- Email: [email protected]
Improving the Lung Cancer Resection Rate in the US Department of Veterans Affairs Health System
o riginal contribution Improving the Lung Cancer Resection Rate in the US Department of Veterans Affairs Health System Mark Thomas Dransfield, Brion Jacob Lock, Robert I. Garver, Jr Abstract BACKGROUND: The optimal treatment for non–small-cell lung cancer (NSCLC) is surgical resection; however, most patients are ineligible because of advanced disease. Although resection rates of 25% have been reported nationally, rates in the Veterans Affairs (VA) system appear lower, perhaps because of limited access to specialized care. We hypothesized that, since the introduction of a specialized Lung Mass Clinic in 1999, the resection rate at the Birmingham VA Medical Center would be comparable with US benchmarks. We also sought to identify the medical and nonmedical factors that influenced the use of surgery. PATIENTS AND METHODS: We reviewed the electronic medical records of all veterans seen in the Lung Mass Clinic from 1999 to 2003 and identified patients with NSCLC. Demographics, comorbidities, diagnostic methods, times to diagnosis/resection, and postoperative survival were recorded. Reasons for nonresection were documented and tabulated, and differences between the resected and nonresected subgroups were examined. RESULTS: One hundred fifty-six patients with NSCLC were identified, and 31 (20%) underwent resection. There were no differences in age, ethnicity, or sex between those undergoing resection and those denied surgery. Patients who underwent resection were less likely to have chronic obstructive pulmonary disease and had better pulmonary function. Eighty-four percent of those who did not undergo resection had advanced disease, poor pulmonary function, or had refused therapy. Although the median time to resection was longer than expected (104 days), overall survival was comparable with other reports (65% at 3 years). CONCLUSION: Since the inception of the Lung Mass Clinic, the resection rate at Birmingham VA Medical Center has improved. The primary limitation to resection was late presentation and not preoperative delays. Clinical Lung Cancer, Vol. 7, No. 4, 268-272, 2006
Key words: Diagnosis, Non–small-cell lung cancer, Practice organization, Thoracic surgery
Introduction
such as ethnicity and socioeconomic status might influence the use of surgical treatments for lung cancer.4 As a result of these limitations, reported resection rates for lung cancer are uniformly < 30%.3,5-9 Unfortunately, comparing these reports is problematic, because investigators have employed a variety of methods for data collection, and although some have reported resection rates for all lung cancers, others have restricted their analyses to non–small-cell lung cancer (NSCLC). Despite these limitations, there does appear to be significant discrepancies between resection rates in the United Kingdom and those observed in Europe and the United States. In a review of 3855 cases of lung cancer that were reported to the Scottish Cancer Registry in 1995, Gregor et al found that only 10.7% underwent surgical resection.7 A 10% resection rate has also been reported among all UK cardiothoracic surgeons, and rates as low as 1% have been observed in some regions without a respiratory physician.9 Even after accounting for the small-cell tumors that were included in these reports, UK resection rates for NSCLC still fall far short of those in the United States, where a survey of
Lung cancer is the leading cause of cancer-related death in the United States and United Kingdom, and 5-year survival remains < 15%.1,2 Multiple factors contribute to this poor prognosis, including the absence of an effective screening test and the fact that most patients present with late-stage disease. As a result, less than half of patients are eligible for surgical resection, the only modality that offers a legitimate hope of a cure.3 Even among those with early-stage tumors, only a fraction undergo surgery, because many have impaired lung function or comorbid illnesses that preclude an operation. Unfortunately, there is also evidence that nonmedical factors Division of Pulmonary, Allergy, and Critical Care Medicine, Birmingham Veterans Affairs Medical Center and University of Alabama at Birmingham Submitted: Dec 20, 2005; Revised: Jan 16, 2006; Accepted: Jan 19, 2006 Address for correspondence: Mark Thomas Dransfield, MD, Division of Pulmonary, Allergy, and Critical Care Medicine, Birmingham Veterans Affairs Medical Center and University of Alabama at Birmingham 215 THT, 1900 University Blvd, Birmingham, AL 35294 Fax: 205-934-6229; email: [email protected]
Electronic forwarding or copying is a violation of US and International Copyright Laws. Authorization to photocopy items for internal or personal use, or the internal or personal use of specific clients, is granted by CIG Media Group, LP, ISSN #1525-7304, provided the appropriate fee is paid directly to Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923 USA 978-750-8400.
268
Clinical Lung Cancer January 2006
hospitals with approved cancer programs found that 28% of NSCLC were resected.5 A review of cancer cases reported to the Rotterdam Cancer Registry demostrated a comparable NSCLC resection rate of 24%.6 There are many possible explanations for these international differences, though it has been suggested that the discrepancy is partly caused by poor access to specialized care in the United Kingdom and inordinate preoperative delays.2 Indeed, 1 study from the United Kingdom found a median time of 109 days from initial presentation to surgery, a delay well beyond the 4- to 8-week period recommended by the American College of Chest Physicians and the British Thoracic Society.10-12 Although there is no clear evidence to suggest that shorter preoperative delays result in better outcomes, 13 recent efforts to streamline the process of care with “fast access investigation units” have shown shorter waiting times and higher rates of surgical resection.8 Although US resection rates of 25%-28% have been reported,3,5 there are no published data about the rate of resection among patients with lung cancer cared for by the Department of Veterans Affairs (VA). The VA system is a network of 162 hospitals and almost 1000 supporting primary care clinics, which in 2002 provided health care for 1.6 million veterans of the US military services.14 The VA system is structured in a manner similar to the UK’s National Health Service in which specialty care, including lung cancer diagnosis and management, is provided by referral hospitals. In the late 1990s, there was concern that lung cancer care among patients at the Birmingham VA Medical Center (BVAMC) was inefficient and that opportunities for resection might be missed. Indeed, a recent review of the BVAMC Cancer Registry confirmed that only 17 of 135 patients (12.6%) diagnosed with NSCLC in 1993 were treated with curative surgical resection (BVAMC Cancer Registry Database). This is in keeping with data from the larger regional Veterans Integrated Service Network (including other facilities in Alabama, Georgia, and South Carolina), which showed that only 62 of 404 patients (15.3%) diagnosed with NSCLC in 1998 underwent wedge resection, segmentectomy, single or bilobectomy, or pneumonectomy for stage I-III disease.15 Because of these concerns about inefficiency, a specialized Lung Mass Clinic was established to provide rapid evaluation and triage of possible lung cancer cases. We hypothesized that, since the inception of the Lung Mass Clinic in 1999, rates of surgical resection at our facility would be comparable with those reported nationally.
