Predictive efficacy of a combined model of radiomics features and clinical signatures for bone metastasis in NSCLC

doi:10.3760/cma.j.cn371439-20251123-00056

Abstract

Abstract:

Objective To explore the predictive efficacy of a combined model integrating intratumoral and peritumoral CT radiomic features and clinical signatures for bone metastasis in non-small cell lung cancer（NSCLC）. Methods A retrospective analysis was conducted on chest arterial-phase CT images and clinical data from 537 pathologically confirmed NSCLC patients at Tangshan People's Hospital between January 2017 and December 2019. Patients were divided into a training cohort（n=376）and a validation set（n=161）in a 7∶3 ratio. Predictive models were developed within the training set， and the predictive efficacy was evaluated in both the training and validation sets respectively， and the clinical application value was verified as well. Using univariate and multivariate logistic regression analysis， the influencing factors for bone metastasis in NSCLC patients were investigated. Radiomics models were established for the tumor interior group， the tumor interior combined with a 3 mm peritumoral group， and the simple 3 mm peritumoral group. The best model was selected and combined with clinical signatures to construct a combined model. The diagnostic efficacy and clinical application value of the model were evaluated using receiver operator characteristic（ROC）curves， the calibration curve， and decision curve analysis（DCA）. Results Among the 537 NSCLC patients， 414 had bone metastasis（290 in training set， 124 in validation set）. Univariate analysis showed that， smoking history， tumor location， T stage， N stage， pathological type， D-dimer， carcinoembryonic antigen（CEA）， cytokeratin fragment antigen 21-1， squamous cell carcinoma antigen， spiculation sign， lobulation sign， pleural indentation sign， and vascular convergence sign were all influencing factors predicting bone metastasis in NSCLC patients（all P<0.05）. Multivariate analysis showed that， T stage（OR=0.69， 95%CI：0.52-0.87， P<0.001）， N stage（OR=0.24， 95%CI：0.13-0.43， P<0.001）， pathological type（OR=6.01， 95%CI：2.83-12.77， P<0.001）， D-dimer（OR=0.32， 95%CI：0.17-0.59， P<0.001）， CEA（OR=0.25， 95%CI：0.14-0.44， P<0.001）， spiculation sign（OR=0.21， 95%CI：0.07-0.65， P=0.007）， and pleural indentation sign（OR=0.32， 95%CI：0.18-0.56， P<0.001）were all independent influencing factors predicting bone metastasis in NSCLC patients. ROC curve analysis showed that， the area under the curve（AUC）for predicting bone metastasis in the training set were 0.81， 0.79， and 0.74 for the models in the tumor interior group， the tumor interior combined with a 3 mm peritumoral group， and the simple 3 mm peritumoral group， respectively. The predictive value of the model in the tumor interior group was higher than that of the models in the tumor interior combined with a 3 mm peritumoral group and the simple 3 mm peritumoral group（Z=1.46， P=0.032； Z=3.01， P=0.024）. In the validation set， the AUC were 0.66， 0.63， and 0.53， respectively， with the predictive value of the model in the tumor interior group being higher than that of the models in the tumor interior combined with a 3 mm peritumoral group and the simple 3 mm peritumoral group（Z=2.37， P=0.025； Z=4.12， P=0.012）. A combined model was established using the radiomic features in the tumor interior group and the influencing factors with statistical significance from the multivariate analysis. The AUC of the combined model was 0.94 in the training set， which was higher than that of the model in the tumor interior group alone（Z=2.43， P=0.023）. In the validation set， the combined model's AUC was 0.92， which was also higher than that of the model in the tumor interior group（Z=3.76， P=0.007）. The calibration curve showed that the actual probabilities of both the training set and the validation set were in relatively good agreement with the predicted probabilities. DCA showed good discrimination ability for the combined model. Conclusions T stage， N stage， pathological type， D-dimer， CEA， spiculation sign， pleural indentation sign are all independent influencing factors predicting bone metastasis in NSCLC patients. Among the radiomics models， the model in the tumor interior group demonstrates higher predictive efficacy for NSCLC bone metastasis. The combined model constructed based on the above factors can further improve the predictive efficacy of bone metastasis in NSCLC patients，showing promising potential for clinical application.

