Construction and validation of a nomogram for predicting overall survival in stage IV non-small cell lung cancer treated with epidermal growth factor receptor tyrosine kinase inhibitors

Introduction

Lung cancer, a highly malignant disease representing a substantial global health burden, predominantly comprises non-small cell lung cancer (NSCLC), representing around 85% of diagnoses (1). Within NSCLC, lung adenocarcinoma emerges as the primary histological subtype, constituting approximately 40% of all instances (2). Epidermal growth factor receptor tyrosine kinase inhibitors (EGFR-TKIs) have been established as a globally recognized treatment for EGFR mutant NSCLC (3).

Gefitinib and icotinib, first-generation EGFR-TKIs, selectively target 19del and L858R. These inhibitors function by reversibly binding to EGFR (4). In China, icotinib has gained widespread application owing to its potent antitumor activity and high selectivity, with robust clinical evidence supporting its safety and tolerability (5). Osimertinib selectively inhibits T790M and EGFR-sensitive mutations, reducing the risk of T790M mutation development as a first-line therapy (6). However, secondary C797S mutations can induce resistance (7).

Certain studies indicate that targeted therapy has contributed to improved survival prospects in patients with stage IV NSCLC. However, the 5-year prognosis remains poor, with survival rates still not exceeding 10% (8). Alarmingly, advanced-stage diagnosis is common among lung cancer patients, posing significant challenges to treatment and prognosis (1). Prior research reveals that the therapeutic efficacy of EGFR-TKIs varies among individuals harboring EGFR mutations, which indicates that additional prognostic factors may influence outcomes in EGFR-TKI-treated patients (9). This study aims to comprehensively explore and leverage patient clinical factors to identify predictors of overall survival (OS) and to develop a nomogram model for predicting OS in EGFR-positive NSCLC patients. We present this article in accordance with the TRIPOD reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2024-2112/rc).

Methods

Patients and data collection

As a single-center study, this study consecutively collected patients diagnosed with advanced NSCLC harboring EGFR mutations, treated at Tongji Hospital from 1 November 2014 to 1 November 2023. All patients were pathologically confirmed as NSCLC and categorized as stage IV based on the ninth edition Tumor Node Metastasis classification system (10). Clinical data, encompassing characteristics including age, sex, staging, smoking history, and pathological type, were retrieved from patient medical records. Detailed information on metastatic sites was documented, including pleural, brain, liver, adrenal, bone, intrapulmonary, and malignant pleural effusions, as well as other metastatic sites. Biomarkers related to lung cancer, including neuron-specific enolase (NSE), carcinoembryonic antigen (CEA), cytokeratin fragment 19 (CYFRA 21-1), squamous cell carcinoma (SCC) antigen, and progastrin-releasing peptide (ProGRP), along with genetic test outcomes and treatment modalities, were documented. Treatment modalities encompassed EGFR-TKIs, immunotherapy, lung resection, metastatic surgery, radiotherapy, and chemotherapy. Follow-up data were gathered via telephone interviews and a review of medical records. Thresholds for each tumor marker were defined according to Tongji Hospital’s maximum normal levels, specifically: 16.3 µg/L for NSE, 5.0 ng/mL for CEA, 3.3 µg/L for CYFRA 21-1, 1.5 ng/mL for SCC, and 65.7 pg/mL for ProGRP. Genetic testing confirmed the presence of EGFR mutations, and they received EGFR-TKI treatments, including osimertinib (80 mg daily), gefitinib (250 mg daily), or icotinib (125 mg three times daily). Patient enrollment for icotinib treatment spanned from November 2014 to September 2022, while enrollment for gefitinib treatment occurred between February 2015 and March 2023. Enrollment for osimertinib treatment spanned from July 2016 to October 2023. Patients with other primary malignancies or incomplete medical records were excluded, resulting in a final sample size of 461 cases and the follow-up period concluded on 12 June 2024. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the institutional ethics board of Tongji Hospital (No. TJ-IRB202408023) and individual consent for this retrospective analysis was waived.

