CLINICAL INVESTIGATION
Weighing Projections of Physical Decline in Lung Cancer Surgery Decisions Samuel Cykert, MD, Paul R. Walker, MD, Lloyd J. Edwards, PhD, Franklin R. McGuire, MD and Peggye Dilworth-Anderson, PhD
atients diagnosed with early-stage non–small cell lung cancer fear debility and death when facing the specter of surgery. These patients assign extremely low value to states of severe postoperative debility such as limitations in activities of daily living or a bed to chair existence while perceptions of states of less debility, such as cutting current walking ability in half, are more acceptable as a treatment result.1 Because lung cancer patients have impaired physical functioning compared
with healthy individuals of similar age2,3 and often have comorbid illnesses, surgical decisions can be equivocal and vary not by strict medical contraindication but by nonmedical factors, such as race or geography, despite the dire consequences of untreated cancer.4,5 Further complicating difficult decisions for complex patients and their doctors is that the interplay of patients’ race, and personal perceptions result in lower surgical rates especially for African Americans particularly when comorbidities are considered.6 Given these circumstances, the use of postoperative physical functioning could emerge as an objective criterion in the calculus of lung cancer surgery decisions. Past reports have compared postoperative with preoperative functional status in lung resection patients and show some decrement,2,3,7 but these decrements are usually small and 1 report describes the return to baseline function within 3 months.7 In addition, a patient with non–small cell lung cancer who does not undergo an operation with curative intent will either face the consequences of progressive disease and/or alternative therapies. Although functional status has been amply described for lung cancer surgery patients relative to populations without cancer or relative to the preoperative state, significant gaps in this literature exist regarding functional status comparisons to similar patients who do not receive surgical care. We, therefore, pose the question, if patients with earlystage non–small cell lung cancer who undergo lung resection surgery are compared with patients with similar disease who do not undergo surgery, will physical functioning significantly differ a year after diagnosis? By comparing functional status to a nonsurgical group with lung cancer, a true decrement in physical functioning would argue that postoperative functional status should be a significant determinant of surgical decisions, whereas similar decrements in functioning between the groups would lead to the opposite conclusion. We recently performed a prospective cohort study examining decisions in lung cancer surgery6; we now report 1-year outcome results to address the described gap regarding functional status change in a nonsurgical lung cancer group compared with a group treated with lung resection for cure.
From the Division of General Internal Medicine and Clinical Epidemiology (SC), North Carolina Area Health Education Centers Program, University of North Carolina School of Medicine, Chapel Hill, North Carolina; Thoracic Oncology Program (PRW), Leo Jenkins Cancer Center, Brody School of Medicine at East Carolina University, Greenville, North Carolina; Department of Biostatistics (LJE), Gillings School of Global Public Health, The University of North Carolina-Chapel Hill, Chapel Hill, North Carolina; Division of Pulmonary and Critical Care Medicine (FRM), The University of South Carolina School of Medicine; and Department of Health Policy and Management (PD-A), Gillings School of Global Public Health, The University of North Carolina-Chapel Hill, Chapel Hill, North Carolina. Submitted April 15, 2014; accepted in revised form August 20, 2014. Supported by a grant from the American Cancer Society, RSGPB-05217-01-CPPB. The authors have no financial or other conflicts of interest to disclose. Correspondence: Samuel Cykert, MD, 145 North Medical Drive, Campus Box #7165, University of North Carolina-Chapel Hill, Chapel Hill, NC 27599-7165 (E-mail: [email protected]).
This study was approved by the Institutional Review Boards of all participating institutions. Full details of recruitment and study protocols are published elsewhere.6 In brief, patients with early-stage non–small cell lung cancer were identified in pulmonary, oncology, thoracic surgery and generalist practices in 5 communities. Patients were included if they were at least 21 years of age, had either biopsy-proven disease or a lung lesion demonstrating a high probability of lung cancer using a Bayesian algorithm and had stage 1 or 2 disease as defined by clinical, radiological and/or mediastinoscopic criteria. Informed consent was obtained. Patients were either referred by the clinical practices or found through a chest computerized tomography review protocol. Patients who were identified using the computerized tomography protocol were
Abstract: Background: Patients with early-stage lung cancer often have comorbid illnesses and fear debility and death when contemplating surgery. However, data that compare physical function of patients who receive surgery with similar patients who do not are sparse. The authors report 1-year outcome results for surgical and nonsurgical patients in a prospective lung cancer cohort to address this gap. Methods: The authors enrolled 386 patients with early-stage lung cancer. A 106-item survey was completed at the time of enrollment including the ShortForm 12 (SF-12) Health Survey to assess functional status. Patients were followed for a year. Chart abstractions were obtained to determine comorbid illnesses and surgical status. Death was ascertained through vital records. The SF-12 was repeated 1 year after the enrollment. Regression models were constructed to identify predictors of 1-year mortality and decline in physical function. Results: Fifty-nine patients (15.3%) died before 1-year follow-up. Mortality in the surgical group was 10.8% compared with 22.8% in the nonsurgical group (P , 0.001). In regression analysis controlling for age and comorbidities, surgical treatment was associated with a reduction in 1-year mortality (odds ratio: 0.5 and 95% confidence interval: 0.3–1.0) but did not worsen physical function relative to the untreated group (average decrease in physical component score of SF-12 5 1.9 for surgery group and 2.5 for no surgery group, P 5 0.66). Conclusions: Functional decline between surgically treated and untreated patients did not differ. This result casts doubt on its value as a treatment determinant. Cancer mortality seems to be a more essential issue in treatment decisions. Key Indexing Terms: Lung cancer; Physical function; Outcomes. [Am J Med Sci 2015;349(1):61–66.]
P
The American Journal of the Medical Sciences
METHODS
Volume 349, Number 1, January 2015
61
Cykert et al
contacted through the ordering practice unless the ordering physician of the scan was based on the emergency department and no documented follow-up occurred. In these cases, the research associate was allowed to contact the patient directly and arrange further care. Pneumonectomies were infrequently performed at the participating study sites (N 5 9). Given our inability to perform meaningful statistical comparisons for pneumonectomy patients, they were excluded from analysis. All participants were administered a 106-item survey at the time of enrollment. Demographic data were obtained while other survey items explored perceptions of patient-physician communication, attitudes about lung cancer and surgery, religiosity, past health care experiences and access to a regular source of care. Race and ethnicity were self-reported. Physical function was ascertained using the Short-Form 12 (SF-12) Health Survey.8 Chart reviews were performed starting 4 months after enrollment. The chart abstraction included lung surgery status, date of surgery if it occurred, preoperative stage, medical comorbidities and preoperative pulmonary function tests. Physical function was reassessed using the SF-12 by telephone 1 year after enrollment. For patients who had died, death and date of death were verified by searching state vital records. The primary outcomes for this analysis are mortality and physical function as assessed by the SF-12 physical component score (PCS) 1 year after lung cancer diagnosis. Significant physical deterioration was defined as a drop of 5 or greater in the PCS 12 months after diagnosis given that a 4.5-point gain in SF-36 PCS was linked to improved (66 m) 6-minute walk distance in pulmonary rehabilitation research9 and the SF-12 PCS is highly correlated with the SF-36.8,10,11 The physical function outcome is reported for the group of 1-year survivors who completed both the initial and 12-month phone assessment. To account for death as a condition of physical deterioration, we also examined a combined outcome of 5-point decrement in PCS and death reflecting a state of global worsening 1 year after diagnosis inclusive of all enrolled patients. Descriptive statistics including selected bivariate associations were computed for demographic variables, surgical rate, religiosity, a sum of comorbid conditions relevant to lung resection surgery (chronic obstructive lung disease, congestive heart failure, renal insufficiency, poor mobility, cerebrovascular disease, obstructive sleep apnea, oxygen dependence), a history of coronary artery disease, known diabetes, the mental component score (MCS) of the SF-12 and the 3 outcomes noted above. All outcome variables were dichotomous, so a generalized linear model for clustered data with multiple predictors was used to assess the simultaneous effect of predictors on outcome.12 Given that the major goal of this analysis was to gauge the effect of surgery on 1-year mortality and physical functioning, receipt of surgery was a variable in all models. Demographic variables were incorporated into all models. Because of the significance of the presence of 2 or more comorbid illnesses and religiosity in our previous work,6 these measures were also included in the regressions. Coronary artery disease and diabetes were prevalent but did not impact mortality or physical function scores in either bivariate or regression analyses, and so these variables were dropped from the final models. Therefore, for the purpose of this analysis, exactly as per our previous work, significant comorbidities included chronic obstructive lung disease, renal insufficiency, poor mobility, congestive heart failure, cerebrovascular disease, obstructive sleep apnea and oxygen dependence.6 The modeling approach fitted a maximum model using main effects only; however, we did assess the effect of several interactions including race and
62
comorbid illnesses, comorbid illnesses and surgery and age and surgery without detecting additional significance beyond the main effects. All statistical analyses were performed using SAS software version 9.3 (SAS Institute Inc, Cary, NC). A priori statistical significance for independent variables and the primary outcomes was set at #0.05 based on a 2-sided test for bivariate and multivariate associations.
