preoperative screening tests, pooled data from 18 studies (n = 20 518) indicated that 21.2% of preoperative chest radiographs were abnormal.³⁶ However, among patients younger than 50 years, only 4.9% of studies were abnormal. In 3.0% of all patients, the results of the chest radiograph changed preoperative management. The authors of a recent systematic review report similar conclusions.³⁷ As to the third potential rationale, 2 studies have reported that an abnormal chest radiograph increases the odds of a PPC even after adjustment for clinical variables,^{14 ,38} but this conclusion is based on limited evidence. I recommend preoperative chest radiography for patients older than 50 years with known cardiopulmonary disease undergoing high-risk surgery.

Blood Tests. In contrast to the limited support for spirometry and chest radiographs, 3 blood tests are significant predictors of PPC risk in multivariable analysis: serum urea nitrogen level higher than 21 mg/dL (7.49 mmol/L), serum creatinine level higher than 1.5 mg/dL (132.6 μmol/L),^{9 - 10 ,39} and a low serum albumin level.¹ A low serum albumin level is the strongest laboratory predictor for PPCs and confers a 2- to 4-fold increase in PPC rates; the reported threshold ranges from 3.0 to 3.9 g/dL.

Risk indexes have been widely used for preoperative cardiac risk stratification for nearly 3 decades but only recently have investigators developed validated pulmonary risk indexes based on large patient samples. Arozullah and colleagues substantially advanced the preoperative pulmonary risk assessment literature with the publication of 2 multifactorial indexes. They separately considered risk factors for pneumonia¹⁰ and respiratory failure.⁹ Using multivariable analysis, the authors assigned a weighted point system to those factors that independently predicted the adverse pulmonary outcome. The principal finding was similar to earlier literature: procedure-related risk factors dominate the analysis, in particular, surgical site. Most risk factors were not modifiable (for example, surgical site, age, functional status). These indexes have yet to be validated independently by other investigators.

Mr A is particularly concerned about failure to wean and the need for long-term mechanical ventilation. Applying the index for respiratory failure,⁹ Mr A receives points for upper abdominal surgery, COPD, and age 60 to 69 years. Therefore, he falls into class 3 (of 5 classes) with an estimated risk of respiratory failure of 5.3%. His actual risk may be greater as the risk index does not allow for adjustment based on the severity of COPD. According to the pneumonia index,¹⁰ Mr A receives points for upper abdominal surgery, COPD, age 60 to 69 years, current smoker, steroid use, and general anesthesia. He is class 3 (of 5 classes) and has an estimated risk of pneumonia of 4.0%.

More recently, the ACP published a guideline based on a systematic review of the evidence that includes 6 principal recommendations.⁵ The guideline recommends that all patients undergoing noncardiothoracic surgery be evaluated for COPD, congestive heart failure, ASA class of 2 or higher, age older than 60 years, and functional dependency because these increase risk of PPCs. Patients undergoing procedures associated with a higher risk of PPCs, including prolonged surgery (>3 hours), abdominal surgery, thoracic surgery, neurosurgery, head and neck surgery, vascular surgery, aortic aneurysm repair, or emergency surgery, or who receive general anesthesia, should be evaluated for other PPC risk factors and receive the recommended interventions to reduce PPCs. Because low serum albumin level (<3.5 mg/dL) is associated with increased risk for PPCs, serum albumin should be measured in patients suspected of having low serum albumin levels (<3.5 g/dL) as well as considered in patients with risk factors for PPCs. Patients at higher risk for PPCs should receive deep breathing exercises or incentive spirometry postoperatively and selective (rather than routine) use of a nasogastric tube (as needed for postoperative nausea or vomiting, inability to tolerate oral intake, or symptomatic abdominal distension). Spirometry and chest x-ray should not be used routinely for predicting risk of PPC but may be appropriate for patients with COPD or asthma. Finally, right heart catheterization, total parenteral nutrition, or total enteral nutrition should not be used with the intent of reducing PPCs.⁴⁰

