Prediction of thirty-day mortality risk after thoracic endovascular aortic repair for intact descending thoracic aortic aneurysms: Derivation of risk calculator in the Vascular Quality Initiative

Using machine learning to predict outcomes of patients with blunt traumatic aortic injuries

BACKGROUND

The optimal management of blunt thoracic aortic injury (BTAI) remains controversial, with experienced centers offering therapy ranging from medical management to TEVAR. We investigated the utility of a machine learning (ML) algorithm to develop a prognostic model of risk factors on mortality in patients with BTAI.

METHODS

The Aortic Trauma Foundation registry was utilized to examine demographics, injury characteristics, management and outcomes of patients with BTAI. A STREAMLINE (A Simple, Transparent, End-To-End Automated Machine Learning Pipeline Facilitating Data Analysis and Algorithm Comparison) model as well as logistic regression (LR) analysis with imputation using chained equations was developed and compared.

RESULTS

From a total of 1018 patients in the registry, 702 patients were included in the final analysis. Of the 258 (37%) patients who were medically managed, 44 (17%) died during admission, 14 (5.4%) of which were aortic related deaths. Four hundred forty-four (63%) patients underwent TEVAR and 343 of which underwent TEVAR within 24 hours of admission. Among TEVAR patients, 39 (8.8%) patients died and 7 (1.6%) had aortic related deaths ( Table 1 ). Comparison of the STREAMLINE and LR model showed no significant difference in ROC curves and high AUCs of 0.869 (95% confidence interval, 0.813-0.925) and 0.840 (95% confidence interval, 0.779-0.900) respectively in predicting in-hospital mortality. Unexpectedly, however, the variables prioritized in each model differed between models. The top 3 variables identified from the LR model were similar to that from existing literature. The STREAMLINE model, however, prioritized location of the injury along the lesser curve, age and aortic injury grade.

CONCLUSION

Machine learning provides insight on prioritization of variables not typically identified in standard multivariable logistic regression. Further investigation and validation in other aortic injury cohorts are needed to delineate the utility of ML models.

LEVEL OF EVIDENCE

Prognostic and Epidemiological; Level III.

Preoperative risk score for mortality within 3 years of visceral segment fenestrated endovascular aortic repair

OBJECTIVE

The purpose of this study was to create a risk score for the event of mortality within 3 years of complex fenestrated visceral segment endovascular aortic repair utilizing variables existing at the time of preoperative presentation.

METHODS

After exclusions, 1916 patients were identified in the Vascular Quality Initiative who were included in the analysis. The first step in development of the risk score was univariable analysis for the primary outcome of mortality within 3 years of surgery. χ2 analysis was performed for categorical variables, and comparison of means with independent Student t-test was performed for ordinal variables. Variables that achieved a univariable P value less than 0.1 were then placed into Cox regression multivariable time dependent analysis for the development of mortality within 3 years. Variables that achieved a multivariable significance of less than 0.1 were utilized for the risk score, with point weighting based on the beta-coefficient. Variables with a beta coefficient of 0.25 to 0.49 were assigned 1 point, 0.5 to 0.74 2 points, 0.75 to 0.99 3 points, and 1.0 to 1.25 4 points. A cumulative score for each patient was then summed, the percentage of patients at each score experiencing mortality within 3 weeks was then calculated, and a comparison of score outcomes was conducted with binary logistic regression. Area under the curve analysis was performed.

RESULTS

The primary outcome of mortality within 3 years of surgery occurred in 12.8% of patients (245/1916). The mean age for the study population was 73.35 years (standard deviation [SD], 8.26 years). The mean maximal abdominal aortic aneurysm (AAA) diameter was 60.43 mm (SD, 10.52 mm). The mean number of visceral vessels stented was 3.3 (SD, 0.76). Variables present at the time of surgery that were included in the risk score were: hemodialysis (3 points); age >87, chronic obstructive pulmonary disease, hypertension, AAA diameter >77 mm (all 2 points); and body mass index <20 kg/m2, female sex, congestive heart failure, active smoking, chronic renal insufficiency, age 80 to 87 years, and AAA diameter 67 to 77 mm (all 1 point). BMI >30 kg/m2 (mean, 34.46 kg/m2) and age <67 years were protective (-1 point). Testing the model resulted in an area under the curve of 0.706. Hosmer and Lemeshow goodness of fit test for logistic regression utilizing the 15 different risk score total groups revealed a model predictive accuracy of 87.3%. Significant escalations in 3-year mortality were noted to occur at scores of 6 and greater. Mean AAA diameter was significantly larger for patients who had higher risk scores (P < .001).