Statistical analysis

The primary endpoint of this study was OS, defined as the duration from the initiation of drug administration to either the study’s conclusion or the patient’s death. Patients were randomized into training and validation cohorts at a 7:3 ratio, and their baseline characteristics were systematically compared to ensure balance. Categorical variables were assessed via Chi-squared tests. Variables were selected through Cox proportional hazards regression analysis, with those yielding P values <0.05 incorporated into the nomogram. The model’s performance was evaluated in both the training and validation cohorts using decision curve analysis (DCA), receiver operating characteristic curves, and calibration plots. Patients were scored based on the nomogram in the training cohort. The various critical thresholds of the risk score were assessed using the surv_cutpoint function from the survminer package in R software to identify the threshold that maximizes the log-rank test statistic. The optimal threshold value was determined to be 221.19. Patients were risk-grouped based on the optimal cut-off value using the surv_categorize function. Survival probabilities were compared across risk groups stratified by the determined cutoff value in the training and validation cohorts. Statistical analyses were conducted using SPSS version 27, while graphical visualizations were generated with R software. The nomogram offers utility for patients’ post-diagnosis, enabling survival prediction and risk assessment before or during treatment.

Results

Comparative analysis of clinical characteristics in stage IV NSCLC patients between training and validation cohorts

In total, 461 eligible NSCLC patients were identified in accordance with predefined criteria for inclusion and exclusion. Participants were assigned to training (n=323) and validation (n=138) groups in a 7:3 ratio randomly. In the training cohort, male and female patients comprised 46.7% and 53.3% of the population, respectively. A comparative analysis of patient attributes in training and validation sets is detailed in Table 1, with no statistically significant variations detected (P>0.05).

Utilizing Cox regression to identify prognostic factors within the training cohort

Prior to conducting Cox proportional hazards analyses in the training cohort, all variables were tested for compliance with the equal proportional risk assumption, and all satisfied this criterion (Table S1).

Predictive elements were identified by Cox regression analyses, with the findings detailed in Table 2. Univariate analysis identified nine variables (stage, brain metastasis, CEA, NSE, ProGRP, CYFRA 21-1, EGFR-TKIs, radiotherapy, and chemotherapy) as potential prognostic factors for OS. All identified variables demonstrated statistical significance (P<0.05). To minimize confounding effects, a multivariate regression was performed on nine key factors mentioned above. Final results indicated that six factors (brain metastasis, NSE, CYFRA 21-1, EGFR-TKIs, radiotherapy, and chemotherapy) demonstrated statistical significance (P<0.05) and could be considered independent prognostic factors.

Establishment and verification of the nomogram for EGFR-TKI treatment in stage IV NSCLC

Following the screening process, six variables (brain metastasis, NSE, CYFRA 21-1, EGFR-TKIs, radiotherapy, and chemotherapy) were selected for inclusion in the nomogram. The nomogram enables direct prediction of OS by assigning scores to each patient variable and calculating a cumulative total score. The nomogram functions as follows: each variable is assigned a specific point, which is marked by drawing a vertical line upwards to the corresponding position on the point axis. The cumulative points across all variables provide the total point, from which a downward projection on the total points axis indicates the predicted survival probability of the patient at 1, 3, and 5 years. For instance, a patient with advanced NSCLC receiving icotinib (100 points), with no brain metastasis (0 points), no radiotherapy (67.5 points), receiving chemotherapy (0 points), NSE-negative (0 points), and positive for cytokeratin 19 fragment (87.5 points), yields a prognostic score of 255, reflecting survival probabilities of 87%, 53%, and 24% at 1, 3, and 5 years, respectively (Figure 1).

Figure 1 Prognostic nomogram model for EGFR-TKIs treatment in stage IV non-small cell lung cancer. CYFRA 21-1, cytokeratin-19 fragment; EGFR-TKI, epidermal growth factor receptor tyrosine kinase inhibitor; NSE, neuron-specific enolase.

C-index and area under the curve (AUC) values were applied to evaluate the accuracy and discriminative power of the nomogram model. In the training cohort, the nomogram achieved a C-index of 0.713, with AUC values of 0.771, 0.772, and 0.768 for 1-, 3-, and 5-year survival, respectively (Figure 2). For the validation cohort, the nomogram displayed a C-index of 0.713, with 1-, 3-, and 5-year AUC values of 0.802, 0.761, and 0.722, respectively (Figure 2). These findings demonstrate a strong concordance between observed outcomes and predicted probabilities. Calibration curves for both cohorts align closely with the diagonal line, indicating an absence of bias in model predictions (Figure 3). DCA was conducted for the training and validation cohorts (Figures 4,5), confirming the nomogram’s strong clinical applicability.

