PD-1/PD-L1与心肌炎的关系：一项孟德尔随机化研究

doi:10.12677/acm.2024.1482389

期刊菜单

PD-1/PD-L1与心肌炎的关系：一项孟德尔随机化研究
Association between PD-1/PD-L1 and Myocarditis: A Mendelian Randomization Study

DOI: 10.12677/acm.2024.1482389, PDF, HTML, XML, 科研立项经费支持
作者: 曾亮嘉：广州医科大学南山学院，广东广州；周若筠：广州医科大学第三临床学院，广东广州；杨文婷, 冯曼婷, 梁英岚^*：广州医科大学第二临床学院，广东广州
关键词: 孟德尔随机化；心肌炎；PD-1；PD-L1；因果关系；Mendelian Randomization； Myocarditis； PD-1； PD-L1； Causal Relationship

摘要: 目的：利用两样本，多变量及双向孟德尔随机化(MR)研究设计，探究程序细胞死亡蛋白1 (programmeddeath-1, PD-1)及其配体(PD-1 ligand, PD-L1)与心肌炎风险的因果关联。方法：从公开全基因组关联研究(GWAS)中提取PD-1/PD-L1与心肌炎的工具变量。采用逆方差加权法(IVW)作为主要的MR分析方法，辅以加权中位数(WM)、Robust adjusted profile score (RAPS)、MR-Egger、孟德尔随机化多效性残差和异常值(MR-PRESSO)、Cochran’s Q检验、留一法、水平多效性分析作为敏感性分析。结果：两样本MR的IVW结果提示，PD-L1与心肌炎具有负向因果关联[比值比(OR)，0.619；95%置信区间(CI)，0.434~0.884；P，0.008]，敏感性分析结果表明因果关系稳健。在多变量MR中，PD-L1与心肌炎的负向因果关系依然存在(OR, 0.658; 95% CI, 0.47~0.92, P, 0.015)。PD-1与心肌炎，心肌炎与PD-1/PD-L1的因果关联没有统计学意义。结论：该研究为PD-L1与心肌炎之间的负向因果关联提供了新的证据，而PD-1与心肌炎，心肌炎与PD-1/PD-L1没有直接因果关系。

Abstract: Objective: To investigate the causal association between programmed death-1 (PD-1) and its ligand (PD-1 ligand, PD-L1) and the risk of myocarditis using a two-sample, multivariate and bidirectional Mendelian randomization (MR) study design. Methods: Instrumental variables for PD-1/PD-L1 and myocarditis were extracted from public genome-wide association studies (GWAS). Inverse variance weighting (IVW) was used as the main MR analysis method, supplemented by weighted median (WM), Robusta adjusted profile score (RAPS), MR-Egger, Mendelian randomization pleiotropic residuals and outliers (MR-PRESSO), Cochran’s Q test, leave-one-out method, and horizontal pleiotropy analysis as sensitivity analyses. Results: The IVW results of the two-sample MR suggested that PD-L1 had a negative causal association with myocarditis [odds ratio (OR), 0.619; 95% confidence interval (CI), 0.434~0.884; P, 0.008], and the sensitivity analysis results showed that the causal relationship was robust. In multivariate MR, the negative causal relationship between PD-L1 and myocarditis still existed (OR, 0.658; 95% CI, 0.47-0.92; P, 0.015). The causal association between PD-1 and myocarditis, and between myocarditis and PD-1/PD-L1 was not statistically significant. Conclusion: This study provides new evidence for the negative causal association between PD-L1 and myocarditis, while there is no direct causal relationship between PD-1 and myocarditis, and between myocarditis and PD-1/PD-L1.

文章引用：曾亮嘉, 周若筠, 杨文婷, 冯曼婷, 梁英岚. PD-1/PD-L1与心肌炎的关系：一项孟德尔随机化研究[J]. 临床医学进展, 2024, 14(8): 1543-1558. https://doi.org/10.12677/acm.2024.1482389

1. 引言

心肌炎是一种非缺血性炎性心脏疾病，可导致猝死和扩张型心肌病等严重后果[1]。根据全球疾病负担数据库统计，1990~2019年全球及中国心肌炎的发病例数和死亡例数均呈逐年上升趋势。2019年，全球心肌炎的患病人数达712,780例，中国心肌炎发病例数已达27.51万，死亡例数为3.24万[2]。心肌炎已成为亟待解决的公共卫生问题之一。

研究发现，程序细胞死亡蛋白1 (programmed death-1, PD-1)及配体 (PD-1 ligand, PD-L1)似乎与心肌炎有着密切联系。有研究报道PD-1或PD-L1缺陷的小鼠会出现致死性心肌炎[3]-[5]，PD-1或PD-L1抑制剂会导致自身免疫性心肌炎[6]-[9]。但上述证据来自动物实验或观察性研究，受到自身免疫背景、癌症等共病及其他混杂因素的影响，并不能可靠地反映PD-1或PD-L1与心肌炎间的因果关系。

传统上，支持因果关系的最有力证据来自随机对照试验，但它们往往规模庞大且成本高昂，且受到伦理的限制[10]。孟德尔随机化研究是近年广泛应用于流行病学中暴露因素与临床疾病之间潜在因果关系评估的方法[11]-[13]。在一个群体中，等位基因的随机分配(根据孟德尔定律)导致一些个体在遗传上受更高或更低水平的暴露因素影响，类似于随机对照试验中对治疗组和对照组的分配[10]。因此，与传统的观察性分析相比，孟德尔随机化分析可以排除混杂因素和反向因果关系的干扰，为研究PD-1或PD-L1和心肌炎之间的潜在因果关系提供更可靠的见解[14] [15]。因此，我们进行了一项两样本孟德尔随机化研究，以探讨PD-1或PD-L1与心肌炎潜在因果关系，为心肌炎的免疫学防治提供新的理论依据。

