Abstract
Background.
The National Comprehensive Cancer Network guidelines recommend surgery for limited stage small cell lung cancer (SCLC). However, there is no literature on minimum acceptable lymph node retrieval in surgery for SCLC.
Methods.
The National Cancer Database was queried for adult patients undergoing lobectomy for limited stage (cT1-2N0M0) SCLC from 2004 to 2015. Patients with unknown survival, staging, or nodal assessment, and patients who received neoadjuvant therapy were excluded. The number of lymph nodes assessed was studied both as a continuous variable and as a categoric variable stratified into distribution quartiles. The primary outcome was overall survival and the secondary outcome was pathologic nodal upstaging.
Results.
A total of 1051 patients met study criteria. In multivariable analysis, only a retrieval of eight to 12 nodes was associated with a significant survival benefit (hazard ratio 0.73; 95% confidence interval, 0.56 to 0.98). However, when modeled as a continuous variable, there was no association between number of nodes assessed and survival (hazard ratio 1.00; 95% confidence interval, 0.98 to 1.02). The overall rate of pathologic nodal upstaging was 19%. Modeled as a continuous variable, more than seven lymph nodes assessed at time of resection was significantly associated with nodal upstaging in multivariable regression (odds ratio 1.03; 95% confidence interval, 1.01 to 1.06).
Conclusions.
In this study, there was no clear difference in survival based on increasing the number of lymph nodes assessed during lobectomy for limited stage SCLC. However, the number of retrieved lymph nodes was associated with pathologic nodal upstaging. Therefore, patients may benefit from retrieval of more than seven lymph nodes during lobectomy for SCLC.
The National Comprehensive Cancer Network guidelines recommend consideration of resection followed by adjuvant chemotherapy for clinical stage I-IIA (cT1-2, N0, M0) small cell lung cancer (SCLC).1 However, in comparison with non-small cell lung cancer, SCLC is much more likely to present with disseminated disease at the time of diagnosis, and there is wide variation between clinical and pathologic staging.2 As a consequence, a significant portion of patients are upstaged and found to have evidence of nodal disease after resection.3 The accuracy of pathologic staging is therefore critical given the limitations of clinical staging as patients with evidence of lymph node metastasis have a worse prognosis4 and may benefit from further intervention beyond adjuvant chemotherapy, including mediastinal radiation therapy or prophylactic cranial irradiation.5,6
Whereas the importance of accurate pathologic nodal staging in limited stage SCLC after resection is clear, no studies to date have provided evidence to guide adequate lymph node harvest at the time of surgery. We therefore undertook an analysis of the National Cancer Database (NCDB) to aid in defining the minimum acceptable lymph node retrieval in surgery for SCLC. We hypothesized that an increased number of assessed lymph nodes would be associated with improved survival and increased pathologic nodal upstaging in patients with limited stage SCLC.
Patients and Methods
Data Source
The NCDB is the result of a collaborative effort between the American Cancer Society and the American College of Surgeons. The database houses information collected by certified tumor registrars in 1500 hospitals. It reflects outcomes associated with nearly 80% of cancers diagnosed across the United States annually.7
Patient Selection
This study was deemed exempt by our Institutional Review Board. Using the NCDB, we identified patients diagnosed with the American Joint Commission on Cancer’s eighth edition clinical T1-2N0M0 SCLC who underwent lobectomy during a study period of 2004 to 2015. Patients who received neoadjuvant therapy or had missing data for staging, survival, or nodal assessment were excluded (Figure 1).
Figure 1.
Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) diagram of patients analyzed. (SCLC, small cell lung cancer.)
Study Design
The number of lymph nodes assessed had an approximately normal distribution (Figure 2A). The median number of lymph nodes harvested was 7 (interquartile range, 4 to 12). In the first part of the study, number of lymph nodes assessed was modeled as a categoric variable, and patients were stratified into four groups based on quartiles (Q): Q1 (0 to 4 nodes), Q2 (5 to 7 nodes), Q3 (8 to 12 nodes), and Q4 (more than 12 nodes). In the second part of the study, number of lymph nodes assessed was considered as a continuous variable. A restricted cubic spline transformation of the number of lymph nodes assessed as a function of survival using four prespecified knots revealed a possible inflection point at 7 nodes, beyond which survival appeared to improve (Figure 2B). As a result, the number of lymph nodes assessed was subsequently modeled in multivariable regression as a continuous variable using piecewise linear splines and a knot at 7 nodes. Restricted cubic splines are cubic polynomial transformations of functions that may be nonlinear or have nonlinear components, and have been shown to better approximate the relationship between the independent and dependent variable than linear models.8 When a knot is placed at a certain value, the value itself is not well approximated by the spline function; therefore, 7 was not strictly included in either spline function.