If the indication for surgery is sufficiently compelling, such as to treat an imminently life-threatening condition or improve long-term survival (such as Mr A's potentially resectable neoplasm), even high-risk patients with COPD can undergo surgery with a degree of morbidity and mortality that may be acceptable to the patient. In a classic study, Kroenke and colleagues determined operative risk among 107 patients with severe COPD.¹⁹ In this study, all patients had an FEV₁ of less than 50% predicted, and an FEV₁/FVC ratio of less than 70%, like Mr A. Clinical lore at the time of this study suggested that risk for such patients was prohibitive. Six patients died and 7 had severe PPCs.

Similarly, Wong and colleagues studied PPC rates among 105 patients with severe COPD (defined as FEV₁ <1.2 L and FEV₁/FVC ratio <75%) undergoing noncardiothoracic surgery.¹⁶ Seven patients died and 39 had at least 1 PPC (the most common being refractory bronchospasm and prolonged intensive care unit stay). There were 7 cases of pneumonia and 8 cases of prolonged mechanical ventilation. Considering these 2 studies, while the risks are nontrivial, they may be acceptable when the value of the planned surgery is great.

Table 4 summarizes potential interventions to reduce PPC rates. The quality of evidence varies considerably. Good evidence from high-quality randomized controlled trials and systematic reviews supports postoperative lung expansion maneuvers.⁴ Evidence from fair-quality randomized controlled trials and systematic reviews supports short-acting neuromuscular blockers, selective postoperative nasogastric decompression, neuraxial blockade, epidural analgesia, and patient-controlled analgesia.⁴ The evidence for laparoscopic surgery and smoking cessation is mixed.

Postoperative lung expansion maneuvers are the only strategy that earned an “A” grade for strength of evidence in the recent ACP guideline.⁴ These maneuvers minimize the expected fall in lung volumes after upper abdominal and thoracic surgery and hence decrease risk for atelectasis and other PPCs. The strategies of deep breathing exercises, chest physical therapy, incentive spirometry, and continuous positive airway pressure each reduce risk by about one half.^{45 ,56} Combinations of 2 or more strategies are no more effective than 1 strategy alone. Lung expansion maneuvers are more effective when begun before surgery.^{57 - 58} Preoperative exercise training decreases dyspnea and improves functional capacity and quality of life for patients with COPD,⁵⁹ but whether it reduces PPC rates is unknown. Despite this lack of evidence, this strategy would be worthwhile for Mr A.

Current cigarette smoking increases the risk of PPCs. Contrary to intuition, several studies have reported that recent quitters (ie, <2 months before surgery) have a higher PPC risk than smokers who have not reduced or stopped smoking shortly before surgery,^{38 ,60 - 61} although a single study of patients undergoing lung resection reported no difference in PPC rates among patients stratified by duration of abstinence.⁶² A single randomized trial evaluated the benefit of a 6- to 8-week smoking cessation program to reduce perioperative complications among 120 patients before knee or hip replacement.⁴¹ The low rate of respiratory failure did not differ (1 patient in each group), but hip and knee surgeries are intrinsically low-risk procedures for PPCs and this study was not powered to detect a difference in PPC rates. If Mr A's surgery can be safely delayed for at least 2 months without a high likelihood of losing the window for curative resection, this would allow him to maximally benefit from a strategy of smoking cessation.

