A "Prediction - Detection - Judgment" framework for sudden water contamination event detection with online monitoring

The contamination detection technology helps in water quality management and protection in surface water. It is important to detect sudden contamination events timely from dynamic variations due to various interference factors in online water quality monitoring data. In this study, a framework named "Prediction - Detection - Judgment" is proposed with a method framework of "Time series increment - Hierarchical clustering - Bayes' theorem model". Time to detection is used as an evaluation index of contamination detection methods, along with the probability of detection and false alarm rate. The proposed method is tested with available public data and further applied in a monitoring site of a river. Results showed that the method could detect the contamination events with a 100% probability of detection, a 17% false alarm rate and a time to detection close to 4 monitoring intervals. The proposed index time to detection evaluates the timeliness of the method, and timely detection ensures that contamination events can be responded to and dealt with in time. The site application also demonstrates the feasibility and practicability of the framework proposed in this study and its potential for extensive implementation.

A machine-learning model based on dynamic contrast-enhanced MRI for preoperative differentiation between hepatocellular carcinoma and combined hepatocellular-cholangiocarcinoma

AIM

To establish a machine-learning model based on dynamic contrast-enhanced (DCE) magnetic resonance imaging (MRI) to differentiate combined hepatocellular-cholangiocarcinoma (cHCC-CC) from hepatocellular carcinoma (HCC) before surgery.

MATERIALS AND METHODS

Clinical and MRI data of 194 patients with histopathologically diagnosed cHCC-CC (n=52) or HCC (n=142) were analysed retrospectively. ITK-SNAP software was used to delineate three-dimensional (3D) lesions and extract high-throughput features. Feature selection was carried out based on Pearson's correlation coefficient and least absolute shrinkage and selection operator (LASSO) regression analysis. A radiomics model (radiomics features), a clinical model (i.e., clinical-image features), and a fusion model (i.e., radiomics features + clinical-image features) were established using six machine-learning classifiers. The performance of each model in distinguishing between cHCC-CC and HCC was evaluated with the receiver operating characteristic (ROC) curve, the area under the ROC curve (AUC), sensitivity, and specificity.

RESULTS

Significant differences in liver cirrhosis, tumour number, shape, edge, peritumoural enhancement in the arterial phase, and lipid were identified between cHCC-CC and HCC patients (p<0.05). The AUC of the fusion model based on logistic regression was 0.878 (95% CI: 0.766-0.949) in the arterial phase in the test set, and the sensitivity/specificity was 0.844/0.714; however, the AUC of the clinical and radiomics models was 0.759 (95% CI: 0.663-0.861) and 0.838 (95% CI: 0.719-0.921) in the test set, respectively.

CONCLUSION

The fusion model based on DCE-MRI in the arterial phase can significantly improve the diagnostic rate of cHCC-CC and HCC as compared with conventional approaches.

Enhancing interpretability and generalizability of deep learning-based emulator in three-dimensional lake hydrodynamics using Koopman operator and transfer learning: Demonstrated on the example of lake Zurich

Three-dimensional lake hydrodynamic model is a powerful tool widely used to assess hydrological condition changes of lake. However, its computational cost becomes problematic when forecasting the state of large lakes or using high-resolution simulation in small-to-medium size lakes. One possible solution is to employ a data-driven emulator, such as a deep learning (DL) based emulator, to replace the original model for fast computing. However, existing DL-based emulators are often black-box and data-dependent models, causing poor interpretability and generalizability in practical applications. In this study, a data-driven emulator is established using deep neural network (DNN) to replace the original model for fast computing of three-dimensional lake hydrodynamics. Then, the Koopman operator and transfer learning (TL) are employed to enhance the interpretability and generalizability of the emulator. Finally, the generalizability of DL-based emulators is comprehensively analyzed through linear regression and correlation analysis. These methods are tested against an existing hydrodynamic model of Lake Zurich (Switzerland) whose data was provided by an open-source web-based platform called Meteolakes/Alplakes. According to the results, (1) The DLEDMD offers better interpretability than DNN because its Koopman operator reveals the linear structure behind the hydrodynamics; (2) The generalization of the DL-based emulators in three-dimensional lake hydrodynamics are influenced by the similarity between the training and testing data; (3) TL effectively improves the generalizability of the DL-based emulators.

