Predicting Outcomes Following Lower Extremity Endovascular Revascularization Using Machine Learning

BACKGROUND

Lower extremity endovascular revascularization for peripheral artery disease carries nonnegligible perioperative risks; however, outcome prediction tools remain limited. Using machine learning, we developed automated algorithms that predict 30-day outcomes following lower extremity endovascular revascularization.

METHODS AND RESULTS

The National Surgical Quality Improvement Program targeted vascular database was used to identify patients who underwent lower extremity endovascular revascularization (angioplasty, stent, or atherectomy) for peripheral artery disease between 2011 and 2021. Input features included 38 preoperative demographic/clinical variables. The primary outcome was 30-day postprocedural major adverse limb event (composite of major reintervention, untreated loss of patency, or major amputation) or death. Data were split into training (70%) and test (30%) sets. Using 10-fold cross-validation, 6 machine learning models were trained using preoperative features. The primary model evaluation metric was area under the receiver operating characteristic curve. Overall, 21 886 patients were included, and 30-day major adverse limb event/death occurred in 1964 (9.0%) individuals. The best performing model for predicting 30-day major adverse limb event/death was extreme gradient boosting, achieving an area under the receiver operating characteristic curve of 0.93 (95% CI, 0.92-0.94). In comparison, logistic regression had an area under the receiver operating characteristic curve of 0.72 (95% CI, 0.70-0.74). The calibration plot showed good agreement between predicted and observed event probabilities with a Brier score of 0.09. The top 3 predictive features in our algorithm were (1) chronic limb-threatening ischemia, (2) tibial intervention, and (3) congestive heart failure.

CONCLUSIONS

Our machine learning models accurately predict 30-day outcomes following lower extremity endovascular revascularization using preoperative data with good discrimination and calibration. Prospective validation is warranted to assess for generalizability and external validity.

Machine Learning to Predict Outcomes of Endovascular Intervention for Patients With PAD

Importance

Endovascular intervention for peripheral artery disease (PAD) carries nonnegligible perioperative risks; however, outcome prediction tools are limited.

Objective

To develop machine learning (ML) algorithms that can predict outcomes following endovascular intervention for PAD.

Design, Setting, and Participants

This prognostic study included patients who underwent endovascular intervention for PAD between January 1, 2004, and July 5, 2023, with 1 year of follow-up. Data were obtained from the Vascular Quality Initiative (VQI), a multicenter registry containing data from vascular surgeons and interventionalists at more than 1000 academic and community hospitals. From an initial cohort of 262 242 patients, 26 565 were excluded due to treatment for acute limb ischemia (n = 14 642) or aneurysmal disease (n = 3456), unreported symptom status (n = 4401) or procedure type (n = 2319), or concurrent bypass (n = 1747). Data were split into training (70%) and test (30%) sets.

Exposures

A total of 112 predictive features (75 preoperative [demographic and clinical], 24 intraoperative [procedural], and 13 postoperative [in-hospital course and complications]) from the index hospitalization were identified.

Main Outcomes and Measures

Using 10-fold cross-validation, 6 ML models were trained using preoperative features to predict 1-year major adverse limb event (MALE; composite of thrombectomy or thrombolysis, surgical reintervention, or major amputation) or death. The primary model evaluation metric was area under the receiver operating characteristic curve (AUROC). After selecting the best performing algorithm, additional models were built using intraoperative and postoperative data.

Results

Overall, 235 677 patients who underwent endovascular intervention for PAD were included (mean [SD] age, 68.4 [11.1] years; 94 979 [40.3%] female) and 71 683 (30.4%) developed 1-year MALE or death. The best preoperative prediction model was extreme gradient boosting (XGBoost), achieving the following performance metrics: AUROC, 0.94 (95% CI, 0.93-0.95); accuracy, 0.86 (95% CI, 0.85-0.87); sensitivity, 0.87; specificity, 0.85; positive predictive value, 0.85; and negative predictive value, 0.87. In comparison, logistic regression had an AUROC of 0.67 (95% CI, 0.65-0.69). The XGBoost model maintained excellent performance at the intraoperative and postoperative stages, with AUROCs of 0.94 (95% CI, 0.93-0.95) and 0.98 (95% CI, 0.97-0.99), respectively.

