Surgical resection is the primary curative treatment modality

Transcription

1 ORIGINAL ARTICLE Use of a Surgical Secimen-Collection Kit to Imrove Mediastinal Lymh-Node Examination of Resectable Lung Cancer Raymond U. Osarogiagbon, MBBS, FACP,* Laura E. Miller, MD, Robert A. Ramirez, DO,* Christoher G. Wang, MD, Thomas F. O Brien, MD, Xinhua Yu, MD, PhD, Alim Khandekar, MD, Glenn P. Schoettle, MD, Samuel G. Robbins, MD, Edward T. Robbins, MD,# and Jeffrey B. Gibson, MD Introduction: Pathologic examination of mediastinal lymh nodes (MLNs) after resection of non small-cell lung cancer is critical in the determination of rognosis and ostoerative management. Although systematic nodal dissection is recommended, the quality of athologic lymh-node staging often falls short of recommendations in ractice. We tested the feasibility of imroving athologic lymhnode staging of resectable non small-cell lung cancer by using a relabeled secimen-collection kit. Methods: Case-control study with comarison of 51 resections, using a secial lymh-node collection kit, with 51 controls matched for surgeon, extent of resection, athologist, and T category. Aroriate statistical methods were used for all comarisons. Results: The median number of MLNs examined increased from one in the control grou, to six in the case grou ( < 0.001). The ercentage of resections attaining the National Comrehensive Cancer Network-recommended quality of MLN examination, and the roortion that would have been eligible for recent landmark ostresection adjuvant theray trials increased significantly ( < 0.001). The duration of surgery and ostoerative comlication rates were similar between cases and controls. Eighteen ercent of kit cases had ositive MLN, comared with 8% of controls. Conclusions: The use of a secialized secimen-collection kit for MLN examination was feasible, markedly imroved MLN staging, *The Multidiscilinary Thoracic Oncology Program, Boston Baskin Cancer Foundation, Batist Cancer Center, Memhis, TN; Deartment of Medicine, University of Tennessee, Memhis, TN; Duckworth Pathology Grou, Memhis, TN; School of Public Health, University of Memhis, Memhis, TN; Division of Cardiothoracic Surgery, University of Tennessee, Memhis, TN; The Cardiovascular Center, Methodist Germantown Hosital, Memhis, TN; and #Cardiothoracic Surgery Associates, Memhis, TN. Suorted by a grant from the Methodist Healthcare Foundation to Dr. Osarogiagbon. Disclosure: Dr. Osarogiagbon has a atent alication ending for the lymhnode secimen-collection kit. The other authors declare no conflict of interest. Address for corresondence: Raymond U. Osarogiagbon, MBBS, FACP, Multidiscilinary Thoracic Oncology Program, Boston Baskin Cancer Foundation, 6027 Walnut Grove Road, Suite 201, Memhis, TN rosarogi@bmg.md ISSN: /12/ and showed a trend toward increased detection of atients with MLN metastasis, with only a modest increase in duration of surgery, and no increase in erioerative morbidity, mortality, or hosital length of stay. Key Words: Staging, Non small-cell lung cancer, Mediastinal lymh nodes, Quality imrovement, Surgical resection. (J Thorac Oncol. 2012;7: ) Surgical resection is the rimary curative treatment modality for non small-cell lung cancer (NSCLC). In the United States, 29% of all atients diagnosed with NSCLC undergo curative-intent surgical resection, an annual case volume of about 60,000 resections. 1,2 Although the vast majority of longterm survivors undergo surgery, only 73% of atients with stage IA, 58% with IB, 46% with IIA, 36% with IIB, and 22% with IIIA NSCLC survive to 5 years. 3 Nodal metastasis sequentially diminishes ostresection 5-year survival rates from 56% in atients with N0 disease to 38% in those with N1, 22% with N2, and 6%with N3. 4 The N-category is thus a owerful determinant of survival in resectable NSCLC. Careful examination of the surgical-resection secimen for lymh-node metastases is imortant for rognostication and selection of ostoerative adjuvant theray. 