Development and Validation of a Nomogram for Predicting Progression-Free Survival in Advanced Non-Small Cell Lung Cancer Patients Treated with Anlotinib

By Zhengyu Wu, Peng Zhou, Yanan Zhao, Junping Wang, Shan Gao

Affiliations

1. Clinical Research Centre, Hefei Cancer Hospital, Chinese Academy of Sciences, Hebei, China

doi: 10.29271/jcpsp.2025.12.1611

ABSTRACT
Objective: To determine factors influencing progression-free survival (PFS) in patients with advanced non-small cell lung cancer (NSCLC) treated with anlotinib and to create a predictive model using a nomogram.
Study Design: A descriptive study.
Place and Duration of the Study: Clinical Research Centre, Hefei Cancer Hospital, Chinese Academy of Sciences, Hebei, China, from July 2020 to 2024.
Methodology: Data of patients with advanced NSCLC receiving anlotinib treatment were retrieved from medical records. Potential predictors of PFS were first identified using univariate Cox proportional hazards regression analysis, and subsequently, multivariate Cox regression analysis was performed to identify key factors influencing PFS. A nomogram was constructed for PFS prediction.
Results: The study included 145 patients. Six factors were identified as significant predictors: combined use of proton pump inhibitors (PPI), nutritional risk score, monocyte percentage, pathological grading, combined use of immune checkpoint inhibitors (ICI), and body temperature. The combined use of PPI and Pathological Grading Ⅳ was found to be an independent risk factor for PFS, whereas the combined use of ICI, a nutritional risk score of 0, a monocyte percentage below 10.04%, and a body temperature under 36.2°C was identified as an independent protective factor. The areas under the ROC curve (AUC) at 1, 3, 6, and 12 months were 0.765 (95% CI: 0.672-0.859), 0.816 (95% CI: 0.759-0.873), 0.784 (95% CI: 0.718-0.849), and 0.773 (95% CI: 0.673-0.874), respectively. The model demonstrated a C-index of 0.71 (95% CI: 0.65-0.76), and the calibration curves indicated a high degree of alignment between the predicted and actual outcomes. A nomogram was successfully created to predict PFS.
Conclusion: Six significant factors affecting PFS in patients with advanced NSCLC treated with anlotinib were identified. A reliable nomogram was successfully constructed for the prediction of PFS, offering a tool to support personalised treatment planning.

Key Words: Anlotinib, Non-small cell lung cancer, Predictors, Cox regression, Nomogram.

INTRODUCTION

Globally, lung cancer accounts for the highest number of cancer-related fatalities, with 80-85% of these cases classified as non- small cell lung cancer (NSCLC).¹ Most patients present with advan-ced disease, resulting in a poor prognosis.² Targeted therapies and immunotherapies have improved overall survival (OS) and quality of life. Anlotinib, an orally administered multi-target tyrosine kinase inhibitor, works by suppressing tumour growth and preventing metastasis. In clinical comparisons with placebo, anlotinib has shown an increase of approximately 4 months in median progression-free  survival (PFS)  and  3  months  in  median  OS.³

Several factors influence disease progression during anlotinib treatment, including genetic mutations, serological indicators, and combination therapies. Risk factors include obesity, high pre-treatment neutrophil-to-lymphocyte (NLR) ratio, elevated baseline AST, NSE, fibrinogen, ANC, squamous cell histology, non- epidermal growth factor receptor (EGFR) driver mutations, and platelet-to-lymphocyte (PLR) ratios.^4,5 Hypertriglyceridaemia and combination therapy with immune checkpoint inhibitors (ICI) can extend PFS, while no correlation exists between EGFR mutations and efficacy.^6-8 Elevated baseline total cholesterol (TC) and LDL levels  are  associated  with  increased  risk  of  progression.⁹

The variability in patient responses to anlotinib, with disease control durations ranging from 1.5 to ≥18 months, highlights the need for predictive biomarkers.¹⁰ Nomograms, which synthesise multiple risk factors for prognostic estimates, have shown utility in lung cancer management. The researchers aimed to develop and validate a nomogram based on the clinical data retrieved from Hefei Cancer Hospital, Chinese Academy of Sciences, to optimise treatment outcome predictions for patients with NSCLC receiving anlotinib.