2. 材料和方法

2.1. 研究设计

MR使用遗传工具变量(IV)来评估暴露和结果之间的因果关系。MR设计的基本原则是遗传变异在受精时已固定，并随机分配给个体。因此，它可以克服传统观察流行病学中常见的混杂和逆向因果关系的问题。在两样本的MR中，它结合来自多个来源的数据，并使用两个不同的研究样本来估计工具变量的风险因素和结果关联。

MR应满足3个基本假设：(1) 关联性假设：遗传变异与暴露强相关；(2) 独立性假设：遗传变异独立于任何潜在的混杂因素；(3) 排他性假设：遗传变异仅通过暴露途径影响结果[16]。

我们研究的设计有三个关键组成部分：(1) 识别用作工具变量的遗传变异；(2) 利用两样本，多变量和双向的MR策略估计因果效应；(3) 评估水平多效性并进行敏感性分析验证结果。

在本研究中，所有相关数据都可以在网站上公开下载，并且可以无限制地使用。

2.2. 数据来源

2.2.1. 暴露数据

PD-1和PD-L1的数据来源于一项全基因组关联研究(GWAS)研究，该研究对来自3301名欧洲血统个体中的2994种血浆蛋白水平的1060万个估算的常染色体变异进行了全基因组检测，这些个体的单核苷酸多态性(SNP)数量为10,534,735 (GWAS ID: prot-a-2214和prot-a-431) [17] (表S1)。

2.2.2. 结局数据

结局数据来自FinnGen数据研究，这是一个大规模的基因组学计划，已分析超过500,000个芬兰生物库样本，并将遗传变异与健康数据相关联，以了解疾病机制和易感性[18]。关于遗传预测的心肌炎风险(GWAS ID: finn-b-I9_MYOCARD)的数据来自117,755名欧洲人(829名心肌炎患者和116,926名健康对照者)，包括16,379,455个SNP (表S1)。

2.3. 工具变量筛选

我们基于以下三个标准获得遗传工具变量(IV)：(1) 对于标准的MR和逆向的MR，与暴露相关的SNP均基于全基因组显著性水平(P < 5 × 10⁻⁶)选择。(2) 基于kb = 10,000、r² < 0.001去除连锁不平衡。(3) 删除重复的SNP和回文的SNP。

我们还计算了F统计量，以衡量MR中的工具变量强度。对于单一遗传变异，F统计量等于与暴露相关的遗传关联的平方除以其标准误的平方，计算公式为F = βi²/se (βi)²。F > 10被认为工具变量强度显著。

2.4. 数据分析

2.4.1. 单变量MR

逆方差加权法(IVW)是我们研究的主要分析方法，这种方法基于所有SNPs都是有效的遗传工具的假设，将回归截距限制为0，以结局方差的倒数作为权重进行拟合，提供一个稳健的结果[19]。对于敏感性分析，我们使用了多种方法，包括MR-Egger回归、Robust adjusted profile score (RAPS)、加权中位数法(WM)、MR-PRESSO法、Cochran’s Q评估、留一法以及水平多效性分析。

MR-Egger通过变异间加权回归的斜率估计因果效应，并将平均多效性效应作为截距。我们计算了MR-Egger的截距，以提供水平多效性的测量，确认变异是否对目标结果有直接影响[19] [20]。RAPS是最近提出的方法，它对系统性和特异性的多效性都很稳健，并且可以为具有许多弱工具变量的MR分析提供稳健的推断[21]。加权中位数法基于遗传变异代表分析中超过50%权重的假设，提供一致的因果效应估计。该方法旨在确保所有工具变量估计的中位数将是一致的估计[22]。MR-PRESSO方法用于测试并在需要时纠正分析中可能存在的水平多效性异常值，通过移除对异质性贡献超出预期的SNP [20]。Cochran’s Q统计量是用于异质性的统计检验，它源自IVW估计，其遵循自由度等于SNP数量减1的χ²分布[20]。留一法通过排除可能在存在一个与暴露特别强相关的遗传变异，为我们提供因果效应的可靠性检验。

2.4.2. 多变量MR

多变量MR适用于使用大量遗传工具的情况，无论它们与暴露是否相关。虽然工具变量可能与多个风险因素相关，但它们必须满足等效工具变量假设[23]。因此，我们采用了这种方法，整合了所有与PD-1和PD-L1有关的工具变量，以确定它们对心肌炎的不同影响。在此分析中，我们使用IVW作为我们的主要分析方法，并使用MR-Egger作为敏感性分析。

2.4.3. 反向MR

我们另外进行了反向MR分析以评估结局对暴露的影响，旨在探究心肌炎是否对PD-1/PD-L1有负面影响，或者两者之间的相关性是否是由于潜在混杂所致。我们反向MR的主要分析是IVW并包括多种敏感性分析(MR-Egger、RAPS、加权中位数方法、MR-PRESSO、Cochran’s Q检验、留一法、水平多效性分析)。

3. 结果

3.1. PD-1/PD-L1与心肌炎的单变量MR分析

Table 1. Univariate MR results of the effect of PD-1/PD-L1 on the risk of myocarditis

