Search

Objective We expanded the previous assessment of a mortality variable suited for real‐world evidence‐focused oncology research. Data source We used a nationwide electronic health record (EHR)‐derived de‐identified database. Data collection We included patients with at least 1 of 18 cancer types between January 1, 2011 and December 31, 2017. Patient‐level structured data (EHRs, obituaries, and Social Security Death Index) and unstructured EHR data (abstracted) were linked to generate a composite mortality variable. Study design We benchmarked sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and ±15‐day agreement against the National Death Index (NDI). Real‐world overall survival (rwOS) was estimated using the Kaplan‐Meier method. We performed sensitivity analyses using a smaller patient cohort that underwent next‐generation sequencing testing. Principal findings Compared with the NDI across 18 cancer types (overall N = 160 436): sensitivity, 83.9%‐91.5% (17/18 cancer types had sensitivity ≥85.0%); specificity, 93.5%‐99.7%; PPV, 96.3%‐98.3%; NPV, 75.0%‐98.7%; ±15‐day agreement, 95.6%‐97.6%; and median rwOS estimates ranging from 2.8% to 12.7% greater. Sensitivity analysis results (n = 17 540) were consistent with the main analysis. Conclusions Across all cancer types analyzed, this composite mortality variable showed high sensitivity, specificity, PPV, NPV, and ±15‐day agreement, and yielded median rwOS values modestly overestimated when compared to NDI‐based results.

To obtain independent navigation results for autonomous underwater vehicles (AUVs) and construct high-resolution consistent seabed maps, a particle filter-based bathymetric simultaneous localization and mapping (BSLAM) method with the mean trajectory map representation is proposed. To reduce the computational consumption, particles only keep the current estimated position of the vehicle, while all historical states of the vehicle are stored in the mean trajectory map. Using this set-up, only the weights of the particles which closed to the mean trajectory map are calculated with newly collected bathymetric data. A hierarchical clustering procedure is also discussed to identify invalid loop closures. The performance of the proposed method is validated using both the simulated data and the field data collected from sea trails. The results demonstrate that the proposed method is 50% more accurate and 50% faster than a state-of-the-art particle filter-based BSLAM method, and it has similar accuracy but 30% faster compared with a graph-based BSLAM method.

Micro- and nanoscale photonic structures and devices play important roles in the development of advanced biophotonic systems, in particular, implantable light sources for optogenetic stimulations. In this paper, we numerically investigate silicon (Si) photonics based microprobes that can achieve multi-site, multi-spectral optical excitation in the deep animal brain. On Si substrates, silicon nitride (Si3N4) based planar waveguides can deliver visible light in the deep tissue with low losses, and couple to grating emitters diffracting light in targeted brain regions. In our model, we combine near-field wave optic and far-field ray tracing simulations, showing that the designed photonic structures spectrally split blue, green and red photons into different locations in the tissue. Furthermore, by introducing dual grating components, photons at different wavelengths can be spatially separated at different depths. Therefore, these photonic probes can be used to selectively activate or inhibit specific neurons and nuclei, when expressing various corresponding light sensitive opsins. We anticipate that such device strategies can find wide applications in the design of advanced implantable photonic systems for neuroscience and neuroengineering.

Objective: We developed an implantable wireless blue micro light-emitting diode (micro-LED) device and evaluated the utility of continuous antimicrobial blue light (aBL) irradiation emitted from th...

Aims To compare the pharmacokinetic (PK) and glucodynamic (GD) characteristics of ultra rapid lispro (URLi; Eli Lilly and Company, Indianapolis, Indiana), Fiasp® (Novo Nordisk, Bagsvaerd, Denmark), Humalog® (Eli Lilly and Company) and NovoRapid® (Novo Nordisk), in patients with type 1 diabetes (T1D). Materials and Methods This was a randomized, double‐blind, four‐period, crossover study, conducted in 68 patients with T1D. Patients received the same individualized subcutaneous dose of each study drug immediately prior to a liquid test meal. For comparison, 12 healthy subjects received the same test meal. Results URLi had a significantly faster insulin absorption compared to the other insulins tested. Early half‐maximal drug concentration was reached 13 minutes after administration of URLi, which was 6 minutes faster than Fiasp, 13 minutes faster than Humalog, and 14 minutes faster than NovoRapid (all P

Background Ultra rapid lispro (URLi) is a novel insulin lispro formulation developed to more closely match physiological insulin secretion and improve postprandial glucose control. This study compared the pharmacokinetics, glucodynamics, safety, and tolerability of URLi and Humalog® in patients with type 1 diabetes mellitus (T1DM). Methods This was a phase I, two-period, randomised, double-blind, crossover glucose clamp study in younger adult (aged 18–45 years; n = 41) and elderly (aged ≥65 years; n = 39) patients with T1DM. At each dosing visit, patients received either URLi or Humalog (15 units subcutaneously) followed by a 10 h automated euglycaemic clamp procedure. Serum insulin lispro and blood glucose were measured. Results Insulin lispro appeared in serum 6 min faster, and exposure was 7.2-fold greater over the first 15 min postdose with URLi versus Humalog in both age groups. Exposure beyond 3 h postdose was 39–41% lower, and exposure duration was reduced by 72–74 min with URLi versus Humalog in both age groups. Onset of insulin action was 11–12 min faster, and insulin action was 3-fold greater over the first 30 min postdose with URLi versus Humalog in both age groups. Insulin action beyond 4 h postdose was 44–54% lower, and duration of action was reduced by 34–44 min with URLi versus Humalog in both age groups. Overall exposure and total insulin action remained similar for both treatments. URLi and Humalog were well tolerated. Conclusion In patients with T1DM, URLi showed ultra-rapid pharmacokinetics and glucodynamics, with the differences between URLi and Humalog in elderly patients mirroring those in younger adults. ClinicalTrials.gov identifier: NCT03166124. Electronic supplementary material The online version of this article (10.1007/s40262-020-00903-0) contains supplementary material, which is available to authorized users.

