Distinct Classes of Complex Structural Variation Uncovered across Thousands of Cancer Genome Graphs.

Cancer genomes often harbor hundreds of somatic DNA rearrangement junctions, many of which cannot be easily classified into simple (e.g., deletion) or complex (e.g., chromothripsis) structural variant classes. Applying a novel genome graph computational paradigm to analyze the topology of junction copy number (JCN) across 2,778 tumor whole-genome sequences, we uncovered three novel complex rearrangement phenomena: pyrgo, rigma, and tyfonas. Pyrgo are "towers" of low-JCN duplications associated with early-replicating regions, superenhancers, and breast or ovarian cancers. Rigma comprise "chasms" of low-JCN deletions enriched in late-replicating fragile sites and gastrointestinal carcinomas. Tyfonas are "typhoons" of high-JCN junctions and fold-back inversions associated with expressed protein-coding fusions, breakend hypermutation, and acral, but not cutaneous, melanomas. Clustering of tumors according to genome graph-derived features identified subgroups associated with DNA repair defects and poor prognosis.

Prospective Study of Risk Factors for Postoperative Atrial Fibrillation After Cardiac Surgery.

INTRODUCTION: To examine risk factors for new-onset postoperative atrial fibrillation (POAF) after cardiac surgery. METHODS: Patients enrolled in the Cardiothoracic Surgical Trials Network multicenter, randomized trial of rate control versus rhythm control for POAF were included. Predictors of POAF were determined using multivariable logistic regression. RESULTS: Among the 2104 patients who were enrolled preoperatively, 695 developed POAF (33.0%). Rates of POAF were 28.1% after isolated coronary artery bypass grafting (CABG), 33.7% after isolated valve repair or replacement, and 47.3% after CABG plus valve repair or replacement. Baseline characteristics associated with an increased risk of POAF identified on multivariable analysis included older age (odds ratio [OR] 1.57; 95% confidence interval [CI] 1.42-1.73, per 10 y), White race or non-Hispanic ethnicity (OR 1.52; CI: 1.11-2.07), history of heart failure (OR 1.55; CI: 1.16-2.08), and history of hypothyroidism (OR 1.42; CI 1.04-1.94). The type of cardiac procedure was associated with an increased risk of POAF with both isolated valve repair or replacement (OR 1.33, CI 1.08-1.64) and combined CABG plus valve repair or replacement (OR 1.64, CI 1.24-2.17) having increased risk of POAF compared to isolated CABG. No preoperative cardiac medication was associated with POAF. CONCLUSIONS: In this prospective cohort of patients, older age, a history of hypothyroidism, a history of heart failure, and valve repair or replacement, with or without CABG, and White non-Hispanic race were associated with an increased risk of POAF.

A comparison of endovascular repair to medical management for acute vs subacute uncomplicated type B aortic dissections.

OBJECTIVE: Thoracic endovascular aortic repair (TEVAR) has emerged as a viable option of treatment for uncomplicated type B aortic dissection (UTBAD) due to the potential for inducing favorable aortic remodeling. The aim of this study is to compare outcomes of UTBAD treated medically or with TEVAR in either the acute (1 to 14 days) or subacute period (2 weeks to 3 months). METHODS: Patients with UTBAD between 2007 and 2019 were identified using the TriNetX Network. The cohort was stratified by treatment type (medical management; TEVAR during the acute period; TEVAR during the subacute period). Outcomes including mortality, endovascular reintervention, and rupture were analyzed after propensity matching. RESULTS: Among 20,376 patients with UTBAD, 18,840 were medically managed (92.5%), 1099 patients were in the acute TEVAR group (5.4%), and 437 patients were in the subacute TEVAR group (2.1%). The acute TEVAR group had higher rates of 30-day and 3-year rupture (4.1% vs 1.5%; P < .001; 9.9% vs 3.6%; P < .001) and 3-year endovascular reintervention (7.6% vs 1.6%; P < .001), similar 30-day mortality (4.4% vs 2.9%; P < .068), and lower 3-year survival compared with medical management (86.6% vs 83.3%; P = .041). The subacute TEVAR group had similar rates of 30-day mortality (2.3% vs 2.3%; P = 1), 3-year survival (87.0% vs 88.8%; P = .377) and 30-day and 3-year rupture (2.3% vs 2.3%; P = 1; 4.6% vs 3.4%; P = .388), with significantly higher rates of 3-year endovascular reintervention (12.6% vs 7.8%; P = .019) compared with medical management. The acute TEVAR group had similar rates of 30-day mortality (4.2% vs 2.5%; P = .171), rupture (3.0% vs 2.5%; P = .666), significantly higher rates of 3-year rupture (8.7% vs 3.5%; P = .002), and similar rates of 3-year endovascular reintervention (12.6% vs 10.6%; P = .380) compared with the subacute TEVAR group. There was significantly higher 3-year survival (88.5% vs 84.0%; P = .039) in the subacute TEVAR group compared with the acute TEVAR group. CONCLUSIONS: Our results found lower 3-year survival in the acute TEVAR group compared with the medical management group. There was no 3-year survival benefit found in patients with UTBAD who underwent subacute TEVAR compared with medical management. This suggests the need for further studies looking at the necessity for TEVAR when compared with medical management for UTBAD as it is non-inferior to medical management. Higher rates of 3-year survival and lower rates of 3-year rupture in the subacute TEVAR group compared with the acute TEVAR group suggest superiority of subacute TEVAR. Further investigations are needed to determine the long-term benefit and optimal timing of TEVAR for acute UTBAD.

