Stimulus-Response Patterns: The Key to Giving Generalizability to Text-based Depression Detection Models.

Text content analysis for depression detection using machine learning techniques has become a prominent area of research. However, previous studies focused mainly on analyzing the textual content, neglecting the fundamental factors driving text generation. Consequently, existing models face the challenge of poor generalization to out-of-domain data as they struggle to capture the crucial features of depression. To address this, we propose a novel computational perspective of "stimulus-response patterns" that brings us closer to the essence of clinical diagnosis of depression. Adopting this computational perspective allows us to conceptually unify diverse datasets and generalize this perspective to common datasets in the field. We introduce the Stimulus-Response Patterns-aware Network (SRP-Net) as an exemplary approach within this computational perspective. To assess the performance of the SRP-Net, we constructed a multi-stimulus dataset and conducted experimental evaluations, demonstrating its exceptional cross-stimulus generalizability. Furthermore, we demonstrated the promising performance of SPR-Net in real medical scenarios and conducted an interpretability analysis of the stimulus-response patterns. Our research investigates the critical role of stimulus-response patterns in enhancing the generalizability of text-based depression detection models, which can potentially facilitate data-driven depression detection to approach the diagnostic accuracy of psychiatrists.

Real-World Maintenance Phase Persistence on Ustekinumab and Adalimumab in Ulcerative Colitis.

Purpose: To describe real-world persistence in bio-naïve and bio-experienced adults with ulcerative colitis (UC) treated with ustekinumab, a recently approved anti-interleukin 12/23 antibody, or adalimumab, an anti-TNF biologic. Methods: This is a descriptive, retrospective cohort study. Patients initiating ustekinumab or adalimumab (index date, between 10/21/2019 and 08/13/2021) were selected from the Komodo Health comprehensive dataset and stratified into bio-naïve and bio-experienced subgroups based on biologic use 12 months pre-index date. Endpoints evaluated at 12-months after maintenance phase start using Kaplan-Meier analysis included 1) persistence; 2) persistence while being corticosteroid-free (<14 consecutive days of corticosteroid supply after day 90 post-index); and, 3) persistence while on monotherapy (no immunomodulators/non-index biologics/advanced therapies). Results: Ustekinumab cohort included 778 patients (236 bio-naïve, 542 bio-experienced) and adalimumab cohort included 1693 patients (1517 bio-naive, 176 bio-experienced). At 12 months after maintenance phase start, 75.5% and 50.5% of bio-naïve patients persisted on ustekinumab and adalimumab and 72.3% and 56.9% of bio-experienced patients persisted on ustekinumab and adalimumab, respectively. Further, 55.1% and 38.2% of bio-naïve patients were persistent and corticosteroid-free with ustekinumab and adalimumab; 43.7% and 33.4% of bio-experienced patients were persistent and corticosteroid-free with ustekinumab and adalimumab, respectively. Moreover, 68.1% and 44.5% of bio-naïve patients were persistent and on monotherapy with ustekinumab and adalimumab; 61.6% and 47.9% of bio-experienced patients were persistent and on monotherapy with ustekinumab and adalimumab, respectively. Conclusion: At 12 months after maintenance phase start, patients with UC treated with ustekinumab had numerically higher persistence, including persistence while corticosteroid-free and persistence while on monotherapy, than patients treated with adalimumab.

Neoadjuvant Chemotherapy With Oxaliplatin and Fluoropyrimidine Versus Upfront Surgery for Locally Advanced Colon Cancer: The Randomized, Phase III OPTICAL Trial.

PURPOSE: The role of neoadjuvant chemotherapy (NAC) in colon cancer remains unclear. This trial investigated whether 3 months of modified infusional fluorouracil, leucovorin, and oxaliplatin (mFOLFOX6) or capecitabine and oxaliplatin (CAPOX) as NAC could improve outcomes in patients with locally advanced colon cancer versus upfront surgery. PATIENTS AND METHODS: OPTICAL was a randomized, phase III trial in patients with clinically staged locally advanced colon cancer (T3 with extramural spread into the mesocolic fat ≥5 mm or T4). Patients were randomly assigned 1:1 to receive six preoperative cycles of mFOLFOX6 or four cycles of CAPOX, followed by surgery and adjuvant chemotherapy (NAC group), or immediate surgery and the physician's choice of adjuvant chemotherapy (upfront surgery group). The primary end point was 3-year disease-free survival (DFS) assessed in the modified intention-to-treat (mITT) population. RESULTS: Between January 2016 and April 2021, of the 752 patients enrolled, 744 patients were included in the mITT analysis (371 in the NAC group; 373 in the upfront surgery group). At a median follow-up of 48.0 months (IQR, 46.0-50.1), 3-year DFS rates were 82.1% in the NAC group and 77.5% in the upfront surgery group (stratified hazard ratio [HR], 0.74 [95% CI, 0.54 to 1.03]). The R0 resection was achieved in 98% of patients who underwent surgery in both groups. Compared with upfront surgery, NAC resulted in a 7% pathologic complete response rate (pCR), significantly lower rates of advanced tumor staging (pT3-4: 77% v 94%), lymph node metastasis (pN1-2: 31% v 46%), and potentially improved overall survival (stratified HR, 0.44 [95% CI, 0.25 to 0.77]). CONCLUSION: NAC with mFOLFOX6 or CAPOX did not show a significant DFS benefit. However, this neoadjuvant approach was safe, resulted in substantial pathologic downstaging, and appears to be a viable therapeutic option for locally advanced colon cancer.

