Development and validation of a nomogram for predicting overall survival of resected N2 non-small cell lung cancer patients undergoing neoadjuvant radiotherapy.

doi:10.21203/rs.3.rs-2976478/v1

Download PDF

Research Article

Development and validation of a nomogram for predicting overall survival of resected N2 non-small cell lung cancer patients undergoing neoadjuvant radiotherapy.

https://doi.org/10.21203/rs.3.rs-2976478/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 05 Jul, 2023

Read the published version in Journal of Cancer Research and Clinical Oncology →

You are reading this latest preprint version

Introduction

Currently, the prognosis of resected N2 non-small cell lung cancer patients undergoing neoadjuvant radiotherapy is poor. The goal of this research was to develop and validate a novel nomogram for exactly predicting the overall survival (OS) of resected N2 NSCLC patients undergoing neoadjuvant radiotherapy.

Methods

The data applied in our research were downloaded from the Surveillance, Epidemiology, and End Results (SEER) database. We divided selected data into a training cohort and a validation cohort using R software, with a ratio of 7:3. Univariate Cox regression and multivariate Cox regression were utilized to select significant variables to build the nomogram. In order to validate our nomogram, calibration curves, receiver operating characteristic curves (ROC), decision curve analysis (DCA), and Kaplan-Meier survival curves were employed. The nomogram model was also compared with the tumor-node-metastasis (TNM) staging system by utilizing Net reclassification index (NRI) and Integrated Discrimination Improvement (IDI).

Results

Eight variables—age, sex, operative type, LN removed number, chemotherapy, AJCC stage, M stage, histology—were statistically significant in the multivariate Cox Regression Analysis and were selected to develop our nomogram. Based on ROC curves, calibration curves, and DCA analysis, our novel nomogram demonstrated good predictive accuracy and clinical utility. Using Kaplan-Meier (KM) survival curves and log-rank tests, the risk stratification system was able to stratify patients based on their estimated mortality risk. The nomogram performed better than the TNM staging system based on the NRI and IDI indexes.

Conclusions

We developed and validated a nomogram to predict prognosis of resected N2 NSCLC patients undergoing neoadjuvant radiotherapy. By using this nomogram, Clinicians may find this nomogram useful in predicting OS of targeted patients and making more appropriate treatment decisions.

Nomogram

NSCLC

Neoadjuvant radiotherapy

Globally, lung cancer is one of the most prevalent cancers. Lung cancer is the principal cause of cancer-specific fatality, and non–small cell lung cancer (NSCLC) accounts for approximately 85% of lung cancer (Siegel et al. 2021). N2 is defined as lung cancer metastasis in the ipsilateral mediastinal lymph nodes and/or subcarinal lymph node(s), accounts for approximately 20%-30% of all non-small cell lung cancers (NSCLC) (Warth et al. 2012; Gao et al. 2020; Goldstraw et al. 2016). Based on The Chinese Society of Clinical Oncology (CSCO) and National Comprehensive Cancer Network (NCCN), neoadjuvant radiotherapy is recommended for N2 (TNM staging system) stage NSCLC (Wang et al. 2021). Patients should receive neoadjuvant therapies as soon as possible to prevent metastases from occurring. Prior to definitive surgery, radiotherapy increases response rates and provides local control during neoadjuvant systemic treatment. Preoperative radiotherapy aims to reduce local tumor recurrences and improve overall survival. In other aspects, the procedure may strengthen resectability by diminishing the tumor, reducing nodal disease, eliminating metastases, and improving local control by removing residual tumor (Sonett et al. 2004; Mauguen et al. 2012; Boschbarrera et al. 2012; Saw et al. 2021; Blumenthal et al. 2018; Wang et al. 2021).

There is wide acceptance for the tumor-node-metastasis (TNM) staging system developed by the Union for International Cancer Control (UICC) and used by the American Joint Committee on Cancer (AJCC) at the moment. The system could predict a patient's prognosis across several cancer types involving tumor invasion (T), regional lymph nodes (N), and distant metastases (M) (Amin et al. 2017). Despite this, TNM staging can be a flawed prognostic system since it lacks precise prediction of some specific patients. So far, there are several prognostic models for NSCLC patients receiving neoadjuvant chemotherapy, postoperative chemotherapy, and postoperative radiotherapy, however, about neoadjuvant radiotherapy is absent. This necessitates developing a new model for predicting prognosis following neoadjuvant radiotherapy for resected N2 NSCLC patients. Medical nomograms are graphical representations of complex mathematical equations. A medical nomogram is a graphic representation of statistical prognostic models that provide a probability of a clinical ending based on clinical and biological parameters (Balachandran et al. 2015). In our study, based on the selected variables, we constructed a novel nomogram, which assists clinician predicting OS of targeted patients and making appropriate treatment decisions.

