Multimodal data fusion using sparse canonical correlation analysis and cooperative learning: a COVID-19 cohort study

Through technological innovations, patient cohorts can be examined from multiple views with high-dimensional, multiscale biomedical data to classify clinical phenotypes and predict outcomes. Here, we aim to present our approach for analyzing multimodal data using unsupervised and supervised sparse linear methods in a COVID-19 patient cohort. This prospective cohort study of 149 adult patients was conducted in a tertiary care academic center. First, we used sparse canonical correlation analysis (CCA) to identify and quantify relationships across different data modalities, including viral genome sequencing, imaging, clinical data, and laboratory results. Then, we used cooperative learning to predict the clinical outcome of COVID-19 patients: Intensive care unit admission. We show that serum biomarkers representing severe disease and acute phase response correlate with original and wavelet radiomics features in the LLL frequency channel (cor(Xu1, Zv1) = 0.596, p value < 0.001). Among radiomics features, histogram-based first-order features reporting the skewness, kurtosis, and uniformity have the lowest negative, whereas entropy-related features have the highest positive coefficients. Moreover, unsupervised analysis of clinical data and laboratory results gives insights into distinct clinical phenotypes. Leveraging the availability of global viral genome databases, we demonstrate that the Word2Vec natural language processing model can be used for viral genome encoding. It not only separates major SARS-CoV-2 variants but also allows the preservation of phylogenetic relationships among them. Our quadruple model using Word2Vec encoding achieves better prediction results in the supervised task. The model yields area under the curve (AUC) and accuracy values of 0.87 and 0.77, respectively. Our study illustrates that sparse CCA analysis and cooperative learning are powerful techniques for handling high-dimensional, multimodal data to investigate multivariate associations in unsupervised and supervised tasks.

Multimodal Biomedical Data Fusion Using Sparse Canonical Correlation Analysis and Cooperative Learning: A Cohort Study on COVID-19

Through technological innovations, patient cohorts can be examined from multiple views with high-dimensional, multiscale biomedical data to classify clinical phenotypes and predict outcomes. Here, we aim to present our approach for analyzing multimodal data using unsupervised and supervised sparse linear methods in a COVID-19 patient cohort. This prospective cohort study of 149 adult patients was conducted in a tertiary care academic center. First, we used sparse canonical correlation analysis (CCA) to identify and quantify relationships across different data modalities, including viral genome sequencing, imaging, clinical data, and laboratory results. Then, we used cooperative learning to predict the clinical outcome of COVID-19 patients. We show that serum biomarkers representing severe disease and acute phase response correlate with original and wavelet radiomics features in the LLL frequency channel (corr(Xu1, Zv1) = 0.596, p-value < 0.001). Among radiomics features, histogram-based first-order features reporting the skewness, kurtosis, and uniformity have the lowest negative, whereas entropy-related features have the highest positive coefficients. Moreover, unsupervised analysis of clinical data and laboratory results gives insights into distinct clinical phenotypes. Leveraging the availability of global viral genome databases, we demonstrate that the Word2Vec natural language processing model can be used for viral genome encoding. It not only separates major SARS-CoV-2 variants but also allows the preservation of phylogenetic relationships among them. Our quadruple model using Word2Vec encoding achieves better prediction results in the supervised task. The model yields area under the curve (AUC) and accuracy values of 0.87 and 0.77, respectively. Our study illustrates that sparse CCA analysis and cooperative learning are powerful techniques for handling high-dimensional, multimodal data to investigate multivariate associations in unsupervised and supervised tasks.

