Evaluation of Comorbidity

Comorbid conditions are best evaluated with use of the Charlson Comorbidity Index ( Table 26.1 ), which has been shown to be an independent predictor of surgical mortality as well as long-term survival. A subset of 1844 patients with lung cancer who had had surgical resection in Norway from 1993 to the end of 2005 was evaluated according to the Charlson Comorbidity Index, and potential factors influencing 30-day mortality were analyzed. The overall mortality rate within 30 days postoperatively was 4.4%. Male gender (odds ratio, 1.76), older age (odds ratio, 3.38 for an age between 70 and 79 years), right-sided tumors (odds ratio, 1.73), and extensive procedures (odds ratio, 4.54 for pneumonectomy) were identified as risk factors for postoperative mortality in multivariate analysis. The Charlson Comorbidity Index was identified as an independent risk factor for postoperative mortality ( p = 0.017).

In a study of 433 consecutive patients (340 men and 93 women) who underwent curative resection for the treatment of nonsmall cell lung cancer, the Charlson Comorbidity Index was used to estimate the risk of mortality. The overall 5-year survival rate was 52% among patients with a Charlson Comorbidity Index of 0, 48% among those with an index of 1 or 2, and 28% among those with an index of 3 or more. Multivariate analysis showed that age; a Charlson comorbidity grade of 1 or 2; a Charlson comorbidity grade of 3 or more; bilobectomy; pneumonectomy; and pathologic stages IB, IIB, IIIA, IIIB, and IV were associated with impaired survival.

Combining the score with physiologic parameters should therefore help in deciding whether a patient may be a candidate for surgery.

Estimation of Cardiac Risk

The risk of major cardiac events, defined as the occurrence of ventricular fibrillation, pulmonary edema, complete heart block, cardiac arrest, or cardiac death, during admission has been reported to be approximately 3% after major anatomic lung resection. Typical candidates for pulmonary surgery for the management of lung cancer usually have both pulmonary and cardiac diseases as a result of cigarette smoking and are potentially at increased risk for perioperative cardiovascular complications. Unfortunately, the available literature specific to cardiac risk in patients undergoing surgery for the management of lung cancer is minimal, and most of what can currently be recommended must be extrapolated from literature on intraabdominal surgery and suprainguinal vascular surgery, both of which, like lung resection, are regarded as high-risk procedures from a cardiac standpoint.

A recent study using Surveillance, Epidemiology, and End Results Medicare data on patients undergoing resection for the management of lung cancer within 1 year after coronary stenting showed that patients who had been treated with stenting had higher rates of major cardiovascular events and mortality (9.3% and 7.7%, respectively) in comparison with those who had not (4.9% and 4.6%, respectively; p < 0.0001 for both comparisons).

Two organizations have produced guidelines on the evaluation and treatment of cardiac risk factors in candidates for lung resection: (1) the European Respiratory Society/European Society of Thoracic Surgeons (ERS/ESTS) joint task force and (2) the American College of Chest Physicians (ACCP).

In general, a detailed evaluation for coronary heart disease is not recommended for patients who have an acceptable exercise tolerance or a Revised Cardiac Risk Index (RCRI) of less than 2.5.

The RCRI, as originally described by Lee et al., is a four-class cardiac risk score with six factors, including a history of coronary artery disease, cerebrovascular accident, insulin-dependent diabetes, congestive heart failure, a serum creatinine level of greater than 2 mg/dL, and high-risk surgery. All factors are equally weighted, and one point is assigned for the presence of each factor.