Patients and Methods The study was approved by the Birmingham VA Institutional Review Board. We performed a retrospective review of the electronic medical records of all veterans who were referred to and seen in the Lung Mass Clinic during the first 4 years of its operation (October 14, 1999 to October 14, 2003). We identified patients who subsequently received a diagnosis of thoracic malignancy. Patients with a history of lung cancer were excluded from the study.
Patients who were found to have a new diagnosis of NSCLC were identified and divided into resected and nonresected subgroups. Baseline demographic information including age, sex, and ethnicity were recorded as were major comorbid medical illnesses including coronary artery disease, congestive heart failure, chronic obstructive pulmonary disease (COPD), cerebrovascular disease, chronic renal insufficiency, and diabetes mellitus. A comorbid illness was considered present if it was documented by an International Classification of Diseases–9 code or described in clinician notes. Pulmonary function data were abstracted from the medical record if they were obtained within 3 months of the Lung Mass Clinic appointment. Additional information obtained included the date and method of lung cancer diagnosis (fiberoptic bronchoscopy, transthoracic needle aspiration, or surgery) and the dates and results of subsequent staging and preoperative procedures, including positron emission tomography (PET), noninvasive cardiac stress testing, and mediastinoscopy. The times from initial Lung Mass Clinic evaluation to diagnosis and to resection (if applicable) were determined. For patients undergoing resection, final pathologic staging was recorded, and among those denied surgery, reasons for nonresection were documented and tabulated. If the explanation for denying resection was not clearly stated in the medical record, it was recorded as unknown. Postoperative survival was also determined. A subject was considered alive if there had been computer-documented contact with the VA system (eg, clinic visit, hospitalization, or medication refill) within 45 days of our review and if there was no computer entry documenting the patient’s death. Differences in baseline data between the resected and nonresected groups were compared with χ2, Fisher exact, or Student t tests as appropriate.
Results Four hundred eighty-seven patients were seen during the study period. Of these, 180 (37%) were found to have thoracic malignancies. This represents approximately 50% of the total number of thoracic malignancies reported to our cancer registry over the same period. The majority of the other cases of lung cancer were referred to our facility for treatment after having outside histologic confirmation of the disease. Therefore, these patients were not seen in the Lung Mass Clinic and were not included in the analysis. Non–small-cell lung cancer was the most common histologic subtype, accounting for 156 (87%) of the total. Other tumor types identified included small-cell lung cancer (n = 18), lymphoma (n = 3), carcinoid (n = 1), melanoma (n = 1), and unknown (n = 1). Of the 156 cases of NSCLC, 31 (20%) underwent surgical resection (29 lobectomies, 1 bilobectomy, and 1 pneumonectomy). Table 1 shows the final pathologic stage distribution of the resected tumors. The stage IIIA tumors were caused by unexpected N2 disease (n = 2) and chest wall invasion with N1 disease (n = 1). There were no significant differences in age, sex, or ethnicity between the resected and nonresected subgroups (Table 2). There were also no significant differences in the incidence of
Clinical Lung Cancer January 2006
269
Lung Cancer Resection Table 1 Final Pathologic Stage Distribution of Resected Tumors Stage
Number (%)
IA IB IIA IIB IIIA
Patient Characteristic
10 (32) 11 (35) 3 (10) 4 (13) 3 (10)
Table 3 Diagnostic Methods and Times to Diagnosis and Resection Resected (n = 31)
Method of Diagnosis Fiberoptic bronchoscopy Transthoracic needle aspiration Surgery PET Scans Performed Noninvasive Cardiac Stress Testing Median Time to Diagnosis (Days) Median Time to Resection (Days) Values in parentheses are percentages.
270
Clinical Lung Cancer January 2006
Nonresected (n = 125)
Resected Nonresected P Value (n = 31) (n = 125) 64.8 ± 8.4
66.8 ± 9
0.28
Ethnicity (% White)
84
74
0.35
Sex (% Male)
97
99
0.7
10 (32) 2 (6) 5 (16) 3 (10) 2 (6) 2 (6) 15 (48) 8 (26) 67 ± 17 71 ± 15
37 (30) 12 (10) 53 (42) 15 (12) 6 (5) 17 (14) 85 (68) 40 (32) 55 ± 17 62 ± 17
1 0.85 0.01 0.96 0.98 0.36 0.07 0.65 0.002 0.02
Age (Years)
major medical comorbidities between the 2 groups, except those resected had a lower incidence of COPD and better pulmonary function than the nonresected group. Patients who underwent resection were most often diagnosed at the time of surgery (61% of cases), and those whose disease was amenable to bronchoscopic diagnosis were usually unresectable (91% of cases; Table 3). Surgery was rarely required for the diagnosis of unresectable disease (7% of cases). Positron emission tomography scans and noninvasive cardiac testing were performed more often in patients who eventually underwent resection. The median time to diagnosis was much shorter among patients who were not resected. The median time to resection from the initial Lung Mass Clinic appointment was 104 days. Table 4 shows the primary reason for nonresection among the 125 patients who were denied surgery. Fifty-three patients had evidence of advanced disease (stage IIIA/IIIB or stage IV) by imaging studies. Of these, 30 (56%) underwent PET scanning. Tissue confirmation of stage IIIA/IIIB or stage IV disease was obtained in an additional 41 cases (33%). One hundred five (84%) of those not resected had evidence of advanced disease by biopsy or imaging studies, prohibitive pulmonary function, or refused therapy. In a minority of cases (9%), a clear reason for denying resection could not be found. With median follow-up approaching 3 years, overall survival among those undergoing resection is 20 of 31 (65%) compared with only 38 of 125 (30%) among those not resected. Figure 1 shows the Kaplan-Meier survival curve for patients who underwent surgical resection. Four deaths occurred in the first 30 days after surgery. Three of these were the result of postoperative pneumonia and its complications, and the cause of death for 1 patient who had been discharged is not known. Of the 11
Methods and Times
Table 2 Baseline Demographics, Comorbid Illness, and Pulmonary Function
Comorbid Illness (%) Coronary artery disease Congestive heart failure COPD Cerebrovascular disease Chronic renal insufficiency Diabetes mellitus Any ≥ 2 Illnesses FEV1 (Mean % Predicted) FVC (Mean % Predicted)
Mean values are presented with standard deviations. Abbreviation: FVC = forced vital capacity
deaths in the series, the cause of death was pneumonia in 4, unknown in 4, and recurrent disease in 3 patients.
