Predicting misdiagnosed adult-onset type 1 diabetes using machine learning.

AIMS: It is now understood that almost half of newly diagnosed cases of type 1 diabetes are adult-onset. However, type 1 and type 2 diabetes are difficult to initially distinguish clinically in adults, potentially leading to ineffective care. In this study a machine learning model was developed to identify type 1 diabetes patients misdiagnosed as type 2 diabetes. METHODS: In this retrospective study, a machine learning model was developed to identify misdiagnosed type 1 diabetes patients from a population of patients with a prior type 2 diabetes diagnosis. Using Ambulatory Electronic Medical Records (AEMR), features capturing relevant information on age, demographics, risk factors, symptoms, treatments, procedures, vitals, or lab results were extracted from patients' medical history. RESULTS: The model identified age, BMI/weight, therapy history, and HbA1c/blood glucose values among top predictors of misdiagnosis. Model precision at low levels of recall (10 %) was 17 %, compared to <1 % incidence rate of misdiagnosis at the time of the first type 2 diabetes encounter in AEMR. CONCLUSIONS: This algorithm shows potential for being translated into screening guidelines or a clinical decision support tool embedded directly in an EMR system to reduce misdiagnosis of adult-onset type 1 diabetes and implement effective care at the outset.

Fourier Transform Infrared Spectroscopic Imaging-Derived Collagen Content and Maturity Correlates with Stress in the Aortic Wall of Abdominal Aortic Aneurysm Patients.

Abdominal aortic aneurysm (AAA) is a degenerative disease of the aorta characterized by severe disruption of the structural integrity of the aortic wall and its major molecular constituents. From the early stages of disease, elastin in the aorta becomes highly degraded and is replaced by collagen. Questions persist as to the contribution of collagen content, quality and maturity to the potential for rupture. Here, using our recently developed Fourier transform infrared imaging spectroscopy (FT-IRIS) method, we quantified collagen content and maturity in the wall of AAA tissues in pairs of specimens with different wall stresses. CT scans of AAAs from 12 patients were used to create finite element models to estimate stress in different regions of tissue. Each patient underwent elective repair of the AAA, and two segments of the AAA tissues from anatomic regions more proximal or distal with different wall stresses were evaluated by histology and FT-IRIS after excision. For each patient, collagen content was generally greater in the tissue location with lower wall stress, which corresponded to the more distal anatomic regions. The wall stress/collagen ratio was greater in the higher stress region compared to the lower stress region (1.01 ± 1.09 vs. 0.55 ± 0.084, p = 0.02). The higher stress region also corresponded to the location with reduced intraluminal thrombus thickness. Further, collagen maturity tended to decrease with increased collagen content (p = 0.068, R = 0.38). Together, these results suggest that an increase in less mature collagen content in AAA patients does not effectively compensate for the loss of elastin in the aortic wall, and results in a reduced capability to endure wall stresses.

Tunable, biodegradable gold nanoparticles as contrast agents for computed tomography and photoacoustic imaging.

Gold nanoparticles (AuNP) have been proposed for many applications in medicine. Although large AuNP (>5.5 nm) are desirable for their longer blood circulation and accumulation in diseased tissues, small AuNP (<5.5 nm) are required for excretion via the kidneys. We present a novel platform where small, excretable AuNP are encapsulated into biodegradable poly di(carboxylatophenoxy)phosphazene (PCPP) nanospheres. These larger nanoparticles (Au-PCPP) can perform their function as contrast agents, then subsequently break down into harmless byproducts and release the AuNP for swift excretion. Homogeneous Au-PCPP were synthesized using a microfluidic device. The size of the Au-PCPP can be controlled by the amount of polyethylene glycol-polylysine (PEG-PLL) block co-polymer in the formulation. Synthesis of Au-PCPP nanoparticles and encapsulation of AuNP in PCPP were evaluated using transmission electron microscopy and their biocompatibility and biodegradability confirmed in vitro. The Au-PCPP nanoparticles were found to produce strong computed tomography contrast. The UV-Vis absorption peak of Au-PCPP can be tuned into the near infrared region via inclusion of varying amounts of AuNP and controlling the nanoparticle size. In vitro and in vivo experiments demonstrated the potential of Au-PCPP as contrast agents for photoacoustic imaging. Therefore, Au-PCPP nanoparticles have high potency as contrast agents for two imaging modalities, as well as being biocompatible and biodegradable, and thus represent a platform with potential for translation into the clinic.

