| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J Thorac Cardiovasc Surg 2005;129:652-660
© 2005 The American Association for Thoracic Surgery
General Thoracic Surgery |
a Division of Thoracic Surgery, Department of Cardiothoracic Surgery,
b Department of Genetic Medicine,
c Department of Public Health,
d Department of Surgery, Weill Medical College of Cornell University, New York, NY
Read at the Eighty-fourth Annual Meeting of The American Association for Thoracic Surgery, Toronto, Ontario, Canada, April 25-28, 2004.
Received for publication April 23, 2004; revisions received September 29, 2004; accepted for publication October 12, 2004. * Address for reprints: Robert J. Korst, MD, Department of Cardiothoracic Surgery, Weill Medical College of Cornell University, 525 East 68th St, Room M-404, New York, NY 10021 (E-mail: rjk2002{at}med.cornell.edu).
 |
Abstract |
|---|
 Top Abstract MethodsResultsDiscussionReferences |
|---|
METHODS: A retrospective analysis was performed of all patients who presented for follow-up in 2002 after complete resection for non–small cell lung cancer. Data collected included demographics, clinicopathologic features of the initial lung cancer, the number and results of surveillance computed tomographic scans performed in 2002, the attending surgeons' impressions of the surveillance scans, the nature of any abnormalities and further diagnostic testing prompted by these abnormalities, and the nature of any lung cancer detected on surveillance scans, as well as the treatment rendered. The cost of surveillance scanning and associated diagnostics was computed by using Medicare fee schedules.
RESULTS: Two hundred thirteen patients met the criteria for inclusion in the study cohort. One hundred sixty-eight surveillance scans were performed in 140 of these patients. One hundred five scans were interpreted as abnormal by the radiologist with regard to pulmonary nodules, adenopathy, or pleural fluid, but the surgeon was suspicious for recurrent or new primary lung cancer in only 32 of 105 scans. Further workup revealed recurrent or new primary lung cancer in 16 of 32 patients, with 6 undergoing resection for localized disease. The cost of the surveillance scans and associated care in the study cohort were 16.6% higher than the cost of care in a hypothetically identical cohort not subjected to surveillance scanning.
CONCLUSIONS: Surveillance computed tomography is frequently abnormal after complete resection for non–small cell lung cancer; however, the majority of these abnormalities are not clinically suspicious. Resectable metachronous lung cancer is detected by using surveillance scanning; however, the use of this modality can be associated with increased cost.
Computed tomography (CT) of the chest has been used extensively to aid in the diagnosis and staging of NSCLC for more than 2 decades. Chest CT has also been shown to be more sensitive than plain chest radiography for detecting both pulmonary nodules and NSCLC in early stages when used as a screening tool in populations thought to be at high risk for development of this disease.6 Finally, although chest CT is also currently used by many clinicians as a surveillance tool after surgical resection with curative intent, a paucity of published data exists regarding both its use in this capacity and the character of abnormalities detected.7-10 With this as a background, the purpose of this study is to determine (1) the prevalence of defined abnormalities detected by means of surveillance CT scans of the chest and upper abdomen in patients after curative resection for NSCLC; (2) the quantity, nature, and cost of further diagnostic testing prompted by findings on surveillance CT scans; and (3) the frequency of recurrent or new primary lung cancer detected by using surveillance CT scans.
|
Methods |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
The follow-up regimen, which had been adopted by the authors before this analysis, was comprised of an office visit with history and physical examination every 3 months for the first year after complete resection, every 6 months for the second year, and yearly thereafter. Radiographic examinations consisted of a posteroanterior and lateral chest radiograph at the time of the office visits at 3, 9, and 18 months after resection. CT generally encompassed the lung apices through the adrenal glands without intravenous contrast and was performed at the time of the office visits at 6 months, 12 months, and yearly thereafter.
Data collection and definitions
Data collected included patient demographics, initial tumor histology and stage, type and year of resection of the initial tumor, and the number and date of routine follow-up office visits with the operating surgeon in 2002. Additional data included the number and results of surveillance CT scans performed in 2002, the attending surgeons' impressions of the surveillance scans, the nature of any abnormalities and further diagnostic testing prompted by these abnormalities on surveillance scans, the number and nature of any lung cancer detected on surveillance scans (surveillance cancers), and the type of treatment rendered for these surveillance cancers. Also recorded was the number and nature of any lung cancer that was detected independent of the surveillance CT scans (interval cancers).