Although the RCRI was cited as the preferred cardiac risk score in the recently published American College of Cardiology/American Heart Association (ACC/AHA) and European Society of Cardiology/European Society of Anaesthesiology guidelines as well as by the joint ERS/ESTS task force on fitness for radical treatment of patients with lung cancer, this score originally was developed from a generic surgical population that included only a small group of thoracic patients. Brunelli et al. recently recalibrated the RCRI in a study involving a large population of candidates for major anatomic lung resection to obtain a more specific tool for our setting. In that study, only four of the original six factors were found to be reliably associated with major cardiac morbidity and these four factors were assigned different weights (history of coronary artery disease, 1.5 points; cerebrovascular disease, 1.5 points; serum creatinine level of greater than 2 mg/dL, 1 point; and pneumonectomy, 1.5 points). The resulting aggregate score, ranging from 0 to 5.5 points and named the ThRCRI, was found to be more accurate than the traditional score in this population (c index, 0.72 compared with 0.62; p = 0.004). The risk of major cardiac events was 23% for patients with an aggregate score of more than 2.5 (class D), compared with 1.5% for those with a score of 0 (class A).

The ThRCRI was subsequently validated in a number of studies. Most recently, the score was tested and validated in a large population of patients who were included in the Society of Thoracic Surgeons (STS) database. Major cardiovascular complications were reported in 4.3% of more than 26,000 patients who underwent pulmonary anatomic resection. The average ThRCRI score for patients without major cardiovascular complications was half of that for patients with complications (0.6 compared with 1.1; p < 0.0001). Incremental differences in the risk of major cardiovascular complications were noted among the score categories (grade A, 2.9%; grade B, 5.8%; grade C, 11.9%; grade D, 11.1%; p < 0.0001). On the basis of this recent evidence, the ACCP guidelines included this parameter in their updated cardiac algorithm.

According to the ACC/AHA guidelines, noninvasive cardiac evaluation is recommended for patients with limited capacity for exercise, those with a ThRCRI of more than 1.5, and those with a known or newly suspected cardiac condition to identify the relatively small proportion of patients who need intensified intervention to control heart failure or arrhythmias or to treat underlying myocardial ischemia.

Appropriately aggressive cardiac interventions should be instituted before surgery only for patients who would need such interventions irrespective of the planned surgery. However, prophylactic coronary revascularization before surgery in patients who otherwise do not need such a procedure does not appear to reduce perioperative risk. McFalls et al. recently demonstrated that, in a population of patients undergoing major elective vascular surgery who had concomitant stenosis of more than 70% in one or more coronary vessels, prophylactic percutaneous coronary intervention or coronary artery bypass grafting did not change the risk of 30-day mortality, postoperative myocardial ischemia, or long-term survival.

Recent data from the Perioperative Ischemic Evaluation study group indicated that although commonly used regimens of perioperative beta-blockers reduced the risk of cardiovascular death and nonfatal myocardial ischemia (hazard ratio, 0.84), they actually increased the risk of stroke (hazard ratio, 2.17) and overall mortality (hazard ratio, 1.33), perhaps by interfering with stress responses in critically ill patients. Therefore, the new institution of a beta-blocker therapy is not recommended for patients with ischemic heart disease who are not already taking them.

Finally, CPET has been shown to be a useful tool for detecting both overt and occult exercise-induced myocardial ischemia with a diagnostic accuracy similar to single-photon emission computed tomographic myocardial perfusion testing and superior to standard electrocardiographic stress testing. For this reason, CPET can be proposed as a noninvasive test for detecting and quantifying myocardial perfusion defects in patients who are at increased risk for coronary artery disease.

Predicted Postoperative Forced Expiratory Volume in 1 Second

ppoFEV ₁ , which is estimated on the basis of the number of functioning, nonobstructed segments to be removed during an operation, traditionally has been used to stratify respiratory risk in candidates for lung resection. The following equations can be applied to estimate the residual lung function.

For candidates for pneumonectomy, the perfusion method is used with the following formula:

A quantitative radionuclide perfusion scan is performed to measure the fraction of total perfusion for the resected lung. For candidates for lobectomy, the anatomic method is used with the following formula:

The number of functional or unobstructed lung segments to be removed is represented by a , and the total number of functional segments is represented by b .

The findings of bronchoscopy and computerized tomographic scanning should be used to assess and estimate the patency of the bronchus and segmental structure.

