August 2023

This randomized control trial compares early surgical aortic valve replacement (SAVR) to non-operative, conservative management in patients with asymptomatic, severe aortic stenosis. They found that patients assigned to surgical intervention had decreased cardiovascular events and mortality (measured by a composite outcome) when compared to those managed without surgery. 

Summary: Prior studies have demonstrated that the median survival for patients with untreated severe, symptomatic aortic stenosis (AS) is 2-3 years. Given the poor natural history of untreated AS, current guidelines recommend aortic valve replacement (surgical or transcatheter) for patients with severe, symptomatic AS. In contrast to patients with symptomatic, severe AS, patients with asymptomatic disease are typically managed conservatively, with watchful waiting until the onset of AS-related symptoms or left ventricular systolic dysfunction (LVEF <50%). Data from observational studies suggest that early intervention may be beneficial in select patients with asymptomatic AS; however, robust randomized data for early intervention are limited. Therefore, the Aortic Valve Replacement versus Conservative Treatment in Asymptomatic Severe Aortic Stenosis (AVATAR) Trial was conducted to evaluate the safety and efficacy of early surgical aortic valve replacement in the treatment of asymptomatic patients with severe AS and normal left ventricular function. 

AVATAR was an international, multicenter randomized controlled trial that recruited patients ≥18 years of age with asymptomatic, severe AS (classified by max velocity >4m/s, mean gradient >40mmHg, aortic valve area <1cm2), and normal left ventricular function. Patients were randomized 1:1 to either early SAVR or conservative management. Of note, patients in the conservative treatment group underwent management of risk factors and comorbidities until the onset of AS-related symptoms, which qualified them for SAVR. Patients were excluded from the trial if they had symptoms with exercise testing, mixed aortic valve regurgitation, concomitant root/ascending aortic aneurysms, mitral valve disease, atrial fibrillation, severe lung disease, or prior history of heart surgery. The primary endpoint of the trial was a composite of all-cause mortality or major adverse cardiovascular events (MACE), defined as acute myocardial infarction, stroke, or unplanned heart failure hospitalization. Pre-specified secondary endpoints included: in hospital mortality, 30-day mortality, time to death, time to first heart failure hospitalization, incidence of overall serious adverse events, repeated MACE, major bleeding events, thromboembolic complications, and repeat aortic valve surgery in operated patients in both groups (accounting for crossover).

A total of 157 patients were randomly assigned to early SAVR (n=78) or conservative treatment (n=79). Mean age was 67 years and median follow up for the study cohort was 32 months. Of note, 25 of 79 patients (31.6%) in the conservative treatment group ultimately underwent SAVR, with new onset symptoms being the most common indication for surgery. At three years, patients who underwent SAVR had a significantly lower incidence of the primary composite endpoint than those assigned to conservative management group (15.2% vs 34.7%; HR 0.46; 95% CI: 0.23-0.90; p=0.02). When looking at the individual components of the composite primary endpoint, early SAVR patients experienced lower rates of MACE (20.5% vs 41.8%, p=0.004) than the conservative treatment group, with no statistically significant difference in all-cause mortality. In addition, there were no statistically significant differences in any prespecified secondary endpoints.

The lower incidence of the composite outcome in the SAVR group was driven mainly by a lower incidence of MACE among these patients. Limitations of this trial, in addition to the use of a composite outcome, include small sample size, event-driven trial design, and the lack of a blinded clinical event adjudication committee. However, these data provide promising preliminary evidence in support of the notion that early SAVR is safe and possibly efficacious for select patients with asymptomatic, severe AS. 

Bottom line: Early surgical aortic valve replacement for asymptomatic, severe aortic stenosis is associated with less major adverse cardiovascular events.

Standardized surgical field development, a surrogate for appropriate surgical field exposure and dissection (i.e., critical view of safety), has become a recent focus of surgical skills assessment tools. This study uses intraoperative videos submitted to Japanese Society for Endoscopic Surgery to construct a deep learning model to recognize standardized surgical field development for laparoscopic sigmoid colectomies. The authors found that automatic predictions from this model are highly correlated with the scores generated from manual expert video review, supporting the feasibility of this artificial intelligence model for use as part of an automatic screening system for endoscopic surgical skill. 

Summary: In this retrospective diagnostic study, authors used 650 videos of laparoscopic sigmoid resection (Lap-S) submitted to Japan Society for Endoscopic Surgery (JSES) between August 2016 – November 2017. Videos submitted to JSES received a technical Endoscopic Surgical Skill Qualification System (ESSQS) Score based on blinded expert evaluation. Authors applied deep learning to 60 videos with high ESQSS scores (under the assumption that surgeons with a high ESSQS score had used a standardized surgical field) to construct a recognition model for Lap-S standardized surgical field development. The model was then validated using a balanced random sample of 60 videos from low, intermediate, and high ESSQS score groups in order to allow for variability of the results. Artificial intelligence confidence scores (AICS) were then calculated for each video reflecting the similarity between surgical field development of the test surgeon compared with that of the highly skilled surgeon used to train the model. The main outcome measures were the correlation between AICS and ESSQS score and the screening performance of AICS for low and high ESSQS score groups. 

