December 2022

5-year results from the PARTNER-2A trial showed that outcomes following transcatheter aortic valve replacement remained non-inferior to surgical aortic valve replacement, despite an interval increase in the incidence of the primary endpoint in the transcatheter group. 

Summary: Surgical aortic valve replacement (SAVR) has for decades been the mainstay approach for treating severe symptomatic aortic stenosis. This was challenged by the advent of transcatheter aortic valve replacements (TAVR), wherein a tissue valve loaded onto an expandable stent frame is deployed into the diseased native aortic valve under fluoroscopic guidance. The benefits of TAVR were most obvious in patients who would otherwise be too high risk to withstand cardiopulmonary bypass for SAVR. Additional studies have since explored the possibility of expanding TAVR to patients presenting with a lower surgical risk. PARTNER-2A was a randomized control trial designed to assess the outcomes following transcatheter aortic valve replacement (TAVR) and surgical aortic valve replacement (SAVR) in patients with severe aortic stenosis with an intermediate risk for surgery.1 

2032 patients were enrolled across 57 centers in the US and Canada. A total of 994 of 1011 patients randomized to TAVR and 944 of 1021 randomized to SAVR underwent the assigned procedure. 76.6 % of the TAVR patients underwent a transfemoral TAVR, and in the remaining 23.3 % the TAVR system was deployed either trans-apically or trans-aortic. The mean STS score was 5.8% in each group. 

The intention-to-treat population was used for all analyses. For the composite outcome of death from any cause or disabling stroke at 5-years, TAVR remained non-inferior to SAVR with incidences of 47.9% and 43.4% respectively (HR 1.09; 95% CI, 0.95 to 1.25; p = 0.21). The incidences of death from any cause in the overall TAVR and SAVR groups were 46.0% and 42.1%, respectively. Stratifying based on TAVR access strategy, the transfemoral access cohort showed slightly improved outcomes (42.7% and 40.5%) vs the transthoracic access cohort (56.9% and 47.3%).2

These results when contextualized with the 2-year PARTNER-2A outcomes are quite revealing. The time to event curves for the primary composite outcome shows that while there is a clear short-term mortality risk associated with SAVR, by 3 years the survival curves intersect, subsequently trending in favor of SAVR by 5 years. This is despite 86 out of the 944 patients (9.11%) who underwent SAVR, had either planned or unplanned concomitant surgical procedures ranging from mitral valve repairs to aortic root enlargements.

Unlike the open surgical approach which involves resecting the diseased aortic valve, TAVR systems must be deployed within heavily calcified and diseased native aortic valves that increase the likelihood of para-valvular regurgitation. This is thought to be the primary driver of the higher long-term mortality with TAVR vs the SAVR group seen in the trial, though questions regarding the long-term durability of TAVR systems remain. The SAPIEN XT valve system that was used for the trial, has since been superseded by the SAPIEN 3 system in contemporary clinical practice, with additional design elements to mitigate paravalvular leaks. Furthermore, transfemoral access is now the default approach, with many centers abandoning the transapical access in favor for other approaches (carotid, subclavian, axillary, caval) owing to the poor outcomes with the transapical route. Finally, the mean age of trial participants was 81 years, and extrapolations to younger age groups must be made with caution.  

In summary the PARTNER-2A trial shows that while TAVR remains non-inferior to SAVR in patients with severe aortic stenosis with an intermediate surgical risk profile, there was a distinct increase in the incidence of the primary endpoint in the TAVR group at 5 years vs 2 years. 

References:

1. Leon, M. B. et al. Transcatheter or Surgical Aortic-Valve Replacement in Intermediate-Risk Patients. N Engl J Med 374, 1609–1620 (2016).
2. Makkar, R. R. et al. Five-Year Outcomes of Transcatheter or Surgical Aortic-Valve Replacement. N Engl J Med 382, 799–809 (2020).

Clinically, the use of intraoperative nerve monitoring is a useful adjunct in the prevention of recurrent laryngeal nerve injury; however, more large-scale, prospective randomized control trials must be performed to show a statistically significant effect.

Summary: 

Introduction: Since the advent of intraoperative nerve monitoring (IONM) in the 1960s, its use in thyroid surgery has increased steadily over the years, with over 90% of endocrine and head & neck surgeons in the US and Europe routinely using IONM to identify the recurrent laryngeal nerve (RLN). However, despite its widespread use and growing popularity, there is still a discrepancy in the literature as to whether its use prevents RLN injury. Previous meta-analyses have conflicting reports, with some supporting the routine use of IONM while others argue its non-significant impact on preventing RLN injury. Thus, this study was the first systematic review and meta-analysis to include only randomized control trials to evaluate the relationship between IONM and RLN injury versus visualization alone (VA) for patients undergoing thyroidectomy.  