Surveillance CT scans were defined as only those performed in 2002 in patients with no signs or symptoms of lung cancer on history and physical examination at the time of the follow-up office visit. Any scans performed to evaluate specific complaints or those in patients with known cancer recurrence were excluded. Abnormalities were defined as either noncalcified pulmonary nodules, mediastinal or hilar adenopathy (>1.5 cm in diameter), or pleural fluid, as read from the radiologists' reports. Findings described by the radiologist that were consistent with postoperative change, including thickening, infiltrates, or hemithoracic volume loss were not considered abnormal.
Although the scan findings were obtained from the radiologists' reports, as described above, scans were only deemed suspicious for recurrent or new primary lung cancer if the attending surgeon altered the previously described follow-up regimen on the basis of the scan findings (eg, further diagnostic testing was pursued). As an example, if the radiologist read a scan as showing a pulmonary nodule but the surgeon did not alter the follow-up regimen, the scan was classified as not suspicious for recurrent or new primary lung cancer. In all cases the attending surgeon based his opinion of the scan (suspicious or not suspicious) on his interpretation of the combination of both the written radiologists' reports and his own impression of the films.
All patients thought to be disease free at the end of 2002 (including those with suspicious surveillance scans in whom recurrent or new primary lung cancer could not be documented) were confirmed to be disease free in 2003 by means of either office visits or telephone contact.
Assessment of cost
Costs for all diagnostic testing-procedures performed, as well as treatments rendered, were obtained from Medicare fee schedules for 2004.11 The cost of facility fees associated with outpatient procedures (flexible bronchoscopy, mediastinoscopy, and transthoracic fine-needle aspiration) were obtained from the hospital billing department and were based on Medicare reimbursements obtained by the institution for actual patients undergoing these procedures. The detailed breakdown of costs, both diagnostic and therapeutic, associated with the performance of surveillance CT scans in 2002 is depicted in Table 1. Treatment of patients who had unresectable stage 3 or 4 disease was universally assigned a cost of $55,185 on the basis of published data from an economic analysis of the Southwest Oncology Group Trial S9509, which determined the cost of the use of carboplatin and paclitaxel, plus associated costs for the treatment of advanced (stage 4 and selected stage 3B) NSCLC adjusted to 2004 dollars.12
|
Second, in this group of 6 patients who underwent reresection for localized metachronous disease, it was assumed that 40% would have been cured of their lung cancer by the second operation, whereas 60% would have recurred with advanced disease after the second procedure. This assumption is based on the results of the 2 largest published series to date on the curability of early-stage metachronous lung cancer.13,14
Third, it was assumed that the cost of the workup of patients with metachronous lung cancer would be similar, regardless of stage. Patients with resectable disease usually require radiographic staging studies, as well as an attempt to obtain a tissue diagnosis in most cases, as do patients with unresectable advanced disease.
Statistical analysis
Because this is an observational study, no statistical analysis was performed.
|
Results |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
|
One hundred five scans were considered abnormal by the radiologists' reports with regard to pulmonary nodules, adenopathy-mediastinal mass, or pleural fluid. Figure 2 depicts the distribution of these abnormalities throughout these 105 scans. Despite this high number of abnormal scans, only 32 (30%) of 105 were thought to be suspicious for recurrent or new primary lung cancer by the attending surgeon such that the follow-up regimen was altered. Follow-up through January 2004 of the patients whose scans were read as abnormal by the radiologist but not deemed suspicious by the surgeon (73/105 scans) revealed 5 additional recurrences: 1 in the brain, 1 in the liver, and 3 in the pleura. Rereview of the 2002 CT scan reports for the 4 patients who had recurrences in the chest and the liver revealed no abnormalities in the pleura and liver in 2002, implying that after an extra year of follow-up, the false-negative rate of the surgeons' impressions of the 2002 scans was 0%.
|
|
|
|
|
|
|
Discussion |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
Abnormalities detected on surveillance CT scans
Surprisingly few published data exist regarding the prevalence of abnormalities on postoperative CT scans, with only one report commenting on the presence of pulmonary nodules. Lamont and colleagues7 found that 45 (44%) of 102 patients had noncalcified pulmonary nodules on their surveillance scans, which is similar to the 90 (54%) of 168 seen in the present study. Together, these 2 reports suggest that the prevalence of pulmonary nodules in patients with previously resected NSCLC might be relatively high compared with previously published screened populations with no history of lung cancer.6 Compared with pulmonary nodules, the presence of pleural fluid or adenopathy was relatively uncommon on the surveillance scans. Although pleural fluid is a common postoperative finding, it usually resolves by the time the initial surveillance scan is performed at 6 months after surgical intervention. For this reason, in the present study pleural fluid was not considered as a postsurgical change.