RESULTS As previously described, 386 patients remained eligible in this early-stage lung cancer prospective cohort.6 All patients with absolute contraindications to surgery as defined by pulmonary function results were excluded. Fifty-nine patients (15.3%) died before 1-year follow-up. Thirty-seven patients either refused or were unable to complete the SF-12 survey 1 year after diagnosis. Therefore, 290 baseline and 1-year surveys were available for paired analyses and 349 were available to assess the combined outcome of significant physical function decline or death. Demographic characteristics, religiosity, comorbid conditions and baseline SF-12 scores plus surgical rates according to these characteristics are summarized in Table 1. Gender, educational attainment and marital status are similar to previous reports.4,13 Because the cited reports are based on combined Surveillance, Epidemiology and End Results-Medicare data, the age of our population is marginally younger. We oversampled African Americans because of initial sample size calculations that allowed enough statistical power to make valid comparisons according to race.6 Note that 13% of participants had 2 or more of the designated comorbidities. The prevalence of each individual comorbidity included in the summary variable was as follows: Chronic Obstructive Pulmonary Disease (COPD): 40%; congestive heart failure: 10%; history of cerebrovascular accident (CVA): 6.2%; renal failure: 5.5%; poor mobility (walking limited to a room): 4.7%; oxygen dependence: 2.6%; and obstructive sleep apnea: 1.8%. Bivariate analyses for all 3 outcomes are shown in Table 2. Age above the median (age, 66 years) is strongly associated with death (1-year survivors averaged 65.8 years of age at enrollment compared with 69.5 years in the group who died [P 5 0.01]), significant physical decline and the combined 1-year outcome of death plus physical decline. Patients who placed in the lowest quartile MCS on the baseline SF-12 experienced twice the mortality within a year compared with those with higher MCS scores. There is no relationship between MCS and physical decline or the composite outcome. The age-old conundrum between surgical benefit and the risk of comorbid conditions remained in the bivariate results. One-year mortality in the surgical group was 10.8% compared with 22.8% in the nonsurgical group (P 5 0.002), but patients with 2 or more significant comorbidities at diagnosis had a 27.5% 1-year mortality compared with 13.5% in the group with ,2 (P 5 0.01). However, when comorbidity was stratified by surgical status, we found that the 2 or more comorbidity surgical group had a 1-year mortality rate of 19% compared with 10% for the group with ,2 (P 5 NS); in the no surgery group, these 1-year mortality rates were 31% and 20%, respectively (P 5 NS). There was no difference in 1-year mortality according to race (African Americans 5 15.0 and whites 5 15.4; P 5 0.9), but African Americans who died were 4 years younger than white patients (66.6 versus 70.6 years; P 5 0.05). The physical function analysis was limited to the 290 patients who survived 1 year and answered the baseline and Volume 349, Number 1, January 2015
Physical Decline and Lung Cancer Care
TABLE 1. Baseline characteristics of early-stage lung cancer cohort combined and by surgical status Patients (N 5 386), Characteristic n (%) P Age, above median, yr Surgical rate, % .66 #66 Male gender Surgical rate, % Male Female Black race (All) Surgical rate, % Black White Education .high school Surgical rate, % .High school #High school Household income, lowest tertile Surgical rate, % Upper tertile Lowest tertile Married (All) Surgical rate, % Married Not married No health insurance Surgical rate, % No insurance Insurance SF-12 MCS, lowest quartile Surgical rate, % Upper 3 quartiles Lowest quartile SF-12 PCS, lowest quartile Surgical rate, % Upper 3 quartiles Lowest quartile High religiosityb (All) Surgical rate, % High religiosity Less religiosity Greater than 2 severe comorbidities Surgical rate, % .2 Comorbidities ,2 Comorbidities History CAD Surgical rate, % CAD No CAD History of DM Surgical rate, % DM No DM
190 (49) 0.24 59 65 215 (56) 0.37 60 65 113 (29) 0.048 55a 66 135 (35) 0.68 64 62 125 (32) 0.02 66a 54 246 (64) 0.16 65 58 29 (8) 0.97 62 63 95 (25) 0.04 65a 54 90 (23)
1-year postdiagnosis SF-12. The average drop in the SF-12 PCS at 1 year was 1.9 for the surgical group and 2.5 in the nonsurgical group (P 5 0.66). Thirty-eight percent of lung cancer patients experienced a significant (5 points or more) fall in their PCS at 1 year. Notably, 37.4% of surgical patients experienced this degree of physical deterioration compared with 40.2% in the nonsurgical group again confirming no difference in this outcome (P 5 0.64). Only 45% of patients treated with surgery demonstrated the combined deterioration outcome compared with 55% in the no surgery group (P 5 0.06) with the difference in 1-year mortality accounting for most of the difference in favor of surgery. Male gender and advanced age were the only statistically significant bivariate associations with significant deterioration of physical function and with the combined outcome for deterioration at 1 year. Educational status, lowest income tertile, marital status, religiosity, history of coronary artery disease and diagnosis of diabetes were not associated with any increase in the adverse outcomes. The regression results are shown in Table 3. Controlling for comorbid illness, patients treated with surgery were half as likely to die within a year compared with the nonsurgical group. Conversely, patients with 2 or more significant comorbid illnesses were more likely to die (odds ratio: 2.2 and 95% confidence interval: 1.01–4.8) whether operated on or not. Age remained a significant predictor as those older than 66 years were about twice as likely to die. Preoperative physical functioning as represented by the PCS was not associated with 1-year mortality, but the lowest quartile of MCS continued to be strongly associated with 1-year mortality independent of age, surgical treatment and comorbid illnesses. Regarding significant physical decline and the composite deterioration outcome, male gender and age older than 66 years continued to be associated with 1-year physical decline in regression models whether patients were treated with surgery or not. Surgery was not associated with loss of physical function using either the 5 or more point drop in PCS or the combined variable. Sensitivity analyses were performed using PCS decline at 1 year of 2 or more points, 10 or more points and 15 or more points as the definition for clinically important functional declines. Lung cancer surgery was not predictive of significant functional loss compared with no surgery patients regardless of the designated cut point.
