Preoperative Risk Stratification of the Thoracic Surgical Patient

David J. Ficke
Jerome M. Klafta

Key Points

1. All patients being considered for lung resection should have pulmonary function tests including spirometry and a DLCO (diffusing capacity of the lung for carbon monoxide) test, from which the predicted postoperative values are calculated. If the results are unfavorable, a measure of exercise capacity or peak oxygen consumption should be obtained.

2. Cardiac evaluation of the thoracic surgical patient should include surgeons, anesthesiologists, and cardiologists. Higher levels of perioperative risk may be acceptable because of the potential curative benefit of surgery for non-small-cell lung cancer.

3. A thorough history of cancer therapy that considers chemotherapy, radiation, and an evaluation of the paraneoplastic effects of the cancer identifies other potential perioperative vulnerabilities.

Case Vignette

A 69-year-old man is scheduled for a left pneumonectomy. A CT-guided biopsy 6 days ago revealed adenocarcinoma. He is obese and has hypertension, type 2 diabetes mellitus, osteoarthritis, and a 55 pack-year smoking history. When a mass was seen on his chest x-ray 2 weeks ago, he quit smoking. Pulmonary function tests show a moderately obstructive ventilatory defect with an FEV₁ of 63% of predicted. He blames limited exercise tolerance on his “bad knees” and has never been evaluated by a cardiologist. How should this case be managed? Are there any other tests that would be helpful for stratifying his perioperative risk?

Thoracic surgery can have profound effects on cardiopulmonary function in the operating room, in the immediate postoperative setting, and in the long-term. The scope of thoracic surgery ranges from a thoracoscopic sympathectomy for a healthy 20-year-old patient to an extrapleural pneumonectomy for an 80-year-old with coronary disease and emphysema. Ever since the first pneumonectomy was described in 1933,¹ physicians have been looking for a simple, effective way to evaluate patients to optimize outcomes. This chapter focuses primarily on the preoperative evaluation of patients who need pulmonary resection, but the principles apply for other thoracic surgeries as well. Esophageal surgery, for example, does not involve resection of lung tissue, but because esophageal pathology is associated with smoking, patients frequently have concurrent pulmonary disease. Several other considerations are noteworthy for esophageal surgery including the frequent presence of reflux and aspiration, poor nutritional status, and preoperative chemotherapy or radiation.

Understanding the surgical approach is critical to preparing for thoracic surgery. For example, if a patient has had coronary bypass surgery with an internal mammary artery, he is at high risk for myocardial ischemia during an ipsilateral extrapleural pneumonectomy. The unique physiology and pathophysiology of pulmonary resection necessitates several other considerations.

Pulmonary resection is generally performed on patients with lung cancer, which accounts for 160,000 deaths per year in the United States.² Five-year survival—only 15% for all lung cancers—is 49% for patients with surgically resectable, localized disease. It is likely that surgery is responsible for most of the long-term survivors. Lung cancers double in size within 30 to 500 days,³ and faster growing tumors are associated with poorer prognosis.⁴ Because of the aggressive nature of lung cancer, many patients (and physicians) are willing to accept higher levels of risk than they might for other types of surgery. Preoperative evaluation, ideally, should not significantly delay a surgery that is potentially curative.

We next provide a framework for the evaluation of the thoracic surgery patient with focus on the respiratory and cardiovascular systems, the potential physiologic impact of other cancer therapies such as chemotherapy and radiation, as well as other unique considerations for thoracic surgery.

Minimally invasive techniques, particularly video-assisted thoracic surgery (VATS), have become increasingly popular in the past decade. A video-assisted thoracoscopic lobectomy may have fewer and less severe complications than the same procedure performed by a conventional thoracotomy,^5,6 but preoperative evaluation of patients should be similar for both open and minimally invasive procedures. The removal of lung parenchyma and the physiologic changes this brings to the cardiorespiratory systems are not significantly different with either surgical technique–lung tissue is still removed. In addition, with VATS there is always the potential for conversion to an open procedure. Therefore, no distinction is made in this chapter between minimally invasive or open techniques with regard to preoperative risk stratification.

