Digital image analysis and machine learning-assisted prediction of neoadjuvant chemotherapy response in triple-negative breast cancer.

BACKGROUND: Pathological complete response (pCR) is associated with favorable prognosis in patients with triple-negative breast cancer (TNBC). However, only 30-40% of TNBC patients treated with neoadjuvant chemotherapy (NAC) show pCR, while the remaining 60-70% show residual disease (RD). The role of the tumor microenvironment in NAC response in patients with TNBC remains unclear. In this study, we developed a machine learning-based two-step pipeline to distinguish between various histological components in hematoxylin and eosin (H&E)-stained whole slide images (WSIs) of TNBC tissue biopsies and to identify histological features that can predict NAC response. METHODS: H&E-stained WSIs of treatment-naïve biopsies from 85 patients (51 with pCR and 34 with RD) of the model development cohort and 79 patients (41 with pCR and 38 with RD) of the validation cohort were separated through a stratified eightfold cross-validation strategy for the first step and leave-one-out cross-validation strategy for the second step. A tile-level histology label prediction pipeline and four machine-learning classifiers were used to analyze 468,043 tiles of WSIs. The best-trained classifier used 55 texture features from each tile to produce a probability profile during testing. The predicted histology classes were used to generate a histology classification map of the spatial distributions of different tissue regions. A patient-level NAC response prediction pipeline was trained with features derived from paired histology classification maps. The top graph-based features capturing the relevant spatial information across the different histological classes were provided to the radial basis function kernel support vector machine (rbfSVM) classifier for NAC treatment response prediction. RESULTS: The tile-level prediction pipeline achieved 86.72% accuracy for histology class classification, while the patient-level pipeline achieved 83.53% NAC response (pCR vs. RD) prediction accuracy of the model development cohort. The model was validated with an independent cohort with tile histology validation accuracy of 83.59% and NAC prediction accuracy of 81.01%. The histological class pairs with the strongest NAC response predictive ability were tumor and tumor tumor-infiltrating lymphocytes for pCR and microvessel density and polyploid giant cancer cells for RD. CONCLUSION: Our machine learning pipeline can robustly identify clinically relevant histological classes that predict NAC response in TNBC patients and may help guide patient selection for NAC treatment.

Thermoreversible Reverse-Phase-Shift Foam for Treatment of Noncompressible Torso Hemorrhage.

BACKGROUND: Noncompressible torso hemorrhage (NCTH) is a leading cause of traumatic exsanguination, requiring emergent damage control surgery performed by a highly trained surgeon in a sterile operating environment. A self-expanding, intraabdominally deployed, thermoreversible foam is one proposed method to potentially task shift temporizing hemostasis to earlier providers and additional settings. The purpose of this study was to assess the feasibility of using Fast Onset Abdominal Management (FOAM) in a lethal swine model of NCTH. METHODS: This was a proof-of-concept study comparing FOAM intervention in large Yorkshire swine to historical control animals in the established Ross-Burns model of NCTH. After animal preparation, a Grade IV liver laceration was surgically induced, followed by a free bleed period of 10 min. FOAM was then deployed to a goal intraabdominal pressure of 60 mm Hg for 5 min, followed by a total 60-min observation period following injury. RESULTS: At the end of the experiment, the FOAM agent was found to be distributed throughout the peritoneal cavity in all animals, without signs of iatrogenic injury. The FOAM group demonstrated a significantly higher mean arterial pressure compared with historical controls and a trend toward improved survival: 82% (9/11) compared with 50% for controls (7/14; P = 0.082). CONCLUSIONS: This is the first study to describe the use of a thermoresponsive foam to manage NCTH and successfully demonstrated proof-of-concept feasibility of FOAM deployment. These results provide strong support for future, higher-powered studies to confirm improved survival with this novel intervention.

Multiple conformational states of the HPK1 kinase domain in complex with sunitinib reveal the structural changes accompanying HPK1 trans-regulation.