表1. PD-1/PD-L1对心肌炎风险效应的单变量MR结果

Exposure	Outcome	SNPS	Methods	Odds ratio	95% CI		P-value	Q-statistics	Ph	Pp
PD-1	Myocarditis	13	Inverse variance weighted	1.151	0.919	1.442	0.221	7.909	0.792
			MR Egger	1.242	0.773	1.998	0.390	7.781	0.733	0.727
			RAPS	1.153	0.909	1.462	0.240
			Weighted median	1.145	0.852	1.538	0.369
			MR Presso	NA	NA	NA	NA
PD-L1	Myocarditis	5	Inverse variance weighted	0.619	0.434	0.884	0.008	1.459	0.834
			MR Egger	0.496	0.090	2.726	0.479	1.390	0.708	0.810
			RAPS	0.616	0.422	0.899	0.012
			Weighted median	0.617	0.394	0.965	0.034
			MR Presso	NA	NA	NA	NA

Abbreviations: RAPS, Robust adjusted profile score; MR-PRESSO, Mendelian Randomization Pleiotropy RESidual Sum and Outlier; CI, confidence interval; Ph, P-value for heterogeneity; Pp, P-value for Pleiotropy; NA, not applicable.

Table 2. Multivariate MR results of the effect of PD-1/PD-L1 on the risk of myocarditis

表2. PD-1/PD-L1对心肌炎风险效应的多变量MR结果

Exposure	Outcome	SNPS	Methods	Odds ratio	95% CI		P-value
PD-1	Myocarditis	18	Inverse variance weighted	1.157	0.927	1.44	0.197
PD-1	Myocarditis	18	MR-Egger	1.195	0.817	1.75	0.358
PD-L1	Myocarditis	18	Inverse variance weighted	0.658	0.47	0.92	0.015
PD-L1	Myocarditis	18	MR-Egger	0.656	0.469	0.919	0.014

Figure 1. Scatter plot of PD-L1 effect on myocarditis

图1. PD-L1对心肌炎效应的散点图

经上述标准筛选后，PD-1的SNP数量为13个，PD-L1为5个(表1)。单变量MR分析中比值比(OR)的统计学效力见表S2，SNP具体信息见表S3，表S4。MR分析结果见表1，IVW分析发现PD-L1与心肌炎具有负向因果关联[OR，0.619；95%置信区间(CI)，0.434~0.884；P，0.008]，这一结果与RAPS (OR, 0.616; 95% CI, 0.422~0.899; P, 0.012)和WM (OR, 0.617; 95% CI, 0.394~0.965; P, 0.034)一致。MR-Egger分析也显示PD-L1对心肌炎的风险具有负向但不显著的效应(OR, 0.496; 95% CI, 0.09~2.726; P, 0.479)。PD-L1对心肌炎效应的散点图和留一图见图1和图2。MR-PRESSO测试未发现PD-L1与心肌炎风险之间的异常值。Cochran’s Q检验未发现异质性(Q, 1.459; P, 0.834)。在MR-Egger回归中未发现水平多效性(P, 0.81)。此外，PD-1与心肌炎之间的关联在统计学上没有显著性。PD-1对心肌炎效应的散点图和留一图在见图S1和图S2。

Figure 2. Leave a picture of the effect of PD-L1 on myocarditis

图2. PD-L1对心肌炎效应的留一图

在考虑了PD-1和PD-L1效应的多变量MR中，PD-L1与心肌炎之间的显著负向因果关系仍然存在(OR, 0.658; 95% CI, 0.47~0.92; P, 0.015) (表2)。MR-Egger分析也提供了一致的结果(OR, 0.657; 95% CI, 0.469~0.919; P, 0.014)。PD-1与心肌炎之间的因果关联依然不显著(表2)。用于多变量MR分析的SNP具体信息见表S5。

3.3. 心肌炎与PD-1/PD-L1的单变量MR分析

Table 3. Single-Variable MR Results of Risk of Myocarditis on PD-1/PD-L1

表3. 心肌炎对PD-1/PD-L1风险效应的单变量MR结果

Exposure	Outcome	SNPS	Methods	Beta	95% CI		P-value	Q-statistics	Ph	Pp
Myocarditis	PD-1	8	Inverse variance weighted	−0.030	−0.096	0.036	0.379	6.863	0.443
			MR Egger	0.023	−0.195	0.242	0.841	6.587	0.361	0.635
			RAPS	−0.037	−0.108	0.035	0.314
			Weighted median	0.005	−0.084	0.094	0.914
			MR Presso	NA	NA	NA	NA
Myocarditis	PD-L1	8	Inverse variance weighted	0.013	−0.085	0.112	0.793	15.568	0.029
			MR Egger	−0.034	−0.368	0.300	0.849	15.350	0.018	0.780
			RAPS	0.014	−0.086	0.113	0.787
			Weighted median	0.018	−0.089	0.125	0.740
			MR Presso	−0.021	−0.110	0.067	0.653

IVW分析结果显示心肌炎与PD-1或PD-L1的因果关联不显著，IVW分析结果与敏感性分析结果一致(表3)。用于反向MR分析的SNP具体信息见表S6，表S7。PD-1/PD-L1对心肌炎效应的散点图和留一图见图S2(a)，图S2(b)，图S3(a)，图S3(b)。

4. 讨论

该研究采用两样本MR，多变量MR和反向MR的方法探究PD-1/PD-L1与心肌炎之间的因果关系，发现PD-L1对心肌炎具有负向的显著的因果关系，而PD-1对心肌炎，心肌炎对PD-1/PD-L1的因果关系不显著。

免疫检查点抑制剂(ICI)因有益于部分癌症患者预后，已经取得了极大的进展。但与此同时，ICI所免疫相关不良事件(irAE)也受到了关注。心肌炎因显著的发病率成为最常报告的心脏irAE，并且死亡率高于其他irAE [24] [25]。ICI相关心肌炎的表现与诊断具有异质性，这对目前的研究造成了一定的局限[8]。此前研究多研究抗PD-1/PD-L1治疗对于心肌炎的影响[26] [27]，由于治疗方案的限制无法分别对PD-1和PD-L1所造成的心肌炎风险进行研究。MR基于孟德尔遗传定律，通过基因变异作为工具变量，即使存在无法观测的混杂因素也可以进行因果效应推断[28]。

