Prediction of the Efficacy and Prognosis of Anlotinib in Lung Cancer Patients Using CT Radiomics

Abstract

Purpose: Our study aimed to develop and validate a predictive model based on chest CT radiomics features combined with clinical variables to evaluate the efficacy of anlotinib and its prognostic value in patients with advanced lung cancer. Methods: This single-center retrospective study included 68 patients with advanced lung cancer who received anlotinib monotherapy at a tertiary grade-A hospital in China between January 2021 and July 2024. All patients had experienced disease progression after receiving standardized chemotherapy, targeted therapy, or immunotherapy prior to enrollment, had not undergone radiotherapy, and had completed all prior treatments at our hospital. Chest CT scans were performed before anlotinib treatment, and radiomics features were extracted. Based on treatment response, patients were grouped, and a radiomics model, a clinical model, and a combined model were constructed. Model performance was evaluated using receiver operating characteristic (ROC) curves, the Hosmer–Lemeshow test, calibration curves, and decision curve analysis (DCA), with internal validation performed via bootstrap resampling (500 iterations). Additionally, Kaplan–Meier survival curves were generated for the high- and low-risk groups predicted by the combined model, and survival differences were compared using the log-rank test. Results: The areas under the ROC curve (AUCs) for the radiomics, clinical, and combined models were 0.721, 0.812, and 0.866, respectively, with the combined model significantly outperforming the other two models (DeLong test, P < .05). Internal validation showed an AUC of 0.866 (95% CI: 0.751-0.967) for the combined model. The integrated model demonstrated good calibration via Hosmer-Lemeshow testing (χ² = 7.81, P = .553) and showed significant clinical net benefit within the threshold probability range of 0.15-0.85 on decision curve analysis. Survival analysis revealed statistically significant differences between the actual treatment-responsive and non-responsive groups (log-rank P < .05), as well as between the model-predicted high-risk and low-risk groups (log-rank P < .05). Multivariable Cox regression confirmed the nomogram score derived from the integrated model as an independent predictor of overall survival (HR = 1.263, 95% CI: 1.090-1.463, P = .002). Conclusion: The combined model incorporating chest CT radiomics features and clinical characteristics demonstrated high accuracy and clinical utility in predicting the efficacy and prognosis of anlotinib treatment in patients with advanced lung cancer.

Keywords

advanced lung cancer anlotinib (ANL)radiomics chest computed tomography predictive model efficacy prognosis

Introduction

Lung cancer remains one of the leading malignancies worldwide in both incidence and mortality.^1,2 Treatment options for patients with advanced-stage lung cancer are extremely limited.^3,4 Anlotinib, an orally administered multi-targeted anti-angiogenic agent, was approved by the China National Medical Products Administration on May 9, 2018, for third-line treatment in patients with advanced non-small cell lung cancer (NSCLC).⁵ In recent years, its use has expanded to include advanced small cell lung cancer as well.⁶ Clinical data have shown that anlotinib significantly prolongs progression-free survival (PFS) and overall survival (OS) in patients with advanced lung cancer.^7,8

Radiomics, an emerging analytical method, extracts high-throughput quantitative features from medical images and combines them with bioinformatics tools for analysis. This approach provides new avenues for disease diagnosis, treatment response evaluation, and prognosis prediction.⁹ It has been widely applied in predicting therapeutic efficacy and prognosis across various cancers.^10,11

However, comprehensive studies combining radiomics and the efficacy of anlotinib in advanced lung cancer remain relatively limited. Specifically, research focusing on survival analysis based on radiomics-driven efficacy prediction is scarce.¹²

Therefore,this study aims to construct a predictive model based on chest CT radiomics features to explore the efficacy of anlotinib in patients with advanced lung cancer and to further analyze its impact on patient survival outcomes.

Materials and Methods

This study was reported in accordance with the TRIPOD guidelines.¹³

Patient Selection and Clinical Data

This retrospective study collected data from 68 patients with advanced lung cancer treated with anlotinib monotherapy at a tertiary grade-A hospital in China between January 2021 and July 2024. The clinical data collected included gender, age, Eastern Cooperative Oncology Group (ECOG) performance status, pretreatment tumor markers such as carcinoembryonic antigen (CEA), neuron-specific enolase (NSE), cytokeratin 19 fragment (CYFRA21-1), and PFS. Additionally, chest CT images of patients before treatment were obtained. Flow-diagram of study was shown in Figure 1.