Hematopoietic progenitor kinase 1 (HPK1 or MAP4K1) is a Ser/Thr kinase that operates via the c-Jun N-terminal kinase (JNK) and extracellular signal-regulated kinase (ERK) signaling pathways to dampen the T-cell response and antitumor immunity. Accordingly, selective HPK1 inhibition is considered a means to enhance antitumor immunity. Sunitinib, a multi-receptor tyrosine kinase (RTK) inhibitor approved for the management of gastrointestinal stromal tumors (GISTs), renal cell carcinoma (RCC), and pancreatic cancer, has been reported to inhibit HPK1 in vitro In this report, we describe the crystal structures of the native HPK1 kinase domain in both nonphosphorylated and doubly phosphorylated states, in addition to a double phosphomimetic mutant (T165E,S171E), each complexed with sunitinib at 2.17-3.00-Å resolutions. The native nonphosphorylated cocrystal structure revealed an inactive dimer in which the activation loop of each monomer partially occupies the ATP- and substrate-binding sites of the partner monomer. In contrast, the structure of the protein with a doubly phosphorylated activation loop exhibited an active kinase conformation with a greatly reduced monomer-monomer interface. Conversely, the phosphomimetic mutant cocrystal structure disclosed an alternative arrangement in which the activation loops are in an extended domain-swapped configuration. These structural results indicate that HPK1 is a highly dynamic kinase that undergoes trans-regulation via dimer formation and extensive intramolecular and intermolecular remodeling of the activation segment.

A CD3-bispecific molecule targeting P-cadherin demonstrates T cell-mediated regression of established solid tumors in mice.

Strong evidence exists supporting the important role T cells play in the immune response against tumors. Still, the ability to initiate tumor-specific immune responses remains a challenge. Recent clinical trials suggest that bispecific antibody-mediated retargeted T cells are a promising therapeutic approach to eliminate hematopoietic tumors. However, this approach has not been validated in solid tumors. PF-06671008 is a dual-affinity retargeting (DART®)-bispecific protein engineered with enhanced pharmacokinetic properties to extend in vivo half-life, and designed to engage and activate endogenous polyclonal T cell populations via the CD3 complex in the presence of solid tumors expressing P-cadherin. This bispecific molecule elicited potent P-cadherin expression-dependent cytotoxic T cell activity across a range of tumor indications in vitro, and in vivo in tumor-bearing mice. Regression of established tumors in vivo was observed in both cell line and patient-derived xenograft models engrafted with circulating human T lymphocytes. Measurement of in vivo pharmacodynamic markers demonstrates PF-06671008-mediated T cell activation, infiltration and killing as the mechanism of tumor inhibition.

Versatile technique for assessing thickness of 2D layered materials by XPS.

X-ray photoelectron spectroscopy (XPS) has been utilized as a versatile method for thickness characterization of various two-dimensional (2D) films. Accurate thickness can be measured simultaneously while acquiring XPS data for chemical characterization of 2D films having thickness up to approximately 10 nm. For validating the developed technique, thicknesses of few-layer graphene (FLG), MoS2 and amorphous boron nitride (a-BN) layer, produced by microwave plasma chemical vapor deposition (MPCVD), plasma enhanced chemical vapor deposition (PECVD), and pulsed laser deposition (PLD) respectively, were accurately measured. The intensity ratio between photoemission peaks recorded for the films (C 1s, Mo 3d, B 1s) and the substrates (Cu 2p, Al 2p, Si 2p) is the primary input parameter for thickness calculation, in addition to the atomic densities of the substrate and the film, and the corresponding electron attenuation length (EAL). The XPS data was used with a proposed model for thickness calculations, which was verified by cross-sectional transmission electron microscope (TEM) measurement of thickness for all the films. The XPS method determines thickness values averaged over an analysis area which is orders of magnitude larger than the typical area in cross-sectional TEM imaging, hence provides an advanced approach for thickness measurement over large areas of 2D materials. The study confirms that the versatile XPS method allows rapid and reliable assessment of the 2D material thickness and this method can facilitate in tailoring growth conditions for producing very thin 2D materials effectively over a large area. Furthermore, the XPS measurement for a typical 2D material is non-destructive and does not require special sample preparation. Therefore, after XPS analysis, exactly the same sample can undergo further processing or utilization.

Slow creep in soft granular packings.

Transient creep mechanisms in soft granular packings are studied numerically using a constant pressure and constant stress simulation method. Rapid compression followed by slow dilation is predicted on the basis of a logarithmic creep phenomenon. Characteristic scales of creep strain and time exhibit a power-law dependence on jamming pressure, and they diverge at the jamming point. Microscopic analysis indicates the existence of a correlation between rheology and nonaffine fluctuations. Localized regions of large strain appear during creep and grow in magnitude and size at short times. At long times, the spatial structure of highly correlated local deformation becomes time-invariant. Finally, a microscale connection between local rheology and local fluctuations is demonstrated in the form of a linear scaling between granular fluidity and nonaffine velocity.