Figure 2 Receiver operating characteristic curves for the nomogram model. Receiver operating characteristic curves for 1-, 3-, and 5-year predictions in the training cohort (A-C) and validation cohort (D-F). AUC, area under the curve.

Figure 3 Calibration curves. Calibration curves for 1-, 3-, and 5-year total survival rates in (A-C) the training group, (D-F) the validation group.

Figure 4 DCA for the training cohort. DCA, decision curve analysis.

Figure 5 DCA for the validation cohort. DCA, decision curve analysis.

The patients in the training group were scored, and the best cutoff value of 221.19 was taken and categorizing them into groups of high and low risk (Figure S1). Compared to the high-risk group, the low-risk group exhibited a notably greater likelihood of survival (P<0.001) (Figure 6). The validation group also confirmed this notable difference (Figure 7). This further validates the accuracy and feasibility of the nomogram model, enabling effective risk stratification and improving patient management.

Figure 6 Comparison of survival probability between high- and low-risk groups in the training cohort.

Figure 7 Comparison of survival probability between high- and low-risk groups in the validation cohort.

Discussion

EGFR-TKI-treated NSCLC patients will unavoidably acquire resistance, resulting in disease advancement. Significantly, there are limited therapies available for patients resistant to osimertinib (11). Despite substantial efforts to develop fourth-generation EGFR-TKIs, these compounds have yet to be commercially approved (7). Furthermore, advanced NSCLC patients exhibit a poorer prognosis and reduced survival duration, underscoring the need for focused investigation into stage IV cases and further research efforts.

This study primarily focused on OS following targeted therapy in advanced NSCLC patients. Due to limitations in available clinical data, we included and examined as many factors as feasible, aiming to achieve a streamlined yet effective approach. Through the integration, grouping, and analysis of clinical data from 461 patients, we developed a nomogram to forecast survival rates at 1, 3, and 5 years for advanced NSCLC patients receiving targeted therapies, intending to provide a helpful tool for clinicians assessing patient prognosis. Six independent prognostic factors for OS were identified through Cox regression analyses: brain metastasis, NSE, CYFRA 21-1, EGFR-TKIs, radiotherapy, and chemotherapy. Moreover, the validation process fully demonstrates the solid predictive accuracy of the nomogram for OS.

Our study demonstrated the following: (I) patients with brain metastases exhibited an elevated mortality risk compared to those without [hazard ratio (HR) =1.66; 95% confidence interval (CI): 1.04–2.66, P=0.03]. (II) NSE-positive patients exhibited an elevated mortality risk than NSE-negative patients (HR =1.46; 95% CI: 1.02–2.09, P=0.04). (III) CYFRA 21-1-positive patients were at an elevated risk of mortality compared to CYFRA 21-1-negative patients (HR =1.70; 95% CI: 1.16–2.50, P=0.007). (IV) Among EGFR-TKIs, icotinib posed a greater mortality risk than osimertinib (HR =2.04; 95% CI: 1.32–3.15, P=0.001), while no similar association was observed with gefitinib. (V) Radiotherapy showed a markedly reduced mortality risk than no radiotherapy (HR =0.66; 95% CI: 0.46–0.93, P=0.02). (VI) Chemotherapy recipients demonstrated a markedly reduced mortality risk compared to non-recipients (HR =0.55; 95% CI: 0.40–0.77, P<0.001).

Prior research has indicated that NSCLC patients receiving gefitinib experienced a progression-free survival (PFS) of approximately 8 months, alongside an OS of around 17 months (12). In contrast, those who underwent treatment with icotinib demonstrated a significantly better OS of 30.50 months and an improved PFS of 11.20 months (13). Moreover, when examining the effectiveness of osimertinib, the median PFS for patients harboring EGFR-sensitive mutations was reported to be 19.17 months. On the other hand, patients who had EGFR T790M-positive mutations exhibited a comparatively shorter median PFS of 10.58 months (14).