Conclusions and Relevance

In this prognostic study, ML models were developed that accurately predicted outcomes following endovascular intervention for PAD, which performed better than logistic regression. These algorithms have potential for important utility in guiding perioperative risk-mitigation strategies to prevent adverse outcomes following endovascular intervention for PAD.

ERICVA Risk Scale simplified as a predictor of amputation in critical limb ischemia

OBJECTIVE

Demonstrate that simplified ERICVA (Valladolid Critical Limb Ischaemia Risk Scale) is useful in predicting amputation in chronic limb-threatening ischemia (CLI) after one year of revascularization.

METHODS

A retrospective cohort study was performed. We analyzed the medical records of 93 patients over the age of 35 with the diagnosis of CLI who were treated in the Department of Internal Medicine, Orthopedics or in the Cardiovascular Surgery Unit of the Víctor Lazarte Echegaray Hospital and the High Complexity Virgen de La Puerta Hospital during the period 2015-2018. The simplified ERICVA score was determined in patients before surgical and endovascular revascularization. We included 31 patients who scored 2 or more points in the exposed group and 62 patients who scored less than 2 points in the group not exposed to amputation risk. The collected data was analyzed with the statistical program SPSS where the Relative Risk and significance was obtained with Pearson's Chi-square. The multivariate analysis was also carried out in order to obtain the adjusted relative risk.

RESULTS

It was identified that the simplified ERICVA score greater than or equal to 2 points was more frequent in those who underwent amputation (90.3%) compared to patients who did not undergo amputation (4.8%), increasing the risk of amputation in those patients with CLI who underwent revascularization (RR: 18.67, P<0.001). It was also possible to identify that within the group of patients that showed a high risk of amputation according to the ERICVA scale, they had a higher risk of major amputation (RR: 9.32, P<0.001) as opposed to the risk of minor amputation (RR: 1, 89, P=0.193). Among the items of the simplified ERICVA scale, the preoperative neutrophil-lymphocyte ratio and hematocrit were significantly higher in the group of amputated patients (P<0.001). In addition, it was possible to identify that the score greater than or equal to 2 was independently associated with the risk of amputation in patients revascularized with CLI (RR: 13.5, P<0.001).

CONCLUSION

In our patient population, the simplified ERICVA scale is useful in predicting major and minor amputation in critical limb ischemia after revascularization. The present data showed that the patients who had a simplified ERICVA score greater than or equal to 2 had a higher risk of major amputation compared to the risk of minor amputation. However, it is important to highlight that the impact on the prediction of minor amputation is greater because in some circumstances major amputation can appear as a complication of CLI.

The comparative and added prognostic value of biomarkers to the Revised Cardiac Risk Index for preoperative prediction of major adverse cardiac events and all-cause mortality in patients who undergo noncardiac surgery

BACKGROUND

The Revised Cardiac Risk Index (RCRI) is a widely acknowledged prognostic model to estimate preoperatively the probability of developing in-hospital major adverse cardiac events (MACE) in patients undergoing noncardiac surgery. However, the RCRI does not always make accurate predictions, so various studies have investigated whether biomarkers added to or compared with the RCRI could improve this.

OBJECTIVES

Primary: To investigate the added predictive value of biomarkers to the RCRI to preoperatively predict in-hospital MACE and other adverse outcomes in patients undergoing noncardiac surgery. Secondary: To investigate the prognostic value of biomarkers compared to the RCRI to preoperatively predict in-hospital MACE and other adverse outcomes in patients undergoing noncardiac surgery. Tertiary: To investigate the prognostic value of other prediction models compared to the RCRI to preoperatively predict in-hospital MACE and other adverse outcomes in patients undergoing noncardiac surgery.

SEARCH METHODS

We searched MEDLINE and Embase from 1 January 1999 (the year that the RCRI was published) until 25 June 2020. We also searched ISI Web of Science and SCOPUS for articles referring to the original RCRI development study in that period.

SELECTION CRITERIA

We included studies among adults who underwent noncardiac surgery, reporting on (external) validation of the RCRI and: - the addition of biomarker(s) to the RCRI; or - the comparison of the predictive accuracy of biomarker(s) to the RCRI; or - the comparison of the predictive accuracy of the RCRI to other models. Besides MACE, all other adverse outcomes were considered for inclusion.