Study population

Surveillance Epidemiology and End Results (SEER) represents approximately 30% of the U.S. population, making it the largest publicly available cancer database. In the present study, data were extracted from SEER 17 registries research database (Nov 2021 Sub, 2000–2019). Due to the fact that SEER does not contain personally identifiable information, this study did not require approval by the review committee and informed consent was not required. The inclusion criteria were outlined below: (1) primary tumor site is lung and Bronchus (ICD-O-3/WHO 2008); C34.0-C34.9 (Primary Site-labeled); (2) diagnosis years from 2004 to 2015; (3) Radiotherapy prior to surgery (Therapy. RX Summ-Surg/Rad Seq); (4) N2 (Derived AJCC N,6th ed); (5) diagnosis as NSCLC; (6) completed follow-up. The exclusion criteria included: (1) AJCC-TNM stage is unknown (2) uncompleted follow-up. Selection of study population is illustrated in Fig. 1.

Study Variables

There were a number of variables included in the analysis, including age, sex, marital status, year of diagnosis, race, location, laterality, histology, AJCC stage, T stage, M stage, operative type, LN removed number, chemotherapy, tumor size. Age was transferred categorical variables (age: <60 years, 60–69 years, and ≥ 70 years). Using X-tile software, we evaluated which cutoff value is optimal for tumor size. Sizes of tumors were classified according to their largest diameters (tumor size: <30 mm, 30-78mm, and > 78mm). Race was divided into white, black and others. Histology was divided into squamous carcinoma, adenocarcinoma and others. Operative type was stratified using SEER code: it was determined that 12,13,15,19–25 were sublobar resections; 30,33,45–48 were lobectomy or extended lobectomy; 55,56,66,70 were pneumonectomy or extended pneumonectomy.

Statistical Analyses

This study used SPSS 26.0 and R software version 4.1.2 (https://www.r-project.org/) to conduct its statistical analysis. R software was used to randomly divide the patients into training and validation cohorts in a 7:3 ratio. Comparisons of baseline characteristics were performed by Chi-square test and Fisher's exact test. In the training cohort, we used univariate Cox regression analyses to identify prognostic variables independently affecting OS. We then performed multivariable Cox regression analyses and calculated odds ratios (ORs) and 95% confidence intervals (CIs) based on the variables with P < 0.05. Considering clinical practice, eight prognostic variables were determined: age, sex, operative type, LN removed number, chemotherapy, AJCC stage, M stage, histology. Next, we constructed a nomogram and based on the determined prognostic variables. In order to evaluate the prediction efficiency of the model, the receiver operating characteristic curve (ROC) with area under the curve (AUC) was used. Calibration plots were used to verify that the actual probability and predicted probability were close together. An appraisal of the clinical utility of the nomogram was conducted using decision curve analysis (DCA). We utilized Youden Index to obtain the best cut-off value. The maximum value corresponding to the Youden Index is the best cutoff. Based on best cut-off value, patients were categorized into high- and low-risk categories. Kaplan-Meier survival analysis with log-rank test was used to compare the OS between the two groups. There was statistical significance with P-values (both sides) less than 0.05. We used “foreign”, “survival”, “survivalROC”, “rms”, “shiny”, “DynNom”, “magrittr”, “ggrisk”, and other R packages to do our works.

Clinical Features

The study enrolled a total of 1889 patients, who were randomly assigned to the training set (n = 1325) and the validation set (n = 564). In both the Chi-square test and Fisher's exact test, there were no significant differences between the validation and training groups (p > 0.05), except for sex (p = 0.041). Table 1 described the baseline data of the patients. Among the 1889 patients, there were 1034 males (54.7%), 1182 married (62.6%), 1624 white people (86.0%). In terms of tumor locational characteristics, there were 1366 regional (72.3%), 1218 right (64.5%). According to the AJCC stage, there were 1098ⅢA (58.1%), 806 T2(42.7%), 1472 M0(77.9%). The majority of patients received chemotherapy (1713, 90.7%). About Operative type, most patients received lobectomy or extended (1055,55.8%), and 1077 patients were removed 4 or more LN (57.0%). At last, more than half of the patients (57.4%) had tumor between 3.0 and 7.8 cm in size.

