The database makes the poison: How the selection of datasets in QSAR models impacts toxicant prediction of higher tier endpoints.

As the United States and the European Union continue their steady march towards the acceptance of new approach methodologies (NAMs), we need to ensure that the available tools are fit for purpose. Critics will be well-positioned to caution against NAMs acceptance and adoption if the tools turn out to be inadequate. In this paper, we focus on Quantitative Structure Activity-Relationship (QSAR) models and highlight how the training database affects quality and performance of these models. Our analysis goes to the point of asking, "are the endpoints extracted from the experimental studies in the database trustworthy, or are they false negatives/positives themselves?" We also discuss the impacts of chemistry on QSAR models, including issues with 2-D structure analyses when dealing with isomers, metabolism, and toxicokinetics. We close our analysis with a discussion of challenges associated with translational toxicology, specifically the lack of adverse outcome pathways/adverse outcome pathway networks (AOPs/AOPNs) for many higher tier endpoints. We recognize that it takes a collaborate effort to build better and higher quality QSAR models especially for higher tier toxicological endpoints. Hence, it is critical to bring toxicologists, statisticians, and machine learning specialists together to discuss and solve these challenges to get relevant predictions.

The physiological and biochemical basis of potency thresholds modeled using human estrogen receptor alpha: implications for identifying endocrine disruptors.

The endocrine system functions by interactions between ligands and receptors. Ligands exhibit potency for binding to and interacting with receptors. Potency is the product of affinity and efficacy. Potency and physiological concentration determine the ability of a ligand to produce physiological effects. The kinetic behavior of ligand-receptor interactions conforms to the laws of mass action. The laws of mass action define the relationship between the affinity of a ligand and the fraction of cognate receptors that it occupies at any physiological concentration. We previously identified the minimum ligand potency required to produce clinically observable estrogenic agonist effects via the human estrogen receptor-alpha (ERα). By examining data on botanical estrogens and dietary supplements, we demonstrated that ERα ligands with potency lower than one one-thousandth that of the primary endogenous hormone 17ß-estradiol (E2) do not produce clinically observable estrogenic effects. This allowed us to propose a Human-Relevant Potency Threshold (HRPT) for ERα ligands of 1 × 10-4 relative to E2. Here, we test the hypothesis that the HRPT for ERα arises from the receptor occupancy by the normal metabolic milieu of endogenous ERα ligands. The metabolic milieu comprises precursors to hormones, metabolites of hormones, and other normal products of metabolism. We have calculated fractional receptor occupancies for ERα ligands with potencies below and above the previously established HRPT when normal circulating levels of some endogenous ERα ligands and E2 were also present. Fractional receptor occupancy calculations showed that individual ERα ligands with potencies more than tenfold higher than the HRPT can compete for occupancy at ERα against individual components of the endogenous metabolic milieu and against mixtures of those components at concentrations found naturally in human blood. Ligands with potencies less than tenfold higher than the HRPT were unable to compete successfully for ERα. These results show that the HRPT for ERα agonism (10-4 relative to E2) proposed previously is quite conservative and should be considered strong evidence against the potential for disruption of the estrogenic pathway. For chemicals with potency 10-3 of E2, the potential for estrogenic endocrine disruption must be considered equivocal and subject to the presence of corroborative evidence. Most importantly, this work demonstrates that the endogenous metabolic milieu is responsible for the observed ERα agonist HRPT, that this HRPT applies also to ERα antagonists, and it provides a compelling mechanistic explanation for the HRPT that is grounded in basic principles of molecular kinetics using well characterized properties and concentrations of endogenous components of normal metabolism.

Kinetically-derived maximal dose (KMD) indicates lack of human carcinogenicity of ethylbenzene.