Background Asthma, a common respiratory disease, is harmful biological effect to our health. As a traditional Chinese medicine for asthma, Majie cataplasm could alleviate the symptoms of asthma and its compositions have immunomodulatory effects. Previous experiments showed that Majie cataplasm was an effective approach to mitigate asthma airway remodeling and had the potential to regulate Th2 cytokines of IL-5 and IL-13. Therefore, our further research focuses on the explanation about the regulatory effect of Majie cataplasm on reshaping Th1/Th2 through their related transcription factors. Methods In this experiment, the launch of asthma model was made by inducing with Ovalbumin (OVA) in C57 mice (n = 40), including 4 groups: the untreated control group (n = 10), the asthma model group (n = 10), the dexamethasone group (n = 10) and the Majie cataplasm group (n = 10). After the intervention, all groups of animals got detected for serum IgE levels, and HE staining of lung tissues was to observe and examine pathological changes. Meanwhile, we analyzed the secretion of IL-4+ T cells and IFN-γ+ T cells in spleen by flow cytometry. The expressions of transcription factor STAT6 mRNA, GATA-3 mRNA and T-bet mRNA in lung tissues was tested by PCR, and western blot had been used to detect levels of JAK2 and STAT3. Results We found that Majie cataplasm eased the content of serum IgE and lung inflammation. It could lower the increased number of IL-4+ T cells and IFN-γ+ T cells (P P P , P P P P Conclusion These data suggest that Majie cataplasm exert an anti-inflammatory effect of Th2 by rebalancing Th1/Th2 through corresponding transcription factor STAT6, GATA-3, STAT3, and T-bet, which providing a strong cornerstone for asthma control.

In practical applications, the rational operation rule derivation can lead to significant improvements in the middle and long-term joint operation of cascade hydropower stations. The key issue of actual optimal operation is to select effective attributions from the deterministic optimal operation results, however, there is still no general and mature method to solve this problem. Firstly, the joint optimal operation model of hydropower reservoirs considering backwater effects are established. Then, the dynamic programming and progressive optimality algorithm are applied to solve the joint optimal operation model and the deterministic optimization results are obtained. Finally, the grey relational analysis method is applied to select more effective factors from the obtained results as the input of a support vector machine for further operation rule derivation. The Xi Luo-du and Xiang Jia-ba cascade reservoirs in the upper Yangtze river of China are selected as a case study. The results show that the proposed method can obtain better input factors to improve the performance of SVM, and smallest value of root mean square error by the proposed method of Xi Luo-du and Xiang Jia-ba are 94.33 and 21.32, respectively. The absolute error of hydropower generation for Xi Luo-du and Xiang Jia-ba are 2.57 and 0.42, respectively. Generally, this study provides a well and promising alternative tool to guide the joint operation of hydropower reservoir systems.

Despite the surges of deep learning-based method in dynamic scene deblurring achieves good performance, the challenges still remain a lot: (a) the running speed is far from the requirement of processing; (b) there will be inevitable information loss along with the deepening of network layers, which will further lead to the deterioration of the quality of the restored pictures. To deal with these challenges, we propose a novel learning-based model. In our method, we integrate two mechanisms for the generator based on the Generative Adversarial Nets (GAN). First, we develop the Enhanced Feature Fusion (EFF) mechanism which aims at providing multi-layer feature information to assist the image restoration. We further design Feature Multi-distillation (FMD) mechanism to filter and well fuse the multi-scale feature maps. By integrating the two mechanisms, the high-level feature maps can be progressively refined and the detailed semantic information can be properly utilized. In addition, we use the double-scale discriminator architecture which could enables the network to observe the image from the perspective of local and global respectively and obtain overall information for the whole image restoration. Extensive experimental results on the GOPRO and Kohler datasets show that our method can approach comparably to the state-of-the-arts in terms of accuracy while consuming much less inference time, which demonstrates that our method acquiring a better trade off between image restoration quality and running speed.

Basal insulin Fc (BIF; LY3209590) is a novel, once-weekly, long-acting IgG Fc-fusion protein assessed for the treatment of diabetes mellitus. A 32-week study evaluating the safety and efficacy of BIF vs. degludec in persons with T2DM previously treated with a basal insulin showed HbA1c non-inferiority of BIF vs. degludec with significantly fewer hypoglycemic events (≤70 mg/dL). Here we present CGM data derived by Dexcom G6, allowing a more detailed assessment of glycemic control of BIF vs. degludec. The study included 2 dosing algorithms for BIF with different fasting glucose (FG) targets: ≤140 mg/dL (BIF-A1) and ≤120 mg/dL (BIF-A2). Degludec was titrated to FG ≤100 mg/dL. Subjects were randomized to 1 of the 3 arms. Subject (N=399) mean age was 60.2 yrs and baseline HbA1c was 8.1%. For the entire 32 weeks, the percent of 24 hrs in range, hyperglycemia and hypoglycemia was similar for the 3 arms (Figure). At Week 32, total duration of hypoglycemia was similar across 7 days post-injection for BIF-A1 and A2, showing that duration of hypoglycemia is independent of day post-injection. CGM data confirm that BIF showed similar glycemic control vs. degludec despite higher FG targets and numerically lower time in hypoglycemia. The flat pharmacokinetic profile enables near peakless insulin concentrations without an increase in hypoglycemia risk at highest exposure. Disclosure C. M. Kazda: Employee; Self; Eli Lilly and Company. J. Chien: Stock/Shareholder; Self; Eli Lilly and Company. Q. Zhang: Employee; Self; Eli Lilly and Company, Stock/Shareholder; Self; Eli Lilly and Company. E. Chigutsa: Employee; Self; Eli Lilly and Company. W. Landschulz: Employee; Self; Lilly Diabetes, Employee; Spouse/Partner; Covance Inc., Stock/Shareholder; Self; Lilly Diabetes. P. Wullenweber: None. A. Haupt: Employee; Self; Eli Lilly and Company, Stock/Shareholder; Self; Eli Lilly and Company. J. P. Frias: Consultant; Self; 89bio, Inc., Altimmune, Axcella Health Inc., Boehringer Ingelheim Pharmaceuticals, Inc., Eli Lilly and Company, Gilead Sciences, Inc., Intercept Pharmaceuticals, Inc., Merck & Co., Inc., Novo Nordisk, Pfizer Inc., Sanofi, Research Support; Self; AstraZeneca, Boehringer Ingelheim Pharmaceuticals, Inc., Bristol-Myers Squibb Company, CymaBay Therapeutics, Eli Lilly and Company, Intercept Pharmaceuticals, Inc., Janssen Pharmaceuticals, Inc., Madrigal Pharmaceuticals, Inc., Merck & Co., Inc., Novartis Pharmaceuticals Corporation, Novo Nordisk, Pfizer Inc., Sanofi, Speaker’s Bureau; Self; Merck & Co., Inc., Sanofi. Funding Eli Lilly and Company

Inorganic semiconductor-based microscale light-emitting diodes (micro-LEDs) have been widely considered the key solution to next-generation, ubiquitous lighting and display systems, with their efficiency, brightness, contrast, stability, and dynamic response superior to liquid crystal or organic-based counterparts. However, the reduction of micro-LED sizes leads to the deteriorated device performance and increased difficulties in manufacturing. Here, we report a tandem device scheme based on stacked red, green, and blue (RGB) micro-LEDs, for the realization of full-color lighting and displays. Thin-film micro-LEDs (size ∼100 μm, thickness ∼5 μm) based on III–V compound semiconductors are vertically assembled via epitaxial liftoff and transfer printing. A thin-film dielectric-based optical filter serves as a wavelength-selective interface for performance enhancement. Furthermore, we prototype arrays of tandem RGB micro-LEDs and demonstrate display capabilities. These materials and device strategies provide a viable path to advanced lighting and display systems.

