自发性丘脑出血后丘脑痛发生的列线图预测模型构建和验证

张茜; 杨鸿阁; 冯宇洋; 江旺; 张相彤; 梁洪生

doi:10.16766/j.cnki.issn.1674-4152.004158

自发性丘脑出血后丘脑痛发生的列线图预测模型构建和验证

doi: 10.16766/j.cnki.issn.1674-4152.004158

哈尔滨医科大学附属第一医院神经外科，黑龙江哈尔滨 150001

基金项目:

国家自然科学基金项目 82271340

哈尔滨医科大学附属第一医院杰出青年基金项目 HYD2020JQ0016

详细信息

通讯作者:
梁洪生，E-mail：lianghongsheng@hrbmu.edu.cn

中图分类号: R743.34
计量
- 文章访问数: 4
- HTML全文浏览量: 2
- PDF下载量: 0
- 被引次数: 0
出版历程
- 收稿日期: 2024-12-24
- 网络出版日期: 2025-11-17

Development and validation of a nomogram for predicting thalamic pain occurrence after spontaneous thalamic hemorrhage

Department of Neurosurgery, the First Affiliated Hospital of Harbin Medical University, Harbin, Heilongjiang 150001, China

摘要

摘要: 目的丘脑痛的发生严重影响患者的生活质量，早期识别可能发生丘脑痛的患者有助于临床上及时采取适当干预和决策以减轻患者痛苦。本研究旨在开发并验证一种列线图模型，用于预测患者自发性丘脑出血后丘脑痛的发生。方法回顾性选取哈尔滨医科大学附属第一医院2021年6月—2022年12月收治的294例自发性丘脑出血患者。按照7∶3的比例，通过计算机生成的随机数将患者随机分配至训练队列(206例)和内部验证队列(88例)，分别用于模型构建和内部验证。使用最小绝对收缩和选择运算符(LASSO)算法，选择最适合用于模型构建的特征变量。将筛选出的变量纳入多变量logistic回归分析构建列线图模型。在训练队列和内部验证队列中使用ROC曲线和决策曲线评估模型的性能。结果共纳入294例患者，其中94例患者发生丘脑痛(31.97%)，200例患者未发生丘脑痛(68.03%)。训练队列中的LASSO回归及logistic回归分析显示，血肿位置(P < 0.001)和改良格雷布(mGS)评分(P < 0.001)是独立预测因素，并纳入列线图预测模型。ROC曲线分析显示，该列线图模型预测性能较高(训练队列AUC：0.890；内部验证队列AUC：0.820)。决策曲线显示该模型对自发性丘脑出血后丘脑痛的发生具有较好的预测能力。结论本研究构建了评估患者自发性丘脑出血后丘脑痛发生风险的列线图模型，且模型预测性能较高，可指导高危患者的治疗决策。
- 自发性丘脑出血 /
- 丘脑痛 /
- 列线图 /
- 预测模型
Abstract: Objective The occurrence of thalamic pain significantly impacts patients' quality of life. Early identification of patients at risk for developing thalamic pain can facilitate timely clinical interventions and decision-making to alleviate patient suffering. This study aims to develop and validate a nomogram model to predict the occurrence of thalamic pain following spontaneous thalamic hemorrhage. Methods A total of 294 patients with spontaneous thalamic hemorrhage were retrospectively included from the First Affiliated Hospital of Harbin Medical University between June 2021 and December 2022. Patients were randomly assigned to the training cohort (n=206) and internal validation cohort (n=88) in a 7∶3 ratio using computer-generated random numbers, with the training cohort used for model development and the internal validation cohort for validation. Feature variables most suitable for model construction were selected using the least absolute shrinkage and selection operator (LASSO) algorithm. Subsequently, the selected features were incorporated into a multivariable logistic regression model to develop the nomogram. The performance of the model was assessed in the training and internal validation cohorts using the area under the receiver operating characteristic curve, and decision curve analysis. Results A total of 294 patients were included in this study, of whom 94 cases (31.97%) developed thalamic pain, while 200 cases (68.03%) did not. The LASSO-logistic regression analysis in the training cohort identified two independent predictive factors: hematoma location (P < 0.001) and modified Graeb score (mGS) score (P < 0.001). These predictors were incorporated into the nomogram prediction model. ROC curve analysis demonstrated high predictive performance of the nomogram model (training cohort AUC: 0.890; internal validation cohort AUC: 0.820). Decision curve analyses further indicated that the model has strong predictive ability for the occurrence of thalamic pain following spontaneous thalamic hemorrhage. Conclusion This study develops a nomogram model to assess the risk of thalamic pain in individual patients following spontaneous thalamic hemorrhage, demonstrating excellent predictive performance. The model provides clinicians with a practical tool to guide treatment decisions for high-risk patients.
- Spontaneous thalamic hemorrhage /
- Thalamic pain /
- Nomogram /
- Prediction model