CONCLUSIONS

A novel risk score for mortality within 3 years of fenestrated visceral segment aortic endograft has been developed that has excellent accuracy in predicting which patients will survive and derive the strongest benefit from intervention. This facilitates risk-benefit analysis and counseling of patients and families with realistic long-term expectations. This potentially enhances patient-centered decision-making.

Thoracic endovascular aortic repair of metachronous thoracic aortic aneurysms following prior infrarenal abdominal aortic aneurysm repair

OBJECTIVE

Thoracic endovascular aortic repair (TEVAR) of metachronous thoracic aortic aneurysms (M-TAAs) following previous infrarenal abdominal aortic aneurysm (AAA) repair has been associated with higher spinal cord ischemia (SCI) risk compared with TEVAR of primary thoracic aortic aneurysms (TAAs). However, data on the impact of the type of prior infrarenal aortic repair on outcomes are scarce. In this study, we examined perioperative outcomes and long-term mortality following TEVAR M-TAA compared with primary TEVAR of TAA.

METHODS

We identified all Vascular Quality Initiative (VQI) patients who underwent TEVAR of TAA in the descending thoracic aorta from 2013 to 2022. Only patients undergoing primary TEVAR or TEVAR following infrarenal open (OAR) or endovascular (EVAR) repair were included. We performed univariate analyses to identify differences in baseline and procedural characteristics, and multivariable analyses for perioperative outcomes and 5-year mortality using logistic and Cox regression, respectively.

RESULTS

We included 1493 patients who underwent primary TEVAR (81%) or TEVAR following prior OAR (9.0%) or prior EVAR (9.7%). Compared with primary TEVAR, patients undergoing TEVAR M-TAA were older, more commonly male, white, and had higher rates of hypertension, smoking, and renal dysfunction. Patients with M-TAA were more likely to be asymptomatic and have larger diameters at presentation but were exposed to greater contrast volume and procedural times relative to primary TEVAR patients. Following risk-adjustment, compared with primary TEVAR, TEVAR after prior EVAR was associated with higher perioperative mortality (9.7% vs 3.9%; odds ratio [OR], 5.3; 95% confidence interval [CI], 2.3-12; P < .001) and 5-year mortality (40% vs 24%; hazard ratio [HR], 2.1; 95% CI, 1.4-3.1; P = .001). Specifically, among octogenarians (n = 375; 25%), the perioperative and 5-year mortality differences were even more pronounced (perioperative mortality: 17% vs 8.4%; OR, 6.7; 95% CI, 2.2-21; P = .001; 5-year mortality: 50% vs 27%; HR, 3.0; 95% CI, 1.5-5.7; P = .010). However, in-hospital complications, including SCI (2.6% vs 2.8%; OR, 1.2; 95% CI, 0.33-3.3; P = .77), were not notably different. In contrast, TEVAR after previous OAR was associated with comparable perioperative mortality (4.4% vs 3.9%; OR, 1.2; 95% CI, 0.32-3.8; P = .73), 5-year mortality (28% vs 24%; HR, 1.3; 95% CI, 0.80-2.1; P = .54), and in-hospital complications, including SCI (2.6% vs 0.7%; OR, 0.21; 95% CI, 0.01-1.1; P = .16).

CONCLUSIONS

Patients undergoing TEVAR of M-TAAs after prior EVAR, particularly octogenarians, have higher perioperative and 5-year mortality and therefore, represent a high-risk group. Future efforts should strive to discern the underlying factors leading to these poorer outcomes; meanwhile, these findings emphasize the need for careful patient selection and appropriate preoperative counseling in these high-risk individuals.