METHODOLOGY

A retrospective review was conducted on patients with meta-static or locally advanced NSCLC who received anlotinib monotherapy at the Clinical Research Centre, Hefei Cancer Hospital, Chinese Academy of Sciences, Hebei, China, from July 2020 to 2024. Patient selection was based on the following requirements: histologically verified Stage III/IV NSCLC, failure of two or more lines of standard therapy, completion of anlotinib treatment exceeding 14 days, and the availability of tumour response documentation through imaging studies (ultrasonography and computed tomography) according to RECIST guidelines. The institutional ethics board granted study approval. The research team collected comprehensive patient information, encompassing demographic details, histopathological findings, therapeutic history, baseline blood parameters, treatment-related toxicities, and clinical outcomes.

Data processing and statistical evaluation were performed using the R computing environment (4.2.0). The model construction consisted of two steps. Step 1 involved preliminary variable screening. Given the large number of variables derived from comprehensive patient information, initial screening was necessary to exclude clearly irrelevant variables. Univariate Cox regression analysis was employed with a statistical significance threshold of p <0.1.¹¹ Variables meeting the criterion were included in step 2.

Step 2 involved multivariate Cox regression analysis. All variables meeting the screening criteria from step 1 were incorporated into the multivariate analysis. Forward selection and backward elimination methods were then applied to adjust for potential confounding factors and to identify the final set of variables, with statistical significance set at p <0.05. Continuous variables were dichotomised using optimal cut-offs from the cut-off R package.^12,13

To validate the prognostic tool, bootstrap resampling (n = 1,000) was employed to generate calibration curves, and discriminative ability was calculated via C-statistics.^14,15 The C-index and receiver operating characteristic (ROC) curve were used to evaluate the discriminative ability of the model. Higher values of C-index and the area under the ROC curve (AUC), approaching 1, indicated superior predictive performance, while calibration plots were generated to compare the observed and predicted probabilities. Statistical significance was drawn using a p-value threshold of 0.05.

RESULTS

A total of 145 patients with NSCLC were administered anlotinib during the study period. Their demographic profile and medical characteristics are presented in Table I.

The cohort of 145 anlotinib-treated patients showed a median PFS of 3.1 months (95% CI: 2.47-3.73). Univariate Cox regression analysis was conducted on 69 variables potentially affecting prognosis to evaluate their correlation with PFS. Table II presents the factors with p <0.1 identified in the univariate analysis results.

Table  I:  Patients’   demographic  and  medical  characteristics.

Characteristics	Patients (%)
Combined use of PPI
Yes	68 (46.90)
No	77 (53.10)
Combined use of ICI
Yes	95 (65.52)
No	50 (34.48)
Age >65	65.16 (9.85)
Yes	80 (55.17)
No	65 (44.83)
Gender
Male	104 (71.72)
Female	41 (28.28)
Pathological
Adenocarcinoma	68 (46.90)
Squamous cell carcinoma	77 (53.10)
Epidermal growth factor receptor mutant
Yes	25 (17.24)
No	120 (82.76)
Mutations
Pathological grading
IV	65 (44.83)
III	80 (55.17)
Metastasis
Yes	104 (71.72)
No	41 (28.28)
Brain metastases
Yes	20 (13.79)
No	125 (86.21)
Radiation
Yes	124 (85.52)
No	24 (16.55)
Number of previous treatment lines
2	91 (62.76)
>2	54 (37.24)
Nutritional risk score = 0
Yes	53 (36.55)
No	92 (63.45)
Karnofsky performance status <70
Yes	32 (22.07)
No	113 (77.93)
Smoking
Yes	10 (06.90)
No	135 (93.10)
Alcohol
Yes	8 (5.52)
No	137 (94.48)
ICI: Immune checkpoint inhibitors; PPI: Proton pump inhibitors.

