Predicting Social Determinants of Health in Patient Navigation: Case Study.

BACKGROUND: Patient navigation (PN) programs have demonstrated efficacy in improving health outcomes for marginalized populations across a range of clinical contexts by addressing barriers to health care, including social determinants of health (SDoHs). However, it can be challenging for navigators to identify SDoHs by asking patients directly because of many factors, including patients' reluctance to disclose information, communication barriers, and the variable resources and experience levels of patient navigators. Navigators could benefit from strategies that augment their ability to gather SDoH data. Machine learning can be leveraged as one of these strategies to identify SDoH-related barriers. This could further improve health outcomes, particularly in underserved populations. OBJECTIVE: In this formative study, we explored novel machine learning-based approaches to predict SDoHs in 2 Chicago area PN studies. In the first approach, we applied machine learning to data that include comments and interaction details between patients and navigators, whereas the second approach augmented patients' demographic information. This paper presents the results of these experiments and provides recommendations for data collection and the application of machine learning techniques more generally to the problem of predicting SDoHs. METHODS: We conducted 2 experiments to explore the feasibility of using machine learning to predict patients' SDoHs using data collected from PN research. The machine learning algorithms were trained on data collected from 2 Chicago area PN studies. In the first experiment, we compared several machine learning algorithms (logistic regression, random forest, support vector machine, artificial neural network, and Gaussian naive Bayes) to predict SDoHs from both patient demographics and navigator's encounter data over time. In the second experiment, we used multiclass classification with augmented information, such as transportation time to a hospital, to predict multiple SDoHs for each patient. RESULTS: In the first experiment, the random forest classifier achieved the highest accuracy among the classifiers tested. The overall accuracy to predict SDoHs was 71.3%. In the second experiment, multiclass classification effectively predicted a few patients' SDoHs based purely on demographic and augmented data. The best accuracy of these predictions overall was 73%. However, both experiments yielded high variability in individual SDoH predictions and correlations that become salient among SDoHs. CONCLUSIONS: To our knowledge, this study is the first approach to applying PN encounter data and multiclass learning algorithms to predict SDoHs. The experiments discussed yielded valuable lessons, including the awareness of model limitations and bias, planning for standardization of data sources and measurement, and the need to identify and anticipate the intersectionality and clustering of SDoHs. Although our focus was on predicting patients' SDoHs, machine learning can have a broad range of applications in the field of PN, from tailoring intervention delivery (eg, supporting PN decision-making) to informing resource allocation for measurement, and PN supervision.

A fully automated high-throughput plasmid purification workstation for the generation of mammalian cell expression-quality DNA.

Early-stage antibody discovery and engineering typically require the cloning, expression, and screening of large numbers of proteins. Normally, DNA fragments encoding proteins of interest are cloned into extra-chromosomal plasmids that are amplified in Escherichia coli. Following purification from the bacteria, the plasmids are introduced into appropriate cells, and the expressed recombinant proteins screened for desired binding or function in a high-throughput manner. Even in a 96-well plate format, plasmid purification from E. coli is typically a labor intensive and time-consuming process. To further accelerate our existing biotherapeutic discovery workflows we designed, qualified, and enabled a fully integrated high-throughput plasmid purification and quantification workstation which we have termed AMPS (Automated Miniprep Plasmid Station). Using components from a commercially available kit, AMPS can purify plasmid preparations from twenty 96-deep-well plates of E. coli cultures, measure DNA absorbance at 260 nm, calculate plasmid concentrations, and prepare 96-deep-well plates for mammalian expression in an operator-independent manner. Plasmid yields and concentrations are equivalent to those obtained off-line. Furthermore, the quality of the DNA purified on the AMPS is equivalent to that obtained off-line in terms of DNA topology, and absence of contaminating bacterial chromosomal DNA and RNA. Most importantly, plasmids purified on the AMPS provide similar antibody titers following transfection in CHO cells as plasmids purified off-line. The AMPS bridges high-throughput E. coli colony picking capabilities typically available in an automation lab with downstream CHO expression needs and will facilitate screening of large numbers of biotherapeutics in binding and cell assay screens.

RNase H-dependent PCR enables highly specific amplification of antibody variable domains from single B-cells.