De Novo Renal Failure and Clinical Outcomes of Patients With Critical Coronavirus Disease 2019.

OBJECTIVES: To describe the characteristics and outcomes associated with concomitant renal and respiratory failure in patients with critical coronavirus disease 2019. DESIGN, SETTING, AND PATIENTS: This is a case series of patients from a U.S. healthcare system in New York City. All adult patients (≥ 18 yr) admitted to the hospital with positive coronavirus disease 2019 testing between March 10, 2020, and March 31, 2020, who required mechanical ventilatory support were included. Patients who remained hospitalized were followed through May 1, 2020. INTERVENTIONS: Renal replacement therapy included at least one session of dialysis, continued venovenous hemofiltration, or peritoneal dialysis. MEASUREMENTS AND MAIN RESULTS: Baseline characteristics, laboratory markers, 30-day in-hospital outcomes, ventilator days, and survival to discharge were included. Multivariate predictors for mortality and need for renal replacement therapy were identified. A total of 330 patients were included in this analysis and were most commonly greater than or equal to 70 years (40%), male (61%), Black or African American (41%), and Hispanic or Latino (38%). Renal replacement therapy was required in 101 patients (29%), most commonly among Blacks or African Americans (50%). Elevated d-dimer, C-reactive protein, and procalcitonin were associated with renal replacement therapy, compared with the nondialysis cohort. Overall, 243 patients (74%) died and 56 (17%) were discharged from the hospital, of which 9 (3%) required renal replacement therapy. Male sex (odds ratio, 2.0; 1.1-3.5; p = 0.020), Black race (odds ratio, 1.8; 1.0-3.1; p = 0.453), and history of hypertension (odds ratio, 2.7; 1.3-5.4; p = 0.005) were predictors for requiring renal replacement therapy. Risk factors for in-hospital mortality included age greater than or equal to 60 years (odds ratio, 6.2; 3.0-13.0; p < 0.0001), male sex (odds ratio, 3.0; 1.4-6.4; p = 0.004), and body mass index greater than or equal to 30 kg/m2 (odds ratio, 2.1; 1.0-4.4; p = 0.039). Concomitant renal failure in critical coronavirus disease 2019 was not a significant predictor of death (odds ratio, 2.3; 0.98-5.5; p = 0.057). CONCLUSIONS: This case series concludes that respiratory failure conveys significant mortality risk in patients with coronavirus disease 2019 and that survival with concomitant renal failure is rare.

Two-Year Outcomes of Surgical Treatment of Moderate Ischemic Mitral Regurgitation.