Table  II:  Univariate  Cox  regression  outcomes.

Characteristics	HR (95% CI)	p-values
Nutritional risk score = 0	0.50 (0.30 ~ 0.81)	0.005
Prealbumin	0.99 (0.99 ~ 0.99)	0.016
Body temperature	1.74 (1.10 ~ 2.75)	0.018
Monocyte percentage	1.04 (1.01 ~ 1.07)	0.05
Pathological grading Ⅳ	1.46 (0.95 ~ 2.24)	0.087
Monocyte count	1.75 (0.91 ~ 3.37)	0.096
Combined use of PPI	2.42 (1.56 ~ 3.77)	<.001
Combined use of ICI	0.43 (0.28 ~ 0.67)	<.001
HR: Hazard ratio; CI: Confidence interval; ICI: Immune checkpoint inhibitors; PPI: Proton pump inhibitors. The p-values were determined by univariate Cox regression analysis; p-value is statistically significant at 0.1.

Table  III:  Results  of  multivariate  Cox  regression  analysis.

Characteristics	HR (95% CI)	p-values
Combined with PPI	2.16 (1.36 ~ 3.43)	0.001
Nutritional risk score = 0	0.53 (0.32 ~ 0.88)	0.014
Monocyte percentage <10.04	0.56 (0.34 ~ 0.93)	0.023
Pathological grading IV	2.03 (1.29 ~ 3.19)	0.002
Combined use of ICI	0.55 (0.35 ~ 0.87)	0.01
Body temperature <36.3℃	0.55 (0.33 ~ 0.92)	0.022
HR: Hazard ratio, CI: Confidence interval; ICI: Immune checkpoint inhibitors; PPI: Proton pump inhibitors. The p-values were determined by multivariate Cox regression analysis; p-value is significant at 0.05.

Table II presents potential prognostic factors for NSCLC patients treated with anlotinib, showing only those with p <0.1. The following optimal cut-off values for continuous variables were selected: prealbumin (160), body temperature (36.3°C), monocyte percentage (10.04%), and monocyte count (0.41).

Figure 1: ROC curves of the multivariate proportional hazards regression model at 1, 3, 6, and 12 months.

Table III shows the results of multivariate analysis. The combined use of PPI and pathological grading Ⅳ was identified as an independent risk factor, while the combined use of ICI, nutritional risk score = 0, monocyte percentage <10.04%, and body temperature <36.2°C was found to be an independent protective factor.

As shown in Figure 1, the AUC at 1, 3, 6, and 12 months were 0.765 (95% CI: 0.672-0.859), 0.816 (95% CI: 0.759-0.873), 0.784 (95% CI: 0.718-0.849), and 0.773 (95% CI: 0.673-0.874), respectively. The model demonstrated a C-index of 0.71 (95% CI: 0.65-0.76). The calibration curves (Figure 2) indicated a high degree of alignment between the predicted and actual outcomes. Based on the multivariate Cox model, a nomogram was created to predict PFS at different time points in NSCLC patients treated with anlotinib.

Figure 3 provides a prognostic tool for predicting PFS in NSCLC patients treated with anlotinib. To use the nomogram, the patient's value for each variable was located on its corresponding axis, and a vertical line was drawn upward to the point axis to determine the score assigned to each variable. These scores were then summed, and the total was located on the total points axis. A vertical line was drawn downward from this point to the survival axes to determine the probability of PFS at 1, 3, 6, and 12 months.

Figure 2: Calibration curves of the multivariate proportional hazards regression model at 1, 3, 6, and 12 months. Figure 3: Nomogram for predicting PFS.