0.01 66a 51 162 (42) ,0.001 52a 70 51 (13) ,0.001 31a 68 100 (26) 0.74 64 62 86 (22)
Difference in surgical status statistically significant (P , 0.05). As determined by agreement with the statement, faith alone can cure disease. CAD, coronary artery disease; DM, diabetes mellitus. a
0.97 63 63
DISCUSSION Physical decline related to lung cancer surgery is an understandable worry for patients facing this decision given the prevalence of smoking-related illnesses in this group. In our initial report, patients who believed that lung cancer surgery would worsen their quality of life in a year were 1 quarter as likely to proceed to surgery.6 Although we know that the extent of lung resection relates to postoperative decline in physical function and that function is generally worse than that of agematched patients,2,14–16 few data are available to compare functional status in similar lung cancer patients who were not surgically treated. The good news for these patients is that physical function 1 year after diagnosis decreases similarly whether a patient has surgery or not. More germane to making
b
Copyright © 2014 by the Southern Society for Clinical Investigation. Unauthorized reproduction of this article is prohibited.
63
Cykert et al
TABLE 2. Bivariate analysis of patient outcomes 1 year after diagnosis of early-stage non–small cell lung cancer according to demographics, functional status, surgical status and comorbid conditions Death Physical function loss of 5 points Combined outcome (death or Characteristic (%) P or more on SF-12 (%) P 5-point loss of physical function) (%) P Age, yr .66 #66 Gender Male Female Race Black White Education HS or less More than HS Income Lowest tertile Top 2 tertiles Married Yes No Health Insurance Yes No MCS Lowest quartile Top 3 quartiles PCS Lowest quartile Top 3 quartiles Religiosity High Low Surgery Yes No CAD Yes No Diabetes Yes No Comorbid conditions $2 ,2
0.01a 20.0 10.7
48.2 29.9
59.5 38.2 0.05a
0.37 16.7 13.5
43.4 32.3 0.93
15.0 15.4
0.17
0.93
0.74
0.74
0.27
0.18
0.68
0.44
0.30 46.7 52.4
0.21 39.4 26.1
0.005a 24.2 12.4
0.34 44.9 50.4
37.5 40.0
15.7 10.3
0.77 49.6 48.0
33.7 40.4
13.4 18.6
0.26 43.9 50.6
37.4 39.4
14.4 15.7
0.03a 53.9 42.3
32.1 40.8
15.2 15.6
,0.001a
0.001a
0.14 49.9 34.6
0.73
0.31
38.9 36.5
47.2 53.5
b
b
0.45 17.8 14.5 0.35 17.3 13.8
0.13 43.4 34.7
0.002a 10.8 22.8
0.65 37.4 40.2
0.91 15.0 15.4
0.13
0.54
0.17 55.2 46.6
0.91 37.7 38.5
0.01a 27.5 13.5
0.06 44.8 55.4
45.9 35.9
17.4 14.7
0.09 54.0 44.8
0.80 50.0 48.4
0.52 33.3 39.2
0.52 53.2 48.2
a
Statistically significant. PCS was not used as an independent variable for analyses related to physical functioning. CAD, coronary artery disease. b
surgical decisions is that lung cancer surgery resulted in nearly a 50% reduction in 1-year mortality while controlling for age and comorbid illness. The tempering statistic, however, is that for the 13% of patients who had 2 or more defined comorbidities, the odds ratio for death at 1 year was 2.2 regardless of surgical status. When examining unadjusted data for patients who underwent surgery, we found that the comorbidity group
64
had a 1-year mortality rate of 19%. Although this death rate was 9 percentage points higher than those who had surgery without comorbid burden, it was a 12-point absolute decrease compared with the 31% mortality experienced by the group with similar comorbidities without surgical intervention. Clearly, early-stage lung cancer patients with these comorbid illnesses have a higher 1-year mortality rate whether treated surgically or not, but Volume 349, Number 1, January 2015
Physical Decline and Lung Cancer Care
TABLE 3. The association of patient characteristics, functional status and lung cancer surgery with outcomes 1 year after diagnosis of stage 1 or 2, non–small cell lung cancer as determined by regression analyses Odds ratio (95% Odds ratio (95% CI) significant Odds ratio (95% CI) combined outcome Characteristic CI) death loss of physical function (death or loss of physical function) Age .66 yr Male gender Black race Education .high school Lowest income tertile Married Has health insurance Upper 3 quartiles on mental component score of SF-12 Upper 3 quartiles of physical component score of SF-12 High religiosityc Had lung cancer surgery Two or more comorbid conditions
2.3 1.5 1.0 1.1 0.94 0.62 1.4 0.40
(1.2–4.3)a (0.83–2.8) (0.49–2.0) (0.59–2.1) (0.48–1.8) (0.33–1.2) (0.38–5.2) (0.21–0.77)a
2.0 1.7 0.67 1.0 0.78 0.71 1.6 0.95
(1.2–3.3)a (1.05–2.9)a (0.36–1.2) (0.61–1.8) (0.45–1.4) (0.41–1.2) (0.56–4.3) (0.51–1.8)
2.2 1.8 0.78 1.0 0.85 0.67 1.6 0.73
(1.4–3.5)a (1.2–2.9)a (0.45–1.3) (0.64–1.7) (0.51–1.4) (0.41–1.1) (0.66–4.0) (0.43–1.2)
1.2 (0.57–2.5)
b
b
0.96 (0.52–1.8) 0.54 (0.29–1.0)a 2.2 (1.01–4.8)a
1.4 (0.83–2.3) 0.87 (0.51–1.5) 0.77 (0.34–1.7)
1.2 (0.78–2.0) 0.72 (0.44–1.1) 1.1 (0.59–2.2)
a
Statistically significant. PCS was not used as an independent variable for analyses related to physical functioning. As determined by agreement with the statement, faith alone can cure disease. CI, confidence interval.
b c
surgery seems to attenuate the death rate in a manner similar to Battafarano’s report.17 A survival analysis reported by Bach demonstrated a doubling of mortality in the nonsurgical group within 1 year similar to our results.4 In addition, recent cohort studies consisting of thousands of early-stage lung cancer patients examined geographic variation in surgical rates and demonstrated a strong relationship between higher surgical rates and survival, again within 1 year.5,18,19 Gray et al5 specifically analyzed a subset of patients who were over the age of 66 with chronic obstructive lung disease. They found a 7% decrease in absolute mortality at 1 year (20% versus 27%) for these sicker patients in high-surgery regions (lung resection rates .79%) compared with low-surgery regions (lung resection rates ,63%) and demonstrated that the excess mortality was lung cancer specific. Although the dilemma of which lung cancer patients are too sick for surgery remains, it seems that more patients should be given the benefit of the doubt when absolute contraindications are not in play. As noted, patients with 2 or more comorbidities have higher death rates, but 1-year death rates are even worse if they do not get an operation. The relationship of lower SF-12 MCS and mortality requires further thought. Reductions in MCS have been shown to correlate with depression,8,20,21 and depression has been associated with worse outcomes in patients with chronic obstructive lung disease including increased mortality.22 Depression has also been associated with increased mortality for cancer in general23 and lung cancer patients specifically.24 Although treatment has not been shown to improve survival in patients with metastatic breast cancer,25 a depression treatment program instituted for older adults in primary care resulted in lower 5-year mortality for patients diagnosed with cancer.26 Given the high mortality rates in patients with early-stage lung cancer, any incremental benefit in survival, especially for those who go through the rigor of surgery, would be valuable. Therefore, an in-depth examination of the diagnosis and treatment of depression in early-stage lung cancer is warranted.