The kinetically-derived maximal dose (KMD) is defined as the maximal external dose at which kinetics are unchanged relative to lower doses, e.g., doses at which kinetic processes are not saturated. Toxicity produced at doses above the KMD can be qualitatively different from toxicity produced at lower doses. Here, we test the hypothesis that neoplastic lesions reported in the National Toxicology Program's (NTP) rodent cancer bioassay with ethylbenzene are a high-dose phenomenon secondary to saturation of elimination kinetics. To test this, we applied Bayesian modeling on kinetic data for ethylbenzene from rats and humans to estimate the Vmax and Km for the Michaelis-Menten equation that governs the elimination kinetics. Analysis of the Michaelis-Menten elimination curve generated from those Vmax and Km values indicated KMD ranges for venous ethylbenzene of 8-17 mg/L in rats and 10-18 mg/L in humans. Those venous concentrations are produced by inhalation concentrations of around 200 ppm ethylbenzene, which is well above typical human exposures. These KMD estimates support the hypothesis that neoplastic lesions seen in the NTP rodent bioassay occur secondary to saturation of ethylbenzene elimination pathways and are not relevant for human risk assessment. Thus, ethylbenzene does not pose a credible cancer risk to humans under foreseeable exposure conditions. Cancer risk assessments focused on protecting human health should avoid endpoint data from rodents exposed to ethylbenzene above the KMD range and future toxicological testing should focus on doses below the KMD range.

Supporting the right ventricle in postcardiotomy renal dysfunction: A case series.

Postcardiotomy RV dysfunction is an under-recognized cause of acute kidney injury (AKI). Insertion of a percutaneous right ventricular assist device (RVAD) reduces central venous hypertension and congestive nephropathy by augmenting cardiac output. In selected patients, percutaneous RVAD insertion may improve renal function and obviate the need for long-term dialysis.

Building cardiac surgical programs in lower-middle income countries.

Objectives: Medical care in low-income countries is often limited by inadequate resources, treatment facilities, and the necessary infrastructure for healthcare delivery. We hypothesized that the development of an independently functioning, internationally supported Kenyan cardiac surgical training program could address these issues through targeted investment. Methods: A review was conducted of the programmatic structure and clinical outcomes from January 2008 to October 2021 at Tenwek Hospital, Bomet, Kenya. Program development phases included (1) cardiovascular care provided by 1 full-time US board-certified cardiothoracic surgeon; (2) short-term volunteer surgical teams from the United States and Canada; and (3) development of a cardiothoracic residency program based on the Society of Thoracic Surgeons training curriculum. Patient demographics and outcomes were analyzed throughout each phase of program development. Results: A total of 817 cardiac procedures were performed during the study period, including 236 congenital (28.8%) and 581 adult (71.1%) procedures. Endemic rheumatic valvular heart disease predominated (581 patients, 62.3%). Local surgical team case volume grew over the study period, overtaking visiting team volume in 2019. Perioperative mortality was 2.1% and consistent between the visiting teams and the locally trained teams. Surgical training via a 3-year cardiothoracic residency is now in its fourth year, with the 2 graduates now retained as full-time teaching staff. Conclusions: Global health partnerships have the potential to address unmet needs in cardiac care within low- and middle-income countries. These data support the concept that acceptable clinical outcomes and consistent growth in volume can be achieved during the transition toward fully independent cardiac surgical care.

Reduced Morbidity and Mortality Associated With Minimally Invasive Single-vessel Coronary Artery Bypass Compared With Conventional Sternotomy.

OBJECTIVE: We aimed to describe the safety and clinical benefits of minimally invasive, nonsternotomy coronary artery bypass grafting (MICABG) using data from The Society of Thoracic Surgeons (STS) National Database. BACKGROUND: MICABG has gained popularity, owing to expected lower perioperative morbidity and shorter recovery. Despite this, concerns remain regarding anastomotic quality and the validity of proposed perioperative benefits. METHODS: We queried the STS National Database for all patients who underwent single-vessel coronary artery bypass grafting (CABG) from January 2014 to December 2016 to compare outcomes of MICABG with conventional CABG. Patients who underwent concomitant or emergent procedures were excluded. Propensity-weighted cohorts were compared by operative approach with adjustment for variability across institutions. RESULTS: Of 12,406 eligible patients, 2688 (21.7%) underwent MICABG, and 9818 (78.3%) underwent conventional CABG. Propensity weighting produced excellent balance in patient characteristics, including completeness of revascularization, body mass index, and STS predictive risk scores. MICABG was associated with significant reduction of in-hospital mortality [odds ratio (OR)=0.32, absolute reduction (AR)=0.91%, P <0.0001]; 30-day mortality (OR=0.51, AR=0.88%, P =0.001), duration of ventilation (8.62 vs 12.6 hours, P <0.0001), prolonged hospitalization (OR=0.77, AR=1.6, P =0.043), deep wound infection (OR=0.33, AR=0.68, P <0.004), postoperative transfusions (OR=0.52, AR=7.7%, P <0.0001), and STS composite morbidity (OR=0.72, AR=1.19%, P =0.008). Subgroup analysis of only off-pump left internal mammary artery-left anterior descending CABG showed similar findings. Major adverse cardiac events and graft occlusion did not differ between groups. CONCLUSIONS: MICABG is associated with lower mortality and perioperative morbidity compared with conventional sternotomy CABG. MICABG may have a role in treating single-vessel disease.

