| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2007;84:182-190
© 2007 The Society of Thoracic Surgeons
a Division of Cardiothoracic Surgery, Department of Surgery, University of Alabama at Birmingham, Birmingham, Alabama
b Department of Epidemiology, School of Public Health, University of Alabama at Birmingham, Birmingham, Alabama
Accepted for publication March 12, 2007.
* Address correspondence to Dr Cerfolio, Division of Cardiothoracic Surgery, University of Alabama at Birmingham, 1900 University Blvd., THT 712, Birmingham, AL 35294 (Email: rcerfolio{at}uab.edu).
Presented at the Forty-third Annual Meeting of The Society of Thoracic Surgeons, San Diego, CA, Jan 29–31, 2007.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
Methods: This study was a retrospective review of a prospective database of consecutive patients who had clinical or pathologic N1 non-small cell lung cancer (NSCLC) who underwent preoperative 2-[(18)F] fluoro-2-deoxy-D-glucose (FDG)-positron emission tomography (PET) scans and complete resection with thoracic lymphadenectomy.
Results: There were 135 patients (88 men). The 5-year disease-free rate was 55%. Kaplan-Meier analysis showed that poor differentiation (p = 0.036), multiple N1 stations (p = 0.010), and the lack of adjuvant chemotherapy (p = 0.039) were all associated with a shorter 5-year disease-free rate. Multivariate disease-free analysis demonstrated that only multiple stations (p = 0.002) were independently associated with recurrence. The overall 5-year survival was 48%. Univariate analysis showed that multiple nodes within one station (p = 0.016), multiple station involvement (p = 0.041), and lack of adjuvant chemotherapy (p = 0.039) and moderate-to-poor tumor differentiation (p = 0.046) were associated with decreased survival. Multivariate analysis found that multiple stations, multiple nodes, and lack of adjuvant chemotherapy were independent predictors of poor survival. Integrated PET-computed tomography (CT) was significantly more sensitive for staging N1 disease than dedicated FDG-PET (p = 0.032). Neoadjuvant chemotherapy given to 48 nonrandomized patients did not seem to impact disease recurrence or overall 5-year survival rates (p = 0.349).
Conclusions: Factors that predict a poor outcome in patients with resected N1 NSCLC are the involvement of multiple N1 stations, multiple N1 nodes, and the lack of adjuvant chemotherapy. Integrated PET-CT is more sensitive for detecting N1 disease then dedicated PET. These data may influence preoperative or postoperative therapy, or both.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
The purpose of this study was to review our experience and to identify our short-term and long-term results with patients with N1 disease who have been carefully clinically staged before surgery and staged intraoperatively with complete thoracic lymphadenectomy.
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
Patients with pathologic N1 (pN1) disease who also had N2 disease after thoracotomy were excluded, as were patients who had N2 disease proven before thoracotomy by mediastinoscopy, esophageal ultrasound with fine-needle aspirate (EUS-FNA), or endoscopic bronchoscopy with ultrasound-guided biopsy (EBUS).
The University of Alabama at Birmingham’s Institutional Review Board approved this study and the prospective database used for this study. Patient consent was obtained for entry into the prospective database.
Staging and Procedures
All patients were initially clinically staged using the T, N, and M, classification system [1]. A clinical stage was assigned for the patient after the CT scan, dedicated PET, or integrated PET/CT scan results by one physician (RJC). FDG-PET or PET/CT scanning was performed using a Discovery LS PET-CT Scanner, (GE, Milwaukee, WI) as previously described [3]. Biopsies were done of N2, N3, or M1 areas (maxSUV > 2.5) suggestive of disease, as previously described at length [4–6].
At the time of thoracotomy, complete thoracic lymphadenectomy was performed and defined as removal of all nodes during right thoracotomy from the 2R, 4R, 7, 8R, and 9R positions as well as the appropriate N1 nodes. On the left side, it was defined as removal of the 5, 6, 7, 8L, 9L, and appropriate N1 nodes. The 4L nodes were removed when possible. Hilar lymph nodes were labeled as 10 and peribronchial lymph nodes were labeled as 11, 12, 13, and 14, as shown in Figure 1. If a particular station had more than one node, they were marked separately despite being from the same station. If one lymph node was removed in four pieces, it was labeled as one lymph node and not four.
|
Disease-free survival was defined as a patient who was alive and had no evidence of recurrence at the end of the follow-up period. Recurrence was defined as biopsy-proven cancer that had the same cell type as the original NSCLC that was previously resected. It was divided into two types: systemic and locoregional. Systemic was defined as metastatic cancer of the same cell type in a different organ system with no other identifiable cause. Locoregional recurrence was defined as recurrent cancer within the ipsilateral chest or mediastinal lymph node chain, or both, that recurred within 2 years of the first operation (ie, was not a second primary).