Several limitations of this study need to be considered. Although our initial report demonstrated an 11% lower surgical rate for African American patients, race was not a predictor of 1-year mortality. This finding might be explained by the fact that surgery itself was the main determinant of survival, overriding any racial effect or the fact that we were simply underpowered to find mortality differences by race. Notably, the 1-year mortality rate was 15% for both racial groups. However, again, African Americans who died were 4 years younger suggesting that the lower operative rate contributed to worse survival as has been described by Bach et al.4 Another limitation is that we did not have pneumonectomy patients in this cohort. Functional impairment and survival are worse after pneumonectomy,15,27 so we cannot apply our findings to those patients with lung cancer who need this more extensive procedure. Finally, comorbidity counts or scores do not necessarily account for the severity of illness associated with a single specific comorbidity. However, use of these scores has been demonstrated to be reliable and valid predictors of outcomes for clinical research generally28 and lung cancer specifically.29 A recent consensus report emphasized the need to incorporate physical functioning into the decision calculus surrounding surgery for high-risk patients with stage I non–small cell disease.30 Although functional status and the decrement attributable to lung resection are important, it should not be considered without factoring in the impact of progressive cancer or radiation therapy. This report highlights the fact that functional loss in the nonoperated cohort is comparable to lung resection. Mortality remains important despite the risk of surgery the survival curves diverge in favor of surgery as early as 1 year after diagnosis.4,5 For individuals without absolute contraindications and less than 2 significant comorbid illnesses, these data suggest that lung cancer surgery should be strongly considered; even within the comorbid illness group, surgical intervention still warrants careful consideration. This tenet holds particularly true for African Americans who are staged invasively less often than similar white patients,13 receive recommendations for surgical
Copyright © 2014 by the Southern Society for Clinical Investigation. Unauthorized reproduction of this article is prohibited.
65
Cykert et al
care less frequently in the context of the absence of absolute contraindications6,13 and suffer lung cancer death at a younger age even when presenting with early-stage disease.4
13. Lathan CS, Neville BA, Earle CC. The effect of race on invasive staging and surgery in non-small-cell lung cancer. J Clin Oncol 2006; 24:413–18.
CONCLUSIONS
14. Bryant AS, Cerfolio RJ, Minnich DJ. Survival and quality of life at least 1 year after pneumonectomy. J Thorac Cardiovasc Surg 2012;144: 1139–45.
The lack of difference in functional decline between surgically treated and untreated patients casts doubt on its value as a treatment determinant at least in the setting of lobectomy and smaller lung resection procedures. Cancer mortality should remain an important consideration in surgical decisions for these patients. ACKNOWLEDGMENTS The Cecil G. Sheps Center for Health Services Research served as the multidisciplinary collaborative center for this work. The authors wish to acknowledge Dr. Richard Feins for his contributions to designing patient recruitment protocols and his insight regarding lung cancer surgery decisions and the role of comorbid illness and communication factors in the decisionmaking process. The authors thank the American Cancer Society (ACS) for their sponsorship of this project. Note that the ACS did not influence the performance of the study and did not play a role in the design of the research or writing of the article. REFERENCES 1. Cykert S, Kissling G, Hansen CJ. Patient preferences regarding possible outcomes of lung resection what outcomes should preoperative evaluations target? Chest 2000;117:1551–9. 2. Handy JR, Asaph JW, Skokan L, et al. What happens to patients undergoing lung cancer surgery?*: outcomes and quality of life before and after surgery. Chest 2002;122:21–30. 3. Myrdal G, Valtysdottir S, Lambe M, et al. Quality of life following lung cancer surgery. Thorax 2003;58:194–7. 4. Bach PB, Cramer LD, Warren JL, et al. Racial differences in the treatment of early-stage lung cancer. N Engl J Med 1999;341:1198–205. 5. Gray SW, Landrum MB, Lamont EB, et al. Improved outcomes associated with higher surgery rates for older patients with early stage nonsmall cell lung cancer. Cancer 2012;118:1404–11. 6. Cykert S, Dilworth-Anderson P, Monroe MH, et al. Factors associated with decisions to undergo surgery among patients with newly diagnosed early-stage lung cancer. JAMA 2010;303:2368–76. 7. Brunelli A, Socci L, Refai M, et al. Quality of life before and after major lung resection for lung cancer: a prospective follow-up analysis. Commentary. Ann Thorac Surg. 2007;84:410–16. 8. Ware J Jr, Kosinski M, Keller SD. A 12-item short-form health survey: construction of scales and preliminary tests of reliability and validity. Med Care 1996;34:220–33. 9. Benzo R, Flume PA, Turner D, et al. Effect of pulmonary rehabilitation on quality of life in patients with COPD: the use of SF-36 summary scores as outcomes measures. J Cardiopulm Rehabil 2000;20:231–4. 10. Ware JE Jr, Gandek B, Kosinski M, et al. The equivalence of SF-36 summary health scores estimated using standard and country-specific algorithms in 10 countries: results from the IQOLA project. J Clin Epidemiol 1998;51:1167–70. 11. Pickard AS, Johnson JA, Penn A, et al. Replicability of SF-36 summary scores by the SF-12 in stroke patients. Stroke 1999;30:1213–17. 12. Diggle P, Liang L, Zeger SL, ed. The analysis of longitudinal data. New York (NY): Oxford University Press; 1994.
66
15. Schulte T, Schniewind B, Dohrmann P, et al. The extent of lung parenchyma resection significantly impacts long-term quality of life in patients with non-small cell lung cancer. Chest 2009;135:322–9. 16. Möller A, Sartipy U. Predictors of postoperative quality of life after surgery for lung cancer. J Thorac Oncol 2012;7:406. 17. Battafarano RJ, Piccirillo JF, Meyers BF, et al. Impact of comorbidity on survival after surgical resection in patients with stage I non-small cell lung cancer. J Thorac Cardiovasc Surg 2002;123:280–7. 18. Riaz SP, Lüchtenborg M, Jack RH, et al. Variation in surgical resection for lung cancer in relation to survival: population-based study in England 2004–2006. Eur J Cancer 2012;48:54–60. 19. Wouters MWJM, Siesling S, Jansen-Landheer ML, et al. Variation in treatment and outcome in patients with non-small cell lung cancer by region, hospital type and volume in the Netherlands. Eur J Surg Oncol 2010;36:S83–92. 20. Lipton R, Hamelsky S, Kolodner K, et al. Migraine, quality of life, and depression: a population-based case–control study. Neurology 2000;55:629–35. 21. McHorney CA, Ware JE Jr, Raczek AE. The MOS 36-Item ShortForm Health Survey (SF-36): II. Psychometric and clinical tests of validity in measuring physical and mental health constructs. Med Care 1993;31:247–63. 22. Ng TP, Niti M, Tan WC, et al. Depressive symptoms and chronic obstructive pulmonary disease effect on mortality, hospital readmission, symptom burden, functional status, and quality of life. Arch Intern Med 2007;167:60–7. 23. Stommel M, Given BA, Given CW. Depression and functional status as predictors of death among cancer patients. Cancer 2002;94: 2719–27. 24. Faller H, Bulzebruck H, Drings P, et al. Coping, distress, and survival among patients with lung cancer. Arch Gen Psychiatry 1999;56:756. 25. Goodwin PJ, Leszcz M, Ennis M, et al. The effect of group psychosocial support on survival in metastatic breast cancer. N Engl J Med 2001;345:1719–26. 26. Gallo JJ, Bogner HR, Morales KH, et al. The effect of a primary care practice–based depression intervention on mortality in older adults: a randomized trial. Ann Intern Med 2007;146:689. 27. Chambers A, Routledge T, Pilling J, et al. In elderly patients with lung cancer is resection justified in terms of morbidity, mortality and residual quality of life? Interact Cardiovasc Thorac Surg 2010;10:1015–21. 28. de Groot V, Beckerman H, Lankhorst GJ, et al. How to measure comorbidity: a critical review of available methods. J Clin Epidemiol 2003;56:221–9. 29. Birim Ö, Kappetein AP, Bogers AJ. Charlson comorbidity index as a predictor of long-term outcome after surgery for nonsmall cell lung cancer. Eur J Cardiothorac Surg 2005;28:759–62. 30. Donington J, Ferguson M, Mazzone P, et al. American College of Chest Physicians and Society of Thoracic Surgeons consensus statement for evaluation and management for high-risk patients with stage I nonsmall cell lung cancer. Consensus statement on lung cancer in high-risk patients. Chest 2012;142:1620–35.