Electroreflectance imaging of gold-H3PO4 supercapacitors. Part I: experimental methodology.

Electroreflectance microscopy is demonstrated as a high-resolution, non-contact method to image dynamic charge distribution in integrated microsupercapacitor structures during fast voltage cycling. Electroreflectance camera images of a gold electrode H3PO4 polymer electrolyte microsupercapacitor reveal time varying charge distribution with submicron spatial resolution, millisecond time resolution, and electroreflectance resolution on the order of 500 nC cm(-2). A model describing changes in the metal electrode's optical constants as a function of free electron concentration shows good agreement with measured electroreflectance. The proposed method can be used for sensitive, non-contact measurements of charge spatial distribution, and defect and performance characterization in electrode-electrolyte microdevices.

Electroreflectance imaging of gold-H3PO4 supercapacitors. Part II: microsupercapacitor ageing characterization.

This microsupercapacitor ageing study demonstrates the usefulness of the electroreflectance technique by quantifying local charge accumulation. Two separate devices with interdigitated electrodes were evaulated over a period of 4.1 million charge/discharge cycles. The key results are spatial mapping of charge accumulation in the gold electrodes derived from variation in the observed electrode reflectance. The nominal device exhibited little change in spatial distribution throughout the ageing cycle and serves as a comparison for the test device, which exhibited some nonuniform charge accumulation behavior. Further, an accelerated ageing test was completed by applying increasing voltage pulses up to 1.46 V to the device. Visual evidence of electrode ageing emerged in the reflectance distribution. An equivalent circuit model was developed to assess the evolution of individual circuit elements that correlate to the physical causes of ageing.

Engineering the electronic bandgaps and band edge positions in carbon-substituted 2D boron nitride: a first-principles investigation.

Modification of graphene to open a robust gap in its electronic spectrum is essential for its use in field effect transistors and photochemistry applications. Inspired by recent experimental success in the preparation of homogeneous alloys of graphene and boron nitride (BN), we consider here engineering the electronic structure and bandgap of C2xB1-xN1-x alloys via both compositional and configurational modification. We start from the BN end-member, which already has a large bandgap, and then show that (a) the bandgap can in principle be reduced to about 2 eV with moderate substitution of C (x < 0.25); and (b) the electronic structure of C2xB1-xN1-x can be further tuned not only with composition x, but also with the configuration adopted by C substituents in the BN matrix. Our analysis, based on accurate screened hybrid functional calculations, provides a clear understanding of the correlation found between the bandgap and the level of aggregation of C atoms: the bandgap decreases most when the C atoms are maximally isolated, and increases with aggregation of C atoms due to the formation of bonding and anti-bonding bands associated with hybridization of occupied and empty defect states. We determine the location of valence and conduction band edges relative to vacuum and discuss the implications on the potential use of 2D C2xB1-xN1-x alloys in photocatalytic applications. Finally, we assess the thermodynamic limitations on the formation of these alloys using a cluster expansion model derived from first-principles.

Inhibition of cholesteryl ester transfer protein by anacetrapib does not impair the anti-inflammatory properties of high density lipoprotein.

Cholesteryl ester transfer protein (CETP) is a target of therapeutic intervention for coronary heart disease. Anacetrapib, a potent inhibitor of CETP, has been shown to reduce LDL-cholesterol by 40% and increase HDL-cholesterol by 140% in patients, and is currently being evaluated in a phase III cardiovascular outcomes trial. HDL is known to possess anti-inflammatory properties, however with such large increases in HDL-cholesterol, it is unclear whether CETP inhibition perturbs HDL functionality such as anti-inflammatory effects on endothelial cells. The purpose of the present study was to determine whether CETP inhibition by anacetrapib affects the anti-inflammatory properties of HDL. HDL was isolated from either hamsters treated with vehicle or anacetrapib for 2weeks, or from normal human subjects treated either placebo, 20mg, or 150mg anacetrapib daily for 2weeks. Anacetrapib treatment increased plasma HDL cholesterol levels by 65% and between 48 and 82% in hamsters and humans, respectively. Pre-incubation of human aortic endothelial cells with HDL isolated from both control and anacetrapib treated hamsters suppressed TNFα induced expression of vascular cell adhesion molecule-1 (VCAM-1), intercellular adhesion molecule-1 (ICAM-1) and E-selectin. Similar results were obtained with human HDL samples pre and post treatment with placebo or anacetrapib. Further, HDL inhibited TNFα-induced MCP-1 secretion, monocyte adhesion and NF-κB activation in endothelial cells, and the inhibition was similar between control and anacetrapib treated groups. These studies demonstrate that anacetrapib treatment does not impair the ability of HDL to suppress an inflammatory response in endothelial cells.