Comparative long-term outcomes of natural orifice specimen extraction surgery and conventional laparoscopic colectomy for left-sided colorectal cancer: a propensity score-matched analysis.

BACKGROUND: Natural orifice specimen extraction surgery (NOSES) is currently widely used in left-sided colorectal cancer. Some clinical comparative studies have been conducted, providing evidence of its safety and oncological benefits. However, these studies are typically characterized by small sample sizes and short postoperative follow-up periods. Consequently, in this research, the authors adopt the propensity score matching method to undertake a large-scale retrospective comparative study on NOSES colectomy for left-sided colorectal cancer, with the goal of further augmenting the body of evidence-based medical support for NOSES. METHODS: This retrospective study involved patients who underwent NOSES colectomy and conventional laparoscopic (CL) colectomy for left-sided colorectal cancer between January 2014 and April 2021. In the NOSES group, specimens were extracted through the anus with the help of a Cai tube (homemade invention: ZL201410168748.2). The patients were matched at a ratio of 1:1 according to age, sex, BMI, tumor diameter, tumor location (descending and splenic flexure colon/ sigmoid colon/ middle and upper rectum), tumor height from anal verge, ASA grade, previous abdominal surgery, clinical pathologic stage, preoperative CEA. After matching, 132 patients in the NOSES group and 132 patients in the CL group were eligible for analysis. RESULTS: Compared with CL group, NOSES group was associated with decreased postoperative maximum pain score (2.6±0.7 vs. 4.7±1.7, P=0.000), less additional analgesia required (6.8 vs. 34.8%, P=0.000), faster time to passage of flatus (2.3±0.6 days vs. 3.3±0.7 days, P=0.000), less wound infection (0.0 vs. 6.1%, P=0.007), and longer operative time (212.5±45.8 min vs. 178.0±43.4 min, P=0.000). No significant differences were observed in estimated blood loss, time to resume regular diet, postoperative hospital stay, conversion to open surgery or conventional minilaparotomy, total morbidity, readmission, mortality, pathologic outcomes, and Wexner incontinence score between groups. After a median follow-up of 63.0 months, the 5-year overall survival rates were 88.3 versus 85.0% (P=0.487), disease-free survival rates were 82.9 versus 83.6% (P=0.824), and the local recurrence rates were 4.4 versus 4.0% (P=0.667) in the NOSES and CL groups, respectively. CONCLUSIONS: This study suggests that NOSES colectomy using a Cai tube for left-sided colorectal cancer is a safe and feasible option with better cosmetic results, less pain, faster recovery of gastrointestinal function, and comparable long-term clinical and oncologic outcomes to CL colectomy.

Burden of chronic corticosteroid use among patients with ulcerative colitis initiated on targeted treatment or conventional therapy in the United States.

BACKGROUND: Chronic corticosteroid use is common in ulcerative colitis (UC); however, real-world evidence of its burden to the health care system is limited. OBJECTIVE: To quantify chronic corticosteroid use burden in UC. METHODS: Adults with UC initiated on targeted treatments (ie, biologics and advanced/small molecule therapies) or conventional therapy (index date) were selected from a deidentified US insurance claims database (January 1, 2004, to September 30, 2021). Targeted treatments and conventional therapy initiators were stratified into chronic (>90 days corticosteroid use 12 months post-index [landmark]) and nonchronic corticosteroid users. Patient characteristics 12 months pre-index were balanced with inverse probability of treatment weighting. Health care resource use, costs (US$ 2021), and corticosteroid-related complications were compared in the 12 months post-landmark. RESULTS: Targeted treatment initiators included 1,886 chronic and 1,911 nonchronic corticosteroid users; conventional therapy initiators included 4,980 chronic and 5,199 nonchronic users. Chronic vs nonchronic users had 94% more inpatient days and 16% more outpatient visits among targeted treatment initiators, and 135% more inpatient days and 30% more outpatient visits among conventional therapy initiators (all P < 0.01). Mean all-cause total costs per patient per year were $73,491 for chronic vs $58,884 for nonchronic users ($14,607 higher; P < 0.01) for targeted treatment initiators, and $39,335 for chronic vs $21,271 for nonchronic users ($18,065 higher; P < 0.01) for conventional therapy initiators. Odds of infection and bone loss were 14% and 113% higher, respectively, in chronic vs nonchronic users among targeted treatment initiators and 29% and 47% higher in chronic vs nonchronic users among conventional therapy initiators (all P < .01). CONCLUSIONS: The results of this study suggest that chronic corticosteroid use is associated with substantial clinical and economic burden and may indicate unmet needs in the management of UC progression.