Peripherally inserted concomitant surgical right and left ventricular support, the Propella, is associated with low rates of limb ischemia, with mortality comparable with peripheral venoarterial extracorporeal membrane oxygenation.

BACKGROUND: Mechanical circulatory support effectively treats adult cardiogenic shock. Whereas cardiogenic shock confers high mortality, acute limb ischemia is a known complication of mechanical circulatory support that confers significant morbidity. We compared our novel approach to peripheral mechanical circulatory support with a conventional femoral approach, with a focus on the incidence of acute limb ischemia. METHODS: This was a retrospective cohort study of patients treated with mechanical circulatory support between January 1, 2015 and December 5, 2021 at our institution. Patients receiving any femoral peripheral venoarterial extracorporeal membrane oxygenation were compared with those receiving minimally invasive, peripherally inserted, concomitant right and left ventricular assist devices. These included the Impella 5.0 (Abiomed, Danvers, MA) left ventricular assist device and the ProtekDuo (LivaNova, London, UK) right ventricular assist device used concomitantly (Propella) approach. The primary outcome was incidence of acute limb ischemia. The baseline patient characteristics, hemodynamic data, and post-mechanical circulatory support outcomes were collected. Fisher exact test and Wilcoxon rank sum test was used for the categorical and continuous variables, respectively. Kaplan-Meier curves and log-rank test were used to estimate overall survival probabilities and survival experience, respectively. RESULTS: Fifty patients were treated with mechanical circulatory support at our institution for cardiogenic shock, with 13 patients supported with the novel Propella strategy and 37 with peripheral venoarterial extracorporeal membrane oxygenation. The baseline characteristics, including patient organ function and medical comorbidities, were similar among the groups. Nine patients suffered mortality in ≤48 hours of mechanical circulatory support initiation and were excluded. Twenty patients (69%) suffered acute limb ischemia in the peripheral venoarterial extracorporeal membrane oxygenation group; 0 patients receiving Propella suffered acute limb ischemia (P < .001). The percentages of patients surviving to discharge in peripheral venoarterial extracorporeal membrane oxygenation and Propella groups were 24% and 69%, respectively (P = .007). CONCLUSION: Patients treated with the Propella experienced a lower incidence of acute limb ischemia compared with patients treated with peripheral venoarterial extracorporeal membrane oxygenation.

Understanding and overcoming the technical challenges in using in silico predictions in regulatory decisions of complex toxicological endpoints - A pesticide perspective for regulatory toxicologists with a focus on machine learning models.

There are many challenges that must be overcome before in silico toxicity predictions are ripe for regulatory decision-making. Today, mandates in the United States of America and the European Union to avoid animal usage in toxicity testing is driving the need to consider alternative technologies, including Quantitative Structure Activity Relationship (QSAR) models, and read across approaches. However, when adopting new methods, it is critical that both new approach developers as well as regulatory users understand the strengths and challenges with these new approaches. In this paper, we identify potential sources of bias in machine learning methods specific to toxicity predictions, that may impact the overall performance of in silico models. We also discuss ways to mitigate these biases. Based on our experiences, the most prevalent sources of bias include class imbalance (differing numbers of "toxic" vs "nontoxic" compounds), limited numbers of chemicals within a particular chemistry, and biases within the studies that make up the database used for model building, as well as model evaluation biases. While this is already complex for repeated dose toxicity, in reproduction and developmental toxicity a further level of complexity is introduced by the need to evaluate effects on individual animal and litter basis (e.g., a hierarchal structure). We also discuss key considerations developers and regulators need to make when they use machine learning models to predict chemical safety. Our objective is for our paper to serve as a desk reference for model developers and regulators as they evaluate machine learning models and as they make decisions using these models.