Discussion Our study confirms the poor prognosis for most patients with lung cancer. As expected, the vast majority of patients presented with late-stage disease that was not amenable to surgical resection. Survival among those not resected was only 30% and will certainly decline further as follow-up is extended. Despite this unfortunate statistic, the fraction of patients who were treated surgically (20%) is higher than reported in most studies done in the United Kingdom and approaches the available estimates for the United States and Europe.3,5-9 This resection rate also represents a significant improvement compared with data from our hospital and region from the 1990s.15 It should also be noted that our resection rate is likely an underestimation, because 4 patients who were eligible for surgery chose to receive additional care in the civilian health-care system. In addition, we would argue that the frequent use of PET scanning in our population lowered the rate of resection. Although PET scanning has been shown to upstage and down-
P Value Table 4 Primary Reasons for Nonresection
10 (32)
96 (77)
< 0.001
2 (6)
19 (15)
0.24
19 (61) 29 (94)
9 (7) 41 (33)
< 0.001 < 0.001
18 (58)
26 (21)
< 0.001
70
8
< 0.001
104
–
–
Reason (n = 125)
Number
Stage IIIA/IIIB or IV Disease by Imaging Stage IIIA/IIIB Disease by Transbronchial Needle Aspiration Biopsy-Proven Stage IV Disease Stage IIIA/IIIB Disease by Mediastinoscopy Prohibitive Pulmonary Function Refusal of Treatment Comorbidities/Functional Status Death During Evaluation Lost to Follow-up or Unknown
53 (42) 28 (22) 7 (6) 6 (5) 6 (5) 5 (4) 5 (4) 4 (3) 11 (9)
Values in parentheses are percentages.
Mark Thomas Dransfield et al Figure 1 Patients Undergoing Surgical Resection 100
Cumulative Survival (%)
stage NSCLC, a randomized trial comparing conventional work-up with conventional work-up plus PET scanning found that the number of patients treated surgically was lower among the latter (65% vs. 81%).16-18 This was primarily on the basis of more frequent preoperative identification of stage III or greater disease and a reduction in the number of “open-and-close” thoracotomies (21% vs. 41%, respectively). No other study has examined the resection rate in a population routinely undergoing PET scanning. It could also be argued that some patients in our study were denied surgery inappropriately on the basis of imaging studies. Indeed, advanced disease based on imaging was the most frequent primary reason for nonresection, though a significant minority did have histologic proof of regional or distant disease. Importantly, 30 of 53 (56%) of those denied surgery on the basis of imaging underwent PET scanning, arguing that their evaluation was quite thorough. In addition, many patients had other contraindications to surgery. There was no evidence in our study that age or ethnicity influenced the use of surgical resection. This is in contrast to a study of Medicare beneficiaries in the United States by Bach et al who found that black patients were less likely to undergo resection for early-stage lung cancer than were white patients.4 This trend held true even after adjustment for comorbid illnesses. There are also data to suggest that, although short-term surgical outcomes for elderly patients with lung cancer approach those of the general population, age bias might unfairly deny thoracotomy and resection to patients aged > 70 years.19-21 In our study, 9 of 31 patients (29%) undergoing resection were aged > 70 years. Underlying COPD significantly increases the risk of thoracic resection in patients with lung cancer, and lobectomy is not well tolerated if the postoperative forced expiratory volume (FEV1) is expected to be < 40% of the predicted normal value.22 In addition, along with tumor size, preoperative FEV1 has been shown to be the best predictor of all-cause mortality among patients undergoing resection of stage I NSCLC.23 Not surprisingly, the incidence of COPD was higher among patients who were denied surgery, and their overall lung function was also worse. A small number of patients were denied surgery on the basis of lung function alone. Comorbid cardiovascular disease increases the perioperative risk of death from 2.5% to 9% and can also reduce 3- and 5-year survival for NSCLC by 15%-20%.24 As a result, most authorities have recommended a thorough cardiac evaluation before lung cancer resection. Although in our study the incidence of coronary artery disease was not different between the resected and nonresected groups, this likely reflects more aggressive preoperative testing in those whose tumors were amenable to resection. The Lung Mass Clinic is staffed by pulmonologists, and thus diagnoses that could be made by bronchoscopy were made quickly. As the yield for fiberoptic bronchoscopy increases with the size and proximal extent of tumors, it follows that many patients whose disease was amenable to bronchoscopy would be unresectable.25 Smaller, more peripheral tumors are often earlier-stage cancers that require surgery for diagnosis and cure. Although rapid diagnoses were made in patients with tumors amenable to bronchoscopy, most patients who underwent resection required other diagnostic procedures. Scheduling these procedures contributed to
80 60 40 20 0
250
500
750
1000
1250
1500
1750
2000
Postoperative Time (Days)
the median time to resection of 104 days, which clearly exceeds the 4- to 8-week delay thought acceptable by the American College of Chest Physicians and the British Thoracic Society.11,12 It is also far longer than the 5-week delay reported by Laroche et al in the study of a fast-access investigation unit at Papworth Hospital in the United Kingdom.8 It is notable that, in the Papworth system, many patients underwent staging computed tomography and bronchoscopy or transthoracic needle aspiration on the same day as their initial referral, thus eliminating some waiting time. In addition, unlike the Lung Mass Clinic, the Papworth Lung Cancer Clinic employed a multidisciplinary team, including respiratory physicians as well as radiologists, pathologists, oncologists, and thoracic surgeons. Each case was reviewed within 3 days of the initial investigation, and plans for additional work-up or treatment were made jointly with rapid scheduling. In the Papworth experience, the overall rate of “open-and-close” thoracotomy was 11%, although in our 4-year study, only 1 was performed. This might relate to the routine use of PET scanning in our population, though limited access to the PET scanner also contributed to our prolonged preoperative delay. The overall survival for patients undergoing resection in our study is currently 65% with median follow-up approaching 3 years. Survival for patients with stage IA disease is 80%. These results are in keeping with other reports of surgically treated patients of comparable age and staging with significant rates of comorbid illness.3,24 Although preoperative delays can lead to unnecessary psychologic stress and dissatisfaction among patients with lung cancer, there is little evidence that overall prognosis is affected.13,26 Although our study was not adequately powered to determine the effects of preoperative delay on survival, to date there is no difference in the risk of death among those with a preoperative delay of > 90 days versus those with a delay < 90 days (hazard ratio [HR], 1.13; P = 0.84 [Cox proportional-hazards test]). Using a similar cutoff of 90 days and studying a larger veteran population, Quarterman et al were also unable to document an effect of delay on 5-year survival (HR, 1.06).13 Interestingly, a study by Myrdal et al found that a shorter pretreatment delay was associated with a poorer prognosis among 466 patients with NSCLC in Sweden.27 This might reflect the fact that patients with significant signs and symptoms of disease receive more rapid treatment.