We study the prediction problem for electricity usage demand, which is crucial for the stable operation of smart grid and the improvement of the users’ experience. Many prediction algorithms for electricity usage demand have been proposed to improve the prediction accuracy of the total electricity consumption in a region. However, most existing methods focus on the long- and medium-term electricity usage demand, and fail to predict the short-term fine-gained electricity usage demand. In this paper, we propose a fine-gained prediction algorithm based on the feature matching for electricity usage demand on smart meters. This paper extracts the electricity usage feature of each user from electricity usage data recorded by the smart meter, and conducts feature matching on the real-time electricity usage data. Finally, the matched features are used to predict electricity consumption. Experiment shows that our model achieves well performance on the fine-gained electricity usage data.

Diabetic retinopathy (DR), one of the leading causes of blindness, is mainly diagnosed based on the vascular pathology of the disease. Current treatment options largely focus on this aspect with mostly insufficient therapeutic long-term efficacy. Mounting evidence implicates mitochondrial dysfunction and oxidative stress in the central etiology of DR. Consequently, drug candidates that aim at normalizing mitochondrial function could be an attractive therapeutic approach. This study compared the mitoprotective compounds, idebenone and elamipretide, side-by-side against two novel short-chain quinones (SCQs) in a rat model of DR. The model effectively mimicked type 2 diabetes over 21 weeks. During this period, visual acuity was monitored by measuring optokinetic response (OKR). Vision loss occurred 5&ndash, 8 weeks after the onset of hyperglycemia. After 10 weeks of hyperglycemia, visual function was reduced by 65%. From this point, the right eyes of the animals were topically treated once daily with the test compounds. The left, untreated eye served as an internal control. Only three weeks of topical treatment significantly restored vision from 35% to 58&ndash, 80%, while visual acuity of the non-treated eyes continued to deteriorate. Interestingly, the two novel SCQs restored visual acuity better than idebenone or elamipretide. This was also reflected by protection of retinal pathology against oxidative damage, retinal ganglion cell loss, reactive gliosis, vascular leakage, and retinal thinning. Overall, mitoprotective and, in particular, SCQ-based compounds have the potential to be developed into effective and fast-acting drug candidates against DR.

Basal insulin Fc (BIF; LY3209590) is a novel, once-weekly, long-acting IgG Fc-fusion protein that is being assessed for the treatment of diabetes mellitus. The presented study evaluated the safety and efficacy of BIF compared to insulin degludec over 32 weeks in patients with T2DM previously treated with oral antidiabetic drugs and a basal insulin. The study design included 2 different dosing algorithms for BIF (BIF-A1 and BIF-A2) with two different fasting glucose (FG) targets of ≤140 mg/dL (BIF-A1) and ≤120 mg/dL (BIF-A2). Insulin degludec was titrated to a FG target of ≤100mg/dL using a modified Riddle treat-to-target algorithm. Study participants (N=399) were randomized in a 1:1:1 ratio to 1 of 3 parallel treatment groups. The average age of participants was 60.2 years, baseline HbA1c was 8.1% and duration of diabetes 14.7 years. There were no statistically significant differences in demographics or baseline characteristics across the 3 treatment groups. Both BIF groups achieved non-inferiority (non-inferiority margin = 0.4%) for the primary endpoint of HbA1c change from baseline to Week 32 with a mean±SE reduction for BIF-A1, BIF-A2 and insulin degludec of 0.6±0.1%, 0.6±0.1% and 0.7±0.1%, respectively. In line with the different fasting serum glucose (FSG) targets, insulin degludec achieved greater FSG lowering from baseline as compared to the BIF arms. Similarly, both BIF dosing groups showed significantly fewer hypoglycemic events compared to insulin degludec (all documented events as well as nocturnal events) when assessing events ≤70 mg/dL (3.9 mmol/L). Hypoglycemic events

Due to the mechanism of the instance-based classification, the feature significance plays an important role in the nearest-neighbour classification tasks. The existence of the irrelevant features would degrade the performance of these algorithms by mischoosing the nearest neighbours. However, these irrelevant features are normally inevitable in the real applications. In this paper, the evolutionary extreme learning machine (E-ELM) algorithm is employed to distinguish the feature significance for the fuzzy-rough nearest-neighbour (FRNN) method. This hybrid learning approach, entitled evolutionary extreme learning machine weighted fuzzy-rough nearest-neighbour algorithm, extracts the feature significance by integrating the parameters from the parallel input node to the output node in E-ELM. Such feature significance is transformed to implement a weighted FRNN method to perform the classification tasks. Systematic experimental results, for both dimensionality reduction and classification problems, demonstrate that the proposed approach generally outperform many state-of-the-art machine learning techniques.

How to plan a path in big and complex environments has always been a key issue. In this paper, we proposed Global Sampling and Local Search Trees(GSLST), a parallel planning algorithm. It uses Connect-RRT to grow two global sampling trees in the entire map while using Closed-Operation and Jump Point Search(JPS) to extract narrow paths and create Local Dynamic Link Trees. GSLST shows both the fastness of the sampling-based algorithm for planning in a wide range of environment and the completeness of the search based algorithm for planning in complex environment with narrow paths. The simulations demonstrate that GSLST is faster than that of sampling-based and search-based algorithms in big complex environments.

Mammogram mass detection is of great clinical significance due to its high proportion in breast cancers. The information from cross views (i.e., mediolateral oblique and cranio-caudal) is highly related and complementary, and is helpful to make comprehensive decisions. However, unlike radiologists who are able to recognize masses with reasoning ability in cross-view images, most existing methods lack the ability to reason under the guidance of domain knowledge, thus it limits the performance. In this paper, we introduce bipartite graph convolutional network to endow existing methods with cross-view reasoning ability of radiologists in mammogram mass detection. The bipartite node sets are constructed by cross-view images respectively to represent relatively consistent regions in breasts, while the bipartite edge learns to model both inherent cross-view geometric constraints and appearance similarities between correspondences. Based on the bipartite graph, the information propagates methodically through correspondences and enables spatial visual features equipped with customized cross-view reasoning ability. Experimental results on DDSM dataset demonstrate that the proposed algorithm achieves state-of-the-art performance. Besides, visual analysis shows the model has a clear physical meaning, which is helpful for radiologists in clinical interpretation.