Improving the interpretability of deep reinforcement learning in urban drainage system operation

Deep reinforcement learning (DRL) has been increasingly used as an adaptive and efficient solution for real-time control (RTC) of the urban drainage system (UDS). Despite the promising potential of DRL, it is a black-box model whose control logic and control consequences are difficult to be understood and evaluated. This leads to issues of interpretability and poses risks in practical applications. This study develops an evaluation framework to analyze and improve the interpretability of DRL-based UDS operation. The framework includes three analysis methods: Sobol sensitivity analysis, tree-based surrogate modelling, and conditional probability analysis. It is validated using two different DRL approaches, i.e., deep Q-learning network (DQN) and proximal policy optimization (PPO), which are trained to reduce combined sewer overflow (CSO) discharges and flooding in a real-world UDS. According to the results, the two DRLs have been shown to perform better than a rule-based control system that is currently being used. Sobol sensitivity analysis indicates that DQN is particularly sensitive to the flow of links and rainfall, while PPO is sensitive to all the states. Tree-based surrogate models effectively reveal the control logic behind the DRLs and indicate that PPO is more comprehensible but DQN is more forward-looking. Conditional probability analysis demonstrates the potential control consequences of the DRLs and identifies three situations where the DRLs are ineffective: a) the storage of UDS is fully utilized; b) peak flows have already passed through actuators; c) a substantial amount of water enters one location simultaneously. The proposed evaluation framework enhances the interpretability of DRL in UDS operations, fostering trust and confidence from operators, stakeholders, and regulators.

Gpnmb silencing protects against hyperoxia-induced acute lung injury by inhibition of mitochondrial-mediated apoptosis

Background: Hyperoxia-induced acute lung injury (HALI) is a complication to ventilation in patients with respiratory failure, which can lead to acute inflammatory lung injury and chronic lung disease. The aim of this study was to integrate bioinformatics analysis to identify key genes associated with HALI and validate their role in H2O2-induced cell injury model.Methods: Integrated bioinformatics analysis was performed to screen vital genes involved in hyperoxia-induced lung injury (HLI). CCK-8 and flow cytometry assays were performed to assess cell viability and apoptosis. Western blotting was performed to assess protein expression.Results: In this study, glycoprotein non-metastatic melanoma protein B (Gpnmb) was identified as a key gene in HLI by integrated bioinformatics analysis of 4 Gene Expression Omnibus (GEO) datasets (GSE97804, GSE51039, GSE76301 and GSE87350). Knockdown of Gpnmb increased cell viability and decreased apoptosis in H2O2-treated MLE-12 cells, suggesting that Gpnmb was a proapoptotic gene during HALI. Western blotting results showed that knockdown of Gpnmb reduced the expression of Bcl-2 associated X (BAX) and cleaved-caspase 3, and increased the expression of Bcl-2 in H2O2 treated MLE-12 cells. Furthermore, Gpnmb knockdown could significantly reduce reactive oxygen species (ROS) generation and improve the mitochondrial membrane potential.Conclusion: The present study showed that knockdown of Gpnmb may protect against HLI by repressing mitochondrial-mediated apoptosis.

A dataset of low-carbon energy transition index for Chinese cities 2003-2019

Cities are at the heart of climate change mitigation as they account for over 70% of global carbon emissions. However, cities vary in their energy systems and socioeconomic capacities to transition to renewable energy. To address this heterogeneity, this study proposes an Energy Transition Index (ETI) specifically designed for cities, and applies it to track the progress of energy transition in Chinese cities. The city-level ETI framework is based on the national ETI developed by the World Economic Forum (WEF) and comprises two sub-indexes: the Energy System Performance sub-index, which evaluates the current status of cities' energy systems in terms of energy transition, and the Transition Readiness sub-index, which assesses their socioeconomic capacity for future energy transition. The initial version of the dataset includes ETI and its sub-indexes for 282 Chinese cities from 2003 to 2019, with annual updates planned. The spatiotemporal data provided by the dataset facilitates research into the energy transition roadmap for different cities, which can help China achieve its energy transition goals.