In the present study no attempt was made to distinguish between new CT findings and those noted to be present on the patients' scans before their first resections; the purpose was simply to determine the prevalence of abnormal findings on surveillance scans. Although some of these abnormalities might have indeed been present on initial preoperative scans, in many instances these initial scans are not available to the clinician caring for the patient. Certainly one reason why some findings on the surveillance scans might not have been deemed suspicious by the attending surgeon might have been because prior scans were available for comparison.
As defined in the "Methods" section, scans were classified as suspicious for recurrent or new primary lung cancer only when the attending surgeon altered the standard follow-up regimen on the basis of either his interpretation of the films, the radiologist's report, or both. By using this definition, only 30% (32/105) of scans in which the radiologist listed a defined abnormality were deemed suspicious by the attending surgeon and subsequently investigated. This raises the obvious question as to the significance of the findings on the remaining 73 unsuspicious scans read as abnormal by the radiologists. Although these patients have been followed into 2004, a clear weakness of this study is its cross-sectional nature, with lack of longer-term follow-up. The only reliable method to answer this important question is with a prospective and longitudinal study, which the authors have initiated.
Once scans were deemed suspicious by the attending surgeon, further testing ranged from more intensive CT surveillance to diagnostic thoracotomies and lung resections. An interesting observation was that recurrent or new primary lung cancer could only be confirmed in 50% (16/32) of these patients, implying that even the combination of an abnormal radiologist's report and a suspicious interpretation by the surgeon only had a specificity of 50% for malignancy. Once again, further prospective data collection is needed to determine whether the remaining 16 patients do indeed have lung cancer in the future and, if so, whether it is related to the surveillance CT scan abnormalities.
Lung cancer detected on surveillance scans
In the present study surveillance scanning detected 16 cases of recurrent or new primary lung cancer in 2002, with 6 (38%) cases being amenable to curative resection. The distinction between recurrent and second primary lung cancer was not defined per se because although criteria have been published to distinguish new primary from recurrent disease,15 these criteria are not universally agreed on, particularly with respect to the disease-free interval. Additionally, the distinction between second primary lung cancer and a solitary intrapulmonary recurrence (metastasis from the first tumor) might not be clinically relevant, especially with the more recent concept of multifocality, which might be relevant to certain types of NSCLC.16 In contrast to the cancers detected with surveillance CT, 9 patients in the study cohort who were being surveyed with CT presented in 2002 with lung cancer independent of surveillance scanning (interval cancer), and none had resectable disease. Although 2 patients presented with brain metastases, several patients had disease in the chest, which might represent a failure of surveillance scanning (Table 5). A weakness of this study is that the majority of patients who were undergoing CT surveillance in 2002 had their first cancer operated on in 2001 or 2002, limiting the follow-up period. Because many metachronous lung cancers will present at a later time, longer follow-up would be ideal for assessing the effect of surveillance CT scans.
Costs associated with surveillance scanning
In the present study cost figures were based on Medicare reimbursements for designated procedures. Although this does not represent the actual realized cost of these procedures that patients and the health care facility incur for these workups, these figures can be viewed as how much these patients' care costs the Medicare system and are useful for comparative purposes.
This cost analysis consists of 2 components. The first portion consists of a tabulation of the cost of the surveillance scans, as well as the additional workup and treatment of abnormalities found on these scans, whereas the second analysis involves a comparison of this dollar figure to that from a hypothetical cohort consisting of the very same patients if they did not undergo surveillance scanning. The cost data suggest that additional cost might be incurred by instituting a surveillance CT regimen. However, because the Medicare fee schedules used in the present study apply to Manhattan, NY, and these schedules tend to be higher than those in many other areas of the country, this increase in cost might be somewhat overestimated. The significance of this additional cost is difficult to interpret because this study is not designed to evaluate the cost-effectiveness of surveillance scanning. Longitudinal follow-up data are necessary to assess the effectiveness of this approach, preferably collected in a prospective fashion. In this regard useful areas of future investigation might include prospective and longitudinal evaluation of a surveillance CT program, creation of a decision-analysis model to assess cost-effectiveness, and finally a randomized prospective trial comparing the use of surveillance CT with the lack of surveillance. Accrual to such a randomization, however, might prove to be difficult given the ease and lack of morbidity associated with surveillance CT scanning.
Discussion
Dr Keith S. Naunheim (St Louis, Mo). Surveillance testing is almost universally performed after lung cancer resection but is highly variable in nature. A few guidelines have been published, but virtually none of these regimens have been demonstrated definitively to be beneficial on the basis of hard outcome data. Several prior publications, including those from my own institution, have questioned the value of intensive postresection surveillance. Such regimens have not been demonstrated to regularly identify recurrent cancer at a stage at which further surgical intervention is likely to be beneficial. Rather, it has been suggested that the real value of such testing is the detection of second primary lesions that might allow for subsequent resection and cure.
Into this admixture of supposition and uncertainty, the authors have introduced a new wrinkle for annual surveillance chest CT. This article reports on the costs and results of the chest CT scans performed routinely in asymptomatic patients after lung cancer resections. It is an observational study and does not purport to evaluate efficacy of recurrent cancer detection or any survival effect of testing. Rather, it seeks to identify the incremental costs of such testing and report the incidence of malignancy, subsequent resection, or both.
I have some comments and follow-up questions. First, regarding the economic analysis, the authors used the term "cost" when it would be more correct to use the term "charges." Charges are very often used as a surrogate for cost, but as you know, cost is very difficult, if not impossible, to actually determine. Also, their computation of costs includes Medicare payments for most physicians' fees but fails to include facility costs for some of the diagnostic procedures, and this would need to be rectified. These are minor points, however.
As for the clinical aspects of their article, I also have some questions. They have suggested that their article deals with both second primary lung cancers and recurrent disease. However, only 4% of the patients examined were 5 years or more out from their resection. Accordingly, it is likely that the overwhelming majority of patients with cancer had recurrent or persistent tumors rather than second primary lesions. Because of this, I must also question their assumption that 40% of second resections would be curative. We have no data to tell us whether these were local or contralateral recurrences, and all but one were resected through segmentectomy or wedge resection. I believe the 40% cure figure to be an overly optimistic one.
None of this, however, greatly detracts from the essential value of the article. It is a first glimpse at very short-term results of annual chest CT surveillance after lung cancer resection. Because of the retrospective nature of this study, the lack of follow-up, and the absence of a control group, no conclusions are possible regarding the clinical value of this strategy.
In conclusion, I would like to ask the authors several questions. First, do they routinely recommend resection for all recurrent cancer presenting as isolated nodules in the postoperative period?
Dr Korst. Thank you, Dr Naunheim, for those constructive comments.
Do we routinely perform resection for any new pulmonary nodule that occurs in these patients? The issue, as you brought up, is clearly whether we think we are dealing with a second primary lung cancer, and it is a relatively easy decision when the disease-free interval is at least more than a couple of years, but it gets a little dicey when you are less than a couple of years our, at which time I think the issue of resection really needs to be individualized on a patient-to-patient basis. I can tell you that the shortest disease-free interval in this study was 13 months, and therefore there was 13 months between the initial cancer, which was a very small stage I cancer in the right middle lobe, and when this gentleman got another small lesion in the left lower lobe and had very terrible lung function, and given that both lesions were very small and appeared to be clinical stage IA, we went ahead and resected that. Now was that a new primary lung cancer? Even though the disease-free interval was short, I think it was. That is obviously debatable. When you look at the disease-free intervals for those 6 patients, they ranged from 13 months to 8 years. Therefore, obviously, for the 8-year interval, you are thinking this is a second primary cancer. But it really has to be individualized when you are dealing with a patient with a short disease-free interval.
Dr Naunheim. Does the laterality, whether it is an ipsilateral or a contralateral lesion, make a difference to you?
Dr Korst. The laterality does not, but clearly the location of the lesion in the chest might make a difference. If you have a lesion that is close to the hilum in a previously resected patient, you might worry a little bit more about a recurrence.
Dr Naunheim. And how about the prior tumor stage, the stage of the first primary lesion?
Dr Korst. All 6 patients that we reresected all had initial primary lesions that were stage I, and therefore I believe that might make a difference, and as you correctly pointed out, I cannot definitively say that on the basis of this observational study, but I think that might come out in any prospective data collection.
Dr Naunheim. How do you decide when to proceed? Your staging algorithm includes, one would assume, positron emission tomographic (PET) scanning. Do you do redo mediastinoscopy, and do you seek a histologic diagnosis on every nodule before resection?
Dr Korst. No. We did some diagnostic resections, and therefore we did not get histology on every patient. We are aggressive about pursuing metastatic disease, though, and all patients get PET scans, and anything that lights up on the PET scan we attempt to biopsy. Therefore we are aggressive to make sure that we are just operating on persons with a solitary focus in the lung.
Dr Naunheim. Do you intend to continue this relatively intensive surveillance regimen with 6-month, 12-month, and yearly computed axial tomographic scans? Do you consider that it might be wasteful and that if other people imitated your institution, we might be throwing away money, or do you see a prospective trial anywhere in the future?
Dr Korst. Clearly we need to have prospective data collection. We are going to continue this for several reasons. The first is that we now have an institutional review board–approved protocol to collect data, and we attempt to enroll every patient who comes in after a complete resection in this protocol. The second reason is that we did pick up resectable metachronous lung cancer in 6 patients. Therefore we are still not clear on whether this is doing any good for any patients. Third, at least where we practice, our belief is that these patients will get CT scans, regardless of whether we perform them. For that reason, I think, as the operating surgeons, that we intimately know the inside of these patients' chests. We know what their previous CT scans showed. I think we are most qualified to know when something is suspicious for either a recurrence or a new primary lesion. Therefore we would like to do that. For those reasons, we are going to continue
Dr Naunheim. I would like to congratulate you and your coauthors on this provocative article. Unfortunately, it raises more questions than it answers, but still, it is very interesting.
Dr Larry R. Kaiser (Philadelphia, Pa). I am just taking the prerogative of the chair to ask one question, and Keith mentioned it just briefly. This issue between costs and charges is a major factor here. It seems to me you were very interested in cost, and if you had cost data, I think it would be even more interesting. As Keith points out, there is more to cost than the charges that the patient is billed on a Medicare fee schedule, and in fact, the costs that are associated with the technical fees are greater by far than the costs that are associated with the professional fees. Therefore, if you could get cost data, this would be particularly interesting to look at in terms of this being a public health issue and in terms of how best to follow patients who are basically asymptomatic; obviously the follow-up data are important. Is it possible for you to get cost data in the future, do you think, as you move forward?
Dr Korst. There are obviously many components of cost, both direct costs of procedures and so forth and a kind of more subtle cost, including such costs as time off from work and so forth. I agree with you. I think that to do a study that aggressively pursues the issue of cost, that really needs to be the primary focus of the study and would be worthwhile. There is no question about that.
Dr Altorki. If your hospital is anything like ours, it is almost impossible to get any information from them about cost, but I agree with you that in principle that would be ideal.
Dr Stephen D. Cassivi (Rochester, Minn). I have one comment and one question.
In terms of defining between a separate primary or recurrent disease, I think you could do that by comparing molecular markers, such as p53. That is the comment.
My question to you is as follows: What is the outcome of those patients in whom the surgeon thought the radiologic finding was not suspicious? I think it will be interesting, in the years to come, to follow up those patients, those 50% who had lesions that the surgeon deemed not suspicious.
Dr Korst. There are 73 patients in whom the radiologist read an abnormality. We did not think it was suspicious. Last week we went back and called those 73 patients, and in the last year, which is now a year after this study, there were 4 more recurrences: 1 in the brain and 3 in the pleura. We went back and looked at the chest CT scan reports in 2002 of the ones that recurred in the pleura, and there were no abnormalities in the pleura read in 2002. Therefore as the years go on, we will get more information.
|
References |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. K. Gould, C. D. Berg, D. R. Aberle, the National Lung Screening Trial Executive Commit, P. B. Bach, S. Kulaga, I. Karp, A. J. Yee, T. J. Lynch, G. A. Silvestri, et al. CT screening for lung cancer. N. Engl. J. Med., February 15, 2007; 356(7): 743 - 744. [Full Text] [PDF]
|
|
|
|
 |
|
 R. J. Korst, A. L. Kansler, J. L. Port, P. C. Lee, and N. K. Altorki Accuracy of surveillance computed tomography in detecting recurrent or new primary lung cancer in patients with completely resected lung cancer. Ann. Thorac. Surg., September 1, 2006; 82(3): 1009 - 1015. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. S. Kent, P. Korn, J. L. Port, P. C. Lee, N. K. Altorki, and R. J. Korst Cost Effectiveness of Chest Computed Tomography After Lung Cancer Resection: A Decision Analysis Model Ann. Thorac. Surg., October 1, 2005; 80(4): 1215 - 1223. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