DATA COLLECTION AND ANALYSIS

We developed a data extraction form based on the CHARMS checklist. Independent pairs of authors screened references, extracted data and assessed risk of bias and concerns regarding applicability according to PROBAST. For biomarkers and prediction models that were added or compared to the RCRI in ≥ 3 different articles, we described study characteristics and findings in further detail. We did not apply GRADE as no guidance is available for prognostic model reviews.

MAIN RESULTS

We screened 3960 records and included 107 articles.   Over all objectives we rated risk of bias as high in ≥ 1 domain in 90% of included studies, particularly in the analysis domain. Statistical pooling or meta-analysis of reported results was impossible due to heterogeneity in various aspects: outcomes used, scale by which the biomarker was added/compared to the RCRI, prediction horizons and studied populations.  Added predictive value of biomarkers to the RCRI Fifty-one studies reported on the added value of biomarkers to the RCRI. Sixty-nine different predictors were identified derived from blood (29%), imaging (33%) or other sources (38%). Addition of NT-proBNP, troponin or their combination improved the RCRI for predicting MACE (median delta c-statistics: 0.08, 0.14 and 0.12 for NT-proBNP, troponin and their combination, respectively). The median total net reclassification index (NRI) was 0.16 and 0.74 after addition of troponin and NT-proBNP to the RCRI, respectively. Calibration was not reported. To predict myocardial infarction, the median delta c-statistic when NT-proBNP was added to the RCRI was 0.09, and 0.06 for prediction of all-cause mortality and MACE combined. For BNP and copeptin, data were not sufficient to provide results on their added predictive performance, for any of the outcomes. Comparison of the predictive value of biomarkers to the RCRI  Fifty-one studies assessed the predictive performance of biomarkers alone compared to the RCRI. We identified 60 unique predictors derived from blood (38%), imaging (30%) or other sources, such as the American Society of Anesthesiologists (ASA) classification (32%). Predictions were similar between the ASA classification and the RCRI for all studied outcomes. In studies different from those identified in objective 1, the median delta c-statistic was 0.15 and 0.12 in favour of  BNP and NT-proBNP alone, respectively, when compared to the RCRI, for the prediction of MACE. For C-reactive protein, the predictive performance was similar to the RCRI. For other biomarkers and outcomes, data were insufficient to provide summary results. One study reported on calibration and none on reclassification. Comparison of the predictive value of other prognostic models to the RCRI   Fifty-two articles compared the predictive ability of the RCRI to other prognostic models. Of these, 42% developed a new prediction model, 22% updated the RCRI, or another prediction model, and 37% validated an existing prediction model. None of the other prediction models showed better performance in predicting MACE than the RCRI. To predict myocardial infarction and cardiac arrest, ACS-NSQIP-MICA had a higher median delta c-statistic of 0.11 compared to the RCRI. To predict all-cause mortality, the median delta c-statistic was 0.15 higher in favour of ACS-NSQIP-SRS compared to the RCRI. Predictive performance was not better for CHADS₂, CHA₂DS₂-VASc, R₂CHADS₂, Goldman index, Detsky index or VSG-CRI compared to the RCRI for any of the outcomes. Calibration and reclassification were reported in only one and three studies, respectively.

AUTHORS' CONCLUSIONS

Studies included in this review suggest that the predictive performance of the RCRI in predicting MACE is improved when NT-proBNP, troponin or their combination are added. Other studies indicate that BNP and NT-proBNP, when used in isolation, may even have a higher discriminative performance than the RCRI. There was insufficient evidence of a difference between the predictive accuracy of the RCRI and other prediction models in predicting MACE. However, ACS-NSQIP-MICA and ACS-NSQIP-SRS outperformed the RCRI in predicting myocardial infarction and cardiac arrest combined, and all-cause mortality, respectively. Nevertheless, the results cannot be interpreted as conclusive due to high risks of bias in a majority of papers, and pooling was impossible due to heterogeneity in outcomes, prediction horizons, biomarkers and studied populations. Future research on the added prognostic value of biomarkers to existing prediction models should focus on biomarkers with good predictive accuracy in other settings (e.g. diagnosis of myocardial infarction) and identification of biomarkers from omics data. They should be compared to novel biomarkers with so far insufficient evidence compared to established ones, including NT-proBNP or troponins. Adherence to recent guidance for prediction model studies (e.g. TRIPOD; PROBAST) and use of standardised outcome definitions in primary studies is highly recommended to facilitate systematic review and meta-analyses in the future.