Economic and clinical burden of chronic corticosteroid use in patients with Crohn[apos]s disease initiated on biologic or conventional therapies in the US: A retrospective claims study.

BACKGROUND: Chronic corticosteroid (CS) use is associated with complications, but estimates of the economic and clinical burden in patients with Crohn's disease (CD) are lacking. OBJECTIVE: To estimate the burden of chronic CS use in CD in the United States in terms of health care resource utilization (HRU), health care costs, and CS-related complications. METHODS: This was a retrospective study of adults with CD initiated on biologics or conventional therapies (index date). Patients from a deidentified insurance claims database (2004-2021) were classified as chronic CS users (>90 days of CS use) or nonchronic CS users based on a 12-month landmark period starting on the index date. Patient baseline characteristics were balanced, and outcomes (HRU, costs [2021 US dollars], and CS-related complications) 12 months after the landmark period were compared between CS groups using regressions with nonparametric bootstrap resampling to estimate confidence intervals and P values. RESULTS: Biologic initiators (mean age: 44 years, 55% female) included 3366 chronic and 3401 nonchronic CS users; conventional therapy initiators (mean age: 51 years, 59% female) included 3657 chronic and 3727 nonchronic CS users. Compared with nonchronic users, chronic users had significantly more inpatient days and outpatient visits (biologic initiators: 37% and 24% more, respectively; conventional therapy initiators: 36% and 17%, respectively; all P<0.05). Chronic users also had significantly higher mean all-cause total costs per-patient-per year (biologic: $72,967 vs. $63,100, mean cost difference [MCD] = $9867; conventional therapy: $40,144 vs. $26,426, MCD = $13,718; all P<0.001), as well as higher odds of infection (biologic: 14% higher; conventional therapy: 20% higher) and bone loss (63% and 41%, respectively) (all P<0.05). CONCLUSION: Chronic CS use in patients with CD is associated with a significant economic and clinical burden including higher HRU, health care costs, and prevalence of complications, suggesting unmet needs in the clinical management of this population.

Evaluation of Treatment Patterns and Maintenance Dose Titration Among Patients With Crohn's Disease Initiating Biologics With 3 Years of Follow-Up.

Background: There is limited real-world evidence on treatment patterns of patients with Crohn's disease (CD) initiating biologics with an extensive follow-up period. This study describes persistence and dose titration among CD patients with 3 years of follow-up. Methods: This retrospective observational study was conducted using the STATinMED RWD Insights all-payer medical and pharmacy data. Adult patients with at least 1 CD medical claim and at least 1 medical/pharmacy claim for a biologic (adalimumab [ADA], certolizumab pegol (CZP), infliximab [IFX] and its biosimilar products [IFX-BS], ustekinumab [UST], and vedolizumab [VDZ]) between September 2016 and October 2018 were identified. Commercially insured patients with continuous capture for at least 12 months before and at least 36 months after biologics initiation were selected. Confirmed CD patients were included in the final cohort. Baseline patient characteristics and treatment patterns over the 3-year follow-up period were evaluated. Results were summarized using means and SD or counts and percentages. Results: A total of 2309 confirmed patients with CD were identified (847 [36.7%] IFX, 534 [23.1%] ADA, 486 [21.1%] VDZ, 394 [17.1%] UST, 85 [3.7%] CZP, and 72 [3.1%] IFX-BS). CZP and IFX-BS were excluded due to small sample sizes. Approximately half of CD patients were between ages 35 and 54. Patients on UST had a higher Charlson Comorbidity Index score. Common comorbidities (>10%) included anemia, anxiety, depression, and hypertension. Persistence over 3 years' follow-up was highest for UST (61.4%) patients, followed by VDZ (58.0% ), ADA (52.1% , and IFX (48.1%). The discontinuation rate without switch or restart was highest for ADA (37.3%), followed by UST (30.7%), IFX (28.1%), and VDZ (25.3%). Over the 3 years of follow-up, the dose titration rate was highest for IFX (76.5%) and lowest for UST (50.8%). In particular, UST had the lowest dose escalation rate (35.5%) and highest dose-reduction rate (16.5%). Conclusions: Patients with CD on UST had the highest persistence and lowest dose escalation across different biologic users over the 3-year follow-up period, possibly suggesting a better clinical response of UST. Future studies with longer follow-up adjusting for confounders are needed to better understand treatment patterns among biologics users.

Cholecystectomy Significantly Alters Gut Microbiota Homeostasis and Metabolic Profiles: A Cross-Sectional Study.