Many studies have investigated the role of ppoFEV ₁ in predicting postoperative complications and in selecting patients for surgery. Olsen et al. were the first to suggest a safety threshold value of 0.8 L as the lower limit for surgical resection. However, Pate et al. found that patients with a mean ppoFEV ₁ of as low as 0.7 L tolerated thoracotomy for the resection of lung cancer. The main limitation of those early studies is that they used an absolute value of ppoFEV ₁ . This method might prevent older patients, patients of small stature, and female patients, all of whom might tolerate a lower absolute FEV ₁ , from having a potentially curative resection for the management of lung cancer.

Markos et al. were the first to propose using a percentage of the predicted value as the cutoff value. They found that half of the patients with a ppoFEV ₁ of less than 40% of the predicted value died in the perioperative period. Other authors confirmed that perioperative risk increases substantially when the ppoFEV ₁ is less than 40% of the predicted normal value. The predictive role of ppoFEV ₁ recently was challenged in investigations that showed an acceptable mortality rate among patients with prohibitive FEV ₁ or ppoFEV ₁ values who underwent lung resection.

Alam et al. demonstrated that the odds ratio for the development of postoperative respiratory complications increased as the ppoFEV ₁ and ppoDLCO decreased (with a 10% increase in the risk of complications for every 5% decrease in predicted postoperative lung function). Brunelli et al. showed that ppoFEV ₁ was not associated with an increased risk of complications in those with FEV ₁ less than 70%.

These findings may be partly explained by the so-called lobar volume reduction effect, which can reduce functional loss in patients with airflow limitations. In candidates for lobectomy with lung cancer and moderate-to-severe COPD, resection of the most affected parenchyma may determine an actual improvement in the elastic recoil, a reduction of the airflow resistance, and an improvement in pulmonary mechanics and ventilation–perfusion matching, similar to what happens in typical candidates for lung volume reduction surgery with end-stage heterogeneous emphysema.

In this regard, many studies already have shown the minimal loss or even improvement of pulmonary function after lobectomy in patients with obstruction, calling into question the traditional operability criteria that are primarily based on pulmonary parameters.

Brunelli et al. recently found that patients with COPD had significantly lower losses of FEV ₁ and DLCO compared with patients without COPD at 3 months after lobectomy for the management of lung cancer (8% compared with 16% and 3% compared with 12%, respectively). In that series, 27% of patients with COPD actually had improvement in FEV ₁ and 34% had improvement in DLCO at 3 months after the operation.

This lobar volume reduction effect takes place very early after lung resection. In fact, 17% of patients with airflow limitation who undergo pulmonary lobectomy actually may have improvement in FEV ₁ at the time of discharge as compared with preoperative measurement.

The early lobar volume reduction effect was confirmed by Varela et al. who showed that the percentage loss of FEV ₁ on the first postoperative day after lobectomy was lower in patients with a higher degree of COPD. These findings indicate that ppoFEV ₁ may not work properly in patients with obstructive disease and cannot be used alone to select patients for surgery, especially those with limited pulmonary function.

Although many studies have shown that ppoFEV ₁ is fairly accurate for predicting the definitive residual FEV ₁ at 3 months to 6 months after surgery, Varela et al. recently demonstrated that it substantially overestimates the actual FEV ₁ in the first postoperative days, when most complications occur. On the first postoperative day, the actual FEV ₁ was measured to be about 30% lower than predicted. This finding may have serious clinical implications whenever ppoFEV ₁ is used for patient selection and risk stratification before surgery.

Carbon Monoxide Lung Diffusion Capacity

In 1988, Ferguson et al. reported that DLCO was a predictor of adverse outcomes after pulmonary resection; in that study, patients with a DLCO of less than 60% had mortality rates as high as 20% and pulmonary complication rates as high as 40%. These findings were subsequently confirmed by other authors.