Table 1

Clinicopathlolgical characteristics of selected patients
Variables	Total	Training cohort	Validation cohort	p-value
	N = 1889	N = 1325	N = 564
Age				0.778
< 60	760(40.2%)	540(40.8%)	220(39.0%)
60–69	709(37.5%)	493(37.2%)	216(38.3%)
≥ 70	420(22.2%)	292(22.0%)	128(22.7%)
Race				0.384
White	1624(86.0%)	1131(85.4%)	493(87.4%)
Black	158(8.4%)	113(8.5%)	45(8.0%)
Other	107(5.7%)	81(6.1%)	26(4.6%)
Sex				0.041
Male	1034(54.7%)	705(53.2%)	329(58.3%)
Female	855(45.3%)	620(46.8%)	235(41.7%)
Marital status				0.691
Unmarried	223(11.8%)	160(12.1%)	63(11.2%)
Married	1182(62.6%)	821(62.0%)	361(64.0%)
other	484(25.6%)	344(25.9%)	140(24.8%)
Year of diagnosis				0.453
2004–2009	1057(56.0%)	734(55.4%)	323(57.3%)
2010–2015	832(44.0%)	591(44.6%)	241(42.7%)
Location				0.304
Distant	523(27.7%)	376(28.4%)	147(26.1%)
Regional	1366(72.3%)	949(71.6%)	417(73.9%)
Laterality				0.94
Right	1218(64.5%)	851(64.2%)	367(65.1%)
Left	654(34.6%)	462(34.9%)	192(34.0%)
Other	17(0.9%)	12(0.9%)	5(0.9%)
Histology				0.109
squamous carcinoma	561(29.7%)	379(28.6%)	182(32.3%)
Adenocarcinoma	820(43.4%)	573(43.2%)	247(43.8%)
Other	508(26.9%)	373(28.2%)	135(23.9%)
AJCC stage				0.8
IIIA	1098(58.1%)	766(57.8%)	332(58.9%)
IIIB	374(19.8%)	261(19.7%)	113(20.0%)
IV	417(22.1%)	298(22.5%)	119(21.1%)
T stage				0.383
T1	314(16.6%)	210(15.8%)	104(18.4%)
T2	806(42.7%)	580(43.8%)	226(40.1%)
T3	225(12.0%)	155(11.7%)	70(12.4%)
T4	544(28.8%)	380(28.7%)	164(29.1%)
M stage				0.505
M0	1472(77.9%)	1027(77.5%)	445(78.9%)
M1	417(22.1%)	298(22.5%)	119(21.1%)
Operative type				0.578
sublobar resection	584(30.9%)	419(31.6%)	165(29.3%)
lobectomy or extended	1055(55.8%)	734(55.4%)	321(56.9%)
pneumonectomy or extended	250(13.2%)	172(13.0%)	78(13.8%)
LN removed number				0.561
None	282(14.9%)	204(15.4%)	78(13.8%)
1 to 3	255(13.5%)	184(13.9%)	71(12.6%)
4 or more	1077(57.0)	751(56.7%)	326(57.8%)
Other	275(14.6%)	186(14.0%)	89(15.8%)
Chemotherapy				0.197
No/Unknown	176(9.3%)	116(8.8%)	60(10.6%)
Yes	1713(90.7%)	1209(91.2%)	504(89.4%)
Tumor size				0.089
< 3.0cm	433(22.9%)	286(21.6%)	147(26.0%)
3.0-7.8cm	1084(57.4%)	769(58.0%)	315(55.9%)
> 7.8cm	372(19.7%)	270(20.4%)	102(18.1%)

Univariate and Multivariate Cox Regression Analysis

In the initial stage of the analysis, univariate Cox regression analysis is used to confirm the prognosis-related risk factors. A patient's survival was associated with age, sex, marital status, location, laterality, AJCC stage, M stage, T stage, histology, operative type, LN removed number, chemotherapy, tumor size. In the next step, we applied the multivariate Cox regression analysis to the selected risk factors. Finally, the results showed that age, sex, operative type, LN removed number, chemotherapy, AJCC stage, histology, Marital status were significant prognostic factors, while M stage is linear dependent covariate (Table 2).