Figure 1.

Flow-Diagram of Study.

Treatment and Evaluation Criteria

Treatment

Patients with advanced lung cancer received anlotinib monotherapy. The prescribed dosage of anlotinib was 14 mg/day, administered for two consecutive weeks followed by a one-week break as one treatment cycle. For patients with intolerable side effects, the dosage could be reduced to 8 mg/day. Efficacy was evaluated after at least two treatment cycles.^14,15 No other medications, such as targeted agents, were administered concurrently during the period of anlotinib treatment.

Efficacy Evaluation

Treatment efficacy was assessed according to the Response Evaluation Criteria in Solid Tumors (RECIST 1.1).¹⁶ Patients were classified into the following categories: complete response (CR), partial response (PR), stable disease (SD), and progressive disease (PD). The specific classifications were as follows: CR: Complete disappearance of all target lesions, sustained for at least 4 weeks. PR: A reduction of at least 30% in the sum of the diameters of target lesions compared with baseline, sustained for at least 4 weeks. SD: Lesion shrinkage that did not meet the criteria for PR, or lesion enlargement that did not meet the criteria for PD. PD: An increase of at least 20% in the sum of the diameters of target lesions compared with the smallest recorded value, with an absolute increase of at least 5 mm, or the appearance of new lesions. In this study, patients with CR, PR, and SD were categorized into the control group, while those with PD were placed in the observation group.

Eastern Cooperative Oncology Group (ECOG) Performance Status

The ECOG performance status was used as one of the study variables. The scale ranges from 0 to 5, with 0 indicating fully active with no limitations; 1 indicating restricted in physically strenuous activity but able to carry out light work; 2 indicating capable of self-care but unable to work, with more than 50% of waking hours spent out of bed; 3 indicating limited self-care, with more than 50% of waking hours spent in bed or sitting; 4 indicating completely disabled and confined to bed or chair; and 5 indicating death.^17,18

Imaging, Segmentation, and Feature Extraction

Imaging

All patients underwent chest CT scans prior to anlotinib treatment. CT scans were performed using a Siemens Somatom CT machine with the following parameters: tube voltage of 120 kV, current of 109 mA, and slice thickness and interval both set to 1 mm.

Segmentation and Feature Extraction

All chest CT images were exported via the PACS system and processed using 3D Slicer software (version 5.2.1) as follows: Lesion segmentation: A radiologist (Reader A) with 5 years of experience in thoracic oncology imaging manually delineated the three-dimensional regions of interest (ROIs) of the tumors. All segmentations were performed in a blinded manner, with the reader unaware of the patients’ treatment responses to anlotinib. Feature extraction: Radiomics features were extracted using the PyRadiomics plugin (version 3.0.1), including four categories: first-order statistical features (18 features), three-dimensional shape features (14 features), texture features (68 features, including GLCM, GLRLM, GLSZM, GLDM, and NGTDM), and wavelet-transformed features (744 features).Reproducibility assessment: One month later, a second radiologist (Reader B) with equivalent qualifications re-segmented 20 randomly selected cases from the training set. Feature stability was evaluated using the intraclass correlation coefficient (ICC), and features with ICC > 0.75 were retained for subsequent analysis.

Feature Selection and Radiomics Model Construction

The extracted features were subjected to screening using the least absolute shrinkage and selection operator (LASSO) method, with L1 regularization applied to reduce redundant features. Ten-fold cross-validation was employed to determine the optimal penalty parameter (λ), achieving the best balance between model complexity and predictive performance. Features with non-zero coefficients were retained for inclusion in binary logistic regression, and a radiomics score (Radscore) was calculated for each patient based on the model formula.