5 8 Accurate collection and analysis of lymh nodes during ulmonary resection is imerative to rovide atients with high-quality care. Examination of most N1 lymh nodes, located within the resected lung, can be regarded as rimarily the resonsibility of the athologist. Examination of hilar and mediastinal lymh nodes (MLN), however, deends on the collaborative efforts of the surgeon and athologist. The relevant lymh nodes must be collected consistently, and labeled using standard nomenclature before they can be examined effectively. There is consensus on the need for examination of lymh nodes from certain stations, including hilar and subcarinal stations in all atients, the lower right aratracheal nodes in those with right-side tumors, and the subaortic and ara-aortic nodes in those with left-side tumors. 9,10 Random examination that fails to encomass these minimum stations and, worse, no examination of MLNs, is associated with significantly worse survival, robably because of understaging Journal of Thoracic Oncology Volume 7, Number 8, August 2012

2 Journal of Thoracic Oncology Volume 7, Number 8, August 2012 Secimen Collection Kit Imroves Staging Our detailed investigation of MLN-examination ractices after lung resection in the Memhis Metroolitan Area revealed quality deficits in the current ractice: 59% of cases did not have hilar nodes, 42% had no MLNs, and 90% had fewer than three MLN stations. 14 Furthermore, 92% of the resections failed to achieve minimum recommended MLNexamination standards in an audit of athology reorts, and 70% of surgical oeration notes described a subotimal MLNexamination rocedure in a blinded audit by an indeendent surgeon. 15 In resonse to these findings, we designed a secial surgical lymh-node secimen-collection kit to imrove the intraoerative collection of lymh nodes during lung resection surgery. Our aim was to imrove the accuracy and quality of MLN staging during surgical resection for lung cancer. MATERIALS AND METHODS We conducted a case-control study with the aroval of the Institutional Review Board of the University of Tennessee (roject aroval number XP), which waived the informed consent requirement for this quality-imrovement study. All oerations were erformed by board-certified cardiothoracic surgeons. We obtained all lymh-node examination information from the final athology reort and abstracted clinical data, including information about atients intra- and ostoerative exerience from oerating-room logs and other clinical records. Patients who had received reoerative chemotheray or radiation theray were ineligible. Data on the kit cases were collected rosectively. Secimen-Collection Kit We designed a self-contained kit with secimen-collection cus relabeled with the standard anatomic name and number of each of the following 12 lymh-node stations: right uer aratracheal (2R), left uer aratracheal (2L), revascular (3a), retrotracheal (3), lower right aratracheal (4R), lower left aratracheal (4L), subaortic (5), ara-aortic or hrenic (6), subcarinal (7), araesohageal (8), ulmonary ligament (9), and hilar (10). One set of kits was labeled for use in right, another set for left, lung resections. In the right-side kits, secimen cus for stations 2R, 4R, 7, 8, 9, and 10R were consicuously marked as mandatory for samling, whereas stations 4L, 5, 6, 7, 8, 9, and 10L were marked as mandatory in the left-side kits. We used the International Association for the Study of Lung Cancer lymh-node ma for lymh-node station nomenclature. 