Brain metastases occur in roughly 30% of NSCLC patients; approximately half of these cases are present at diagnosis, while the remainder develop during treatment (15). A study analyzing the long-term survival of advanced EGFR-mutant NSCLC patients found that cerebral metastases were linked to poorer OS compared to non-brain metastases (16). Wang et al. demonstrated that radiotherapy is safe for individuals with stage IV NSCLC harboring EGFR mutations, and enhances both survival rate and life quality. The average survival reached 17.4 months for those on TKI alone and 25.5 months when combining TKI with radiotherapy (P<0.05) (17). Combination chemotherapy with EGFR-TKIs inhibits extracellular signal-regulated kinase and Akt activation, promotes coordinated induction of apoptosis, and delays drug resistance (8). Multiple researches have demonstrated that combining EGFR-TKIs with chemotherapy markedly enhances OS and PFS in patients with EGFR-mutant advanced NSCLC (18,19).

Within the realm of EGFR-TKI research for lung cancer, most studies on NSE focus on its function in small cell lung cancer transformation, where NSE plays the role of a predictor of small-cell transformation in NSCLC patients undergoing EGFR-TKI therapy (20-22). Dong et al. identified NSE as an independent predictive marker for predicting PFS in advanced NSCLC patients undergoing targeted therapy (23). However, their study did not examine OS, a gap that our research addresses. Elevated CYFRA 21-1 levels are indicative of a negative prognosis for NSCLC (24). In contrast to the normal levels observed, individuals with increased CYFRA 21-1 levels exhibited shorter OS (385 vs. 607 days, P=0.001) and PFS (99 vs. 123.5 days, P=0.01) (25).

In the context of targeted therapies for advanced NSCLC, most prior studies utilizing nomograms have primarily focused on PFS (26) or short-term efficacy outcomes (27). According to our knowledge, researches on predicting OS in patients with advanced EGFR-mutated NSCLC treated with EGFR-TKIs using nomograms are limited. Ng et al. have developed a nomogram to predict OS in metastatic and recurrent NSCLC populations treated with EGFR-TKIs, their study included only 199 participants and failed to account for the potential influence of lung cancer biomarkers. Additionally, it did not incorporate patients treated with osimertinib, nor did it consider the effects of chemotherapy (28). Our study addresses these gaps. The clinical characteristics included in our research are readily accessible across hospitals of various levels and provide a highly convenient and effective tool for physicians and patients.

We acknowledge that treating the treatment category (radiotherapy, chemotherapy, immunotherapy) as a single variable underestimates the complexity of the treatment regimen, but because of the complexity of retrospective data collection and difficulties in ensuring data integrity, we ultimately treated it as a single variable. If radiotherapy combined with chemotherapy was included in the nomogram model, for patients who received both radiotherapy and chemotherapy, the model would calculate the radiotherapy score, chemotherapy score, and radiotherapy combined with chemotherapy score separately. This would cause the model to be unreasonable, and repeat scoring would overestimate patients’ prognosis. The purpose of nomograms is to provide clinicians with intuitive prediction tools, and repeated scoring can complicate and reduce utility. This is a limitation of our study, which we will refine in future studies. The nomogram can be used to predict the OS of stage IV EGFR-mutated NSCLC patients who received EGFR-TKIs as first-line treatment. As long as patients have received first-line EGFR-TKIs treatment, the nomogram can be used for long-term and dynamic prediction, regardless of subsequent treatment.

Notably, this study is retrospective and has a relatively limited participant pool. Secondly, the analysis did not incorporate data regarding tumor size, lymph node metastasis, and primary tumor location. Thirdly, the absence of an external validation cohort to evaluate the model’s accuracy represents a further limitation. Additionally, this study focused on icotinib, gefitinib, and osimertinib as the primary research subjects, as these drugs are widely prescribed in China due to their inclusion in the national medical insurance reimbursement scheme. Although erlotinib, afatinib, and other EGFR-TKIs are more frequently utilized globally, their clinical adoption in China is limited by the constraints of the national medical insurance reimbursement policies. Consequently, the proportion of patients opting for these therapies remains relatively low, and they were excluded from the current investigation. The findings of this study are primarily relevant to China and other regions where the analyzed drugs are widely used, and their generalizability to other populations may be limited. Despite being covered by health insurance, other third-generation EGFR-TKIs, such as furmonertinib and almonertinib, were excluded from this study due to their limited clinical use (less than 2 years) and insufficient follow-up data. Future research endeavors will conduct multicenter clinical studies and incorporate a broader range of international patient data to validate and expand upon the current findings.