Sequential Bayesian-optimized graphene synthesis by direct solar-thermal chemical vapor deposition.

This work reports the use of a high-flux solar simulator that mimics the solar spectrum and a cold-wall CVD reactor to demonstrate the feasibility of utilizing a renewable energy resource in synthesizing graphene under various conditions. A parametric study of process parameters was conducted using a probabilistic approach. Gaussian process regression serves as a surrogate to establish a prior for Bayesian optimization, and an information acquisition function is employed to identify conditions that yield high-quality products. Backscattered electron images and Raman mapping were used to assess the effects of growth conditions on graphene characteristic sizes, film quality, and uniformity. We report the synthesis of high-quality single-layer graphene (SLG) and AB-stacked bilayer graphene films in a one-step, short-time process with [Formula: see text] ratios of 0.21 and 0.14, respectively. Electron diffraction analysis shows peak intensities that resemble SLG and AB-bilayer graphene with up to 5 and 20 [Formula: see text]m grain sizes, respectively. The optical transmissivities of SLG and AB-bilayer graphene fall between 0.959-0.977 and 0.929-0.953, whereas the sheet resistances measured by a 4-point probe with 1 mm spacing are 15.5 ± 4.6 and 3.4 ± 1.5 k[Formula: see text]/sq, respectively. Further scale-up of the optimized graphene growth area was achieved by flattening the insolation profile, leading to spatial uniformity up to 13 mm in radius. Direct solar capture for CVD synthesis enable a practical and sustainable option for synthesizing graphene films applicable for photonic and electronic applications.

Local Thermochemical Mechanisms in Direct Solar Graphite Synthesis from Methane.

Methane pyrolysis is known to produce hydrogen and solid carbon in a variety of thermal processes. However, the generated carbon product typically belongs to a low-value amorphous type. Here, we elucidate the thermochemical mechanisms of a reaction that produces high-quality graphite via direct solar methane pyrolysis on a porous substrate. By comparing graphite deposition rates and local reaction zone temperatures of exposed and shadowed regions from the same experiment, we clarify the effects of thermolysis and photolysis in this emission-free process that both decarbonizes a fuel and produces a critical material for the sustainable energy transition.

Multivariable modeling: A retrospective cohort study exploring the impact of socioeconomic status and distance to a rural academic center on all-cause preterm delivery.

OBJECTIVE: Hospital-based labor and delivery units are closing at increasing rates in the rural US, significantly impacting maternal and newborn health. The objective of this study to determine if rurality-measured as distance from the hospital-and socioeconomic status-measured as insurance payor-impact both spontaneous and medically indicated preterm birth incidence at a single rural academic institution. METHODS: This was a retrospective cohort study using electronic medical records of patients with singleton pregnancies delivering at a single rural academic institution between 2016-2018. The primary outcome was preterm delivery (PTD) and secondary outcomes included low birth weight (LBW) and intensive care nursery (ICN) admission. The primary exposures included (1) travel time from a patient's address to the hospital and (2) insurance carrier as a proxy for socioeconomic status. Bivariate analyses indicated that travel time, insurance status, race, ethnicity, marital status, number of prenatal visits, gravida and para, and smoking status were significant predictors of all outcomes (LBW, ICN admission, and PTD). Therefore, these predictors were included in the multivariable logistic models. RESULTS: Within the multivariable logistic model, patients traveling 1-1.5 hours had approximately twice the odds of PTD (Odds Ratio, OR: 2.08, 95% Confidence Interval CI, 1.32, 3.29, p = .002), birth of a LBW neonate (OR: 2.15; 95% CI: 1.29-3.58, p = .005), and infant admission to the ICN (OR 1.83, 95% CI: 1.22-2.76, p = .004) compared to patients traveling under 30 minutes,. Insurance carrier status was not associated with increased odds of PTD, LBW, or ICN admission. CONCLUSION: Patients living 1-to-1.5 hours from the hospital had an increased risk for LBW, ICN admission, and PTD, despite living in zip codes with less social deprivation than zip codes further away from the hospital.