Volume 349, Number 1, January 2015
Weighing Projections of Physical Decline in Lung Cancer Surgery Decisions Samuel Cykert, MD, Paul R. Walker, MD, Lloyd J. Edwards, PhD, Franklin R. McGuire, MD and Peggye Dilworth-Anderson, PhD
atients diagnosed with early-stage non–small cell lung cancer fear debility and death when facing the specter of surgery. These patients assign extremely low value to states of severe postoperative debility such as limitations in activities of daily living or a bed to chair existence while perceptions of states of less debility, such as cutting current walking ability in half, are more acceptable as a treatment result.1 Because lung cancer patients have impaired physical functioning compared
with healthy individuals of similar age2,3 and often have comorbid illnesses, surgical decisions can be equivocal and vary not by strict medical contraindication but by nonmedical factors, such as race or geography, despite the dire consequences of untreated cancer.4,5 Further complicating difficult decisions for complex patients and their doctors is that the interplay of patients’ race, and personal perceptions result in lower surgical rates especially for African Americans particularly when comorbidities are considered.6 Given these circumstances, the use of postoperative physical functioning could emerge as an objective criterion in the calculus of lung cancer surgery decisions. Past reports have compared postoperative with preoperative functional status in lung resection patients and show some decrement,2,3,7 but these decrements are usually small and 1 report describes the return to baseline function within 3 months.7 In addition, a patient with non–small cell lung cancer who does not undergo an operation with curative intent will either face the consequences of progressive disease and/or alternative therapies. Although functional status has been amply described for lung cancer surgery patients relative to populations without cancer or relative to the preoperative state, significant gaps in this literature exist regarding functional status comparisons to similar patients who do not receive surgical care. We, therefore, pose the question, if patients with earlystage non–small cell lung cancer who undergo lung resection surgery are compared with patients with similar disease who do not undergo surgery, will physical functioning significantly differ a year after diagnosis? By comparing functional status to a nonsurgical group with lung cancer, a true decrement in physical functioning would argue that postoperative functional status should be a significant determinant of surgical decisions, whereas similar decrements in functioning between the groups would lead to the opposite conclusion. We recently performed a prospective cohort study examining decisions in lung cancer surgery6; we now report 1-year outcome results to address the described gap regarding functional status change in a nonsurgical lung cancer group compared with a group treated with lung resection for cure.
From the Division of General Internal Medicine and Clinical Epidemiology (SC), North Carolina Area Health Education Centers Program, University of North Carolina School of Medicine, Chapel Hill, North Carolina; Thoracic Oncology Program (PRW), Leo Jenkins Cancer Center, Brody School of Medicine at East Carolina University, Greenville, North Carolina; Department of Biostatistics (LJE), Gillings School of Global Public Health, The University of North Carolina-Chapel Hill, Chapel Hill, North Carolina; Division of Pulmonary and Critical Care Medicine (FRM), The University of South Carolina School of Medicine; and Department of Health Policy and Management (PD-A), Gillings School of Global Public Health, The University of North Carolina-Chapel Hill, Chapel Hill, North Carolina. Submitted April 15, 2014; accepted in revised form August 20, 2014. Supported by a grant from the American Cancer Society, RSGPB-05217-01-CPPB. The authors have no financial or other conflicts of interest to disclose. Correspondence: Samuel Cykert, MD, 145 North Medical Drive, Campus Box #7165, University of North Carolina-Chapel Hill, Chapel Hill, NC 27599-7165 (E-mail: [email protected]).
This study was approved by the Institutional Review Boards of all participating institutions. Full details of recruitment and study protocols are published elsewhere.6 In brief, patients with early-stage non–small cell lung cancer were identified in pulmonary, oncology, thoracic surgery and generalist practices in 5 communities. Patients were included if they were at least 21 years of age, had either biopsy-proven disease or a lung lesion demonstrating a high probability of lung cancer using a Bayesian algorithm and had stage 1 or 2 disease as defined by clinical, radiological and/or mediastinoscopic criteria. Informed consent was obtained. Patients were either referred by the clinical practices or found through a chest computerized tomography review protocol. Patients who were identified using the computerized tomography protocol were
Abstract: Background: Patients with early-stage lung cancer often have comorbid illnesses and fear debility and death when contemplating surgery. However, data that compare physical function of patients who receive surgery with similar patients who do not are sparse. The authors report 1-year outcome results for surgical and nonsurgical patients in a prospective lung cancer cohort to address this gap. Methods: The authors enrolled 386 patients with early-stage lung cancer. A 106-item survey was completed at the time of enrollment including the ShortForm 12 (SF-12) Health Survey to assess functional status. Patients were followed for a year. Chart abstractions were obtained to determine comorbid illnesses and surgical status. Death was ascertained through vital records. The SF-12 was repeated 1 year after the enrollment. Regression models were constructed to identify predictors of 1-year mortality and decline in physical function. Results: Fifty-nine patients (15.3%) died before 1-year follow-up. Mortality in the surgical group was 10.8% compared with 22.8% in the nonsurgical group (P , 0.001). In regression analysis controlling for age and comorbidities, surgical treatment was associated with a reduction in 1-year mortality (odds ratio: 0.5 and 95% confidence interval: 0.3–1.0) but did not worsen physical function relative to the untreated group (average decrease in physical component score of SF-12 5 1.9 for surgery group and 2.5 for no surgery group, P 5 0.66). Conclusions: Functional decline between surgically treated and untreated patients did not differ. This result casts doubt on its value as a treatment determinant. Cancer mortality seems to be a more essential issue in treatment decisions. Key Indexing Terms: Lung cancer; Physical function; Outcomes. [Am J Med Sci 2015;349(1):61–66.]
P
The American Journal of the Medical Sciences
METHODS
Volume 349, Number 1, January 2015
61
Cykert et al
contacted through the ordering practice unless the ordering physician of the scan was based on the emergency department and no documented follow-up occurred. In these cases, the research associate was allowed to contact the patient directly and arrange further care. Pneumonectomies were infrequently performed at the participating study sites (N 5 9). Given our inability to perform meaningful statistical comparisons for pneumonectomy patients, they were excluded from analysis. All participants were administered a 106-item survey at the time of enrollment. Demographic data were obtained while other survey items explored perceptions of patient-physician communication, attitudes about lung cancer and surgery, religiosity, past health care experiences and access to a regular source of care. Race and ethnicity were self-reported. Physical function was ascertained using the Short-Form 12 (SF-12) Health Survey.8 Chart reviews were performed starting 4 months after enrollment. The chart abstraction included lung surgery status, date of surgery if it occurred, preoperative stage, medical comorbidities and preoperative pulmonary function tests. Physical function was reassessed using the SF-12 by telephone 1 year after enrollment. For patients who had died, death and date of death were verified by searching state vital records. The primary outcomes for this analysis are mortality and physical function as assessed by the SF-12 physical component score (PCS) 1 year after lung cancer diagnosis. Significant physical deterioration was defined as a drop of 5 or greater in the PCS 12 months after diagnosis given that a 4.5-point gain in SF-36 PCS was linked to improved (66 m) 6-minute walk distance in pulmonary rehabilitation research9 and the SF-12 PCS is highly correlated with the SF-36.8,10,11 The physical function outcome is reported for the group of 1-year survivors who completed both the initial and 12-month phone assessment. To account for death as a condition of physical deterioration, we also examined a combined outcome of 5-point decrement in PCS and death reflecting a state of global worsening 1 year after diagnosis inclusive of all enrolled patients. Descriptive statistics including selected bivariate associations were computed for demographic variables, surgical rate, religiosity, a sum of comorbid conditions relevant to lung resection surgery (chronic obstructive lung disease, congestive heart failure, renal insufficiency, poor mobility, cerebrovascular disease, obstructive sleep apnea, oxygen dependence), a history of coronary artery disease, known diabetes, the mental component score (MCS) of the SF-12 and the 3 outcomes noted above. All outcome variables were dichotomous, so a generalized linear model for clustered data with multiple predictors was used to assess the simultaneous effect of predictors on outcome.12 Given that the major goal of this analysis was to gauge the effect of surgery on 1-year mortality and physical functioning, receipt of surgery was a variable in all models. Demographic variables were incorporated into all models. Because of the significance of the presence of 2 or more comorbid illnesses and religiosity in our previous work,6 these measures were also included in the regressions. Coronary artery disease and diabetes were prevalent but did not impact mortality or physical function scores in either bivariate or regression analyses, and so these variables were dropped from the final models. Therefore, for the purpose of this analysis, exactly as per our previous work, significant comorbidities included chronic obstructive lung disease, renal insufficiency, poor mobility, congestive heart failure, cerebrovascular disease, obstructive sleep apnea and oxygen dependence.6 The modeling approach fitted a maximum model using main effects only; however, we did assess the effect of several interactions including race and
62
comorbid illnesses, comorbid illnesses and surgery and age and surgery without detecting additional significance beyond the main effects. All statistical analyses were performed using SAS software version 9.3 (SAS Institute Inc, Cary, NC). A priori statistical significance for independent variables and the primary outcomes was set at #0.05 based on a 2-sided test for bivariate and multivariate associations.