BACKGROUND: In a trial comparing coronary-artery bypass grafting (CABG) alone with CABG plus mitral-valve repair in patients with moderate ischemic mitral regurgitation, we found no significant difference in the left ventricular end-systolic volume index (LVESVI) or survival after 1 year. Concomitant mitral-valve repair was associated with a reduced prevalence of moderate or severe mitral regurgitation, but patients had more adverse events. We now report 2-year outcomes. METHODS: We randomly assigned 301 patients to undergo either CABG alone or the combined procedure. Patients were followed for 2 years for clinical and echocardiographic outcomes. RESULTS: At 2 years, the mean (±SD) LVESVI was 41.2±20.0 ml per square meter of body-surface area in the CABG-alone group and 43.2±20.6 ml per square meter in the combined-procedure group (mean improvement over baseline, -14.1 ml per square meter and -14.6 ml per square meter, respectively). The rate of death was 10.6% in the CABG-alone group and 10.0% in the combined-procedure group (hazard ratio in the combined-procedure group, 0.90; 95% confidence interval, 0.45 to 1.83; P=0.78). There was no significant between-group difference in the rank-based assessment of the LVESVI (including death) at 2 years (z score, 0.38; P=0.71). The 2-year rate of moderate or severe residual mitral regurgitation was higher in the CABG-alone group than in the combined-procedure group (32.3% vs. 11.2%, P<0.001). Overall rates of hospital readmission and serious adverse events were similar in the two groups, but neurologic events and supraventricular arrhythmias remained more frequent in the combined-procedure group. CONCLUSIONS: In patients with moderate ischemic mitral regurgitation undergoing CABG, the addition of mitral-valve repair did not lead to significant differences in left ventricular reverse remodeling at 2 years. Mitral-valve repair provided a more durable correction of mitral regurgitation but did not significantly improve survival or reduce overall adverse events or readmissions and was associated with an early hazard of increased neurologic events and supraventricular arrhythmias. (Funded by the National Institutes of Health and Canadian Institutes of Health Research; ClinicalTrials.gov number, NCT00806988.).

Surgical ablation of atrial fibrillation during mitral-valve surgery.

BACKGROUND: Among patients undergoing mitral-valve surgery, 30 to 50% present with atrial fibrillation, which is associated with reduced survival and increased risk of stroke. Surgical ablation of atrial fibrillation has been widely adopted, but evidence regarding its safety and effectiveness is limited. METHODS: We randomly assigned 260 patients with persistent or long-standing persistent atrial fibrillation who required mitral-valve surgery to undergo either surgical ablation (ablation group) or no ablation (control group) during the mitral-valve operation. Patients in the ablation group underwent further randomization to pulmonary-vein isolation or a biatrial maze procedure. All patients underwent closure of the left atrial appendage. The primary end point was freedom from atrial fibrillation at both 6 months and 12 months (as assessed by means of 3-day Holter monitoring). RESULTS: More patients in the ablation group than in the control group were free from atrial fibrillation at both 6 and 12 months (63.2% vs. 29.4%, P<0.001). There was no significant difference in the rate of freedom from atrial fibrillation between patients who underwent pulmonary-vein isolation and those who underwent the biatrial maze procedure (61.0% and 66.0%, respectively; P=0.60). One-year mortality was 6.8% in the ablation group and 8.7% in the control group (hazard ratio with ablation, 0.76; 95% confidence interval, 0.32 to 1.84; P=0.55). Ablation was associated with more implantations of a permanent pacemaker than was no ablation (21.5 vs. 8.1 per 100 patient-years, P=0.01). There were no significant between-group differences in major cardiac or cerebrovascular adverse events, overall serious adverse events, or hospital readmissions. CONCLUSIONS: The addition of atrial fibrillation ablation to mitral-valve surgery significantly increased the rate of freedom from atrial fibrillation at 1 year among patients with persistent or long-standing persistent atrial fibrillation, but the risk of implantation of a permanent pacemaker was also increased. (Funded by the National Institutes of Health and the Canadian Institutes of Health Research; ClinicalTrials.gov number, NCT00903370.).

A Comparison of Inflammatory Responses Between Robotically Enhanced Coronary Artery Bypass Grafting and Conventional Coronary Artery Bypass Grafting: Implications for Hybrid Revascularization.

OBJECTIVE: The inflammatory response elicited by robotically enhanced coronary artery bypass grafting (r-CABG) has not been well described. When r-CABG is performed as part of hybrid coronary revascularization, the inflammatory milieu and the timing of percutaneous coronary intervention may affect the stent patency negatively in the short and long term. The goal of this study was to describe the extent and time course of cytokine release after r-CABG compared with conventional CABG (c-CABG) and to elucidate the optimal timing for r-CABG in the setting of hybrid coronary revascularization for a future study. DESIGN: Prospective, observational study. SETTING: Tertiary-care center in a university hospital. PARTICIPANTS: The study comprised patients scheduled to undergo r-CABG or c-CABG from October 2012 to November 2014. INTERVENTIONS: Cytokine levels of interleukin (IL)-6, IL-8, IL-10; tumor necrosis factor-α; and C-reactive protein (CRP) were measured at the following time points: preprocedure; at the end of the procedure; and at 4, 8, 12, 24, and 48 hours after the procedure. MEASUREMENTS AND MAIN RESULTS: Twenty-eight patients undergoing r-CABG and 10 patients undergoing c-CABG were enrolled. The levels of cytokines after r-CABG and c-CABG were compared using the mixed-effect linear regression model for longitudinal data. Cytokine release in the r-CABG group was comparatively less for IL-6, IL-10, tumor necrosis factor, and CRP levels. They all trended toward the baseline by the 48th hour in both groups, except CRP levels, which reached their peak at 48 hours in both groups. CONCLUSIONS: The inflammatory response to r-CABG was blunted compared with that of c-CABG. The high CRP levels on the second postoperative day after r-CABG were a cause for concern in regard to percutaneous coronary intervention performed at that time period, but additional studies are necessary.