Table 2

Univariate and Multivariate Cox Regression Analysis in training cohort.
Variables	Univariate analysis		Multivariate analysis
	HR (95% CI)	p-value	HR (95% CI)	p-value
Age
< 60	Reference		Reference
60–69	1.377(1.195–1.586)	< 0.001	1.430(1.236–1.654)	< 0.001
≥ 70	1.829(1.557–2.149)	< 0.001	1.800(1.519–2.133)	< 0.001
Race
White	Reference
Black	0.920(0.733–1.154)	0.469
Other	1.196(0.940–1.521)	0.144
Sex
Female	Reference		Reference
Male	1.272(1.125–1.439)	< 0.001	1.225(1.075–1.396)	0.002
Marital status
Married	Reference		Reference
Unmarried	0.986(0.809–1.201)	0.885	0.992(0.812–1.213)	0.941
other	1.227(1.065–1.414)	0.005	1.224(1.052–1.423)	0.009
Year of diagnosis
2004–2009	Reference
2010–2015	1.007(0.886–1.144)	0.912
Location
Distant	Reference		Reference
Regional	0.386(0.338–0.440)	< 0.001	0.942(0.702–1.263)	0.689
Laterality
Right	Reference		Reference
Left	1.056(0.929–1.201)	0.403	0.995(0.873–1.135)	0.945
Other	2.108(1.161–3.829)	0.014	0.895(0.484–1.655)	0.724
Histology
squamous carcinoma	Reference		Reference
Adenocarcinoma	0.850(0.732–0.987)	0.033	0.825(0.703–0.968)	0.018
Other	1.105(0.942–1.296)	0.218	1.060(0.899–1.250)	0.489
AJCC stage
IIIA	Reference		Reference
IIIB	1.219(1.036–1.434)	0.017	0.951(0.700-1.293)	0.751
IV	3.205(2.765–3.715)	< 0.001	1.718(1.222–2.418)	0.002
T stage
T1	Reference		Reference
T2	1.052(0.876–1.264)	0.588	1.093(0.847–1.410)	0.495
T3	1.042(0.818–1.327)	0.739	1.048(0.773–1.422)	0.761
T4	1.401(1.155–1.699)	0.001	1.109(0.804–1.529)	0.529
M stage
M0	Reference
M1	3.052(2.650–3.515)	< 0.001
Operative type
sublobar resection	Reference		Reference
lobectomy or extended	0.292(0.255–0.335)	< 0.001	0.495(0.413–0.593)	< 0.001
pneumonectomy or extended	0.409(0.336–0.499)	< 0.001	0.707(0.560–0.893)	0.004
LN removed number
None	Reference		Reference
1 to 3	0.379(0.305–0.470)	< 0.001	0.662(0.521–0.841)	0.001
4 or more	0.262(0.221–0.310)	< 0.001	0.536(0.431–0.667)	< 0.001
Other	0.638(0.519–0.785)	< 0.001	0.811(0.655–1.004)	0.054
Chemotherapy
No/Unknown	Reference		Reference
Yes	0.333(0.272–0.407)	< 0.001	0.515(0.415–0.638)	< 0.001
Tumor size
< 3.0cm	Reference		Reference
3.0-7.8cm	1.152(0.983–1.351)	0.081	1.041(0.830–1.304)	0.73
> 7.8cm	1.484(1.227–1.795)	< 0.001	1.175(0.911–1.516)	0.214

