Significance of Radiomics in Predicting Local Control for Patients with Malignant Liver Tumors Treated Using Stereotactic Body Radiotherapy

PURPOSE/OBJECTIVE(S)

We routinely deliver Stereotactic Body Radiotherapy (SBRT) in malignant liver tumors using planning computed tomography (CT) and Magnetic Resonance Images (MRI) to aid target definition. In this study, we extracted radiomic features from the MR images to predict local control (LC) post-SBRT.

MATERIALS/METHODS

We retrospectively analyzed patients with either hepatocellular cancers (HCCs) or liver metastases (Mets) treated with SBRT between Aug 2014 and Aug 2020. All patients had CT simulation followed by 1.5 Tesla planning MRI in treatment position. Contrast enhanced T1 VIBE and T2 Haste MR sequences were registered with planning CT for target definition. Radiomic features were extracted from Gross Tumor Volumes (GTV) masked out of 60 seconds post contrast T1 VIBE MR images using the Radiomics calculator tool RaCaT. The output included 480 (408 textural, 50 intensity and 22 morphological) features for each target. Principal Component Analysis of the outputs obtained from all the targets yielded 20 radiomic feature clusters after computational prioritization. These clusters were correlated to LC outcomes at various time points post-SBRT. LC was defined as non-progressive disease. Accuracy of predictions was measured by area under (AUC) receiver operating characteristic curve. Cox regression analysis was done to find univariate and multivariate clinical [HCCs vs. Mets, single vs. multiple lesions, previous local therapy (yes vs. no), GTV volume (≤40 vs. >40 cc)], radiomic and dosimetric predictors (continuous) of LC.

RESULTS

In total, 97 patients received SBRT to 122 lesions. The median dose prescribed was 45 Gy (range, 30-50 Gy). Median age was 69 years (interquartile range, IQR 61-73 yrs.). 59 patients had HCCs and 38 had Mets. 24 lesions had prior ablative therapy. 75 patients had one target, and 22 had multiple targets. Median GTV was 43.5 cc (IQR 23.4-78.6 cc). Median follow up was 16.6 months (IQR 9.7-27.2 mths). Median LC was 13.6 months (IQR 8.0-23.5 mths). On univariate analysis, histology (HCCs vs. Mets; Hazard ratio (HR) 2.9, 95% CI 1.4-6.4; p < 0.006), radiomic clusters (p < 0.006) and the max., mean, and min. doses to GTV and Planning Target Volumes correlated with improved LC (all p-values < 0.05). On multivariate analysis, histology (HCCs vs. Mets; HR 4.4, 95% CI 1.6-12.3; p = 0.004), radiomic clusters (p = 0.034) and prescription dose (p = 0.048) were significant covariates. Specifically, the 20 radiomic clusters were predictive of LC, and the accuracy of predictions showed promise with AUC values of 0.74, 0.80, and 0.81 at 12, 24, and 36 months post-SBRT, respectively. AUC values for LC in HCCs vs. Mets at 12, 24, and 36 months were 0.83, 0.77, and 0.70, and 0.66, 0.77, and 0.88, respectively.

CONCLUSION

MR-based radiomics predict LC post-SBRT in patients with malignant liver tumors. Further research focused on independent validation of the model is required to explore its clinical use.

Large-scale protein-protein post-translational modification extraction with distant supervision and confidence calibrated BioBERT

MOTIVATION

Protein-protein interactions (PPIs) are critical to normal cellular function and are related to many disease pathways. A range of protein functions are mediated and regulated by protein interactions through post-translational modifications (PTM). However, only 4% of PPIs are annotated with PTMs in biological knowledge databases such as IntAct, mainly performed through manual curation, which is neither time- nor cost-effective. Here we aim to facilitate annotation by extracting PPIs along with their pairwise PTM from the literature by using distantly supervised training data using deep learning to aid human curation.