Comparison of Index Hospitalization Costs Between Robotic CABG and Conventional CABG: Implications for Hybrid Coronary Revascularization.

OBJECTIVES: To compare the direct costs of the index hospitalization and 30-day morbidity and mortality incurred during robotic and conventional coronary artery bypass grafting at a single institution based on hospital clinical and financial records. DESIGN: Retrospective study, propensity-matched groups with one-to-one nearest neighbor matching. SETTING: University hospital, a tertiary care center. PARTICIPANTS: Two thousand eighty-eight consecutive patients who underwent primary coronary artery bypass grafting (CABG) from January 2007 to March 2012. INTERVENTIONS: One hundred forty-one matched pairs were created and analyzed. MEASUREMENTS AND MAIN RESULTS: Robotic CABG was associated with a decrease in operative time (5.61±1.1 v 6.6±1.15 hours, p<0.001), a lower need for blood transfusion (12.8% v 22.6%, p = 0.04), a shorter length of stay (6 [4-9]) v 7 [5-11] days, p = 0.001), a shorter ICU stay (31 [24-49] hours v 52 [32-96.5] hours, p<0.001) and lower NY state complications composite rate (4.26% v 13.48%, p = 0.01). In spite of that, the cost of robotic procedures was not significantly different from matched conventional cases ($18,717.35 [11,316.1-34,550.6] versus $18,601 [13,137-50,194.75], p = 0.13), except 26 hybrid coronary revascularizations in which angioplasty was performed on the same admission (hybrid 25,311.1 [18,537.1-41,167.85] versus conventional 18,966.13 [13,337.75-56,021.75], p = 0.02). CONCLUSION: Robotically assisted CABG does not increase the cost of the index hospitalization when compared to conventional CABG unless hybrid revascularization is performed on the same admission.

Iatrogenic Ventricular Septal Defect Following Transcatheter Aortic Valve Replacement: A Systematic Review.

BACKGROUND: Ventricular septal defects (VSD) are rarely reported as a complication following transcatheter aortic valve replacement (TAVR). We sought to characterise the patients, clinical management, and outcomes regarding this rare phenomenon. METHODS: Relevant articles were identified by a systematic search of MEDLINE and EMBASE databases from January, 2002 to September, 2015. RESULTS: A total of 18 case reports, including 20 patients, were identified. The median age was 83 years and six were male. Twelve were performed by trans-femoral approach. Pre-dilation was performed in 12 patients and post-dilation in four. Balloon expandable valves were used in the majority (85%) of cases. The clinical presentation varied from asymptomatic to progressive heart failure. The timing of the diagnosis also varied significantly from immediately post valve implantation to one year afterwards. There were two cases of Gerbode-type defect while the rest were inter-ventricular defects. The location was mostly membranous or perimembranous (79%) and adjacent to the valve landing zone. A total of seven interventions (one open surgery and six percutaneous closure) were performed. Four patients died during the same hospital admission. Sixteen survived past discharge (range 12 days to two years). CONCLUSIONS: Ventricular septal defects post-TAVR were seen more with balloon expandable valves and with pre-dilation or post-dilation. Percutaneous treatment of the VSD was preferred over open cardiac surgery given the high surgical risk in this patient population. Some, but not all, patients survived TAVR and VSD and had a good prognosis for both patient groups with or without VSD closure.

Open, endovascular or hybrid repair of aortic arch disease: narrative review of diverse strategies with diverse options.