Figure 2.
(A) Histogram of number of lymph nodes assessed in the overall cohort. (B) Unadjusted restricted cubic spline transformation of number of lymph nodes assessed. Arrows denote prespecified knots selected in unadjusted analysis. The Y-axis demonstrates the unadjusted log hazard of mortality, and the X-axis the number of lymph nodes assessed. Dotted lines reflect bounds of the 95% confidence interval.
The primary outcome was overall survival. Continuous and categoric variables were compared using the Wilcoxon rank sum and Pearson’s χ2 tests, respectively. Survival was estimated using the Kaplan-Meier and Cox proportional hazards methods. The secondary outcome was the odds of pathologic nodal upstaging. That was studied using multivariable logistic regression. Variables included in the Cox and logistic regression models were selected a priori based on clinical significance and known prognostic importance from prior studies: age, sex, race, diagnosis year, Charlson-Deyo comorbidity index score, insurance status, academic or nonacademic type of treatment center, and tumor size. Because an “intention-to-treat” approach was used, only covariates known at the time of surgery were used in the Cox model. The proportional hazards assumption for variables in the Cox model, and the overall model was checked using visual and quantitative representations of Schoenfeld residuals; the proportional hazards assumption was not violated in any model (Schoenfeld P > .05).
We performed a separate analysis to explore the interaction between number of nodes assessed, receipt of adjuvant therapy, and survival. For this, we created a Cox regression model adjusted for age, sex, race, Charlson score, insurance status, pathologic tumor size (T status), pathologic N status, margin status, and receipt of adjuvant radiation, and included an interaction term between number of nodes assessed (modeled using restricted cubic splines with four prespecified knots) and receipt of adjuvant chemotherapy. Missing data were handled with complete case analysis in regression. A two-sided P value less than or equal to .05 was considered significant. All analyses were performed using R version 3.5.1 (R Foundation for Statistical Computing, Vienna, Austria).
Results
A total of 1051 patients met study criteria. The demographic characteristics of study patients, stratified by quartile, are summarized in Table 1. Patients in the highest quartile were more likely to have larger tumors at presentation and be treated at academic centers compared with patients in the lowest quartile. Although there was no difference based on quartile, only approximately 40% of patients in this cohort received adjuvant chemotherapy. The median number of lymph nodes harvested was 7. The median survival for patients in Q1, Q2, Q3, and Q4 was 54 months (95% confidence interval [CI], 42 to 69), 47 months (95% CI, 37 to 77), 61 months (95% CI, 45 to 88), and 45 months (95% CI, 37 to 55), respectively. There was no significant difference in survival between groups in our unadjusted model (Figure 3).
Table 1.
Demographic Characteristics of Study Patients
| Demographics | Quartile 1 (n = 295) | Quartile 2 (n = 237) | Quartile 3 (n = 283) | Quartile 4 (n = 236) | P Value |
|---|---|---|---|---|---|
| Median age, y | 66 | 67 | 68 | 67 | .41 |
| Sex, female | 176 (60) | 136 (57) | 152 (54) | 123 (52) | .28 |
| Race | |||||
| White | 257 (88) | 212 (90) | 266 (95) | 214 (91) | .16 |
| Black | 26 (9) | 18 (8) | 10 (4) | 15 (6) | |
| Other | 10 (3) | 5 (2) | 5 (2) | 7 (3) | |
| Year of diagnosis | 2007 (2003-2009) | 2007 (2004-2010) | 2008 (2005-2010) | 2008 (2004-2010) | .004 |
| CDCI score | .90 | ||||
| 0 | 109 (48) | 86 (45) | 110 (45) | 87 (45) | .90 |
| 1 | 89 (39) | 71 (38) | 98 (40) | 81 (42) | |
| 2+ | 31 (13) | 32 (17) | 36 (15) | 24 (13) | |
| Insurance status | |||||
| Private | 94 (32) | 86 (37) | 80 (29) | 76 (33) | .61 |
| Government | 193 (66) | 147 (63) | 196 (70) | 155 (66) | |
| None | 5 (2) | 2 (1) | 3 (1) | 3 (1) | |
| Academic/research program | 78 (26) | 75 (32) | 105 (37) | 110 (47) | <.001 |
| VATS | 16 (27) | 14 (29) | 22 (29) | 18 (33) | .90 |
| Tumor size, mm | 20 (15-27) | 20 (15-29) | 22 (16-28) | 25 (18-33) | <.001 |
| Pathologic nodal status | |||||
| N0 | 234 (84) | 200 (87) | 219 (82) | 169 (73) | .001 |
| N1 | 28 (10) | 23 (10) | 42 (16) | 42 (18) | |
| N2 | 17 (6) | 7 (3) | 7 (3) | 20 (9) | |
| 90-day postoperative mortality | 13 (7) | 16 (7) | 10 (4) | 11 (5) | .27 |
| Adjuvant chemotherapy | 94 (32) | 89 (38) | 125 (44) | 115 (49) | <.001 |
| Adjuvant thoracic radiation | 64 (22) | 60 (25) | 64 (23) | 67 (28) | .28 |
Values are n (%) or median (interquartile range) unless otherwise indicated. Quartile 1, 0 to 4 nodes; Quartile 2, 5 to 7 nodes; Quartile 3, 8 to 12 nodes; and Quartile 4, more than 12 nodes.
