This single-center, retrospective observational study included 992 consecutive patients with TBAD undergoing TEVAR from January 2010 to December 2017. The diagnosis of TBAD was validated through enhanced computed tomography angiography (CTA). The inclusion criteria were the patients of TBAD undergoing TEVAR. Exclusion of patients occurred due to the following factors: (1) blunt traumatic aortic injury, (2) malignant tumor, (3) disorders of connective tissue, (4) prior surgical intervention on the aorta, (5) insufficient data for nutritional assessment (Supplementary Fig. 1). The final analysis included the remaining 881 participants. The ethics committee of Guangdong Provincial People’s Hospital (#201807) gave approval for this study and waived the requirement for informed consent.

Blood samples were harvested at admission and blood routine examination, lipid profile test, and other laboratory indicators analysis were performed in the central laboratory of the hospital. The CONUT score was created and confirmed as a tool for evaluating the nutritional status of patients admitted to the hospital. It is calculated by adding up the scores of total lymphocytes, albumin level, and total cholesterol levels (Supplementary Table 1) [10]. Scores on a scale of 0 to 12, with higher scores indicating a more unfavorable condition. To detect sarcopenia, the skeletal muscle mass index (SMI) was determined using the following formulas: for males, 0.220 multiplied by the body mass index (BMI) and then added to 2.991; for females, 0.141 multiplied by the BMI and then added to 3.377 [12]. Complex TBAD was defined as TBAD accompanied by persistent pain, unresponsive hypertension despite maximum medication, rapid aortic enlargement, malperfusion syndromes, and signs of rupture (such as hemothorax, increasing periaortic and mediastinal hematoma) [1].

All patients received with standardized medications and TEVAR treatment following the current guideline and consensus [1, 13]. Patients who had uncomplicated TBAD were subjected to TEVAR if the aortic diameter exceeded 40 mm, primary entry tear diameter $>$10 mm, false lumen (FL) diameter $>$22 mm and a patent or partially thrombosed FL [14]. The details of the procedures at our center have been described elsewhere [15]. Briefly, the stent-graft was inserted in a reverse manner through percutaneous femoral artery entry to close the initial tear and the stent-graft size was typically larger by 5% to 10%. When needed, the left subclavian artery (LSA) and/or left common carotid artery (LCCA) were occluded in order to achieve a minimum proximal landing zone of 1.5 centimeters. The method to reconstructing the arch vessels (including chimney or hybrid techniques like TEVAR combined with supra-arch bypass) was determined by the operating surgeon based on the specific characteristics of the aortic pathologies.

All in-hospital survival patients received clinical and the image of CTA follow-up at 3, 6, 12 months, and subsequently on an annual basis. The assessment of the patient’s state was carried out either by visiting the outpatient clinic or by conducting a telephone interview. The primary outcome were thirty-day death and follow-up mortality. The secondary results included early outcomes that happened during the hospital stay or within 30 days after the procedure. These outcomes encompassed mortality, extended stay in the intensive care unit (ICU), confusion, stroke caused by lack of blood supply to the brain, reduced blood flow to limbs or organs, reduced blood flow to the spinal cord, acute kidney damage, the need for further intervention, and subsequent intervention or stroke during follow-up.

Mean $\pm$ standard deviation or median and interquartile range (IQR) are used to express continuous variables based on their distributions and compared using Student’s t-test or the Manne-Whitney U test, as appropriate. The presentation of categorical variables is in the form of n (%) and they were compared using either the chi-square test or Fisher’s exact test. To evaluate the associations between patient characteristics and pre-operative nutritional status, Spearman’s rank correlation test was utilized.

The primary predictor was pre-operative CONUT score modeled as a continuous variable. The CONUT score ($>$5) before surgery was used as a categorical variable in the secondary prediction model. Youden’s index was used to determine the optimal threshold value of the CONUT score (CONUT = 5) for predicting post-operative mortality, which was obtained from the receiver operating characteristic (ROC) curves [16]. High CONUT ($>$5) was defined as the malnutrition. To examine the connection between the nutritional status before surgery and the initial results, logistic regression models were utilized.