A novel, reverse-phase-shifting, thermoreversible foaming hydrogel containing antibiotics for the treatment of thermal burns in a swine model - A pilot study.

BACKGROUND: This study compared a novel topical hydrogel burn dressing (CI-PRJ012) to standard of care (silver sulfadiazine) and to untreated control in a swine thermal burn model, to assess for wound healing properties both in the presence and absence of concomitant bacterial inoculation. METHODS: Eight equal burn wounds were created on six Yorkshire swine. Half the wounds were randomized to post-burn bacterial inoculation. Wounds were subsequently randomized to three treatments groups: no intervention, CI-PRJ012, or silver sulfadiazine cream. At study end, a blinded pathologist evaluated wounds for necrosis and bacterial colonization. RESULTS: When comparing CI-PRJ012 and silver sulfadiazine cream to no treatment, both agents significantly reduced the amount of necrosis and bacteria at 7 days after wound creation (p < 0.01, independently for both). Further, CI-PRJ012 was found to be significantly better than silver sulfadiazine (p < 0.02) in reducing bacterial colonization. For wound necrosis, no significant difference was found between silver sulfadiazine cream and CI-PRJ012 (p = 0.33). CONCLUSIONS: CI-PRJ012 decreases necrosis and bacterial colonization compared to no treatment in a swine model. CI-PRJ012 appeared to perform comparably to silver sulfadiazine. CI-PRJ012, which is easily removed with the application of room-temperature water, may provide clinical advantages over silver sulfadiazine.

Optical properties of ordered carbon nanotube arrays grown in porous anodic alumina templates.

We have synthesized ordered carbon nanotube (CNT) arrays in porous anodic alumina (PAA) matrix, and have characterized their total optical reflectance and bi-directional reflectance distribution function after each processing step of the microwave plasma chemical vapor deposition process (MPCVD). For a PAA sample without CNT growth, the reflectance shows an oscillating pattern with wavelength that agrees reasonably with a multilayer model. During the MPCVD process, heating the sample significantly reduces the reflectance by 30-40%, the plasma treatment reduces the reflectance by another 5-10%, and the CNT growth further reduces the reflectance by 2-3%. After an atomic layer deposition (ALD) process, the reflectance increases to the embedded CNT arrays. After etching and exposure of CNT tips, the reflectance almost returns to the original pattern with slightly higher reflectance. Bi-directional reflectance distribution function (BRDF) measurements show that the CNT-PAA surface is quite specular as indicated by a large lobe at the specular angle, while the secondary lobe can be attributed to surface roughness.

Digital image analysis and machine learning-assisted prediction of neoadjuvant chemotherapy response in triple-negative breast cancer.

Background: Pathological complete response (pCR) is associated with favorable prognosis in patients with triple-negative breast cancer (TNBC). However, only 30-40% of TNBC patients treated with neoadjuvant chemotherapy (NAC) show pCR, while the remaining 60-70% show residual disease (RD). The role of the tumor microenvironment (TME) in NAC response in patients with TNBC remains unclear. In this study, we developed a machine learning-based two-step pipeline to distinguish between various histological components in hematoxylin and eosin (H&E)-stained whole slide images (WSIs) of TNBC tissue biopsies and to identify histological features that can predict NAC response. Methods: H&E-stained WSIs of treatment-naïve biopsies from 85 patients (51 with pCR and 34 with RD) were separated through a stratified 8-fold cross validation strategy for the first step and leave one out cross validation strategy for the second step. A tile-level histology label prediction pipeline and four machine learning classifiers were used to analyze 468,043 tiles of WSIs. The best-trained classifier used 55 texture features from each tile to produce a probability profile during testing. The predicted histology classes were used to generate a histology classification map of the spatial distributions of different tissue regions. A patient-level NAC response prediction pipeline was trained with features derived from paired histology classification maps. The top graph-based features capturing the relevant spatial information across the different histological classes were provided to the radial basis function kernel support vector machine (rbfSVM) classifier for NAC treatment response prediction. Results: The tile-level prediction pipeline achieved 86.72% accuracy for histology class classification, while the patient-level pipeline achieved 83.53% NAC response (pCR vs. RD) prediction accuracy. The histological class pairs with the strongest NAC response predictive ability were tumor and tumor tumor-infiltrating lymphocytes for pCR and microvessel density and polyploid giant cancer cells for RD. Conclusion: Our machine learning pipeline can robustly identify clinically relevant histological classes that predict NAC response in TNBC patients and may help guide patient selection for NAC treatment.