Comparative Analysis of Morbidity and Mortality Outcomes in Elderly and Nonelderly Patients Undergoing Elective TEVAR: A Systematic Review and Meta-Analysis

OBJECTIVE

Due to an ever-increasing ageing population and limited available data around the use of thoracic endovascular aortic repair (TEVAR) in elderly patients, investigating its efficacy and safety in this age cohort is of vital importance. We thus reviewed the existing literature on this topic to assess the feasibility of TEVAR in elderly patients with severe thoracic aortic pathologies.

METHODS

We identified all original research studies that assessed TEVAR in elderly patients published up to 2023. Morbidity, as assessed by neurological and respiratory complications, endoleaks, and length of stay, was the primary endpoint. Short-term mortality and long-term survival were the secondary endpoints. The Mantel-Haenszel random and fixed effects methods were used to calculate the odds ratios for each outcome. Further sensitivity and subgroup analyses were performed to validate the outcomes.

RESULTS

Twelve original studies that evaluated elective TEVAR outcomes in elderly patients were identified. Seven studies directly compared the use of TEVAR between an older and a younger patient group. Apart from a shorter hospital stay in older patients, no statistically significant difference between the morbidity outcomes of the two different cohorts was found. Short-term mortality and long-term survival results favoured the younger population.

CONCLUSIONS

The present meta-analysis indicates that, due to a safe perioperative morbidity profile, TEVAR should not be contraindicated in patients based purely on old age. Further research using large patient registries to validate our findings in elderly patients with specific aortic pathologies and both elective and emergency procedures is necessary.

Outcomes and survival following thoracic endovascular repair in patients with aortic aneurysms limited to the descending thoracic aorta

BACKGROUND

Thoracic endovascular aortic repair (TEVAR) is a well-established therapy for descending aortic aneurysms (DTA). There is a paucity of large series reporting the mid- and long-term outcomes from this era. The main aim of this study was to evaluate the outcomes of TEVAR with regards to the effect of aortic morphology and procedure-related variables on survival, reintervention and freedom from endoleaks.

METHODS

In this retrospective single center study, we evaluated the clinical outcomes among 158 consecutive patients with DTA than underwent TEVAR between 2006 and 2019 at our center. The cohort included 51% patients with device landing zones proximal to the subclavian artery and 25.9% patients undergoing an emergent or urgent TEVAR. The primary outcome was survival, and secondary outcomes were reintervention and occurrence of endoleaks.

RESULTS

Median follow-up was 33 months [IQR 12 to 70] while 50 patients (30.6%) had longer than 5-year follow-up. With a median patient age of 74 years, post-operative Kaplan Meyer survival estimates were 94.3% (95%CI 90.8-98.0, SE 0.018%) at 30 days, 76.4% (95%CI 70.0-83.3, SE 0.034%) at one year and, 52.9% (95%CI 45.0-62.2, SE 0.043%) at five years. Freedom from reintervention at 30 days, one year, and five years was 92.9% (95%CI 89.0-97.1, SE 0.021%), 80.0% (95%CI 72.6-88.1, SE 0.039%), and 52.8% (95%CI 41.4-67.4, SE 0.065%), respectively. On cox regression analysis greater aneurysm diameter, and the use of device landing zones in aortic regions 0-1 were associated with an increased probability of all-cause mortality, and with reintervention during follow-up. Independent of aneurysm size undergoing urgent or emergent TEVAR was associated with higher mortality risk for the first three years post-operative but not on long-term follow-up.

CONCLUSIONS

Larger aneurysms and those requiring stent-graft landing in aortic zones 0 or 1, are associated with higher risk for mortality and reintervention. There remains a need to optimize clinical management and device design for larger proximal aneurysms.

Anemia is associated with higher mortality and morbidity after thoracic endovascular aortic repair

BACKGROUND

It is uncertain whether preoperative anemia is independently associated with thoracic endovascular aortic repair (TEVAR) outcomes. Using a national vascular surgery database, we evaluated the associations between preoperative anemia and 30-day mortality, postoperative complications, and 1-year survival for patients undergoing TEVAR.