Improved sensitivity of computed tomography towards iodine and gold nanoparticle contrast agents via iterative reconstruction methods.

Computed tomography is a widely used medical imaging technique that has high spatial and temporal resolution. Its weakness is its low sensitivity towards contrast media. Iterative reconstruction techniques (ITER) have recently become available, which provide reduced image noise compared with traditional filtered back-projection methods (FBP), which may allow the sensitivity of CT to be improved, however this effect has not been studied in detail. We scanned phantoms containing either an iodine contrast agent or gold nanoparticles. We used a range of tube voltages and currents. We performed reconstruction with FBP, ITER and a novel, iterative, modal-based reconstruction (IMR) algorithm. We found that noise decreased in an algorithm dependent manner (FBP > ITER > IMR) for every scan and that no differences were observed in attenuation rates of the agents. The contrast to noise ratio (CNR) of iodine was highest at 80 kV, whilst the CNR for gold was highest at 140 kV. The CNR of IMR images was almost tenfold higher than that of FBP images. Similar trends were found in dual energy images formed using these algorithms. In conclusion, IMR-based reconstruction techniques will allow contrast agents to be detected with greater sensitivity, and may allow lower contrast agent doses to be used.

Correlations between transmural mechanical and morphological properties in porcine thoracic descending aorta.

Determination of correlations between transmural mechanical and morphological properties of aorta would provide a quantitative baseline for assessment of preventive and therapeutic strategies for aortic injuries and diseases. A multimodal and multidisciplinary approach was adopted to characterize the transmural morphological properties of descending porcine aorta. Histology and multi-photon microscopy were used for describing the media layer micro-architecture in the circumferential-radial plane, and Fourier Transform infrared imaging spectroscopy was utilized for determining structural protein, and total protein content. The distributions of these quantified properties across the media thickness were characterized and their relationship with the mechanical properties from a previous study was determined. Our findings indicate that there is an increasing trend in the instantaneous Young׳s modulus (E), elastic lamella density (ELD), structural protein (SPR), total protein (TPR), and elastin and collagen circumferential percentage (ECP and CCP) from the inner towards the outer layers. Two regions with equal thickness (inner and outer halves) were determined with significantly different morphological and material properties. The results of this study represent a substantial step toward anatomical characterization of the aortic wall building blocks and establishment of a foundation for quantifying the role of microstructural components on the functionality of aorta.

Nanodisco balls: control over surface versus core loading of diagnostically active nanocrystals into polymer nanoparticles.

Nanoparticles of complex architectures can have unique properties. Self-assembly of spherical nanocrystals is a high yielding route to such systems. In this study, we report the self-assembly of a polymer and nanocrystals into aggregates, where the location of the nanocrystals can be controlled to be either at the surface or in the core. These nanospheres, when surface decorated with nanocrystals, resemble disco balls, thus the term nanodisco balls. We studied the mechanism of this surface loading phenomenon and found it to be Ca(2+) dependent. We also investigated whether excess phospholipids could prevent nanocrystal adherence. We found surface loading to occur with a variety of nanocrystal types including iron oxide nanoparticles, quantum dots, and nanophosphors, as well as sizes (10-30 nm) and shapes. Additionally, surface loading occurred over a range of polymer molecular weights (∼30-3000 kDa) and phospholipid carbon tail length. We also show that nanocrystals remain diagnostically active after loading onto the polymer nanospheres, i.e., providing contrast in the case of magnetic resonance imaging for iron oxide nanoparticles and fluorescence for quantum dots. Last, we demonstrated that a fluorescently labeled protein model drug can be delivered by surface loaded nanospheres. We present a platform for contrast media delivery, with the unusual feature that the payload can be controllably localized to the core or the surface.