CDCI, Charlson-Deyo comorbidity index; VATS, video-assisted thoracoscopic surgery.
Figure 3.
Kaplan-Meier survival curves for patients with pT1-2N0M0 small cell lung cancer, stratified by number of lymph nodes assessed as grouped in quartiles (Q): Q1, 0 to 4 (red line); Q2, 5 to 7 (blue line); Q3, 8 to 12 (green line); and Q4, more than 12 (yellow). The P value represents the result of the log rank test.
The effect of number of lymph nodes harvested on survival and pathologic nodal upstaging was then modeled in a multivariable Cox proportional hazards regression. There was a significant survival benefit associated with retrieval of 8 to 12 lymph nodes at the time of surgery, although the other quartiles were not associated with survival (Table 2). The overall rate of pathologic upstaging across quartiles was 18.6%. Harvest of more than 12 lymph nodes was associated with a significantly higher rate of pathologic nodal upstaging (Table 3).
Table 2.
Cox Multivariable Regression of Factors Independently Associated With Survival With Number of Assessed Nodes Modeled as Categoric Variable by Quartile
| Variable | HR | 95% CI | P Value | |
|---|---|---|---|---|
| Lower | Upper | |||
| Age, per year | 1.02 | 1.01 | 1.04 | .001 |
| Sex, female | 0.85 | 0.69 | 1.04 | .11 |
| Race (reference: white) | ||||
| Black | 0.86 | 0.56 | 1.31 | .48 |
| Other | 1.21 | 0.71 | 2.04 | .49 |
| Year of diagnosis, per year | 1.03 | 0.98 | 1.08 | .21 |
| CDCI score (reference: 0) | ||||
| 1 | 1.03 | 0.83 | 1.28 | .79 |
| 2+ | 1.20 | 0.89 | 1.60 | .23 |
| Insurance status (reference: private) | ||||
| Government | 1.05 | 0.82 | 1.35 | .70 |
| None | 0.40 | 0.10 | 1.66 | .21 |
| Facility type (reference: nonacademic) | ||||
| Academic/research program | 1.05 | 0.85 | 1.30 | .66 |
| Tumor size, per mm | 1.02 | 1.01 | 1.03 | .001 |
| Number of nodes assessed (reference: Q1) | ||||
| Q2 | 1.09 | 0.82 | 1.44 | .55 |
| Q3 | 0.74 | 0.56 | 0.98 | .04 |
| Q4 | 0.91 | 0.69 | 1.21 | .52 |
CDCI, Charlson-Deyo comorbidity index; CI, confidence interval; HR, hazard ratio; Q, quartile.
Table 3.