Clinical Lung Cancer January 2006
271
Lung Cancer Resection Conclusion In recent years, the VA has been criticized for inadequacies in care that might result from the inefficiency of a centralized healthcare system, with the under-use of coronary revascularization procedures being the most notable example.28 Since the inception of the Lung Mass Clinic, the BVAMC has performed as well as other health-care systems in the surgical management of lung cancer. Similar specialized clinics have also improved care for patients with lung cancer in the United Kingdom.8,26 We believe that complex medical diseases can be managed effectively in these centralized systems, but they might require the establishment of problemfocused clinics as well as significant administrative support.
Acknowledgement The study was supported by institutional funds at the BVAMC.
References 1. Jemal A, Clegg L, Ward E, et al. Annual report to the nation on the status of cancer, 1975-2001, with a special feature regarding survival. Cancer 2004; 101:3-27. 2. Janssen-Heijnen M, Gatta G, Forman D, et al, and the EUROCARE Working Group. Variation in survival of patients with lung cancer in Europe, 1985-1989. Eur J Cancer 1998; 34:2191-2196. 3. Fry W, Menck H, Winchester D. The national cancer data base report on lung cancer. Cancer 1996; 77: 1947-1955. 4. Bach P, Cramer L, Warren J, et al. Racial differences in the treatment of early stage lung cancer. N Engl J Med 1999; 341:1198-1205. 5. Humphrey E, Smart C, Winchester D, et al. National survey of the pattern of care for carcinoma of the lung. J Thorac Cardiovasc Surg 1990; 100:837-843. 6. Damhuis R, Schutte, P. Resection rates and postoperative mortality in 7,899 patients with lung cancer. Eur Respir J 1996; 9:7-10. 7. Gregor A, Thomson C, Brewster D, et al. Management and survival of patients with lung cancer in Scotland diagnosed in 1995: results of a national population based study. Thorax 2001; 56:212-217. 8. Laroche C, Wells F, Coulden R, et al. Improving surgical resection rate in lung cancer. Thorax 1998; 53:445-449. 9. Society of Cardiothoracic Surgeons of Great Britain and Ireland. UK Thoracic Surgical Register, 1994. 10. Billing J, Wells F. Delays in the diagnosis and treatment of lung cancer. Thorax 1996; 51:903-906
272
Clinical Lung Cancer January 2006
11. Alberts W, Bepler G, Hazelton T, et al. Practice organization. Chest 2003; 123:332S-337S. 12. BTS Recommendations to respiratory physicians for organising the care of patients with lung cancer. Thorax 1998; 53(suppl 1):S1-S8. 13. Quarterman R, McMillan A, Ratcliffe M, et al. Effect of preoperative delay on the prognosis for patients with early stage non-small cell lung cancer. J Thorac Cardiovasc Surg 2003; 125:108-114. 14. United States Department of Veterans Affairs. Available at: http://www.va.gov. Accessed November 4, 2005. 15. United States Department of Veterans Affairs Central Cancer Registry. Available at: http://www1.va.gov/cancer. Accessed November 4, 2005. 16. Hoekstra C, Stroobants S, Hoekstra O, et al. The value of [18F] fluoro-2-deoxy-D-glucose positron emission tomography in the selection of patients with stage IIIA-N2 nonsmall cell lung cancer for combined modality treatment. Lung Cancer 2003; 39:151157 17. Kalff V, Hicks R, Macmanus M, et al. Clinical impact of (18)F fluorodeoxyglucose positron emission tomography in patients with non-small cell lung cancer: a prospective study. J Clin Oncol 2001; 19:111-118. 18. van Tinteren H, Hoekstra OS, Smit EF, et al. Effectiveness of positron emission tomography in the preoperative assessment of patients with suspected non-small-cell lung cancer: the PLUS multicentre randomised trial. Lancet 2002; 359:1388-1393. 19. Port J, Kent M, Korst R, et al. Surgical resection for lung cancer in the octogenarian. Chest 2004; 126:733-738. 20. Yamamoto K, Padilla A, Calvo M, et al. Surgical results of stage I non-small cell lung cancer: comparison between elderly and younger patients. Eur J Cardiothorac Surg 2003; 23:21-25. 21. Peake M, Thompson S, Lowe D, et al. Ageism in the management of lung cancer. Age Ageing 2003; 32:171-177. 22. Beckles M, Spiro S, Colice G, et al. The physiologic evaluation of patients with lung cancer being considered for resectional surgery. Chest 2003; 123:105S-114S. 23. Iizasa T, Suzuki M, Yasufuku K, et al. Preoperative pulmonary function as a prognostic factor for stage I non-small cell lung carcinoma. Ann Thorac Surg 2004; 77:1896-1903. 24. Ambrogi V, Pompeo E, Elia S, et al. The impact of cardiovascular comorbidity on the outcome of surgery for stage I and II non-small lung cancer. Eur J Cardiothorac Surg 2003; 23:811-817. 25. Schreiber G, McCrory D. Performance characteristics of different modalities for diagnosis of suspected lung cancer: summary of published evidence. Chest 2003; 123:115S128S. 26. Murray P, O’Brien M, Sayer R, et al. The pathway study: results of a pilot feasibility study in patients suspected of having lung carcinoma investigated in a conventional chest clinic setting compared to a centralised two-stop pathway. Lung Cancer 2003; 42:283-290. 27. Myrdal G, Lambe M, Hillerdal G, et al. Effect of delays on prognosis in patients with non-small cell lung cancer. Thorax 2004; 59:45-49. 28. Petersen L, Mormand S, Leape L, et al. Regionalization and the underuse of angiography in the Veterans Affairs health care system as compared to fee-for-service system. N Engl J Med 2003; 348:2209-2217.