METHOD

We use the IntAct PPI database to create a distant supervised dataset annotated with interacting protein pairs, their corresponding PTM type, and associated abstracts from the PubMed database. We train an ensemble of BioBERT models-dubbed PPI-BioBERT-x10-to improve confidence calibration. We extend the use of ensemble average confidence approach with confidence variation to counteract the effects of class imbalance to extract high confidence predictions.

RESULTS AND CONCLUSION

The PPI-BioBERT-x10 model evaluated on the test set resulted in a modest F1-micro 41.3 (P =5 8.1, R = 32.1). However, by combining high confidence and low variation to identify high quality predictions, tuning the predictions for precision, we retained 19% of the test predictions with 100% precision. We evaluated PPI-BioBERT-x10 on 18 million PubMed abstracts and extracted 1.6 million (546507 unique PTM-PPI triplets) PTM-PPI predictions, and filter [Formula: see text] (4584 unique) high confidence predictions. Of the 5700, human evaluation on a small randomly sampled subset shows that the precision drops to 33.7% despite confidence calibration and highlights the challenges of generalisability beyond the test set even with confidence calibration. We circumvent the problem by only including predictions associated with multiple papers, improving the precision to 58.8%. In this work, we highlight the benefits and challenges of deep learning-based text mining in practice, and the need for increased emphasis on confidence calibration to facilitate human curation efforts.

Healthcare Utilization Survey in the Hybrid Model of the Surveillance for Enteric Fever in India (SEFI) Study: Processes, Monitoring, Results, and Challenges

Background

Lack of reliable data in India drove the “Surveillance of Enteric Fever in India” (SEFI) concept. Hybrid surveillance, combining facility-based surveillance for the crude incidence, and a community-based healthcare utilization survey (HCUS) to calculate the factor needed to arrive at the adjusted incidence, was used in 6 sites. The HCUS aimed to determine the percentage of utilization of study facilities by the catchment population for hospitalizations due to febrile illness.

Methods

Population proportional to size sampling and systematic random sampling, in 2 stages, were used to survey 5000 households per site. Healthcare utilization was assessed.

Results

Febrile illness accounted for 20% of admissions among 137 990 individuals from 30 308 households. Only 9.6%–38.3% of those admitted with febrile illness sought care in the study hospitals. The rate of rural utilization of the private sector for hospitalization was 67.6%. The rate of hospitalization for febrile illness, per 1000 population, ranged from 2.6 in Manali to 9.6 in Anantapur; for 25.8% of the deaths associated with febrile illness, no facility was used before death.

Conclusions

One in 5 hospitalizations were associated with fever. Rural utilization of the private sector for hospitalization due to febrile illness was more than that of the public sector. Healthcare utilization patterns for hospital admissions due to febrile illness varied across sites. A meticulously performed HCUS is pivotal for accurate incidence estimation in a hybrid surveillance.

Clinical Trials Registration

ISRCTN72938224.

Findings from the 2017 HIV estimation round & trend analysis of key indicators 2010-2017: Evidence for prioritising HIV/AIDS programme in India

Background & objectives:

The National AIDS Control Organisation (NACO) and the ICMR-National Institute of Medical Statistics, the nodal agency for conducting HIV estimations in India, have been generating HIV estimates regularly since 2003. The objective of this study was to describe India's biennial HIV estimation 2017 process, data inputs, tool, methodology and epidemiological assumptions used to generate the HIV estimates and trends of key indicators for 2010-2017 at national and State/Union Territory levels.

Methods:

Demographic Projection (DemProj) and AIDS Impact Modules (AIM) of Spectrum 5.63 software recommended by the United Nations Programme on HIV and AIDS Global Reference Group on HIV Estimates, Modelling and Projections, were used for generating HIV estimations on key indicators. HIV sentinel surveillance, epidemiological and programme data were entered into Estimation Projection Package (EPP), and curve fitting was done using EPP classic model. Finally, calibration was done using the State HIV prevalence of two rounds of National Family Health Survey (NFHS) -3 and -4 and Integrated Biological and Behavioural Surveillance (IBBS), 2014-2015.