EVALUATION OF RESPIRATORY FUNCTION

The majority of pulmonary resections for lung cancer are performed in patients with some degree of respiratory impairment. Historically, maximum voluntary ventilation (MVV) was used to determine fitness for pulmonary resection.⁷ MVV is defined as the maximal amount of air a patient can inhale and exhale in 12 seconds. In 1955, Gaensler found that patients with low MVV had a higher mortality, and subsequent studies have confirmed a correlation between MVV and perioperative mortality.^8,9 Because one of the limitations of MVV is that it depends entirely on effort, it has largely been replaced by other tests.

Spirometry

Spirometry has become the primary method by which patients are evaluated before thoracic surgery. Spirometry, which is relatively noninvasive, evaluates lung mechanics without the need for expensive equipment. The patient exhales air as fast as possible into a device that measures the pressure, flow, and volume of air exhaled. Several spirometric tests have been shown to correlate with outcome in thoracic surgery. Forced expiratory volume in 1 second (FEV₁) and forced vital capacity (FVC) are two such tests, with FEV₁ having the highest predictive value for complications.^10–12 Guidelines from the American College of Chest Physicians (ACCP)¹³ and the British Thoracic Society (BTS)¹⁴ suggest that after maximum bronchodilator therapy, an FEV₁ more than 2 L before a pneumonectomy or FEV₁ more than 1.5 L before a lobectomy is sufficient for a patient to tolerate surgery, assuming no significant dyspnea on exertion or no interstitial lung disease. The difficulty with using an absolute number for cutoffs is that FEV₁ depends on age, gender, and patient size. Because absolute values may unnecessarily exclude an elderly, small woman from surgery, FEV₁ and FVC are reported as a percentage of predicted value, which takes other factors into account. These same guidelines recommend an FEV₁ of more than 80% as sufficient to tolerate either pneumonectomy or lobectomy without further workup.

If a patient does not meet the above criterion, a more detailed approach is needed to estimate the patient’s predicted postoperative FEV₁ (ppo-FEV₁). Several methods can be used to estimate ppo-FEV₁. The most basic (the anatomic method)¹⁵ involves counting the number of segments to be removed. See Figure 9–1.

Figure 9–1. Example of anatomic method for calculating predicted postoperative FEV_1. RUL = right upper lobe, RML = right middle lobe, RLL = right lower lobe, LUL = left upper lobe, LLL = left lower lobe

The calculation is as follows:

One advantage of this method is that postoperative lung mechanics can be calculated after a second lobectomy or completion pneumonectomy. An alternative method using radionuclide perfusion^16,17 may be better at predicting the ppo-FEV₁ after pneumonectomy than the anatomic method, which may underestimate the actual ppo-FEV₁.¹⁸

The risk of perioperative complications increases when the ppo-FEV₁ less than 40%.^19,20 The ACCP guidelines¹³ suggest that patients with a ppo-FEV₁ less than 40% undergo exercise testing for further risk stratification (see below). A low ppo-FEV₁ is not an absolute contraindication to resection; Linden et al²¹ showed that patients with a ppo-FEV₁ less than 35% could tolerate lung resections. In fact, the newly published guidelines of the European Respiratory Society and the European Society of Thoracic Surgery (ERS/ESTS) recommend a ppo-FEV₁ of 30% as the threshold to define high-risk patients.²² Emerging studies also show that patients with extremely poor lung function may benefit from combined lung volume reduction surgery (LVRS) and resection of a malignant tumor.²³ It appears that the ideal candidate for combined LVRS and lung cancer resection has upper lobe emphysema with a tumor in the emphysematous upper lobe (see Chapter 15 on LVRS.). As anesthetic, surgical, and postoperative techniques improve, continuing studies are needed to determine the lowest spirometry values compatible with an acceptable surgical risk.