Involvement of microcirculation in critical ischemia: how to identify it?

Critical limb ischemia represents the most severe pattern of peripheral arterial disease (PAD) associated with the high risk of major amputation, cardiovascular events and death. The diagnosis and management of critical limb ischemia (CLI) is often challenging. Systolic ankle and toe pressure measurements are considered to be the basic techniques for the identification of PAD. However, they provide rough insight into the dependent local tissue perfusion. Furthermore, those techniques do not enable investigation of microcirculation which has crucial role in the pathogenesis of CLI. Some patients with mild deterioration of macrocirculation develop CLI if microcirculation is affected. Investigation of perfusion on macro- and local microcirculatory level enables more effective treatment: revascularization of the angiosome-related artery. The technologies capable of assessing limb tissue oxygenation or perfusion on microcirculatory level enable direct assessment of distant tissue oxygenation. Transcutaneous oxygen tension (TcPO2) measurement which was introduced in clinical practice represents one of the objective criteria for the diagnosis of CLI. Main weakness of this technique as well as laser Doppler flow measurement is low penetrance from the skin surface. Measurement of tissue blood flow on microcirculatory level can be performed with indocyanine green fluorescent imaging (ICG), contrast-enhanced magnetic resonance and vital microscopy. ICG is promising method which provides excellent informative image of tissue perfusion. However, it offers little quantitative information. Investigation of microcirculation in patients with CLI is of outmost importance because it enables insight in local tissue perfusion and oxygenation, which represents the basis of identification of most ischemic regions and provide more successful angiosome related revascularization of an affected artery.

High-pressure, non compliant balloon angioplasty for long and calcified infrapopliteal and inframalleolar lesions is feasible

BACKGROUND

To evaluate the safety, feasibility and effectiveness of high-pressure, noncompliant balloon angioplasty in the management of long infrapopliteal calcified lesions.

METHODS

Consecutive patients, presenting with chronic limb-threatening ischemia (CLTI) and long (>100 mm) calcified infrapopliteal lesions who were treated with a high pressure, noncompliant balloon (JADE, OrbusNeich, Hong Kong) between January 2016 and July 2016 were retrospectively analyzed. Angioplasty was performed by inflating the balloon to a pressure of 22 to 24 atm for 90 seconds. Primary outcome was technical success. Secondary outcomes were procedure-related complications, limb salvage, amputation-free survival (AFS), wound healing, overall survival, freedom from clinically driven target lesion reintervention (CD-TLR), and resolution of CLTI at 2 and 3 years.

RESULTS

Overall, 23 lesions in 21 limbs of 20 patients were treated. All patients had tissue loss (Rutherford 5 or 6). The mean lesion length was 374.8 mm. Of all lesions, 56.5% were occlusions, 91.3% were classified as TransAtlantic Inter-Society Consensus (TASC) C and D lesions, and 78.3% had severe calcification classification. Of all lesions, 52.2% extended into the below-the-ankle arteries. Technical success was achieved in 22 lesions (95.7%). There were no procedure-related complications. No bailout stenting was required. At 2 and 3 years, limb salvage was 84.7% and 78.7%, AFS was 71.4% and 56.1%, wound healing was 81.0% and 85.7%, overall survival was 75.0% and 64.3% and freedom from CD-TLR was 77.6% and 63.5%, respectively. Resolution of CLTI without TLR was 81.0% at 2 and 3 years.

CONCLUSIONS

This study is the first to analyze safety and feasibility of a high-pressure, noncompliant balloon for long, calcified infrapopliteal and inframalleolar lesions.

Nutritional status and outcomes of superficial femoral artery stenting due to intermittent claudication

BACKGROUND

The pathogenesis of in-stent restenosis is still not clear. The aim of this study was to determine the nutritional status of patients with lower limb ischemia and the risk of target lesion revascularization (TLR) after superficial femoral artery (SFA) stenting.

METHODS

Numerous parameters of nutritional status assessment were compared between 70 patients undergoing SFA endovascular intervention with a self-expandable plane stent due to life-limiting intermittent claudication and 40 patients undergoing carotid artery stenting (CAS). All subjects were followed up for at least 1 year in relation to outcomes such as clinically driven TLR occurrence.