URLi is a novel insulin lispro formulation that was developed to more closely match physiological insulin secretion. This 2-site, randomized, 2-period crossover, double-blind study evaluated the pharmacokinetics and glucodynamics during a liquid test meal, after a single 0.2 U/kg SC dose of URLi or Humalog in 13 children, 14 adolescents, 15 adults with T1D. Onset of insulin appearance was faster with URLi vs. Humalog in children (1.1 vs. 6.5 min; p=0.0002), adolescents (1.9 vs. 6.4 min; p=0.001), and adults (0.9 vs. 4.8 min; p=0.004). Early exposure (AUC0-15min) was greater with URLi vs. Humalog: 7-fold (p In summary, URLi accelerated insulin lispro absorption, reduced late exposure and early PPG following a test meal vs. Humalog in children, adolescents, and adults with T1D. Disclosure R. Aronson: None. H. Linnebjerg: Employee; Self; Eli Lilly and Company. Stock/Shareholder; Self; Eli Lilly and Company. J. Leohr: None. E.S. LaBell: Employee; Self; Eli Lilly and Company. Stock/Shareholder; Self; Eli Lilly and Company, Johnson & Johnson, Novartis AG. D.E. Coutant: Employee; Self; Eli Lilly and Company. Employee; Spouse/Partner; Eli Lilly and Company. Q. Zhang: Employee; Self; Eli Lilly and Company. Stock/Shareholder; Self; Eli Lilly and Company. T. Danne: Advisory Panel; Self; AstraZeneca, Boehringer Ingelheim International GmbH, Eli Lilly and Company, Medtronic, Novo Nordisk A/S, Sanofi. R.K. Pollom: Employee; Self; Eli Lilly and Company. Stock/Shareholder; Self; Eli Lilly and Company. Funding Eli Lilly and Company

URLi is a novel prandial insulin lispro formulation developed to more closely match physiological insulin secretion. PRONTO-T2D, a Phase 3, double-blind, treat-to-target study in pts with T2D, showed URLi, in a basal-bolus regimen with glargine or degludec, was non-inferior for HbA1c with superior PPG control vs. Humalog. Subgroup analyses were performed in pts Disclosure A.M. Chang: Employee; Self; Eli Lilly and Company. Stock/Shareholder; Self; Eli Lilly and Company. T. Hardy: Other Relationship; Self; Eli Lilly and Company. Q. Zhang: Employee; Self; Eli Lilly and Company. Stock/Shareholder; Self; Eli Lilly and Company. Funding Eli Lilly and Company

The forthcoming mobile cellular networks (5G) and beyond need to address the challenges of performance requirements in diverse technical scenarios, such as seamless wide-area coverage, hot-spot high-capacity, and low power massive connections, etc. It is widely recognized that Heterogeneous Network (HetNet) is an effective network architecture for addressing these challenges, especially dealing with the access control issue of massive machine-type communication (mMTC). In this paper, we propose a device-to-device (D2D) communication-assisted MTC access mechanism with aim of maximizing the number of admitted MTCDs (MTCDs) in the system while ensuring the quality of service (QoS) requirements of MTCDs. In our proposed scheme, we first determine the access mode for the devices that fail in directly accessing the base station. Specifically, another MTCD which has been successfully connected to the base station (BS) is selected as the relay for a specific MTCD via device-to-device (D2D) communications. We first formulate the access control problem of maximizing the number of admitted MTCDs and prove its NP-hardness. We then resort to genetic algorithm (GA) to solve the optimal relay MTCD selection problem. We conduct simulation experiments to evaluate the performance of our proposed algorithm and numerical results show that the proposed algorithm outperforms the traditional algorithms in terms of the number of successful access MTCDs, transmission rate, and system throughput.

Acquiring large scale annotations is challenging in medical image analysis because of the limited number of qualified annotators. Thus, it is essential to achieve high performance using a small number of labeled data, where the key lies in mining the most informative samples to annotate. In this paper, we propose two effective metrics which leverage anatomical priors to evaluate multi-class segmentation methods without ground truth (GT). Together with our smooth margin loss, these metrics can help to mine the most informative samples for training. In experiments, first we demonstrate the proposed metrics can clearly distinguish samples with different degree of errors in the task of pulmonary lobe segmentation. And then we show that our metrics synergized with the proposed loss function can reach the Pearson Correlation Coefficient (PCC) of 0.7447 with mean surface distance (MSD) and −0.5976 with Dice score, which implies the proposed metrics can be used to evaluate segmentation methods. Finally, we utilize our metrics as sample selection criteria in an active learning setting, which shows that the model trained with our anatomy based query achieves comparable performance with the one trained with random query and uncertainty based query using more annotated training data.

OBJECTIVE To evaluate the efficacy and safety of ultra rapid lispro (URLi) versus lispro in patients with type 2 diabetes on a basal-bolus insulin regimen. RESEARCH DESIGN AND METHODS This was a phase 3, treat-to-target, double-blind 26-week study. After an 8-week lead-in to optimize basal insulin glargine or degludec in combination with prandial lispro treatment, patients were randomized to blinded URLi (n = 336) or lispro (n = 337) injected 0–2 min prior to meals. Patients could continue metformin and/or a sodium–glucose cotransporter 2 inhibitor. The primary end point was change in HbA1c from baseline to 26 weeks (noninferiority margin 0.4%), with multiplicity-adjusted objectives for postprandial glucose (PPG) excursions during a standardized meal test. RESULTS HbA1c improved for both URLi and lispro, and noninferiority was confirmed: estimated treatment difference (ETD) 0.06% (95% CI −0.05; 0.16). Mean change in HbA1c was −0.38% for URLi and −0.43% for lispro, with an end-of-treatment HbA1c of 6.92% and 6.86%, respectively. URLi was superior to lispro in controlling 1- and 2-h PPG excursions: 1-h ETD, −0.66 mmol/L (95% CI −1.01, −0.30); 2-h ETD, −0.96 mmol/L (−1.41, −0.52). Significantly lower PPG excursions were evident from 0.5 to 4.0 h postmeal with URLi treatment. There were no significant treatment differences in rates of severe or documented hypoglycemia ( CONCLUSIONS URLi compared with lispro in a basal-bolus regimen was confirmed to be noninferior for HbA1c and superior to lispro for PPG control in patients with type 2 diabetes.