免疫检查点多方面地参与维护心血管稳态，有关心脏irAE机制尚不完全清楚[29] [30]。通常认为，PD-1保护心肌细胞免受部分炎症损伤[5]，而PD-1抑制剂可通过极化巨噬细胞诱导心脏损伤[31]。巨噬细胞介导的免疫反应主要造成了心肌炎，而PD-1/CTLA4在功能上协同催化了这一过程[7] [32]。PD-L1与炎症关系复杂。PD-L1在心脏组织上低水平表达[33]，而炎症可能上调其表达[29]；动物实验证明了PD-L1的表达缺失会促进心脏炎症发生[34] [35]。PD-1/PD-L1不同的表达水平也可能是导致心肌炎的触发因素[36]。

此前一项基于多中心的回顾性研究报告心肌炎病例基本由抗PD-1治疗引起[37]，然而观察性研究并不能对因果关系进行推断。也有研究报道PD-1抑制剂与PD-L1抑制剂毒性相似，但这项系统评价研究并未针对心肌炎进行研究[38]。目前尚无大型随机对照试验评估PD-1/PD-L1表达与心肌炎之间的因果联系。我们的研究揭示了PD-L1对心肌炎具有保护性因果效应，这项结果在进行多变量分析后依然成立。至于PD-1与心肌炎并不具有显著的因果效应，其原因可能是由循环标志物调节，本身和心肌炎无直接联系[39]，也可能相关的遗传标记导致了广泛的心肌炎易感性，而不起决定性作用。同时，心肌炎与PD-1/PD-L1也不具备直接的因果联系，其中具体的机制有待进一步的探索。

该研究具有几个显著的优势。首先，我们深入探讨了PD-1/PD-L1与心肌炎之间的相互因果联系，利用双向MR，减轻了混杂因素、逆向因果关系和暴露偏见的影响[33]。此外，该研究进行了包括MR-Egger、加权中位数法和MR-PRESSO等在内的敏感性评估，以增强结果的一致性和稳健性。并且，我们运用了多变量MR来调整混杂变量，增强了PD-1/PD-L1与心肌炎之间因果联系推断的可靠性。尽管具有这些优势，但我们必须承认该研究存在不足。由于SNP有限，为了避免相关性被削弱，我们根据P < 5 × 10⁻⁶选择SNP [40] [41]，并通过F统计量(所有SNP的F值大于10)验证了不存在弱工具变量偏移。此外，我们的研究重点是欧洲人口，因此应谨慎将我们的结果推广到其他人口统计组。最后，由于本研究中使用的GWAS数据是汇总级别的数据，我们缺乏按性别和年龄分层或个体水平数据的访问权。因此，我们的研究无法探索PD-1/PD-L1与心肌炎在各种年龄和性别亚组之间的潜在因果关联。

综上所述，本研究采用两样本、多变量和反向MR方法，结果表明PD-L1与心肌炎呈负相关的因果关联，PD-1对心肌炎无因果关联，心肌炎对PD-1/PD-L1无因果关联，为心肌炎的免疫学防治提供了新的理论依据。

基金项目

2023年国家级大学生创新创业训练计划(项目编号：202310570047)。

附录

Table S2. The statistical power of the univariable MR for detecting odds ratios at 5% type I error

表S2. 单变量MR在5%类型I错误下检测比值比的统计功效

Exposure	Outcome	SNPs	R² EXPOSURE(SUM)	N cases Outcome	N controls Outcome	Ratio of cases to controls	Sample size Outcome	OR(IVW)	Power (%)
PD-1	Myocarditis	13	0.120056046	829	116,926	0.007089954	117,755	1.151178545	28.8
PD-L1	Myocarditis	5	0.039743955	829	116,926	0.007089954	117,755	0.619154841	78.3

Table S3. Independent instruments for PD-1 use in myocarditis: harmonized data

表S3. 心肌炎中PD-1使用的独立工具：协调数据

SNP	effect_ allele	other_ allele	chr	pos	beta. exposure	beta. outcome	eaf. exposure	eaf. outcome	se. exposure	se. outcome	pval. exposure	pval. outcome	F	R²_ exposure
rs10,444,703	T	C	14	38,719,281	0.1717	0.0027	0.14077	0.09778	0.0372	0.0848	3.80E−06	0.9744	21.30368395	0.007131652
rs111,808,226	G	A	5	135,215,385	−0.6272	−0.3135	0.01036	0.01108	0.136	0.2528	3.98E−06	0.215	21.2683737	0.008066388
rs117444695	G	A	6	138,046,690	0.5472	−0.0126	0.01306	0.007755	0.1115	0.2936	9.33E−07	0.9659	24.08476664	0.007718912
rs12,073,392	A	C	1	222,926,678	−0.1455	−0.074	0.81272	0.8017	0.0314	0.0628	3.72E−06	0.2384	21.47171285	0.006444487
rs147,652,769	T	C	2	8,057,238	0.5534	0.1043	0.01737	0.006941	0.1068	0.2994	2.19E−07	0.7275	26.84947537	0.010454377
rs34,777,990	G	A	19	3,563,017	−0.1367	−0.019	0.26609	0.2987	0.0299	0.0554	4.90E−06	0.732	20.90232771	0.00729858
rs5,030,928	G	A	10	71,176,637	0.1509	0.0232	0.26207	0.2479	0.0293	0.0582	2.69E−07	0.6904	26.52425771	0.008807263
rs55,720,497	G	A	16	25,686,871	−0.4143	−0.6042	0.02258	0.004345	0.0832	0.3805	6.31E−07	0.1123	24.79609057	0.007576437
rs5,757,973	G	T	22	22,712,467	0.3788	−0.0595	0.89028	0.9033	0.0399	0.0862	2.00E−21	0.4899	90.13099164	0.028032534
rs71,362,039	C	A	18	8,678,954	0.3356	0.1535	0.03838	0.02221	0.0663	0.1703	4.17E−07	0.3674	25.62222541	0.00831347
rs72,692,455	G	A	4	182,108,656	0.1246	−0.0422	0.31944	0.3001	0.027	0.0549	3.98E−06	0.4417	21.29651578	0.00675028
rs77,326,121	T	C	3	73,247,608	0.3256	0.1797	0.03447	0.0329	0.0682	0.1429	1.82E−06	0.2085	22.79292404	0.007056768
rs9,958,590	G	A	18	66,092,216	0.1239	0.0164	0.29656	0.3292	0.0271	0.053	4.90E−06	0.7576	20.9027791	0.006404898