Key words: Carcinoma, non-small-cell lung, Neoplasm metastasis, Radiomics, Chest-enhanced computed tomography, Prediction model

Li Jianing, Yao Xuemin, Wang Jinyun, Jia Jinghao, Sun Guogui. Predictive efficacy of a combined model of radiomics features and clinical signatures for bone metastasis in NSCLC[J]. Journal of International Oncology, 2026, 53(6): 346-354.

Figures/Tables 10

References 19

[1]	Bray F, Laversanne M, Sung H, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries[J]. CA Cancer J Clin, 2024, 74(3): 229-263. DOI: 10.3322/caac.21834.
[2]	Li Y, Yan B, He S. Advances and challenges in the treatment of lung cancer[J]. Biomed Pharmacother, 2023, 169: 115891. DOI: 10.1016/j.biopha.2023.115891.
[3]	朱志朋, 虞淦军, 许炜. 地舒单抗联合免疫检查点抑制剂在非小细胞肺癌骨转移中的临床研究进展[J]. 中国肿瘤生物治疗杂志, 2024, 31(5): 528-534. DOI: 10.3872/j.issn.1007-385x.2024.05.014.
[4]	Broder MS, Gutierrez B, Cherepanov D, et al. Burden of skeletal-related events in prostate cancer: unmet need in pain improvement[J]. Support Care Cancer, 2015, 23(1): 237-247. DOI: 10.1007/s00520-014-2437-3. pmid: 25270847
[5]	She K, Yang W, Li M, et al. FAIM2 promotes non-small cell lung cancer cell growth and bone metastasis by activating the Wnt/β-catenin pathway[J]. Front Oncol, 2021, 11: 690142. DOI: 10.3389/fonc.2021.690142.
[6]	Lubner MG, Smith AD, Sandrasegaran K, et al. CT texture analysis: definitions, applications, biologic correlates, and challenges[J]. Radiographics, 2017, 37(5): 1483-1503. DOI: 10.1148/rg.2017170056. pmid: 28898189
[7]	Koo HJ, Xu H, Choi CM, et al. Preoperative CT predicting recurrence of surgically resected adenocarcinoma of the lung[J]. Medicine (Baltimore), 2016, 95(2): e2513. DOI: 10.1097/MD.0000000000002513.
[8]	Vansteenkiste J, De Ruysscher D, Eberhardt WE, et al. Early and locally advanced non-small-cell lung cancer (NSCLC): ESMO clinical practice guidelines for diagnosis, treatment and follow-up[J]. Ann Oncol, 2013, 24 Suppl 6: vi89-vi98. DOI: 10.1093/annonc/mdx222.
[9]	钟建, 胡必富, 童随阳. 人附睾蛋白4等肿瘤标志物联合全身骨显像对非小细胞肺癌骨转移的诊断价值[J]. 国际呼吸杂志, 2023, 43(7): 789-795. DOI: 10.3760/cma.j.cn131368-20221112-01018.
[10]	Li W, Guo Z, Zou Z, et al. Development and validation of a prognostic nomogram for bone metastasis from lung cancer: a large population-based study[J]. Front Oncol, 2022, 12: 1005668. DOI: 10.3389/fonc.2022.1005668.
[11]	邱小雯, 宋轶鹏. 非小细胞肺癌脑转移危险因素及治疗进展研究[J]. 中国医刊, 2022, 57(10): 1059-1063. DOI: 10.3969/j.issn.1008-1070.2022.10.005.
[12]	Knapp BJ, Devarakonda S, Govindan R. Bone metastases in non-small cell lung cancer: a narrative review[J]. J Thorac Dis, 2022, 14(5): 1696-1712. DOI: 10.21037/jtd-21-1502. pmid: 35693589
[13]	许尧, 张超, 郭旭, 等. 肺癌不同病理类型与骨转移临床特征相关性研究[J]. 中华骨科杂志, 2019, 39(6): 329-335. DOI: 10.3760/cma.j.issn.0253-2352.2019.06.002.
[14]	Hu H, Chen WJ, Sun C, et al. D-D and PLR, NLR combined forecasting first-line treatment: the short-term curative effect and prognosis of patients with SCLC[J]. Medicine (Baltimore), 2023, 102(45): e34877. DOI: 10.1097/MD.0000000000034877.
[15]	秦晓娟, 袁媛. 血清CA125、CEA、NSE和SII联合检测对原发性肺癌患者骨转移的预测价值[J]. 国际呼吸杂志, 2024, 44(5): 563-567. DOI: 10.3760/cma.j.cn131368-20240329-00166.
[16]	马海洋, 吴琼, 张勇. 临床特征联合血清标志物检测预警模型在肺癌骨转移诊断中的临床价值[J]. 中国骨与关节杂志, 2023, 12(5): 351-356. DOI: 10.3969/j.issn.2095-252X.2023.05.006.
[17]	Numata T, Endo T, Yanai H, et al. Serum CEA and CYFRA levels in ALK-rearranged NSCLC patients: correlation with distant metastasis[J]. In Vivo, 2020, 34(4): 2095-2100. DOI: 10.21873/invivo.12013. pmid: 32606188
[18]	陈亮亮, 陈基明, 丁俊, 等. CT影像组学特征预测非小细胞肺癌同步脑转移的价值[J]. 放射学实践, 2022, 37(4): 419-425. DOI: 10.13609/j.cnki.1000-0313.2022.04.002.
[19]	Yilmaz A, Duyar SS, Cakir E, et al. Clinical impact of visceral pleural, lymphovascular and perineural invasion in completely resected non-small cell lung cancer[J]. Eur J Cardiothorac Surg, 2011, 40(3): 664-670. DOI: 10.1016/j.ejcts.2010.12.059. pmid: 21334917