Development and validation of the nomogram

On the basis of clinical experience and the OS-related prognostic factors identified, eight variables, including age, sex, operative type, LN removed number, chemotherapy, AJCC stage, M stage and histology were selected to develop a nomogram that predicted 1-, 2-, and 3-year OS in resected N2 NSCLC patients undergoing neoadjuvant radiotherapy (Fig. 2). Here, we offer detailed explanation for our nomogram. Age: 0 means༜60, 1 means 60–69, 2 means ≥ 70. Sex: 0 means female, 1 means male. AJCC stage: 0 means ⅢA, 1 means ⅢB, 2 means Ⅳ. M stage: 0 means M0, 1 means M1. Operative type: 0 means sublobar resection, 1 means lobectomy or extended, 2 means pneumonectomy or extended. LN removed number: 0 means none, 1 means one to three, 2 means four or more, 3 means other. Chemotherapy: 0 means no/unknown, 1 means yes. Histology: 0 means squamous carcinoma, 1 means Adenocarcinoma, 2 means other. Each factor contributed different amounts to OS according to its line length in the nomogram. A nomogram shows that Operative type is the most significant prognostic factor for selected patients. The second most important prognostic factor is chemotherapy. And LN removed number, M stage, age, LN removed number, AJCC stage, sex, histology were followed. The more LN were removed, the higher the patient's survival probability. Prediction accuracy of the model was tested at 1-, 2-, and 3-year intervals using the area under the ROC. ROC curve results showed that AUC in the training cohort was 0.776 at 1 year, 0.725 at 2 years, and 0.712 at 3 years, and in the validation cohort was 0.734 at 1 year, 0.670 at 2 years, and 0.670 at 3 years, showing good discrimination capability of the nomogram (Fig. 3A and 3B). There was good agreement between prediction and actual values in both training and validation cohorts (Fig. 4A-4F). Further evaluation of the clinical utility of this nomogram was conducted through decision curve analysis (Fig. 5A-5F). Then, we compared our model with traditional AJCC staging. For 1-, 2-, and 3-year follow-up NSCLC examinations in the training cohort, the Net reclassification index (NRI) values were 0.568 (95% CI = 0.417–0.719), 0.478 (95% CI = 0.350–0.604), and 0.421 (95% CI = 0.324–0.571) respectively. The NRI estimates for the validation cohort were respectively 0.584 (95% CI = 0.406–0.853), 0.467 (95% CI = 0.275–0.702), and 0.588 (95% CI = 0.309–0.736). Likewise, Integrated Discrimination Improvement (IDI) values were reported for 1-, 2-, and 3-year NSCLC follow-up examinations in the training cohort of 0.104 (95% CI = 0.007–0.137), 0.080 (95% CI = 0.057–0.105), and 0.067 (95% CI = 0.048–0.093), and 0.135 (95% CI = 0.092–0.188), 0.086 (95% CI = 0.049–0.127), and 0.084 (95% CI = 0.051–0.120) in the validation cohort (Table 3).

Table 3

NRI and IDI analysis between our predictive model and traditional AJCC staging.
Index	training cohort			validation cohort
	Estimate	95%CI	p-value	Estimate	95%CI	p-value
NRI (vs. AJCC TNM staging)
For 1–year OS	0.568	0.417–0.719		0.584	0.406–0.853
For 2–year OS	0.478	0.350–0.604		0.467	0.275–0.702
For 3–year OS	0.421	0.324–0.571		0.588	0.309–0.736
IDI (vs. AJCC TNM staging)
For 1–year OS	0.104	0.007–0.137	< 0.01	0.135	0.092–0.188	< 0.01
For 2–year OS	0.08	0.057–0.105	< 0.01	0.086	0.049–0.127	< 0.01
For 3–year OS	0.067	0.048–0.093	< 0.01	0.084	0.051–0.120	< 0.01

Risk stratification system

For the purpose of risk stratification, we used the best cut-off value obtained by Youden Index. Two groups of patients were grouped for training and validation: those with low risk and those with high. In both the training cohort and validation cohort, KM survival curves and log-rank tests revealed significant differences between the two risk groups (p < 0.001), suggesting the nomogram has the potential to be an effective tool in stratifying our selected specific patients according to their estimated mortality risk. (Fig. 6A and 6B). Then, we constructed distribution plot and heatmap to further demonstrate effectivity of our risk stratification system (Fig. 7A -7C).

Patients with NSCLC are currently staged using the AJCC TNM staging system to predict their prognoses (Balachandran et al. 2017). Despite this, TNM-based prediction methods cannot sufficiently predict the outcome of neoadjuvant radiotherapy for resected N2 NSCLC patients. There have been few studies dedicated to predicting the outcome of resected N2 non-small cell lung cancer patients undergoing neoadjuvant radiotherapy. Hence, a new nomogram model based on routinely used clinicopathological variables should be developed that better predicts the outcome of resected N2 NSCLC patients receiving neoadjuvant radiotherapy. Here, our novel nomogram is based on a large group of patients from the SEER database, which is convenient for clinicians and patients to utilize. Our nomogram model, consisting of demographic characteristics, tumor information, therapy modalities, predicts the probability of overall survival for resected N2 NSCLC patients undergoing neoadjuvant radiotherapy very well.