DISCUSSION

In this study, a nomogram was developed to predict PFS in NSCLC patients treated with anlotinib. Previous studies have suggested that the use of PPIs combined with tyrosine kinase inhibitors (TKIs) may negatively impact patient survival.¹⁶ The solubility of anlotinib is pH-dependent, and at pH 6.5, its solubility falls below the required level.^17,18 Reduced absorptions of anlotinib when co-administered with PPIs may contribute to the shortened PFS observed in these patients.

This study identified the combination of anlotinib and ICI as a protective factor for prolonged PFS (HR: 0.54; 95% CI: 0.34-0.86), consistent with previous studies. For instance, Yu et al. reported improved PFS with an anlotinib-immuno-therapy combination versus anlotinib monotherapy (HR: 0.68; 95% CI: 0.68-0.97).¹⁹

Univariate Cox analysis showed a 1.74-fold increase in risk for each 1°C rise in body temperature. Otsuka et al. suggested that elevated body temperature promotes cancer cell growth and extracellular vesicle secretion, linking it to poor prognosis in breast cancer.²⁰ Thus, increased body temperature may contribute to disease progression.

This study identified a nutritional risk score of 0 as a protective factor for PFS. Similarly, in a prognostic investigation, Chen et al. discovered that NSCLC patients exhibiting reduced nutritional status (PNI ≤42.48) experienced markedly abbreviated PFS relative to those with optimal nutritional indices (1.5 vs. 4.0 months; p = 0.010),²¹ indicating that nutritional status influences PFS in NSCLC patients treated with anlotinib.

Prior studies indicated that elevated NLR (>7.75), LDH (>254.65 U/L), AST, NSE, and fibrinogen levels were predictive of poor survival outcomes.²² In this study, univariate analysis failed to demonstrate such correlations. The calculated hazard ratios with 95% confidence intervals (0.96-1.04, 0.99-1.01, and 0.99-1.02, respectively) showed no statistical significance for these parameters. Prealbumin (HR 0.99, 95% CI 0.98-0.99) and monocyte count (HR 1.04, 95% CI 1.01-1.07) showed mild statistical significance.

Prealbumin reflects the protein levels and nutritional status of the body, aligning with studies linking higher albumin levels to improved survival in cancer patients.²³ Monocyte count, part of the pan-immune inflammatory index, may predict survival and immune-related adverse events in late- stage NSCLC.²⁴

This investigation identified the following variables potentially predictive of anlotinib efficacy: the concomitant use of PPI, nutritional risk score, monocyte percentage, grade IV histology, immunotherapy combination, and body temperature. A nomogram was subsequently constructed to forecast PFS in post-chemotherapy metastatic NSCLC patients. This prognostic tool generates individualised survival probabilities, facilitating personalised treatment strategies and identifying candidates for intensified surveillance or therapy modification.

However, this study has certain limitations. The single- centre, retrospective design potentially constrains the gene-ralisability of results and yields lower-level evidence compared to prospective approaches. The nomogram demonstrated consistent predictive performance at 6 and 12 months but may have limitations in predicting long-term outcomes. Additionally, due to insufficient data on the OS, the model is restricted to predicting PFS only.

CONCLUSION

This investigation identified prognostic factors influencing PFS among patients with metastatic NSCLC receiving anlotinib therapy. Based on these prognostic factors, a nomogram was developed to help clinicians predict PFS and guide clinical decisions, such as closer monitoring or treatment adjustments before and during therapy.

FUNDING:
This study was funded by the Young Medical Talents programme at Hefei Cancer Hospital, Chinese Academy of Sciences, Hebei, China.

ETHICAL APPROVAL:
This study was conducted with the approval of the Medical Ethics Committee of the Hefei Cancer Hospital, Chinese Academy of Sciences, Hebei, China (Ref. No: PJ-KY2024-042).

PATIENTS’ CONSENT:
All patients signed a general informed consent form upon admission, agreeing to the use of their medical record data for scientific research.