Immunization-based antibody discovery platforms require robust and effective protocols for the amplification, cloning, expression, and screening of antibodies from large numbers of B-cells in order to effectively capture the diversity of an experienced Ig-repertoire. Multiplex PCR using a series of forward and reverse primers designed to recover antibodies from a range of different germline sequences is challenging because primer design requires the recovery of full length antibody sequences, low starting template concentrations, and the need for all the primers to function under the same PCR conditions. Here we demonstrate several advantages to incorporating RNase H2-dependent PCR (rh-PCR) into a high-throughput, antibody-discovery platform. Firstly, rh-PCR eliminated primer dimer synthesis to below detectable levels, thereby eliminating clones with a false positive antibody titer. Secondly, by increasing the specificity of PCR, the rh-PCR primers increased the recovery of cognate antibody variable regions from single B-cells, as well as downstream recombinant antibody titers. Finally, we demonstrate that rh-PCR primers provide a more homogeneous sample pool and greater sequence quality in a Next Generation Sequencing-based approach to obtaining DNA sequence information from large numbers of cloned antibody cognate pairs. Furthermore, the higher specificity of the rh-PCR primers allowed for a better match between native antibody germline sequences and the VL/VH fragments amplified from single B-cells.

Comparing domain interactions within antibody Fabs with kappa and lambda light chains.

IgG antibodies are multi-domain proteins with complex inter-domain interactions. Human IgG heavy chains (HCs) associate with light chains (LCs) of the κ or λ isotype to form mature antibodies capable of binding antigen. The HC/LC interaction involves 4 domains: VH and CH1 from the HC and VL and CL from the LC. Human Fabs with κ LCs have been well characterized for their unfolding behaviors and demonstrate a significant level of cooperativity and stabilization when all 4 domains are intact. Very little is known regarding the thermodynamic properties of human Fabs with λ LCs. Here, we dissect the domain contributions to Fab stability for both κ and λ LC-containing Fabs. We find the cooperativity of unfolding between the constant domains, CH1/Cλ, and variable domains, VH/Vλ, within λ LC-containing Fabs is significantly weaker than that of κ LC-containing Fabs. The data suggests there may not be an evolutionary necessity for strong variable/constant domain cooperativity within λ LC-containing Fabs. After investigating the biophysical properties of Fabs with mismatched variable and constant domain subunits (e.g., VH/Vκ paired with CH1/Cλ or T cell receptor Cα/Cß), the major role of the constant domains for both κ- and λ-containing Fabs may be to reduce the hydrophobic exposure at the VH/VL interface. Even though Fabs with these non-native pairings were thermodynamically less stable, they secreted well from mammalian cells as well behaved monodisperse proteins, which was in contrast to what was observed with the VH/Vκ and VH/Vλ scFvs that secreted as a mixture of monomer and aggregates.

Convergence of eicosanoid and integrin biology: 12-lipoxygenase seeks a partner.

BACKGROUND: Integrins and enzymes of the eicosanoid pathway are both well-established contributors to cancer. However, this is the first report of the interdependence of the two signaling systems. In a screen for proteins that interacted with, and thereby potentially regulated, the human platelet-type 12-lipoxygenase (12-LOX, ALOX12), we identified the integrin ß4 (ITGB4). METHODS: Using a cultured mammalian cell model, we have demonstrated that ITGB4 stimulation leads to recruitment of 12-LOX from the cytosol to the membrane where it physically interacts with the integrin to become enzymatically active to produce 12(S)-HETE, a known bioactive lipid metabolite that regulates numerous cancer phenotypes. RESULTS: The net effect of the interaction was the prevention of cell death in response to starvation. Additionally, regulation of ß4-mediated, EGF-stimulated invasion was shown to be dependent on 12-LOX, and downstream Erk signaling in response to ITGB4 activation also required 12-LOX. CONCLUSIONS: This is the first report of an enzyme of the eicosanoid pathway being recruited to and regulated by activated ß4 integrin. Integrin ß4 has recently been shown to induce expansion of prostate tumor progenitors and there is a strong correlation between stage/grade of prostate cancer and 12-LOX expression. The 12-LOX enzymatic product, 12(S)-HETE, regulates angiogenesis and cell migration in many cancer types. Therefore, disruption of integrin ß4-12LOX interaction could reduce the pro-inflammatory oncogenic activity of 12-LOX. This report on the consequences of 12-LOX and ITGB4 interaction sets a precedent for the linkage of integrin and eicosanoid biology through direct protein-protein association.

Platelets and cancer: a casual or causal relationship: revisited.