Table S4. Independent instruments for PD-L1 use in myocarditis: harmonized data

表S4. 心肌炎中PD-L1使用的独立工具：协调数据

SNP	effect_ allele	other_ allele	chr	pos	beta. exposure	beta. outcome	eaf. exposure	eaf. outcome	se. exposure	se. outcome	pval. exposure	pval. outcome	F	R²_ exposure
rs11,011,804	A	G	10	20,394,779	0.3091	−0.317	0.0353	0.0214	0.0671	0.1711	4.07E−06	0.0639794	21.22037087	0.006507213
rs13,322,229	T	C	3	35,068,626	0.1742	−0.0577	0.1322	0.1835	0.0369	0.0643	2.40E−06	0.3698	22.2865872	0.006962695
rs140,094,912	A	G	20	12,182,666	−0.3337	0.0626	0.0354	0.02176	0.0719	0.1748	3.55E−06	0.7205	21.54044309	0.00760489
rs822,341	C	T	9	5,453,396	0.1731	−0.0895	0.7347	0.7264	0.0278	0.0559	4.47E−10	0.1095	38.7707805	0.011680769
rs9,859,911	G	A	3	107,733,600	0.1981	−0.0952	0.9012	0.8824	0.0417	0.0776	2.09E−06	0.22	22.5681958	0.006988389

Table S5. Multivariable instruments for exposures PD-1 and PD-L1 on myocarditis: harmonized data

表S5. 心肌炎PD-1和PD-L1暴露的多变量工具：协调数据

SNP	beta_PD-1	beta_PD-L1	se_PD-1	se_PD-L1	P-value_PD-1	P-value_PD-L1	beta_out	se_out	P-value_out
rs10,444,703	0.1717	−0.0063	0.0372	0.0373	3.80E−06	0.870964	0.0027	0.0848	0.9744
rs11,011,804	−0.0022	0.3091	0.0673	0.0671	0.977237	4.07E−06	−0.317	0.1711	0.0639794
rs111,808,226	−0.6272	0.089	0.136	0.1365	3.98E−06	0.512861	−0.3135	0.2528	0.215
rs117,444,695	0.5472	−0.0252	0.1115	0.1119	9.33E−07	0.812831	−0.0126	0.2936	0.9659
rs12,073,392	−0.1455	−0.0742	0.0314	0.0315	3.72E−06	0.0186209	−0.074	0.0628	0.2384
rs13,322,229	−0.0071	0.1742	0.0371	0.0369	0.851138	2.40E−06	−0.0577	0.0643	0.3698
rs140,094,912	−0.0957	−0.3337	0.0722	0.0719	0.186209	3.55E−06	0.0626	0.1748	0.7205
rs147,652,769	0.5534	−0.0267	0.1068	0.1072	2.19E−07	0.794328	0.1043	0.2994	0.7275
rs34,777,990	−0.1367	0.0689	0.0299	0.03	4.90E−06	0.0218776	−0.019	0.0554	0.732
rs5,030,928	0.1509	0.0335	0.0293	0.0295	2.69E−07	0.25704	0.0232	0.0582	0.6904
rs55,720,497	−0.4143	−0.0689	0.0832	0.0835	6.31E−07	0.40738	−0.6042	0.3805	0.1123
rs5,757,973	0.3788	0.0098	0.0399	0.0404	2.00E−21	0.812831	−0.0595	0.0862	0.4899
rs71,362,039	0.3356	−0.0238	0.0663	0.0666	4.17E−07	0.724436	0.1535	0.1703	0.3674
rs72,692,455	0.1246	0.0173	0.027	0.0271	3.98E−06	0.524807	−0.0422	0.0549	0.4417
rs77,326,121	0.3256	−0.016	0.0682	0.0685	1.82E−06	0.812831	0.1797	0.1429	0.2085
rs822,341	0.0514	0.1731	0.0279	0.0278	0.0660693	4.47E−10	−0.0895	0.0559	0.1095
rs9,859,911	−0.1	0.1981	0.0419	0.0417	0.0169824	2.09E−06	−0.0952	0.0776	0.22
rs9,958,590	0.1239	0.0184	0.0271	0.0272	4.90E−06	0.501187	0.0164	0.053	0.7576