Clinical Model Construction and Combined Model

Clinical data, including age, gender, smoking history, ecog score, CEA, NSE, CYFRA21-1, neutrophil-to-lymphocyte ratio (NLR), and platelet-to-lymphocyte ratio (PLR), were subjected to univariate analysis. Variables with a P-value < .1 were included in a multivariate logistic regression model using forward and backward stepwise selection to identify the final predictive factors for constructing the clinical model. The radiomics score (Radscore) was combined with the selected clinical factors to build the combined model.

Follow-Up

Patients underwent follow-up chest CT scans every two months (after two treatment cycles) to assess tumor response. Efficacy was evaluated according to RECIST 1.1 criteria. OS, the secondary endpoint of this study, was defined as the time from the start of treatment to either tumor progression, death, or the last follow-up. The last follow-up was on July 1, 2024. Kaplan-Meier curves were generated to compare survival outcomes between patients classified as having progressed disease based on treatment response after two cycles and those classified as having progression by the optimal model prediction.

Statistical Analysis

Data analysis was performed using R version 4.4.1 and SPSS version 27. Descriptive statistics were presented as mean ± standard deviation or median and interquartile range for normally and non-normally distributed continuous variables, respectively. Group comparisons for normally distributed data were conducted using t-tests, while the Mann-Whitney U test was used for non-normally distributed data. Categorical variables were summarized as percentages and compared using the Chi-square test. Binary logistic regression was used for variables with statistical significance. The optimal model was visualized using a nomogram, and receiver operating characteristic (ROC) curves were generated to evaluate model discrimination, calibration curves for model calibration, and decision curve analysis (DCA) for assessing clinical net benefit at various risk thresholds. DeLong's test was used to compare the area under the curve (AUC) between models. A P-value < .05 was considered statistically significant.

Outcomes

Baseline Characteristics

The baseline characteristics of the 68 patients with advanced lung cancer treated with anlotinib monotherapy are presented in Table 1. After at least two treatment cycles, 47 patients (classified as complete response, partial response, or stable disease) showed no disease progression, while 27 patients experienced disease progression. Significant differences were observed in ecog scores, NLR, and PLR between the two groups (P < .05).

Table 1.

Baseline Characteristics of Patients Stratified by Treatment Response.

Variable	Effective Group (n = 47)	Ineffective Group (n = 21)	Statistic (t / χ² / Z)	P-value
Gender			0.083	.880
Male	36 (76.6%)	15 (71.4%)
Female	11 (23.4%)	6 (28.6%)
Smoking status			0.119	.653
Smoker	20 (42.6%)	7 (33.3%)
Non-smoker	27 (57.4%)	14 (66.7%)
Age (years)	66.7 ± 11.9	64.4 ± 12.1	0.021	.477
ECOG score	2 [1-2]	2 [2-2]	−0.204	.007
Pathological type			3.090	.378
Small cell lung cancer	9 (19.1%)	5 (23.8%)
Adenocarcinoma	21 (41.7%)	10 (47.6%)
Squamous cell carcinoma	16 (34.0%)	4 (19.0%)
Other/Unknown	1 (2.1%)	2 (9.5%)
CEA (ng/mL)	4.56 [2.71-11.2]	7.40 [3.78-20.0]	−0.820	.067
NSE (ng/mL)	17.9 [14.7-24.3]	21.6 [14.0-33.8]	−0.896	.524
CYFRA21-1 (ng/mL)	4.82 [3.10-15.4]	5.60 [3.46-16.4]	16.23 ± 36.78	.360
Fibrinogen (g/L)	4.08 ± 1.03	4.55 ± 0.87	−1.938	.059
NLR	3.45 [2.23-5.11]	5.63 [3.84-17.4]	−2.954	.003
PLR	161 [121-216]	252 [172-449]	−2.131	.033

Construction of the Radiomics Model

Radiomics Analysis

The extracted features were subjected to screening using the LASSO method, with L1 regularization applied to reduce redundant features. Ten-fold cross-validation was employed to determine the optimal penalty parameter (λ), achieving the best balance between model complexity and predictive performance. Features with non-zero coefficients were retained for inclusion in binary logistic regression, and a radiomics score (Radscore) was calculated for each patient based on the model formula (Figure 2).

Figure 2.

(A) Selection of the Tuning Parameter λ in the LASSO Model. (B) Three Non-Zero Coefficient Features Obtained Based On the Optimal λ.