4 Controls We designed a set of controls to enable us account for otentially confounding factors during our study. We created a historical control cohort using a redetermined hierarchical set of matching criteria in the following riority order: surgeon, extent of resection, athologist, and American Joint Committee on Cancer T category. When multile matches were available, we selected the control closest in time to the case. We selected the historical controls from the Memhis Metroolitan Area Quality of Surgical Resection database of lung resection cases from 2004 to We recruited six surgeons to articiate in the study. All other surgeons who oerated at the study institution were identified as nonarticiating surgeons. To adjust for confounding unrelated system-wide ractice changes during the study eriod, we also comared lung cancer resections erformed by articiating surgeons within the year immediately receding kit deloyment with cases erformed by the same surgeons using the kit. In addition, we monitored resections erformed by nonarticiating surgeons before and after deloyment of the kit in the study institution. Finally, we evaluated resections erformed by articiating surgeons after the onset of kit deloyment but which, although eligible, were erformed without the kit because of inadvertent logistic failure of kit requisition. The data on all control oerations erformed after the time of study onset were collected rosectively. Statistical Analysis We used descritive analysis to comare atient characteristics, lymh-node examination characteristics, and erioerative comlications between cases and controls. Unadjusted statistical significance was assessed using McNemar s matched χ 2 test for categorical variables, and matched t test for continuous variables. Because our study matched cases with controls in a hierarchical order by surgeon, extent of resection, athologist, and T category, we used matched Poisson regression for count variables (e.g., number of lymh nodes examined and number of lymh nodes with metastasis), and matched logistic regression for binary variables (e.g., yes/no status of MLN examined, lymh nodes with metastasis, and surgery comlications), and matched linear regression for continuous variables. Age, race, sex, and insurance status were also included in the multivariate model. For unmatched analysis, we used the Pearson χ 2 test, t test, and Poisson regression. We used Stata 12 t module for statistical analysis. RESULTS From November 8, 2010 to October 11, 2011, 51 atients underwent surgical resection with intraoerative collection of their MLN secimen with the lymh-node collection kit (kit cases). We matched these cases with 51 historical controls. The demograhic, clinical, and tumor-resection characteristics of the cases and controls were very similar (Table 1). All atients had resection with negative margins. Number and Distribution of Examined Lymh Nodes Significantly more hilar, mediastinal, and total lymh nodes were examined in the kit cases (Table 2). All kit cases had at least one MLN examined, whereas 49% of the controls had no MLN examined, including 6% with neither N1 nor N2 lymh nodes. The imrovement in lymh-node examination was largely because of an increase in the number of stations from which MLN were collected, from one in the controls to four stations in the cases. The majority of the cases achieved samling of lymh nodes from stations 7, 8, and 9, whereas these stations were examined in less than 25% of the controls. Station 10 was 1277

3 Osarogiagbon et al. Journal of Thoracic Oncology Volume 7, Number 8, August 2012 TABLE 1. Comarison of Key Characteristics Between Cases with Surgical Kits and Matched Controls Variable a N = 51 N = 51 Case Control Patient demograhics Age, median (range) 67 (49 87) 64 (36 82) 0.06 Male sex (%) Race (%) 0.40 African American White Insurance status (%) 0.76 Third arty (± Medicare) Medicare No insurance 10 6 Disease characteristics Histology (%) 0.51 Adenocarcinoma Squamous Other 14 8 Differentiation (%) 0.51 Well/moderate Poorly/undifferentiated Not reorted Tumor size, median 2.8 ( ) 2.5 ( ) 0.08 (interquartile range) T category (%) / Preoerative staging evaluation (%) CT scan PET-CT scan <0.001 Preoerative mediastinoscoy Resection characteristics (%) Extent Pneumonectomy Bilobectomy 12 8 Lobectomy Margin negative status (%) Technique (%) Oen thoracotomy Video-assisted thoracic surgery (VATS) CT, comuted tomograhy; PET, ositron emission tomograhy. a All variables reorted as ercentage of cohort, unless