The use of pancuronium for neuromuscular blockade increases the likelihood of postoperative residual neuromuscular blockade. Among this subset of patients, there is a 3-fold increase in PPC rates.⁴² Intraoperative neuraxial blockade, as opposed to general anesthesia, may reduce PPC rates, but this remains controversial.^{4 ,25}

Postoperative pain can cause splinting and impair deep breathing, contributing to PPCs after high-risk chest, abdominal, and aortic surgeries. In a systematic review, postoperative epidural analgesia reduced PPC rates relative to systemic opioids for atelectasis (epidural opioids: OR, 0.53; 95% CI, 0.33-0.83), pulmonary infection (epidural local anesthetics: OR, 0.36; 95% CI, 0.21-0.65), and respiratory failure (epidural local anesthetics: OR, 0.58; 95% CI, 0.42-0.80).⁴⁸ Postoperative patient-controlled analgesia also reduces PPC rates when compared with on-demand analgesia.⁴⁹

The benefits to patients of laparoscopic vs open abdominal surgery are well known and include less postoperative pain, shorter hospital stays, and quicker return to normal activities.⁶³ A meta-analysis of 12 randomized controlled trials of laparoscopic vs open surgery for colon cancer identified no difference in PPC rates (OR, 0.65; 95% CI, 0.28-1.49) but was underpowered to identify significant differences.⁴³ Therefore, it is unknown if this strategy would reduce Mr A's PPC risk.

The presence of a nasogastric tube after surgery may increase PPC rates. Selective nasogastric decompression after abdominal surgery refers to its use for abdominal distension or nausea rather than as a routine practice. A systematic review of 19 randomized controlled trials (N = 2892) reported a trend toward fewer PPCs with selective use when compared with routine use (relative benefit increase, 1.35; 95% CI, 0.98-1.86).⁴⁷

Strategies that have proven ineffective include routine use of total parenteral nutrition, enhanced enteral nutrition (“immunonutrition”), and perioperative pulmonary artery catheter–guided therapy.^{4 ,44 ,53 - 54}

For Mr A, the risk of the procedure must be balanced against the risk of doing nothing (ie, the natural history of the untreated condition). The prognosis for Mr A's presumed colon cancer can only be established with certainty at the time of surgery as the extent of invasion and lymph node status remain unknown. However, if he is fortunate to have early-stage colon cancer, this carries a favorable prognosis when resected. The 5-year survival is 93% for stage I and 85% for stage IIA lesions.⁶⁴ Mr A's colon cancer, if left untreated, likely would cause symptoms within several years and ultimately lead to his death.

His PPC risk is increased by virtue of COPD, ASA class higher than 2, age, current cigarette use, upper abdominal surgery, and (probably) general anesthesia. Using a composite of the scores from Arozullah et al^{9 - 10} and estimates from the ACP systematic review,¹ his expected rate for all pulmonary complications is in the range of 10% to 20%. His mortality due to untreated colon cancer would be higher than his expected mortality from the planned surgery. I suggest that Mr A proceed to the planned surgery.

Mr A and his physician, Dr N, will need to weigh the risk of PPCs and the benefit that Mr A could gain if he were able to stop smoking for 2 months before surgery vs the likelihood that his colon cancer will progress over 2 months and no longer be potentially curable by resection. Whether Mr A's concern about possible PPCs outweighs his concern about having cancer is unclear. Regardless, he should continue to receive standard medical therapy for his COPD. If Dr N finds that his COPD is worse than usual as he approaches his scheduled surgery, Mr A could receive a course of oral corticosteroids (eg, prednisone 60 mg daily with a 10-14 day taper) in an attempt to reduce inflammation that may be worsening his airway obstruction. According to evidence in the literature, this treatment will not increase his risk of wound complications or respiratory infection.^{65 - 66} However, the recommendation for preoperative corticosteroids is based on expert opinion because no trial data specifically address their value. He should receive training in incentive spirometry or deep breathing exercises before surgery and begin one of these strategies as soon as possible after surgery.^{57 - 58} Exercise training before surgery may also be beneficial,⁵⁹ but should be part of a supervised rehabilitation program in light of his coronary artery disease.

Regarding the surgery itself, he should speak with his surgeon to determine whether laparoscopic surgery is technically possible, although the benefit of this strategy to reduce PPC rates is unknown. If this is not possible, a lower abdominal incision, rather than an incision that extends to the upper abdomen, would confer a lower PPC risk. He should receive postoperative epidural analgesia to improve postoperative pain and reduce PPC rates. He should receive a postoperative nasogastric tube only if needed for abdominal distension, nausea, or vomiting.