Background: Basal insulin peglispro (BIL) has a peripheral-to-hepatic distribution of action that resembles endogenous insulin and a prolonged duration of action with a flat pharmacokinetic/pharmacodynamic profile at steady state, characteristics that tend to reduce hypoglycemia risk compared to insulin glargine (GL). The primary objective was to demonstrate that clinically significant hypoglycemia (blood glucose ≤54 mg/dL [3.0 mmol/L] or symptoms of severe hypoglycemia) occurred less frequently within 84 h after a double dose (DD) of BIL than a DD of GL. Methods: This was a randomized, double-blind, two-period crossover study in patients with type 2 diabetes (T2D) previously treated with insulin (N = 68). For the first 3 weeks of each of the two crossover periods, patients received an individualized dose of BIL or GL once nightly (stable dose for 2 weeks/period). Then, during a 7-day inpatient stay with frequent blood glucose monitoring and standardized meals, one DD of study insulin was given. Glucose was infused if blood glucose was ≤54 mg/dL (3.0 mmol/L) or for symptoms of severe hypoglycemia. Results: Within 84 h after the DD, a significantly smaller proportion of patients experienced clinically significant hypoglycemia with BIL compared to GL (BIL, 6.6%; GL, 35.5%; odds ratio for BIL/GL 0.13 [95% confidence interval 0.04–0.39]; P

Nuclear medicine imaging has been developed as a powerful diagnostic approach for cancers by detecting gamma rays directly or indirectly from radionuclides to construct images with beneficial characteristics of high sensitivity, infinite penetration depth and quantitative capability. Current nuclear medicine imaging modalities mainly include single-photon emission computed tomography (SPECT) and positron emission tomography (PET) that require administration of radioactive tracers. In recent years, a vast number of radioactive tracers have been designed and constructed to improve nuclear medicine imaging performance toward early and accurate diagnosis of cancers. This review will discuss recent progress of nuclear medicine imaging tracers and associated biomedical imaging applications. Radiolabeling nanomaterials for rational development of tracers will be comprehensively reviewed with highlights on radiolabeling approaches (surface coupling, inner incorporation and interface engineering), providing profound understanding on radiolabeling chemistry and the associated imaging functionalities. The applications of radiolabeled nanomaterials in nuclear medicine imaging-related multimodality imaging will also be summarized with typical paradigms described. Finally, key challenges and new directions for future research will be discussed to guide further advancement and practical use of radiolabeled nanomaterials for imaging of cancers.

URLi is a novel prandial insulin lispro formulation developed to more closely match physiological insulin secretion. This phase 3, multicenter, treat-to-target, 26-week trial evaluated the efficacy and safety of URLi vs. lispro in patients with T2D. Primary endpoint was A1C change from baseline to week 26, with multiplicity adjusted objectives for PPG excursions (PPGE) with meal test. After 8-week lead-in to optimize basal insulin glargine or degludec and transfer to prandial lispro treatment, patients were randomized to double-blind URLi (n=336) or lispro (n=337) injected 0 to 2 minutes prior to meals. Patients could continue metformin and/or SGLT2 inhibitor. Non-inferiority for primary endpoint was confirmed for URLi: estimated treatment difference (ETD) 0.06 [95% CI -0.05; 0.16], with mean change in A1C of -0.38% URLi and -0.43% lispro (Figure). Mean A1C at week 26 was 6.92% URLi and 6.86% lispro. URLi was superior to lispro in controlling 1- and 2-h PPGE with standardized meal test at week 26, with lower PPGE from 0.5-4 h with URLi (Figure). There were no significant treatment differences in rates of severe or overall hypoglycemia. Incidence of overall treatment-emergent adverse events was similar between treatments. URLi in a basal-bolus regimen led to superior PPG control vs. lispro in patients with T2D. Disclosure T. Blevins: Research Support; Self; Amgen Inc., Gan and Lee Pharma, Lilly Diabetes, Mylan, Novo Nordisk Inc. Speaker's Bureau; Self; Abbott, AstraZeneca, Boehringer Ingelheim Pharmaceuticals, Inc., Janssen Pharmaceuticals, Inc., Lilly Diabetes, Sanofi US. Other Relationship; Self; Medtronic MiniMed, Inc. Q. Zhang: Employee; Self; Eli Lilly and Company. Stock/Shareholder; Self; Eli Lilly and Company. J.P. Frias: Advisory Panel; Self; Becton, Dickinson and Company, Eli Lilly and Company, Gilead Sciences, Inc., Sanofi. Consultant; Self; Echosens, Genentech, Inc., Johnson & Johnson Diabetes Institute, Novo Nordisk Inc., Zafgen, Inc. Research Support; Self; AbbVie Inc., Akcea Therapeutics, Allergan, Amgen Inc., AstraZeneca, Bayer US, Boehringer Ingelheim Pharmaceuticals, Inc., Bristol-Myers Squibb Company, Cirius Therapeutics, Elcelyx Therapeutics, Inc., Eli Lilly and Company, Enanta Pharmaceuticals, Inc., GENFIT, Intarcia Therapeutics, Inc., Intercept Pharmaceuticals, Inc., Ionis Pharmaceuticals, Inc., Janssen Pharmaceuticals, Inc., Merck & Co., Inc., NGM Biopharmaceuticals, Novartis Pharmaceuticals Corporation, Novo Nordisk A/S, Oramed Pharmaceuticals, Pfizer Inc., Sanofi, TaiwanJ Pharmaceuticals Co., Ltd., Theracos, Inc. Speaker's Bureau; Self; Merck & Co., Inc., Sanofi. H. Jinnouchi: None. A.M. Chang: Employee; Self; Eli Lilly and Company. Funding Eli Lilly and Company