Results:

The national adult prevalence of HIV was estimated to be 0.22 per cent in 2017. Mizoram, Manipur and Nagaland had the highest prevalence over one per cent. An estimated 2.1 million people were living with HIV in 2017, with Maharashtra estimated to have the highest number. Of the 88 thousand annual new HIV infections estimated nationally in 2017, Telangana accounted for the largest share. HIV incidence was found to be higher among key population groups, especially people who inject drugs. The annual AIDS-related deaths were estimated to be 69 thousand nationally. For all indicators, geographic variation in levels and trends between States existed.

Interpretation & conclusions:

With a slow decline in annual new HIV infections by only 27 per cent from 2010 to 2017 against the national target of 75 per cent by 2020, the national target to end AIDS by 2030 may be missed; although at the sub-national level some States have made better progress to reduce new HIV infection. It calls for reinforcement of HIV prevention, diagnosis and treatment efforts by geographical regions and population groups.

Overview of the BioCreative VI Precision Medicine Track: mining protein interactions and mutations for precision medicine

The Precision Medicine Initiative is a multicenter effort aiming at formulating personalized treatments leveraging on individual patient data (clinical, genome sequence and functional genomic data) together with the information in large knowledge bases (KBs) that integrate genome annotation, disease association studies, electronic health records and other data types. The biomedical literature provides a rich foundation for populating these KBs, reporting genetic and molecular interactions that provide the scaffold for the cellular regulatory systems and detailing the influence of genetic variants in these interactions. The goal of BioCreative VI Precision Medicine Track was to extract this particular type of information and was organized in two tasks: (i) document triage task, focused on identifying scientific literature containing experimentally verified protein-protein interactions (PPIs) affected by genetic mutations and (ii) relation extraction task, focused on extracting the affected interactions (protein pairs). To assist system developers and task participants, a large-scale corpus of PubMed documents was manually annotated for this task. Ten teams worldwide contributed 22 distinct text-mining models for the document triage task, and six teams worldwide contributed 14 different text-mining systems for the relation extraction task. When comparing the text-mining system predictions with human annotations, for the triage task, the best F-score was 69.06%, the best precision was 62.89%, the best recall was 98.0% and the best average precision was 72.5%. For the relation extraction task, when taking homologous genes into account, the best F-score was 37.73%, the best precision was 46.5% and the best recall was 54.1%. Submitted systems explored a wide range of methods, from traditional rule-based, statistical and machine learning systems to state-of-the-art deep learning methods. Given the level of participation and the individual team results we find the precision medicine track to be successful in engaging the text-mining research community. In the meantime, the track produced a manually annotated corpus of 5509 PubMed documents developed by BioGRID curators and relevant for precision medicine. The data set is freely available to the community, and the specific interactions have been integrated into the BioGRID data set. In addition, this challenge provided the first results of automatically identifying PubMed articles that describe PPI affected by mutations, as well as extracting the affected relations from those articles. Still, much progress is needed for computer-assisted precision medicine text mining to become mainstream. Future work should focus on addressing the remaining technical challenges and incorporating the practical benefits of text-mining tools into real-world precision medicine information-related curation.

BioCreative VI Precision Medicine Track system performance is constrained by entity recognition and variations in corpus characteristics