Range of the perfluorooctanoate (PFOA) safe dose for human health: An international collaboration.

Many government agencies and expert groups have estimated a dose-rate of perfluorooctanoate (PFOA) that would protect human health. Most of these evaluations are based on the same studies (whether of humans, laboratory animals, or both), and all note various uncertainties in our existing knowledge. Nonetheless, the values of these various, estimated, safe-doses vary widely, with some being more than 100,000 fold different. This sort of discrepancy invites scrutiny and explanation. Otherwise what is the lay public to make of this disparity? The Steering Committee of the Alliance for Risk Assessment (2022) called for scientists interested in attempting to understand and narrow these disparities. An advisory committee of nine scientists from four countries was selected from nominations received, and a subsequent invitation to scientists internationally led to the formation of three technical teams (for a total of 24 scientists from 8 countries). The teams reviewed relevant information and independently developed ranges for estimated PFOA safe doses. All three teams determined that the available epidemiologic information could not form a reliable basis for a PFOA safe dose-assessment in the absence of mechanistic data that are relevant for humans at serum concentrations seen in the general population. Based instead on dose-response data from five studies of PFOA-exposed laboratory animals, we estimated that PFOA dose-rates 10-70 ng/kg-day are protective of human health.

Thoracoabdominal Normothermic Regional Perfusion for Cardiac Procurement.

In donation after circulatory death donors, warm ischemia time is a significant threat to successful cardiac transplantation. The ability to perfuse these organs during the minutes after death, until cardiac evaluation is completed to satisfaction, is crucial in limiting total warm ischemic time. Thoracoabdominal normothermic regional perfusion (TANRP) has emerged as a promising strategy for recovering and monitoring these hearts. We propose a series of clinical practice pearls that we follow for all donation after circulatory death procurements to streamline the process of setting up a TANRP circuit and ensuring all team members present at time procurement are familiar with the procedure. Bicaval cannulation is achieved via the abdomen for aortic cannulation, and via the chest for right atrial cannulation, avoiding deairing maneuvers and providing the shortest possible duration from incision to initiation of cardiopulmonary bypass. Here, we describe a series of practice techniques which we have utilized in our early experience with TANRP.

Protocolized screening effectively identifies myocardial recovery following destination therapy left ventricular assist device implantation.

BACKGROUND: Myocardial recovery following left ventricular assist device (LVAD) implantation has been of interest in transplant candidates with non-ischemic cardiomyopathy but is rare. Evidence suggests that a combination of left ventricular unloading and pharmacologic reverse remodeling is beneficial. Recovery in non-transplant candidates (i.e., destination therapy [DT]) patients is believed to be even rarer. METHODS: All DT LVADs between January 1, 2017 and November 23, 2020 were reviewed. All patients were subjected to an institutional protocol consisting of combined pharmacologic remodeling and mechanical unloading with proactive screening for recovery. The primary outcome of interest was the cumulative incidence of myocardial recovery. Baseline characteristics and operative outcomes were compared between recovered and non-recovered DT patients using non-parametric tests to identify predictive factors. RESULTS: A total of 49 patients received DT LVADs. Nine patients were identified as myocardial recovery candidates using the protocol screening criteria. Overall, 11 patients underwent formal confirmatory testing for recovery, of which 10 were deemed recovered and underwent LVAD explant, defunctionalization, or transplantation. 37.5% of patients that had a concomitant coronary artery bypass during LVAD implantation achieved recovery. An equal proportion of ischemic and non-ischemic cardiomyopathy patients achieved recovery. The cumulative incidence of myocardial recovery was 25.1% at 36 months. No factors were identified as being predictive of recovery. CONCLUSION: Myocardial recovery in DT LVAD patients can be achieved at a higher rate than previously reported. Revascularization at the time of LVAD is safe and may be beneficial. LVAD therapy may not be the final destination in these patients.