Key words: Diagnosis, Non–small-cell lung cancer, Practice organization, Thoracic surgery
Introduction
such as ethnicity and socioeconomic status might influence the use of surgical treatments for lung cancer.4 As a result of these limitations, reported resection rates for lung cancer are uniformly < 30%.3,5-9 Unfortunately, comparing these reports is problematic, because investigators have employed a variety of methods for data collection, and although some have reported resection rates for all lung cancers, others have restricted their analyses to non–small-cell lung cancer (NSCLC). Despite these limitations, there does appear to be significant discrepancies between resection rates in the United Kingdom and those observed in Europe and the United States. In a review of 3855 cases of lung cancer that were reported to the Scottish Cancer Registry in 1995, Gregor et al found that only 10.7% underwent surgical resection.7 A 10% resection rate has also been reported among all UK cardiothoracic surgeons, and rates as low as 1% have been observed in some regions without a respiratory physician.9 Even after accounting for the small-cell tumors that were included in these reports, UK resection rates for NSCLC still fall far short of those in the United States, where a survey of
Lung cancer is the leading cause of cancer-related death in the United States and United Kingdom, and 5-year survival remains < 15%.1,2 Multiple factors contribute to this poor prognosis, including the absence of an effective screening test and the fact that most patients present with late-stage disease. As a result, less than half of patients are eligible for surgical resection, the only modality that offers a legitimate hope of a cure.3 Even among those with early-stage tumors, only a fraction undergo surgery, because many have impaired lung function or comorbid illnesses that preclude an operation. Unfortunately, there is also evidence that nonmedical factors Division of Pulmonary, Allergy, and Critical Care Medicine, Birmingham Veterans Affairs Medical Center and University of Alabama at Birmingham Submitted: Dec 20, 2005; Revised: Jan 16, 2006; Accepted: Jan 19, 2006 Address for correspondence: Mark Thomas Dransfield, MD, Division of Pulmonary, Allergy, and Critical Care Medicine, Birmingham Veterans Affairs Medical Center and University of Alabama at Birmingham 215 THT, 1900 University Blvd, Birmingham, AL 35294 Fax: 205-934-6229; email: [email protected]
Electronic forwarding or copying is a violation of US and International Copyright Laws. Authorization to photocopy items for internal or personal use, or the internal or personal use of specific clients, is granted by CIG Media Group, LP, ISSN #1525-7304, provided the appropriate fee is paid directly to Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923 USA 978-750-8400.
268
Clinical Lung Cancer January 2006
hospitals with approved cancer programs found that 28% of NSCLC were resected.5 A review of cancer cases reported to the Rotterdam Cancer Registry demostrated a comparable NSCLC resection rate of 24%.6 There are many possible explanations for these international differences, though it has been suggested that the discrepancy is partly caused by poor access to specialized care in the United Kingdom and inordinate preoperative delays.2 Indeed, 1 study from the United Kingdom found a median time of 109 days from initial presentation to surgery, a delay well beyond the 4- to 8-week period recommended by the American College of Chest Physicians and the British Thoracic Society.10-12 Although there is no clear evidence to suggest that shorter preoperative delays result in better outcomes, 13 recent efforts to streamline the process of care with “fast access investigation units” have shown shorter waiting times and higher rates of surgical resection.8 Although US resection rates of 25%-28% have been reported,3,5 there are no published data about the rate of resection among patients with lung cancer cared for by the Department of Veterans Affairs (VA). The VA system is a network of 162 hospitals and almost 1000 supporting primary care clinics, which in 2002 provided health care for 1.6 million veterans of the US military services.14 The VA system is structured in a manner similar to the UK’s National Health Service in which specialty care, including lung cancer diagnosis and management, is provided by referral hospitals. In the late 1990s, there was concern that lung cancer care among patients at the Birmingham VA Medical Center (BVAMC) was inefficient and that opportunities for resection might be missed. Indeed, a recent review of the BVAMC Cancer Registry confirmed that only 17 of 135 patients (12.6%) diagnosed with NSCLC in 1993 were treated with curative surgical resection (BVAMC Cancer Registry Database). This is in keeping with data from the larger regional Veterans Integrated Service Network (including other facilities in Alabama, Georgia, and South Carolina), which showed that only 62 of 404 patients (15.3%) diagnosed with NSCLC in 1998 underwent wedge resection, segmentectomy, single or bilobectomy, or pneumonectomy for stage I-III disease.15 Because of these concerns about inefficiency, a specialized Lung Mass Clinic was established to provide rapid evaluation and triage of possible lung cancer cases. We hypothesized that, since the inception of the Lung Mass Clinic in 1999, rates of surgical resection at our facility would be comparable with those reported nationally.
Patients and Methods The study was approved by the Birmingham VA Institutional Review Board. We performed a retrospective review of the electronic medical records of all veterans who were referred to and seen in the Lung Mass Clinic during the first 4 years of its operation (October 14, 1999 to October 14, 2003). We identified patients who subsequently received a diagnosis of thoracic malignancy. Patients with a history of lung cancer were excluded from the study.