Construction of the Radiomics Model

The three radiomic features selected in 3.2.1 were: wavelet-HLL-glcm-Imc1, wavelet-HHL-glszm-LowGrayLevelZoneEmphasis, and wavelet-HHH-ngtdm-Coarseness. These features were incorporated into a multivariate logistic regression model, resulting in the following radiomics model expression:

radmodel = 0.431-0.032wavelet-HLL-glcm-Imc1-12.66wavelet-HHL-glszm-LowGrayLevelZoneEmphasis (Note: The feature wavelet-HHH-ngtdm-Coarseness was excluded from the final model due to a near-zero coefficient).

The radiomics score (RadScore) was calculated using this formula for subsequent integration into the combined model.

Construction of the Clinical Model

Clinical variables with a P-value < .1 from univariate analysis were included in a forward and backward stepwise multivariate logistic regression. Two significant variables, the NLR and Eastern Cooperative Oncology Group (ECOG) score, were selected. The clinical model was constructed as follows:

climodel = 1.188 ecog + 0.19 NLR - 4.092

Construction of the Combined Model

The combined model was developed by integrating the RadScore with the significant clinical factors. The combined model expression was as follows:

combinemodel = 0.89 1 radsore + 1.211 ecog + 0.173 NLR - 3.279

A nomogram was constructed to visualize the combined model, as shown in Figure 3.

Figure 3.

Nomogram of the Combined Model Predicting the Efficacy of Anlotinib. “Radscore” Represents the Radiomics Score (Radscore), “nlr” Represents the Neutrophil-to-Lymphocyte Ratio, and “Ecog” Represents the Ecog Score.

Model Evaluation

Evaluation Using Receiver Operating Characteristic (ROC) Curves

The ROC curve analysis showed that the AUC for the radiomics model (RadModel) was 0.721 (95% CI: 0.589-0.854). The AUC for the clinical model (CliModel) was 0.812 (95% CI: 0.699-0.924). The AUC for the combined model (CombineModel) was 0.866 (95% CI: 0.757-0.975). Pairwise Delong tests revealed that the differences between the models were statistically significant (P < .05 for all comparisons). See Figure 4A.

Figure 4.

(A) ROC Curves for the Radiomics Model, Clinical Model, and Combined Model Predicting the Efficacy of Anlotinib. (B) ROC Curve for the Combined Model, Internally Validated Using 500 Bootstrap Resamples, with an AUC of 0.866 (95% CI: 0.751-0.967). The Shaded Area Represents the 95% Confidence Interval of the AUC.

The best-performing model, the combined model (CombineModel), was further validated internally using bootstrap sampling with 500 iterations. In the validation set, the AUC of the combined model was 0.866 (95% CI: 0.751-0.967), as shown in Figure 4B.

Model Calibration Evaluation

The calibration of the clinical model, radiomics model, and combined model was assessed using the Hosmer-Lemeshow test. The results were as follows: for the clinical model, X² = 7.178623, P = .6185282; for the radiomics model, X² = 7.57307, P = .0404625; and for the combined model, X² = 7.810795, P = .5533204. The calibration curve for the combined model is shown in Figure 5.

Figure 5.

Calibration Curve for Predicting the Efficacy of Anlotinib in Advanced Lung Cancer Patients Using the Combined Model.

Clinical Decision Curve

The decision curve analysis showed that the combined model provided a net benefit and clinical value within the threshold range of 0.15 to 0.85 (Figure 6).

Figure 6.

Clinical Decision Curve for Predicting the Efficacy of Anlotinib in Advanced Lung Cancer Patients Using the Combined Model.

Survival Prediction

As of July 1, 2024, all patients had completed follow-up. TEfficacy-based subgroup analysis revealed significantly prolonged median PFS in the anlotinib-responsive group (409 days; 95% CI: 224 days-not reached) compared to the non-responsive group (72 days; 95% CI: 59-176 days), with this difference being statistically significant (log-rank P = 5.02 × 10⁻¹¹).Risk-stratified analysis using the combined model demonstrated superior median PFS in the low-risk group (321 days; 95% CI: 206-not reached) versus the high-risk group (90 days; 95% CI: 69-211 days; log-rank P = .00114). The 6-month PFS rate was markedly higher in the anlotinib-responsive group (75.02%) than in the non-responsive group (20.17%), with log-rank testing confirming statistically significant between-group differences (P < .0001). Similarly, the model-predicted 6-month PFS rates showed significant discrimination between risk groups (low-risk: 47.3% vs high-risk: 73%; log-rank P = .0011), indicating robust risk stratification capability (Figure 7).