Table S6. Independent instruments for myocarditis use in PD-1: harmonized data

表S6. PD-1中心肌炎使用的独立工具：协调数据

SNP	effect_ allele	other_ allele	chr	pos	beta. exposure	beta. outcome	eaf. exposure	eaf. outcome	se. exposure	se. outcome	pval. exposure	pval. outcome	F	R2_ exposure
rs1,0254,255	A	G	7	96,468,009	−0.3395	−0.0034	0.8513	0.81587	0.0711	0.0318	1.77E−06	0.912011	22.80028921	0.029181201
rs117,132,123	G	T	10	119,756,612	0.7716	0.0686	0.02527	0.05058	0.1681	0.0569	4.44E−06	0.229087	21.06924683	0.029329456
rs2,041,604	G	C	17	65,757,750	0.3726	−0.0515	0.1218	0.11333	0.0789	0.0434	2.34E−06	0.234423	22.30137778	0.029699998
rs4,765,412	T	C	12	127,065,845	0.372	−0.0651	0.1126	0.09858	0.0808	0.0411	4.15E−06	0.112202	21.19645133	0.027655002
rs62,348,040	T	C	5	17,227,808	−0.2445	−0.0012	0.344	0.34219	0.0527	0.0304	3.45E−06	0.977237	21.52466973	0.026980501
rs72,808,762	C	T	2	46,347,515	0.4032	−0.0373	0.09678	0.14246	0.0864	0.0363	3.07E−06	0.301995	21.77777778	0.02842171
rs80,227,756	G	A	3	173,351,196	0.5246	−0.0465	0.06167	0.08821	0.1084	0.0454	1.30E−06	0.30903	23.42059953	0.03185049
rs8,112,871	T	G	19	17,024,672	0.2888	0.0109	0.2416	0.24983	0.0591	0.0284	1.02E−06	0.707946	23.87918037	0.030564664

Table S7. Independent instruments for myocarditis use in PD-L1: harmonized data

表S7. PD-L1中心肌炎使用的独立工具：协调数据

SNP	effect_ allele	other_ allele	chr	pos	beta. exposure	beta. outcome	eaf. exposure	eaf. outcome	se. exposure	se. outcome	pval. exposure	pval. outcome	F	R²_ exposure
rs10,254,255	A	G	7	96,468,009	−0.3395	−0.0054	0.8513	0.81587	0.0711	0.0319	1.77E−06	0.870964	22.80028921	0.029181201
rs117,132,123	G	T	10	119,756,612	0.7716	−0.1091	0.02527	0.05058	0.1681	0.0568	4.44E−06	0.0549541	21.06924683	0.029329456
rs2,041,604	G	C	17	65,757,750	0.3726	0.0262	0.1218	0.11333	0.0789	0.0434	2.34E−06	0.549541	22.30137778	0.029699998
rs4,765,412	T	C	12	127,065,845	0.372	0.0753	0.1126	0.09858	0.0808	0.0411	4.15E−06	0.0676083	21.19645133	0.027655002
rs62,348,040	T	C	5	17,227,808	−0.2445	0.0371	0.344	0.34219	0.0527	0.0304	3.45E−06	0.223872	21.52466973	0.026980501
rs72,808,762	C	T	2	46,347,515	0.4032	0.0904	0.09678	0.14246	0.0864	0.0363	3.07E−06	0.0125893	21.77777778	0.02842171
rs80,227,756	G	A	3	173,351,196	0.5246	0.0136	0.06167	0.08821	0.1084	0.0454	1.30E−06	0.758578	23.42059953	0.03185049
rs8,112,871	T	G	19	17,024,672	0.2888	−0.0201	0.2416	0.24983	0.0591	0.0284	1.02E−06	0.47863	23.87918037	0.030564664