Human platelets arise as subcellular fragments of megakaryocytes in bone marrow. The physiologic demand, presence of disease such as cancer, or drug effects can regulate the production circulating platelets. Platelet biology is essential to hemostasis, vascular integrity, angiogenesis, inflammation, innate immunity, wound healing, and cancer biology. The most critical biological platelet response is serving as "First Responders" during the wounding process. The exposure of extracellular matrix proteins and intracellular components occurs after wounding. Numerous platelet receptors recognize matrix proteins that trigger platelet activation, adhesion, aggregation, and stabilization. Once activated, platelets change shape and degranulate to release growth factors and bioactive lipids into the blood stream. This cyclic process recruits and aggregates platelets along with thrombogenesis. This process facilitates wound closure or can recognize circulating pathologic bodies. Cancer cell entry into the blood stream triggers platelet-mediated recognition and is amplified by cell surface receptors, cellular products, extracellular factors, and immune cells. In some cases, these interactions suppress immune recognition and elimination of cancer cells or promote arrest at the endothelium, or entrapment in the microvasculature, and survival. This supports survival and spread of cancer cells and the establishment of secondary lesions to serve as important targets for prevention and therapy.

Molecular engineering of silk-elastinlike polymers for matrix-mediated gene delivery: biosynthesis and characterization.

The unique advantage of genetic engineering techniques for the design and development of polymers for controlled gene delivery lies in exquisite control over polymer structure. In this article we report the biosynthesis and characterization of a series of new silk-elastinlike protein polymers (SELPs), namely, SELP415K, with larger elastin blocks per monomer unit than SELP47K previously studied for matrix-mediated gene delivery. A new cloning strategy was used, where a block of eight elastin units (8E) was integrated into the existing DNA sequence of SELP47K monomer genes using appropriate restriction endonuclease recognition sites. Following random multimerization, multimer gene segments of desired size were selected, expressed, and purified on Ni-agarose columns. The molecular weight and sequence composition of the purified SELPs were determined by MALDI-TOF and amino acid analysis, respectively. The influence of structural changes on the rheological properties of the polymers was investigated. In addition, hydrogel disks were prepared from 47K and 415K-8mer polymer solutions, and the effects of cure time and environmental conditions on the hydrogel equilibrium swelling ratio as a function of polymer composition were studied. DNA sequencing and agarose gel electrophoresis confirmed the successful cloning of the monomer gene segment of SELP415K consisting of 312 bp. Random concatemerization of SELP415K monomer gene segments resulted in a library of SELP415K multimer sequences of 6, 8, and 10 repeats respectively, each yielding a polymer with exact molecular weight and sequence. Rheometric measurements showed that both complex shear modulus (G*) and gelation point were influenced by polymer composition. Equilibrium swelling studies on hydrogel disks prepared from 47K and 415K-8mer polymer solutions showed that changes in polymer composition resulted in different gelation patterns and increased sensitivity toward changes in temperature and ionic strength but not pH. Together these results demonstrate the potential of recombinant techniques in engineering polymers with defined structures which allows the study of the structural parameters affecting matrix-mediated delivery of genes and bioactive agents.

Genetic engineering of stimuli-sensitive silkelastin-like protein block copolymers.

Differentially charged analogues of block copolymers containing repeating sequences from silk (GAGAGS) and elastin (GVGVP) were synthesized using genetic engineering techniques by replacing a valine residue with glutamic acid. The sensitivity to pH and temperature was examined at various polymer concentrations, ionic strengths, and polymer lengths. The polymers transitioned from soluble to precipitate state over narrow temperature ranges. The transition temperature T(t) (the temperature at which half-maximal spectrophotometric absorption was observed) increased with increasing pH up to pH 7.0 and leveled off above this value for the Glu-containing polymer (17E)(11). T(t) was independent of pH for the Val-containing polymer (17V)(11). It decreased with increasing ionic strength, polymer concentration, and polymer length for both polymers. These results suggest that by substituting charged amino acids for neutral amino acids at strategic locations in the polymer backbone and by control of the length of silkelastin-like block copolymers using genetic engineering techniques, it is possible to precisely control sensitivity to pH, temperature, and ionic strength.

Genetic synthesis and characterization of pH- and temperature-sensitive silk-elastinlike protein block copolymers.