Figure 7.

(A) Shows the Kaplan-Meier Survival Curves Based On the Actual Efficacy of Anlotinib in Advanced Lung Cancer Patients (Group = 0 Indicates Effective, Group = 1 Indicates Ineffective). (B) Shows the Kaplan-Meier Survival Curves Based On the Predicted Efficacy of Anlotinib by the Combined Model (Predmodel1 = High Indicates Effective, Predmodel1 = Low Indicates Ineffective). The X-Axis Represents Time in Days, and the Shaded Area Represents the 95% Confidence Interval of Survival Probability.

Discussion

We developed and validated a novel integrated model combining thin-section chest CT radiomics with clinical features to predict treatment response and prognosis in advanced lung cancer patients receiving anlotinib. As a broad-spectrum oral antiangiogenic agent developed in China, anlotinib has gained widespread clinical application for advanced lung cancer.^19–22 While chest CT remains the standard imaging modality for therapeutic evaluation, radiomics provides a transformative approach by extracting quantitative imaging features that capture tumor heterogeneity beyond visual assessment.^23,24

Radiomics Model Construction

The LASSO regression-selected radiomics model achieved an AUC of 0.721 (95% CI: 0.589-0.854), demonstrating the predictive value of wavelet-HLL-glcm–Imc1 and wavelet-HHL-glszm-LowGrayLevelZoneEmphasis features. These wavel- et-transformed features may reflect tumor vascularity - particularly relevant si-nce anlotinib primarily exerts antitumor effects through angiogenesis inhibition. Our findings align with Cui's research suggesting wavelet features can indirectl-y characterize tumor vasculature.²⁵ However, as radiomics alone may not fully capture patients’ clinical status, we pursued multimodal integration.^26,27

The Clinical Model in This Study

Through univariate analysis and stepwise regression, we identified NLR (reflecting systemic inflammation) and ECOG performance status as key predictors. The clinical model showed superior performance (AUC = 0.812, 95% CI: 0.699-0.924), underscoring the prognostic value of host factors.²⁸

The Combined Model in the Study

The combined radiomics-clinical model demonstrated significantly enhanced predictive accuracy (AUC = 0.866, 95% CI: 0.757-0.975), providing a robust tool for personalized therapeutic decision-making.^29,30

Model Validation

Bootstrap internal validation confirmed model stability (AUC = 0.866, 95% CI: 0.751-0.967). Excellent calibration was evidenced by Hosmer-Lemeshow testing (χ² = 7.81, P = .553). Decision curve analysis revealed substantial clinical net benefit across threshold probabilities of 15%-85%.

Survival Analysis

The model effectively stratified patients into distinct prognostic groups (log-rank P < .05). While predicted median survival (low-risk: 321 days; high-risk: 95 days) showed some deviation from observed outcomes (responders: 409 days; non-responders: 72 days), the model maintained strong discriminative ability. Multivariable Cox regression confirmed the nomogram score as an independent OS predictor (HR = 1.263, 95% CI: 1.090-1.463, P = .002).

Model Benefits, Challenges and Next Steps

The combined model developed in this study demonstrates certain innovations and clinical value. A review of the literature shows that most predictive models focus on NSCLC, such as the anlotinib prediction model for NSCLC developed by Chen et al²³ In contrast, this study includes patients with small cell lung cancer (SCLC), offering the first radiomics-based efficacy prediction exploration for SCLC patients. Although this study includes advanced lung cancer patients with different pathological types (including SCLC, squamous cell carcinoma, and adenocarcinoma), a statistical test was performed on the pathological distribution between the effective and ineffective anlotinib groups. The results showed a Z-value of 3.090 and a P-value of .378, suggesting no significant difference in pathological type distribution between the two groups. Therefore, despite not including pathological type as a model variable, this result indicates its limited influence on treatment efficacy grouping and is unlikely to be a significant source of bias.