Molecular classification and biomarkers of clinical outcome in breast ductal carcinoma in situ: Analysis of TBCRC 038 and RAHBT cohorts

Ductal carcinoma in situ (DCIS) is the most common precursor of invasive breast cancer (IBC), with variable propensity for progression. We perform multiscale, integrated molecular profiling of DCIS with clinical outcomes by analyzing 774 DCIS samples from 542 patients with 7.3 years median follow-up from the Translational Breast Cancer Research Consortium 038 study and the Resource of Archival Breast Tissue cohorts. We identify 812 genes associated with ipsilateral recurrence within 5 years from treatment and develop a classifier that predicts DCIS or IBC recurrence in both cohorts. Pathways associated with recurrence include proliferation, immune response, and metabolism. Distinct stromal expression patterns and immune cell compositions are identified. Our multiscale approach employed in situ methods to generate a spatially resolved atlas of breast precancers, where complementary modalities can be directly compared and correlated with conventional pathology findings, disease states, and clinical outcome.

Abstract GS4-07: The Breast PreCancer Atlas DCIS genomic signatures define biology and correlate with clinical outcomes: An analysis of TBCRC 038 and RAHBT cohorts

Background. DCIS consists of a molecularly heterogeneous group of premalignant lesions, with variable risk of invasive progression. Understanding biomarkers for invasive progression could help individualize treatment recommendations based upon tumor biology. As part of the NCI Human Tumor Atlas Network (HTAN), we conducted comprehensive genomic analyses on two large DCIS case-control cohorts. Methods. We performed smart3-seq and low-pass whole genome sequencing on two independent, retrospective, longitudinally sampled DCIS case-control cohorts. TBCRC 038 was a multicenter cohort diagnosed with DCIS between 1998 and 2016 at one of the Translational Breast Cancer Research sites; the RAHBT (Resource of Archival Human Breast Tissue) cohort included women identified through the St. Louis Breast Tissue Repository, and the Women’s Health Repository diagnosed between 1997 and 2001. We studied the spectrum of molecular changes present and sought genomic predictors of subsequent ipsilateral breast events (iBEs: DCIS recurrence or invasive progression) in both DCIS epithelium and stroma in formalin fixed paraffin embedded tissue. We generated de novo tumor and stroma-centric subtypes for DCIS that represents fundamental transcriptomic organization. Copy number analysis was performed using low-pass DNA sequencing. Non-negative matrix factorization (NMF) was applied to the RNA expression of all coding genes to identify clusters. A negative-binomial regression model was used to identify differentially expressed genes. Results. We analyzed 677 DCIS samples from 481 patients with 7.1 years median follow-up. In TBCRC samples, we identified three clusters via NMF in TBCRC referred to as ER low, quiescent, and ER high. The ER-low cluster had significantly higher levels of ERBB2 and lower levels of ESR1 compared to quiescent and ER-high clusters. Quiescent cluster lesions were less proliferative and less metabolically active than ER high and ER low subtypes. These findings were replicated in the RAHBT cohort. Focusing on the stromal component of DCIS from laser capture microdissection in RAHBT samples, we identified four distinct DCIS-associated stromal clusters. A “normal-like” stromal cluster with ECM organization and PI3K-AKT signaling; a “collagen-rich” stromal cluster; a “desmoplastic” stromal cluster with high fibroblast and total myeloid abundance, mostly associated with macrophages and myeloid dendritic cells (mDC); and an “immune-dense” stromal cluster. Further, we compared differentially expressed genes in patients with or without subsequent iBEs within 5 years of diagnosis. Hypothesizing that the resulting 812 DE genes (DESeq2) represent multiple routes to subsequent iBEs, we leveraged NMF to identify paths to progression. In both TBCRC and RAHBT cohorts, poor outcome groups exhibited increased ER, MYC signaling, and oxidative phosphorylation, supporting that these pathways are important for DCIS recurrence and progression. Conclusion. Comprehensive genomic profiling in two independent DCIS cohorts with longitudinal outcomes shows distinct DCIS stromal expression patterns and immune cell composition. RNA expression profiles reveal underlying tumor biology that is associated with later iBEs in both cohorts. These studies provide new insight into DCIS biology and will guide the design of diagnostic strategies to prevent invasive progression.