Many variables affect the survival of resected N2 NSCLC patients undergoing neoadjuvant radiotherapy. These 16 clinicopathological factors are included in the nomogram model developed in this study: race, age, year of diagnosis, marital status, sex, stage, location, histology, laterality, AJCC stage, T stage, M stage, operative type, LN removed number, chemotherapy, tumor size. We have included more variables in our nomogram than in the 6th edition of AJCC TNM staging system, so it has a greater ability to predict. The prognosis of cancer patients can be predicted independently by their age in numerous studies. It has been shown that age > 54 is associated with a worse OS in a prognostic nomogram compiled by Dai et al. Previous studies have reported that race had an impact on the prognosis of NSCLC. It was accounted that white and black people have a shorter survival time than Asians or Hispanics (Soneji et al. 2017; Patel et al. 2016). Many studies demonstrated that sex is an independent factor about predicting prognosis of NSCLC. For example, as far as we know, Liang et al. found that female have a better prognosis than male (Liang et al. 2015). NSCLC patient’s survivorship is also influenced by their marital status. The reason for this may be that married cancer patients have the support and encouragement of their spouses, which is better than single patients both psychologically and physically. An analysis of a large population of patients with NSCLC by Wu et al. discovered that married patients tend to live longer than unmarried patients (Wu et al. 2017). Stage, histological type and laterality were shown to have some effect on OS of NSCLC (Liang et al. 2015; Zuo et al. 2021). Zuo et al. reported that Squamous cell carcinoma had a worse OS than Adenocarcinoma and other types. As we all know, TNM stage is a very important factor affecting prognosis of NSCLC. Many previous studies showed proved it. For example, a nomogram was constructed by Liang et al. for predicting survival of NSCLC patients who had undergone resection, and T stage and N stage contributed the most to prognosis (Liang et al. 2015). Wang et al. have found lymph node ratio (LNR) and log odds of positive lymph nodes (LODDS) had a greatly high predictive accuracy for OS of patients with resected lung adenocarcinoma (Wang et al. 2021). And Wang et al. also found that LODDS could effectively differentiate patients’ prognoses (Wang et al. 2022). Treatment modalities, including operative types, LN removed number, chemotherapy, were vitally affected the prognosis of N2 NSCLC patients experiencing neoadjuvant radiotherapy. Zuo et al. have reported that patients experiencing lobectomy had a better OS than others, and resecting more LNs also had a better prognosis (Zuo et al. 2021). DuMa et al. have claimed that chemotherapy could downstage of the tumor and improved tolerability, thus it is of necessity for our selected patients (Duma et al. 2019). Qu et al. have done a study examining the effect of neoadjuvant radiotherapy on survival after pneumonectomy for patients with stage IIIA-N2 NSCLC (Qu et al. 2022). And they demonstrated that N2 NSCLC patients could get benefit from NART (Neoadjuvant Radiotherapy), however, for N0-N1 NSCLC patients, NART is not very effective. Wang et al. have reported that patients at T1/2 & N2/3 may get benefit from neoadjuvant or adjuvant RT (Wang et al. 2018).

ROC curve results showed that AUC in the training cohort was 0.776 at 1 year, 0.725 at 2 years, and 0.712 at 3 years, and in the validation cohort was 0.734 at 1 year, 0.670 at 2 years, and 0.670 at 3 years, showing good discrimination capability of the nomogram. Moreover, our model was shown to be efficacious by DCA analysis, a calibration plot and a Kaplan-Meier survival curve. Using a risk system, we stratified patients with NSLCL based on their estimated mortality risk. In comparison with the TNM staging system, both the NRI and IDI index demonstrated that the nomogram performed better.

As far as we know, this is the first nomogram to predict survival in resected N2 NSCLC patients undergoing neoadjuvant radiotherapy. However, this study had several limitations. First, there are some other significant variables are lacked, including molecular or genetic information, more specific clinic characters and biochemical indicators, however, the SEER database did not provide these data. Second, our study only depended on internal validation, which is not accurately enough. In the future, we will address these shortcomings to further improve our nomogram.

As a result of our work, we have developed a novel nomogram model to predict survival of resected N2 NSCLC patients undergoing neoadjuvant radiotherapy. The nomogram demonstrated good accuracy and discriminative ability. As compared to TNM staging, this nomogram model was better at predicting prognosis of specific patients. By using this nomogram, Clinicians may find this nomogram useful in predicting OS of targeted patients and making more appropriate treatment decisions.

NSCLC

Non-small cell lung cancer

SEER

Surveillance, Epidemiology, and End Results

Overall survival

AJCC

American Joint Committee on Cancer

ROC

Receiver operating characteristic curves

DCA

Decision curve analysis

TNM

tumor-node-metastasis

NRI

Net reclassification index

IDI

Integrated discrimination improvement

Kaplan-Meier

CSCO

The Chinese Society of Clinical Oncology

NCCN

National Comprehensive Cancer Network

UICC

Union for International Cancer Control

Acknowledgements The authors would like to express their gratitude to SEER for providing free access to the database.