COMPETING INTEREST:
The authors declared no conflict of interest.

AUTHORS' CONTRIBUTION:
ZW: Experimental design, data analysis, and manuscript writing.
PZ, YZ: Experimental data extraction and data analysis.
JW: Experimental data extraction and manuscript writing.
SG: Data analysis and manuscript writing.
All authors approved the final version of the manuscript to be published.

REFERENCES

1. Xiong Q, Qin B, Xin L, Yang B, Song Q, Wang Y, et al, Real- world efficacy and safety of anlotinib with and without immunotherapy in advanced non-small cell lung cancer. Front Oncol 2021; 11:659380. doi: 10.3389/fonc.2021. 659380.
2. Cantelmo AR, Dejos C, Kocher F, Hilbe W, Wolf D, Pircher A. Angiogenesis inhibition in non-small cell lung cancer: A critical appraisal, basic concepts and updates from American Society for Clinical Oncology 2019. Curr Opin Oncol 2020; 32(1):44-53. doi: 10.1097/CCO.0000000000000591.
3. Gao Y, Liu P, Shi R. Anlotinib as a molecular targeted therapy for tumors. Oncol Lett 2020; 20(2):1001-14. doi: 10.3892/ol.2020.11685.
4. Xiong A, Nie W, Cheng L, Zhong H, Chu T, Zhong R, et al. Association between obesity and poor prognosis in patients receiving anlotinib for advanced non-small cell lung cancer. Front Pharmacol 2022; 13:812555. doi: 10.3389/fphar. 2022.812555.
5. Zhou J, Zhao J, Jia Q, Chu Q, Zhou F, Chu X, et al. Peripheral blood autoantibodies against to tumor-associated antigen predict clinical outcome to immune checkpoint inhibitor-based treatment in advanced non-small cell lung cancer. Front Oncol 2021; 11:625578. doi: 10.3389/fonc.2021. 625578.
6. Wang J, Zhao Y, Wang Q, Zhang L, Shi J, Wang Z, et al. Prognostic factors of refractory NSCLC patients receiving anlotinib hydrochloride as the third- or further-line treatment. Cancer Biol Med 2018; 15(4):443-51. doi: 10.20892/j.issn. 2095-3941.2018.0158.
7. Chen Q, Li Y, Zhang W, Wang C, Yang S, Guo Q. Safety and efficacy of ICI plus anlotinib vs. anlotinib alone as third-line treatment in extensive-stage small cell lung cancer: A retrospective study. J Cancer Res Clin Oncol 2022; 148(2):401-8. doi: 10.1007/s00432-021-03858-2.
8. Zhao Y, Wang Q, Zhang L, Shi J, Wang Z, Cheng Y, et al. The efficacy of anlotinib as third-line treatment for non-small cell lung cancer by EGFR mutation status: A subgroup analysis of the ALTER0303 randomized phase 3 study. Transl Lung Cancer Res 2022; 11(5):776-85. doi: 10.21037/tlcr-22-320.
9. Tang M, Song C, Zhang Y, Xu X, Wang C, Zhang Z, et al. Levels of pretreatment blood lipids are prognostic factors in advanced NSCLC patients treated with anlotinib. Lipids Health Dis 2021; 20(1):165. doi: 10.1186/s12944-021-01596-5.
10. Huang A, Wang W, Qin R, Chen G, Gu A, Han B. Efficacy, safety, and prognostic factors of anlotinib treatment in advanced non-small cell lung cancer patients. J Cancer Res Ther 2022; 18(5):1299-305. doi: 10.4103/jcrt.jcrt_835_21.
11. Liu Z, He M, Wang X. A nomogram for predicting the overall survival in rectal cancer patients after total neoadjuvant therapy. Tech Coloproctol 2024; 28(1):107. doi: 10.1007/ s10151-024-02986-4.