Methods: An electronic search for relevant studies was performed across multiple platforms. Eligible studies included those with clear research methodology, including prospective randomization of patients to IONM and VA during thyroid surgery and those that report on the primary outcome measure of overall RLN injury rates following thyroid surgery. RLN injury was further subdivided into transient (those that resolved within 6 month post-operatively) and permanent (unresolved injuries 6-months or longer post-operatively). 

Results: 973 studies were identified of which 8 met inclusion criteria and were included in the systematic review and meta-analysis between the years of 2009 and 2021. Of the 8 studies, 6 of them were conducted at European institutions. Between the 8 studies, there were 2521 patients with 4977 nerves at risk. Most patients underwent total thyroidectomy (1995/2251, 79.1%) and 49.8% of RLNs underwent IONM (2480/4978) while 50.2% underwent VA (2497/4978). Overall, the RLN injury rate was 2.8% (138/4977), which was higher for those who underwent VA compared to those who underwent IOM but not a statistically significant difference (VA: 3.2% (80/2497) vs. IONM: 2.3% (58/2480), P = 0.069). The overall odds ratio for the meta-analysis was 0.72, but also not significant (95% CI: 0.51 – 1. 02, p = 0.060). There was no significant difference in transient RLN injury rates (OR: 0.69, 95% CI: 0.42–1.06, P = 0.090) while the rate of permanent RLN injury rate were slightly higher in those who underwent VA, but again, a difference that was not statistically significant (OR 0.76, 95% CI: 0.36 – 1.59, p = 0.470). 

Discussion: This systematic review and meta-analysis of randomized control trials showed similar RLN injury rates between patients who had IOMN during thyroidectomy versus those who had just visualization alone. However, the authors mention how this difference “trended towards” significance. There were several limitations of this study (and studies of this nature in general), one being that the number of RLN injury events in each group were small and so it is inherently more difficult to perform a prospective RCT (and recruit enough patients) in order to obtain statistically significant results or see a clinically meaningful difference in the two groups. Additionally, it is difficult to blind surgeons to an intervention and thus, this potentially introduces bias into these types of studies. Lastly, one of the studies provided over 60% of the data given it was the largest RCT and thereby could have potentially skewed the results.

Resident wellbeing correlates more strongly with perception of service vs education ratio than it does with duty hours worked. Higher service obligation scores led to overall lower wellness scores regardless of the number of hours reported. The quality of the time spent in the clinical learning environment matters more than the quantity of time. 

Summary: Concerns regarding physician wellness, burnout, suicide, sleep deprivation, and impaired patient outcomes have led the ACGME to emphasize duty-hour restrictions and wellness initiatives. Despite this, the FIRST trial demonstrated no significant impact on patient outcomes or physician wellness. Additional studies on the enforcement of duty hours have shown mixed effects on resident wellbeing. 

The authors sought to evaluate the impact of duty hours and service vs education on wellness after they were given 2 citations by the ACGME in 2019, one for duty hour violations, and one for service obligations taking precedence over education activities. 

The study took place at UT Southwestern (5-year program, 13 residents/year, optional lab time). After their citations, the program developed a weekly screening tool to track progress on both dimensions. Each week, General Surgery residents logged their hours, rated the service-to-education ratio (SVE; 1-5 scale), and recorded their wellness levels (via a Wellbeing “Fuel Guage” from 1 [empty] to 5 [full]). Surveys from the 73 residents were collected over a 24-week period. Work hour reporting was mandatory (program coordinators contact trainees if not entered within 1 week) and survey responses were voluntary. 

Data for an individual resident-week was considered complete if the duty hour log, Fuel Gauge score, and SVE score were completed for the same week. Any incomplete data sets were excluded from the analysis. 273 complete data sets were collected over the 24-week study period. Spearman’s rank correlation was used to describe the relationship between duty hours, Fuel Gauge and SVE. The median Fuel Gauge score was 4, the median SVE score was 4, and 8.8% of resident-weeks exceeded 80 hours. Fuel Gauge assessment scores demonstrated a moderately positive correlation with SVE (rho 0.64, p <0.001) but not with duty hours (rho -0.13, p 0.035). There was a weak negative correlation between SVE and duty hours (rho -0.20, p <0.001). For residents with low SVE perception, 50% has a low Fuel Gauge score (1-2), vs 32.1% with medium SVE and 17.9% with high SVE. 

This data shows that focusing on service vs education ratio may be more impactful than focusing on duty hours. The quality of the time spent in the clinical learning environment matters more than the quantity of time. Additionally, a hyperfocus on hours may have unintended consequences including service restructuring and a loss of resident autonomy. The main takeaway is that wellness is multifactorial. 

The main limitation of this study is that it was done at a single institution with voluntary survey-based tool. The lack of mandated reporting may skew the results towards those choosing to fill out the survey. Additionally, the data (including duty hours) is self-reported and may not be accurate. Also, the Fuel Gauge is a surrogate measure of wellness and burnout and has not been validated like the Maslach Burnout Inventory.