stated otherwise. examined in 43% of controls, comared with 76% of cases ( = 0.009). In right-side cases, examination of stations 2R and 4R increased significantly ( = and <0.001, resectively); in left-side cases, examination of stations 5 and 6 also imroved ( = 0.02 and 0.06, resectively). Examination of lymh nodes from stations not tagged as mandatory remained infrequent. For examle, only 3 cases (6%) had samling of lymh nodes from station 3, no controls had lymh nodes examined from this station, and no cases or controls had lymh nodes from contralateral mediastinal stations. TABLE 2. Comarison of the Number of Lymh Nodes Examined Case Control N = 51 N = 51 Number of lymh nodes examined All cases, median (IQR) From N1 stations 5 (3 7) 3 (1 6) From N2 stations 6 (4 9) 1 (0 3) <0.001 From all stations 12 (9 15) 5 (2 8) <0.001 From oen cases N1 5 (4 7) 3 (1 6) N2 6 (4 9) 0 (0 2) <0.001 All 13 (9 15) 5 (2 8) <0.001 In VATS cases N1 5 (2 6) 4 (2 5) 0.66 N2 5 (3 6) 3 (2 3) 0.33 All 9 (7 10) 7 (6 8) 0.29 Number of N2 stations examined All cases Median 4 0 <0.001 Interquartile range (IQR) Range Oen thoracotomy cases, 4 (3 5) 0 (0 1) <0.001 median (IQR) VATS cases, median (IQR) 4 (3 4) 2 (1 2) <0.001 Examination of mandatory stations (%) All cases < Right-sided cases only N = 34 N = 32 2R R <0.001 Left-sided cases only N = 17 N = Proortion of atients without lymh nodes (%) From N1 stations From N2 stations 0 43 <0.001 From anywhere Detection of lymh node metastasis Number of lymh nodes with metastasis median (range) N1 stations 0 (0 1) 0 (0 10) 0.67 Station 10 0 (0 4) 0 (0 10) 0.40 N2 stations 0 (0 5) 0 (0 3) 0.26 Total 0 (0 7) 0 (0 11) 0.19 Number of N2 stations with metastasis 0 (0 2) 0 (0 3) 0.38 IQR, interquartile range; VATS, video-assisted thoracoscoic surgery. The Influence of Technique of Resection on Lymh-Node Examination Most atients had oen thoracotomy, but 12% of kit cases and 10% of controls had video-assisted thoracoscoic surgery. Kit cases with oen thoracotomy had a median of 6 MLN 1278

4 Journal of Thoracic Oncology Volume 7, Number 8, August 2012 Secimen Collection Kit Imroves Staging TABLE 3. Comarison of Kit Cases to Various Controls to Account for Unrelated Secular Changes in Practice Particiating Surgeons Before Kit ( ) N = 33 After Kit N = 51 Number of lymh nodes, median (range; IQR) N1 4 (1 15; 2 7) 5 (1 13; 3 7) N2 3 (0 15; 1 5) 6 (2 19; 4 9) <0.001 Total 7 (1 27; 4 12) 12 (4 30; 9 15) <0.001 Number of N2 stations 1 (0 4; 0 2) 4 (1 6; 2 5) <0.001 Kit Cases vs. Concurrent Nonarticiating Surgeons Cases Kit Cases N = 51 Nonarticiating Surgeons Cases After Kit Deloyment, N = 14 Number of lymh nodes, median (range; IQR) N1 5 (1 13; 3 7) 4 (0 18; 1 4) N2 6 (2 19; 4 9) 1 (0 14; 0 3) <0.001 Total 12 (4 30; 9 15) 6 (1 18; 2 8) <0.001 Number of N2 stations 4 (1 6; 2 5) 0 (0 4; 0 1) <0.001 Nonarticiating Surgeons Before Kit Deloyment Date N = 10 After Kit Deloyment Date N = 14 Number of lymh nodes, median (range; IQR) N1 3 (2 7; 2 5) 4 (0 18; 1 4) 0.49 N2 2 (0 9; 1 7) 1 (0 14; 0 3) 0.04 Total 6 (2 14; 4 9) 6 (1 18; 2 8) 0.08 Number of N2 stations 1 (0 2; 0 1) 0 (0 4; 0 1) 0.87 IQR, interquartile range. (interquartile range [IQR], 4 9), comared with 0 (0 2) nodes in controls ( < 0.001). A median of four MLN stations were examined in the kit cases with oen thoracotomy comared with a median of zero stations in controls ( < 0.001). Although the number of lymh nodes examined in video-assisted thoracoscoic surgery resections was similar between cases and controls, the number of MLN stations from which lymh nodes were obtained increased from a median of two in controls, to four stations in kit cases ( < 0.001) (Table 2). Imact of Secimen-Collection Kit on Particiating Surgeons Lymh-Node Counts To account for unrelated institutional changes in quality of surgical lymh-node staging, we comared the kit cases with several other controls. In 33 resections erformed by articiating surgeons in the 12 months immediately receding study activation, a