The purpose of this work was to synthesize and characterize a pH- and temperature-sensitive block copolymer containing repeating sequences from silk (Gly-Ala-Gly-Ala-Gly-Ser) and elastin (Gly-Val-Gly-Val-Pro) protein. The monomer contained one repeat of silk and eight repeat units of elastin, with the first valine in one of the elastin repeats being replaced by glutamic acid. The copolymer was synthesized using genetic engineering techniques. The sensitivity of the copolymer to pH and temperature was examined at various polymer concentrations and ionic strengths. Turbidity measurements were carried out over a temperature range of 20 to 100 degrees C at various pH, concentration, and ionic strength values. The introduction of an ionizable residue (glutamic acid) rendered the copolymer sensitive to changes in pH. The transition termperature (T(t)), the temperature at which the polymer became insoluble upon increase in temperature, was modulated by changing the pH. In general, the T(t) value, was found: (1) to increase with an increase in pH, (2) to decrease with increasing ionic strength, and (3) to decrease with increasing concentration. Results of these studies suggest that by strategic placement of charged amino acids in genetically engineered silk-elastinlike protein block copolymers it is possible to precisely control sensitivity to stimuli such as pH and temperature.

Prostate Cancer - Old Problems and New Approaches. (Part II. Diagnostic and Prognostic Markers, Pathology and Biological Aspects).

Diagnostic and prognostic markers for prostatic cancer (PCa) include conventional protein markers (e.g., PAP, PSA, PSMA, PIP, OA-519, Ki-67, PCNA, TF, collagenase, and TIMP 1), angiogenesis indicator (e.g., factor VIII), neuroendocrine differentiation status, adhesion molecules (E-cadherin, integrin), bone matrix degrading products (e.g., ICPT), as well as molecular markers (e.g., PSA, PSMA, p53, 12-LOX, and MSI). Currently, only PSA is used clinically for early diagnosis and monitoring of PCa. The histological differential diagnosis of prostatic adenocarcinoma includes normal tissues such as Cowper's gland, paraganglion tissue and seminal vesicle or ejaculatory duct as well as pathological conditions such as atypical adenomatous hyperplasia, atrophy, basal cell hyperplasia and sclerosing adenosis. A common PCa is characterized by a remarkable heterogeneity in terms of its differentiation, microscopic growth patterns and biological aggressiveness. Most PCa are multifocal with signi ficant variations in tumor grade between anatomically separated tumor foci. The Gleason grading system which recognizes five major grades defined by patterns of neoplastic growth has gained almost uniform acceptance. In predicting the biologic behavior of PCa clinical and pathological stages are used as the major prognostic indicators. Among the cell proliferation and death regulators androgens are critical survival factors for normal prostate epithelial cells as well as for the androgen-dependent human prostatic cancer cells. The androgen ablation has been shown to increase the apoptotic index in prostatic cancer patients and castration also promotes apoptotic death of human prostate carcinoma grown in mice. The progression of PCa, similarly to other malignancies, is a multistep process, accompanied by genetic and epigenetic changes, involving phenomenons as adhesion, invasion and angiogenesis (without prostate specific features).

Prostate Cancer Old Problems and New Approaches. Part III. Prevention and Treatment.

In Part Three of this review, we begin with an analysis of prevention strategies for prostate cancer followed by a discussion of the clinical use of molecular techniques for the evaluation and treatment of patients with clinically localized prostate cancer. New developments in neutron and photon therapy of prostate cancer are addressed as well as the use of systemic radiotherapy for the treatment of bone metastases. Finally, we conclude with the role of hormonal therapy in the treatment of prostate cancer and the current status of development of chemo therapeutic regimens for the treatment of prostate cancer.

Mechanical Relaxation in Polyethylene Crystallized With Various Degrees of Lamellar Orientation.

The mechanical relaxation behavior of a set of well-characterized samples of polyethylene crystallized with different degrees of lamellar orientation is reported. The various samples ranged in morphology from unoriented isotropic samples to ones which showed a high degree of orientation of the b-axis along the sample growth direction. The mechanical measurements were made using a torsion pendulum apparatus of standard design, the direction of shear being normal to the b-axis for the oriented samples. The temperature range covered was from 100 to 400 °K. No definite effects attributable to orientation were observed for either the γ or ß relaxation process, whereas for the α relaxation results for G″ indicate that a slight decrease in peak height resulted from the presence of lamellar orientation, particularly on the high temperature side of this peak. Data for the real and imaginary parts of the complex shear compliance are also discussed.