METHODS

We retrospectively analyzed all patients in the Vascular Quality Initiative who had undergone TEVAR for aortic dissection, aortic aneurysm, penetrating aortic ulcer, hematoma, or thrombus between January 2011 and December 2019. We excluded patients with a ruptured aneurysm, traumatic dissection, emergent repair, treated aorta distal to zone 5, polycythemia, transfusion of >4 U of packed red blood cells intraoperatively or postoperatively, and missing data on hemoglobin level or surgical indications. The final study cohort was dichotomized into two groups: normal/mild anemia (women, ≥10 g/dL; men, ≥12 g/dL) and moderate/severe anemia (women, <10 g/dL; male, <12 g/dL). Propensity scores by stratification were used to control for confounding in the analysis of the association between the outcomes of 30-day mortality, postoperative complications, and 1-year survival and a binary indicator variable of moderate/severe anemia vs normal/mild anemia. Kaplan-Meier analysis and log-rank tests were used to compare the 1-year survival between the two groups. A Cox regression model was fitted to assess the associations between anemia and survival outcomes.

RESULTS

A total of 3391 patients were analyzed, 958 (28.3%) of whom had had moderate/severe anemia. After adjustment for multiple clinical factors using propensity score stratification, moderate/severe anemia was associated with a 141% increased odds of 30-day mortality (adjusted odds ratio [aOR], 2.41; 95% confidence interval [CI], 1.15-5.05; P = .019), 58% increased odds of any in-hospital complication (aOR, 1.58; 95% CI, 1.17-2.13; P = .003), 281% increased odds of intraoperative transfusion (aOR, 3.81; 95% CI, 2.68-5.53; P < .001). In addition, moderate/severe anemia was associated with significantly worse survival within the first year after TEVAR (log-rank P < .001; 1-year survival rate using Kaplan-Meier estimates, 86.4% ± 1.3% standard error vs 92.5% ± 0.6% standard error) and with an increased risk of mortality in the first postoperative year (adjusted hazard ratio, 1.81; 95% CI, 1.16-2.82; P = .009).

CONCLUSIONS

We found that moderate or severe anemia is associated with significantly increased odds of mortality, postoperative complications, and worse 1-year survival after TEVAR. Future studies are needed to evaluate the effect of anemia correction on the outcomes of TEVAR.

A systematic review and meta-analysis of thoracic endovascular aortic repair with the proximal landing zone 0

Background

Thoracic endovascular aortic repair, initially intended for thoracic aortic disease treatment, has extended its application to the proximal zone of the aorta. However, the safety and surgical outcomes of extending the proximal landing zone into the ascending aorta (zone 0) in selected cases remain unknown. Thus, we performed a systematic review and meta-analysis of zone 0 thoracic endovascular aortic repair (TEVAR) to obtain a deeper understanding of its safety, outcomes, and trends over time.

Methods

A literature search was performed using PubMed, EMBASE, and Web of Science databases in accordance with the preferred reporting items for systematic reviews and meta-analyses guidelines, from January, 1997 to January, 2022. Only studies involving zone 0 TEVAR were included. The retrieved data from the eligible studies included basic study characteristics, 30-day/in-hospital mortality rate, indications, comorbidities, stent grafts, techniques, and complications. Summary effect measures of the primary outcomes were obtained by logarithmically pooling the data with an inverse variance-weighted fixed-effects model.

Results

Fifty-three studies with 1,013 patients were eligible for analysis. The pooled 30-day/in-hospital mortality rate of zone 0 TEVAR was 7.49%. The rates of post-operative stroke, type Ia endoleak, retrograde type A aortic dissection, and spinal cord ischemia were 8.95, 9.01, 5.72, and 4.12%, respectively.

Conclusions

Although many novel stent grafts and techniques targeting zone 0 TEVAR are being investigated, a consensus on technique and device selection in zone 0 TEVAR is yet to be established in current practice. Furthermore, the post-operative stroke rate is relatively high, while other complication rates and perioperative death rate are comparable to those of TEVAR for other aortic zones.