Interval to death or recurrence of disease was defined as the time between the operation and death or discovery of locoregional recurrence by biopsy. Patients alive or disease free at the end of the follow-up period were censored.
Survival calculations were performed on the 87 patients who did not receive neoadjuvant therapy, unless otherwise specified. Patients with metastatic disease in more than one lymph node at a lymph node station were called "having multiple nodes." Patients with metastatic nodal disease that involved more than one station were called "having multiple stations." Patients with both were defined as having both.
Complete responders from neoadjuvant chemotherapy were defined as patients who underwent preoperative chemotherapy and had no viable cancer left in the removed specimen after resection. Partial responders were defined as those who had 10% to 90% cellular death of the removed specimen after neoadjuvant therapy. Adjuvant cisplatinum-based chemotherapy was offered to patients younger than 75 years of age who had pathologic stage IB or II disease or greater and who had sufficient creatinine clearance.
Data Collection and Follow-Up
Patients were actively followed up after pathologic staging and resection to the end of this study (June 2006). 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, including clinic letters, follow-up scans, hospital computer information systems, tumor registry, Social Security Death Index, and telephone calls and letters from oncologists and other physicians. All information was entered into our prospective database.
Statistical Methods
Data were imported from the prospective Excel database (Microsoft Corp, Redmond, WA) into SAS 9.0 (SAS Institute, Cary, NC). Kaplan-Meier survival analysis was used to identify the covariates that significantly affected survival. Cox-proportional hazards analysis was used to identify the independent predictors of survival and prognosis. Univariate predictors significant with a value of p 
0.10 were entered into a step-wise multivariable model. A two-tailed p 0.05 was considered to indicate a statistically significant outcome unlikely due to chance.
Sensitivity of PET scan was calculated for patients who went directly to surgery and did not receive neoadjuvant therapy. Patients who underwent preoperative chemotherapy were eliminated because a patient who was a complete responder may have had N1 disease at the time of the initial PET scan but was pT0 N0 M0 after resection. Sensitivity is defined by the number of patients who were true positive (predicted by PET scan to have cN1 disease and confirmed by pathology to have pN1) divided by the sum of patients with the disease. The sum of patients with the disease includes true positives and false negative (those patients with pathologic N1 disease who were PET negative). Accuracy was not calculated in this study because it is defined as the sum of the true negative and the true positives divided by the sum of the true negatives, true positives, false negatives, and false positives. Because not all of these numbers were included in this study population accuracy could not be calculated for this trial.
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
|
|
|
|
|
|
|
|
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
Most agree that biopsy-proven N2 disease before thoracotomy or definitive resection is best treated by neoadjuvant therapy; however, the ideal treatment for biopsy-proven N1 disease before resection is unknown. We reviewed our experience in attempt to identify factors that may predict which patients with N1 disease are at increased risk for recurrence or death after resection. We also had a unique chance to evaluate the results of neoadjuvant therapy for some patients with N1 disease because we prescribed this treatment in almost 50 patients who came to resection.
Recent technologic advances have increased our ability to diagnosis N1 disease before definitive resection. EUBS for N1 nodes has gained significant popularity. Although it is very useful in patients with a lung mass who do not have a tissue diagnosis and for biopsy of N2 nodes, its role in the patient whose lung mass has already been proven to be a NSCLC looking for N1 disease is yet undefined. Because there are no data to suggest neoadjuvant chemotherapy followed by resection offers improved survival over resection followed by adjuvant chemotherapy, the advantage of proving N1 disease before resection remains controversial. The role of EBUS and EUS-FNA are clear for pathologically proving N2 disease.
PET scanning has improved detection of N2 disease compared with CT scanning [8, 9]. One of the strengths of this study is the low likelihood of missed N2 disease, because all patients received an integrated PET/CT or PET scan, 93% received a mediastinoscopy, and many also had an EUS-FNA. The role of PET for detecting N1 disease remains questionable, however. In 2004, Reed and colleagues [10] reported results of the Z0050 trial that evaluated the efficacy of PET in staging 303 patients with NSCLC, where they found FDG-PET was 42% correct for the detection of N1disease compared with 13% for CT scan. Other studies have shown marginal accuracy ranging from 70% to 80% [8, 11]. Integrated PET/CT has been shown to be a more promising tool for the identification of N1 disease compared with dedicated FDG-PET [12]. Lardinois and colleagues [13] in 2003 also reported the superiority of integrated PET/CT versus dedicated PET in the clinical staging of patients with NSCLC.