Gas Exchange

Unlike the measures of respiratory mechanics obtained by spirometry, other tests evaluate the capacity for gas exchange in the alveoli, including arterial oxygenation (PaO₂), arterial carbon dioxide (PaCO₂), and the diffusing capacity for carbon monoxide (DLCO). Historically, a PaO₂ less than 60 mm Hg breathing ambient air has been considered a contraindication for pulmonary resection. This number should be interpreted cautiously because PaO₂ may improve after lung resection when ventilation-perfusion matching has improved.²⁴ Similarly, a PaCO₂ more than 45 mm Hg has historically been the upper limit of acceptable hypercapnea before lung resection, but some studies have shown that complications do not necessarily increase with a PaCO₂ more than 45 mm Hg.²⁵

Because of the limitations of PaO₂ and PaCO₂ values, DLCO is now considered the most useful test for evaluating gas exchange in the alveoli. The value is relatively easy to obtain and is often performed with other pulmonary function tests. To measure DLCO, the patient inhales a small amount of carbon monoxide and air and holds his breath for 10 seconds. When the patient exhales, the amount of exhaled carbon monoxide is measured and the diffusing capacity is calculated as the difference between the inhaled and exhaled amount.

In a retrospective analysis, Ferguson et al²⁶ found that DLCO correlated with surgical morbidity and mortality, perhaps even more so than FEV₁. The predicted postoperative DLCO (ppo-DLCO) can be calculated in the same manner as ppo-FEV₁ (see page 152). In another recent study, ppo-DLCO correlated with morbidity and mortality even in patients with normal spirometric values.²⁷

As with ppo-FEV₁, several studies have shown that if the ppo-DLCO is less than 40%, perioperative risk is significantly increased.^19,28 The ACCP guidelines¹³ suggest further risk stratification with formal exercise testing in patients with a ppo-DLCO less than 40%. Several groups have suggested that a product of % ppo-FEV₁ × % ppo-DLCO less than 1650 may be even more sensitive for revealing patients at high risk for perioperative complications.^28,29

Split Lung Function Tests and Ventilation-Perfusion Scans

Given the strain on the cardiopulmonary system after pulmonary resection, a number of techniques have been developed to try to simulate this change and predict the body’s response to the resection of a portion of lung parenchyma. Techniques involving temporary occlusion of a bronchus or pulmonary artery have been described.³⁰ If a pulmonary artery or lobar branch is occluded and pulmonary artery pressure does not change significantly, it is presumed that the remaining pulmonary vasculature is able to accommodate. These tests are invasive, resource intensive, and not widely used. They also may be misleading because pulmonary artery pressure may remain constant due to a failing right ventricle rather than accommodation of the pulmonary vasculature.³¹

Ventilation-perfusion scintigraphy scans (V/Q scans) have also been used in preoperative assessment to determine the relative contribution of each lung to overall ventilation.³² A V/Q scan has two parts. The first part measures ventilation after a patient inhales a radioactive isotope that shows which parts of the lung are ventilated. The second part measures the perfusion of the lung after a separate radioactive isotope is injected to reveal which areas of the lung are perfused. V/Q scans appear to have reasonable correlation for predicting ppo-FEV₁ and ppo-FVC.³³ With this technique, an obstructed or underperfused area of lung parenchyma can be detected and the ppo-FEV₁ adjusted accordingly. If a patient has a ppo-FEV₁ less than 40% by the anatomic method, a V/Q scan may adjust the ppo-FEV₁ upwards. For example, if the patient from Figure 9–1 had a V/Q scan that showed his left lung received only 42% of the perfusion, then his revised ppo-FEV₁ would be 0.63 × (1–0.42) or 36.5%. If the revised ppo-FEV₁ still identifies the patient as high risk, exercise testing is generally recommended for further risk stratification (see page 155).

Flow-Volume Loops

Flow-volume loops are occasionally obtained before thoracic surgery to supplement other tests and are performed in the same manner as spirometry. They may be useful in the evaluation of a mediastinal mass.³⁴ (See also Chapter 12 on mediastinal masses.) A flow-volume loop can identify an intrathoracic airway obstruction by showing airflow limitation in the expiratory limb. Flow-volume loops were ordered frequently in the past, before more sophisticated imaging techniques of the intrathoracic airway were developed. Despite other advances in imaging, flow-volume loops have value because they measure airflow limitations throughout the entire respiratory cycle rather than at a single point in time. If a patient has few or no comorbidities and does not describe positional dyspnea or coughing, these tests are often omitted.