OBJECTIVES: The management of aortic arch disease is complex. Open surgical management continues to evolve, and the introduction of endovascular repair is revolutionizing aortic arch surgery. Although these innovative techniques have generated the opportunity for better outcomes in select patients, they have also introduced confusion and uncertainty regarding best practices. METHODS: In New York, we developed a collaborative group, the New York Aortic Consortium, as a means of cross-linking knowledge and working together to better understand and treat aortic disease. In our meeting in May 2023, regional aortic experts and invited international experts discussed the contemporary management of aortic arch disease, differences in interpretation of the available literature and the integration of endovascular technology into disease management. We summarized the current state of aortic arch surgery in this review article. RESULTS: Approaches to aortic arch repair have evolved substantially, whether it be methods to reduce cerebral ischaemia, improve haemostasis, simplify future operations or expand options for high-risk patients with endovascular approaches. However, the transverse aortic arch remains challenging to repair. Among our collaborative group of cardiac/aortic surgeons, we discovered a wide disparity in our practice patterns and our management strategies of patients with aortic arch disease. CONCLUSIONS: It is important to build unique institutional expertise in the context of complex and evolving management of aortic arch disease with open surgery, endovascular repair and hybrid approaches, tailored to the risk profiles and anatomical specifics of individual patients.

Elevated glycosylated hemoglobin levels are associated with severe acute kidney injury following coronary artery bypass surgery.

BACKGROUND: Diabetic patients are at increased risk of acute kidney injury (AKI) following surgery. The significance of uncontrolled diabetes on kidney function after coronary artery bypass grafting (CABG) remains controversial. Our aim was to study the association between pre-operative hemoglobin A1c (HbA1c) and severe cardiac surgery-associated AKI (CSA-AKI) following CABG. METHODS: A single-center, retrospective cohort study including patients who underwent isolated CABG from 2010 to 2018 was performed. Patients were grouped into pre-operative HbA1c of <6.5 %, 6.5-8.5 %, and ≥8.5 %. Postoperative serum creatinine levels were queried for up to 30 days, and the 30-day risk of severe AKI was compared among groups. Multivariable logistic regression was used to study factors associated with severe CSA-AKI and the association of severe CSA-AKI with postoperative outcomes. Cox regression was used to study the association between severe CSA-AKI and all-cause mortality from the time of surgery to the last follow-up or death. RESULTS: A total of 2424 patients met the inclusion criteria. Patients were primarily male (70.5 %), with a median age of 64 years (IQR 57-71). Median bypass and cross-clamp times were 95 (IQR 78-116) and 78 min (IQR 63-95). Severe CSA-AKI occurred within 30 days in 5.7 %, 6.7 %, and 9.1 % of patients with pre-op HbA1c of <6.5 %, 6.5-8.5 %, and ≥8.5 %, respectively. After adjusting for covariates, HbA1c >8.5 %, was independently associated with severe CSA-AKI 30 days after CABG (aOR 1.59, 95%CI 1.06-2.40). In addition, severe CSA-AKI was associated with increased 30- (aOR 15.83,95%CI 7.94-31.56) and 90- day mortality (aOR 9.54, 95%CI 5.46-16.67), prolonged length of stay (aOR 3.46,95%CI 2.41-4.96) and unplanned 30-day readmission (aOR 2.64, 95%CI 1.77-3.94). Lastly, severe CSA-AKI was associated with increased all-cause mortality (aHR 3.19, 95%CI 2.43-4.17). CONCLUSION: Elevated preoperative HbA1c (≥8.5 %) was independently associated with an increased 30-day risk of severe CSA-AKI, which is a consistent predictor of adverse outcomes after CABG. Delaying surgery to achieve optimal glycemic control in an elective setting may be reasonable.

Low-Dose vs Standard Warfarin After Mechanical Mitral Valve Replacement: A Randomized Trial.

BACKGROUND: Current guidelines recommend a target international normalized ratio (INR) range of 2.5 to 3.5 in patients with a mechanical mitral prosthesis. The Prospective Randomized On-X Anticoagulation Clinical Trial (PROACT) Mitral randomized controlled noninferiority trial assessed safety and efficacy of warfarin at doses lower than currently recommended in patients with an On-X (Artivion, Inc) mechanical mitral valve. METHODS: After On-X mechanical mitral valve replacement, followed by at least 3 months of standard anticoagulation, 401 patients at 44 North American centers were randomized to low-dose warfarin (target INR, 2.0-2.5) or standard-dose warfarin (target INR, 2.5-3.5). All patients were prescribed aspirin, 81 mg daily, and encouraged to use home INR testing. The primary end point was the sum of the linearized rates of thromboembolism, valve thrombosis, and bleeding events. The design was based on an expected 7.3% event rate and 1.5% noninferiority margin. RESULTS: Mean patient follow-up was 4.1 years. Mean INR was 2.47 and 2.92 (P <.001) in the low-dose and standard-dose warfarin groups, respectively. Primary end point rates were 11.9% per patient-year in the low-dose group and 12.0% per patient-year in the standard-dose group (difference, -0.07%; 95% CI, -3.40% to 3.26%). The CI >1.5%, thus noninferiority was not achieved. Rates (percentage per patient-year) of the individual components of the primary end point were 2.3% vs 2.5% for thromboembolism, 0.5% vs 0.5% for valve thrombosis, and 9.13% vs 9.04% for bleeding. CONCLUSIONS: Compared with standard-dose warfarin, low-dose warfarin did not achieve noninferiority for the composite primary end point. (PROACT Clinicaltrials.gov number, NCT00291525).