A novel approach to calculating the kinetically derived maximum dose.

The kinetically derived maximal dose (KMD) provides a toxicologically relevant upper range for the determination of chemical safety. Here, we describe a new way of calculating the KMD that is based on sound Bayesian, theoretical, biochemical, and toxicokinetic principles, that avoids the problems of relying upon the area under the curve (AUC) approach that has often been used. Our new, mathematically rigorous approach is based on converting toxicokinetic data to the overall, or system-wide, Michaelis-Menten curve (which is the slope function for the toxicokinetic data) using Bayesian methods and using the "kneedle" algorithm to find the "knee" or "elbow"-the point at which there is diminishing returns in the velocity of the Michaelis-Menten curve (or acceleration of the toxicokinetic curve). Our work fundamentally reshapes the KMD methodology, placing it within the well-established Michaelis-Menten theoretical framework by defining the KMD as the point where the kinetic rate approximates the Michaelis-Menten asymptote at higher concentrations. By putting the KMD within the Michaelis-Menten framework, we leverage existing biochemical and pharmacological concepts such as "saturation" to establish the region where the KMD is likely to exist. The advantage of defining KMD as a region, rather than as an inflection point along the curve, is that a region reflects uncertainty and clarifies that there is no single point where the curve is expected to "break;" rather, there is a region where the curve begins to taper off as it approaches the asymptote (Vmax in the Michaelis-Menten equation).

Catheter-based system for the treatment of left ventricular assist device thrombosis.

BACKGROUND: Thrombotic complications continue to pose challenges to patients on left ventricular assist device (LVAD) support. The Hoplon system was developed to administer catheter-based lytic therapy with a novel approach to embolic protection. METHODS: Two porcine non-survival surgeries were performed in which off-pump LVAD insertion was followed by injection of thrombus into the impeller, isolation of the pump using the Hoplon system, and administration of lytic therapy to the pump chamber. Successful thrombus resolution was confirmed by pathological examination of the LVAD and brain tissue after animal sacrifice. RESULTS: Limitations of the prototype design resulted in the extrusion of thrombus from around the catheter in the first animal. Subsequent device modifications resulted in the resolution of LVAD thrombus as confirmed on removal and examination of the pump. Pathological examination of the brain tissue revealed the absence of any embolic or hemorrhagic complications. CONCLUSIONS: Early animal studies suggest feasibility in restoring function to an LVAD while at the same time preventing cerebroembolic events using the Hoplon system.

Do Right Heart Hemodynamic Improvements Persist After Pulmonary Thromboendarterectomy?

The survival benefits of pulmonary thromboendarterectomy (PTE) for the treatment of chronic thromboembolic pulmonary hypertension have been well described. However, the significance of right heart hemodynamic changes and their impact on survival remains poorly understood. We sought to characterize the effects of these changes. We conducted a single center, retrospective review of 159 patients who underwent PTE between 1993 and 2015. Echocardiographic and right heart catheterization data were compared longitudinally before and after PTE in order to establish the extent of hemodynamic response to surgery. Kaplan Meier estimates were used to characterize patient survival over time. Univariable and multivariable Cox proportional hazards regression models were used to assess factors associated with long-term mortality. Among the 159 patients studied, 74 (46.5%) were male with a median age of 55 (IQR: 42-66). One-, 5-, 10-, and 15-year survival was 91.0% (95% CI: 86.6-95.6), 79.6% (73.5-86.3), 66.5% (59.2-74.7), and 56.2% (48.1-65.8). Of the 9 candidate risk factors that were evaluated, only advanced age and increased cardiopulmonary bypass time were found to be significantly associated with increased risk of mortality. Pre- and postsurgical echocardiographic imaging data, when available, revealed a median reduction in right ventricular systolic pressure of 29.0 mm Hg (P < 0.0001) and improvement of tricuspid regurgitation (P < 0.0001), both of which appeared to be sustained across long-term follow-up. Improvements in right heart hemodynamics and tricuspid valvular regurgitation persist on long term surveillance following PTE. While patient selection is often driven by the distribution of disease, close postoperative follow up may improve outcomes.

Transcarotid Approach to Placement of an Impella 5.0.