Cholecystectomy (CCE) is a standard clinical treatment for conditions like gallstones and cholecystitis. However, its link to post-CCE syndrome, colorectal cancer, and nonalcoholic fatty liver disease has raised concerns. Additionally, studies have demonstrated the disruptive effects of CCE on gut microbiota homeostasis and bile acid (BA) metabolism. Considering the role of gut microbiota in regulating host metabolic and immune pathways, the use of dietary and probiotic intervention strategies to maintain a stable gut ecosystem after CCE could potentially reduce associated disease risks. Inter-study variations have made it challenging to identify consistent gut microbiota patterns after CCE, a prerequisite for targeted interventions. In this study, we first meta-analyzed 218 raw 16S rRNA gene sequencing datasets to determine consistent patterns of structural and functional changes in the gut microbiota after CCE. Our results revealed significant alterations in the gut microbiota's structure and function due to CCE. Furthermore, we identified characteristic gut microbiota changes associated with CCE by constructing a random model classifier. In the validation cohort, this classifier achieved an area under the receiver operating characteristic curve (AUC) of 0.713 and 0.683 when distinguishing between the microbiota of the CCE and healthy groups at the family and genus levels, respectively. Further, fecal metabolomics analysis demonstrated that CCE also substantially modified the metabolic profile, including decreased fecal short-chain fatty acid levels and disrupted BA metabolism. Importantly, dietary patterns, particularly excessive fat and total energy intake, influenced gut microbiota and metabolic profile changes post-CCE. These dietary habits were associated with further enrichment of the microbiota related to BA metabolism and increased levels of intestinal inflammation after CCE. In conclusion, our study identified specific alterations in gut microbiota homeostasis and metabolic profiles associated with CCE. It also revealed a potential link between dietary patterns and gut microbiota changes following CCE. Our study provides a theoretical basis for modulating gut microbiota homeostasis after CCE using long-term dietary strategies and probiotic interventions.

Long-term persistence and other treatment patterns among bio-naïve patients with Crohn's disease treated with ustekinumab or adalimumab.

OBJECTIVE: To estimate long-term persistence among bio-naïve patients with CD initiated on ustekinumab or adalimumab. METHODS: Adults with CD initiating ustekinumab or adalimumab (index date, between September 23, 2016 and August 1, 2019) were sampled from the IBM MarketScan Commercial Database. Patients without CD-indicated biologics (bio-naïve) and with no diagnoses for other autoimmune diseases 12 months pre-index date (baseline) were included. Cohorts were balanced on baseline characteristics with inverse probability of treatment weighting. Persistence was defined as the absence of therapy exposure gaps >120 days (ustekinumab) or >60 (adalimumab) between days of supply. Composite endpoints were persistence and being corticosteroid-free (no corticosteroids >14 days of supply after day 90 post-index) and persistence while on monotherapy (no immunomodulators/non-index biologics). Persistence was analyzed using Kaplan-Meier and Cox's models. RESULTS: Ustekinumab and adalimumab cohorts included 671 and 2,975 patients. At 12 months post-index, ustekinumab patients were significantly more persistent (hazard ratio [HR] = 1.60; 95% confidence interval [CI] = 1.33-1.93), persistent while on monotherapy (HR = 1.43; 95% CI = 1.24-1.65), and trended toward being more persistent and corticosteroid-free (HR = 1.14; 95% CI = 0.99-1.30) vs adalimumab. At 24 months post-index, ustekinumab patients were significantly more persistent (HR = 1.66; 95% CI = 1.40-1.97), persistent while on monotherapy (HR = 1.44; 95% CI = 1.26-1.64), and persistent and corticosteroid-free (HR = 1.15; 95% CI = 1.01-1.31) vs adalimumab. CONCLUSIONS: Bio-naïve patients with CD initiated on ustekinumab demonstrated significantly more persistence than patients initiated on adalimumab at 12 and 24 months of treatment. Long-term persistence is a measure of a drug's real-world performance and findings may aid clinical decision-making.

Choosing a treatment on which a patient can stay over a long period of time is key for the successful management of chronic conditions such as Crohn's disease. Information on whether and how long patients stay on treatment can help physicians make the right therapeutic choice. This study examined whether adults with Crohn's disease, who have not previously taken biologics, stay on treatment longer when given the biologic ustekinumab or adalimumab. At 12 and 24 months after starting the treatment, a larger proportion of patients were still using ustekinumab compared with adalimumab. The proportion of patients using the biologic without immunomodulators or other biologics was also higher with ustekinumab. The results suggest that patients without previous biologic experience stay on treatment longer with ustekinumab than with adalimumab.

Metabolomic analysis reveals key metabolites and metabolic pathways in Suaeda salsa under salt and drought stress.

Suaeda salsa is an important salt- and drought-tolerant plant with important ecological restoration roles. However, little is known about its underlying molecular regulatory mechanisms. Therefore, understanding the response mechanisms of plants to salt and drought stress is of great importance. In this study, metabolomics analysis was performed to evaluate the effects of salt and drought stress on S. salsa . The experiment consisted of three treatments: (1) control (CK); (2) salt stress (Ps); and (3) drought stress (Pd). The results showed that compared with the control group, S. salsa showed significant differences in phenotypes under salt and drought stress conditions. First, a total of 207 and 292 differential metabolites were identified in the Ps/CK and Pd/CK groups, respectively. Second, some soluble sugars and amino acids, such as raffinose, maltopentoses, D -altro-beptulose, D -proline, valine-proline, proline, tryptophan and glycine-L -leucine, showed increased activity under salt and drought stress conditions, suggesting that these metabolites may be responsible for salt and drought resistance in S. salsa . Third, the flavonoid biosynthetic and phenylalanine metabolic pathways were significantly enriched under both salt and drought stress conditions, indicating that these two metabolic pathways play important roles in salt and drought stress resistance in S. salsa . The findings of this study provide new insights into the salt and drought tolerance mechanisms of S. salsa .