Patients who were found to have a new diagnosis of NSCLC were identified and divided into resected and nonresected subgroups. Baseline demographic information including age, sex, and ethnicity were recorded as were major comorbid medical illnesses including coronary artery disease, congestive heart failure, chronic obstructive pulmonary disease (COPD), cerebrovascular disease, chronic renal insufficiency, and diabetes mellitus. A comorbid illness was considered present if it was documented by an International Classification of Diseases–9 code or described in clinician notes. Pulmonary function data were abstracted from the medical record if they were obtained within 3 months of the Lung Mass Clinic appointment. Additional information obtained included the date and method of lung cancer diagnosis (fiberoptic bronchoscopy, transthoracic needle aspiration, or surgery) and the dates and results of subsequent staging and preoperative procedures, including positron emission tomography (PET), noninvasive cardiac stress testing, and mediastinoscopy. The times from initial Lung Mass Clinic evaluation to diagnosis and to resection (if applicable) were determined. For patients undergoing resection, final pathologic staging was recorded, and among those denied surgery, reasons for nonresection were documented and tabulated. If the explanation for denying resection was not clearly stated in the medical record, it was recorded as unknown. Postoperative survival was also determined. A subject was considered alive if there had been computer-documented contact with the VA system (eg, clinic visit, hospitalization, or medication refill) within 45 days of our review and if there was no computer entry documenting the patient’s death. Differences in baseline data between the resected and nonresected groups were compared with χ2, Fisher exact, or Student t tests as appropriate.
Results Four hundred eighty-seven patients were seen during the study period. Of these, 180 (37%) were found to have thoracic malignancies. This represents approximately 50% of the total number of thoracic malignancies reported to our cancer registry over the same period. The majority of the other cases of lung cancer were referred to our facility for treatment after having outside histologic confirmation of the disease. Therefore, these patients were not seen in the Lung Mass Clinic and were not included in the analysis. Non–small-cell lung cancer was the most common histologic subtype, accounting for 156 (87%) of the total. Other tumor types identified included small-cell lung cancer (n = 18), lymphoma (n = 3), carcinoid (n = 1), melanoma (n = 1), and unknown (n = 1). Of the 156 cases of NSCLC, 31 (20%) underwent surgical resection (29 lobectomies, 1 bilobectomy, and 1 pneumonectomy). Table 1 shows the final pathologic stage distribution of the resected tumors. The stage IIIA tumors were caused by unexpected N2 disease (n = 2) and chest wall invasion with N1 disease (n = 1). There were no significant differences in age, sex, or ethnicity between the resected and nonresected subgroups (Table 2). There were also no significant differences in the incidence of
Clinical Lung Cancer January 2006
269
Lung Cancer Resection Table 1 Final Pathologic Stage Distribution of Resected Tumors Stage
Number (%)
IA IB IIA IIB IIIA
Patient Characteristic
10 (32) 11 (35) 3 (10) 4 (13) 3 (10)
Table 3 Diagnostic Methods and Times to Diagnosis and Resection Resected (n = 31)
Method of Diagnosis Fiberoptic bronchoscopy Transthoracic needle aspiration Surgery PET Scans Performed Noninvasive Cardiac Stress Testing Median Time to Diagnosis (Days) Median Time to Resection (Days) Values in parentheses are percentages.
270
Clinical Lung Cancer January 2006
Nonresected (n = 125)
Resected Nonresected P Value (n = 31) (n = 125) 64.8 ± 8.4
66.8 ± 9
0.28
Ethnicity (% White)
84
74
0.35
Sex (% Male)
97
99
0.7
10 (32) 2 (6) 5 (16) 3 (10) 2 (6) 2 (6) 15 (48) 8 (26) 67 ± 17 71 ± 15
37 (30) 12 (10) 53 (42) 15 (12) 6 (5) 17 (14) 85 (68) 40 (32) 55 ± 17 62 ± 17
1 0.85 0.01 0.96 0.98 0.36 0.07 0.65 0.002 0.02
Age (Years)
major medical comorbidities between the 2 groups, except those resected had a lower incidence of COPD and better pulmonary function than the nonresected group. Patients who underwent resection were most often diagnosed at the time of surgery (61% of cases), and those whose disease was amenable to bronchoscopic diagnosis were usually unresectable (91% of cases; Table 3). Surgery was rarely required for the diagnosis of unresectable disease (7% of cases). Positron emission tomography scans and noninvasive cardiac testing were performed more often in patients who eventually underwent resection. The median time to diagnosis was much shorter among patients who were not resected. The median time to resection from the initial Lung Mass Clinic appointment was 104 days. Table 4 shows the primary reason for nonresection among the 125 patients who were denied surgery. Fifty-three patients had evidence of advanced disease (stage IIIA/IIIB or stage IV) by imaging studies. Of these, 30 (56%) underwent PET scanning. Tissue confirmation of stage IIIA/IIIB or stage IV disease was obtained in an additional 41 cases (33%). One hundred five (84%) of those not resected had evidence of advanced disease by biopsy or imaging studies, prohibitive pulmonary function, or refused therapy. In a minority of cases (9%), a clear reason for denying resection could not be found. With median follow-up approaching 3 years, overall survival among those undergoing resection is 20 of 31 (65%) compared with only 38 of 125 (30%) among those not resected. Figure 1 shows the Kaplan-Meier survival curve for patients who underwent surgical resection. Four deaths occurred in the first 30 days after surgery. Three of these were the result of postoperative pneumonia and its complications, and the cause of death for 1 patient who had been discharged is not known. Of the 11
Methods and Times
Table 2 Baseline Demographics, Comorbid Illness, and Pulmonary Function
Comorbid Illness (%) Coronary artery disease Congestive heart failure COPD Cerebrovascular disease Chronic renal insufficiency Diabetes mellitus Any ≥ 2 Illnesses FEV1 (Mean % Predicted) FVC (Mean % Predicted)
Mean values are presented with standard deviations. Abbreviation: FVC = forced vital capacity
deaths in the series, the cause of death was pneumonia in 4, unknown in 4, and recurrent disease in 3 patients.