Predicting neoadjuvant treatment response in triple-negative breast cancer using machine learning.

Background: Neoadjuvant chemotherapy (NAC) is the standard treatment for early-stage triple negative breast cancer (TNBC). The primary endpoint of NAC is a pathological complete response (pCR). NAC results in pCR in only 30%â€"40% of TNBC patients. Tumor-infiltrating lymphocytes (TILs), Ki67 and phosphohistone H3 (pH3) are a few known biomarkers to predict NAC response. Currently, systematic evaluation of the combined value of these biomarkers in predicting NAC response is lacking. In this study, the predictive value of markers derived from H&E and IHC stained biopsy tissue was comprehensively evaluated using a supervised machine learning (ML)-based approach. Identifying predictive biomarkers could help guide therapeutic decisions by enabling precise stratification of TNBC patients into responders and partial or non-responders. Methods: Serial sections from core needle biopsies (n=76) were stained with H&E, and immunohistochemically for the Ki67 and pH3 markers, followed by whole slide image (WSI) generation. The resulting WSI triplets were co-registered with H&E WSIs serving as the reference. Separate mask region-based CNN (MRCNN) models were trained with annotated H&E, Ki67 and pH3 images for detecting tumor cells, stromal and intratumoral TILs (sTILs and tTILs), Ki67 + , and pH3 + cells. Top image patches with a high density of cells of interest were identified as hotspots. Best classifiers for NAC response prediction were identified by training multiple ML models, and evaluating their performance by accuracy, area under curve, and confusion matrix analyses. Results: Highest prediction accuracy was achieved when hotspot regions were identified by tTIL counts and each hotspot was represented by measures of tTILs, sTILs, tumor cells, Ki67 + , and pH3 + features. Regardless of the hotspot selection metric, a complementary use of multiple histological features (tTILs, sTILs) and molecular biomarkers (Ki67 and pH3) resulted in top ranked performance at the patient level. Conclusions: Overall, our results emphasize that prediction models for NAC response should be based on biomarkers in combination rather than in isolation. Our study provides compelling evidence to support the use of ML-based models to predict NAC response in patients with TNBC.

Predicting Neoadjuvant Treatment Response in Triple-Negative Breast Cancer Using Machine Learning.

BACKGROUND: Neoadjuvant chemotherapy (NAC) is the standard treatment for early-stage triple negative breast cancer (TNBC). The primary endpoint of NAC is a pathological complete response (pCR). NAC results in pCR in only 30-40% of TNBC patients. Tumor-infiltrating lymphocytes (TILs), Ki67 and phosphohistone H3 (pH3) are a few known biomarkers to predict NAC response. Currently, systematic evaluation of the combined value of these biomarkers in predicting NAC response is lacking. In this study, the predictive value of markers derived from H&E and IHC stained biopsy tissue was comprehensively evaluated using a supervised machine learning (ML)-based approach. Identifying predictive biomarkers could help guide therapeutic decisions by enabling precise stratification of TNBC patients into responders and partial or non-responders. METHODS: Serial sections from core needle biopsies (n = 76) were stained with H&E and immunohistochemically for the Ki67 and pH3 markers, followed by whole-slide image (WSI) generation. The serial section stains in H&E stain, Ki67 and pH3 markers formed WSI triplets for each patient. The resulting WSI triplets were co-registered with H&E WSIs serving as the reference. Separate mask region-based CNN (MRCNN) models were trained with annotated H&E, Ki67 and pH3 images for detecting tumor cells, stromal and intratumoral TILs (sTILs and tTILs), Ki67+, and pH3+ cells. Top image patches with a high density of cells of interest were identified as hotspots. Best classifiers for NAC response prediction were identified by training multiple ML models and evaluating their performance by accuracy, area under curve, and confusion matrix analyses. RESULTS: Highest prediction accuracy was achieved when hotspot regions were identified by tTIL counts and each hotspot was represented by measures of tTILs, sTILs, tumor cells, Ki67+, and pH3+ features. Regardless of the hotspot selection metric, a complementary use of multiple histological features (tTILs, sTILs) and molecular biomarkers (Ki67 and pH3) resulted in top ranked performance at the patient level. CONCLUSIONS: Overall, our results emphasize that prediction models for NAC response should be based on biomarkers in combination rather than in isolation. Our study provides compelling evidence to support the use of ML-based models to predict NAC response in patients with TNBC.