HTML全文

图 1 LASSO回归交叉验证图

Figure 1. Cross-validation graph for LASSO regression

下载: 全尺寸图片幻灯片

图 2 LASSO回归系数路径图

Figure 2. Coefficient path diagram of LASSO regression

下载: 全尺寸图片幻灯片

图 3 自发性丘脑出血后丘脑痛发生的列线图预测模型

Figure 3. Nomogram prediction model for the occurrence of thalamic pain after spontaneous thalamic hemorrhage

下载: 全尺寸图片幻灯片

图 4 列线图预测模型的ROC曲线

Figure 4. ROC curves of the nomogram prediction model

下载: 全尺寸图片幻灯片

图 5 列线图预测模型的训练队列及内部验证队列决策曲线

Figure 5. Decision curves of the training queue and internal validation queue of the nomogram prediction model

下载: 全尺寸图片幻灯片

表 1 2组STH患者临床资料比较

Table 1. Comparison of clinical data of two groups of STH patients

项目	非丘脑痛组(n=200)	丘脑痛组(n=94)	统计量	P值
梗阻性脑积水[例(%)]			8.209^a	0.004
否	168(84.0)	90(95.7)
是	32(16.0)	4(4.3)
IVH[例(%)]			19.368^a	< 0.001
否	40(20.0)	42(44.7)
是	160(80.0)	52(55.3)
血肿位置[例(%)]			43.095^a	< 0.001
1	77(38.5)	72(76.6)
2	27(13.5)	12(12.8)
3	96(48.0)	10(10.6)
GCS评分[M(P₂₅, P₇₅), 分]	13.0(9.0, 15.0)	14.0(12.0, 15.0)	-2.091^b	0.032
初始血肿体积[M(P₂₅, P₇₅), mL]	7(4, 12)	6(3, 9)	2.757^b	0.006
中线移位[M(P₂₅, P₇₅), mm]	0.00(0.00, 5.69)	0.00(0.00, 0.00)	3.471^b	< 0.001
mGS评分[M(P₂₅, P₇₅), 分]	7.0(2.0, 11.0)	2.0(0.0, 4.0)	6.610^b	< 0.001
WBC[M(P₂₅, P₇₅)]	9.3(7.5, 12.0)	8.6(7.4, 10.2)	-1.870^b	0.049
NLR[M(P₂₅, P₇₅)]	6.8(4.2, 11.0)	5.1(3.8, 7.9)	-1.645^b	0.025
PNR[M(P₂₅, P₇₅)]	27(20, 37)	30(23, 43)	-5.078^b	0.042
SII[M(P₂₅, P₇₅)]	1 415(766, 2 433)	1 110(775, 1 817)	-1.861^b	0.048
注：^a为χ²值，^b为Z值。本表仅列出差异有统计学意义的项目。

下载: 导出CSV

表 2 变量赋值情况

Table 2. Variable assignment situation

变量	赋值方法
血肿位置	丘脑后侧和丘脑外侧=(0，0)；全丘脑=(0，1)；丘脑的其他位置=(1，0)
mGS评分	连续性变量，以实际值赋值
丘脑痛发生	是=1，否=0

下载: 导出CSV

表 3 基于训练队列的多变量logistic回归分析

Table 3. Multivariate logistic regression analysis based on the training cohort

变量	B	SE	Waldχ²	P值	OR值	95% CI
血肿位置
1	3.459	0.602	5.744	< 0.001	31.792	10.809~119.809
2	4.378	0.842	5.201	< 0.001	79.647	16.637~465.215
mGS评分	-0.371	0.066	-5.598	< 0.001	0.690	0.599~0.778

下载: 导出CSV

参考文献(21)