The mean (SD) ESSQS score of all 650 intraoperative videos was 66.2 points, and those of the 60 model-construction and 60 validation videos were 79.3 points and 62.4 points, respectively. The accuracy of the recognition model was 78.2%.  Spearman rank correlation coefficient between the AICS and ESSQS score was 0.81 (p < 0.001). For the low-score group, the model performed with specificity 93.3%, sensitivity 82.2% and area under receiver operating characteristic (AUROC) of 0.93, which is considered excellent performance. Similarly, for the high-score group, the model performed with specificity 93.3%, sensitivity 86.7%, AUROC 0.94. 

These results demonstrate that AICS is strongly correlated with ESSQS score. While the sample size was small, this proof-of-concept study supports the feasibility of a standardized surgical field recognition model as a tool for automatic screening of surgical skill. This method may be useful in efficiently identifying high and low-score groups without the need for human raters or expert review. Additionally, when used together with other evaluative tools, it can help to create a more comprehensive surgical skill assessment system. Lastly, since this model evaluates the progression of surgery, it is broadly applicable to other types of endoscopic procedures, especially ones that don’t have a defined “critical view of safety.” However, one of the drawbacks of this model is that it is not a mechanism for feedback but rather an evaluation tool (for example, it does not tell you if a certain view was achieved but rather if the surgeon had an appropriate view during the progression of the procedure).  

Bottom line: The use of artificial intelligence and deep learning models for standardized surgical field recognition during minimally invasive procedures may allow for automatic screening of surgical skill and contribute to a more comprehensive and objective system of laparoscopic surgical skill assessment.

This retrospective propensity-matched cohort study compared several outcomes in patients with a diagnosis of small bowel neuroendocrine tumor presenting with synchronous metastases who underwent either upfront small bowel resection (USBR) or nonoperative management (NOM). They found that those who received USBR had a lower incidence of unplanned acute care admissions and subsequent small bowel-related surgery, as well as improved overall survival. 

Summary: This is a retrospective cohort study that used linked administrative healthcare datasets from the province of Ontario, Canada through the Institute for Clinical Evaluative Sciences to identify 1,000 patients with a new diagnosis of primary small bowel neuroendocrine tumor (jejunum or ileum) presenting with synchronous metastases (diagnosed ≤6 months from initial date of diagnosis) from 2001 to 2017. Of the 1,000 patients, 785 (78.5%) underwent USBR and 215 (21.5%) were treated with initial NOM. To address confounding by indication, a 2:1 propensity score matching was performed in which 558 patients were matched, 384 (68.8%) of which were in the treatment group (USBR) and 174 (31.2%) were in the control group (NOM). The primary study outcomes were unplanned acute care admissions and subsequent small bowel-related surgery for both the entire cohort as well as the propensity-matched cohort. The secondary outcome was median overall survival.  

After 4.6 years of follow up, when looking at the entire cohort, patients with USBR had lower cumulative incidence of unplanned acute care admissions at 1-year (48.1 % vs 81.3%, p <0.001) and 3-years (71.1% vs 84.6%, p <0.001), compared to those with NOM (HR 0.71, 95% CI: 0.58-0.87). They also had a lower cumulative incidence of subsequent small bowel-related surgery at 1 year and 3 years (HR 0.47, 95% CI: 0.36-0.61). The median overall survival for the entire cohort was 13.2 years for USBR compared to 7.1 years for NOM. In the propensity-matched cohort, patients with USBR were less likely to have unplanned acute care admissions than those with NOM at 1 year (49.1% vs 82.2%, p <0.001) and 3 years (72.6% vs 86.4%, p <0.001) with an adjusted HR of 0.72 (95% CI: 0.57-0.91) and lower cumulative incidence of subsequent small bowel surgery at 1 year and at 3 years (HR 0.44, 95% CI: 0.29-0.67). The median overall survival for the propensity matched cohort was 11.6 years (95% CI, 9.3- 13.9) for USBR and 6.2 years (95% CI, 4.6- 10.6) for NOM.  

In a sensitivity analysis excluding emergency surgery cases, the USBR group showed lower cumulative rates of unplanned acute care admissions and subsequent small bowel-related surgeries compared to the NOM group. In the propensity-matched cohort, USBR had significantly lower rates of such admissions at 1-year (45.2%) and 3-year (70.0%), as well as subsequent small bowel-related surgeries at 1-year (15.0%) and 3-year (18.3%). They also calculated E-values to determine the strength of the unmeasured confounding necessary to invalidate the observed association between exposure and outcome; they found in the propensity-matched cohort, the E-value for the association between USBR and outcomes was 1.85 for unplanned acute care admission and 2.75 for subsequent small bowel related surgery. 

This methodologically rigorous study suggests USBR may prevent loco-regional complications from small bowel neuroendocrine tumors when compared to NOM. On the other hand, both USBR and NOM groups had prolonged survival in patients presenting with metastatic disease (beyond 5 years), highlighting the importance of minimizing disease morbidity and prioritizing patients’ quality of life when considering different interventions for neuroendocrine tumors. 

Bottom line: Upfront small bowel resections for metastatic small bowel neuroendocrine tumors is associated with less unplanned acute care admissions and less subsequent small bowel surgeries compared to non-operative management.