Multivariable Logistic Regression of Factors Independently Associated With Pathologic Nodal Upstaging With Number of Assessed Nodes Modeled as Categoric Variable By Quartile
| Variable | OR | 95% CI |
P Value |
|
|---|---|---|---|---|
| Lower | Upper | |||
| Age, per year | 1.01 | 0.98 | 1.03 | .49 |
| Sex, female | 0.94 | 0.65 | 1.34 | .72 |
| Race (reference: white) | ||||
| Black | 1.37 | 0.70 | 2.67 | .35 |
| Year of diagnosis, per year | 0.94 | 0.87 | 1.01 | .10 |
| CDCI score (reference: 0) | ||||
| 1 | 1.10 | 0.75 | 1.61 | .63 |
| 2+ | 0.87 | 0.49 | 1.53 | .62 |
| Insurance status (reference: private) | ||||
| Government | 0.92 | 0.60 | 1.41 | .69 |
| None | 0.47 | 0.06 | 3.97 | .49 |
| Facility type (reference: nonacademic) | ||||
| Academic/research program | 1.22 | 0.84 | 1.77 | .30 |
| Tumor size, per mm | 1.02 | 1.00 | 1.04 | .03 |
| Number of nodes assessed (reference: Q1) | ||||
| Q2 | 0.77 | 0.44 | 1.34 | .35 |
| Q3 | 1.10 | 0.67 | 1.80 | .72 |
| Q4 | 1.64 | 0.99 | 2.71 | .05 |
CDCI, Charlson-Deyo comorbidity index; CI, confidence interval; HR, hazard ratio; Q, quartile.
We then modeled the effect of number of lymph nodes assessed as a continuous variable using piecewise linear splines with a knot at 7 lymph nodes. In multivariable Cox proportional hazards regression, neither assessment of more than 7 lymph nodes nor fewer than 7 lymph nodes was associated with a significant difference in survival (Table 4). For pathologic nodal upstaging, however, sampling of more than 7 lymph nodes resulted in a significantly higher likelihood of pathologic upstaging (Table 5).
Table 4.
Cox Multivariable Regression of Factors Independently Associated With Survival With Number of Assessed Nodes Modeled as Continuous Variable Using Linear Splines
| Variable | HR | 95% CI | P Value | |
|---|---|---|---|---|
| Lower | Upper | |||
| Age, per year | 1.02 | 1.01 | 1.04 | .001 |
| Sex, female | 0.84 | 0.69 | 1.03 | .10 |
| Race (reference: white) | ||||
| Black | 0.88 | 0.58 | 1.36 | .57 |
| Other | 1.23 | 0.73 | 2.09 | .43 |
| Year of diagnosis, per year | 1.03 | 0.98 | 1.08 | .23 |
| CDCI score (reference: 0) | ||||
| 1 | 1.03 | 0.83 | 1.28 | .77 |
| 2+ | 1.21 | 0.91 | 1.62 | .20 |
| Insurance status (reference: private) | ||||
| Government | 1.04 | 0.81 | 1.34 | .74 |
| None | 0.39 | 0.10 | 1.60 | .19 |
| Facility type (reference: nonacademic) | ||||
| Academic/research program | 1.04 | 0.84 | 1.29 | .69 |
| Tumor size, per mm | 1.02 | 1.01 | 1.03 | .002 |
| Number of nodes assessed, per node | ||||
| <7 nodes | 0.97 | 0.92 | 1.03 | .30 |
| >7 nodes | 1.00 | 0.98 | 1.02 | .89 |
CDCI, Charlson-Deyo comorbidity index; CI, confidence interval; HR, hazard ratio.
Table 5.
Multivariable Logistic Regression of Factors Independently Associated With Pathologic Nodal Upstaging With Number of Assessed Nodes Modeled as Continuous Variable With Piecewise Linear Splines and Knot at Seven Nodes
| Variable | HR | 95% CI | P Value | |
|---|---|---|---|---|
| Lower | Upper | |||
| Age, per year | 1.01 | 0.98 | 1.03 | .53 |
| Sex, female | 0.93 | 0.65 | 1.34 | .71 |
| Race (reference: white) | ||||
| Black | 1.36 | 0.70 | 2.65 | .36 |
| Year of diagnosis, per year | 0.93 | 0.87 | 1.01 | .07 |
| CDCI score (reference: 0) | ||||
| 1 | 1.11 | 0.76 | 1.62 | .60 |
| 2+ | 0.86 | 0.49 | 1.51 | .60 |
| Insurance status (reference: private) | ||||
| Government | 0.92 | 0.60 | 1.42 | .72 |
| None | 0.49 | 0.06 | 4.12 | .51 |
| Facility type (reference: nonacademic) | ||||
| Academic/research program | 1.22 | 0.84 | 1.77 | .31 |
| Tumor size, per mm | 1.02 | 1.00 | 1.04 | .04 |
| Number of nodes assessed, per node | ||||
| <7 nodes | 1.01 | 0.91 | 1.13 | .79 |
| >7 nodes | 1.03 | 1.00 | 1.06 | .02 |
CDCI, Charlson-Deyo comorbidity index; CI, confidence interval; HR, hazard ratio.
