Faculty Bibliography

OBJECTIVE:Ischemic complications (including in the lower extremity, visceral, spinal, and pelvic territories) following standard endovascular aortic repair (EVAR) are well recognized but fortunately uncommon. The incidence of such complications following fenestrated and branched aortic repair (F/BEVAR) has not been well defined in the literature. The objective of this study was to compare the incidence of ischemic complications between EVAR and F/BEVAR and to elucidate potential risk factors for these complications. METHODS:We identified all patients who underwent EVAR from 2003 to 2017 or F/BEVAR from 2012 to 2017 in the national Vascular Quality Initiative database. We assessed differences in perioperative ischemic outcomes with methods including logistic regression and inverse probability of treatment propensity score weighting, using a composite endpoint of lower extremity ischemia, intestinal ischemia, stroke, or new dialysis as the primary endpoint. RESULTS:The data comprised 35,379 EVAR patients and 3374 F/BEVAR patients. F/BEVAR patients were more likely to be female, have had previous aneurysm repairs, and be deemed unfit for open aneurysm repair; they were less likely to have ruptured aneurysms; and they had higher estimated blood losses, contrast volumes, and fluoroscopy and procedure times. The incidence of any ischemic event (7.7% vs 2.2%) as well as the incidences of the component endpoints of lower extremity ischemia (2.3% vs 1.0%), intestinal ischemia (2.7% vs 0.7%), stroke (1.5% vs 0.3%), and new hemodialysis (3.1% vs 0.4%) were all significantly increased (all P < .001) in F/BEVAR compared with standard EVAR. After propensity adjustment, F/BEVAR conferred increased odds of any ischemic complication (1.8), intestinal ischemia (2.0), lower extremity ischemia (1.3), new hemodialysis (10.2), and stroke (2.3). CONCLUSIONS:Rates of lower extremity ischemia, intestinal ischemia, new dialysis, and stroke each range from 0% to 1% for standard EVAR and 1% to 3% for F/BEVAR. The incidence of perioperative ischemic complications following F/BEVAR is significantly increased compared to EVAR. The real-world data in this study should help guide decision-making for surgeons and patients as well as serve as one metric for progress in device and technique development. Improvements in ischemic complications may come from continued technology development such as smaller sheaths, improved imaging to decrease procedure time and contrast volume, embolic protection, and increased operator skill with wire and catheter manipulation.

OBJECTIVE:Prior authorization (PA) is a process used by payers for safety and cost savings purposes, but it has received criticism for being time-consuming and costly because of administrative burden. Our study evaluated efficacy of PA applied to in-office lower extremity superficial venous procedures. METHODS:All in-office lower extremity venous procedures scheduled to be performed at our institution in 2017 were included in the study. Variables of interest were type of procedure, initial PA status (approved or denied), rationale for the decision, and final status after appeal. Cost analysis was performed using Centers for Medicare and Medicaid Services allowable rates to approximate billing and reimbursement data (proprietary) as well as calculated using industry averages. RESULTS:For 2017, of 1959 procedures scheduled, 57.9% (n = 1134) required PA. Of these, only 6.1% (n = 69) received initial PA denial, and nearly 40% of the denials (n = 27) were overturned after appeal. Of the 42 denials that were upheld, 15 resulted in cancellations; the remainder were performed by patient self-pay (n = 11) or by the provider pro bono (n = 16). Overturned denials were a result of either submission of incomplete clinical data on initial PA request or insufficient documentation of clinical necessity. When Centers for Medicare and Medicaid Services allowable rates were applied for cost analysis, the denials resulted in <$60,000 payer savings. Administrative expenses totaled >$110,000 when industry standards were applied, which far exceeds any calculated payer savings using the same methods. The 15 denials resulting in procedure cancellations (1.3% of all PAs) could be considered a net savings to the health care system but only approximated a mere $30,000. CONCLUSIONS:Our study demonstrates that PA is not a cost-effective measure for utilization management of outpatient superficial venous procedures when surgeon practices are already well aligned with insurance guidelines. For these physicians and physician groups, the administrative cost associated with the PA process exceeds the savings seen by the insurance companies.