Conclusions

We created and confirmed the accuracy of a nomogram using data from 461 patients in predicting OS rates of patients with advanced NSCLC carrying EGFR mutations undergoing treatment through EGFR-TKIs. This model enables physicians to conduct personalized prognostic assessments and supports clinical decision-making, offering meaningful implications for clinical practice.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the TRIPOD reporting checklist. Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2024-2112/rc

Data Sharing Statement: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2024-2112/dss

Peer Review File: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2024-2112/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2024-2112/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the institutional ethics board of Tongji Hospital (No. TJ-IRB202408023) and individual consent for this retrospective analysis was waived.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Li H, Li Y, Zheng X, et al. RBM15 facilitates osimertinib resistance of lung adenocarcinoma through m6A-dependent epigenetic silencing of SPOCK1. Oncogene 2025;44:307-21. [Crossref] [PubMed]
2. Gu Z, Huang P, Zhao J, et al. Bilateral diffuse metastases in advanced lung adenocarcinoma harboring EGFR mutations was associated with a favorable prognosis to EGFR-TKIs. Int J Cancer 2024;154:1979-86. [Crossref] [PubMed]
3. Liu Y, Xiong ZC, Sun X, et al. Impact of apatinib in combination with osimertinib on EGFR T790M-positive lung adenocarcinoma. Transl Cancer Res 2019;8:2151-63. [Crossref] [PubMed]
4. He J, Huang Z, Han L, et al. Mechanisms and management of 3rd generation EGFR TKI resistance in advanced non small cell lung cancer Int J Oncol 2021;59:90. (Review). [Crossref] [PubMed]
5. Tan X, Wang S, Xia H, et al. Prognosis prediction of icotinib as targeted therapy for advanced EGFR-positive non-small cell lung cancer patients. Invest New Drugs 2023;41:463-72. [Crossref] [PubMed]
6. Leonetti A, Sharma S, Minari R, et al. Resistance mechanisms to osimertinib in EGFR-mutated non-small cell lung cancer. Br J Cancer 2019;121:725-37. [Crossref] [PubMed]
7. Pan Q, Lu Y, Xie L, et al. Recent Advances in Boosting EGFR Tyrosine Kinase Inhibitors-Based Cancer Therapy. Mol Pharm 2023;20:829-52. [Crossref] [PubMed]
8. Kuang L, Wang P, Zhou L, et al. Strategies and influencing factors for the treatment of advanced non-small cell lung cancer based on epidermal growth factor receptor tyrosine kinase inhibitors: a narrative review. Transl Cancer Res 2024;13:5123-40. [Crossref] [PubMed]
9. Hsu WH, Yang JC, Mok TS, et al. Overview of current systemic management of EGFR-mutant NSCLC. Ann Oncol 2018;29:i3-9. [Crossref] [PubMed]
10. Rami-Porta R, Nishimura KK, Giroux DJ, et al. The International Association for the Study of Lung Cancer Lung Cancer Staging Project: Proposals for Revision of the TNM Stage Groups in the Forthcoming (Ninth) Edition of the TNM Classification for Lung Cancer. J Thorac Oncol 2024;19:1007-27. [Crossref] [PubMed]
11. Laface C, Maselli FM, Santoro AN, et al. The Resistance to EGFR-TKIs in Non-Small Cell Lung Cancer: From Molecular Mechanisms to Clinical Application of New Therapeutic Strategies. Pharmaceutics 2023;15:1604. [Crossref] [PubMed]
12. Noronha V, Patil VM, Joshi A, et al. Gefitinib Versus Gefitinib Plus Pemetrexed and Carboplatin Chemotherapy in EGFR-Mutated Lung Cancer. J Clin Oncol 2020;38:124-36. [Crossref] [PubMed]