Relation of Community-Level Socioeconomic Status to Delayed Diagnosis of Acute Type A Aortic Dissection.

Acute type A aortic dissection requires timely diagnosis and intervention. Previous studies have examined risk factors associated with delayed diagnosis; however, the effect of socioeconomic status (SES) has not been previously studied. Our study examined the impact of various SES measures on time to diagnosis. We examined time to diagnosis in consecutive cases of acute type A aortic dissection at a single institution. SES variables included race/ethnicity, Medicaid eligibility, and residence in a zip code with an increased Distressed Communities Index-an aggregate measure of community SES. Delayed diagnosis was defined as time to diagnosis in the upper quartile of the study population (>6.6 hours). A model predicting risk factors for delayed diagnosis was created using multivariable logistic regression. Our study included 124 patients with a median time to diagnosis of 3.36 hours (interquartile range [IQR] 1.83 to 6.63). A total of 92 patients were in the nondelayed cohort (median diagnosis time of 2.59 hours, IQR 1.49 to 4.18) and 32 patients were in the delayed cohort (median diagnosis time of 15.57 hours, IQR 9.34 to 28.75). In multivariable logistic regression, residence in a high-Distressed Communities Index zip code was associated with diagnostic delay (adjusted odds ratio [aOR] 5.108, p = 0.008). Patient age (aOR 0.944, p = 0.011), chest pain at presentation (aOR 0.099, p = 0.004), back pain at presentation (aOR 0.247, p = 0.012), evidence of malperfusion syndrome (aOR 0.040, p <0.001), history of hyperlipidemia (aOR 3.507, p = 0.026), and history of congestive heart failure (aOR 0.061, p = 0.036) were also significantly associated. In conclusion, our findings suggest community-level SES affects time to diagnosis in acute type A aortic dissection.

WITHDRAWN: Low-Dose Versus Standard Warfarin After Mechanical Mitral Valve Replacement: A Randomized Controlled Trial.

This article has been withdrawn at the request of the author(s) and/or editor. The Publisher apologizes for any inconvenience this may cause. The full Elsevier Policy on Article Withdrawal can be found at https://www.elsevier.com/about/our-business/policies/article-withdrawal.

Attention Flows: Analyzing and Comparing Attention Mechanisms in Language Models.

Advances in language modeling have led to the development of deep attention-based models that are performant across a wide variety of natural language processing (NLP) problems. These language models are typified by a pre-training process on large unlabeled text corpora and subsequently fine-tuned for specific tasks. Although considerable work has been devoted to understanding the attention mechanisms of pre-trained models, it is less understood how a model's attention mechanisms change when trained for a target NLP task. In this paper, we propose a visual analytics approach to understanding fine-tuning in attention-based language models. Our visualization, Attention Flows, is designed to support users in querying, tracing, and comparing attention within layers, across layers, and amongst attention heads in Transformer-based language models. To help users gain insight on how a classification decision is made, our design is centered on depicting classification-based attention at the deepest layer and how attention from prior layers flows throughout words in the input. Attention Flows supports the analysis of a single model, as well as the visual comparison between pre-trained and fine-tuned models via their similarities and differences. We use Attention Flows to study attention mechanisms in various sentence understanding tasks and highlight how attention evolves to address the nuances of solving these tasks.

Incidence and Risk Factors for Permanent Pacemaker Implantation Following Mitral or Aortic Valve Surgery.