In a separate multivariable regression, there was no significant interaction between number of lymph nodes assessed and receipt of adjuvant chemotherapy (P = .15) with respect to their effect on survival, suggesting that the number of lymph nodes does not mediate the relationship between adjuvant chemotherapy and survival (Supplemental Figure 1). However, using a multivariable logistic regression examining factors associated with receipt of adjuvant chemotherapy, sampling more than 7 nodes (odds ratio [OR] 1.04 per node; 95% CI, 1.01 to 1.07; P = .01) but not fewer than 7 nodes (OR 1.05; 95% CI, 0.96 to 1.16; P = .30) was associated with increased odds of receiving adjuvant chemotherapy. In addition, upstaging to pathologic N1 (OR 2.22; 95% CI, 1.36 to 3.62; P = .001) but not N2 (OR 1.22; 95% CI, 0.56 to 2.64; P = .62) was associated with increased odds of receiving chemotherapy.
Comment
In this analysis of the NCDB, we found that harvest of 8 to 12 lymph nodes at the time of resection was associated with decreased mortality when the number of lymph nodes assessed was modeled as a categoric variable based on distribution quartiles. However, there was no association between mortality and any other quartile, nor was there an association between number of nodes assessed and survival when modeled as a continuous variable. Therefore, in aggregate, there was no clear survival benefit associated with an increasing number of lymph nodes harvested in cT1-2N0M0 SCLC. However, we found that examination of more than 7 nodes was significantly associated with an increased odds of nodal upstaging and receipt of adjuvant therapy. Our results would therefore suggest that patients may benefit having more than 7 lymph nodes harvested at the time of surgery for patients with limited stage SCLC who are undergoing lobectomy to ensure accurate pathologic nodal staging.
This is the first study to offer insight into the minimum number of lymph nodes necessary to sample during surgery for patients with limited stage SCLC. Despite finding that sampling more than 7 lymph nodes during surgery for SCLC was associated with increased pathologic nodal upstaging, the median number of lymph nodes assessed in our cohort was 7, suggesting that many patients are receiving an inadequate lymphadenectomy. Furthermore, the rate of pathologic upstaging in our sample was 19%, which corroborates findings in the current literature and lends external validity to our data.3,4 Given the number of patients who received an insufficient lymph node retrieval and the high likelihood for discrepancies with clinical staging, our analysis suggests that many patients with SCLC are being undertreated, and that may be further exacerbated by inaccurate pathologic staging as a consequence of inadequate lymph node harvest at the time of resection. That is manifest in our cohort as only approximately 40% of these patients received adjuvant chemotherapy, when all should have based on current guidelines.1
Although the etiology of this discordance with treatment guidelines is unclear, we would hypothesize that, as in non-small cell lung cancer, a mixture of socioeconomic factors, surgical factors, and tumor characteristics including nodal stage might play a role.9 In the context of other malignancies, efforts to define an adequate lymph node harvest at the time of surgery have been useful in providing a means for standardization of the extent of resection, which in turn better ensures accurate pathologic staging and access to appropriate adjuvant therapy for surgical candidates.10,11 Although survival in our cohort being independent of the number of lymph nodes collected may reflect the biology of SCLC and poor efficacy of currently available therapeutic modalities, there is potential that better guideline concordance may result in improved outcomes.
Our study has several important limitations. For one, given the retrospective nature of our analysis we have a limited ability to control for all possible confounding variables that may influence our results. For instance, the NCDB does not contain information on nodal stations sampled during surgery, which might introduce variation and consequently confound our analysis. The NCDB also does not catalogue if patients received intraoperative frozen sections, which might have dictated the extent of lymphadenectomy. Furthermore, we have no data regarding how patients were clinically staged. This information would likely provide useful insight into whether the limited extent of lymphadenectomy at surgery for many patients in our cohort was due to previous negative invasive mediastinal staging, with prior studies demonstrating the low prevalence of mediastinal staging in the United States.12,13 Similarly, given the lack of data available in the NCDB, we have no ability to comment on recurrence or disease-free survival. Another potential limitation is that there may be inaccuracies in the number of lymph nodes collected as fragmentation of nodes can occur during harvest. As such, there is potential that one lymph node may result in several fragments and be counted more than once, falsely elevating the count. Finally, a substantial portion of patients who underwent resection for limited stage SCLC were excluded from our cohort owing to missing data and may have biased our study.