Secondly, the features used in the model are all derived from routine CT scans performed before medication. This allows clinicians to perform risk assessments and predict survival time before making treatment decisions. Furthermore, since anlotinib is a high-cost drug and patients often face a significant financial burden, this model provides evidence-based treatment recommendations for advanced-stage patients. Clinically, the combined model can be used to predict the efficacy of anlotinib for a specific advanced lung cancer patient. If the model's prediction is favorable, a doctor can consider prescribing anlotinib and predicting the patient's survival time, based on the patient's overall condition and financial capacity. This can help avoid unnecessary financial expenses and enhance the patient's confidence in treatment. Conversely, if the prediction suggests poor efficacy, alternative treatment plans can be considered.

Despite the significant predictive capabilities of the combined model, several limitations must be addressed. First, due to the relatively small sample size and the diverse tumor morphologies of patients receiving anlotinib, there were challenges in outlining ROIs and extracting features. This, combined with the limited sample size, may increase the risk of overfitting in the radiomics model. Moreover, image delineation was performed manually, and while consistency was evaluated through repeated segmentation, the potential for subjectivity cannot be entirely avoided.³¹

Additionally, the Eastern Cooperative Oncology Group (ECOG) performance status has a strong subjective component, which can lead to inter-rater variability. The NLR is susceptible to various non-tumor factors, such as infections or stress responses, which may interfere with its accuracy in cancer prognosis.^32,33 This study was a retrospective analysis, which may be subject to selection bias, and some clinical data were incomplete. Finally, this study did not perform a stratified analysis on dose adjustments, so the potential impact of dose on radiomics and biomarkers cannot be completely ruled out. Future research should consider controlling for dose in the study design or evaluating its effect through subgroup analysis. Therefore, future studies should consider conducting prospective, multi-center, large-scale studies to further validate the predictive performance and clinical application value of the model.

Conclusion

This study demonstrates that a combined model based on chest CT radiomics and clinical features has significant clinical value in predicting the efficacy and prognosis of anlotinib in advanced lung cancer patients. Despite some limitations, the results of this study provide a valuable reference for developing personalized treatment strategies in the future. Larger-scale prospective studies will help confirm these findings and advance the application of precision medicine in lung cancer treatment.^34–36

Footnotes

Abbreviations

ORCID iD

Yan Zhang

Ethics Approval

This study was approved by the Ethics Committee of Dazhou Central Hospital (Approval No. 2022-YX-071). All procedures involving human participants were conducted in accordance with the ethical standards of the institutional research committee and the Declaration of Helsinki, as well as any relevant national regulation.

Consent to Participate

Informed consent was obtained from all participants included in the study.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the the Key Research and Development Program of Sichuan Science and Technology Department, (grant number 2023YFS0469).

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability

The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.

References

Thai

Solomon

Sequist

Gainor

Heist

. Lung cancer. Lancet. 2021;398(10299):535-554. doi:10.1016/S0140-6736(21)00312-3

Wang

Zhou

. Lung cancer in China: current and prospect. Curr Opin Oncol. 2021;33(1):40-46. doi:10.1097/CCO.0000000000000703

Zugazagoitia

Paz-Ares

. Extensive-stage small-cell lung cancer: first-line and second-line treatment options. J Clin Oncol. 2022;40(6):671-680. doi:10.1200/JCO.21.01881

Meyer

Fitzgerald

Paz-Ares

, et al. New promises and challenges in the treatment of advanced non-small-cell lung cancer. Lancet. 2024;404(10454):803-822. doi:10.1016/S0140-6736(24)01029-8

Han

Zhang

Bao

Liu

Wang

Qian

. Anlotinib enhances the antitumor immunity of radiotherapy by activating cGAS/STING in non-small cell lung cancer. Cell Death Discov. 2022;8(1):468. Published 2022 Nov 28. doi:10.1038/s41420-022-01256-2.