Citation Format: Siri H Strand, Belén Rivero-Gutiérrez, Kathleen E Houlahan, Jose A Seoane, Lorraine M King, Tyler Risom, Lunden Simpson, Sujay Vennam, Aziz Khan, Timothy Hardman, Bryan E Harmon, Fergus J Couch, Kristalyn Gallagher, Mark Kilgore, Shi Wei, Angela DeMichele, Tari King, Priscilla F McAuliffe, Julie Nangia, Joanna Lee, Jennifer Tseng, Anna Maria Storniolo, Alastair Thompson, Gaorav Gupta, Robyn Burns, Deborah J Veis, Katherine DeSchryver, Chunfang Zhu, Magdalena Matusiak, Jason Wang, Shirley X Zhu, Jen Tappenden, Daisy Yi Ding, Dadong Zhang, Jingqin Luo, Shu Jiang, Sushama Varma, Cody Straub, Sucheta Srivastava, Christina Curtis, Rob Tibshirani, Robert Michael Angelo, Allison Hall, Kouros Owzar, Kornelia Polyak, Carlo Maley, Jeffrey R Marks, Graham A Colditz, E Shelley Hwang, Robert B West. The Breast PreCancer Atlas DCIS genomic signatures define biology and correlate with clinical outcomes: An analysis of TBCRC 038 and RAHBT cohorts [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr GS4-07.

The Effectiveness of Multitask Learning for Phenotyping with Electronic Health Records Data

Electronic phenotyping is the task of ascertaining whether an individual has a medical condition of interest by analyzing their medical record and is foundational in clinical informatics. Increasingly, electronic phenotyping is performed via supervised learning. We investigate the effectiveness of multitask learning for phenotyping using electronic health records (EHR) data. Multitask learning aims to improve model performance on a target task by jointly learning additional auxiliary tasks and has been used in disparate areas of machine learning. However, its utility when applied to EHR data has not been established, and prior work suggests that its benefits are inconsistent. We present experiments that elucidate when multitask learning with neural nets improves performance for phenotyping using EHR data relative to neural nets trained for a single phenotype and to well-tuned baselines. We find that multitask neural nets consistently outperform single-task neural nets for rare phenotypes but underperform for relatively more common phenotypes. The effect size increases as more auxiliary tasks are added. Moreover, multitask learning reduces the sensitivity of neural nets to hyperparameter settings for rare phenotypes. Last, we quantify phenotype complexity and find that neural nets trained with or without multitask learning do not improve on simple baselines unless the phenotypes are sufficiently complex.

The influence of a history of clinical depression on peri-operative outcomes in elective total shoulder arthroplasty: a ten-year national analysis

AIMS

Depression can significantly affect quality of life and is associated with higher rates of medical comorbidities and increased mortality following surgery. Although depression has been linked to poorer outcomes following orthopaedic trauma, total joint arthroplasty and spinal surgery, we wished to examine the impact of depression in elective total shoulder arthroplasty (TSA) as this has not been previously explored.

PATIENTS AND METHODS

The United States Nationwide Inpatient Sample (NIS) was used to identify patients undergoing elective TSA over a ten-year period. Between 2002 and 2012, 224 060 patients underwent elective TSA.

RESULTS

Among the identified patients who had undergone TSA, 12.4% had a diagnosis of a history of depression. A diagnosis of depression was twice as common in women compared with men (16.0% vs 8.0%, p < 0.001), and more frequent in those with low income and Medicaid insurance (p < 0.001). A diagnosis of depression was an independent risk factor for post-operative delirium (odds ratio (OR) 2.29, p < 0.001), anaemia (OR 1.65, p < 0.001), infection (2.09, p = 0.045) and hospital discharge to a placement other than home (OR 1.52, p < 0.001) CONCLUSION: A history of clinical depression is present in 12.4% of patients undergoing elective TSA and the disease burden is projected to increase further in the future. Depression is often underdiagnosed and pre-operative screening and appropriate peri-operative management of patients is encouraged.

TAKE HOME MESSAGE

The awareness that clinical depression is associated with increased complications following total shoulder arthroplasty provides physicians an opportunity for early intervention in this at-risk population. Cite this article: Bone Joint J 2016;98-B:818-24.

