Stanford B型创伤性主动脉夹层患者术后血清NF-κB水平变化对早期主动脉不良事件的预测价值

The predictive value of postoperative serum NF-κB level changes for early aortic adverse events in patients with Stanford type B traumatic aortic dissection

摘要:

目的]探究Stanford B型创伤性主动脉夹层(TBAD)患者术后血清核因子κB(NF-κB)水平变化与早期主动脉不良事件(AAE)发生的相关性及其预测价值,为临床早期识别术后AAE发生风险提供理论依据。 [方法]回顾性选取2021年4月至2024年8月我院收治的159例TBAD患者,经纳入排除标准筛选后最终纳入146例；根据术后是否发生AAE,将患者分为AAE组(n＝47)和非AAE组(n＝99)。另从同期在我院接受胸主动脉腔内修复术治疗的364例非创伤性主动脉夹层患者中,采用随机数字表法抽取150例作为非TBAD组。采用重复测量方差分析比较患者术前、术后1个月及术后3个月的血清NF-κB水平；绘制剂量依赖图,分析血清NF-κB下降水平(ΔNF-κB)与假腔血栓化形成概率的关联；通过广义混合效应模型(GLMM)探讨ΔNF-κB与AAE发生风险的关系；利用ROC曲线评估ΔNF-κB对AAE的预测效能；采用限制性立方样条(RCS)模型分析ΔNF-κB与AAE发生风险的关联；借助R包Interplot绘制ΔNF-κB对假腔血栓化程度与AAE发生风险的交互作用图；使用Kaplan-Meier法绘制生存曲线,评估不同ΔNF-κB水平患者的AAE发生情况。 [结果]AAE组术后1个月NF-κB下降水平(ΔNF-κB₁)分别是非TBAD组和非AAE组的0.69倍、0.76倍；AAE组术后3个月NF-κB下降水平(ΔNF-κB₂)分别是非TBAD组和非AAE组的0.74倍、0.80倍(P＜0.05)。剂量依赖图结果显示,随着AAE组ΔNF-κB₁和ΔNF-κB₂的升高,假腔血栓化形成概率逐渐上升；当ΔNF-κB₁＞0.62 μg/L或ΔNF-κB₂＞0.69 μg/L时,假腔血栓化形成概率趋于稳定。GLMM分析结果显示,与ΔNF-κB₁＜0.15 μg/L相比,ΔNF-κB₁＜0.56 μg/L时与AAE发生风险呈正相关,ΔNF-κB₁≥0.56 μg/L时与AAE发生风险呈负相关(P＜0.05)；与ΔNF-κB₂＜0.18 μg/L相比,ΔNF-κB₂＜0.66 μg/L时与AAE发生风险呈正相关,ΔNF-κB₂≥0.66 μg/L时与AAE发生风险呈负相关。ROC曲线分析结果显示,ΔNF-κB₁预测AAE发生的AUC值为0.784(95%CI:0.710～0.859,P＜0.001),最佳截断值为0.62 μg/L；ΔNF-κB₂预测AAE发生的AUC值为0.811(95%CI:0.741～0.881,P＜0.001),最佳截断值为0.69 μg/L,且ΔNF-κB₂对AAE发生的预测效能更强。RCS模型分析结果显示,随着ΔNF-κB₁或ΔNF-κB₂的升高,AAE发生风险呈下降趋势(P＜0.05)。交互作用图显示,随着NF-κB下降水平的升高,假腔血栓化程度与AAE发生风险的负相关关系随之增强。Kaplan-Meier生存曲线结果显示,不同ΔNF-κB₁、ΔNF-κB₂水平患者AAE发生率差异有统计学意义(χ²＝5.064,P＝0.028),高ΔNF-κB水平患者AAE未发生率更高。 [结论]TBAD患者术后血清NF-κB的下降水平可有效预测早期AAE的发生风险,且二者呈负相关关系,这一效应可能通过促进假腔血栓化实现。

Abstract:

Aim To explore the correlation between postoperative serum nuclear factor-κB (NF-κB) level changes and early aortic adverse events (AAE) as well as its predictive value in patients with Stanford type B traumatic aortic dissection (TBAD), so as to provide a theoretical basis for early clinical identification of postoperative AAE risk. Methods A total of 159 patients with TBAD admitted to our hospital from April 2021 to August 2024 were retrospectively enrolled, and 146 patients were finally included according to the inclusion and exclusion criteria. All patients were divided into AAE group (n＝47) and non-AAE group (n＝99) based on the occurrence of postoperative AAE. Meanwhile, 150 cases were randomly selected via the random number table method from 364 patients with non-traumatic aortic dissection receiving thoracic endovascular aortic repair, who were assigned to the non-TBAD group. Repeated-measures analysis of variance was used to compare NF-κB levels before surgery, 1 month and 3 months after surgery. Dose-response plots were generated to analyze the correlation between the reduction in NF-κB levels (ΔNF-κB) and the probability of false lumen thrombosis. A generalized linear mixed model (GLMM) was employed to analyze the correlation between ΔNF-κB and the risk of AAE. ROC curves were used to evaluate the predictive performance of ΔNF-κB for AAE. Restricted cubic spline (RCS) models were used to analyze the association between ΔNF-κB and AAE risk. The R package Interplot was used to generate interaction plots of ΔNF-κB with false lumen thrombosis degree on AAE risk. The Kaplan-Meier method was used to plot survival curves to evaluate the occurrence of AAE in patients with different levels of ΔNF-κB. Results The 1-month postoperative NF-κB decreased level (ΔNF-κB₁) in the AAE group was 0.69-fold and 0.76-fold that of the non-TBAD and non-AAE groups, respectively, while the 3-month postoperative decreased level (ΔNF-κB₂) was 0.74-fold and 0.80-fold, respectively (P＜0.05). The dose-dependent plot results showed that the probability of false lumen thrombosis formation gradually increased with the elevation of ΔNF-κB₁ and ΔNF-κB₂ in the AAE group, and tended to stabilize when ΔNF-κB₁＞0.62 μg/L or ΔNF-κB₂＞0.69 μg/L. GLMM analysis showed that compared with ΔNF-κB₁＜0.15 μg/L, ΔNF-κB₁＜0.56 μg/L was positively associated with AAE risk, and ΔNF-κB₁≥0.56 μg/L was negatively associated with AAE risk (P＜0.05). Compared with ΔNF-κB₂＜0.18 μg/L, ΔNF-κB₂＜0.66 μg/L was positively correlated with AAE risk, and ΔNF-κB₂≥0.66 μg/L was negatively correlated with AAE risk. ROC analysis showed that the AUC of ΔNF-κB₁ for predicting AAE was 0.784 (95%CI:0.710～0.859, P＜0.001), and the best cut-off value was 0.62 μg/L; the AUC of ΔNF-κB₂ for predicting AAE was 0.811 (95%CI:0.741～0.881, P＜0.001), and the best cut-off value was 0.69 μg/L, suggesting that ΔNF-κB₂ had a stronger predictive performance for AAE. RCS model analysis showed that with the increase of ΔNF-κB₁ and ΔNF-κB₂, the risk of AAE showed a decreasing trend (P＜0.05). The interaction diagram showed that the negative correlation between false lumen thrombosis degree and AAE risk was enhanced along with the increase of NF-κB reduction magnitude. The Kaplan-Meier survival curve showed that there was a statistically significant difference in the incidence of AAE among patients with different levels of ΔNF-κB₁ and ΔNF-κB₂ (χ²＝5.064, P＝0.028), and patients with higher ΔNF-κB levels had a lower incidence of AAE. Conclusion Postoperative reduction of serum NF-κB level can effectively predict the risk of early AAE in TBAD patients, which is negatively correlated with the AAE risk, and this mechanism may be mediated by promoting false lumen thrombosis.