RESULTS

Patients undergoing SFA stenting had a lower prevalence of overweight and obesity than those who underwent CAS (51.43% vs. 72.50%; P=0.031). An increase in Ankle-Brachial Index of >0.15 after SFA stenting (early end-point) was positively associated with greater handgrip strength (HGS), fat-free mass, skeletal muscle mass and waist-to-hip ratio. Freedom from TLR (late end-point) was significantly related to a higher waist-to-height ratio (WHtR), HGS and Geriatric Nutritional Risk Index (GNRI) Score. The 1-year risk of TLR for patients with a WHtR of ≥61.39 amounted to odds ratio; 95% confidence interval: 0.21; 0.05-0.25; P=0.021.

CONCLUSIONS

Parameters of nutritional status assessment were associated with early and late outcomes of SFA stenting in patients with intermittent claudication. Abdominal fat distribution and higher HGS and GNRI scores lowered the 1-year risk of TLR. Further study is needed to determine the pathomechanism of the obesity paradox, sarcopenia and undernutrition in relation to outcomes of endovascular interventions.

Comparison of patency between two different peripheral self-expandable stents, absolute Pro® versus complete SE® in femoropopliteal occlusive disease

BACKGROUND

Self-expandable stents, Absolute Pro® (ABS) and Complete SE® (COM), demonstrated safety and efficacy in previous studies. We aimed to determine which stent is more effective for the endovascular treatment of femoropopliteal artery disease.

METHODS

We enrolled patients who underwent endovascular intervention from 2010 to 2015 in our hospital. The primary endpoint was clinical primary patency, which is a composite of freedom from restenosis or clinically driven target lesion revascularization (TLR).

RESULTS

A total of 197 patients and 210 limbs were analyzed, with 96 limbs of 88 patients assigned to the ABS group and 114 limbs of 109 patients assigned to the COM group. Baseline and lesion characteristics were similar between the two groups. The number of stents per limbs were 1.28±0.55 and 1.29±0.51 in the ABS and COM groups, respectively (P=0.92). The postprocedure ankle-brachial index was significantly improved in both groups compared with the preprocedural one (P<0.01), but there were no differences between the both groups at 6, 12, and 24 months after the index procedure. There were no significant differences in clinical primary patency rate (68.7% in ABS vs. 66.7% in COM, P=0.68) and TLR (9.4% in ABS vs. 14.0% in COM, P=0.41) between the two groups. There was no interaction of the clinical primary patency rate of the two stents and the patients' characteristics or lesion characteristics.

CONCLUSIONS

In this retrospective single-center study, ABS and COM showed no difference in clinical efficacy. Both stents can be effectively used for the endovascular intervention of femoropopliteal artery disease when it is necessary.

PLX-PAD Cell Treatment of Critical Limb Ischaemia: Rationale and Design of the PACE Trial

BACKGROUND

Critical limb ischaemia (CLI) is a life threatening condition with a considerable risk of major amputation and death. Besides revascularisation, no treatment has been proven to reduce the risks. Therapeutic angiogenesis by gene or cell therapy has not demonstrated definitive evidence in randomised controlled trials. PLX-PAD is an "off the shelf" allogeneic placental derived, mesenchymal like cell therapy, which, in preclinical studies, has shown pro-angiogenic, anti-inflammatory, and regenerative properties. Favourable one year amputation free survival (AFS), and trends in reduction of pain scores and increase of tissue perfusion have been shown in two small, open label, phase I trials.

METHODS

The PACE study is a phase III randomised, double blind, multicentre, multinational placebo controlled, parallel group study to evaluate the efficacy, tolerability, and safety of intramuscular injections of PLX-PAD cells to treat patients with atherosclerotic CLI with minor tissue loss (Rutherford Category 5) up to the ankle level, who are unsuitable for revascularisation or carry an unfavourable risk benefit for that treatment. The study will enroll 246 patients, who after screening are randomised in a ratio of 2:1 to treatment with intramuscular injections of PLX-PAD 300 × 10⁶ cells or placebo on two occasions, eight weeks apart. The primary efficacy endpoint is time to major amputation or death (amputation free survival), which will be assessed in follow up of at least 12 months and up to 36 months.

CONCLUSIONS

Based on favourable pre-clinical and initial clinical study results, the PACE phase III randomised controlled trial will evaluate placenta derived PLX-PAD cell treatment in patients with critical limb ischaemia, with an unfavourable risk benefit for revascularisation. Clinicaltrials.gov: NCT03006770.