Precision medicine aims to provide personalized treatments based on individual patient profiles. One critical step towards precision medicine is leveraging knowledge derived from biomedical publications-a tremendous literature resource presenting the latest scientific discoveries on genes, mutations and diseases. Biomedical natural language processing (BioNLP) plays a vital role in supporting automation of this process. BioCreative VI Track 4 brings community effort to the task of automatically identifying and extracting protein-protein interactions (PPi) affected by mutations (PPIm), important in the precision medicine context for capturing individual genotype variation related to disease.We present the READ-BioMed team's approach to identifying PPIm-related publications and to extracting specific PPIm information from those publications in the context of the BioCreative VI PPIm track. We observe that current BioNLP tools are insufficient to recognise entities for these two tasks; the best existing mutation recognition tool achieves only 55% recall in the document triage training set, while relation extraction performance is limited by the low recall performance of gene entity recognition. We develop the models accordingly: for document triage, we develop term lists capturing interactions and mutations to complement BioNLP tools, and select effective features via a feature contribution study, whereas an ensemble of BioNLP tools is employed for relation extraction.Our best document triage model achieves an F-score of 66.77% while our best model for relation extraction achieved an F-score of 35.09% over the final (updated post-task) test set. Impacting the document triage task, the characteristics of mutations are statistically different in the training and testing sets. While a vital new direction for biomedical text mining research, this early attempt to tackle the problem of identifying genetic variation of substantial biological significance highlights the importance of representative training data and the cascading impact of tool limitations in a modular system.

Molecular interactions in ethyl acetate-chlorobenzene binary solution: Dielectric, spectroscopic studies and quantum chemical calculations

Dielectric studies using Time Domain Reflectometry method has been carried out on the binary solution of Ethyl acetate (EA) with Chlorobenzene (CBZ) over the entire composition range. Spectroscopic (FTIR and ¹³C NMR) signatures of neat EA, CBZ and their equimolar binary solution have also been recorded. The results of the spectroscopic studies favour the presence of (CBZ) CH⋯OC (EA), (EA) methylene CH⋯π electrons (CBZ) and (EA) methyl CH⋯Cl (CBZ) contacts which have been validated using quantum chemical calculations. Dimerization of CBZ has been identified. Presence of β-clusters has been identified in all the solutions. Although EA and CBZ molecules have nearly equal molar volumes, CBZ molecules experience larger hindrance for the rotation than EA molecules. Very small excess dielectric constant (ε^E) values may be correlated with weak heteromolecular forces and/or closed heteromolecular association.

A facile graphene oxide based sensor for electrochemical detection of prostate anti-cancer (anti-testosterone) drug flutamide in biological samples

An electrochemical sensor based on graphene oxide modified glassy carbon electrode for the determination of anti-cancer drug flutamide.

Abstract B1-40: Biological network analysis using an in vitro model of androgen-resistance in prostate cancer

The development of androgen resistance is a major limitation to androgen deprivation treatment in prostate cancer (PCa). We have developed an in vitro model of androgen-resistance using the androgen sensitive cell line LNCaP to characterize the phenotypic and transcriptomic changes occurring as androgen resistance develops. Our aim is to understand biological network profiles of transcriptomic changes during the transition to androgen-resistance and to validate these changes between our in vitro model and previously published clinical datasets (paired samples before and after androgen-deprivation therapy of patients with advanced PCa) (1). Methods: PCa cells, LNCaP, expressing mutated AR which are androgen-dependent (2,3), are used in the development of an androgen-resistant subline. Subline cells were established by prolonged cultures in media + 10% CS-FBS to mimic the clinical course of PCa. Cell proliferation, cell motility and invasion, morphology, AR expression were examined. RNA-sequencing was performed using the parental LNCaP cells and an androgen-resistant subline (LNCaP-AI) established by chronic exposure to the androgen-deprivation. Reads from cells and clinical samples (1), pre- vs. post-treatment, were processed through the same standard pipeline and quality control. Data outputs were analysed as differential expression [DEG] (EdgeR) (4) and top scoring protein-protein interaction (PPI) networks [PINA2 (5) and BioNet (6)]. Data outputs from the cell line model and clinical samples were compared. Results: LNCaP cells initially showed poor growth after prolonged exposure to androgen-deprived conditions but later adapted and started to grow well. After 24 weeks of androgen-deprivation, LNCaP-AI's growth was no longer responsive to addition of androgen [0.1 - 10 nM]. AR expression was not different in LNCaP and LNCaP-AI (P&gt;0.05). LNCaP-AI cells had increased proliferation and cell invasion compared to LNCaP. We identified key genes that overlap between our cell line and clinical RNAseq (1) datasets and analyzed the overlapping PPI network that showed the same pattern of behavior in both datasets. The network revealed several potential mechanisms and gene interactions that warrant further investigation, including cooperative behaviors of other nuclear receptors, TP63 mediated signalling pathway and Aryl hydrocarbon receptor transcriptional pathway. Conclusion: Cell line model of androgen-resistance will be used for further longitudinal study of the mechanism of castrate resistant prostate cancer (CRPC). Our approach allows for better characterization of biological processes of CRPC. Knowledge of the genetic profiles during transition to androgen resistance will improve our understanding of this common clinical scenario and may lead to biomarker discovery.