URLi is a novel prandial insulin lispro formulation developed to more closely match physiological insulin secretion. This randomized, double-blind, 2 period, crossover, 10 h euglycemic clamp study assessed PK and glucodynamics after one 15U SC dose of URLi and lispro in 41 patients with T1D (mean ± SD age, 32.0 ± 6.8 y; duration of T1D, 17.7 ± 9.3 y; HbA1c, 7.2 ± 0.6%; BMI, 24.8 ± 2.4 kg/m2). After dosing URLi, onset of insulin lispro appearance in the serum was faster (1.3 vs. 6.8 min; p Onset of insulin action was also faster (20 vs. 31 min; p=0.0007) and early action was increased 2.8 fold (p Total insulin lispro exposure and glucose infused during the clamp did not differ and tolerability was similar between URLi vs. lispro. In patients with T1D, URLi accelerated insulin absorption with a reduced late exposure and overall shorter duration vs. lispro, resulting in faster onset of insulin action and reduced late insulin action. Disclosure H. Linnebjerg: Employee; Self; Eli Lilly and Company. Stock/Shareholder; Self; Eli Lilly and Company. Q. Zhang: Employee; Self; Eli Lilly and Company. Stock/Shareholder; Self; Eli Lilly and Company. E.S. LaBell: Employee; Self; Eli Lilly and Company. Stock/Shareholder; Self; Eli Lilly and Company, Johnson & Johnson, Novartis AG. S. Reddy: Employee; Self; Eli Lilly and Company. Stock/Shareholder; Self; Eli Lilly and Company. D.E. Coutant: Employee; Self; Eli Lilly and Company. Employee; Spouse/Partner; Eli Lilly and Company. U. Hoevelmann: None. L. Plum-Moerschel: None. T. Herbrand: None. J. Leohr: Employee; Spouse/Partner; Accenture. Employee; Self; Eli Lilly and Company. Stock/Shareholder; Spouse/Partner; Accenture. Stock/Shareholder; Self; Eli Lilly and Company. Funding Eli Lilly and Company

The comparison of bilateral mammogram images is important for finding masses especially in dense breasts. However, most existing mammogram mass detection algorithms only considered unilateral image. In this paper, we propose a deep model called contrasted bilateral network (CBN) to take bilateral information into consideration. In CBN, Mask R-CNN is used as a basic framework, upon which two major modules are developed to exploit the bilateral information: distortion insensitive comparison module and logic guided bilateral module. The former one is designed to be robust to nonrigid distortion of bilateral registration, while the latter one integrates the bilateral domain knowledge of radiologist. Experimental results on DDSM dataset demonstrate that the proposed algorithm achieves the state-of-the-art performance.

For the problems of privacy data protection caused by massive data sharing in the construction of power Internet of things, a data sensitivity measurement and classification model based on information entropy and BP neural network is proposed. Firstly, a recognition matching algorithm is proposed to identify the sensitive level of attributes in the dataset, and the information entropy is used to determine the weight of attributes sensitivity level, so as to calculate the sensitivity measurement value of records in the dataset; finally, the BP neural network is used to output the data classification results. The experimental results show that the model can achieve accurate measurement and classification of data, with low incorrect judgment rate and error rate.

Aims To compare 24-hour fixed-time basal insulin peglispro (BIL) dosing with 8- to 40-hour variable-time BIL dosing for glycaemic control and safety in patients with type 1 diabetes. Primary outcome was non-inferiority of BIL variable-time dosing compared with fixed-time dosing for glycated haemoglobin (HbA1c) change after 12-week treatment (margin = 0.4%). Materials and methods This Phase 3, open-label, randomized, cross-over study (N = 212) was conducted at 20 centres in the United States. During the 12-week lead-in phase, patients received BIL daily at fixed-times. Two 12-week randomized cross-over treatment phases followed, where patients received BIL dosed at either fixed- or variable-times. During the 4-week safety follow-up, patients received conventional insulins. Results During the lead-in period, least-squares mean HbA1c decreased from 7.5% to 6.8%. For BIL, variable-time dosing was non-inferior to fixed-time dosing for HbA1c change [least-squares mean difference = 0.06%, 95% confidence interval (−0.01, 0.13)]. In both regimens, HbA1c increased slightly during the cross-over periods, but remained significantly below baseline. Variable- and fixed-time dosing regimens had similar rates of total hypoglycaemia (10.4 ± 0.62 and 10.5 ± 0.67 events/patient/30 days, P = .947) and nocturnal hypoglycaemia (1.3 ± 0.11 and 1.5 ± 0.13 events/patient/30days, P = .060). Comparable proportions of patients achieved HbA1c < 7.0% with variable- [91 (54.5%)] and fixed-time dosing [101 (60.5%)]. Conclusions Treatment with BIL allows patients to use flexible dosing intervals from 8 to 40 hours. Glycaemic efficacy (HbA1c), glycaemic variability and hypoglycaemia are similar to fixed-time dosing, suggesting that BIL could potentially provide flexibility in dosing for patients who miss their daily basal insulin.

Decreasing risk of cardiovascular (CV) disease remains a challenge to survival in type 2 diabetes.The objective was to assess the association between demographic, glycemic, and other clinical factors and CV risk in the Hyperglycemia and Its Effect After Acute Myocardial Infarction on Cardiovascular Outcomes in Patients With Type 2 Diabetes Mellitus trial. DESIGN, SETTINGS, PARTICIPANTS, AND INTERVENTION: We used discrete-time survival tree analysis to examine data collected for up to 4.6 years in 1115 patients with type 2 diabetes mellitus experiencing acute myocardial infarction (MI) less than or equal to 18 days before enrollment.The primary objective was to identify demographic, glycemic, and CV risk factors best separating survival curves over time for a composite end point: CV death, nonfatal MI, nonfatal stroke, hospitalization for acute coronary syndromes, or coronary revascularization planned after randomization.Average change across visits in mean 2-hour blood glucose level after meals was associated with the greatest difference in event-free survival probability for the primary end point: mean time to 75% event-free survival for an average change across visits less than or equal to -0.14 mmol/L, 73.48 weeks; for visits with average change more -0.14 mmol/L, 29.10 weeks. An average change across visits in the hemoglobin A1c level less than or equal to -0.92% (-10.06 mmol/mol) and the absence of a history of stroke or acute MI increased CV event-free survival time further. Fasting blood glucose and randomized insulin treatment strategy were weak predicting factors of event-free survival.Postprandial glycemia should be considered a potential target in trials to reduce CV morbidity and mortality in type 2 diabetes mellitus.