The Kaplan-Meier method was used to present survival data and the log-rank test was employed to compare differences in survival. To evaluate the impact of preoperative nutritional condition on subsequent overall mortality, Cox proportional-hazards regression models were employed. Formal tests were conducted to examine the assumption of proportional hazard, utilizing the techniques outlined by Grambsch and Therneau [17]. There was no indication of any breaches of this assumption. An initial multivariable Cox regression model included demographic characteristics, comorbidities, laboratory tests, and imaging findings. Variables that had a p value less than 0.1 in the univariable analysis were included in the multivariable models using a forward stepwise technique. Additionally, to reduce bias and mimic randomization, two propensity-score methods were used to account for potential confounding by characteristics influencing outcomes. A multivariate logistic regression model with covariates was used to estimate individual tendencies for each subject.

The primary analysis used inverse probability of treatment weighting (IPTW) [18]. The stabilized IPTW weight was calculated using the predicted probabilities derived from the propensity-score model in the IPTW analysis. Cox regression models that used the IPTW weights were reported. In the same Cox regression model, we conducted a secondary analysis incorporating the propensity score as an extra covariate.

To further analyze the data, the CONUT score threshold was modified, categorizing it into different levels: normal (0–1), mild malnutrition (2–4), moderate malnutrition (5–8), and severe malnutrition ($>$8) [10]. R software (version 4.0.3, R Foundation for Statistical Computing, Vienna, Austria) and IBM SPSS 25.0 (SPSS 25 Inc., Armonk, NY, USA) were utilized for conducting all statistical analyses. A significance level of less than 0.05 was deemed significant.

Of the 881 patients enrolled, the average age was 54.2 $\pm$ 10.9 years and the majority of the participants (86.6%) were male. The average BMI was 24.5 $\pm$ 3.7 kg/m${}^2$, and the most common comorbid condition was hypertension (84.8%). Based on the optimal CONUT score threshold, patients were divided into two groups: the high CONUT group ($>$5, n = 702) and the low CONUT group ($\le$5, n = 179). Additional details of baseline characteristics are presented in Table 1. The prevalence of malnutrition defined as CONUT score greater than 5 was 20.3%. Patients with a BMI below 18.5 kg/m${}^2$ had the highest rate of malnutrition, accounting for 36.8% (Fig. 1). More than 10% of patients (48/359) suffered malnutrition even in the overweight/obesity group. The distribution of malnutrition did not significantly differ between men and women (19.8% vs. 23.7%; p = 0.322).

Fig. 1.

Distribution of malnutrition by subgroups of patients according to body mass index. CONUT, controlling nutritional status.

Individuals belonging to the high CONUT category exhibited a greater tendency towards advanced age, elevated occurrences of subacute individuals, anemia, the engagement of visceral arteries, the engagement of renal arteries, pericardial effusion, pleural effusion, creatinine levels surpassing 2 mg/dL, and an estimated glomerular filtration rate (eGFR) 60 mL/min/1.73 m${}^2$. Additionally, they exhibited reduced BMI, SMI, lymphocyte count, albumin concentration, total cholesterol (TC) concentration, triglyceride concentration, and low-density lipoprotein cholesterol (LDL-c) concentration. More data on the baseline characteristics were detailed in Table 1.

The CONUT score was significantly associated with age, BMI, SMI, hemoglobin, albumin, lymphocyte, TC, triglyceride, LDL-c, creatinine, and eGFR (p 0.05 for all), as indicated in Table 2. Multivariable logistic regression analysis revealed that anemia (odds ratio [OR], 1.76; 95% confidence interval [CI], 1.07–2.90) and serum albumin (OR, 0.55; 95% CI, 0.50–0.60) were the independent predictors for malnutrition.