Author contributions LZ, JS and BP designed the study. JS wrote and edited the manuscript. CW, XZ, TL, RX, JS created the figures and tables. XC handled the partial data. KW, ZS, CX, JS collected the data. All authors contributed to the artical and approved the submitted version.

Availability of data and materials SEER Stat 8.401 software was used to extract our data online, and data in our study is available at SEER database: https://seer.cancer.gov/

Conflict of Interest The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Ethics approval and consent to participate Since the patient information in the SEER database is deidentifed and publicly available, our study was exempt from Institutional Review Board approval. In addition, we declared all methods were carried out in accordance with relevant guidelines and regulations.

Consent for publication Each author agreed to publish the paper.

Siegel RL, Miller KD, Fuchs HE, Jemal A (2021) Cancer Statistics, 2021. CA Cancer J Clin 71(1):7–33. doi: 10.3322/caac.21654.
Warth A, Muley T, Meister M, Stenzinger A, Thomas M, Schirmacher P et al (2012) The novel histologic International Association for the Study of Lung Cancer/American Thoracic Society/European Respiratory Society classification system of lung adenocarcinoma is a stage-independent predictor of survival. J Clin Oncol 30(13):1438–46. doi: 10.1200/JCO.2011.37.2185.
Gao F, Li N, Xu Y, Yang G (2020) Evaluation of Postoperative Radiotherapy Effect on Survival of Resected Stage III-N2 Non-small Cell Lung Cancer Patients. Front Oncol 10:1135. doi: 10.3389/fonc.2020.01135.
Goldstraw P, Chansky K, Crowley J, Rami-Porta R, Asamura H, Eberhardt WE, Nicholson AG, Groome P, Mitchell A, Bolejack V (2016) The IASLC Lung Cancer Staging Project: Proposals for Revision of the TNM stage Groupings in the Forthcoming (Eighth) Edition of the TNM Classification for Lung Cancer. J Thorac Oncol 11(1):39–51. doi: 10.1016/j.jtho.2015.09.009.
Wang FH, Zhang XT, Li YF, Tang L, Qu XJ, Ying JE et al (2021) The Chinese Society of Clinical Oncology (CSCO): Clinical guidelines for the diagnosis and treatment of gastric cancer, 2021. Cancer Commun (Lond) 41(8):747–795. doi: 10.1002/cac2.12193.
Sonett JR, Suntharalingam M (2004) Pulmonary resection after curative intent radiotherapy (> 59 Gy) and concurrent chemotherapy in non-small-cell lung cancer. Ann Thorac Surg 78(4):1200. doi: 10.1016/j.athoracsur.2004.04.085.
Mauguen A, Le PC, Saunders MI, Schild SE, Turrisi AT, Baumann M et al (2012) Hyperfractionated or accelerated radiotherapy in lung cancer: an individual patient data meta-analysis. J Clin Oncol 30(22):2788. doi: 10.1200/JCO.2012.41.6677.
Boschbarrera J, Garcíafranco C, Guilléngrima F, Morenojiménez M, Lópezpicazo JM, Gúrpide A et al (2012) The multimodal management of locally advanced N2 non-small cell lung cancer: is there a role for surgical resection? A single institution's experience. Clin Transl Oncol 14(11):835–841. doi: 10.1007/s12094-012-0874-3.
Saw SPL, Ong BH, Chua KLM, Takano A, Tan DSW (2021) Revisiting neoadjuvant therapy in non-small-cell lung cancer. Lancet Oncol 22(11):e501-e516. doi: 10.1016/S1470-2045(21)00383-1.
Blumenthal GM, Bunn PA Jr, Chaft JE, McCoach CE, Perez EA, Scagliotti GV et al (2018) Current Status and Future Perspectives on Neoadjuvant Therapy in Lung Cancer. J Thorac Oncol 13(12):1818–1831. doi: 10.1016/j.jtho.2018.09.017.
Wang YQ, Liu XD, Bai WL, Li SQ (2021) Identification of Resectable N2 in NSCLC: A Single Center Experience and Review of the SEER Database. Front Oncol 11:647546. doi: 10.3389/fonc.2021.647546.
Amin MB, Greene FL, Edge SB, Compton CC, Gershenwald JE, Brookland RK et al (2017) The Eighth Edition AJCC Cancer Staging Manual:Continuing to build a bridge from a population-based to a more “personalized”approach to cancer staging. CA Cancer J Clin 67(2):93–9. doi: 10.3322/caac.21388.