Discussion Our study confirms the poor prognosis for most patients with lung cancer. As expected, the vast majority of patients presented with late-stage disease that was not amenable to surgical resection. Survival among those not resected was only 30% and will certainly decline further as follow-up is extended. Despite this unfortunate statistic, the fraction of patients who were treated surgically (20%) is higher than reported in most studies done in the United Kingdom and approaches the available estimates for the United States and Europe.3,5-9 This resection rate also represents a significant improvement compared with data from our hospital and region from the 1990s.15 It should also be noted that our resection rate is likely an underestimation, because 4 patients who were eligible for surgery chose to receive additional care in the civilian health-care system. In addition, we would argue that the frequent use of PET scanning in our population lowered the rate of resection. Although PET scanning has been shown to upstage and down-
P Value Table 4 Primary Reasons for Nonresection
10 (32)
96 (77)
< 0.001
2 (6)
19 (15)
0.24
19 (61) 29 (94)
9 (7) 41 (33)
< 0.001 < 0.001
18 (58)
26 (21)
< 0.001
70
8
< 0.001
104
–
–
Reason (n = 125)
Number
Stage IIIA/IIIB or IV Disease by Imaging Stage IIIA/IIIB Disease by Transbronchial Needle Aspiration Biopsy-Proven Stage IV Disease Stage IIIA/IIIB Disease by Mediastinoscopy Prohibitive Pulmonary Function Refusal of Treatment Comorbidities/Functional Status Death During Evaluation Lost to Follow-up or Unknown
53 (42) 28 (22) 7 (6) 6 (5) 6 (5) 5 (4) 5 (4) 4 (3) 11 (9)
Values in parentheses are percentages.
Mark Thomas Dransfield et al Figure 1 Patients Undergoing Surgical Resection 100
Cumulative Survival (%)
stage NSCLC, a randomized trial comparing conventional work-up with conventional work-up plus PET scanning found that the number of patients treated surgically was lower among the latter (65% vs. 81%).16-18 This was primarily on the basis of more frequent preoperative identification of stage III or greater disease and a reduction in the number of “open-and-close” thoracotomies (21% vs. 41%, respectively). No other study has examined the resection rate in a population routinely undergoing PET scanning. It could also be argued that some patients in our study were denied surgery inappropriately on the basis of imaging studies. Indeed, advanced disease based on imaging was the most frequent primary reason for nonresection, though a significant minority did have histologic proof of regional or distant disease. Importantly, 30 of 53 (56%) of those denied surgery on the basis of imaging underwent PET scanning, arguing that their evaluation was quite thorough. In addition, many patients had other contraindications to surgery. There was no evidence in our study that age or ethnicity influenced the use of surgical resection. This is in contrast to a study of Medicare beneficiaries in the United States by Bach et al who found that black patients were less likely to undergo resection for early-stage lung cancer than were white patients.4 This trend held true even after adjustment for comorbid illnesses. There are also data to suggest that, although short-term surgical outcomes for elderly patients with lung cancer approach those of the general population, age bias might unfairly deny thoracotomy and resection to patients aged > 70 years.19-21 In our study, 9 of 31 patients (29%) undergoing resection were aged > 70 years. Underlying COPD significantly increases the risk of thoracic resection in patients with lung cancer, and lobectomy is not well tolerated if the postoperative forced expiratory volume (FEV1) is expected to be < 40% of the predicted normal value.22 In addition, along with tumor size, preoperative FEV1 has been shown to be the best predictor of all-cause mortality among patients undergoing resection of stage I NSCLC.23 Not surprisingly, the incidence of COPD was higher among patients who were denied surgery, and their overall lung function was also worse. A small number of patients were denied surgery on the basis of lung function alone. Comorbid cardiovascular disease increases the perioperative risk of death from 2.5% to 9% and can also reduce 3- and 5-year survival for NSCLC by 15%-20%.24 As a result, most authorities have recommended a thorough cardiac evaluation before lung cancer resection. Although in our study the incidence of coronary artery disease was not different between the resected and nonresected groups, this likely reflects more aggressive preoperative testing in those whose tumors were amenable to resection. The Lung Mass Clinic is staffed by pulmonologists, and thus diagnoses that could be made by bronchoscopy were made quickly. As the yield for fiberoptic bronchoscopy increases with the size and proximal extent of tumors, it follows that many patients whose disease was amenable to bronchoscopy would be unresectable.25 Smaller, more peripheral tumors are often earlier-stage cancers that require surgery for diagnosis and cure. Although rapid diagnoses were made in patients with tumors amenable to bronchoscopy, most patients who underwent resection required other diagnostic procedures. Scheduling these procedures contributed to
80 60 40 20 0
250
500
750
1000
1250
1500
1750
2000
Postoperative Time (Days)
the median time to resection of 104 days, which clearly exceeds the 4- to 8-week delay thought acceptable by the American College of Chest Physicians and the British Thoracic Society.11,12 It is also far longer than the 5-week delay reported by Laroche et al in the study of a fast-access investigation unit at Papworth Hospital in the United Kingdom.8 It is notable that, in the Papworth system, many patients underwent staging computed tomography and bronchoscopy or transthoracic needle aspiration on the same day as their initial referral, thus eliminating some waiting time. In addition, unlike the Lung Mass Clinic, the Papworth Lung Cancer Clinic employed a multidisciplinary team, including respiratory physicians as well as radiologists, pathologists, oncologists, and thoracic surgeons. Each case was reviewed within 3 days of the initial investigation, and plans for additional work-up or treatment were made jointly with rapid scheduling. In the Papworth experience, the overall rate of “open-and-close” thoracotomy was 11%, although in our 4-year study, only 1 was performed. This might relate to the routine use of PET scanning in our population, though limited access to the PET scanner also contributed to our prolonged preoperative delay. The overall survival for patients undergoing resection in our study is currently 65% with median follow-up approaching 3 years. Survival for patients with stage IA disease is 80%. These results are in keeping with other reports of surgically treated patients of comparable age and staging with significant rates of comorbid illness.3,24 Although preoperative delays can lead to unnecessary psychologic stress and dissatisfaction among patients with lung cancer, there is little evidence that overall prognosis is affected.13,26 Although our study was not adequately powered to determine the effects of preoperative delay on survival, to date there is no difference in the risk of death among those with a preoperative delay of > 90 days versus those with a delay < 90 days (hazard ratio [HR], 1.13; P = 0.84 [Cox proportional-hazards test]). Using a similar cutoff of 90 days and studying a larger veteran population, Quarterman et al were also unable to document an effect of delay on 5-year survival (HR, 1.06).13 Interestingly, a study by Myrdal et al found that a shorter pretreatment delay was associated with a poorer prognosis among 466 patients with NSCLC in Sweden.27 This might reflect the fact that patients with significant signs and symptoms of disease receive more rapid treatment.