Thermoreversible Reverse-Phase-Shift Foam for Treatment of Noncompressible Torso Hemorrhage, a Safety Trial in a Porcine Model.

INTRODUCTION: Noncompressible torso hemorrhage is the leading cause of exsanguination on the battlefield. A self-expanding, intraperitoneal deployed, thermoreversible foam has been developed that can be easily administered by a medic in austere settings to temporarily tamponade noncompressible torso hemorrhage. The purpose of this study was to assess the long-term safety and physical characteristics of using Fast Onset Abdominal Management (FOAM; Critical Innovations LLC) in swine. MATERIALS AND METHODS: Yorkshire swine (40-60 kg) were sedated, intubated, and placed on ventilatory support. An external jugular catheter was placed for sampling of blood. Continuous heart rate, temperature, saturation of peripheral oxygen, end-tidal carbon dioxide, and peak airway pressures were monitored for a 4-hour period after intervention (i.e., FOAM agent injection or a sham introducer without agent delivery). The FOAM agent was injected to obtain an intra-abdominal pressure of 60 mmHg for at least 10 minutes. After 4 hours, the animals were removed from ventilatory support and returned to their housing for a period of 7-14 days. Group size analysis was not performed, as this was a descriptive safety study. Blood samples were obtained at baseline and at 1-hour post-intervention and then on days 1, 3, 7, and 14. Euthanasia, necropsy, and harvesting of samples for histologic analysis (from kidneys, terminal ilium, liver, pancreas, stomach, spleen, and lungs) were performed upon expiration. Histologic scoring for evidence of ischemia, necrosis, and abdominal compartment sequela was blinded and reported by semi-quantitative scale (range 0-4; 0 = no change, 1 = minimal, 2 = mild, 3 = moderate, and 4 = marked). Oregon Health & Science University's Institutional Animal Care and Use Committee, as well as the U.S. Army Animal Care and Use Review Office, approved this protocol before the initiation of experiments (respectively, protocol numbers IP00003591 and MT180006.e002). RESULTS: Five animals met a priori inclusion criteria, and all of these survived to their scheduled endpoints. Two animals received sham injections of the FOAM agent (one euthanized on day 7 and one on day 14), and three animals received FOAM agent injections (one euthanized on day 7 and two on day 14). A transitory increase in creatinine and lactate was detected during the first day in the FOAM injected swine but resolved by day 3. No FOAM agent was observed in the peritoneal cavity upon necropsy at day 7 or 14. Histologic data revealed no clinically relevant differences in any organ system between intervention and control animals upon sacrifice at day 7 or 14. CONCLUSIONS: This study describes the characteristics, survival, and histological analysis of using FOAM in a porcine model. In our study, FOAM reached the desired intra-abdominal pressure endpoint while not significantly altering basic hematologic parameters, except for transient elevations of creatinine and lactate on day 1. Furthermore, there was no clinical or histological relevant evidence of ischemia, necrosis, or intra-abdominal compartment syndrome. These results provide strong support for the safety of the FOAM device and will support the design of further regulatory studies in swine and humans.

Design and Synthesis of Functionally Active 5-Amino-6-Aryl Pyrrolopyrimidine Inhibitors of Hematopoietic Progenitor Kinase 1.

Immune activating agents represent a valuable class of therapeutics for the treatment of cancer. An area of active research is expanding the types of these therapeutics that are available to patients via targeting new biological mechanisms. Hematopoietic progenitor kinase 1 (HPK1) is a negative regulator of immune signaling and a target of high interest for the treatment of cancer. Herein, we present the discovery and optimization of novel amino-6-aryl pyrrolopyrimidine inhibitors of HPK1 starting from hits identified via virtual screening. Key components of this discovery effort were structure-based drug design aided by analyses of normalized B-factors and optimization of lipophilic efficiency.