Mr A also has significant cardiac risk. A recent randomized controlled trial of preoperative coronary revascularization for patients with stable coronary artery disease reported no difference in cardiac outcomes after high-risk noncardiac surgery when compared with medical therapy.³⁰ While based mostly on retrospective data, several trials and a recent meta-analysis suggest a reduction in postoperative cardiac events among perioperative statin users.^{67 - 70} Alpha-2 agonists, including clonidine, also appear to reduce mortality and myocardial infarction after vascular surgery.⁷¹ It is possible, though unproven, that a similar benefit would exist for other types of high-risk surgery, such as colon cancer surgery. Finally, the value of perioperative β-blockers has been recently challenged by several large negative trials and a recent meta-analysis.^{72 - 74} A consensus is emerging that they be used primarily for high-risk patients undergoing vascular surgery.⁷⁵ Given Mr A's potential risk for bronchospasm and recent trial data, I suggest that Dr N consider perioperative statins and clonidine for Mr A.

QUESTION: You never once mentioned the word “blood gas.” And the other thing that made me a little nervous about steroids is that you say most of the data come from asthmatics. And they're a lot younger usually than 60- to 69-year-old people, and I wonder how well those data are generalizable to Mr A.

DR SMETANA: There had been some suggestion in the earlier literature that either hypercarbia, in particular, or hypoxemia might be relative or absolute contraindications to surgery. But more recently, that hasn't been shown to be the case. A few reports have looked at hypercapnia, for example, in high-risk procedures such as lung resection surgery, and found that it does not increase risk and is not an absolute contraindication.^{31 ,76}

In some studies, hypoxemia was an independent predictor of PPCs, but abnormal values did not necessarily represent an absolute contraindication to surgery.^{33 ,77} In general, blood gas abnormalities correlate fairly well with severity of COPD based on clinical assessment. They're not usually unexpected or surprising findings.

With regard to the steroid data, you're correct that most of the evidence is in asthmatic patients who tend to be younger and have fewer comorbidities. There may be some difficulty in extrapolating it to the COPD population. But the risk of having undertreated COPD and undergoing elective surgery with a COPD exacerbation would most likely be substantially higher than that of a short course of corticosteroids.

QUESTION: Could you comment on the functional assessment before operation, such as using a 6-minute walk test or a cardiopulmonary exercise study?

DR SMETANA: Much of the literature about functional testing exists in the lung resection surgery population. But there is some literature looking at patients undergoing nonresective surgery, such as high-risk abdominal surgery.

There appears to be a trend toward increased pulmonary complications among patients who report poor exercise capacity, but it's not significant in most of the literature.⁷⁸ In studies that have looked at cardiopulmonary stress testing to stratify risk, the results have generally suggested that patients who perform poorly on such tests have a higher risk of PPCs.^{79 - 80} It's likely that clinicians could identify patients who would perform poorly on cardiopulmonary stress testing based on clinical grounds alone, but the literature has not addressed this question.

QUESTION: This man has tried to quit smoking multiple times; do you think he should really delay surgery for the 2 months? Or maybe he should focus on a few weeks of the incentive spirometry and know that probably, given the anxiety of the surgery, this man's probably not going to stop smoking, and leave it at that. I mean, how would you balance that with his internist?

DR SMETANA: This is a place where the primary care physician could play an important role. Mr A's success may depend to some extent on how well informed he is about the risks associated with smoking and what benefit he might gain if he's able to stop smoking for a period of a couple of months. And maybe it's never been presented to him in such a fashion where he sees such a potential concrete benefit.

In the single randomized controlled trial of a smoking cessation intervention, the investigators were actually able to get over one half of the patients to stop smoking before surgery, including many long-term smokers.⁴¹ So in a situation such as this, where Mr A is faced with a high-risk scenario, he may have more motivation than we’d see in another setting.