References:

1. Rajan P, Sudbery IM, Villasevil ME, Mui E, Fleming J, Davis M, et al. Next-generation sequencing of advanced prostate cancer treated with androgen-deprivation therapy. Eur Urol 2013;66(1):32-9.

2. Marques RB, van Weerden WM, Erkens-Schulze S, de Ridder CM, Bangma CH, Trapman J, et al. The human PC346 xenograft and cell line panel: A model system for prostate cancer progression. Eur Urol 2006;49(2):245-57.

3. Pienta KJ, Bradley D. Mechanisms underlying the development of androgen-independent prostate cancer. Clin Cancer Res 2006;12(6):1665-71.

4. Robinson MD, McCarthy DJ, Smyth GK. edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics 2010;26(1):139-40.

5. Cowley MJ, Pinese M, Kassahn KS, Waddell N, Pearson JV, Grimmond SM, et al. PINA v2.0: mining interactome modules. Nucleic Acids Res 2012;40(Database issue):D862-5.

6. Beisser D, Klau GW, Dandekar T, Müller T, Dittrich M. BioNet: an R-Package for the functional analysis of biological networks. Bioinformatics 2010;26(8):1129-30.

Citation Format: Sujitra Detchokul, Aparna Elangovan, Melissa J. Davis, Geoff Macintyre, Edmund J. Crampin, Albert G. Frauman. Biological network analysis using an in vitro model of androgen-resistance in prostate cancer. [abstract]. In: Proceedings of the AACR Special Conference on Computational and Systems Biology of Cancer; Feb 8-11 2015; San Francisco, CA. Philadelphia (PA): AACR; Cancer Res 2015;75(22 Suppl 2):Abstract nr B1-40.

A probe on the intermolecular forces in diisopropyl ether-n-butyric acid mixture by dielectric, FTIR studies and quantum chemical calculations

The results of FTIR spectral measurement on equimolar diisopropyl ether-butyric acid binary mixture and quantum chemical calculations on the complex molecule have been presented. Dielectric studies have been carried out on the binary mixture over the entire composition range and at four different temperatures 303 K, 308 K, 313 K and 318 K. n-Butyric acid seems to prefer less polar ether to interact with it. It appears that the usual interpretation of variation of static dielectric constant and positive deviation of excess permittivity from ideal mixture behavior needs to be relooked.

Exploring the relation between body mass index, diet, and dental caries among 6-12-year-old children

BACKGROUND AND AIM

Childhood overweight and obesity are becoming a major public health concern all over the world. Change in lifestyles and economic growth have led to sedentary lifestyle and altered dietary patterns. There are conflicting reports in the literature regarding the association between body mass index (BMI) and dental caries from various parts of the world. The aim of the present study was to determine if there is an association between BMI-for-age and dental caries in children and to find out the role of diet with respect to BMI-for-age and dental caries.

MATERIALS AND METHODS

Demographics and anthropometric measurements were obtained for 600 children and BMI-for-age was calculated. Clinical examination for dental caries was carried out following WHO criteria. A diet recording sheet was prepared and children/parents were asked to record the dietary intake for 3 days. Data obtained were statistically analyzed using Chi-square, analysis of variance (ANOVA), and multiple linear regression.