Heroin abuse and nitric oxide, oxidation, peroxidation, lipoperoxidation

To further reveal the risks of heroin abuse to human body, and to determine the injuries of oxidation, peroxidation and lipoperoxidation induced by nitric oxide and other free radicals to heroin abusers, we determined and compared plasma values of lipoperoxides (LPO), nitric oxide (NO), vitamin C (VC), vitamin E (VE), beta-carotene (beta-CAR) and erythrocyte values of LPO, superoxide dismutase (SOD), catalase (CAT), and glutathione peroxidase (GSH-Px) in 114 heroin abusers and 100 healthy volunteers. Using linear regression and correlation as well as stepwise regression and correlation, we also analyzed the effect of the abusing duration, and daily abusing quantity on the above-mentioned biochemical parameters in the heroin abusers. The results showed that, compared with the healthy volunteer groups, the average plasma values of LPO, and NO, and the average erythrocyte value of LPO in the heroin abuser group were significantly increased (P < 0.0001), and the average plasma values of VC, VE, and beta-CAR and the average erythrocyte values of SOD, CAT, and GSH-Px were significantly decreased (P < 0.0001). Analysis of linear regression and correlation showed that with prolonged heroin abusing and with increased daily quantity in the heroin abusers, the plasma values of LPO, and NO, and the erythrocyte value of LPO were gradually increased (P < 0.001), whereas the plasma values of VC, VE, and beta-CAR and the erythrocyte values of SOD, CAT, and GSH-Px were gradually decreased (P < 0.001). Analysis of stepwise regression and correlation indicated that the plasma values of NO, VC and VE were closely correlated with the abusing duration and daily abusing quantity. These results indicate that the balance between oxidation and antioxidation in the heroin abusers was seriously disturbed, and the injuries induced by nitric oxide and other free radicals, through oxidation, peroxidation and lipoperoxidation to the bodies of heroin abusers exacerbated. It is therefore necessary that in abstaining from heroin dependence, the heroin abusers should acquire sufficient quantities of antioxidants such as VC, VE and beta-CAR.

Effects of cigarette smoking and smoking cessation on plasma constituents and enzyme activities related to oxidative stress

In order to study effects of cigarette smoking and smoking cessation on plasma constituents and enzyme activities related to oxidative stress, 1255 smokers and 524 healthy non-smokers were investigated in terms of plasma levels of lipoperoxides (LPO), nitric oxide (NO), vitamin C (VC), vitamin E (VE) and beta-carotene (beta-CAR). Additionally, erythrocytes were examined to determine the level of LPO, the activities of superoxide dismutase (SOD), catalase (CAT) and glutathione peroxidase (GSH-Px). The results showed that, when compared with the average values of the non-smoker group, the average plasma values of LPO, NO and the average erythrocyte value of LPO in the smoker group were significantly increased (P < 0. 001), while the average plasma values of VC, VE, beta-CAR, and the average erythrocyte activities of SOD, CAT, GSH-Px were significantly decreased (P < 0.001). A linear regression and correlation analysis for 65 male smokers who were all 40 years old showed that with longer smoking duration and greater daily smoking quantity, the plasma values of LPO, NO and the erythrocyte value of LPO were elevated, while the plasma values of VC, VE, beta-CAR and erythrocyte values of SOD, CAT, GSH-Px were decreased. In a group of 73 smokers who stopped smoking completely for six months, the average plasma values of LPO, NO and the average erythrocyte value of LPO decreased, although they were still significantly higher than those in the matched non-smoker group (P < 0.05). Additionally, the average plasma values of VC, VE, beta-CAR and the average erythrocyte values of SOD, CAT, GSH-Px increased, although they were still significantly lower than those in the matched non-smoker group (P < 0.05). However, after smoking cessation for one year the above average values were not significantly different from those in the matched non-smoker group (P > 0.05). This finding indicates that the markedly increased oxidative stress in smokers might gradually return to normal but only after a long period of smoking cessation. In conclusion, in the bodies of smokers a series of free radical chain reactions were gravely aggravated, the dynamic balance between oxidation and antioxidation was seriously disrupted, and oxidative stress was clearly exacerbated, which is closely related to many disorders or diseases in smokers. The present study underscored the need, urgency and importance of complete smoking cessation.