Objective: In oncology research, mortality, as a variable, and overall survival (OS), as an efficacy endpoint, are critical. As real-world data gains use in supporting regulatory decisions, assessing the validity of mortality data is important. Mortality data derived solely from electronic health records (EHRs) has known gaps. We developed a mortality variable based on multiple sources of death data and benchmarked it against the National Death Index (NDI) data as a gold standard within 4 cancer types (Curtis et al, 2018). We refreshed and expanded validity assessment to 18 cancer types. Methods: Patients diagnosed with at least 1 of 18 cancer types between 1/1/2011 and 12/31/2017 were included from the Flatiron Health EHR-derived de-identified US database. To develop a composite mortality variable, we amalgamated multiple sources of real-world death data, linking patient-level structured (EHR, commercial, Social Security Death Index) and unstructured data curated via technology-enabled abstraction. We calculated validity metrics (sensitivity, specificity, ±15-day accuracy) by benchmarking against the NDI data in each cancer type, as a composite variable, as well as by individual and combinations of death data sources. OS was estimated using the Kaplan-Meier method. Results: There were 160,436 patients included in the study cohort. In the 18 cancer types, sensitivity ranged from 83.9 - 91.5% (17 out of 18 had sensitivity ≥85%), specificity ranged from 93.5 - 99.7%, and ±15-day accuracy ranged from 95.6 - 97.6%, compared to the NDI (Table). Median OS estimates for the composite mortality variable when compared to that from the NDI ranged from 3 - 13% greater across all cancer types. Cancer Type Note: some patients could have >1 cancer typeNSensitivity (%)Specificity (%)+/- 15-day Accuracy (%)Solid TumorBreast Cancer (Early)166983.9 (77.4, 90.3)99.7 (99.5, 100)96.3 (92.7, 99.9)Breast Cancer (Metastatic)1647388.3 (87.6, 88.9)97.7 (97.4, 98.0)96.9 (96.6, 97.3)Colorectal Cancer (Metastatic)1723286.4 (85.7, 87.1)97.1 (96.8, 97.5)96.3 (95.9, 96.8)Gastro-Esophageal Cancer (Advanced)716988.5 (87.6, 89.4)94.3 (93.3, 95.3)97.1 (96.6, 97.6)Head and Neck Cancer (Advanced)527189.3 (88.2, 90.3)95.9 (95.0, 96.8)96.9 (96.3, 97.5)Hepatocellular Carcinoma278485.0 (83.3, 86.7)95.7 (94.5, 96.9)95.6 (94.5, 96.6)Malignant Pleural Mesothelioma170091.0 (89.4, 92.6)95.6 (93.7, 97.4)97.6 (96.7, 98.5)Melanoma (Advanced)703189.5 (88.4, 90.6)98.7 (98.4, 99.1)97.4 (96.8, 98)Non-small Cell Lung Cancer (Advanced)4507090.4 (90.1, 90.7)95.1 (94.8, 95.5)97.2 (97.0, 97.4)Ovarian Cancer496486.8 (85.2, 88.3)98.7 (98.3, 99.1)96.8 (95.9, 97.6)Pancreatic Cancer (Metastatic)545890.2 (89.3, 91.1)93.5 (92.2, 94.8)97.2 (96.7, 97.7)Prostate Cancer (Metastatic)849589.0 (88.0, 90)98.7 (98.4, 99.0)97.1 (96.5, 97.7)Renal Cell Carcinoma (Metastatic)577088.5 (87.4, 89.6)97.7 (97.1, 98.3)97.2 (96.6, 97.8)Small Cell Lung Cancer472490.4 (89.4, 91.4)95.8 (94.7, 96.8)97.5 (97.0, 98.1)Urothelial Cancer (Advanced)629390.2 (89.3, 91.1)96.6 (95.8, 97.4)97.5 (97.0, 98.0)Liquid TumorChronic Lymphocytic Leukemia903591.5 (90.4, 92.7)99.3 (99.1, 99.5)96.8 (96.1, 97.6)Diffuse Large B-cell Lymphoma434489.1 (87.4, 90.8)99 (98.7, 99.4)96.1 (95.0, 97.3)Multiple Myeloma780388.5 (87.3, 89.7)99.0 (98.8, 99.3)97.0 (96.3, 97.7) Conclusions: We observed high sensitivity, specificity, and date accuracy of our composite mortality variable across all cancer types. Future research will investigate the validity of our mortality variable stratified by demographic and clinical characteristics to understand performance in subpopulations. Citation Format: Qianyi Zhang, Anala Gossai, Shirley Monroe, Nathan C. Nussbaum, Christina M. Parrinello. Validation analysis of a composite real-world mortality endpoint for US cancer patients [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 5772.

Data on cytokeratin-18 (K-18) and enhanced liver fibrosis (ELF) score in insulin-treated diabetes patients with non-alcoholic fatty liver disease (NAFLD) are limited. This study analyzed phase III data comparing basal insulin peglispro (BIL) and insulin glargine in type 1 (T1D), and type 2 diabetes (T2D) (insulin-naïve and insulin-treated). Alanine aminotransferase (ALT), K-18, ELF scores and liver fat content (LFC), measured by MRI, were obtained longitudinally. Baseline K-18 (U/L) was higher in T2D (range: 207-247) than T1D (range: 148-183), correlated with ALT in all populations (r (range) 0.264-0.637, p

The microalgal biotech sector began in the 1940s, where the first modern attempts to grow microalgae focused on finding alternative sources of chemicals for use in munition manufacturing and fuels during World War II, by examining the production of lipids by various microalgae (Harder and Witsch 1942; Burlew 1953). In the postwar years, microalgae were seen as one solution to a shortfall in food production before the advent of the “Green Revolution.” Later, during the oil crisis of the 1970s, when the price of petroleum was high, microalgae were revisited for their potential use in biofuels based on their ability to accumulate oil, which is usually in the form of triacylglycerols (Borowitzka 2013a). Microalgae have high theoretical productivities, 92partly due to their unicellular nature, which leads to an efficient resource allocation into storage products as opposed to structure, their rapid growth rate, and their carbon-concentrating mechanisms, which increase the efficiency of CO2 utilization (Giordano et al. 2005).

Background: We set out to study the relationship between different measures of glycemic variability and the rate of hypoglycemia in patients with type 2 diabetes. Subjects and Methods: Data were pooled from three 24-week insulin trials including patients on twice-daily (BID) insulin lispro mix 75/25 (75% insulin lispro protamine suspension, 25% insulin lispro) (n = 805), daily (QD) insulin glargine (n = 1,019), insulin lispro protamine suspension (n = 353) (QD or BID), and insulin detemir (n = 166) (QD or BID), all with continuation of prestudy oral antihyperglycemic medications. Glycemic variability measures were derived from seven-point self-monitored blood glucose profiles. Results: At baseline, mean (+/- SD) age was 56.9 +/- 9.7 years, duration of type 2 diabetes was 9.5 +/- 6.1 years, hemoglobin A1c (HbA1c) was 8.9 +/- 1.1%, and 51.9% were male. Intra-day glucose coefficient of variation (CV), fasting blood glucose, intra-day minimum glucose and inter-day glucose CV at 24 weeks, and intra-day glucose CV at baseline were significantly correlated with the rate of hypoglycemia events between Weeks 12 to 24 (P

Optical fibers and waveguides in general effectively control and modulate light propagation, and these tools have been extensively used in communication, lighting and sensing. Recently, they have received increasing attention in biomedical applications. By delivering light into deep tissue via these devices, novel applications including biological sensing, stimulation and therapy can be realized. Therefore, implantable fibers and waveguides in biocompatible formats with versatile functionalities are highly desirable. In this review, we provide an overview of recent progress in the exploration of advanced optical fibers and waveguides for biomedical applications. Specifically, we highlight novel materials design and fabrication strategies to form implantable fibers and waveguides. Furthermore, their applications in various biomedical fields such as light therapy, optogenetics, fluorescence sensing and imaging are discussed. We believe that these newly developed fiber and waveguide based devices play a crucial role in advanced optical biointerfaces.