Pneumonitis After Concurrent Chemoradiation and Immune Checkpoint Inhibition in Patients with Locally Advanced Non-small Cell Lung Cancer

AIMS

Pneumonitis is a common and potentially deadly complication of combined chemoradiation and immune checkpoint inhibition (CRT-ICI) in patients with locally advanced non-small cell lung cancer (LA-NSCLC). In this study we sought to identify the risk factors for pneumonitis with CRT-ICI therapy in LA-NSCLC cases and determine its impact on survival.

MATERIALS AND METHODS

We conducted a retrospective chart review of 140 patients with LA-NSCLC who underwent curative-intent CRT-ICI with durvalumab between 2018 and 2021. Pneumonitis was diagnosed by a multidisciplinary team of clinical experts. We used multivariable cause-specific hazard models to identify risk factors associated with grade ≥2 pneumonitis. We constructed multivariable Cox proportional hazard models to investigate the impact of pneumonitis on all-cause mortality.

RESULTS

The median age of the cohort was 67 years; most patients were current or former smokers (86%). The cumulative incidence of grade ≥2 pneumonitis was 23%. Among survivors, 25/28 patients had persistent parenchymal scarring. In multivariable analyses, the mean lung dose (hazard ratio 1.14 per Gy, 95% confidence interval 1.03-1.25) and interstitial lung disease (hazard ratio 3.8, 95% confidence interval 1.3-11.0) increased the risk for pneumonitis. In adjusted models, grade ≥2 pneumonitis (hazard ratio 2.5, 95% confidence interval 1.0-6.2, P = 0.049) and high-grade (≥3) pneumonitis (hazard ratio 8.3, 95% confidence interval 3.0-23.0, P < 0.001) were associated with higher all-cause mortality.

CONCLUSIONS

Risk factors for pneumonitis in LA-NSCLC patients undergoing CRT-ICI include the mean radiation dose to the lung and pre-treatment interstitial lung disease. Although most cases are not fatal, pneumonitis in this setting is associated with markedly increased mortality.

Protocol in a Day: An Educational Institutional Workshop for Protocol Development

PURPOSE/OBJECTIVE(S)

The Protocol-in-a-Day (PIAD) workshop was developed to support junior faculty and residents with clinical trial protocol design, with the main goal of providing initial feedback during development to reduce time for review and approval from institutional oversight committees. Our objectives are to mentor and educate participants and to evaluate the time to institutional approval by oversight committees.

MATERIALS/METHODS

PIAD provided concurrent educational feedback on 6 key elements of trial design. These included: (1) regulatory aspects; (2) institutional scientific review committee (SRC) and institutional review board (IRB); (3) clinical research and data coordination (including nursing); (4) statistics; (5) correlatives including imaging, biospecimens, and health services research/patient-reported outcomes; and (6) operations. The average number of days from submission to IRB approval or study activation for PIAD protocols was compared to other protocols submitted between January 2018 - January 2022 within the Division of Radiation Oncology. Participants were also given a 15-question survey to assess their perspective of the impact of the workshop.