Transcriptomic study of Suaeda salsa in response to salt and drought stress.

Drought and salinity are the main factors limiting agricultural production. Improving crop resistance to relieve land stress is a major challenge in agriculture. The salt-tolerant species Suaeda salsa is a typical indicator of saline soil. It has a strong drought tolerance and can be used as a model plant to study salt and drought tolerance in plants. In this study, transcriptome sequencing and bioinformatic analysis were performed to study gene expression changes in S. salsa under salt and drought stresses, and to screen out differentially expressed genes. The genetic changes were most abundant in cellular processes, metabolic processes, ion binding, signalling, post-translational modifications, protein conversion, and molecular chaperones, suggesting that the above methods may play a significant role in the response of S. salsa to external salt and drought stress. Enrichment analysis showed that carbohydrate metabolic processes, oxidoreductase activity, transmembrane transport, kinase activity, cellular protein modification processes, and ion-binding pathways are involved in the stress response of S. salsa .

Persistence and other treatment patterns among bio-experienced patients with Crohn's disease initiated on ustekinumab or adalimumab.

BACKGROUND: Real-world data on persistence on ustekinumab and adalimumab among bio-experienced patients with Crohn's disease (CD) are limited. OBJECTIVE: To compare treatment persistence and describe switching, restart, and dose titration among bio-experienced patients with CD initiated on ustekinumab or adalimumab. METHODS: IBM MarketScan Commercial Database was used to identify bio-experienced adults with CD who were assigned to either the ustekinumab or adalimumab cohort based on the agent first initiated (index date) after September 23, 2016. Cohorts were balanced using inverse probability of treatment weights-average treatment effect on treated. Persistence on index agent (absence of exposure gap > 120 days for ustekinumab or > 60 days for adalimumab), persistence while corticosteroid-free, and persistence while receiving monotherapy were assessed at 12 months after index date and compared between cohorts using weighted Kaplan-Meier and Cox proportional hazards model analyses. RESULTS: Among 903 patients in the ustekinumab cohort and 525 patients in the adalimumab cohort, baseline characteristics were balanced after weighting. At 12 months post-index, ustekinumab was associated with higher persistence (80.1% vs 64.6%; hazard ratio = 2.02 [95% CI = 1.60-2.56]; P < 0.001) and persistence while receiving monotherapy (51.6% vs 40.0%; 1.51 [1.28-1.78]; P < 0.001) vs adalimumab. Persistence while corticosteroid-free was similar in the ustekinumab vs adalimumab cohort (50.1% vs 48.2%; 1.19 [1.00-1.41]; P = 0.0516). CONCLUSIONS: This retrospective real-world study demonstrated that among bio-experienced patients with CD, initiation of ustekinumab was associated with better persistence at 12 months of follow-up, including persistence while receiving monotherapy, compared with adalimumab. DISCLOSURES: This study was funded by Janssen Scientific Affairs, LLC. Drs Zhao, Ding, and Kachroo are employees of Janssen Scientific Affairs, LLC, and stockholders of Johnson & Johnson. Dr Manceur, Mr Lefebvre, Ms Zhdanava, and Mr Pilon are employees of Analysis Group, Inc., a consulting company that has provided paid consulting services to Janssen Scientific Affairs, LLC, which funded the development and conduct of this study and article. Mr Holiday was an employee of Analysis Group, Inc., at the time of study conduct.

Persistence Among Patients with Crohn Disease Previously Treated with an Anti-tumor Necrosis Factor Inhibitor and Switching or Cycling to Another Biologic Agent.

PURPOSE: Nonresponse to an anti-tumor necrosis factor (TNF) agent in patients with Crohn disease (CD) is often managed by either a switch to a different class of biologic (ie, ustekinumab, vedolizumab) or by cycling to another anti-TNF agent (ie, adalimumab, infliximab, certolizumab pegol). Persistence after a switch to a different biologic class or after cycling within the anti-TNF class was assessed in patients with nonresponse to an anti-TNF agent. METHODS: Adults with CD who discontinued from an anti-TNF agent and either switched to a different class of biologic (ie, anti-interleukin/integrin; the switching cohort) or cycled within the anti-TNF class (the cycling cohort) between September 23, 2016, and August 1, 2019, were selected from a commercial database. The index date was defined as the date of the first claim of the subsequent-line biologic (index biologic) after an anti-TNF. The switching and cycling cohorts were balanced with regard to baseline characteristics, using inverse probability of treatment weights-average treatment effect (IPTW-ATE). Persistence with the index biologic was defined as consistent use with no gaps of >120 days (ustekinumab, vedolizumab, infliximab) or of >60 days (adalimumab, certolizumab pegol) in biologic supply. Composite end points were persistence while being corticosteroid-free (defined as no use of corticosteroids with ≥14 days of supply after day 90 post-index) and persistence while on monotherapy (no immunomodulators/nonindex biologics). Weighted Kaplan-Meier and Cox models were used to assess outcomes at 12 months post-index. FINDINGS: There were 444 patients in the weighted switching cohort (mean age, 40.4 years; 56.3% female) and 441 in the weighted cycling cohort (mean age, 39.5 years; 58.4% female). At 12 months post-index, the rate of persistence with the index biologic was 75.7% in the switching cohort compared to 67.5% in the cycling cohort (log-rank P = 0.023); the rate of persistence while on monotherapy was 58.2% compared to 44.2%, respectively (log-rank P < 0.001). The rate of persistence was 44% greater in the switching compared to that in the cycling cohort (hazard ratio [HR] = 1.44; 95% CI, 1.11-1.88; P = 0.007); the rate of persistence while on monotherapy was 56% greater in the switching cohort (HR = 1.56; 95% CI, 1.28-1.90; P < 0.001). The between-cohort difference in persistence while being corticosteroid-free was not statistically significant (HR = 1.08; 95% CI, 0.89-1.32; P = 0.426). IMPLICATIONS: Patients with CD who switched to a different biologic class were more persistent than were patients who cycled to another anti-TNF agent. These findings may be useful for physicians when considering the treatment of patients who have experienced nonresponse or loss of response to the first-line anti-TNF agent.