Microaxial left ventricular assist devices (mLVADs) have traditionally been placed through a transfemoral or transaxillary arterial approach. Transfemoral access is restrictive, significantly limiting postoperative patient ambulation. Transaxillary placement is preferred but not feasible in a subset of patients due to small arterial diameter or tight angulation of the thoracic outlet. Transcarotid delivery has been utilized for other cardiovascular device deployment with good success; however, this approach has not been described for mLVAD support. We present a case series of transcarotid placement of mLVADs in cases where a transaxillary and transfemoral approach was not feasible. From May 2017 to April 2019, six patients in cardiogenic shock required mLVAD support achieved via a transcarotid approach. Technical success was achieved in all patients. One patient was directly weaned from mLVAD support and two patients died on mLVAD support. Escalation to venoarterial extracorporeal membrane oxygenation (VA-ECMO) was required for three patients, two of whom subsequently died. There were no bleeding or valvular complications related to device placement, and no obvious or known neurologic complications related to mLVAD support. Transcarotid placement of mLVADs expands the utility of these devices as an alternative to traditional support strategies or prohibitive arterial anatomy; however, further study is needed to determine its efficacy.

How Does Multiple Listing Affect Lung Transplantation? A Retrospective Analysis.

The impact of multiple listing (ML) strategies on lung transplantation is unknown. Retrospective review of United Network for Organ Sharing (UNOS) registry for lung transplantation between May 1, 2005 and March 31, 2017 was performed. Characteristics of single (SL) and ML candidates were compared, and incidence density matching was used to select up to 10 controls for each case. Overall survival was evaluated using Cox regression stratified by matched sets. Nelson-Aalen estimators were used to estimate the cumulative incidence (CI) of transplant, death on the waiting list, and removal from wait-list as competing risks; Gray's test was used to compare wait list outcomes between groups. 23,445 subjects listed for lung transplant, of which 467 (2%) subjects listed at 2+ centers; 206 matched sets. There was no difference in overall survival of matched cases and controls at 1 year (ML 83.7%, SL 90.2%), 3 years (ML 63.9%, SL 68%), and 5 years (ML 51.9%, SL 49.3%) (p=0.24). The CIs of receiving a lung transplant at 2 years for ML and SL were 83.6% and 71%, respectively. Multi-listing increased the probability of receiving a transplant (p<0.001) but was not associated with waitlist mortality (p=0.13). There was no difference in post-transplant survival between ML and SL candidates (HR=0.82, p=0.32). ML was associated with a substantial increase in probability of lung transplantation, but there was no difference in overall survival, post-transplant, or wait-list mortality. Our study permits more informed decision-making for patients considering the ML strategy.

Landward zones of mangroves are sinks for both land and water borne anthropogenic debris.

The hotspots for mangrove diversity and plastic emissions from rivers overlap in Asia, however very few studies have investigated anthropogenic marine debris (AMD) pollution in these threatened coastal ecosystems. Despite Hong Kong's position at the mouth of the Pearl River, a major source of mismanaged waste in Asia, the mangroves in Hong Kong have never been extensively surveyed for AMD. Here we assessed the patterns of AMD abundance within 18 mangrove forests across Hong Kong surveying both their landward and seaward zones. We recorded and categorised, according to their material and potential uses, both the amount of debris items and area they covered, to better quantify its potential impact on the mangroves. Across Hong Kong mangroves, the average abundance of debris was 1.45 ± 0.38 (SE) items m-2, with an average coverage of 6.05 ± 1.59%. Plastic formed a high proportion of AMD accounting for 70.31% by number of items and 49.71% by area covered, followed by glass/ceramics and wood/bamboo. Disposable food packaging, fishing gear and industrial and construction related waste were the major sources of AMD we documented. On average, we recorded about six times more debris items m-2 at the landward sites than at the seaward one, but these abundances varied between the East and the West coastlines of Hong Kong. Our data confirms the hypothesis that landward areas of mangrove forests act as traps and retain marine borne debris, but they also suggest that direct dumping of waste from the land could represent a serious impact for these forests placed in between the land and the sea. More research is needed to ascertain the impact of land disposed debris on mangrove degradation, and this study strongly advocates for a cultural shift about the perception of these forests by the public.