median of four (IQR, 2 7) N1, three (IQR 1 5) N2, seven (IQR 4 12) total lymh nodes were examined along with one (IQR 0 2) MLN station. These were all significantly less than the kit cases ( = for the N1 comarison and < for all others). The number of lymh nodes and lymh-node stations examined were significantly higher in the kit cases than in the 14 cases erformed by nonarticiating surgeons during the time of the study. The attern and quality of lymh-node examination in nonarticiating surgeons cases did not imrove during the time of our study. Finally, there were eight cases erformed by articiating surgeons after the date of study commencement, which would have been eligible for use of the kit, but in which logistic glitches FIGURE 1. Particiating surgeon cases in the year immediately receding kit introduction and after kit introduction (with and without use of the kit). with kit deloyment revented use of the kit. The number of lymh nodes examined in these oststudy activation nonkit cases by articiating surgeons was significantly less than those in the kit cases (Table 3). Indeed, it was very similar to the resections erformed in the year immediately receding the study (Fig. 1). Attainment of Recommended Quality Criteria Fifty-seven ercent of kit cases met the International Association for the Study of Lung Cancer criteria (Table 4) for comlete resection, defined as the resence of all the following: negative resection margin, examination of at least three lymh nodes each from N1 and N2 stations, examination of a minimum of six total lymh nodes, lymh nodes from at least 1279

5 Osarogiagbon et al. Journal of Thoracic Oncology Volume 7, Number 8, August 2012 TABLE 4. Attainment of Quality Mediastinal Lymh-Node Examination Criteria Proortion Attaining Criteria Criteria Case Control RADIANT (>1 N2 lymh node <0.001 stations) NCCN (>2 N2 lymh node stations) <0.001 ECOG 63 8 <0.001 Right: 4R <0.001 Left: ACOSOG Systematic samling 26 0 <0.001 Right: 2R + 4R R Left: L Mediastinal lymh node dissection 20 0 <0.001 Right: 2R + 4R R Left: L Resection status a Comlete Indeterminate ACOSOG, American College of Surgeons Oncology Grou; ECOG, Eastern Cooerative Oncology Grou; NCCN, National Comrehensive Cancer Network; RADIANT, OSI Pharmaceuticals RADIANT study. a As defined by International Association for the Study of Lung Cancer 9 and Euroean Society of Thoracic Surgeons, 10 comlete resection must meet all the following criteria: negative resection margin, examination of at least three lymh nodes each from N1 and N2 stations, examination of a minimum of six total lymh nodes, lymh nodes from at least three different N2 stations, at least one lymh node from station 7 and negative highest mediastinal lymh node. Indeterminate resection: negative margins, with one or more of other criteria not met. three different N2 stations, at least one lymh node from station 7 and negative highest MLN. Only 4% of the controls achieved this ( = 0.03). Majority of the controls were classified as indeterminate resections. Achievement of good quality resection criteria was significantly imroved in the kit cases ( < 0.001). For examle, the National Comrehensive Cancer Network recommendation for examination of lymh nodes from at least three mediastinal stations was met in 76% of kit cases, comared with 12% of controls; 94% of kit cases would have met surgical-quality eligibility criteria for the RADIANT trial, whereas 73% of controls would have been disqualified by the requirement for examination of at least two MLN stations. Furthermore, although 20% of the kit cases met the American College of Surgeons Oncology Grou definition of MLN dissection, no control atient attained this. Detection of Lymh-Node Metastasis There were few MLNs with metastasis detected (Table 5). The median number of MLNs with metastasis was zero in both cases and controls (95th ercentile 2 versus 1, range, 0 5 versus 0 3, resectively). Nine kit cases (18%) had MLN metastasis, comared with four controls (8%). Imact