BACKGROUND: Risk factors for post-operative conduction disturbances after cardiac valve surgery requiring a permanent pacemaker (PPM) are poorly characterized. OBJECTIVES: The aim of this study was to investigate the timing and risk factors for PPM implantation after mitral or aortic valve surgery. METHODS: All patients who underwent open aortic or mitral valve surgery between January 1996 and December 2014 were reviewed using New York State's mandatory hospital discharge database. Patients with prior cardiac surgery or pre-existing PPM were excluded. The primary endpoint was PPM implantation within 1 year. RESULTS: Among 77,882 patients, 63.8% (n = 49,706) underwent aortic valve replacement (AVR), 18.9% (n = 14,686) underwent mitral valve replacement (MVR), 10.5% (n = 8,219) underwent mitral valve repair (MVr), 5.4% (n = 4,202) underwent AVR plus MVR, and 1.4% (n = 1,069) underwent AVR plus MVr. The 1-year PPM implantation rate was 4.5% after MVr, 6.6% after AVR, 9.3% after AVR plus MVr, 10.5% after MVR, and 13.3% after AVR plus MVR (p < 0.001). Across all groups, the majority of PPMs were implanted during the index hospitalization (79.9%). MVr was associated with the lowest risk for PPM and AVR plus MVR with the highest risk. Older age, history of arrhythmias, pre-operative conduction disturbances, and concomitant index procedures were associated with increased risk for PPM during the index hospitalization. Conversely, beyond 30 days, chronic comorbidities were associated with increased risk for PPM. CONCLUSIONS: Conduction disturbances requiring PPM remain a common adverse event after valve surgery. Identifying patients at risk for PPM will help facilitate perioperative planning and inform clinical decision making regarding post-operative rhythm surveillance.

Biatrial maze procedure versus pulmonary vein isolation for atrial fibrillation during mitral valve surgery: New analytical approaches and end points.

OBJECTIVE: To use novel statistical methods for analyzing the effect of lesion set on (long-standing) persistent atrial fibrillation (AF) in the Cardiothoracic Surgical Trials Network trial of surgical ablation during mitral valve surgery (MVS). METHODS: Two hundred sixty such patients were randomized to MVS + surgical ablation or MVS alone. Ablation was randomized between pulmonary vein isolation and biatrial maze. During 12 months postsurgery, 228 patients (88%) submitted 7949 transtelephonic monitoring (TTM) recordings, analyzed for AF, atrial flutter (AFL), or atrial tachycardia (AT). As previously reported, more ablation than MVS-alone patients were free of AF or AF/AFL at 6 and 12 months (63% vs 29%; P < .001) by 72-hour Holter monitoring, without evident difference between lesion sets (for which the trial was underpowered). RESULTS: Estimated freedom from AF/AFL/AT on any transmission trended higher after biatrial maze than pulmonary vein isolation (odds ratio, 2.31; 95% confidence interval, 0.95-5.65; P = .07) 3 to 12 months postsurgery; estimated AF/AFL/AT load (ie, proportion of TTM strips recording AF/AFL/AT) was similar (odds ratio, 0.90; 95% confidence interval, 0.57-1.43; P = .6). Within 12 months, estimated prevalence of AF/AFL/AT by TTM was 58% after MVS alone, and 36% versus 23% after pulmonary vein isolation versus biatrial maze (P < .02). CONCLUSIONS: Statistical modeling using TTM recordings after MVS in patients with (long-standing) persistent AF suggests that a biatrial maze is associated with lower AF/AFL/AT prevalence, but not a lower load, compared with pulmonary vein isolation. The discrepancy between AF/AFL/AT prevalence assessed at 2 time points by Holter monitoring versus weekly TTM suggests the need for a confirmatory trial, reassessment of definitions for failure after ablation, and validation of statistical methods for assessing atrial rhythms longitudinally.

Sex-Based Differences in Outcomes After Mitral Valve Surgery for Severe Ischemic Mitral Regurgitation: From the Cardiothoracic Surgical Trials Network.