Balachandran VP, Gonen M, Smith JJ, DeMatteo RP (2015) Nomograms in oncology: more than meets the eye. Lancet Oncol 16(4):e173-80. doi: 10.1016/S1470-2045(14)71116-7.
Soneji S, Tanner NT, Silvestri GA et al (2017) Racial and Ethnic Disparities in Early-Stage Lung Cancer Survival. Chest 152:587–97. doi: 10.1016/j.chest.2017.03.059.
Patel MI, Wang A, Kapphahn K et al (2016) Racial and Ethnic Variations in Lung Cancer Incidence and Mortality: Results From the Women's Health Initiative. J Clin Oncol 34:360–8. doi: 10.1200/JCO.2015.63.5789.
Liang W, Zhang L, Jiang G, Wang Q, Liu L, Liu D et al (2015) Development and validation of a nomogram for predicting survival in patients with resected non-small-cell lung cancer. J Clin Oncol 10;33(8):861–9. doi: 10.1200/JCO.2014.56.6661.
Wu Y, Ai Z, Xu G (2017) Marital status and survival in patients with non-small cell lung cancer: an analysis of 70006 patients in the SEER database. Oncotarget 8:103518–34. doi: 10.18632/oncotarget.21568.
Zuo Z, Zhang G, Song P, Yang J, Li S, Zhong Z et al (2021) Survival Nomogram for Stage IB Non-Small-Cell Lung Cancer Patients, Based on the SEER Database and an External Validation Cohort. Ann Surg Oncol 28(7):3941–3950. doi: 10.1245/s10434-020-09362-0.
Wang S, Yu Y, Xu W et al (2021) Dynamic nomograms combining N classification with ratio-based nodal classifications to predict long-term survival for patients with lung adenocarcinoma after surgery: a SEER population-based study. BMC Cancer 21:653. doi:10.1186/s12885-021-08410-6
Wang Q, Wang S, Sun Z, Cao M, Zhao X (2022) Evaluation of log odds of positive lymph nodes in predicting the survival of patients with non-small cell lung cancer treated with neoadjuvant therapy and surgery: a SEER cohort-based study. BMC Cancer 22(1):801. doi: 10.1186/s12885-022-09908-3.
Duma N, Santana-Davila R, Molina JR (2019) Non-Small Cell Lung Cancer: Epidemiology, Screening, Diagnosis, and Treatment. Mayo Clin Proc 94(8):1623–1640. doi: 10.1016/j.mayocp.2019.01.013
Qu C, Li R, Han J, Yue W, Tian H (2022) Effects of Neoadjuvant Radiotherapy on Survival in Patients with Stage IIIA-N2 Non-Small-Cell Lung Cancer Following Pneumonectomy. J Clin Med 11(23):7188. doi: 10.3390/jcm11237188.
Wang X, Yin C, Su S, Li X, Wang C, Zhang C et al (2018) Long-term effects of neoadjuvant radiotherapy, adjuvant radiotherapy, and chemotherapy-only on survival of locally advanced non-small cell lung Cancer undergoing surgery: a propensity-matched analysis. BMC Cancer 18(1):1067. doi: 10.1186/s12885-018-4900-x.

No competing interests reported.

Download PDF

Journal Publication

published 05 Jul, 2023

Read the published version in Journal of Cancer Research and Clinical Oncology →

Editorial decision: Accepted
29 Jun, 2023
Reviewers agreed at journal
02 Jun, 2023
Reviewers invited by journal
27 May, 2023
Editor assigned by journal
27 May, 2023
Submission checks completed at journal
25 May, 2023
First submitted to journal
24 May, 2023

You are reading this latest preprint version

Development and validation of a nomogram for predicting overall survival of resected N2 non-small cell lung cancer patients undergoing neoadjuvant radiotherapy.

Status:

Journal Publication

Version 1

Abstract

Figures

Introduction

Material and methods

Study population

Study Variables

Statistical Analyses

Results

Clinical Features

Univariate and Multivariate Cox Regression Analysis

Development and validation of the nomogram

Risk stratification system

Discussion

Conclusions

Abbreviations

Declarations

References

Additional Declarations

Status:

Journal Publication

Version 1