Despite these limitations, in our analysis of the NCDB, we found that an increased number of lymph nodes assessed was not associated with survival for patients with T1-2N0M0 SCLC. However, sampling more than 7 lymph nodes at the time of surgery was associated with increased pathologic nodal upstaging and receipt of adjuvant chemotherapy. In aggregate, our findings suggest that more than 7 lymph nodes should be harvested at the time of lobectomy for limited stage SCLC.
Supplementary Material
Acknowledgments
The American College of Surgeons is in a Business Associate Agreement that includes a data use agreement with each of its Commission on Cancer accredited hospitals. The data used in the study are derived from a deidentified National Cancer Data Base file. The American College of Surgeons and the Commission on Cancer have not verified and are not responsible for the analytic or statistical methodology used or the conclusions drawn from these data by the investigators. This work received no direct funding. Drs Raman and Voigt are supported by a National Institutes of Health T-32 grant 5T32CA093245 in surgical oncology. Dr Jawitz is supported by a National Institutes of Health T-32 grant 5T32HL069749 in clinical research.
Footnotes
Presented at the Sixty-sixth Annual Meeting of the Southern Thoracic Surgical Association, Marco Island, FL, Nov 6-9, 2019.
References
- 1.National Comprehensive Cancer Network. Small cell lung cancer (version 2.2019). August 5, 2019. Available at: https://www.nccn.org/professionals/physician_gls/pdf/sclc_blocks.pdf. Accessed August 5, 2019.
- 2.Jett JR, Schild SE, Kesler KA, Kalemkerian GP. Treatment of small cell lung cancer. Chest. 2013;143:e400S–e419S. [DOI] [PubMed] [Google Scholar]
- 3.Vallieres E, Shepherd F, Crowley J, et al. The IASLC Lung Cancer Staging Project: proposals regarding the relevance of TNM in the pathologic staging of small cell lung cancer in the forthcoming (seventh) edition of the TNM classification for lung cancer. J Thorac Oncol. 2009;4:1049–1059. [DOI] [PubMed] [Google Scholar]
- 4.Nicholson AG, Chansky K, Crowley J, et al. The International Association for the Study of Lung Cancer Lung Cancer Staging Project: proposals for the revision of the clinical and pathologic staging of small cell lung cancer in the forthcoming eighth edition of the TNM classification for lung cancer. J Thorac Oncol. 2016;11:300–311. [DOI] [PubMed] [Google Scholar]
- 5.Stinchcombe TE. Current treatments for surgically resectable, limited-stage, and extensive-stage small cell lung cancer. Oncologist. 2017;22:1510–1517. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Yang Y, Zhang D, Zhou X, et al. Prophylactic cranial irradiation in resected small cell lung cancer: a systematic review with meta-analysis. J Cancer. 2018;9:433–439. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Bilimoria KY, Stewart AK, Winchester DP, Ko CY. The National Cancer Data Base: a powerful initiative to improve cancer care in the United States. Ann Surg Oncol. 2008;15:683–690. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Harrell F Multivariable modeling strategies. In: Harrell FE, ed. Regression Modeling Strategies: With Applications to Linear Models, Logistic, and Ordinal Regression, and Survival Analysis. New York, NY: Springer International Publishing; 2015:63–102. [Google Scholar]
- 9.Farrow NE, An SJ, Speicher PJ, et al. Disparities in guideline-concordant treatment for node-positive, non-small cell lung cancer following surgery. J Thorac Cardiovasc Surg. 2020;160:261–271. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Samayoa AX, Pezzi TA, Pezzi CM, et al. Rationale for a minimum number of lymph nodes removed with non-small cell lung cancer resection: correlating the number of nodes removed with survival in 98,970 patients. Ann Surg Oncol. 2016;23(suppl 5):1005–1011. [DOI] [PubMed] [Google Scholar]
- 11.Vielva LR, Jaen MW, Alcacer JA, Cardona MC. State of the art in surgery for early stage NSCLC—does the number of resected lymph nodes matter? Transl Lung Cancer Res. 2014;3:95–99. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Thornblade LW, Wood DE, Mulligan MS, et al. Variability in invasive mediastinal staging for lung cancer: a multicenter regional study. J Thorac Cardiovasc Surg. 2018;155:2658–2671.e2651. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Krantz SB, Howington JA, Wood DE, et al. Invasive mediastinal staging for lung cancer by The Society of Thoracic Surgeons Database participants. Ann Thorac Surg. 2018;106:1055–1062. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.