Exercise Testing

Patients may also undergo exercise testing before thoracic surgery. Rather than measuring isolated respiratory mechanics or gas exchange, exercise testing examines the function of the integrated cardiopulmonary system. Historically, stair climbing has been used as one functional assessment of overall cardiorespiratory status. The ability to climb three flights of stairs indicated the ability to tolerate a lobectomy; climbing five flights of stairs indicated the ability to tolerate a pneumonectomy. Surgical complications and mortality have been shown to correlate with inability to climb stairs.³⁵

The lack of standardization for stair climbing makes the test somewhat problematic: the speed of ascent, duration of climbing, and number of stairs per flight may vary. Nevertheless, stair climbing still provides an easy, inexpensive estimate of the patient’s exercise tolerance. In one study, patients who climbed fewer than 12 meters owing to symptoms of dyspnea had increased complications and mortality compared to those who could climb higher than 22 meters.³⁶ When combined with pulse oximetry, stair climbing may increase the sensitivity of predicting postoperative complications.³⁷

A more objective measurement of exercise tolerance is the measure of maximal oxygen consumption (VO₂max), the gold standard for evaluation of exercise tolerance. The patient exercises—often on a treadmill—while wearing a mask that measures the volume and concentration of inhaled and exhaled gases. Many studies have shown that postoperative complications and mortality increase as preoperative VO₂max decreases.^38–41 A VO₂max less than 10 mL/kg/min is a marker for increased risk of complications and mortality. Patients with a low ppo-FEV₁ or ppo-DLCO with a VO₂max between 10 and 15 mL/kg/min also have a high rate of adverse events. Indeed the ACCP guidelines recommend that patients with a VO₂max less than 10 mL/kg/min or a VO₂max less than 15 mL/kg/min with both ppo-FEV₁ less than 40% and ppo-DLCO less than 40% be counseled about nonstandard surgery (ie, segmentectomy) or nonoperative therapies.¹³

Measuring VO₂max is time-consuming and resource intensive. Several other methods have been devised as alternatives for formal measurement of VO₂max. The 6-minute walk test,⁴² which measures how far a patient can walk in 6 minutes, correlates well with VO₂max and predicts complications after pulmonary resection.^43,44 A distance of less than 2000 feet (610 meters) correlates to a VO₂max of less than 15mL/kg/min.⁴⁵

The shuttle walk test is another surrogate for VO₂max. In this test, the patient walks between two markers 10 meters apart, paced by an audio signal, until too tired to continue.⁴⁶ The shuttle walk test is used frequently in the nonsurgical assessment of pulmonary exercise tolerance. This test shows reasonable correlation to VO₂max.⁴⁷ Somewhat limited data suggest that a patient unable to complete 25 shuttles on two occasions is likely to have a VO₂max less than 10 mL/kg/min.⁴⁸

Summary Recommendations for Respiratory Evaluation

Given the poor prognosis of cancer without surgery and patients’ willingness to accept higher levels of risk, every effort should be made to optimize the medical condition of a patient so that surgery can be considered. All patients should have spirometry and DLCO testing. If the FEV₁ and DLCO are more than 80% predicted, no further evaluation is needed. For patients with FEV₁ less than 80% or DLCO less than 80%, a ppo-FEV₁ and ppo-DLCO should be calculated using the anatomic method. If either ppo-FEV₁ or ppo-DLCO is less than 40%, a V/Q scan can be considered to further refine predicted postoperative function. Alternatively, formal exercise testing with measurement of VO₂max should be performed. If the institution does not have the capacity to measure VO₂max, either stair climbing, the shuttle walk test, or the 6-minute walk test can be substituted. For those patients who, after exercise testing, are still at high risk for complications or death, nonstandard surgery (segmentectomy, combined LVRS/cancer resection) or nonoperative therapy should be discussed. A decision is made on a case by case basis. The algorithm in Figure 9–2 summarizes these recommendations.