Cheng

Wang

, et al. Anlotinib vs placebo as third- or further-line treatment for patients with small cell lung cancer: a randomised, double-blind, placebo-controlled phase 2 study. Br J Cancer. 2021;125(3):366-371. doi:10.1038/s41416-021-01356-3

Shen

Zheng

Ren

, et al. Anlotinib: a novel multi-targeting tyrosine kinase inhibitor in clinical development. J Hematol Oncol. 2018;11(1):120. Published 2018 Sep 19. doi:10.1186/s13045-018-0664-7.

Luo

, et al. Anlotinib induces a T cell-inflamed tumor microenvironment by facilitating vessel normalization and enhances the efficacy of PD-1 checkpoint blockade in neuroblastoma. Clin Cancer Res. 2022;28(4):793-809. doi:10.1158/1078-0432.CCR-21-2241

Chen

Copley

Viola

Aboagye

. Radiomics and artificial intelligence for precision medicine in lung cancer treatment. Semin Cancer Biol. 2023;93:97-113. doi:10.1016/j.semcancer.2023.05.004

10.

Jochems

Refaee

, et al. Structural and functional radiomics for lung cancer. Eur J Nucl Med Mol Imaging. 2021;48(12):3961-3974. doi:10.1007/s00259-021-05242-1

11.

Chetan

Gleeson

. Radiomics in predicting treatment response in non-small-cell lung cancer: current status, challenges and future perspectives. Eur Radiol. 2021;31(2):1049-1058. doi:10.1007/s00330-020-07141-9

12.

Tunali

Gillies

Schabath

. Application of radiomics and artificial intelligence for lung cancer precision medicine. Cold Spring Harb Perspect Med. 2021;11(8):a039537. Published 2021 Aug 2. doi:10.1101/cshperspect.a039537.

13.

Collins

Reitsma

Altman

Moons

. Transparent reporting of a multivariable prediction model for individual prognosis or diagnosis (TRIPOD): the TRIPOD statement. Br Med J. 2015;350:g7594. doi: 10.1136/bmj.g7594

14.

Han

Wang

, et al. Effect of anlotinib as a third-line or further treatment on overall survival of patients with advanced non-small cell lung cancer: the ALTER 0303 phase 3 randomized clinical trial [published correction appears in JAMA Oncol. 2018;4(11):1625. doi: 10.1001/jamaoncol.2018.5345]. JAMA Oncol. 2018;4(11):1569-1575. doi:10.1001/jamaoncol.2018.3039

15.

Yuan

Cheng

Xiao

Wang

Fan

. Efficacy and safety of anlotinib in the treatment of small cell lung cancer: a real-world observation study. Front Oncol. 2022;12:917089. Published 2022 Jun 20. doi:10.3389/fonc.2022.917089.

16.

Seymour

Bogaerts

Perrone

, et al. iRECIST: guidelines for response criteria for use in trials testing immunotherapeutics [published correction appears in Lancet Oncol. 2019;20(5):e242. doi: 10.1016/S1470-2045(19)30240-2]. Lancet Oncol. 2017;18(3):e143-e152. doi:10.1016/S1470-2045(17)30074-8

17.

Cheson

Fisher

Barrington

, et al. Recommendations for initial evaluation, staging, and response assessment of Hodgkin and non-Hodgkin lymphoma: the Lugano classification. J Clin Oncol. 2014;32(27):3059-3068. doi:10.1200/JCO.2013.54.8800

18.

Dewys

Begg

Lavin

, et al. Prognostic effect of weight loss prior to chemotherapy in cancer patients. Eastern Cooperative Oncology Group. Am J Med. 1980;69(4):491-497. doi:10.1016/s0149-2918(05)80001-3

19.

Zhou

Chen

Zhang

, et al. China National Medical Products Administration approval summary: anlotinib for the treatment of advanced non-small cell lung cancer after two lines of chemotherapy. Cancer Commun (Lond). 2019;39(1):36. Published 2019 Jun 20. doi:10.1186/s40880-019-0383-7.

20.

Liang

Hui

, et al. Autophagy inhibition potentiates the anti-angiogenic property of multikinase inhibitor anlotinib through JAK2/STAT3/VEGFA signaling in non-small cell lung cancer cells. J Exp Clin Cancer Res. 2019;38(1):71. Published 2019 Feb 12. doi:10.1186/s13046-019-1093-3.