Ensemble learning with speaker embeddings in multiple speech task stimuli for depression detection.

Introduction: As a biomarker of depression, speech signal has attracted the interest of many researchers due to its characteristics of easy collection and non-invasive. However, subjects' speech variation under different scenes and emotional stimuli, the insufficient amount of depression speech data for deep learning, and the variable length of speech frame-level features have an impact on the recognition performance. Methods: The above problems, this study proposes a multi-task ensemble learning method based on speaker embeddings for depression classification. First, we extract the Mel Frequency Cepstral Coefficients (MFCC), the Perceptual Linear Predictive Coefficients (PLP), and the Filter Bank (FBANK) from the out-domain dataset (CN-Celeb) and train the Resnet x-vector extractor, Time delay neural network (TDNN) x-vector extractor, and i-vector extractor. Then, we extract the corresponding speaker embeddings of fixed length from the depression speech database of the Gansu Provincial Key Laboratory of Wearable Computing. Support Vector Machine (SVM) and Random Forest (RF) are used to obtain the classification results of speaker embeddings in nine speech tasks. To make full use of the information of speech tasks with different scenes and emotions, we aggregate the classification results of nine tasks into new features and then obtain the final classification results by using Multilayer Perceptron (MLP). In order to take advantage of the complementary effects of different features, Resnet x-vectors based on different acoustic features are fused in the ensemble learning method. Results: Experimental results demonstrate that (1) MFCC-based Resnet x-vectors perform best among the nine speaker embeddings for depression detection; (2) interview speech is better than picture descriptions speech, and neutral stimulus is the best among the three emotional valences in the depression recognition task; (3) our multi-task ensemble learning method with MFCC-based Resnet x-vectors can effectively identify depressed patients; (4) in all cases, the combination of MFCC-based Resnet x-vectors and PLP-based Resnet x-vectors in our ensemble learning method achieves the best results, outperforming other literature studies using the depression speech database. Discussion: Our multi-task ensemble learning method with MFCC-based Resnet x-vectors can fuse the depression related information of different stimuli effectively, which provides a new approach for depression detection. The limitation of this method is that speaker embeddings extractors were pre-trained on the out-domain dataset. We will consider using the augmented in-domain dataset for pre-training to improve the depression recognition performance further.

Cerebral small-vessel disease at high altitude: A comparison of patients from plateau and plain.

Background and purpose: Cerebral small-vessel disease (CSVD) is prevalent worldwide and one of the major causes of stroke and dementia. For patients with CSVD at high altitude, a special environmental status, limited information is known about their clinical phenotype and specific neuroimaging change. We investigated the clinical and neuroimaging features of patients residing at high altitude by comparing with those in the plain, trying to explore the impact of high altitude environment on CSVD. Methods: Two cohorts of CSVD patients from the Tibet Autonomous Region and Beijing were recruited retrospectively. In addition to the collection of clinical diagnoses, demographic information and traditional vascular risk factors, the presence, location, and severity of lacunes and white matter hyperintensities were assessed by manual counting and using age-related white matter changes (ARWMC) rating scale. Differences between the two groups and influence of long-term residing in the plateau were analyzed. Results: A total of 169 patients in Tibet (high altitude) and 310 patients in Beijing (low altitude) were enrolled. Fewer patients in high altitude group were found with acute cerebrovascular events and concomitant traditional vascular risk factors. The median (quartiles) ARWMC score was 10 (4, 15) in high altitude group and 6 (3, 12) in low altitude group. Less lacunes were detected in high altitude group [0 (0, 4)] than in low altitude group [2 (0, 5)]. In both groups, most lesions located in the subcortical (especially frontal) and basal ganglia regions. Logistic regressions showed that age, hypertension, family history of stroke, and plateau resident were independently associated with severe white matter hyperintensities, while plateau resident was negatively correlated with lacunes. Conclusion: Patients of CSVD residing at high altitude showed more severe WMH but less acute cerebrovascular events and lacunes in neuroimaging, comparing to patients residing at low altitude. Our findings suggest potential biphasic effect of high altitude on the occurrence and progression of CSVD.