Targeted delivery of vascular endothelial growth factor improves stem cell therapy in a rat myocardial infarction model.

Rebuilding of infarcted myocardium by mesenchymal stem cells (MSCs) has not been successful because of poor cell survival due in part to insufficient blood supply after myocardial infarction (MI). We hypothesize that targeted delivery of vascular endothelial growth factor (VEGF) to MI can help regenerate vasculature in support of MSC therapy in a rat model of MI. VEGF-encapsulated immunoliposomes targeting overexpressed P-selectin in MI tissue were infused by tail vein immediately after MI. One week later, MSCs were injected intramyocardially. The cardiac function loss was moderated slightly by targeted delivery of VEGF or MSC treatment. Targeted VEGF+MSC combination treatment showed highest attenuation in cardiac function loss. The combination treatment also increased blood vessel density (80%) and decreased collagen content in post-MI tissue (33%). Engraftment of MSCs in the combination treatment group was significantly increased and the engrafted cells contributed to the restoration of blood vessels. FROM THE CLINICAL EDITOR: VEGF immunoliposomes targeting myocardial infarction tissue resulted in significantly higher attenuation of cardiac function loss when used in combination with mesenchymal stem cells. MSCs were previously found to have poor ability to restore cardiac tissue, likely as a result of poor blood supply in the affected areas. This new method counterbalances that weakness by the known effects of VEGF, as demonstrated in a rat model.

Fourier transform infrared spectroscopy to quantify collagen and elastin in an in vitro model of extracellular matrix degradation in aorta.

Extracellular matrix (ECM) is a key component and regulator of many biological tissues including aorta. Several aortic pathologies are associated with significant changes in the composition of the matrix, especially in the content, quality and type of aortic structural proteins, collagen and elastin. The purpose of this study was to develop an infrared spectroscopic methodology that is comparable to biochemical assays to quantify collagen and elastin in aorta. Enzymatically degraded porcine aorta samples were used as a model of ECM degradation in abdominal aortic aneurysm (AAA). After enzymatic treatment, Fourier transform infrared (FTIR) spectra of the aortic tissue were acquired by an infrared fiber optic probe (IFOP) and FTIR imaging spectroscopy (FT-IRIS). Collagen and elastin content were quantified biochemically and partial least squares (PLS) models were developed to predict collagen and elastin content in aorta based on FTIR spectra. PLS models developed from FT-IRIS spectra were able to predict elastin and collagen content of the samples with strong correlations (RMSE of validation = 8.4% and 11.1% of the range respectively), and IFOP spectra were successfully used to predict elastin content (RMSE = 11.3% of the range). The PLS regression coefficients from the FT-IRIS models were used to map collagen and elastin in tissue sections of degraded porcine aortic tissue as well as a human AAA biopsy tissue, creating a similar map of each component compared to histology. These results support further application of FTIR spectroscopic techniques for evaluation of AAA tissues.

Targeted delivery of VEGF to treat myocardial infarction.

Noninvasive injection of pro-angiogenic compounds such as vascular endothelial growth factor (VEGF) has shown promising results in regenerating cardiac microvasculature. However, these results have failed to translate into successful clinical trials in part due to the short half-life of VEGF in circulation. Increasing the dose of VEGF may increase its availability to the target tissue, but harmful side-effects remain a concern. Encapsulating and selectively targeting VEGF to the MI border zone may circumvent these problems. Anti-P-selectin conjugated immunoliposomes containing VEGF were developed to target the infarct border zone in a rat MI model. Targeted VEGF therapy significantly improves vascularization and cardiac function after an infarction.

Targeted Delivery of VEGF after a Myocardial Infarction Reduces Collagen Deposition and Improves Cardiac Function.