on Oerative and Postoerative Comlications Other than a median increase in oeration time of 27 minutes, there was no difference in intra- or ostoerative TABLE 5. Comarison of Stage Distribution and Imact on Eligibility for Postoerative Adjuvant Theray Case Control Nodal stage distribution (%) 0.23 N N N Nx (not determined) 0 6 Overall stage distribution (%) 0.27 I II III Eligibility for ostoerative adjuvant theray (%) Chemotheray Radiation TABLE 6. Oerative and Postoerative Comlication Rates and Duration of Hosital Admission Clinical Exerience Case Control Duration of surgery (minutes) Median Interquartile range Range Estimated blood loss (cc) Median Interquartile range Range Intraoerative blood transfusion units Median (range) 0 (0 2) 0 (0 3) 0.50 Duration of hosital admission (days) Median Interquartile range Range Duration of chest tube drainage (days) Median Interquartile range Range Duration of ICU admission 0.77 (days) Median 3 3 Interquartile range Range Frequency of ostoerative comlications (%) Arrhythmia Myocardial infarction Pneumonia day mortality 2 4 ICU, intensive care unit. exerience or comlication rates between the cases and controls. Bleeding (as judged by estimated blood loss or use of blood transfusion), length of stay in the intensive care unit or hosital, duration of chest tube drainage, ostoerative cardioulmonary comlication rates, and 30-day mortality rates, were similar (Table 6). 1280

6 Journal of Thoracic Oncology Volume 7, Number 8, August 2012 Secimen Collection Kit Imroves Staging DISCUSSION We demonstrate the feasibility of imroving athologic staging of lung cancer by means of a secifically designed lymh-node secimen-collection kit. The kit was readily adoted by the cardiothoracic surgery oerating-room team. Imrovements included elimination of nonexamination of MLNs, a general increase in the number of lymh nodes examined, increase in the number of stations from which a lymh-node secimen was obtained, and increase in likelihood of examination of lymh nodes from certain secific stations, which are widely agreed to be required for the accurate staging of resected lung cancer. The result of all this was a significant imrovement in the quality of athologic nodal staging. This was achieved at the exense of a median 27-minute increase in the duration of the surgical rocedure, but no increase in erioerative comlication rates or duration of intensive care unit or hosital admission. The imact on stage distribution, which a feasibility study such as this is not owered to demonstrate, although not statistically significant, is interesting. The roortion of atients with N2 disease increased from 8% to 18%, and the roortion of atients who would have been identified as otential beneficiaries of ostoerative adjuvant chemotheray or radiation theray, increased. Imortantly, in this era of intense research into ways of imroving adjuvant-theray results, disqualification of atients from eligibility for contemorary ostoerative adjuvant theray trials, such as the recently comleted RADIANT study 16 or the ongoing Eastern Cooerative Oncology Grou 1505 study, 17 because of subotimal MLN examination, would have been significantly reduced by use of this simle device (Table 4). Use of the kit was associated with imrovement in surgeon erformance in the oerating room, and imrovement in the quality of the athology examination. There aeared to be a causal relationshi between use of the kit and imrovement in the quality of lymh-node examination. This observation is suorted by the regression in quality of lymh-node examination in eligible cases erformed without the kit by articiating surgeons, to the level in the eriod immediately receding introduction of the kit (Fig. 1). We attribute this henomenon to the fact that the quality of lymh-node examination is influenced by the combination of surgeon factors, athology ractice, and less obvious factors such as the adequacy of secimen identification, and security of the chain of secimen handling between the oerating room and the athology laboratory. 