Objective: Abdominal aortic aneurysm (AAA) is a fatal vascular disease on rupture with still limited mechanistic knowledge of the pathophysiologic process. We sought to determine the heterogeneous cell subtypes and to characterize the spectrum of transcriptome signatures in each cell population within the aneurysmal wall by unbiased single-cell RNA sequencing (scRNA-seq) of human AAA tissue.
Method(s): Aortic specimens were collected from AAA and control healthy organ donor. Samples were processed by enzymatic digestion and mechanical disruption to generate single-cell suspension. Single-cell RNA libraries were prepared after generation of single-cell beads in emulsion. Sequencing was performed on a NovaSeq 6000 platform (Illumina, San Diego, Calif). After alignment, barcode assignment, and sample de-multiplexing, data analysis was performed on t-distributed stochastic neighbor embedding charts of cell transcriptome. Cell clusters were identified by unsupervised proximity based on Euclidian distance and supervised identification of biologic markers within clusters. Pathway analysis algorithms were used to outline biologically relevant networks.
Result(s): Unbiased analysis of scRNA-seq data sets showed 19 different cell clusters with unique transcriptomic signatures in AAA. A total of 8826 significant differentially expressed genes were identified in AAA vs control. Notably, gene transcription-associated extracellular matrix remodeling (COL1A1, COL3A1, COL1A2, LUM), Wnt signaling modulation (SFRP2), and synthetic cellular phenotypes (RPS29, RPS27, RPL13A, RPL28) were among the top increased profiles in AAA tissue. Pathway enrichment analysis of AAA vs control libraries revealed significant modulation of cell proliferation, cell-extracellular matrix interaction, neoangiogenesis, and inflammation. Five novel cell clusters with distinct immune synthetic phenotypes were predominantly abundant in AAA wall compared with the healthy aorta. A robust enrichment in immune cell entities was identified in AAA but not in control tissues, including expansion of CD19⁺ B lymphocytes and a subset of CD3E⁺ T lymphocytes significantly expressing IL32 and CCL5. In contrast, smooth muscle cell (ACTA2⁺MYH11⁺) number declined in AAA but revealed increased transcription of the protease ADAMTS4 and inflammatory signals (CCL19, CCL21, IL6, CCL2). Intercluster pathway analysis revealed enrichment of eukaryotic initiation factor 2 and mechanistic target of rapamycin signaling in the AAA macrophage population along with an increased number of inflammatory and T-cell activation cascades.
Conclusion(s): To the best of our knowledge, this is the first report of scRNA-seq analysis on human AAA. This cutting-edge technique uncovered novel cell clusters and provided a comprehensive understanding of cellular spatiotemporal changes within the AAA wall. Here we provide novel interconnected mechanistic insights into this complex disease to enrich our understanding of AAA development.
Copyright

CONTEXT/BACKGROUND:Over the past decade, a number of endovascular approaches have evolved to treat aortic aneurysms with anatomy that is not amenable to traditional endovascular repair, although the optimal practice and referral patterns remain in question. The Zenith fenestrated (Z-Fen) endograft (Cook Medical) represents the first commercially available fenestrated graft product in the United States. OBJECTIVE:We aim to quantify practice patterns in Z-Fen use during the first 5 years of commercial availability, and we identify predictors of high and low uptake. DESIGN, SETTING, AND PATIENTS/METHODS:This is a retrospective review of complete order records for Z-Fen endografts since June 2012. We performed univariate and multivariate regressions of predictors that surgeons and centers would be in the top and bottom quartiles of annual Z-Fen use. RESULTS:Since June 15, 2012, 744 surgeons have been trained to use Z-Fen, and 4133 cases have been performed at 409 trained centers. The average annual number of cases per trained surgeon was 4.46 [95% confidence interval (CI), 3.58-5.70]; however, many surgeons performed few or no cases following training, and there was a skew toward users with low average annual volumes (25th percentile 1.23, 50th percentile 2.35, 75th percentile 4.93, and 99th percentile 33.29). Predictors of high annual use in the years following training included academic center (aOR 5.87, P = .001) and training within the first 2 years of availability (aOR 46.23, P < .001). CONCLUSION/CONCLUSIONS:While there is literature supporting the safety and efficacy of Z-Fen, adoption has been relatively slow in an era when the vast majority of vascular surgeons have advanced endovascular skills. Given the training and resources required to use fenestrated or branched aortic endovascular devices, referral patterns should be determined and training should be focused on centers with high expected volumes.