12. Cutoff: Seek the Significant Cutoff Value. (Accessed on 3/6/ 2025). Available from: httpss://cran.r-project.org/web/packages/cutoff/index.html.
13. Altman DG, Lausen B, Sauerbrei W, Schumacher M. Dangers of using optimal cutpoints in the evaluation of prognostic factors. J Natl Cancer Inst 1994; 86(11):829-35. doi: 10. 1093/jnci/86.11.829.
14. Wang H, Ou Y, Fan T, Zhao J, Kang M, Dong R, et al. Development and internal validation of a nomogram to predict mortality during the ICU stay of thoracic fracture patients without neurological compromise: An analysis of the MIMIC-III clinical database. Front Public Health 2021; 9:818439. doi: 10.3389/fpubh.2021.818439.
15. Cui Y, Lin J, Sun D, Zhang H, Diao T, Fu Q. Nomogram for predicting the overall survival and cancer-specific survival of patients with intraductal carcinoma of the prostate. J Cancer Res Clin Oncol 2024; 150(2):45. doi: 10.1007/s00432-023- 05582-5.
16. Lee CH, Shen MC, Tsai MJ, Chang JS, Huang YB, Yang YH, et al. Proton pump inhibitors reduce the survival of advanced lung cancer patients with therapy of gefitinib or erlotinib. Sci Rep 2022; 12(1):7002. doi: 10.1038/s41598-022-10938-x.
17. Van Leeuwen RWF, Jansman FGA, Hunfeld NG, Peric R, Reyners AKL, Imholz ALT, et al. Tyrosine kinase inhibitors and proton pump inhibitors: An evaluation of treatment options. Clin Pharmacokinet 2017; 56(7):683-88. doi: 10. 1007/s40262-016-0503-3.
18. Jin Z, He Q, Zhu X, Zhu M, Wang Y, Wu XA, et al. Application of physiologically based pharmacokinetic modelling for the prediction of drug-drug interactions involving anlotinib as a perpetrator of cytochrome P450 enzymes. Basic Clin Pharmacol Toxicol 2022; 130(5):592-605. doi: 10.1111/bcpt.13721.
19. Yu L, Xu J, Qiao R, Han B, Zhong H, Zhong R. Efficacy and safety of anlotinib combined with PD-1/PD-L1 inhibitors as second-line and subsequent therapy in advanced small-cell lung cancer. Cancer Med 2023; 12(5):5372-83. doi: 10.1002/ cam4.5360.
20. Otsuka K, Yamamoto Y, Ochiya T. Uncovering temperature-dependent extracellular vesicle secretion in breast cancer. J Extracell Vesicles 2020; 10(2):e12049. doi: 10. 1002/jev2.12049.
21. Chen T, Liang G, Xiang Z, He J, Xu X, Tang M. Prognostic value of prognostic nutritional index and its variations in advanced non-small-cell lung cancer patients treated with anlotinib monotherapy. J Clin Lab Anal 2022; 36(4):e24300. doi: 10.1002/jcla.24300.
22. Zhang C, Wang J, Wang X, Meng Z, Cheng Y, Li K. Peripheral blood indices to predict PFS/OS with anlotinib as a subsequent treatment in advanced small-cell lung cancer. Cancer Biol Med 2021; 19(8):1249-58. doi: 10.20892/j.issn.2095- 3941.2020.0727.
23. Cehreli R, Yavuzsen T, Ates H, Akman T, Ellidokuz H, Oztop I. Can inflammatory and nutritional serum markers predict chemotherapy outcomes and survival in advanced stage nonsmall cell lung cancer patients? Biomed Res Int 2019; 2019:1648072. doi: 10.1155/2019/1648072.
24. Chen Y, Gong L, Gu P, Hua Y, Sun Y, Ni S, et al. Pan-immune- inflammation and its dynamics: Predictors of survival and immune-related adverse events in patients with advanced NSCLC receiving immunotherapy. BMC Cancer 2023; 23(1): 944. doi: 10.1186/s12885-023-11366-4.