NOTES

^*通讯作者。

参考文献

[1]	刘方圆, 高艳, 王文先, 史荣超, 李莎, 王锡明. 基于心脏磁共振晚期钆增强的心肌信号强度异质性对心肌炎患者预后价值的研究[J]. 磁共振成像, 2024, 15(3): 68-73.
[2]	师维, 崔旋旋, 周伍明, 华天凤, 杨旻. 1990-2019年中国心肌炎疾病负担水平及其变化分析[J]. 医学新知, 2023, 33(5): 325-333.
[3]	Lucas, J.A., Menke, J., Rabacal, W.A., Schoen, F.J., Sharpe, A.H. and Kelley, V.R. (2008) Programmed Death Ligand 1 Regulates a Critical Checkpoint for Autoimmune Myocarditis and Pneumonitis in MRL Mice. The Journal of Immunology, 181, 2513-2521. https://doi.org/10.4049/jimmunol.181.4.2513
[4]	Wang, J., Okazaki, I., Yoshida, T., Chikuma, S., Kato, Y., Nakaki, F., et al. (2010) PD-1 Deficiency Results in the Development of Fatal Myocarditis in MRL Mice. International Immunology, 22, 443-452. https://doi.org/10.1093/intimm/dxq026
[5]	Tarrio, M.L., Grabie, N., Bu, D., Sharpe, A.H. and Lichtman, A.H. (2012) PD-1 Protects against Inflammation and Myocyte Damage in T Cell-Mediated Myocarditis. The Journal of Immunology, 188, 4876-4884. https://doi.org/10.4049/jimmunol.1200389
[6]	Tsuruoka, K., Wakabayashi, S., Morihara, H., Matsunaga, N., Fujisaka, Y., Goto, I., et al. (2020) Exacerbation of Autoimmune Myocarditis by an Immune Checkpoint Inhibitor Is Dependent on Its Time of Administration in Mice. International Journal of Cardiology, 313, 67-75. https://doi.org/10.1016/j.ijcard.2020.04.033
[7]	Dong, H., Qi, Y., Kong, X., Wang, Z., Fang, Y. and Wang, J. (2022) PD-1/PD-L1 Inhibitor-Associated Myocarditis: Epidemiology, Characteristics, Diagnosis, Treatment, and Potential Mechanism. Frontiers in Pharmacology, 13, Article 835510. https://doi.org/10.3389/fphar.2022.835510
[8]	Moslehi, J., Lichtman, A.H., Sharpe, A.H., Galluzzi, L. and Kitsis, R.N. (2021) Immune Checkpoint Inhibitor—Associated Myocarditis: Manifestations and Mechanisms. Journal of Clinical Investigation, 131, e145186. https://doi.org/10.1172/jci145186
[9]	汪潇潇, 陈万一. PD-1/PD-L1抑制剂相关心肌炎不良反应——基于三大报告系统的真实世界研究[J]. 中国新药杂志, 2019, 28(11): 1404-1408.
[10]	Levin, M.G. and Burgess, S. (2024) Mendelian Randomization as a Tool for Cardiovascular Research: A Review. JAMA Cardiology, 9, 79-89. https://doi.org/10.1001/jamacardio.2023.4115
[11]	于天琦, 徐文涛, 苏雅娜, 李静. 孟德尔随机化研究基本原理、方法和局限性[J]. 中国循证医学杂志, 2021, 21(10): 1227-1234.
[12]	夏晓琳, 刘晔, 徐晓冰, 孟俊彤, 王艳. 甲状腺功能减退症与冠心病因果关系的孟德尔随机化研究[J]. 现代预防医学, 2024, 51(3): 412-416.
[13]	杨梦霞, 赵春铭, 陈腾飞, 徐霄龙, 刘清泉. 脓毒症与心律失常因果关联的两样本孟德尔随机化研究[J]. 中国急救医学, 2024, 44(2): 142-147.
[14]	Davies, N.M., Holmes, M.V. and Davey Smith, G. (2018) Reading Mendelian Randomisation Studies: A Guide, Glossary, and Checklist for Clinicians. BMJ, 362, k601. https://doi.org/10.1136/bmj.k601
[15]	杜冬杰, 陈健, 王诗悦, 等. 血残粒胆固醇水平与冠心病和心肌梗死风险的两样本孟德尔随机化研究[J]. 中国动脉硬化杂志, 2023, 31(6): 510-516.
[16]	Larsson, S.C., Butterworth, A.S. and Burgess, S. (2023) Mendelian Randomization for Cardiovascular Diseases: Principles and Applications. European Heart Journal, 44, 4913-4924. https://doi.org/10.1093/eurheartj/ehad736
[17]	Sun, B.B., Maranville, J.C., Peters, J.E., Stacey, D., Staley, J.R., Blackshaw, J., et al. (2018) Genomic Atlas of the Human Plasma Proteome. Nature, 558, 73-79. https://doi.org/10.1038/s41586-018-0175-2
[18]	Kurki, M.I., Karjalainen, J., Palta, P., Sipilä, T.P., Kristiansson, K., Donner, K.M., et al. (2023) Finngen Provides Genetic Insights from a Well-Phenotyped Isolated Population. Nature, 613, 508-518. https://doi.org/10.1038/s41586-022-05473-8
[19]	Burgess, S., Scott, R.A., Timpson, N.J., Davey Smith, G. and Thompson, S.G. (2015) Using Published Data in Mendelian Randomization: A Blueprint for Efficient Identification of Causal Risk Factors. European Journal of Epidemiology, 30, 543-552. https://doi.org/10.1007/s10654-015-0011-z
[20]	Verbanck, M., Chen, C., Neale, B. and Do, R. (2018) Detection of Widespread Horizontal Pleiotropy in Causal Relationships Inferred from Mendelian Randomization between Complex Traits and Diseases. Nature Genetics, 50, 693-698. https://doi.org/10.1038/s41588-018-0099-7
[21]	Zhao, Q., Wang, J., Hemani, G., Bowden, J. and Small, D.S. (2020) Statistical Inference in Two-Sample Summary-Data Mendelian Randomization Using Robust Adjusted Profile Score. The Annals of Statistics, 48, 1742-1769. https://doi.org/10.1214/19-aos1866
[22]	Bowden, J., Davey Smith, G., Haycock, P.C. and Burgess, S. (2016) Consistent Estimation in Mendelian Randomization with Some Invalid Instruments Using a Weighted Median Estimator. Genetic Epidemiology, 40, 304-314. https://doi.org/10.1002/gepi.21965
[23]	Sanderson, E., Davey Smith, G., Windmeijer, F. and Bowden, J. (2018) An Examination of Multivariable Mendelian Randomization in the Single-Sample and Two-Sample Summary Data Settings. International Journal of Epidemiology, 48, 713-727. https://doi.org/10.1093/ije/dyy262
[24]	Jo, W., Won, T., Daoud, A. and Čiháková, D. (2024) Immune Checkpoint Inhibitors Associated Cardiovascular Immune-Related Adverse Events. Frontiers in Immunology, 15, Article 1340373. https://doi.org/10.3389/fimmu.2024.1340373