[1]	许莹宁, 穆琼. PPARγ激动剂对雄性大鼠脑出血后血肿周围小胶质细胞极化的影响[J]. 中华全科医学, 2024, 22(5): 781-785. doi: 10.16766/j.cnki.issn.1674-4152.003502 XU Y N, MU Q. Effect of PPARγ agonist on microglia polarization around hematoma after cerebra hemorrhage in male rats[J]. Chinese Journal of General Practice, 2024, 22(5): 781-785. doi: 10.16766/j.cnki.issn.1674-4152.003502
[2]	LI L, MURTHY S B. Cardiovascular events after intracerebral hemorrhage[J]. Stroke, 2022, 53(7): 2131-2141. doi: 10.1161/STROKEAHA.122.036884
[3]	BARBOSA L M, DA SILVA V A, DE LIMA RODRIGUES A L, et al. Dissecting central post-stroke pain: a controlled symptom-psychophysical characterization[J]. Brain Commun, 2022, 4(3): fcac090. DOI: 10.1093/braincomms/fcac090.
[4]	KLIT H, FINNERUP N B, JENSEN T S. Central post-stroke pain: clinical characteristics, pathophysiology, and management[J]. Lancet Neurol, 2009, 8(9): 857-868. doi: 10.1016/S1474-4422(09)70176-0
[5]	LIAMPAS A, VELIDAKIS N, GEORGIOU T, et al. Prevalence and management challenges in central post-stroke neuropathic pain: a systematic review and meta-analysis[J]. Adv Ther, 2020, 37(7): 3278-3291. doi: 10.1007/s12325-020-01388-w
[6]	RI S. The management of poststroke thalamic pain: update in clinical practice[J]. Diagnostics(Basel), 2022, 12(6): 1439. DOI: 10.3390/diagnostics12061439.
[7]	MOHANAN A T, NITHYA S, NOMIER Y, et al. Stroke-induced central pain: overview of the mechanisms, management, and emerging targets of central post-stroke pain[J]. Pharmaceuticals(Basel), 2023, 16(8): 1103. DOI: 10.3390/ph16081103.
[8]	FANG H, LI M, YANG J, et al. Repressing iron overload ameliorates central post-stroke pain via the Hdac2-Kv1.2 axis in a rat model of hemorrhagic stroke[J]. Neural Regen Res, 2024, 19(12): 2708-2722. doi: 10.4103/NRR.NRR-D-23-01498
[9]	FINNERUP N B, KUNER R, JENSEN T S. Neuropathic pain: from mechanisms to treatment[J]. Physiol Rev, 2021, 101(1): 259-301. doi: 10.1152/physrev.00045.2019
[10]	DAI J, LIU D, LI X, et al. Prediction of hematoma expansion in hypertensive intracerebral hemorrhage by a radiomics nomogram[J]. Pak J Med Sci, 2023, 39(4): 1149-1155.
[11]	YOO J, HYUN S H, LEE J, et al. Prognostic significance of 18 F-FDG PET/CT radiomics in patients with resectable pancreatic ductal adenocarcinoma undergoing curative surgery[J]. Clin Nucl Med, 2024, 49(10): 909-916. doi: 10.1097/RLU.0000000000005363
[12]	YOU C, SU G H, ZHANG X, et al. Multicenter radio-multiomic analysis for predicting breast cancer outcome and unravelling imaging-biological connection[J]. NPJ Precis Oncol, 2024, 8(1): 193. DOI: 10.1038/s41698-024-00666-y.
[13]	舒存洪, 张姚萍, 张元斌. LASSO回归筛选骨质疏松性椎体压缩骨折患者术后残余疼痛的预测因子及模型构建[J]. 中华全科医学, 2024, 22(11): 1880-1883, 1980. doi: 10.16766/j.cnki.issn.1674-4152.003755 SHU C H, ZHANG Y P, ZHANG Y B. LASSO regression to identify predictors and model construction for postoperative residual pain in patients with osteoporotic vertebral compression fracture[J]. Chinese Journal of General Practice, 2024, 22(11): 1880-1883, 1980. doi: 10.16766/j.cnki.issn.1674-4152.003755
[14]	CHENG Y, WU B, HUANG J, et al. Research progress on the mechanisms of central post-stroke pain: a review[J]. Cell Mol Neurobiol, 2023, 43(7): 3083-3098. doi: 10.1007/s10571-023-01360-6
[15]	HIRATO M, MIYAGISHIMA T, TAKAHASHI A, et al. Thalamic anterior part of the ventral posterolateral nucleus and central lateral nucleus in the genesis of central post-stroke pain[J]. Acta Neurochir, 2021, 163(8): 2121-2133. doi: 10.1007/s00701-021-04743-0
[16]	ROSNER J, DE ANDRADE D C, DAVIS K D, et al. Central neuropathic pain[J]. Nat Rev Dis Primers, 2023, 9(1): 73. DOI: 10.1038/s41572-023-00484-9.
[17]	LIU P F, WANG Y, XU L, et al. Modulation of itch and pain signals processing in ventrobasal thalamus by thalamic reticular nucleus[J]. iScience, 2022, 25(1): 103625. DOI: 10.1016/j.isci.2021.103625.
[18]	ZHU X, TANG H D, DONG W Y, et al. Distinct thalamocortical circuits underlie allodynia induced by tissue injury and by depression-like states[J]. Nat Neurosci, 2021, 24(4): 542-553. doi: 10.1038/s41593-021-00811-x
[19]	BISSON D A, FLAHERTY M L, SHATIL A S, et al. Original and modified graeb score correlation with intraventricular hemorrhage and clinical outcome prediction in hyperacute intracranial hemorrhage[J]. Stroke, 2020, 51(6): 1696-1702. doi: 10.1161/STROKEAHA.120.029040
[20]	傅佳, 冯富媛, 欧册华, 等. 中枢卒中后疼痛的神经机制与临床管理策略[J]. 西南医科大学学报, 2024, 47(6): 553-558. FU J, FENG F Y, OU C H, et al. Neurological mechanisms and clinical management strategies of central post-stroke Pain[J]. Journal of Southwest Medical University, 2024, 47(6): 553-558.
[21]	YUAN X Q, HU S Y, FAN X C, et al. Central post-stroke pain: advances in clinical and preclinical research[J]. Stroke Vasc Neurol, 2024: svn-2024-003418. DOI: 10.1136/svn-2024-003418.