13. Shi YK, Wang L, Han BH, et al. First-line icotinib versus cisplatin/pemetrexed plus pemetrexed maintenance therapy for patients with advanced EGFR mutation-positive lung adenocarcinoma (CONVINCE): a phase 3, open-label, randomized study. Ann Oncol 2017;28:2443-50. [Crossref] [PubMed]
14. Yi L, Fan J, Qian R, et al. Efficacy and safety of osimertinib in treating EGFR-mutated advanced NSCLC: A meta-analysis. Int J Cancer 2019;145:284-94. [Crossref] [PubMed]
15. Yen CT, Wu WJ, Chen YT, et al. Surgical resection of brain metastases prolongs overall survival in non-small-cell lung cancer. Am J Cancer Res 2021;11:6160-72.
16. Tompkins W, Grady CB, Hwang WT, et al. Characteristics of Long-Term Survivors With EGFR-Mutant Metastatic NSCLC. JTO Clin Res Rep 2024;5:100669. [Crossref] [PubMed]
17. Wang XS, Bai YF, Verma V, et al. Randomized Trial of First-Line Tyrosine Kinase Inhibitor With or Without Radiotherapy for Synchronous Oligometastatic EGFR-Mutated Non-Small Cell Lung Cancer. J Natl Cancer Inst 2023;115:742-8. [Crossref] [PubMed]
18. Lin J, Li M, Chen S, et al. Efficacy and Safety of First-Generation EGFR-TKIs Combined with Chemotherapy for Treatment-Naïve Advanced Non-Small-Cell Lung Cancer Patients Harboring Sensitive EGFR Mutations: A Single-Center, Open-Label, Single-Arm, Phase II Clinical Trial. J Inflamm Res 2021;14:2557-67. [Crossref] [PubMed]
19. Hosomi Y, Morita S, Sugawara S, et al. Gefitinib Alone Versus Gefitinib Plus Chemotherapy for Non-Small-Cell Lung Cancer With Mutated Epidermal Growth Factor Receptor: NEJ009 Study. J Clin Oncol 2020;38:115-23. [Crossref] [PubMed]
20. Chen X, Li D, Miao K, et al. Small cell lung cancer transformation after EGFR-TKIs treatment in lung adenocarcinoma: A case report and literatures review. Medicine (Baltimore) 2023;102:e32697. [Crossref] [PubMed]
21. Zhang Y, Li XY, Tang Y, et al. Rapid increase of serum neuron specific enolase level and tachyphylaxis of EGFR-tyrosine kinase inhibitor indicate small cell lung cancer transformation from EGFR positive lung adenocarcinoma? Lung Cancer 2013;81:302-5. [Crossref] [PubMed]
22. Jiang Y, Shou L, Guo Q, et al. Small-cell lung cancer transformation from EGFR-mutant adenocarcinoma after EGFR-TKIs resistance: A case report. Medicine (Baltimore) 2021;100:e26911. [Crossref] [PubMed]
23. Dong J, Tong S, Shi X, et al. Progastrin-Releasing Peptide Precursor and Neuron-Specific Enolase Predict the Efficacy of First-Line Treatment with Epidermal Growth Factor Receptor (EGFR) Tyrosine Kinase Inhibitors Among Non-Small-Cell Lung Cancer Patients Harboring EGFR Mutations. Cancer Manag Res 2020;12:13607-16. [Crossref] [PubMed]
24. Xu Y, Xu L, Qiu M, et al. Prognostic value of serum cytokeratin 19 fragments (Cyfra 21-1) in patients with non-small cell lung cancer. Sci Rep 2015;5:9444. [Crossref] [PubMed]
25. Takeuchi A, Oguri T, Sone K, et al. Predictive and Prognostic Value of CYFRA 21-1 for Advanced Non-small Cell Lung Cancer Treated with EGFR-TKIs. Anticancer Res 2017;37:5771-6. [Crossref] [PubMed]
26. Keam B, Kim DW, Park JH, et al. Nomogram Predicting Clinical Outcomes in Non-small Cell Lung Cancer Patients Treated with Epidermal Growth Factor Receptor Tyrosine Kinase Inhibitors. Cancer Res Treat 2014;46:323-30. [Crossref] [PubMed]
27. Qi H, Hou Y, Zheng Z, et al. Clinical characteristics and MRI based radiomics nomograms can predict iPFS and short-term efficacy of third-generation EGFR-TKI in EGFR-mutated lung adenocarcinoma with brain metastases. BMC Cancer 2024;24:362. [Crossref] [PubMed]
28. Ng IK, Kumarakulasinghe NB, Syn NL, et al. Development, internal validation and calibration of a risk score to predict survival in patients with EGFR-mutant non-small cell lung cancer. J Clin Pathol 2021;74:116-22. [Crossref] [PubMed]