[25]	Salem, J., Manouchehri, A., Moey, M., Lebrun-Vignes, B., Bastarache, L., Pariente, A., et al. (2018) Cardiovascular Toxicities Associated with Immune Checkpoint Inhibitors: An Observational, Retrospective, Pharmacovigilance Study. The Lancet Oncology, 19, 1579-1589. https://doi.org/10.1016/s1470-2045(18)30608-9
[26]	Moslehi, J.J., Salem, J., Sosman, J.A., Lebrun-Vignes, B. and Johnson, D.B. (2018) Increased Reporting of Fatal Immune Checkpoint Inhibitor-Associated Myocarditis. The Lancet, 391, 933. https://doi.org/10.1016/s0140-6736(18)30533-6
[27]	Zhang, C., Wei, F., Ma, W. and Zhang, J. (2024) Immune-related Cardiovascular Toxicities of PD-1/PD-L1 Inhibitors in Solid Tumors: An Updated Systematic Review and Meta-Analysis. Frontiers in Immunology, 15, Article 1255825. https://doi.org/10.3389/fimmu.2024.1255825
[28]	Sanderson, E., Glymour, M.M., Holmes, M.V., Kang, H., Morrison, J., Munafò, M.R., et al. (2022) Mendelian randomization. Nature Reviews Methods Primers, 2, Article No. 6. https://doi.org/10.1038/s43586-021-00092-5
[29]	Baik, A.H., Oluwole, O.O., Johnson, D.B., Shah, N., Salem, J., Tsai, K.K., et al. (2021) Mechanisms of Cardiovascular Toxicities Associated with Immunotherapies. Circulation Research, 128, 1780-1801. https://doi.org/10.1161/circresaha.120.315894
[30]	Rubio-Infante, N., Ramírez-Flores, Y.A., Castillo, E.C., Lozano, O., García-Rivas, G. and Torre-Amione, G. (2021) Cardiotoxicity Associated with Immune Checkpoint Inhibitor Therapy: A Meta-Analysis. European Journal of Heart Failure, 23, 1739-1747. https://pubmed.ncbi.nlm.nih.gov/34196077/
[31]	Xia, W., Zou, C., Chen, H., Xie, C. and Hou, M. (2020) Immune Checkpoint Inhibitor Induces Cardiac Injury through Polarizing Macrophages via Modulating microRNA-34a/Kruppel-Like Factor 4 Signaling. Cell Death & Disease, 11, Article No. 575. https://doi.org/10.1038/s41419-020-02778-2
[32]	Wei, S.C., Meijers, W.C., Axelrod, M.L., Anang, N.A.S., Screever, E.M., Wescott, E.C., et al. (2021) A Genetic Mouse Model Recapitulates Immune Checkpoint Inhibitor–associated Myocarditis and Supports a Mechanism-Based Therapeutic Intervention. Cancer Discovery, 11, 614-625. https://doi.org/10.1158/2159-8290.cd-20-0856
[33]	Yamaguchi, S., Morimoto, R., Okumura, T., Yamashita, Y., Haga, T., Kuwayama, T., et al. (2018) Late-Onset Fulminant Myocarditis with Immune Checkpoint Inhibitor Nivolumab. Canadian Journal of Cardiology, 34, 812.e1-812.e3. https://doi.org/10.1016/j.cjca.2018.03.007
[34]	Juchem, K.W., Sacirbegovic, F., Zhang, C., Sharpe, A.H., Russell, K., McNiff, J.M., et al. (2018) PD-L1 Prevents the Development of Autoimmune Heart Disease in Graft-versus-Host Disease. The Journal of Immunology, 200, 834-846. https://doi.org/10.4049/jimmunol.1701076
[35]	Bracamonte-Baran, W., Gilotra, N.A., Won, T., Rodriguez, K.M., Talor, M.V., Oh, B.C., et al. (2021) Endothelial Stromal PD-L1 (Programmed Death Ligand 1) Modulates CD8⁺ T-Cell Infiltration after Heart Transplantation. Circulation: Heart Failure, 14, e007982. https://doi.org/10.1161/circheartfailure.120.007982
[36]	Bishawi, M., Bowles, D., Pla, M.M., Oakes, F., Chiang, Y., Schroder, J., et al. (2021) PD-1 and PD-L1 Expression in Cardiac Transplantation. Cardiovascular Pathology, 54, Article ID: 107331. https://doi.org/10.1016/j.carpath.2021.107331
[37]	Mahmood, S.S., Fradley, M.G., Cohen, J.V., Nohria, A., Reynolds, K.L., Heinzerling, L.M., et al. (2018) Myocarditis in Patients Treated with Immune Checkpoint Inhibitors. Journal of the American College of Cardiology, 71, 1755-1764. https://doi.org/10.1016/j.jacc.2018.02.037
[38]	Pillai, R.N., Behera, M., Owonikoko, T.K., Kamphorst, A.O., Pakkala, S., Belani, C.P., et al. (2017) Comparison of the Toxicity Profile of PD‐1 versus PD‐L1 Inhibitors in Non-Small Cell Lung Cancer: A Systematic Analysis of the Literature. Cancer, 124, 271-277. https://doi.org/10.1002/cncr.31043
[39]	Li, W. and Wang, W. (2023) Unraveling the Genetic Associations between PD-1/PD-L1 and 13 Circulating Biomarkers Linked to Physiological and Pathological Processes. Clinical and Translational Oncology, 26, 1157-1169. https://doi.org/10.1007/s12094-023-03333-z
[40]	Ference, B.A., Majeed, F., Penumetcha, R., Flack, J.M. and Brook, R.D. (2015) Effect of Naturally Random Allocation to Lower Low-Density Lipoprotein Cholesterol on the Risk of Coronary Heart Disease Mediated by Polymorphisms in NPC1L1, HMGCR, or Both: A 2 × 2 Factorial Mendelian Randomization Study. Journal of the American College of Cardiology, 65, 1552-1561. https://doi.org/10.1016/j.jacc.2015.02.020
[41]	Zou, X., Wang, S., Wang, L., Xiao, L., Yao, T., Zeng, Y., et al. (2021) Childhood Obesity and Risk of Stroke: A Mendelian Randomisation Analysis. Frontiers in Genetics, 12, Article 727475. https://doi.org/10.3389/fgene.2021.727475

为你推荐

友情链接