Treatment persistence among bio-naïve patients with Crohn's disease initiated on ustekinumab or adalimumab.

OBJECTIVES: To compare persistence and describe dose titration among bio-naïve patients with Crohn's disease (CD) initiated on ustekinumab or adalimumab. METHODS: Bio-naïve adults with CD who initiated ustekinumab or adalimumab (index date) from 23 September 2016 (ustekinumab US approval for CD) to 1 August 2019 were selected from IQVIA PharMetrics Plus. Cohorts were balanced on baseline characteristics measured over 12 months pre-index using inverse probability of treatment weights. Persistence was defined as no gaps (ustekinumab: >120 days; adalimumab: >60 days) between days of supply. Dose escalation was defined as ≥2 consecutive sub-cutaneous claims 100% above the US label daily dose in the maintenance phase; de-escalation was a return to the daily dose for ≥2 consecutive claims. Outcomes were described using weighted Kaplan-Meier models; persistence outcomes were compared using Cox's proportional hazards models. RESULTS: At 12 months post-index, patients in the ustekinumab (n = 948) versus adalimumab (n = 4143) cohort had a significantly higher rate of persistence on index biologic (hazard ratio [HR] 1.50; 95% confidence interval [CI]: 1.29-1.74). A total of 830 (87.6%) patients in the ustekinumab cohort and 3713 (89.6%) in the adalimumab cohort began the maintenance phase; within 12 months, 11.2% and 16.9%, underwent a dose escalation, and 26.6% and 6.3%, respectively, subsequently de-escalated to the per US label daily exposure. CONCLUSIONS: Bio-naïve patients with CD initiated on ustekinumab were more persistent than patients initiated on adalimumab; moreover, these patients had numerically lower dose escalation and higher de-escalation rates than patients initiated on adalimumab. Findings support the use of ustekinumab as a first-line treatment for bio-naïve patients with CD.

Clinical outcomes of laparoscopic-assisted natural orifice specimen extraction colectomy using a Cai tube for left-sided colon cancer: a prospective randomized trial.

BACKGROUND: The role of laparoscopic-assisted natural orifice specimen extraction (LA-NOSE) colectomy in the treatment of left-sided colon cancer has not been well defined, and there remains confusion about how to conveniently exteriorize specimens through natural orifices. Therefore, we introduced a homemade invention, the Cai tube, to facilitate the extraction of specimens and compared the clinical outcomes of LA-NOSE with conventional laparoscopic (CL) colectomy for left-sided colon cancer. METHODS: From March 2015 to August 2017, patients with left-sided colon cancer were randomly divided into LA-NOSE and CL groups. Specimens were extracted through the anus with the help of a Cai tube (Patent Number: ZL201410168748.2) in the LA-NOSE group. The primary outcome measure was postoperative pain. Secondary outcomes were the duration of operation, postoperative recovery, surgical morbidity, pathological quality of the specimen, and long-term outcomes, including 3-year overall survival, disease-free survival, local recurrence, and overall recurrence. RESULTS: A total of 60 patients (30 per group) were recruited for this study. None of the patients required emergency conversion to conventional laparoscopic or open surgery during the operation. The postoperative maximum pain score was significantly lower in the LA-NOSE group (mean 2.5 vs. 5.1, P = 0.001), as was the additional analgesia requirement (mean 2/30 vs. 10/30, P = 0.021). Patients in the LA-NOSE group experienced a shorter first time to passage of flatus (mean 2.2 vs. 3.1 days, P = 0.026). All patients could control their defecation at 6 months after surgery. The comparison between the two groups showed no significant differences in the operative time, bleeding volume, postoperative hospital stay, surgical morbidity rates, number of lymph nodes harvested, or resection margin status. The mean follow-up was 48 months (range 7-59) and was similar in both groups. The results showed no differences in long-term outcomes between the two groups. CONCLUSION: In the treatment of left-sided colon cancer, compared with conventional laparoscopic colectomy, LA-NOSE colectomy using the Cai tube exhibited lower postoperative pain, shorter recovery of gastrointestinal function, and similar long-term outcomes. REGISTRATION NUMBER: ChiCTR-OOR-15007060 ( http://www.chictr.org.cn/ ).

Dual-Stream Multiple Instance Learning for Depression Detection With Facial Expression Videos.

