Gloviczki, Péter

Background: Objective performance assessment is essential in surgical skill training, yet current methods are labor-intensive and focus on observing the trainee rather than the end-product of the procedure. Machine learning (ML) methods offer reproducible feedback but have mainly relied on kinematic or video data, often reducing assessment to binary or ternary classification. Our objective was to compare ML regression models predicting expert-assigned scores of vascular anastomoses from computational fluid dynamics (CFD) features of the final product. Additionally, we aimed to assess biomechanical plausibility of predictions. Methods: A total of 146 participants performed 419 end-to-side anastomoses on case-specific three-dimensional (3D) printed simulators. Anastomoses were digitized via 3D scanning, ranked by experts, and characterized using CFD-derived hemodynamic features. These served as input for linear models (Ridge, Partial Least Squares), support vector machines, and tree-based ensembles (Random Forest, Extremely Randomized Trees, and Extreme Gradient Boosting [XGBoost]), evaluated using 10-fold nested cross-validation with genetic hyperparameter optimization. Results: Inter-rater reliability of expert indicated strong agreement (intraclass correlation coefficient ICC3k = 0.846). XGBoost achieved the lowest mean root mean squared error of 0.758 (95% bootstrap CI: 0.722–0.799) and a coefficient of determination (R2) of 0.673 (0.617–0.725), with the most stable performance across folds. Shapley additive explanations (SHAP) identified the wall shear stress gradient, transverse wall shear stress, and maximum pressure as the most influential features—variables associated with intimal hyperplasia and atherosclerotic remodeling. Conclusions: Tree-based ensemble methods, particularly XGBoost, effectively modeled biomechanical properties against expert scores. Combining CFD and ML can provide reproducible, mechanistically relevant feedback in vascular surgical skill training.

Experiments were designed to compare perioperative blood loss, early thrombogenicity and morphologic and functional characteristics of the neointima of three types of prosthetic materials used for carotid artery patch angioplasty. Bilateral carotid patch angioplasties were performed in 20 dogs, using 20 gelatin-impregnated fluoropassivated Dacron (GIF), 10 untreated knitted Dacron and 10 expanded polytetrafluoroethylene (PTFE) patches (5 cm2). Intraoperative blood loss, platelet deposition at 24 h and neointimal morphology at 6 weeks after the operation were assessed. Bioassay of the neointima was performed at 6 weeks in 16 patches. Mean (s.e.m.) blood loss was significantly less in GIF patches (14.7 (2.7) ml) compared with either PTFE (75.6 (24) ml) or untreated Dacron (64. 3 (9.5)) ( P < 0.005). Mean (s.e.m.) platelet deposition in GIF patches (1380 (328) platelets/cm2) was approximately 50% less at 24 h than in untreated Dacron (2562 (1035) platelets/cm2) or PTFE (2140 (998) platelets/cm2) patches ( P < 0.05). Neointimal coverage was greater in PTFE (94 (1.3%)) compared with GIF (79 (2.7%)) or untreated Dacron (86(2.4%)) patches ( P < 0.05). The thickness of the neointimal layer of PTFE (0.5 (0.01) mm) patches was greater than other patch types; GIF (0.2 (0.01) mm) or untreated Dacron (0.3 (0.01) mm) ( P < 0.05). Under bioassay conditions, acetylcholine caused release of vasoactive relaxing factor(s) from all patches. However, relaxations from baseline were less with GIF patches (−37.9 (11.7)% versus −54.5 (9.6) for untreated Dacron; −50.2 (15.2)% for PTFE) ( P = n.s.). Endothelin-1 release occurred from all patches and was increased with the extent of neointimal coverage. These data demonstrate that GIF patches caused the least perioperative bleeding, were the least thrombogenic at 24 h and developed the thinnest neointima at 6 weeks. All patch materials developed a functioning neointima.

Increased shear, pressure, and oxygen tension in vein grafts may alter production of endothelium-derived vasoactive and anti-mitogenic factors such as nitric oxide which subsequently affect development of neointimal hyperplasia. This study was designed to determine whether or not nitric oxide mediates endothelium-dependent responses in femoral in situ vein grafts. Non-reversed, canine femoral vein grafts were placed bilaterally to bypass a ligated segment of the femoral artery in dogs. After 6 weeks, the grafts were removed, cut into rings, and suspended in organ chambers for measurement of isometric force. In some rings the endothelium was removed deliberately. In the presence of indomethacin, the synthetic analog of L-arginine, L-NG-monomethyl-arginine (10−4 M: L-NMMA) did not cause a significant change in baseline tension of rings with endothelium. L-NMMA reduced only contractions of rings with endothelium to the alpha-adrenergic agonist UK 14,304. The analog also reduced maximal relaxations to the calcium-ionophore, A23187 in rings with endothelium. In addition, L-NMMA reduced relaxations of rings without endothelium to adenosine diphosphate by 35%. Positive immunostaining for nitric oxide synthase was present in both the myointima and media of histological sections of grafts. In conclusion, these results suggest that in situ vein grafts exhibit two unique properties which are unlike unoperated arteries or veins: (i) alpha2-adrenergic receptors may be coupled to the release of contractile endothelium-derived factors associated with production of nitric oxide; and (ii) nitric oxide may be released by the smooth muscle in response to purinergic stimulation. The presence of nitric oxide synthase throughout the wall of the graft may result in production of nitric oxide in response to adenosine diphosphate released by platelets and to circulating catecholamines.