A heterogenous group of patients present with N1 NSCLC. Factors that have been shown to influence the survival in patients with N1 disease include location and number of the N1 lymph node stations involved, macroscopic versus microscopic disease [14], the method of spread (direct extension or via lymphatics), and methodologic issues such as lead time bias and length time bias. Tanaka and colleagues [15] in 2001, Riquet and colleagues [16] in 1999, and van Velzen and colleagues [17] in 1996 showed hilar node (stations 10R, 10L) involvement portended a poor prognosis compared with patients with only pulmonary nodal involvement (stations 11, 12, or 13). Similarly in our series, patients with hilar involvement had a decreased overall and disease-free survival period than those with pulmonary nodal involvement alone. However, this difference did not achieve statistical significance because of the small number of patients in the latter group. Osaki and colleagues [14] in 2004 showed that number of involved nodes (two or more nodes) was a significant predictor of poor 5-year overall survival. We also found that multiple nodal involvement was a significant predictor of poor outcome for overall survival.
Of interest was our finding that adjuvant chemotherapy improved survival. Adjuvant chemotherapy was primarily used during the latter part of this study because it did not become the standard of care in our practice until January 2004 after the International Adjuvant Lung Cancer Trail (IALCT) results were published in the New England Journal of Medicine [18]. Our findings in this study are consistent with those from the larger prospective IALCT study, as well as with more recent trials such as the National Cancer Institute of Canada intergroup study JBR.10 (NCIC JBR10) [19] and the Adjuvant Navelbine International Trialist Association trial (ANITA) in 2005 [20] that demonstrated adjuvant chemotherapy increased survival after resection for patients with stages IB-III NSCLC.
The role of neoadjuvant chemotherapy for patients with N1 disease is more controversial. The S9900 study started in 1999. It was a multiinstitutional trial that was designed to randomize patients with stage IB or stage II NSLCC to receive resection alone or preoperative chemotherapy before resection. Approximately 300 patients were enrolled in this study, and 93 were from our practice. Unfortunately, the S9900 study was stopped early because of the publication of the studies mentioned above that proved the advantage of adjuvant chemotherapy. Although the results from this study remain unpublished, our own experience from the patients with N1 disease that we placed into that trial shows no advantage for neoadjuvant chemotherapy.
Similarly in this current study, 48 patients received preoperative chemotherapy. These patients were not randomized like those in the S9900 study and thus cannot be compared fairly with patients who did not get preoperative chemotherapy without the risk of a differential misclassification error. For example, we were more likely to offer preoperative chemotherapy to patients with bulky N1 disease than we would be to those with one small suggestive N1 node on PET/CT. Moreover, we were more likely to offer it to a younger patient or one whose primary lung mass had a high maxSUV.
With these limitations in mind, we compared the two groups and found no advantage for those 48 patients who received neoadjuvant chemotherapy compared with the 87 patients who did not. However, a significant survival advantage for the 8 patients who were complete responders of the chemotherapy was identified when compared with partial responders and nonresponders collectively. All complete responders are currently alive, without recurrence, and their median follow-up is 39 months (range, 28 to 64 months). Patients who were partial responders enjoy better survival as well, although it is not statistically significant. This finding fits with the findings of Pisters and colleagues [21] in 1993.21
In conclusion, factors that predict a poor outcome in patients with resected N1 NSCLC are the involvement of multiple N1 stations, multiple N1 nodes, and the lack of adjuvant chemotherapy. The identification of these factors may help direct postoperative treatment such as more aggressive adjuvant therapy or more diligent follow-up after resection. Integrated PET/CT is more sensitive for detecting N1 disease then dedicated PET. Adjuvant chemotherapy after resected stage II from N1 NSCLC is the standard of care; however, the role of neoadjuvant therapy for patients with N1 disease is less defined and unproven.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
DR BRYANT: Thank you for your questions. Patients who had clinically suspected N1 disease on PET/CT scan or biopsy-proven N1 disease went on to receive neoadjuvant therapy. All of these patients had N2 disease ruled out via mediastinoscopy and EUS-FNA. Following neoadjuvant chemotherapy, we restaged the patients using a repeat integrated PET/CT scan, and for some, repeat EUS-FNA. If they were N2-negative, they went on for a complete resection. Compete resection was defined as tumor removed, margin negative, bronchial margin negative, and all N2 and N1 nodes removed.
DR VAN SCHIL: So it could be that some of these patients initially had N2 disease?
DR BRYANT: No, they could not have. The only way they could have had N2 disease initially is if they had a false negative med and a false negative EUS-FNA, then got chemotherapy for assumed or proven N1 disease and that made their unproven N2 disease go away since patients with N2 disease after thoracotomy and lymph node resection are not in this study. That is the only way, and we believe that is very unlikely.