Depression is a common mental illness which has brought great harm to the individuals. With recent evidence that many objective physiological signals are associated with depression, automated detection of depression is urgent and important for the growing concern of mental illness. We investigate the problem of classifying depression by facial expressions, which may aid in online diagnosis and rehabilitation engineering of depression. In this work, We propose a weakly supervised learning approach employing multiple instance learning (MIL) on 150 videos data from 75 depressed and 75 healthy subjects. In addition, we present a novel MIL dual-stream aggregator that considers both the instance-level and the bag-level in order to emphasize the information with symptoms. Specifically, our method named ADDMIL uses max-pooling at the instance level to capture symptom information and further integrates the contribution of each instance at the bag level using attention weights. Our method achieves 74.7% accuracy and 74.5% recall on the collected dataset, which not only improves 10.1% accuracy and 9.8% recall over the baseline but also exceeds the best accuracy result of MIL-based method by 2.1%. Our work achieves results that are comparable to the state-of-the-art methods and demonstrates that multiple instance learning has great potential for depression classification. We present for the first time a weakly supervised learning approach in the detection of depression through raw facial expressions, which may provide a new framework for other psychiatric disorders detection methods.

The Real-World Effectiveness and Safety of Ustekinumab in the Treatment of Crohn's Disease: Results From the SUCCESS Consortium.

INTRODUCTION: We evaluated the real-world effectiveness and safety of ustekinumab (UST) in patients with Crohn's disease (CD). METHODS: This study used a retrospective, multicenter, multinational consortium of UST-treated CD patients. Data included patient demographics, disease phenotype, disease activity, treatment history, and concomitant medications. Cumulative rates of clinical, steroid-free, endoscopic, and radiographic remissions were assessed using time-to-event analysis, and clinical predictors were assessed by using multivariate Cox proportional hazard analyses. Serious infections and adverse events were defined as those requiring hospitalization or treatment discontinuation. RESULTS: A total of 1,113 patients (51.8% female, 90% prior antitumor necrosis factor exposure) were included, with a median follow-up of 386 days. Cumulative rates of clinical, steroid-free, endoscopic, and radiographic remissions at 12 months were 40%, 32%, 39%, and 30%, respectively. Biologic-naive patients achieved significantly higher rates of clinical and endoscopic remissions at 63% and 55%, respectively. On multivariable analyses, prior antitumor necrosis factor (hazard ratio, 0.72; 95% confidence interval, 0.49-0.99) and vedolizumab exposure (hazard ratio, 0.65; 95% confidence interval, 0.48-0.88) were independently associated with lower likelihoods of achieving endoscopic remission. In patients who experienced loss of remission, 77 of 102 (75%) underwent dose optimization, and 44 of 77 (57%) achieved clinical response. An additional 152 of 681 patients (22.3%) were dose-optimized because of primary nonresponse incomplete response to UST, of whom 40.1% (61 of 152) responded. Serious infections occurred in 3.4% of patients while other noninfectious adverse events (lymphoma [n = 1], arthralgia [n = 6], rash [n = 6], headache [n = 3], hepatitis [n = 3], hair loss [n = 3], neuropathy [n = 1], and vasculitis [n = 1]) occurred in 2.4% of patients. DISCUSSION: UST represents a safe and effective treatment option for CD, with 40% of patients from a highly refractory cohort achieving clinical remission by 12 months. The greatest treatment effect of UST was seen in biologic-naive patients, and dose escalation may recapture clinical response.

Racial or ethnic differences on treatment adherence and persistence among patients with inflammatory bowel diseases initiated with biologic therapies.

BACKGROUND: Inflammatory bowel disease (IBD) is a chronic disease with the potential for significant morbidity in case of suboptimal treatment (e.g. low treatment adherence). In spite of immense research in IBD, literature on association of IBD with race/ethnicity is fragmented. In this study, we aimed to evaluate the association between race/ethnicity and treatment adherence and persistence among patients with Crohn's disease (CD) or ulcerative colitis (UC) initiated with biologic therapies. METHODS: This observational, retrospective study utilized the Optum Clinformatics (Optum) Extended Data Mart Socioeconomic Status (SES) database. Adult patients with ≥ 2 medical claims for CD or UC diagnosis, ≥ 1 medical or pharmacy claim for corresponding FDA-approved biologic therapy, and a ≥ 12-month pre-index (index date: date of the first biologic medical/pharmacy claim) continuous health plan enrollment were included. Treatment adherence was measured as the proportion of days covered of ≥ 80% and treatment persistence by the number of days from the index date to the biologics discontinuation date. Switching among biologics was allowed for both treatment adherence and treatment persistence. Multivariable regression analyses were performed to evaluate the association between race/ethnicity and treatment adherence/persistence. RESULTS: Among patients with CD (N = 1430) and UC (N = 1059) included, majority were White (CD: 80.3%, UC: 78.3%), followed by African Americans (AA; CD: 10.5%, UC: 9.7%). Among patients with CD, AA were significantly less likely to adhere to biologics (adjusted OR [95%CI]: 0.61 [0.38; 0.99]) and more likely to discontinue biologics earlier (adjusted HR [95%CI]: 1.52 [1.16; 2.0]) during the follow-up period compared to Whites, after adjusting for other patient sociodemographic and clinical characteristics. Among patients with UC, no significant differences in the treatment adherence/persistence were observed between different races/ethnicities. CONCLUSIONS: Patients with CD were found to display racial differences in the treatment adherence and persistence of biologics, with significantly lower adherence and earlier discontinuation in AA compared to Whites. Such differences were not observed in patients with UC. Future studies are warranted to understand the possible reasons for racial differences, particularly in patients with CD.