Construction Of Nomogram For Acute Aortic Dissection Complicated With Hypoxemia And Exploration Of Inflammatory Indicators

Background:Acute aortic dissection(AAD)is one of the most common emergency and critical cardiovascular diseases,which is often associated with serious complications such as malperfusion syndromes,hypoxemia,pericardial and pleural effusion,acute renal insufficiency and so on.Hypoxemia(HXA)is one of the common complications of AAD.At the same time,the occurrence of HXA can further aggravate the condition of AAD patients and affect their prognosis.Screening the clinical risk factors associated with AAD combined with HXA,exploring the relevant indicators causing the inflammatory response,trying to elaborate the relevant inflammatory response,and constructing a columnar line graph prediction model for the occurrence of HXA in AAD,which can help to provide early warning and early intervention,and provide ideas for clinical guidance and research experiments.Objective:Through statistical analysis of the clinical data and characteristics of patients with acute aortic coarctation during hospitalization,we screened the independent risk factors associated with HXA,constructed a prediction model of HXA column line graph,and initially elucidated the possible mechanisms causing HXA to provide clinical reference for early prevention and prognosis.Methods:The clinical datum of 348 patients with AAD diagnosed by transthoracic aortic CTA and admitted to the Cardiovascular Medical Center of Huizhou Hospital,Guangzhou Medical University from December 2016 to October 2022 were retrospectively analyzed.According to the inclusion and exclusion criteria,the obtained patients were divided into hypoxic group(oxygenation index≤200 mm Hg)and non-hypoxic group according to the lowest oxygenation index during hospitalization(within24 hours postoperatively excluded).The risk factor indexes that might be related to HXA were included in the database,and then subjected to univariate analysis to screen out the meaningful risk factors.Next,multifactorial logistic regression analysis was performed to obtain independent risk factors.The obtained independent risk factors were analyzed by establishing regression equations and constructing a column line graph prediction model.We assessed the discrimination and calibration of the model using the receiver operating characteristic curve(ROC curve)and the goodness-of-fit(Hosmer-Lemeshow)test,and conducted clinical validity tests,while using a total of28 AAD patients from October to December in 2022 to verify external validation of the model.Finally,plasma specimens from 40 patients with AAD(21 hypoxic vs.19 non-hypoxic)and 20 healthy individuals with physical examination were sent for heparin-binding protein,myeloperoxidase,matrix metalloproteinase 2,matrix metalloproteinase 9 and neutrophil elastase to analyze the variability of inflammatory factors expression among the three groups.Results:1.The prevalence of HXA among the included AAD patients was about 41.95%.2.The results of multifactorial binary logistic regression analysis suggested that the AAD patients with BMI(OR:0.120;95%CI:0.852-0.941;P=0.120),smoking history(OR:0.844;95%CI:1.249-4.327;P=0.008),neutrophil count(OR:0.502;95%CI:1.453～1.880;P<0.01),CRP(OR:0.023;95%CI:1.003～1.021;P<0.01),and ALB(OR:-0.127;95%CI:0.825～0.941;P<0.01)were independent risk factors for the occurrence of HXA.3.Columnar graph prediction models were constructed by the screened independent risk factors and validated for discrimination,calibration,and clinical validity.The area under the ROC curve of the prediction model was 0.887,and the Hosmer-Lemeshow test showedχ~2=3.461(P>0.05),which indicated that the prediction model had good calibration.The valve probability in the DCA curve was 34%,and the net clinical benefit was 65%.4.HBP was statistically different among the healthy,hypoxic,and non-hypoxic groups(P<0.05).MMP-2 and MMP-9 were statistically different(P<0.05)between healthy vs hypoxic and healthy vs non-hypoxic groups,but not between hypoxic and non-hypoxic groups.Conclusions:1.CRP,neutrophil count,BMI and smoking were independent risk factors for the occurrence of HXA during hospitalization in patients with aortic coarctation.ALB was independently a protective factor.2.The established logistic regression model has high predictive efficacy for the occurrence of HXA during hospitalization in patients with aortic coarctation,and the visualized column line graphs have good discrimination,calibration and clinical validity,which are of clinical guidance.3.HBP,MMP-2 and MMP-9 may be involved in the early inflammatory cascade response,which have certain guiding significance for clinical guidance,column line graph construction improvement and basic research.