In addition to being a good predictor of immediate postoperative complications, DLCO is probably the objective parameter that is most closely associated with postoperative quality of life.

ppoDLCO, calculated in the same manner as ppoFEV ₁ , was first shown to be a reliable predictor of pulmonary complications and mortality in 1995. In that series, patients with a ppoDLCO of less than 40% had a mortality rate as high as 23%. Those results were subsequently confirmed by Santini et al. who found an inverse linear correlation between pulmonary complications and ppoDLCO. Patients with a ppoDLCO of less than 30% may have a risk of pulmonary complications of greater than 80%. Recent studies have shown that FEV ₁ and DLCO are poorly correlated and that more than 40% of patients with a normal FEV ₁ (an FEV ₁ of more than 80%) may have a DLCO of less than 80% and that 7% of patients with FEV ₁ >80% may have a ppoDLCO of 40%. Other studies have demonstrated that a reduced ppoDLCO is a reliable predictor of cardiopulmonary morbidity and mortality not only in patients with reduced FEV ₁ but also in those with normal respiratory function. In a recent large study involving approximately 8000 patients from the STS General Thoracic Surgery Database who were treated with lung resection, the percentage of predicted DLCO was strongly associated with the occurrence of pulmonary complications. This association was independent from the COPD status.

On the basis of this evidence, recent functional guidelines have recommended the measurement of DLCO in all candidates for lung resection, regardless of the preoperative FEV ₁ value.

Many patients undergoing major lung resection for the management of cancer receive preoperative chemotherapy. Recent reports have suggested that chemotherapy can be associated with a 10% to 20% reduction in DLCO despite stable or improved spirometric values. These changes are associated with drug-induced structural lung damage and have been associated with an increase in postoperative respiratory complications. Therefore reassessment of pulmonary status and DLCO after induction therapy and prior to resection is recommended to ensure that the operative risk has not increased as a result of newly impaired DLCO.

Video-Assisted Thoracoscopic Surgery

Several reports have shown reduced rates of morbidity among patients who are treated with minimally invasive video-assisted thoracoscopic surgery (VATS) lobectomy. This finding is likely explained by the minimal impact of this operation on the chest wall mechanics. This effect is particularly evident in patients with compromised pulmonary function. Berry et al. reported that, in patients with an FEV ₁ of less than 60% or a DLCO of less than 60% who underwent pulmonary lobectomy with use of either thoracotomy or VATS, thoracotomy was a strong predictor of complications on multivariable analysis (odds ratio, 3.46; p = 0.0007). FEV ₁ and DLCO remained predictors of complications in patients undergoing thoracotomy but not thoracoscopy.

Similarly, in a large population of patients in the STS database who underwent lobectomy, multivariable regression analysis showed that thoracotomy (odds ratio, 1.25; p < 0.001), decreasing FEV ₁ % predicted (odds ratio, 1.01 per unit; p < 0.001), and DLCO% predicted (odds ratio, 1.01 per unit; p < 0.001) independently predicted pulmonary complications. Among patients with an FEV ₁ of less than 60%, the rate of pulmonary complications was markedly increased among those who underwent thoracotomy as compared with those who underwent VATS ( p = 0.023). No significant difference was noted among patients with an FEV ₁ of more than 60% of the predicted value. More recently, Burt et al. (Burt BM, Kosinski AS, Shrager JB, Onaitis MW, Weigel T. Thoracoscopic lobectomy is associated with acceptable morbidity and mortality in patients with predicted postoperative forced expiratory volume in 1 second or diffusing capacity for carbon monoxide less than 40% of normal. J Thorac Cardiovasc Surg. 2014 Jul;148(1):19–28) have shown that patients operated on through VATS and with prohibitive ppoFEV ₁ or ppoDLCO below 40% of predicted value have a far lower risk of mortality compared with case-matched counterparts operated on through thoracotomy. For patients with ppoFEV ₁ % less than 40%, mortality was 4.8% in the open group versus 0.7% in the VATS group ( p = 0.003). Similar results were seen for ppoDLCO% less than 40% (5.2% open, 2.0% VATS, p = 0.003).

Other studies have shown better preservation of pulmonary function compared with preoperative values in patients undergoing VATS lobectomy than in those undergoing thoracotomy.

At this time, the evidence is still too limited to justify a change in the current functional guidelines. However, it appears likely that with the increasing number of patients who are treated with the VATS approach, we will be able to verify whether traditional pulmonary thresholds of operability (mostly derived from series of patients undergoing thoracotomy) should be updated.