Clinical Lung Cancer January 2006
271
Lung Cancer Resection Conclusion In recent years, the VA has been criticized for inadequacies in care that might result from the inefficiency of a centralized healthcare system, with the under-use of coronary revascularization procedures being the most notable example.28 Since the inception of the Lung Mass Clinic, the BVAMC has performed as well as other health-care systems in the surgical management of lung cancer. Similar specialized clinics have also improved care for patients with lung cancer in the United Kingdom.8,26 We believe that complex medical diseases can be managed effectively in these centralized systems, but they might require the establishment of problemfocused clinics as well as significant administrative support.
Acknowledgement The study was supported by institutional funds at the BVAMC.
References 1. Jemal A, Clegg L, Ward E, et al. Annual report to the nation on the status of cancer, 1975-2001, with a special feature regarding survival. Cancer 2004; 101:3-27. 2. Janssen-Heijnen M, Gatta G, Forman D, et al, and the EUROCARE Working Group. Variation in survival of patients with lung cancer in Europe, 1985-1989. Eur J Cancer 1998; 34:2191-2196. 3. Fry W, Menck H, Winchester D. The national cancer data base report on lung cancer. Cancer 1996; 77: 1947-1955. 4. Bach P, Cramer L, Warren J, et al. Racial differences in the treatment of early stage lung cancer. N Engl J Med 1999; 341:1198-1205. 5. Humphrey E, Smart C, Winchester D, et al. National survey of the pattern of care for carcinoma of the lung. J Thorac Cardiovasc Surg 1990; 100:837-843. 6. Damhuis R, Schutte, P. Resection rates and postoperative mortality in 7,899 patients with lung cancer. Eur Respir J 1996; 9:7-10. 7. Gregor A, Thomson C, Brewster D, et al. Management and survival of patients with lung cancer in Scotland diagnosed in 1995: results of a national population based study. Thorax 2001; 56:212-217. 8. Laroche C, Wells F, Coulden R, et al. Improving surgical resection rate in lung cancer. Thorax 1998; 53:445-449. 9. Society of Cardiothoracic Surgeons of Great Britain and Ireland. UK Thoracic Surgical Register, 1994. 10. Billing J, Wells F. Delays in the diagnosis and treatment of lung cancer. Thorax 1996; 51:903-906
272
Clinical Lung Cancer January 2006
11. Alberts W, Bepler G, Hazelton T, et al. Practice organization. Chest 2003; 123:332S-337S. 12. BTS Recommendations to respiratory physicians for organising the care of patients with lung cancer. Thorax 1998; 53(suppl 1):S1-S8. 13. Quarterman R, McMillan A, Ratcliffe M, et al. Effect of preoperative delay on the prognosis for patients with early stage non-small cell lung cancer. J Thorac Cardiovasc Surg 2003; 125:108-114. 14. United States Department of Veterans Affairs. Available at: http://www.va.gov. Accessed November 4, 2005. 15. United States Department of Veterans Affairs Central Cancer Registry. Available at: http://www1.va.gov/cancer. Accessed November 4, 2005. 16. Hoekstra C, Stroobants S, Hoekstra O, et al. The value of [18F] fluoro-2-deoxy-D-glucose positron emission tomography in the selection of patients with stage IIIA-N2 nonsmall cell lung cancer for combined modality treatment. Lung Cancer 2003; 39:151157 17. Kalff V, Hicks R, Macmanus M, et al. Clinical impact of (18)F fluorodeoxyglucose positron emission tomography in patients with non-small cell lung cancer: a prospective study. J Clin Oncol 2001; 19:111-118. 18. van Tinteren H, Hoekstra OS, Smit EF, et al. Effectiveness of positron emission tomography in the preoperative assessment of patients with suspected non-small-cell lung cancer: the PLUS multicentre randomised trial. Lancet 2002; 359:1388-1393. 19. Port J, Kent M, Korst R, et al. Surgical resection for lung cancer in the octogenarian. Chest 2004; 126:733-738. 20. Yamamoto K, Padilla A, Calvo M, et al. Surgical results of stage I non-small cell lung cancer: comparison between elderly and younger patients. Eur J Cardiothorac Surg 2003; 23:21-25. 21. Peake M, Thompson S, Lowe D, et al. Ageism in the management of lung cancer. Age Ageing 2003; 32:171-177. 22. Beckles M, Spiro S, Colice G, et al. The physiologic evaluation of patients with lung cancer being considered for resectional surgery. Chest 2003; 123:105S-114S. 23. Iizasa T, Suzuki M, Yasufuku K, et al. Preoperative pulmonary function as a prognostic factor for stage I non-small cell lung carcinoma. Ann Thorac Surg 2004; 77:1896-1903. 24. Ambrogi V, Pompeo E, Elia S, et al. The impact of cardiovascular comorbidity on the outcome of surgery for stage I and II non-small lung cancer. Eur J Cardiothorac Surg 2003; 23:811-817. 25. Schreiber G, McCrory D. Performance characteristics of different modalities for diagnosis of suspected lung cancer: summary of published evidence. Chest 2003; 123:115S128S. 26. Murray P, O’Brien M, Sayer R, et al. The pathway study: results of a pilot feasibility study in patients suspected of having lung carcinoma investigated in a conventional chest clinic setting compared to a centralised two-stop pathway. Lung Cancer 2003; 42:283-290. 27. Myrdal G, Lambe M, Hillerdal G, et al. Effect of delays on prognosis in patients with non-small cell lung cancer. Thorax 2004; 59:45-49. 28. Petersen L, Mormand S, Leape L, et al. Regionalization and the underuse of angiography in the Veterans Affairs health care system as compared to fee-for-service system. N Engl J Med 2003; 348:2209-2217.