Genetic polymorphism of 4'-hydroxylation of S-mephenytoin in 148 Chinese of Han nationality

AIM

To study genetic polymorphism of S-mephenytoin (S-Mep) 4'-hydroxylation in the Chinese population of Han nationality.

METHODS

The lg metabolic ratio (MR) and lg hydroxylation index (HI) in the urine (0-12 h) after oral administration with 100 mg of racemic Mep tablet were determined by HPLC method in 148 consangeously unrelated native Chinese subjects and 21 individuals of 5 families.

RESULTS

The lg MR and lg HI showed a bimodal distribution with an antimode of -1.00 and 1.50, respectively. The occurrence of poor metabolizers (PM) was 13.5% in the population. The pedigree analysis in 5 families indicated that deficient S-Mep hydroxylation was an autosomal recessive trait.

CONCLUSIONS

The occurrence of PM for S-Mep 4-hydroxylation in Chinese was higher than that of the Caucasians, and both genetic modes were of autosomal recessive trait.

Effects of naloxone on tissue oxygen supply and somatosensory evoked potentials in cat brain during focal cerebral ischemia

The effects of naloxone on local tissue oxygen partial pressure (pO2) and on the somatosensory evoked potentials (SEP) were studied in the brain of cat during focal cerebral ischemia. Ischemia, produced by clamping of the middle cerebral artery (MCA) by a transorbital approach, was performed in two series of cats for 3 h. In one group of cats (n = 5), naloxone 5 mg.kg-1 was injected i.v. 0.5 h after clamping. The pO2 was continuously recorded on different depths (0-5000 microns) of the median gyrus by a polarographic oxygen microelectrode. After MCA clamping, pO2 (depth of 0-1000 microns) decreased markedly and hypoxia occurred in the ischemic area. But in the deeper brain (1001-5000 microns) pO2 did not change significantly. The amplitude of SEP decreased after MCA clamping, while the conduction time of SEP did not significantly decrease. The mean pO2 values in the ischemic area were increased as compared to the control group after naloxone, especially at the depths of 0-1000 microns, and the hypoxia was improved. The amplitude of SEP was increased after naloxone in comparison to the situation of ischemia without naloxone. The conduction time of SEP was not improved significantly. We conclude that naloxone can improve the oxygen supply and the electrical activity of neurons in the ischemic region of the brain.

Effect of nicardipine on somatosensory evoked potentials in patients with acute cerebral infarction

We evaluated the effect of nicardipine, a calcium channel blocker, on somatosensory evoked potentials (SEP) in 26 patients with acute cerebral infarction. Post treatment, 58% (15/26) of the N20 and P25 latencies were prolonged in the affected hemispheres; 8% (2/26) were shortened; and 35% (9/26) did not change. The mean N20 and P25 latencies were significantly prolonged two hours post treatment in the affected hemisphere (N20, P less than 0.01, P25 P less than 0.01). Nicardipine (Ni) had no effect on SEP components in the intact hemispheres. Seventy five per cent of the 12 patients with hypertension had a decrease in blood pressure (BP) after taking nicardipine, but there were no undesirable side effects or worsening of neurological signs. Our study demonstrates that nicardipine prolongs the latencies of short-latency components of SEP in the affected hemisphere after acute ischaemic stroke and also decreases BP. These observations suggest that nicardipine therapy might impair neuronal function in the ischaemic zone.

Electrophysiological measures during acupuncture-induced surgical analgesia

Electrophysiological recordings (electroencephalograms, somatosensory-evoked potentials, cognitive-evoked potentials) were made in five patients during acupuncture-induced analgesia for removal of a thyroid tumor. The electrophysiological measures were unchanged during the operation. Acupuncture's modes of action in producing analgesia are not revealed in this study.