The development of adjunctive therapies which attenuate adverse remodeling and improve LV function post myocardial infarction (MI) is of significant clinical interest. Previously, we have shown that targeted delivery of therapeutic vascular endothelial growth factor (VEGF) to the infarct border zone significantly increases vascular perfusion and results in improvements in LV function. In this study, we tested the hypothesis that improvements in cardiac function observed with this novel targeted drug delivery system strongly correlate with reductions in collagen deposition in the scar tissue after an MI. Rats received anti-P-selectin conjugated immunoliposomes containing VEGF immediately post-MI. Over 4 weeks, evolutionary changes in LV geometry and function were correlated with collagen deposition and infarct size quantified by Gomori's trichrome and picrosirius red staining. Targeted VEGF treated hearts showed a 37% decrease in collagen deposition in the anterior wall, as well as significant improvements in LV filling pressures. Multi-regression analysis showed that the extent of collagen deposition post MI can be predicted by a linear combination of normalized LV mass and ejection fraction. Targeted delivery of VEGF post-MI results in significant decreases in collagen deposition and adverse remodeling. Improvements in cardiac function in this model are related to degree of collagen deposition and extent of scar formation.

Fourier transform infrared spectroscopic imaging of cardiac tissue to detect collagen deposition after myocardial infarction.

Myocardial infarction often leads to an increase in deposition of fibrillar collagen. Detection and characterization of this cardiac fibrosis is of great interest to investigators and clinicians. Motivated by the significant limitations of conventional staining techniques to visualize collagen deposition in cardiac tissue sections, we have developed a Fourier transform infrared imaging spectroscopy (FT-IRIS) methodology for collagen assessment. The infrared absorbance band centered at 1338 cm(-1), which arises from collagen amino acid side chain vibrations, was used to map collagen deposition across heart tissue sections of a rat model of myocardial infarction, and was compared to conventional staining techniques. Comparison of the size of the collagen scar in heart tissue sections as measured with this methodology and that of trichrome staining showed a strong correlation (R=0.93). A Pearson correlation model between local intensity values in FT-IRIS and immuno-histochemical staining of collagen type I also showed a strong correlation (R=0.86). We demonstrate that FT-IRIS methodology can be utilized to visualize cardiac collagen deposition. In addition, given that vibrational spectroscopic data on proteins reflect molecular features, it also has the potential to provide additional information about the molecular structure of cardiac extracellular matrix proteins and their alterations.

Towards a targeted multi-drug delivery approach to improve therapeutic efficacy in breast cancer.

IMPORTANCE OF THE FIELD: Significant improvements in breast cancer treatments have resulted in a significant decrease in mortality. However, current breast cancer therapies, for example, chemotherapy, often result in high toxicity and nonspecific side effects. Other treatments, such as hormonal and antiangiogenic therapies, often have low treatment efficacy if used alone. In addition, acquired drug resistance decreases further the treatment efficacy of these therapies. Intra-tumor heterogeneity of the tumor tissue may be a major reason for the low treatment efficacy and the development of chemoresistance. Therefore, targeted multi-drug therapy is a valuable option for addressing the multiple mechanisms that may be responsible for reduced efficacy of current therapies. AREAS COVERED IN THIS REVIEW: In this article, different classes of drugs for treating breast cancer, the possible reasons for the drug resistance in breast cancer, as well as different targeted drug delivery systems are summarized. The current targeting strategies used in cancer treatment are discussed. WHAT THE READER WILL GAIN: This article considers the current state of breast cancer therapy and the possible future directions in targeted multi-drug delivery for treating breast cancer. TAKE HOME MESSAGE: A better understanding of tumor biology and physiological responses to nanoparticles, as well as advanced nanoparticle design, are needed to improve the therapeutic outcomes for treating breast cancer using nanoparticle-based targeted drug delivery systems. Moreover, selective delivery of multi-drugs to tumor tissue using targeted drug delivery systems may reduce systemic toxicity further, overcome drug resistances, and improve therapeutic efficacy in treating breast cancer.