DR CERFOLIO: Ayesha is doing great as usual. Her answer is partially correct. Some of these patients that were given neoadjuvant chemo had bulky N1 and had EBUS-proven N1, but not all of them did. If they had obvious N1—you know, looking at the ratio of the maxSUV of the primary tumor to the maxSUV of the N1 node, and they were young and had bulky disease, we selected them to get chemo. Some had biopsy-proven N1; some did not. But it was bulky young patients. All got mediastinoscopy and EUS-FNA to rule out N2 disease to ensure that was negative. If they got neoadjuvant chemotherapy, we did not repeat mediastinoscopy, but we did repeat EUS-FNA or EBUS to ensure there was no N2. But if they had N1, we still proceeded to resect.
DR YOUNG TAE KIM (Seoul, Korea): I’ll direct my question to Dr Cerfolio. In your practice, how many cases do you think you lost among the patients you had identified N1 disease by EBUS? And how many patients did you find finally having true N1 disease?
DR CERFOLIO: I’m not sure what you mean by "lost." Because if they are N2-positive, we don’t consider them lost; but rather, we would go on to give them neoadjuvant chemoradiotherapy and then restage them with repeat EUS-FNA, and if down-staged, then we resect them.
DR KIM: I am asking selection of the N1 group. You selected some of N1 patient group with endoscopic bronchoscopy, right?
DR CERFOLIO: Correct. And that number is in the manuscript. But there was a significant number of patients, about 25% of patients, who had clinically suggested N1—did not have clinically suggested N2 after integrated PET/CT—but had pathologic unsuspected N2 after the med and/or EUS-FNA. We just presented this type of data in an article in the December issue of Chest. So about 25% to 30% were N2 and were not in the data you saw today.
DR KIM: I’m just asking, how many patients did you perform EBUS in and in how many of them did you miss N1 disease?
DR CERFOLIO: All patients clinically staged as N1 received EUS-FNA. All patients with suggested N1 disease in my practice get mediastinoscopy and EUS-FNA, not EBUS. EBUS was something we have just started doing very, very recently. And that is to really prove N1 and to look for N2 in people who are going to get neoadjuvant so I can do my mediastinoscopy after neoadjuvant. That’s the big advantage for EBUS in my opinion.
DR KIM: So, you don’t have conclusive data for EBUS at this point, do you?
DR CERFOLIO: No, we do not, but we are developing it. It will be quite accurate, we believe.
DR G. ALEXANDER PATTERSON (St. Louis, MO): How many of these patients were identified as N1 only on the basis of the final pathology and the resected specimen? In other words, you had no idea when you were in the operating room that the patient had N1 disease. Is that a significant number here?
DR CERFOLIO: It is a significant number, especially of the 87. Certainly of the 48, it’s none of those. But of the 87 people who only had pathologic N1, probably about 30% to 40% had unsuspected N1, single station N1, missed by dedicated PET, but more importantly, missed by integrated PET/CT. This type of microscopic N1 disease is quite different from bulky N1, just as can be said of N2 disease; hence, the heterogeneity of these patients.
DR CHING TZAO (Taipei, Taiwan): Congratulations on your excellent and interesting study, Dr Bryant. I have a quick question for you. In your investigation, you have found out that the prognostic factor included nodal station or multiple nodal involvements. Do you think it is worthwhile to analyze the nodal distribution in relation to prognosis as it might give us an insight on the whether patients presenting different locations of metastatic nodes may have different outcome.
DR BRYANT: Previous studies have shown that nodal station impacted outcome. However, in our patient population, nodal station was a nonsignificant risk factor for poor outcome. Keep in mind that only 18% to 20% of our patient population was positive at other nodal stations (stations 11, 12, and 13) besides the 10 station, thus making almost any association unlikely to be significant.
DR RICARDO S. SANTOS (Pittsburgh, PA): My question is in regard to the molecular staging and the micrometastases. Did you perform any further analysis to find any markers, molecular markers, or to identify the best responders to cisplatin-based therapy? Did you order those markers to refine staging? I know Dr Cerfolio is familiar that the International Lung Adjuvant Lung Trial (IALT) is moving towards a more comprehensive biological staging. I’d like to hear your comments about that.
DR BRYANT: In this trial, no, we did not evaluate for markers.
DR CERFOLIO: But we do have markers in our prospective database, in our last, I think, 300 or 400 lobectomies, and so that is coming. We are doing that now. We will have to look at it retrospectively, but we are putting the data in prospectively. We are sending all of our tumors off to Oncotech and we are getting all of those analyses at Oncotech.
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