项目	无骨转移（n=123）	骨转移（n=414）	χ²值	P值	项目	无骨转移（n=123）	骨转移（n=414）	χ²值	P值
年龄（岁）					D-二聚体（ng/ml）
≤60	47	155	0.02	0.877	≤230	62	340	50.70	<0.001
>60	76	259	0.02	0.877	>230	61	74	50.70	<0.001
性别					CEA（ng/ml）
男	66	247	1.41	0.236	≤5	31	316	108.41	<0.001
女	57	167	1.41	0.236	>5	92	98	108.41	<0.001
吸烟史					CYFRA21-1（ng/ml）
无	97	283	5.06	0.025	≤3.3	47	248	18.02	<0.001
有	26	131	5.06	0.025	>3.3	76	166	18.02	<0.001
家族史					SCC-Ag（ng/ml）
无	115	388	0.01	0.929	≤1.5	104	309	5.25	0.022
有	8	26	0.01	0.929	>1.5	19	105	5.25	0.022
肿瘤位置					毛刺征
左肺上叶	37	94		0.100	无	4	149	49.89	<0.001
左肺下叶	24	88			有	119	265	49.89	<0.001
右肺上叶	44	127			分叶征
右肺中叶	3	18			无	30	269	63.30	<0.001
右肺下叶	15	87			有	93	145	63.30	<0.001
T分期					胸膜凹陷征
T_1-2	103	374	4.16	0.041	无	41	306	68.30	<0.001
T_3-4	20	40	4.16	0.041	有	82	108	68.30	<0.001
N分期					血管集束征
N_0-1	27	289	89.67	<0.001	无	83	342	13.15	<0.001
N_2-3	96	125	89.67	<0.001	有	40	72	13.15	<0.001
病理类型
鳞状细胞癌	16	170	32.97	<0.001
腺癌	107	244	32.97	<0.001