Background Osteosarcoma is the most frequent type of malignant bone tumor in children and adolescents and is associated with a high propensity for lung metastasis. Recent experiments have indicated that PLA2G16 contributes to osteosarcoma progression and metastasis in both mouse and human osteosarcoma cell lines. The aim of this study was to compare the expression of PLA2G16 in non-metastatic and metastatic osteosarcomas to determine whether PLA2G16 expression can serve as a biomarker of osteosarcoma prognosis and metastasis. Methods Quantitative real-time PCR was used to examine PLA2G16 mRNA in primary osteosarcoma patients (18 patients without metastases and 17 patients with metastases), and immunohistochemistry (IHC) staining of PLA2G16 was performed on tissue microarrays from 119 osteosarcoma patients. Tumor metastatic behavior and survival of the patients were followed up for a minimum of 36 months and a maximum of 171 months. The prognostic value of PLA2G16 expression was evaluated by the Kaplan–Meier method and a log-rank test. Multivariate Cox regression analysis was used to identify significant independent prognostic factors. Results Osteosarcoma patients with metastasis showed a higher expression of PLA2G16 at both the mRNA and protein levels (both at P values< 0.05) than did patients without metastasis. Osteosarcoma patients with positive IHC staining of PLA2G16 expression at primary sites had shorter overall survival and metastasis-free survival (both at P values

This paper proposes a novel super-resolution framework to reconstruct high-resolution fundus images from multiple low-resolution video frames in retinal fundus imaging. Natural eye movements during an examination are used as a cue for super-resolution in a robust maximum a-posteriori scheme. In order to compensate heterogeneous illumination on the fundus, we integrate retrospective illumination correction for photometric registration to the underlying imaging model. Our method utilizes quality self-assessment to provide objective quality scores for reconstructed images as well as to select regularization parameters automatically. In our evaluation on real data acquired from six human subjects with a low-cost video camera, the proposed method achieved considerable enhancements of low-resolution frames and improved noise and sharpness characteristics by 74%. In terms of image analysis, we demonstrate the importance of our method for the improvement of automatic blood vessel segmentation as an example application, where the sensitivity was increased by 13% using super-resolution reconstruction.

We examined whether identification of patients with placebo-remitter characteristics and placebo-nonremitter characteristics enhances the ability to identify drug-placebo treatment differences and, perhaps, differences between agents in major depressive disorder (MDD). We hypothesized: 1) that drug-placebo differences in remission rates would be greater for both duloxetine and selective serotonin reuptake inhibitors (SSRIs) among placebo nonremitters than placebo remitters and: 2) that the difference between active treatments would also be greater in placebo nonremitters than placebo remitters.Data were obtained from seven randomized, parallel, double-blind MDD studies which compared the effects of duloxetine (N = 795), placebo (N = 510), and SSRIs (N = 551). The 17-item Hamilton depression rating scale (HAMD) was used to assess depression severity. The classification of participants as having placebo-remitter or placebo-nonremitter characteristics was based on age, duration of current MDD episode, HAMD anxiety score, and HAMD core score. Odds ratios (ORs) for remission comparing active treatment and placebo were obtained from logistic regression models. Tests of homogeneity were used for between-group comparisons.For placebo nonremitters, both duloxetine (OR = 3.52 [95% CI: 2.21, 5.62; P 0.0001]) and SSRIs (OR = 2.38 [95% CI: 1.45, 3.89; P = 0.0006]) showed significantly higher remission rates compared to placebo. Drug-placebo differences in remission were significantly greater for placebo nonremitters than for placebo remitters for both duloxetine (P = 0.02) and SSRIs (P = 0.049). For placebo remitters, remission with duloxetine (OR = 1.83 [95% CI: 1.35, 2.47; P = 0.0001]), but not the SSRIs selected for this study (OR = 1.31 [95% CI: 0.93, 1.84; P = 0.12]), was significantly greater than placebo. Contrary to expectation, the advantage of duloxetine over the SSRIs was not greater in placebo nonremitters. However, the fact that our analysis required patient-level data limited the number of agents studied, compromising generalization.Our study suggests that drug-placebo differences in remission rates will be greater in subjects with placebo-nonremitter than with placebo-remitter characteristics.

To identify predictors of remission with placebo treatment in double-blind randomized controlled trials (RCTs) in major depressive disorder (MDD).A total of 1017 placebo-treated patients with baseline Hamilton Depression rating scale (HAMD) total ≥15 from eight duloxetine RCTs were included. Remission was defined as endpoint (7-8 weeks) HAMD total ≤7. Data were randomly split into training data (N = 813, 80%) for model selection and test data (N = 204, 20%) for validation. Logistic regression and classification and regression tree (CART) methods were used to identify predictors of remission. Predictive accuracy of models was assessed by Receiver Operator Characteristic (ROC) curves.Baseline predictors for remission with placebo consistently identified with the logistic regression and CART analysis were less severe depressive symptoms (based on HAMD core symptoms), younger age, less anxiety (based on HAMD anxiety/somatization), and shorter current MDD episode duration. Associated cut-off values from the CART method characterized patient groups according to their remission likelihood. However, the predictive accuracy was modest for both methods with areas under the ROC curve of 0.6-0.65 based on test data.The derived models, although of limited value for predicting remission in individual patients, may be useful for adjusting for placebo effects in clinical trials.

Ensemble methods have become popular for QSAR modeling, but most studies have assumed balanced data, consisting of approximately equal numbers of active and inactive compounds. Cheminformatics data are often far from being balanced. We extend the application of ensemble methods to include cases of imbalance of class membership and to more adequately assess model output. Based on the extension, we propose an ensemble method called MBEnsemble that automatically determines the appropriate tuning parameters to provide reliable predictions and maximize the F-measure. Results from multiple data sets demonstrate that the proposed ensemble technique works well on imbalanced data.