RESULTS As previously described, 386 patients remained eligible in this early-stage lung cancer prospective cohort.6 All patients with absolute contraindications to surgery as defined by pulmonary function results were excluded. Fifty-nine patients (15.3%) died before 1-year follow-up. Thirty-seven patients either refused or were unable to complete the SF-12 survey 1 year after diagnosis. Therefore, 290 baseline and 1-year surveys were available for paired analyses and 349 were available to assess the combined outcome of significant physical function decline or death. Demographic characteristics, religiosity, comorbid conditions and baseline SF-12 scores plus surgical rates according to these characteristics are summarized in Table 1. Gender, educational attainment and marital status are similar to previous reports.4,13 Because the cited reports are based on combined Surveillance, Epidemiology and End Results-Medicare data, the age of our population is marginally younger. We oversampled African Americans because of initial sample size calculations that allowed enough statistical power to make valid comparisons according to race.6 Note that 13% of participants had 2 or more of the designated comorbidities. The prevalence of each individual comorbidity included in the summary variable was as follows: Chronic Obstructive Pulmonary Disease (COPD): 40%; congestive heart failure: 10%; history of cerebrovascular accident (CVA): 6.2%; renal failure: 5.5%; poor mobility (walking limited to a room): 4.7%; oxygen dependence: 2.6%; and obstructive sleep apnea: 1.8%. Bivariate analyses for all 3 outcomes are shown in Table 2. Age above the median (age, 66 years) is strongly associated with death (1-year survivors averaged 65.8 years of age at enrollment compared with 69.5 years in the group who died [P 5 0.01]), significant physical decline and the combined 1-year outcome of death plus physical decline. Patients who placed in the lowest quartile MCS on the baseline SF-12 experienced twice the mortality within a year compared with those with higher MCS scores. There is no relationship between MCS and physical decline or the composite outcome. The age-old conundrum between surgical benefit and the risk of comorbid conditions remained in the bivariate results. One-year mortality in the surgical group was 10.8% compared with 22.8% in the nonsurgical group (P 5 0.002), but patients with 2 or more significant comorbidities at diagnosis had a 27.5% 1-year mortality compared with 13.5% in the group with ,2 (P 5 0.01). However, when comorbidity was stratified by surgical status, we found that the 2 or more comorbidity surgical group had a 1-year mortality rate of 19% compared with 10% for the group with ,2 (P 5 NS); in the no surgery group, these 1-year mortality rates were 31% and 20%, respectively (P 5 NS). There was no difference in 1-year mortality according to race (African Americans 5 15.0 and whites 5 15.4; P 5 0.9), but African Americans who died were 4 years younger than white patients (66.6 versus 70.6 years; P 5 0.05). The physical function analysis was limited to the 290 patients who survived 1 year and answered the baseline and Volume 349, Number 1, January 2015
Physical Decline and Lung Cancer Care
TABLE 1. Baseline characteristics of early-stage lung cancer cohort combined and by surgical status Patients (N 5 386), Characteristic n (%) P Age, above median, yr Surgical rate, % .66 #66 Male gender Surgical rate, % Male Female Black race (All) Surgical rate, % Black White Education .high school Surgical rate, % .High school #High school Household income, lowest tertile Surgical rate, % Upper tertile Lowest tertile Married (All) Surgical rate, % Married Not married No health insurance Surgical rate, % No insurance Insurance SF-12 MCS, lowest quartile Surgical rate, % Upper 3 quartiles Lowest quartile SF-12 PCS, lowest quartile Surgical rate, % Upper 3 quartiles Lowest quartile High religiosityb (All) Surgical rate, % High religiosity Less religiosity Greater than 2 severe comorbidities Surgical rate, % .2 Comorbidities ,2 Comorbidities History CAD Surgical rate, % CAD No CAD History of DM Surgical rate, % DM No DM
190 (49) 0.24 59 65 215 (56) 0.37 60 65 113 (29) 0.048 55a 66 135 (35) 0.68 64 62 125 (32) 0.02 66a 54 246 (64) 0.16 65 58 29 (8) 0.97 62 63 95 (25) 0.04 65a 54 90 (23)
1-year postdiagnosis SF-12. The average drop in the SF-12 PCS at 1 year was 1.9 for the surgical group and 2.5 in the nonsurgical group (P 5 0.66). Thirty-eight percent of lung cancer patients experienced a significant (5 points or more) fall in their PCS at 1 year. Notably, 37.4% of surgical patients experienced this degree of physical deterioration compared with 40.2% in the nonsurgical group again confirming no difference in this outcome (P 5 0.64). Only 45% of patients treated with surgery demonstrated the combined deterioration outcome compared with 55% in the no surgery group (P 5 0.06) with the difference in 1-year mortality accounting for most of the difference in favor of surgery. Male gender and advanced age were the only statistically significant bivariate associations with significant deterioration of physical function and with the combined outcome for deterioration at 1 year. Educational status, lowest income tertile, marital status, religiosity, history of coronary artery disease and diagnosis of diabetes were not associated with any increase in the adverse outcomes. The regression results are shown in Table 3. Controlling for comorbid illness, patients treated with surgery were half as likely to die within a year compared with the nonsurgical group. Conversely, patients with 2 or more significant comorbid illnesses were more likely to die (odds ratio: 2.2 and 95% confidence interval: 1.01–4.8) whether operated on or not. Age remained a significant predictor as those older than 66 years were about twice as likely to die. Preoperative physical functioning as represented by the PCS was not associated with 1-year mortality, but the lowest quartile of MCS continued to be strongly associated with 1-year mortality independent of age, surgical treatment and comorbid illnesses. Regarding significant physical decline and the composite deterioration outcome, male gender and age older than 66 years continued to be associated with 1-year physical decline in regression models whether patients were treated with surgery or not. Surgery was not associated with loss of physical function using either the 5 or more point drop in PCS or the combined variable. Sensitivity analyses were performed using PCS decline at 1 year of 2 or more points, 10 or more points and 15 or more points as the definition for clinically important functional declines. Lung cancer surgery was not predictive of significant functional loss compared with no surgery patients regardless of the designated cut point.