RESULTS

After excluding improperly filled diet recording sheets, 510 children were included in the study. Caries prevalence was more in obese children than in other BMI groups. Caries scores increased as BMI-for-age increased, though this was not statistically significant. Consumption of fatty foods and snacks was more with obese children compared to other groups. A correlation was found between caries and snacks.

CONCLUSION

Dental caries scores showed no relationship between BMI-for-age in children. Both snacks and fatty food items were consumed more by obese children, which seeks attention.

Molecular interaction study of the diisopropyl ether-propionic acid mixture by spectroscopic and dielectric studies

FTIR and 13C NMR spectral studies have been carried out on diisopropyl ether-propionic acid binary mixture to probe the molecular interactions and stoichiometry of complexation. Density functional theory (DFT) calculations of vibrational frequencies of pure acid and ether-acid binary mixtures have also been performed. In addition, Kirkwood-correlation factors, excess permittivity and excess free energy of mixing have been obtained at various concentrations and at four different temperatures from the dielectric measurements. Excess permittivity is found to have positive deviation and excess free energy deviates negatively from ideal behaviour.

Modelling epidemiology of leprosy

A simulation model for leprosy transmission and control has been developed with specific objectives. Several sensitivity experiments have been carried out by altering the various inputs based on empirical data combined with intelligent guessing. The outputs generated through these exercises were on the expected lines. While incremental exercises would improve the model, it can be used even at the existing stage as a tool for programme managers.

Efficacy of a Bacillus sphaericus formulation as influenced by the quality of Culex quinquefasciatus breeding waters

To understand the physico-chemical factors that influence the efficacy of B. sphaericus formulation in the breeding sites of Culex quinquefasciatus, a study was carried out in Mayiladuturai area of Tamil Nadu (India). The factors studied were hydrogen ion concentration (pH), acidity, alkalinity, chlorides, phosphates, total hardness, sulphates, total solids, dissolved solids, suspended solids, nitrate nitrogen, ammoniacal nitrogen, dissolved oxygen, biological oxygen demand (BOD) and chemical oxygen demand (COD). Efficacy of the formulation was assessed in terms of reduction in larval population in the treated habitats. pH of water in the treated sites was around neutral range (mean +/- SD 7.65 +/- 0.23). Phosphate content was low (2.27 +/- 1.34 ppm) whereas chlorides (326.1 +/- 55.8 ppm) and sulphates (38.9 +/- 23.8 ppm) were high. Total hardness ranged from 206 to 462.5 ppm with a mean of 312.1 +/- 80.5 ppm. The chlorides and sulphates, though present in considerable quantity, did not have any influence on the efficacy of B. sphaericus formulation. However, the proportion of insoluble chlorides and sulphates which contribute to total hardness seemed to influence the formulation adversely.

Efficacy of Bacillus sphaericus in different breeding habitats of Culex quinquefasciatus

'Spherifix', an alginate based slow release formulation of Bacillus sphaericus was field tested in different types of breeding habitats of Culex quinquefasciatus at the dose of 15 kg ai/ha at bimonthly interval. The efficacy of the formulation was higher in most of the months except in rainy and post-rainy months. The mean percentage reduction +/-SD during the treatment phase of one year was 31.2 +/- 17.9, 50 +/- 29.4, 28.3 +/- 17.6, 30.3 +/- 21.1, 66 +/- 22.5 and 53 +/- 20.4 in larval density and 49 +/- 20.8, 65.1 +/- 26.1, 30.3 +/- 21.9, 59.8 +/- 22.6, 63.1 +/- 21.9 and 47.7 +/- 24.2 in pupal density respectively in cement tanks, cesspools, cesspits, disused wells, unlined drains and cement lined drains. The reduction in immature density was relatively higher in undisturbed, debris free and shallow habitats such as cesspools, unlined drains and cement lined drains. After withdrawal of treatment, the effect of the formulation could be seen for a period of four months.