Patients classified in the high CONUT group had significantly higher rates of 30-day prolonged ICU stay, mortality, and post- surgical confusion (p 0.05 for all), However, they exhibited comparable occurrences of stroke, limb and organ ischemia, spinal cord ischemia, and re-intervention (Table 3). Moreover, the prevalence of post-operative acute kidney injury (AKI) was greater in the high CONUT category (30.7%) compared to the low CONUT category (24.5%), although this disparity did not achieve statistical significance.

Multivariable logistic analyses indicated that the CONUT score, which assessed pre-operative nutritional status, was a significant independent predictor of prolonged ICU stay (OR, 1.09; 95% CI, 1.02–1.17; p = 0.015), 30-day death (OR, 1.43; 95% CI, 1.19–1.72; p 0.001), delirium (OR, 1.11; 95% CI, 1.01–1.23; p = 0.035) and AKI (OR, 1.09; 95% CI, 1.01–1.16; p = 0.027) (Table 4). Likewise, an elevated CONUT score ($>$5) was found to be linked with a longer duration of stay in the ICU (OR, 1.69; 95% CI, 1.15–2.50; p = 0.008), while it did not show any association with other initial unfavorable results (Table 4).

Over a period of 70.8 months (with an interquartile range of 46.1–90.8 months), 102 (11.8%) patients died after the procedures, with 39 (21.8%) and 63 (9.0%) patients in the high and low CONUT group, respectively (p 0.001; Table 3). The overall survival rate for follow-up all-cause mortality was considerably greater in the high CONUT category compared to the low CONUT category (21.8% vs. 9.0%; log-rank p 0.001; Fig. 2). Furthermore, the occurrence of subsequent stroke and repeat procedure were comparable in both groups (p $>$ 0.05 for both; Table 3).

Fig. 2.

Kaplan-Meier curves for all-cause mortality by the CONUT score. CONUT, controlling nutritional status.

After adjusting for confounding factors (Table 4), Cox multivariate analysis showed that the hazard ratios (HR) for follow-up mortality were 1.13 (95% CI, 1.05–1.23; p = 0.002) for CONUT score as a continuous variable. Additional significant factors included BMI (HR, 0.88; 95% CI, 0.83–0.94; p 0.001), creatinine levels $>$2 mg/dL (HR, 2.34; 95% CI, 1.40–3.92; p = 0.001), maximum aortic diameter $>$40 mm (HR, 1.62; 95% CI, 1.07–2.45; p = 0.023), presence of abdominal aortic aneurysm (HR, 2.26; 95% CI, 1.13–4.51; p = 0.021), occurrence of post-operative AKI (HR, 1.81; 95% CI, 1.21–2.72; p = 0.004) and development of delirium (HR, 2.21; 95% CI, 1.33–3.67; p = 0.002) (Fig. 3). Furthermore, the association remained (HR, 1.68; 95% CI, 1.11–2.53; p = 0.014) even after including the CONUT as a categorical factor (CONUT score $>$5; Table 4).

Fig. 3.

Multivariate analysis results of follow-up mortality. HR, hazard ratio; CI, confidence interval; CONUT, controlling nutritional status; AKI, acute kidney injury.

In the multivariable analysis using IPTW based on the propensity score, the CONUT score $>$5 remained as the independent risk factor of follow-up mortality (HR, 1.76; 95% CI, 1.10–2.81; p = 0.018; Table 5). A comparable outcome was noted when the propensity score was included in the identical model (HR, 1.68; 95% CI, 1.11–2.53; p = 0.014; Table 5). Moreover, there was no correlation between the CONUT score and subsequent re-interventions (p = 0.193) or the incidence of stroke (p = 0.290).

Moreover, the threshold for the CONUT score was modified, and malnutrition was categorized as follows: a CONUT score of 0 to 1 was classified as normal, while scores of 2 to 4, 5 to 8, and 9 to 12 were designated as mild, moderate, and severe malnutrition, respectively [10]. Pre-operative significant undernourishment persisted as a separate forecaster for death within 30 days, delirium after surgery, acute kidney injury after surgery, and mortality during follow-up (p 0.05 for all; see Table 4).