18 The kit is designed to rectify otential roblems in each of these utative quality breakdown sites. Correct anatomical maing of the lymh nodes was done in all the kit cases. Prelabeling the kit imroved the communication between surgeon and athologist, which is vital to a thorough and anatomically accurate final athology reort. However, a few roblems remain. The attern of lymh-node examination in the kit cases seems more similar to systematic samling than a MLN dissection in most cases, and few cases included lymh nodes beyond the minimum mandated stations. However, the results of the American College of Surgeons Oncology Grou Z0030 study, in which there was equivalent survival between atients randomized to systematic samling or MLN dissection suggests that this issue should not cause much concern. 19 Furthermore, the kit will accommodate the roduct of a more extensive dissection. Pathology errors ersist. For examle, in one case, the athologist erroneously attributed a documented ositive N2 lymh node as N1 (understaging error). However, such errors are easily recognized when athologists use anatomic nomenclature to identify every lymh node s station of origin. Limitations of our study include the case-control design and relatively small samle size. We attemted to minimize the otential for bias by comaring our intervention cohort with several different controls. We comared each kit case with controls that we matched for key otentially confounding factors including surgeon, extent of resection, athologist, and T category. We also attemted to control for indeendent secular changes in ractice by comaring kit cases with nonkit cases erformed by articiating surgeons in the 12 months immediately receding deloyment of the kit. Finally, we used nonarticiating surgeons cases as an external control to detect any unrelated system-wide imrovement in lymhnode staging ractice. We found no imrovement in the nonarticiating surgeons cases during the time of our study. The key characteristics of the cases and controls were very similar, excet for the identified rate of reoerative ositron emission tomograhy (PET)/comuted tomograhy (CT) staging. We were only able to verify the use of reoerative PET/CT in 37% of controls, comared with 92% of kit cases. This discordance may reflect evolution of ractice, but may just as likely indicate our inability to identify all control atients who had a reoerative PET/CT scan outside the institution where their lung resection was done. It is imortant to oint out that any true bias from this discreancy (less thorough reoerative staging in the control atients) would be exected to lead to a higher rate of detection of athologic N2 disease in the control atients. This would tend to blunt the effect size of the imroved surgical MLN staging associated with use of the kit. Our reort may therefore be conservative. The relevance of our findings to surgeons oerating in high-volume centers of excellence can be debated. However, because most lung cancer surgery in the United States is erformed outside such institutions, our results are robably relevant to the majority of lung cancer surgeons in the country. 1 Besides, athologists in high-volume centers would robably areciate the increased recision in secimen identification and labeling rovided by the kit. Obviously, the main rationale for imroving hilar and MLN examination is to identify all atients with lymh-node metastasis. Although we doubled the detection of atients with N2 disease, this achievement was not statistically significant. Furthermore, we have rovided no data on atient outcomes. Our study was not designed to achieve these aims. Rather, we aimed to estimate the effect size of this corrective intervention, to estimate the samle size that would be required in a rosective randomized controlled trial that can examine these imortant questions. We are currently designing such a trial. However, the imrovement in quality of surgical lymhnode staging is incontrovertible. Routine use of this secimencollection kit can be justified as a quality imrovement tool. 1281