Renal fusion or ectopia can present formidable challenges during aortic surgery. To evaluate morbidity and define optimal management, the clinical histories of 120 patients with renal fusion or ectopia who underwent 21 aortic procedures at the authors' institution over a 37-year period were reviewed. Indications for surgery included aortic aneurysm in 16 patients (infrarenal in 15 and thoracoabdominal in one) and aortoiliac occlusive disease in five (with renovascular hypertension in two). The abnormal kidney was detected before surgery in 13 patients (65%) by excretory urography, arteriography, computed tomography, or ultrasonography. Arteriography revealed multiple and/or anomalous renal arteries in nine of 12 patients studied. At surgery, 15 patients (75%) were found to have multiple or anomalous renal arteries. Six required renal revascularization (reimplantation four, endarterectomy one, aortorenal bypass one). The renal symphysis was divided in two patients. There were no operative deaths. Six major complications included bleeding requiring reoperation, renal failure requiring short-term dialysis, pancreatitis, gastrointestinal bleeding, pneumonia and thrombophlebitis. Preoperative aortography is recommended in patients with renal fusion or ectopia because of the high incidence of associated renal artery anomalies. The surgeon must be prepared to preserve or revascularize these anomalous renal arteries. Division of the renal symphysis is rarely required. Although perioperative morbidity is raised, aortic reconstruction in patients with renal fusion or ectopia can be safely performed without increased mortality.

Acute iliofemoral deep vein thrombosis and chronic iliofemoral venous obstruction cause substantial patient harm and are increasingly managed with endovascular venous interventions, including percutaneous mechanical thrombectomy and stent placement. However, studies of these treatment elements have not been designed and reported with sufficient rigor to support confident conclusions about their clinical utility. In this project, the Trustworthy consensus-based statement approach was utilized to develop, via a structured process, consensus-based statements to guide future investigators of venous interventions. Thirty statements were drafted to encompass major topics relevant to venous study description and design, safety outcome assessment, efficacy outcome assessment, and topics specific to evaluating percutaneous venous thrombectomy and stent placement. Using modified Delphi techniques for consensus achievement, a panel of physician experts in vascular disease voted on the statements and succeeded in reaching the predefined threshold of >80% consensus (agreement or strong agreement) on all 30 statements. It is hoped that the guidance from these statements will improve standardization, objectivity, and patient-centered relevance in the reporting of clinical outcomes of endovascular interventions for acute iliofemoral deep venous thrombosis and chronic iliofemoral venous obstruction in clinical studies and thereby enhance venous patient care.

Background: Hybrid debranching repair of pararenal and thoracoabdominal aortic aneurysms was initially designed as a better alternative to standard open repair, addressing the limitations of endovascular repair involving the visceral aorta. We reviewed the collective outcomes of hybrid debranching repairs using extra-anatomic, open surgical debranching of the renal-mesenteric arteries, followed by endovascular aortic stenting. Methods: Data from patients who underwent hybrid repair in 14 North American institutions during 10 years were retrospectively reviewed. Society of Vascular Surgery scores were used to assess comorbidity risk. Early and late outcomes, including mortality, morbidity, reintervention, and patency were analyzed. Results: A total of 208 patients (118 male; mean age, 71±8 years old) were treated by hybrid repair with extraanatomic reconstruction of 657 renal and mesenteric arteries (mean 3.2 vessels/patient). Mean aneurysm diameter was 6.6±1.3 cm. Thoracoabdominal aortic aneurysms were identified in 163 (78%) patients and pararenal aneurysms in 45 (22%). A single-stage repair was performed in 92 (44%) patients. The iliac arteries were the most common source of inflow (n=132; 63%), and most (n=150; 72%) had 3 or more bypasses. There were 30 (14%) early deaths, ranging widely across sites (0%–21%). A Society of Vascular Surgery comorbidity score >15 was the primary predictor of early mortality ( P <0.01), whereas mortality was 3% in a score ≤9. Early complications occurred in 140 (73%) patients and included respiratory complications in 45 patients (22%) and spinal cord ischemia in 22 (11%), of whom 10 (45%) fully recovered. At 5 years, survival was 61±5%, primary graft patency was 90±2%, and secondary patency was 93±2%. The most significant predictor of late mortality was renal insufficiency ( P <0.0001). Conclusions: Mortality after hybrid repair and visceral debranching is highly variable by center, but strongly affected by preoperative comorbidities and the centers’ experience with the technique. With excellent graft patency at 5 years, the outcomes of hybrid repair done at centers of excellence and in carefully selected patients may be comparable (or better) than traditional open or even totally endovascular approaches. However, in patients already considered as high-risk for surgery, it may not offer better outcomes.