21.

Yang

Zhang

Wang

. Emerging therapies for small cell lung cancer. J Hematol Oncol. 2019;12(1):47. Published 2019 May 2. doi:10.1186/s13045-019-0736-3.

22.

Lee

Saxena

Giaccone

. Advancements in small cell lung cancer. Semin Cancer Biol. 2023;93:123-128. doi:10.1016/j.semcancer.2023.05.008

23.

Wei

Zhang

Zhao

. Development and validation of a radiomics nomogram for diagnosis of malignant pleural effusion. Discov Oncol. 2023;14(1):213. Published 2023 Nov 24. doi:10.1007/s12672-023-00835-8.

24.

Zhou

Zhang

Guo

, et al. Multiparametric MRI radiomics in prostate cancer for predicting Ki-67 expression and Gleason score: a multicenter retrospective study. Discov Oncol. 2023;14(1):133. Published 2023 Jul 20. doi:10.1007/s12672-023-00752-w.

25.

Yang

Zhang

Zhao

, et al. Radiomic analysis based on magnetic resonance imaging for the prediction of VEGF expression in hepatocellular carcinoma pa-tients. Abdom Radiol (NY). 2024;49(11):3824-3833. doi:10.1007/s00261-024-04427-0

26.

Chen

Wang

, et al. CT-based radiomics can predict the efficacy of anlotinib in advanced non-small-cell lung cancer. J Oncol. 2022;2022:4182540. Published 2022 Dec 26. doi:10.1155/2022/4182540.

27.

Felfli

Liu

Zerka

, et al. Systematic review, meta-analysis and radiomics quality score assessment of CT radiomics-based models predicting tumor EGFR mutation status in patients with non-small-cell lung cancer. Int J Mol Sci. 2023;24(14):11433. Published 2023 Jul 14. doi:10.3390/ijms241411433.

28.

Huang

Prasad

, et al. Association of neutrophil-lymphocyte ratio with survival in peripheral early-stage non-small cell lung cancer after stereotactic body radiation therapy. BMC Cancer. 2023;23(1):254. Published 2023 Mar 18. doi:10.1186/s12885-023-10719-3.

29.

Sengupta

Banerjee

Biswas

, et al. The incidence of hyponatraemia and its effect on the ECOG performance status among lung cancer patients. J Clin Diagn Res. 2013;7(8):1678-1682. doi:10.7860/JCDR/2013/5900.3225

30.

Schuler

Bölükbas

Darwiche

Theegarten

Herrmann

Stuschke

. Personalized treatment for patients with lung cancer. Dtsch Arztebl Int. 2023;120(17):300-310. doi:10.3238/arztebl.m2023.0012

31.

Wang

Herbst

Boshoff

. Toward personalized treatment approaches for n-on-small-cell lung cancer. Nat Med. 2021;27(8):1345-1356. doi:10.1038/s41591-021-01450-2

32.

Sørensen

Klee

Palshof

Hansen

. Performance status assessment in cancer patients. An inter-observer variability study. Br J Cancer. 1993;67(4):773-775. doi:10.1038/bjc.1993.140

33.

Zahorec

. Neutrophil-to-lymphocyte ratio, past, present and future perspectives. Bratisl Lek Listy. 2021;122(7):474-488. doi:10.4149/BLL_2021_078

34.

Capone

Giannarelli

Mallardo

, et al. Baseline neutrophil-to-lymphocyte ratio (NLR) and derived NLR could predict overall survival in patients with advanced melanoma treated with nivolumab. J Immunother Cancer. 2018;6(1):74. Published 2018 Jul 16. doi:10.1186/s40425-018-0383-1.

35.

Sicklick

Kato

Okamura

, et al. Molecular profiling of cancer patients enables personalized combination therapy: the I-PREDICT study. Nat Med. 2019;25(5):744-750. doi:10.1038/s41591-019-0407-5

36.

Jackson

Chester

. Personalised cancer medicine. Int J Cancer. 2015;137(2):262-266. doi:10.1002/ijc.28940