0.01 66a 51 162 (42) ,0.001 52a 70 51 (13) ,0.001 31a 68 100 (26) 0.74 64 62 86 (22)
Difference in surgical status statistically significant (P , 0.05). As determined by agreement with the statement, faith alone can cure disease. CAD, coronary artery disease; DM, diabetes mellitus. a
0.97 63 63
DISCUSSION Physical decline related to lung cancer surgery is an understandable worry for patients facing this decision given the prevalence of smoking-related illnesses in this group. In our initial report, patients who believed that lung cancer surgery would worsen their quality of life in a year were 1 quarter as likely to proceed to surgery.6 Although we know that the extent of lung resection relates to postoperative decline in physical function and that function is generally worse than that of agematched patients,2,14–16 few data are available to compare functional status in similar lung cancer patients who were not surgically treated. The good news for these patients is that physical function 1 year after diagnosis decreases similarly whether a patient has surgery or not. More germane to making
b
Copyright © 2014 by the Southern Society for Clinical Investigation. Unauthorized reproduction of this article is prohibited.
63
Cykert et al
TABLE 2. Bivariate analysis of patient outcomes 1 year after diagnosis of early-stage non–small cell lung cancer according to demographics, functional status, surgical status and comorbid conditions Death Physical function loss of 5 points Combined outcome (death or Characteristic (%) P or more on SF-12 (%) P 5-point loss of physical function) (%) P Age, yr .66 #66 Gender Male Female Race Black White Education HS or less More than HS Income Lowest tertile Top 2 tertiles Married Yes No Health Insurance Yes No MCS Lowest quartile Top 3 quartiles PCS Lowest quartile Top 3 quartiles Religiosity High Low Surgery Yes No CAD Yes No Diabetes Yes No Comorbid conditions $2 ,2
0.01a 20.0 10.7
48.2 29.9
59.5 38.2 0.05a
0.37 16.7 13.5
43.4 32.3 0.93
15.0 15.4
0.17
0.93
0.74
0.74
0.27
0.18
0.68
0.44
0.30 46.7 52.4
0.21 39.4 26.1
0.005a 24.2 12.4
0.34 44.9 50.4
37.5 40.0
15.7 10.3
0.77 49.6 48.0
33.7 40.4
13.4 18.6
0.26 43.9 50.6
37.4 39.4
14.4 15.7
0.03a 53.9 42.3
32.1 40.8
15.2 15.6
,0.001a
0.001a
0.14 49.9 34.6
0.73
0.31
38.9 36.5
47.2 53.5
b
b
0.45 17.8 14.5 0.35 17.3 13.8
0.13 43.4 34.7
0.002a 10.8 22.8
0.65 37.4 40.2
0.91 15.0 15.4
0.13
0.54
0.17 55.2 46.6
0.91 37.7 38.5
0.01a 27.5 13.5
0.06 44.8 55.4
45.9 35.9
17.4 14.7
0.09 54.0 44.8
0.80 50.0 48.4
0.52 33.3 39.2
0.52 53.2 48.2
a
Statistically significant. PCS was not used as an independent variable for analyses related to physical functioning. CAD, coronary artery disease. b
surgical decisions is that lung cancer surgery resulted in nearly a 50% reduction in 1-year mortality while controlling for age and comorbid illness. The tempering statistic, however, is that for the 13% of patients who had 2 or more defined comorbidities, the odds ratio for death at 1 year was 2.2 regardless of surgical status. When examining unadjusted data for patients who underwent surgery, we found that the comorbidity group
64
had a 1-year mortality rate of 19%. Although this death rate was 9 percentage points higher than those who had surgery without comorbid burden, it was a 12-point absolute decrease compared with the 31% mortality experienced by the group with similar comorbidities without surgical intervention. Clearly, early-stage lung cancer patients with these comorbid illnesses have a higher 1-year mortality rate whether treated surgically or not, but Volume 349, Number 1, January 2015
Physical Decline and Lung Cancer Care
TABLE 3. The association of patient characteristics, functional status and lung cancer surgery with outcomes 1 year after diagnosis of stage 1 or 2, non–small cell lung cancer as determined by regression analyses Odds ratio (95% Odds ratio (95% CI) significant Odds ratio (95% CI) combined outcome Characteristic CI) death loss of physical function (death or loss of physical function) Age .66 yr Male gender Black race Education .high school Lowest income tertile Married Has health insurance Upper 3 quartiles on mental component score of SF-12 Upper 3 quartiles of physical component score of SF-12 High religiosityc Had lung cancer surgery Two or more comorbid conditions
2.3 1.5 1.0 1.1 0.94 0.62 1.4 0.40
(1.2–4.3)a (0.83–2.8) (0.49–2.0) (0.59–2.1) (0.48–1.8) (0.33–1.2) (0.38–5.2) (0.21–0.77)a
2.0 1.7 0.67 1.0 0.78 0.71 1.6 0.95
(1.2–3.3)a (1.05–2.9)a (0.36–1.2) (0.61–1.8) (0.45–1.4) (0.41–1.2) (0.56–4.3) (0.51–1.8)
2.2 1.8 0.78 1.0 0.85 0.67 1.6 0.73
(1.4–3.5)a (1.2–2.9)a (0.45–1.3) (0.64–1.7) (0.51–1.4) (0.41–1.1) (0.66–4.0) (0.43–1.2)
1.2 (0.57–2.5)
b
b
0.96 (0.52–1.8) 0.54 (0.29–1.0)a 2.2 (1.01–4.8)a
1.4 (0.83–2.3) 0.87 (0.51–1.5) 0.77 (0.34–1.7)
1.2 (0.78–2.0) 0.72 (0.44–1.1) 1.1 (0.59–2.2)
a
Statistically significant. PCS was not used as an independent variable for analyses related to physical functioning. As determined by agreement with the statement, faith alone can cure disease. CI, confidence interval.
b c
surgery seems to attenuate the death rate in a manner similar to Battafarano’s report.17 A survival analysis reported by Bach demonstrated a doubling of mortality in the nonsurgical group within 1 year similar to our results.4 In addition, recent cohort studies consisting of thousands of early-stage lung cancer patients examined geographic variation in surgical rates and demonstrated a strong relationship between higher surgical rates and survival, again within 1 year.5,18,19 Gray et al5 specifically analyzed a subset of patients who were over the age of 66 with chronic obstructive lung disease. They found a 7% decrease in absolute mortality at 1 year (20% versus 27%) for these sicker patients in high-surgery regions (lung resection rates .79%) compared with low-surgery regions (lung resection rates ,63%) and demonstrated that the excess mortality was lung cancer specific. Although the dilemma of which lung cancer patients are too sick for surgery remains, it seems that more patients should be given the benefit of the doubt when absolute contraindications are not in play. As noted, patients with 2 or more comorbidities have higher death rates, but 1-year death rates are even worse if they do not get an operation. The relationship of lower SF-12 MCS and mortality requires further thought. Reductions in MCS have been shown to correlate with depression,8,20,21 and depression has been associated with worse outcomes in patients with chronic obstructive lung disease including increased mortality.22 Depression has also been associated with increased mortality for cancer in general23 and lung cancer patients specifically.24 Although treatment has not been shown to improve survival in patients with metastatic breast cancer,25 a depression treatment program instituted for older adults in primary care resulted in lower 5-year mortality for patients diagnosed with cancer.26 Given the high mortality rates in patients with early-stage lung cancer, any incremental benefit in survival, especially for those who go through the rigor of surgery, would be valuable. Therefore, an in-depth examination of the diagnosis and treatment of depression in early-stage lung cancer is warranted.