Background: Repair of thoracoabdominal aortic aneurysms (TAAAs) continues to be a challenging task. Hemorrhagic shock, cardiac arrest and multisystem organ failures are the most frequent causes of death, and paraplegia and renal failure are the most devastating complications. Methods: Flawless surgical technique and the use of adjuncts to protect key organs including the brain, heart, spinal cord, liver and kidneys affect outcome. Perfection in exposure and suturing technique decreases bleeding complications, shortens cross-clamp time and assures optimal, visceral, renal and lower extremity perfusion. Technical details include retroperitoneal abdominal aortic exposure, double thoracotomy for Type I and Type II aneurysms, and preservation of the diaphragm. The kidneys are protected by perfusion of iced lactated Ringers; visceral ischemia in Type 1 and Type II, aneurysms is diminished by using pulmonary vein-femoral artery pump with sequential clamping. Spinal cord protection is attempted by spinal fluid pressure monitoring and drainage, moderate general hypothermia, selective left heart bypass, reimplantation of critical intercostal arteries, monitoring somatosensory and somatomotor evoked potentials and epidural cooling of the spinal cord. Results: Outcome in 203 patients (Group I) who underwent repair of TAAAs without epidural cooling was compared with outcome in 97 patients, 27 with thoracic aortic aneurysms and 70 with TAAA (Group II) who underwent repair using epidural cooling. In Group II paraplegia/parapesis occurred in 11.6%. not different from the 8.9% in Group I. Thirty day mortality for elective cases decreased from 14.6% (Group I) to 7.2% (Group II, P<0.05). Conclusions: Open surgical repair of TAAA carries elevated mortality and complication rates. The etiology of ischemic and reperfusion injury to the spinal cord is multifactorial and its prevention remains a formidable and as yet unresolved task. To select patients for surgical repair, the risk of TAAA rupture should be balanced against risks of perioperative mortality, paraplegia and renal failure.

Between 1978 and 1990, 32 patients (15 men and 17 women; mean age 48 (range 15–83) years) underwent 35 operations for renal artery aneurysm (4.2% of 829 renal artery repairs). Eleven patients presented with acute symptoms (nine with hematuria, eight with abdominal pain, two with acute hypertension). Twenty-eight of the 32 patients had chronic hypertension. The diagnosis was confirmed by angiography in all but two. The mean diameter of the renal artery aneurysm was 1.7 (range 0.7–9.0) cm. Seventeen patients had concomitant renal artery stenosis; none of the aneurysms ruptured. Nephrectomy was performed in seven patients and excision of the aneurysm without reconstruction in five. Twenty patients underwent 23 reconstructions using lateral suture (three procedures), vein patch (three), saphenous vein (13), Dacron (three) or composite (vein and hypogastric artery) graft (one). Seven patients underwent ex vivo renal artery repair. There was no perioperative death or secondary nephrectomy. One postoperative graft occlusion was successfully revised. Hypertension improved in 50% of patients. The presence of hypertension, enlargement of a renal artery aneurysm, solitary kidney, bilateral involvement, acute hematuria or potential loss of kidney or renal function may be indications for surgical treatment of an aneurysm > 1.5 cm in diameter. A renal artery aneurysm of any size should be repaired in women who may become pregnant. Where there is branch involvement, ex vivo repair is the procedure of choice for renal salvage.

Despite the publication of several national/international guidelines, the optimal management of patients with asymptomatic carotid stenosis (AsxCS) remains controversial. This article compares 3 recently released guidelines (the 2020 German–Austrian, the 2021 European Stroke Organization [ESO], and the 2021 Society for Vascular Surgery [SVS] guidelines) vs the 2017 European Society for Vascular Surgery (ESVS) guidelines regarding the optimal management of AsxCS patients. The 2017 ESVS guidelines defined specific imaging/clinical parameters that may identify patient subgroups at high future stroke risk and recommended that carotid endarterectomy (CEA) should or carotid artery stenting (CAS) may be considered for these individuals. The 2020 German–Austrian guidelines provided similar recommendations with the 2017 ESVS Guidelines. The 2021 ESO Guidelines also recommended CEA for AsxCS patients at high risk for stroke on best medical treatment (BMT), but recommended against routine use of CAS in these patients. Finally, the SVS guidelines provided a strong recommendation for CEA+BMT vs BMT alone for low-surgical risk patients with >70% AsxCS. Thus, the ESVS, German–Austrian, and ESO guidelines concurred that all AsxCS patients should receive risk factor modification and BMT, but CEA should or CAS may also be considered for certain AsxCS patient subgroups at high risk for future ipsilateral ischemic stroke.