Figure 9–2. Preoperative physiologic assessment for lung resection. (Data modified from Figure 1 in Colice GL, Shafazand S, Griffin JP, et al.¹³)

CARDIAC EVALUATION

Although guidelines exist for cardiac evaluation before noncardiac surgery^49,50 (see also Figure 9–3), they are not specific to thoracic surgery. The objective of this section is to clarify some important aspects that are specific to thoracic surgical patients.

Figure 9–3. ACC/AHA guidelines for preoperative cardiac evaluation for noncardiac surgery. Thoracic surgery is categorized as an intermediate risk procedure, and preoperative cardiac evaluation is generally not indicated unless the patient exhibits an active cardiac condition, even in the presence of multiple risk factors for coronary disease. (Reprinted with permission from Circulation. 2007;116:e418-e500. ©2007 American Heart Association, Inc.)

In general, the guidelines of the American College of Cardiology and the American Heart Association (ACC/AHA) classify intrathoracic surgery as a procedure of intermediate cardiac risk. That is, the risk of cardiac death or nonfatal myocardial infarction is 1% to 5%.⁴⁹ In our view, this is somewhat of an oversimplification. Even among pulmonary operations, not all resections will generate the same amount of stress on the cardiovascular system. An extrapleural pneumonectomy, for example, with its potential for blood loss and major effects on pulmonary vascular resistance, is a higher risk surgery than a wedge resection.

The revised ACC/AHA guidelines⁴⁹ recommend that if a thoracic surgical patient has an exercise tolerance of at least 4 metabolic equivalents (the ability to walk up one flight of steps), no further coronary evaluation is necessary, even in the presence of multiple risk factors, unless such evaluation would change management. As with ordering any test, the physician should consider what is to be done with the results. A positive stress test is often followed by cardiology consultation and coronary angiography. If a flow-limiting obstruction is found during angiography, then angioplasty and stenting is considered. Stents obligate the patient to a period of anticoagulation, which may necessitate delaying surgery (see below).

Stress testing is still warranted for a patient with significant risk factors and poor exercise tolerance as an additional means of risk stratification.

We recommend involving a cardiologist in a multidisciplinary approach for patients considered at high risk for cardiovascular complications after thoracic surgery to help define the pathology, further quantify risk, and optimize preoperative management.

Coronary Stents

Patients with a flow-limiting coronary obstruction have an area of myocardium at risk for ischemia that perhaps could be relieved by stenting. Some enthusiasm for stenting in patients with stable coronary disease has been tempered by the COURAGE trial, which showed no significant reduction in death or rate of acute coronary syndrome with bare metal stents compared to medical therapy.⁵¹ This trial, however, did not specifically focus on the perioperative setting, nor did it evaluate drug-eluting stents. As perioperative ischemia has a different pathophysiology than nonperioperative ischemia (supply-demand ratio mismatch rather than a plaque rupture and acute thrombus, in most cases), many practitioners still consider relieving the flow-limiting obstruction with stents.

When a stent is placed, the patient is obligated to a period of anticoagulation with multiple platelet-inhibiting drugs. Prematurely stopping platelet inhibitors can lead to acute thrombosis of the stent and ST elevation myocardial infarction. The practitioner must weigh the risk of stopping anticoagulation before a surgical procedure with the risk of delaying surgery. This decision is best made in consultation with a cardiologist. Compared to bare-metal stents (BMS), drug-eluting stents (DES) take a longer period of time to endothelialize,⁵² which makes them prone to thrombosis for longer. Because of reports of acute thrombosis in patients 1 year after placement of DES,⁵³ the ACC/AHA have revised the guidelines on antiplatelet therapy after coronary stents. Dual antiplatelet therapy is recommended for at least 1 month after BMS placement, and for 1 year following a DES.⁵⁴ Again, in patients with recent stents, the risk of discontinuing antiplatelet therapy must be weighed against the risk of delaying potentially curative surgery or operating on an anticoagulated patient and necessarily foregoing postoperative epidural analgesia.