Several limitations of this study need to be considered. Although our initial report demonstrated an 11% lower surgical rate for African American patients, race was not a predictor of 1-year mortality. This finding might be explained by the fact that surgery itself was the main determinant of survival, overriding any racial effect or the fact that we were simply underpowered to find mortality differences by race. Notably, the 1-year mortality rate was 15% for both racial groups. However, again, African Americans who died were 4 years younger suggesting that the lower operative rate contributed to worse survival as has been described by Bach et al.4 Another limitation is that we did not have pneumonectomy patients in this cohort. Functional impairment and survival are worse after pneumonectomy,15,27 so we cannot apply our findings to those patients with lung cancer who need this more extensive procedure. Finally, comorbidity counts or scores do not necessarily account for the severity of illness associated with a single specific comorbidity. However, use of these scores has been demonstrated to be reliable and valid predictors of outcomes for clinical research generally28 and lung cancer specifically.29 A recent consensus report emphasized the need to incorporate physical functioning into the decision calculus surrounding surgery for high-risk patients with stage I non–small cell disease.30 Although functional status and the decrement attributable to lung resection are important, it should not be considered without factoring in the impact of progressive cancer or radiation therapy. This report highlights the fact that functional loss in the nonoperated cohort is comparable to lung resection. Mortality remains important despite the risk of surgery the survival curves diverge in favor of surgery as early as 1 year after diagnosis.4,5 For individuals without absolute contraindications and less than 2 significant comorbid illnesses, these data suggest that lung cancer surgery should be strongly considered; even within the comorbid illness group, surgical intervention still warrants careful consideration. This tenet holds particularly true for African Americans who are staged invasively less often than similar white patients,13 receive recommendations for surgical
Copyright © 2014 by the Southern Society for Clinical Investigation. Unauthorized reproduction of this article is prohibited.
65
Cykert et al
care less frequently in the context of the absence of absolute contraindications6,13 and suffer lung cancer death at a younger age even when presenting with early-stage disease.4
13. Lathan CS, Neville BA, Earle CC. The effect of race on invasive staging and surgery in non-small-cell lung cancer. J Clin Oncol 2006; 24:413–18.
CONCLUSIONS
14. Bryant AS, Cerfolio RJ, Minnich DJ. Survival and quality of life at least 1 year after pneumonectomy. J Thorac Cardiovasc Surg 2012;144: 1139–45.
The lack of difference in functional decline between surgically treated and untreated patients casts doubt on its value as a treatment determinant at least in the setting of lobectomy and smaller lung resection procedures. Cancer mortality should remain an important consideration in surgical decisions for these patients. ACKNOWLEDGMENTS The Cecil G. Sheps Center for Health Services Research served as the multidisciplinary collaborative center for this work. The authors wish to acknowledge Dr. Richard Feins for his contributions to designing patient recruitment protocols and his insight regarding lung cancer surgery decisions and the role of comorbid illness and communication factors in the decisionmaking process. The authors thank the American Cancer Society (ACS) for their sponsorship of this project. Note that the ACS did not influence the performance of the study and did not play a role in the design of the research or writing of the article. REFERENCES 1. Cykert S, Kissling G, Hansen CJ. Patient preferences regarding possible outcomes of lung resection what outcomes should preoperative evaluations target? Chest 2000;117:1551–9. 2. Handy JR, Asaph JW, Skokan L, et al. What happens to patients undergoing lung cancer surgery?*: outcomes and quality of life before and after surgery. Chest 2002;122:21–30. 3. Myrdal G, Valtysdottir S, Lambe M, et al. Quality of life following lung cancer surgery. Thorax 2003;58:194–7. 4. Bach PB, Cramer LD, Warren JL, et al. Racial differences in the treatment of early-stage lung cancer. N Engl J Med 1999;341:1198–205. 5. Gray SW, Landrum MB, Lamont EB, et al. Improved outcomes associated with higher surgery rates for older patients with early stage nonsmall cell lung cancer. Cancer 2012;118:1404–11. 6. Cykert S, Dilworth-Anderson P, Monroe MH, et al. Factors associated with decisions to undergo surgery among patients with newly diagnosed early-stage lung cancer. JAMA 2010;303:2368–76. 7. Brunelli A, Socci L, Refai M, et al. Quality of life before and after major lung resection for lung cancer: a prospective follow-up analysis. Commentary. Ann Thorac Surg. 2007;84:410–16. 8. Ware J Jr, Kosinski M, Keller SD. A 12-item short-form health survey: construction of scales and preliminary tests of reliability and validity. Med Care 1996;34:220–33. 9. Benzo R, Flume PA, Turner D, et al. Effect of pulmonary rehabilitation on quality of life in patients with COPD: the use of SF-36 summary scores as outcomes measures. J Cardiopulm Rehabil 2000;20:231–4. 10. Ware JE Jr, Gandek B, Kosinski M, et al. The equivalence of SF-36 summary health scores estimated using standard and country-specific algorithms in 10 countries: results from the IQOLA project. J Clin Epidemiol 1998;51:1167–70. 11. Pickard AS, Johnson JA, Penn A, et al. Replicability of SF-36 summary scores by the SF-12 in stroke patients. Stroke 1999;30:1213–17. 12. Diggle P, Liang L, Zeger SL, ed. The analysis of longitudinal data. New York (NY): Oxford University Press; 1994.
66
15. Schulte T, Schniewind B, Dohrmann P, et al. The extent of lung parenchyma resection significantly impacts long-term quality of life in patients with non-small cell lung cancer. Chest 2009;135:322–9. 16. Möller A, Sartipy U. Predictors of postoperative quality of life after surgery for lung cancer. J Thorac Oncol 2012;7:406. 17. Battafarano RJ, Piccirillo JF, Meyers BF, et al. Impact of comorbidity on survival after surgical resection in patients with stage I non-small cell lung cancer. J Thorac Cardiovasc Surg 2002;123:280–7. 18. Riaz SP, Lüchtenborg M, Jack RH, et al. Variation in surgical resection for lung cancer in relation to survival: population-based study in England 2004–2006. Eur J Cancer 2012;48:54–60. 19. Wouters MWJM, Siesling S, Jansen-Landheer ML, et al. Variation in treatment and outcome in patients with non-small cell lung cancer by region, hospital type and volume in the Netherlands. Eur J Surg Oncol 2010;36:S83–92. 20. Lipton R, Hamelsky S, Kolodner K, et al. Migraine, quality of life, and depression: a population-based case–control study. Neurology 2000;55:629–35. 21. McHorney CA, Ware JE Jr, Raczek AE. The MOS 36-Item ShortForm Health Survey (SF-36): II. Psychometric and clinical tests of validity in measuring physical and mental health constructs. Med Care 1993;31:247–63. 22. Ng TP, Niti M, Tan WC, et al. Depressive symptoms and chronic obstructive pulmonary disease effect on mortality, hospital readmission, symptom burden, functional status, and quality of life. Arch Intern Med 2007;167:60–7. 23. Stommel M, Given BA, Given CW. Depression and functional status as predictors of death among cancer patients. Cancer 2002;94: 2719–27. 24. Faller H, Bulzebruck H, Drings P, et al. Coping, distress, and survival among patients with lung cancer. Arch Gen Psychiatry 1999;56:756. 25. Goodwin PJ, Leszcz M, Ennis M, et al. The effect of group psychosocial support on survival in metastatic breast cancer. N Engl J Med 2001;345:1719–26. 26. Gallo JJ, Bogner HR, Morales KH, et al. The effect of a primary care practice–based depression intervention on mortality in older adults: a randomized trial. Ann Intern Med 2007;146:689. 27. Chambers A, Routledge T, Pilling J, et al. In elderly patients with lung cancer is resection justified in terms of morbidity, mortality and residual quality of life? Interact Cardiovasc Thorac Surg 2010;10:1015–21. 28. de Groot V, Beckerman H, Lankhorst GJ, et al. How to measure comorbidity: a critical review of available methods. J Clin Epidemiol 2003;56:221–9. 29. Birim Ö, Kappetein AP, Bogers AJ. Charlson comorbidity index as a predictor of long-term outcome after surgery for nonsmall cell lung cancer. Eur J Cardiothorac Surg 2005;28:759–62. 30. Donington J, Ferguson M, Mazzone P, et al. American College of Chest Physicians and Society of Thoracic Surgeons consensus statement for evaluation and management for high-risk patients with stage I nonsmall cell lung cancer. Consensus statement on lung cancer in high-risk patients. Chest 2012;142:1620–35.
Volume 349, Number 1, January 2015