OBJECTIVES: This study investigated sex-based differences in outcomes after mitral valve (MV) surgery for severe ischemic mitral regurgitation (SIMR). BACKGROUND: Whether differences in outcomes exist between men and women after surgery for SIMR remains unknown. METHODS: Patients enrolled in a randomized trial comparing MV replacement versus MV repair for SIMR were included and followed for 2 years. Endpoints for this analysis included all-cause mortality, major adverse cardiovascular and cerebrovascular events (MACCE) (defined as the composite of death, stroke, hospitalization for heart failure, worsening New York Heart Association functional class or MV re-operation), quality of life (QOL), functional status, and percentage of change in left ventricular end-systolic volume index (LVESVI) from baseline through 2 years. RESULTS: Of 251 patients enrolled in the trial, 96 (38.2%) were women. Compared with men, women had smaller LV volumes and effective regurgitant orifice areas (EROA) but greater EROA/left ventricular (LV) end-diastolic volume ratios. At 2 years, women had higher rates of all-cause mortality (27.1% vs. 17.4%, respectively; adjusted hazard ratio [adjHR]: 1.85; 95% confidence interval [CI]: 1.05 to 3.26; p = 0.03) and of MACCE (49.0% vs. 38.1%, respectively; adjHR: 1.58; 95% CI: 1.06 to 2.37; p = 0.02). Women also reported worse QOL and functional status at 2 years. There were no significant differences in the percentage of change over 2 years in LVESVI between women and men (adjß: -10.4; 95% CI: -23.4 to 2.6; p = 0.12). CONCLUSIONS: Women with SIMR displayed different echocardiographic features and experienced higher mortality and worse QOL after MV surgery than men. There were no significant differences in the degree of reverse LV remodeling between sexes. (Comparing the Effectiveness of Repairing Versus Replacing the Heart's Mitral Valve in People With Severe Chronic Ischemic Mitral Regurgitation [Severe Ischemic Mitral Regurgitation]; NCT00807040).

Pacemaker Implantation After Mitral Valve Surgery With Atrial Fibrillation Ablation.

BACKGROUND: The incidence of permanent pacemaker (PPM) implantation is higher following mitral valve surgery (MVS) with ablation for atrial fibrillation (AF) compared with MVS alone. OBJECTIVES: This study identified risk factors and outcomes associated with PPM implantation in a randomized trial that evaluated ablation for AF in patients who underwent MVS. METHODS: A total of 243 patients with AF and without previous PPM placement were randomly assigned to MVS alone (n = 117) or MVS + ablation (n = 126). Patients in the ablation group were further randomized to pulmonary vein isolation (PVI) (n = 62) or the biatrial maze procedure (n = 64). Using competing risk models, this study examined the association among PPM and baseline and operative risk factors, and the effect of PPM on time to discharge, readmissions, and 1-year mortality. RESULTS: Thirty-five patients received a PPM within the first year (14.4%), 29 (83%) underwent implantation during the index hospitalization. The frequency of PPM implantation was 7.7% in patients randomized to MVS alone, 16.1% in MVS + PVI, and 25% in MVS + biatrial maze. The indications for PPM were similar among patients who underwent MVS with and without ablation. Ablation, multivalve surgery, and New York Heart Association functional (NYHA) functional class III/IV were independent risk factors for PPM implantation. Length of stay post-surgery was longer in patients who received PPMs, but it was not significant when adjusted for randomization assignment (MVS vs. ablation) and age (hazard ratio [HR]: 0.81; 95% confidence interval [CI]: 0.61 to 1.08; p = 0.14). PPM implantation did not increase 30-day readmission rate (HR: 1.43; 95% CI: 0.50 to 4.05; p = 0.50). The need for PPM was associated with a higher risk of 1-year mortality (HR: 3.21; 95% CI: 1.01 to 10.17; p = 0.05) after adjustment for randomization assignment, age, and NYHA functional class. CONCLUSIONS: AF ablation, multivalve surgery, and NYHA functional class III/IV were associated with an increased risk for permanent pacing. PPM implantation following MVS was associated with a significant increase in 1-year mortality. (Surgical Ablation Versus No Surgical Ablation for Patients With Atrial Fibrillation Undergoing Mitral Valve Surgery; NCT00903370).

Missed Testing Opportunities for HIV Screening and Early Diagnosis in an Urban Tertiary Care Center.

Newark, New Jersey, is disproportionally affected by HIV with one of the highest prevalence rates in the United States. Rutgers New Jersey Medical School is a major healthcare provider to Newark's underserved population and has implemented a HIV testing program that can diagnose and link newly diagnosed individuals to care. We conducted a retrospective chart review of all new patients seen in the Infectious Disease Practice from January 1, 2013, to December 31, 2014, to determine the proportion of patients with a missed testing opportunity (MTO) (patients with a new HIV diagnosis with an encounter at the institution in the 1 year prior to their first appointment). 117 newly diagnosed patients were identified. 36 (31%) had at least one MTO. A total of 34 (29%) of newly diagnosed patients had AIDS at presentation and 17% had CD4 counts of 50 cells/µL (p value 0.5). The two most common locations of a missed testing opportunity were the hospital ED (45%) and subspecialty clinics (37%). This study demonstrates that, even in a high prevalence institution with HIV counseling, testing, and referral service, HIV screening is lacking at multiple points of care and patients are missing opportunities for earlier diagnosis and treatment.