项目	训练集（n=376）	验证集（n=161）	χ²值	P值	项目	训练集（n=376）	验证集（n=161）	χ²值	P值
年龄（岁）					D-二聚体（ng/ml）
≤60	144	58	0.25	0.618	≤230	282	120	0.01	0.909
>60	232	103	0.25	0.618	>230	94	41	0.01	0.909
性别					CEA（ng/ml）
男	209	104	3.77	0.052	≤5	235	112	2.46	0.117
女	167	57	3.77	0.052	>5	141	49	2.46	0.117
吸烟史					CYFRA21-1（ng/ml）
无	269	111	0.37	0.544	≤3.3	215	80	2.56	0.110
有	107	50	0.37	0.544	>3.3	161	81	2.56	0.110
家族史					SCC-Ag（ng/ml）
无	353	150	0.10	0.755	≤1.5	301	112	6.98	0.008
有	23	11	0.10	0.755	>1.5	75	49	6.98	0.008
肿瘤位置					毛刺征
左肺上叶	95	36			无	108	45	0.03	0.856
左肺下叶	75	37			有	268	116	0.03	0.856
右肺上叶	129	42	6.86	0.143	分叶征
右肺中叶	12	9			无	193	106	9.62	0.002
右肺下叶	65	37			有	183	55	9.62	0.002
T分期					胸膜凹陷征
T_1-2	342	135	5.74	0.017	无	237	110	1.38	0.240
T_3-4	34	26	5.74	0.017	有	139	51	1.38	0.240
N分期					血管集束征
N_0-1	219	97	0.19	0.666	无	282	143	13.04	<0.001
N_2-3	157	64	0.19	0.666	有	94	18	13.04	<0.001
病理类型
鳞状细胞癌	108	78	19.37	<0.001
腺癌	268	83	19.37	<0.001

因素	单因素分析			多因素分析
因素	OR值	95%CI	P值	OR值	95%CI	P值
年龄	1.00	0.66~1.51	0.993
性别	0.78	0.52~1.17	0.236
吸烟史	1.73	1.07~2.79	0.026	1.31	0.69~2.49	0.414
家族史	0.96	0.42~2.19	0.929
肿瘤位置
左肺上叶	Ref
左肺下叶	1.44	0.80~2.60	0.223
右肺上叶	1.14	0.68~1.90	0.625
右肺中叶	2.36	0.66~8.49	0.188
右肺下叶	2.28	1.17~4.45	0.015	-	-	-
T分期	0.57	0.32~1.01	0.043	0.69	0.52~0.87	<0.001
N分期	0.12	0.08~0.20	<0.001	0.24	0.13~0.43	<0.001
病理类型	4.66	2.66~8.16	<0.001	6.01	2.83~12.77	<0.001
D-二聚体（ng/ml）	0.22	0.14~0.34	<0.001	0.32	0.17~0.59	<0.001
CEA（ng/ml）	0.10	0.07~0.17	<0.001	0.25	0.14~0.44	<0.001
CYFRA21-1（ng/ml）	0.41	0.27~0.63	<0.001	0.65	0.35~1.18	0.156
SCC-Ag（ng/ml）	1.86	1.09~3.18	0.023	1.91	0.87~4.19	0.106
毛刺征	0.06	0.02~0.17	<0.001	0.21	0.07~0.65	0.007
分叶征	0.17	0.11~0.26	<0.001	-	-	-
胸膜凹陷征	0.18	0.11~0.27	<0.001	0.32	0.18~0.56	<0.001
血管集束征	0.44	0.28~0.69	<0.001	-	-	-

模型	AUC	95%CI	敏感性（%）	特异性（%）	准确性（%）
训练集（n=376）
肿瘤内部组模型	0.81	0.76~0.86	75.61	71.42	63.32
肿瘤内部联合瘤周3 mm组模型	0.79	0.74~0.84	77.93	71.70	76.69
瘤周3 mm组模型	0.74	0.69~0.80	87.24	52.18	71.58
联合模型	0.94	0.89~0.95	94.65	83.14	83.64
验证集（n=161）
肿瘤内部组模型	0.66	0.57~0.75	67.68	62.96	65.48
肿瘤内部联合瘤周3 mm组模型	0.63	0.54~0.72	64.92	62.90	76.25
瘤周3 mm组模型	0.53	0.43~0.62	83.81	34.75	86.53
联合模型	0.92	0.89~0.98	86.54	91.12	82.28