7 Osarogiagbon et al. Journal of Thoracic Oncology Volume 7, Number 8, August 2012 ACKNOWLEDGMENTS The authors thank Matthew Smeltzer for biostatistics suort. Suorted by a grant from the Methodist Healthcare Foundation to Dr. Osarogiagbon. REFERENCES 1. Little AG, Rusch VW, Bonner JA, et al. Patterns of surgical care of lung cancer atients. Ann Thorac Surg 2005;80: ; discussion Jemal A, Siegel R, Ward E, Hao Y, Xu J, Thun MJ. Cancer statistics, CA Cancer J Clin 2009;59: Goldstraw P, Crowley J, Chansky K, et al.; International Association for the Study of Lung Cancer International Staging Committee; Particiating Institutions. The IASLC Lung Cancer Staging Project: roosals for the revision of the TNM stage grouings in the forthcoming (seventh) edition of the TNM Classification of malignant tumours. J Thorac Oncol 2007;2: Rusch VW, Crowley J, Giroux DJ, et al.; International Staging Committee; Cancer Research and Biostatistics; Observers to the Committee; Particiating Institutions. The IASLC Lung Cancer Staging Project: roosals for the revision of the N descritors in the forthcoming seventh edition of the TNM classification for lung cancer. J Thorac Oncol 2007;2: The International Adjuvant Lung Cancer Trial Collaborative Grou. Cislatin-based adjuvant chemotheray in atients with comletely resected non-small-cell lung cancer. N Engl J Med 2004;350: Winton T, Livingston R, Johnson D, et al.; National Cancer Institute of Canada Clinical Trials Grou; National Cancer Institute of the United States Intergrou JBR.10 Trial Investigators. Vinorelbine lus cislatin vs. observation in resected non-small-cell lung cancer. N Engl J Med 2005;352: Pignon JP, Tribodet H, Scagliotti GV, et al.; LACE Collaborative Grou. Lung adjuvant cislatin evaluation: a ooled analysis by the LACE Collaborative Grou. J Clin Oncol 2008;26: Lally BE, Zelterman D, Colasanto JM, Haffty BG, Detterbeck FC, Wilson LD. Postoerative radiotheray for stage II or III non-small-cell lung cancer using the surveillance, eidemiology, and end results database. J Clin Oncol 2006;24: Rami-Porta R, Wittekind C, Goldstraw P; International Association for the Study of Lung Cancer (IASLC) Staging Committee. Comlete resection in lung cancer surgery: roosed definition. Lung Cancer 2005;49: Lardinois D, De Leyn P, Van Schil P, et al. ESTS guidelines for intraoerative lymh node staging in non-small cell lung cancer. Eur J Cardiothorac Surg 2006;30: Wu Y, Huang ZF, Wang SY, Yang XN, Ou W. A randomized trial of systematic nodal dissection in resectable non-small cell lung cancer. Lung Cancer 2002;36: Gajra A, Newman N, Gamble GP, Kohman LJ, Graziano SL. Effect of number of lymh nodes samled on outcome in atients with stage I nonsmall-cell lung cancer. J Clin Oncol 2003;21: Varlotto JM, Recht A, Nikolov M, Flickinger JC, Decam MM. Extent of lymhadenectomy and outcome for atients with stage I nonsmall cell lung cancer. Cancer 2009;115: Allen JW, Farooq A, O Brien TF, Osarogiagbon RU. Quality of surgical resection for nonsmall cell lung cancer in a US metroolitan area. Cancer 2011;117: Osarogiagbon RU, Allen JW, Farooq A, Wu JT. Objective review of mediastinal lymh node examination in a lung cancer resection cohort. J Thorac Oncol 2011;7: RADIANT. Available at htt:// ?term=RADIANT&rank=4. Accessed December 15, E1505 Trial Overview. Available at htt:// general/e1505_trial_overview.df. Accessed December 15, Makary MA, Estein J, Pronovost PJ, Millman EA, Hartmann EC, Freischlag JA. Surgical secimen identification errors: a new measure of quality in surgical care. Surgery 2007;141: Darling GE, Allen MS, Decker PA, et al. Randomized trial of mediastinal lymh node samling versus comlete lymhadenectomy during ulmonary resection in the atient with N0 or N1 (less than hilar) non-small cell carcinoma: results of the American College of Surgery Oncology Grou Z0030 Trial. J Thorac Cardiovasc Surg 2011;141: