| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Division of Cardiothoracic Surgery, University of Alabama at Birmingham, Birmingham, Alabama
Accepted for publication May 13, 2009.
* Address correspondence to Dr Cerfolio, Division of Cardiothoracic Surgery, University of Alabama at Birmingham, 703 19th St S, ZRB 739, Birmingham, AL 35294 (Email: robert.cerfolio{at}ccc.uab.edu).
Presented at the Forty-fifth Annual Meeting of The Society of Thoracic Surgeons, San Francisco, CA, Jan 26–28, 2009.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
Methods: A prospective database was retrospectively reviewed. All patients had an integrated positron-emission tomography/computed tomography (CT) and CT scan, an R0 pulmonary resection with lung palpation, and complete thoracic lymphadenectomy.
Results: From August 2002 until July 2008, 2171 patients presented with presumed, resectable NSCLC. Of these, 721 were clinically (c)staged I, and 1450 were (c)staged II, III, or IV. Of the 721 (c)stage I, 405 (56%) had (p)stage I disease; 101 (14%) were clinically over-staged (benign nodules). Of those with NSCLC, 32% were clinically under-staged (stage II or higher on path). The 5-year Kaplan-Meier survival rates were 80% for (p)stage IA, 72% for (p)stage IB (p = 0.026), and 87% for the 721 with (c)stage I disease. The median-follow up was 3.8 years.
Conclusions: When patients with NCSLC are accurately staged preoperatively and undergo complete thoracic lymphadenectomy, the 5-year survival is 80% for (p)stage IA tumors and 87% for (c)stage I disease. About 32% of patients are under-staged (most commonly from nonimaged N2 disease) despite the liberal application of all of the techniques that assess mediastinal lymph nodes preoperatively. Thus surgical intervention offers improved staging with resection of unsuspected nodal or parenchymal disease. If stereotactic radiation and radiofrequency ablation are considered for patients with clinically staged I NSCLC, these results should be considered.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
Although these techniques hold promise and have a clear role in the rare patient who is not a surgical candidate, their results as a curative form or replacement therapy for low-risk surgical patients remain unproven. A Japanese study by Onishi and colleagues [2] in 2004 that included nearly 100 patients who were deemed medically operable and who underwent SRT showed a surprisingly high 5-year overall survival rate of 70%. Another report by Uematsu and colleagues [3] in 2001 documented a 3-year overall survival of 86% in the 29 patients who were deemed medically operable. Other series report findings on nonmedically operable patients with survival ranges from 30% to 70% [4, 5]. Several important cautions include that the follow-up was not complete and the definitions of local recurrence were vague. Moreover, NSCLC may represent a different biologically and genetic disease in North America compared with Asia.
Because these new techniques attempt to replace surgical resection, the improved staging that surgical intervention offers must be considered. This advantage is often ignored. Despite the use of preoperative integrated positron emission tomography (PET) using 2-dexoy-2-18F-fluoro-D-glucose (FDG) with computed tomography (integrated PET/CT), 5-mm columinated fine-cut CT scanning with contrast, endobronchial ultrasound with fine needle aspirate (EBUS-FNA), esophageal ultrasound with FNA (EUS-FNA) and mediastinoscopy, thoracotomy and surgical resection stills improves staging.
The objectives of this study were to determine the number of patients who actually had pathologic (p)stage I disease but were clinically (c)staged I NSCLC after integrated PET/CT scanning, fine-cut CT scan with contrast, and the liberal use of all of the mediastinal lymph node biopsy techniques currently available. In addition, we wanted to report the 5-year survival of patients who underwent careful preoperative and intraoperative staging using a thoracotomy with lung palpation and complete thoracic lymphadenectomy and had resected (p)stage I. These values could serve as realistic benchmarks to compare the 5-year survival for nonsurgical therapy in operable patients.
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
If PET/CT and CT scan suggested different stages, the integrated PET/CT scan stage was used; however, if pulmonary nodules were less than 0.8 mm, then the CT scan stage was used. All suspicious N2, N3, or M1 areas (maximum standardized uptake value > 2.5) were biopsied before pulmonary resection. In addition, EBUS or mediastinoscopy, or both, was used for tumors that were centrally located or that had a maximum standardized uptake value of 9 or greater. A combination of mediastinoscopy, EBUS, or EUS-FNA were used to biopsy suspicious lymph nodes. Mediastinoscopy was used to biopsy lymph nodes stations 2R, 4R, 2L, 4L, and 7. EUS was used for suspicious lymph nodes in station 4R, 4L, 7, 8, and 9. EBUS was used to evaluate lymph node stations 2R, 4R, 4L, and 7. EUS and EBUS were performed under conscious sedation and as previously described [7].
Analysis was conducted using SAS 9.01software (SAS Institute, Cary, NC). Continuous data are presented as medians or means, and categoric data are presented as percentages. The Fisher exact test or the Pearson 
2 test was used to assess categoric data and the Wilcoxon test to evaluate continuous variables. Survival estimates were derived by Kaplan-Meier analysis, and Mantel-Haenszel log-rank test was used to assess differences in survival amongst groups. A two-sided value of p < 0.05 was considered statistically significant and unlikely due to chance.
Efficacy was computed using standard calculations for sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy. A "true positive" was defined as any patient who was both clinically and pathologically staged 1 NSCLC. A "true negative" was any patient who was staged anything other than stage 1 both clinically and pathologically. "False negative" was any patient clinically staged as having stage II, III, or IV NSCLC but found to have pathologic stage 1 NSCLC on resection. A "false positive" was any patient who was clinically staged I but found to have any stage other than I on pathology after resection, including benign, stage II, III, IV NSCLC, or metastatic disease.
Patients alive at the end of the study were censored for purposes of survival analysis. Death from any cause was used to determine the overall survival rate. Local, regional, or distant recurrence of cancer was used to determine the disease-free survival rate. Local recurrence was defined as recurrence of tumor within the same lobe; regional recurrence was defined as the involvement of mediastinal lymph nodes ipsilateral to the side of surgery; and distant recurrence was defined as recurrence in the contralateral lung or in the remaining lung on the same side the operation was performed, or tumor recurrence elsewhere in the body. Operative deaths were not eliminated from the survival data set and were included as deaths in the survival analyses.
Follow-up consisted of chest and abdominal CT every 6 months for the first 2 years and yearly afterwards. Data were obtained from multiple sources such as clinic letters, follow-up scans, hospital computer information systems, tumor registry, Social Security Death Index, telephone calls, and letters from referring physicians or oncologists. The final data acquisition point was July 2008, and patients were monitored to January 2009.
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
|
The remaining 715 patients underwent thoracotomy. As shown in Figure 1, 101 of these 715 patients had benign nodules, 405 had (p)stage I disease, 70 had (p)stage II, 106 had (p)stage III, and 17 had (p)stage IV NSCLC disease secondary to unsuspected malignant pulmonary nodules of the same pathology in a different lobe. In addition, 16 patients had unsuspected metastatic disease from solid organ tumors thought to be primary lung cancers, comprising 5 patients with colorectal cancer metastases, 3 with renal cell cancer, 2 with hepatocellular carcinoma, and 6 with other types of solid organ metastases to the lung. Synchronous stage I tumors from nodules of different pathology that were discovered in a different lobe were resected in 25 patients, who were considered (p)stage I for the analysis.
In addition, 99 patients who were (c)stage greater than stage I disease were found to have (p)stage I NSCLC (Fig 1). These 504 patients (405 plus 99) comprise the patients used for the survival analysis for this study. Table 1 summarizes the raw data used to calculate the efficacy of integrated PET/CT for predicting stage I disease. The sensitivity was 80%, the specificity was 81%, the PPV was 56%, the NPV was 93%, and the accuracy was 81%.
|
Of the 504 patients used for the survival analysis in this study, 220 had (p)stage IA and 284 had (p)stage IB NSCLC. Their characteristics are reported in Table 2. The only significant difference between the two groups was type of resection. Significantly more lobectomies were done in the stage IB group compared with the stage IA group (p = .011). There were 5 (2%) operative deaths in the patients with (p)stage IA NSCLC and 4 (1%) in those with (p)stage IB.
|
|
|
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
These findings are consistent with previous reports as well. In 2005 we showed that integrated PET/CT and CT scans only correctly predicted the actual stage in 53% of patients [10]. Toloza and Harpole [11] from Duke reported similar findings in 2003 when they documented a low accuracy of CT imaging for metastatic disease in the mediastinum with a sensitivity of only 57% and a specificity of 82% [11] even after improvements were made in CT scan resolution. The low NPV and PPV of CT scan for direct tumor invasion into the chest wall or of mediastinal structures (T3 and T4 involvement), or both, is also well documented.[12] An important multi-intuitional study by Reed and colleagues [13] in 2004 showed a NPV of only 87% and a PPV of 56% for detection of nodal involvement by PET scan.
The findings of these studies must be carefully considered, especially if nonsurgical therapies for NSCLC are to be considered for patients who are at low-risk for pulmonary resection. If SRS, which is most commonly and effectively delivered using the Cyberknife (Accuray, Sunnyvale, CA) or RFA procedures, or both, are chosen, patients need to be carefully counseled about the modalities advantages and disadvantages. This includes the frequent clinical under-staging and the improved staging offered by surgical resection. Careful postprocedural follow-up is warranted.
Patients who are offered nonsurgical therapy for stage I disease who are labeled "medically inoperable" should have been carefully evaluated by at least one dedicated general thoracic surgeon. The reason for medical inoperability must be carefully documented, and the label should be reconsidered after patients complete smoking cessation programs as well as pulmonary, cardiac, and nutritional rehabilitation programs. Then, if nonsurgical therapy is chosen, the general thoracic surgeon must continue to be involved in these nonsurgical procedures. We should work in tandem with our colleagues in radiation oncology and medical oncology.
The definitive question is: If patients are offered Cyberknife, is their survival lowered by not having surgical therapy? That question will only be accurately answered by well-designed, prospective randomized studies, some of which are on the way. In addition, the answer depends partially on one's belief of what is the best treatment of lung cancer for each stage. In this study, we found the most common cause for misstaging was from unsuspected or missed N2 disease. Because it is unknown if microscopic unsuspected N2 disease is better treated by neoadjuvant therapy followed by resection, compared with resection followed by adjuvant therapy, the clinical effect of this type of misstaging is unknown [14]. The answers to these questions also depend on the number of patients who had T3 disease instead of T1 or T2, because theoretically, SRS using Cyberknife or RFA may offer adequate therapy despite the clinical misstaging.
This retrospective study documented a 5-year survival of 80% for those patients with (p)stage IA disease and 72% for those with (p)stage IB NSCLC. The International Association for the Study of Lung Cancer Lung Cancer Staging Project reported 5-year survival rates of 77% and 71% after radical excision of pT1 N0 tumors 2 cm or smaller (pT1a; n = 1816 patients) and tumors 2 to 3 cm, respectively (pT1b; n = 1653 patients). The corresponding 5-year survival rates for pathologically staged T2 N0 tumors ([p]stage IB) measuring 3 to 5 cm (pT2a; n = 2822), 5 to 7 cm (pT2b; n = 825), and exceeding 7 cm (pT2c, now considered T3 in the new staging classification system; n = 364) were 58%, 49%, and 35% (p < 0.0001), respectively [15]. Our survival rate was higher, but this is probably because all of our patients underwent PET/CT before resection and all had complete thoracic lymphadenectomy. Thus, more of our patients were probably truly stage I and not under-staged. This may explain the higher survival rates demonstrated in our study.
In 2007 Simon and colleagues [16] reported a 5-year survival rate of 27% for patients with stage I NSCLC who underwent RFA [16]. Similarly, Dupuy and colleagues [17] in 2006 reported a 5-year survival rate of 39% for patients with stage I NSCLC who underwent RFA. The cause of death in all 3 of the 14 patients who died was recurrent cancer or metastatic disease. These low survival rates may be secondary to the patient's underlying disease that led to the nonsurgical option being selected. In addition, as in most reports, the survival is not cancer-specific. We similarly report overall survival and not cancer-specific survival because the true cause of death in many patients is unknown. A patient may die secondary to cancer recurrence, but it is unknown; it may be mistakenly listed as natural causes when the patient actually died of nondiagnosed metastatic disease. Therefore, we chose to report the overall survival.
The strengths of this study are the careful and similar preoperative and intraoperative staging the patients received as well as the careful postoperative follow-up. Other strengths include the large number of patients, the follow-up was active, and all patients had complete thoracic lymphadenectomy. Among the limitations are that this was a retrospective single-center study from one surgeon. Another limitation is that the survival data are not cancer-specific.
In conclusion, we have shown that about a third of patients with NSCLC who are thought to have stage I disease after careful clinical staging actually have disease that pathologically stage II or higher. The most common misstaging is secondary to surprise N2 disease that probably is not adequately treated with SRS or RFA procedures. The 5-year survival is 80% for patients with pathologic stage IA NSCLC and 72% for pathologic stage IB. Those with clinical stage I disease have a 5-year survival of 87%. These findings concerning misstaging and survival add to the growing body of benchmark data for comparison of surgical vs nonsurgical therapy.
Carefully designed prospective, multi-intuitional randomized studies are needed to better evaluate the efficacy, morbidity, mortality, recovery time, and survival of the medical therapies for patients with clinically staged I NSCLC patients compared with the surgical therapies.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
DR CERFOLIO: You are right. That is a very, very good point and some might think that is one way to do it. You saw through the ruse. I wanted to set the bar high and thus we looked at only those with true pathologic stage I. However, since our clinical stage Is who were not received superior care, the comparison is not fair since they had more treatment then just to the primary lesion.
DR BRYAN FITCH MEYERS (St. Louis, MO): I guess I saw through it, too, because I was going to say the same thing.
DR CERFOLIO: Yes. I mean, it is an important point, one that is discussed in the paper.
DR HIROYUKI ITO (Yokohama, Japan): Thank you for your good presentation. So 20% of the operations you did perform the segmentectomy. Why did you decide to do this operation for stage IA?
DR CERFOLIO: That is a very good question. During most of the time frame of this study, 2002 to 2008, and even today, I still believe in lobectomy for a 1.2-cm nodule if the patient has good pulmonary functions.
Now I admit, maybe now if I have someone with a 7- or 8-mm nodule, even with good PFTs [pulmonary function tests], I might consider a segment. But in general, let me make it clear, that we only did a segment in this series, 99% of the time because their pulmonary function tests were depressed. And my idea of low or poor PFT's is probably quite different than yours. I mean, an FEV1 [forced expiratory volume in 1 second] of less than 30%, a DLCO [diffusion of the lung for carbon monoxide] of less than 30%, we routinely will operate and do lobes, especially if the tumor is in the upper lobes if there is emphysema there.
DR ITO: Thank you very much. And one more question. Is there any differences in survival between lobectomy and segmentectomy?
DR CERFOLIO: The survival was not different, but the local recurrence was higher in those who had a segment compared to those who had a lobectomy.
DR ITO: Thank you very much.
DR PAUL C. LEE (New York, NY): I really enjoyed your talk. A couple of questions.
Of the 715 clinically stage I lung cancer patients, it seems like your rate of benign disease is quite high at 100 patients. How hard do you try to get a tissue diagnosis before you take those patients to the operating room either by a needle biopsy, bronchoscopy?
And in how many of those patients do you end up doing a lymphadenectomy that might not be necessary at the time of surgery?
DR CERFOLIO: How many had a what?
DR LEE: How many patients did you end up doing a lymphadenectomy, a lymph node dissection at the time of surgery?
DR CERFOLIO: So your first question is how often do I do a needle? You see the answer. I think a needle is a big waste of time.
Now, sure, I have taken some benign nodules out, but these were patients with growing pulmonary nodules on CT [computed tomography], most were smokers and the lesion was also PET [positron-emission tomography]-positive so if the needle was negative would you really watch them? The bottom line is I am trying to convince many pulmonologists that the needle does not change the management. For example, if you had a 72-year-old guy who smoked his brains out for 40 years, he's got a growing nodule on CT, and he's PET-positive. You can stick a needle in it all day long. I am going to take it out, assuming the mediastinum is quiet or silent. So we don't do a lot of needles. The argument that the needle is done to ensure it is no small cell is silly, because the mediastinum is almost always full of metastatic disease, and thus; it has to be assessed first, and it if is negative, then this may be the one time a resection may help a patient with small cell.
Now, some of those you may start off with the VATS [video-assisted thoracoscopic], and for this study we threw those people out. So these were suspicious nodules that were not VATable. Most of them were active histo, some were TB [tuberculosis], but they were all were PET-positive, I think, except some of them that were growing on CT scans. But it's a high rate; I concur.
The second question is about the lymphadenectomy. I am not patient in the operating room, so I will be honest, the nodule gets wedged out, and I'll have the lymph nodes all removed by the time I get the report back. Did we get burned with that? I don't think so.
There is a risk with lymphadenectomy, and we do a complete lymph node removal. I remember I was telling someone out in the hall, I remember a nurse, that I got an apraxia on a T1 N0 M0 lesion of her recurrent laryngeal nerve that was out for 3 months.
So I got my complications and a few chylothoraces, but I think they are very minor.
DR LEE: Thank you.
DR MALCOLM M. DECAMP (Boston, MA): So, Cerf, I am going to return the favor from our SUV [standard uptake value] paper. Do you have data about how many occult N1 and N2 patients? You are telling us that you are refining the denominator down to absolute N0 patients by doing lymphadenectomies, but you are doing these people whose PET scans show the mediastinum is quiet. How many patients in the IAs and IBs had occult nodal disease and then weren't part of this?
DR CERFOLIO: When you mean "occult," you mean there were ...
DR DECAMP: Like, their lymphadenectomy had disease where their PET scan didn't.
DR CERFOLIO: Yes. Well, there were 69 people in that series, so 69 had occult N2. Now, we just presented at this meeting ...
DR DECAMP: IA versus IB, big tumors, little tumors, evenly across the, ...
DR CERFOLIO: Oh, yeah, I see what you are saying. I am not too smart, now I get your question. I don't know if I can tell you the difference based on the T status for this study, but we have reported lot so data in our studies that the maxSUV and the size do predict the risk of unsuspected N1 and N2 disease. For this study, I'd have to go back and look at that. We didn't look at that.
DR DECAMP: Because a bunch of people have demonstrated that in the past. As T stage goes up, our absolute size changes.
DR CERFOLIO: Sure.
DR DECAMP: That the occult N2 rate is going to go up.
DR CERFOLIO: It is a great point. I don't know the answer to that, but we should go back and look at that.
DR MICHAEL S. MULLIGAN (Seattle, WA): Cerf, two quick questions. One, did you see any differences in your IIBs that were either IIB because there was visceropleural invasion or by size?
The second question is, does palpation really add anything? There are probably four or five good VATS series that demonstrate comparable survival to yours for stage I disease.
DR CERFOLIO: So your second question first. It is a great point. I don't know if it adds anything because the question is, are these nodules we are finding, are they clinically significant? Yes, they are malignant, but does it make a difference? You know, maybe you save the person a little anxiety when they come back at 1 or 2 years and that nodule has grown. They are going to have a reoperation, that has risk, cost, and mental anguish.
But does that impact on survival if they have close follow-up? I don't know. They have to get another operation. They have to get another work-up, more expense, repeat PET, maybe another procedure with risk. But you are right, we don't know the clinical significance.
Your other question is very good, and I took that out for time's sake. We did a subanalysis of the T2 lesions looking at visceropleural invasions, and we did a subanalysis of the T1s looking less than 2 cm or greater. But we did find a difference that the visceropleural invasion is a bad actor, and it was worse. And the T1s that were less than 2 cm did better than those that were T2 to T3. So the new classification, the T1a versus TIb is real and works for our data.
EVELYN M. BARAM-CLOTHIER (Philadelphia, PA): Did you have any interest in following CEA [carcinoembryonic antigen] levels during pre- and post-lobectomy?
DR CERFOLIO: CEA?
MS BARAM-CLOTHIER: Yes.
DR CERFOLIO: No, I have not. I mean, some of our pulmonologists and oncologists really believe in that. I don't do that routinely in my practice.
MS BARAM-CLOTHIER: Okay. Fine.
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. C. Tomaszek, Y. Kim, S. D. Cassivi, M. R. Jensen, K.-H. R. Shen, F. C. Nichols, C. Deschamps, and D. A. Wigle Bronchial resection margin length and clinical outcome in non-small cell lung cancer Eur J Cardiothorac Surg, November 1, 2011; 40(5): 1151 - 1156. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. J. Cerfolio, A. S. Bryant, and D. J. Minnich Starting a Robotic Program in General Thoracic Surgery: Why, How, and Lessons Learned Ann. Thorac. Surg., June 1, 2011; 91(6): 1729 - 1737. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. J. Cerfolio, A. S. Bryant, M. A. Eloubeidi, P. A. Frederick, D. J. Minnich, K. C. Harbour, and M. T. Dransfield The True False Negative Rates of Esophageal and Endobronchial Ultrasound in the Staging of Mediastinal Lymph Nodes in Patients With Non-Small Cell Lung Cancer Ann. Thorac. Surg., August 1, 2010; 90(2): 427 - 434. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. D. Timmerman Surgery Versus Stereotactic Body Radiation Therapy for Early-Stage Lung Cancer: Who's Down for the Count? J. Clin. Oncol., February 20, 2010; 28(6): 907 - 909. [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 |