RESULTS

A total of 25 protocols went through the PIAD workshop between January 2018-January 2022. Of the 25 protocols, 7 (28%) were excluded from this study due to not being submitted possibly after participants benefited from education on the limitations of their design. Eighteen protocols were included in our final analyses. These protocols included phase II (n = 11), phase 1 (n = 5), and phase III (n = 2). At the time of this report, all protocols (n = 18) have received IRB approval and have been activated. Protocol elements that could impact study activation included protocols requiring investigational new drug (IND) approval (n = 8) and multicenter studies (n = 1). Analyzing the time of submission to request for activation showed a decrease in time for protocols that went through PIAD vs those that did not [PIAD protocols, 254 days vs All other protocols, 262 days]. Likewise, those who attended PIAD had a lower average time from submission to IRB Approval [ PIAD protocols, 40 days vs All other protocols, 59 days]. All participants (100%) of the PIAD workshop responded that the educational program "improved the overall quality of the study design." The most commonly cited changes were protocol language (n = 17), statistics (n = 15), consent language (n = 8), and study design (n = 8). Aspects participants identified as the most educational included mentorship from regulatory, clinical research finance, and IRB review.

CONCLUSION

PIAD from participant surveys provided high educational value in the areas of improving trial quality, language and statistical design. When analyzing the average time, from 'submission to IRB initial approval' and 'submission to activation', PIAD protocols had a shorter time for approval, and thus suggests PIAD is effective in improving the overall design of protocols.

Definitive Radiotherapy for Oligometastatic and Oligoprogressive Thyroid Cancer

PURPOSE/OBJECTIVE(S)

Local consolidative radiotherapy (LCT) for oligometastatic disease is a promising paradigm improving outcomes for various malignancies but has been underexplored for metastatic thyroid cancer. We hypothesize that LCT to distant sites with definitive RT doses can yield favorable outcomes and defer systemic therapy escalation for these patients.

MATERIALS/METHODS

We reviewed 96 thyroid cancer patients who received 175 LCT courses from 2010-2022 to 228 metastatic sites, including: thorax (45%), bone (40%), brain (6%), head/neck (5%), and abdomen (3%). Common prescriptions were 50-55Gy/4-5fxs or 56-70Gy/8-10fxs for lung; 52.5-60Gy/15fxs for mediastinum; and 18-24Gy/1fx or 27-30Gy/3fxs for bone. RECIST v1.1 and CTCAE v5.0 were used to define progression and toxicities, respectively. Outcomes were evaluated via Kaplan-Meier and associations examined via Cox proportional hazards modeling.

RESULTS

Median age was 63 years (range: 26-92), with 62 oligometastatic cases (total 1-5 sites) and 34 oligoprogressive (with 1-5 growing sites). Primary disease was controlled in all patients, with 39% receiving post-op RT and 66% prior RAI. Histologies included papillary (40%), anaplastic (25%), follicular (12%), medullary (9%), Hurthle (7%), and poorly-differentiated (7%). Median time from initial diagnosis to LCT was 3 yrs (IQR 1-8), and median follow-up from 1st LCT was 21 mos (IQR 9-51). Patients received an average 2 LCT courses (range 1-8) treating 1-4 sites. Median survival (OS) from 1st LCT was 9 yrs (95% CI = 5-14). On multivariable analysis (MVA), worse OS was associated with anaplastic histology (HR 4.6, p<.01), but longer OS was associated with prior RAI (HR 0.33, p = .02) and oligometastatic disease (HR 0.3, p = .01). For anaplastic histology, median OS was 1.2 years vs. 9.3 years for non-anaplastic; 3-yr OS was 36% vs. 88% (log-rank, p<.01). Five-year OS for oligometastatic cases was 75% vs 53% for oligoprogressive (log-rank, p = .04). Median progression free survival (PFS) from 1st LCT was 15.5 mos (95% C I = 11-20). On MVA for all LCT courses, time to any progression (TTP) was negatively associated with anaplastic histology (HR 1.7, p = .02) and 2nd or higher LCT course (HR 1.45, p = .05), but favorably associated with thoracic site (HR 0.49, p<.01). Following later LCT courses, median TTP was 11 mos vs 17 mos for initial LCT course (log-rank, p = .03). After LCT to lung/chest, TTP was 18.6 mos vs 9.5 mos for non-thoracic sites (log-rank, p<.01). Only 6% of failures occurred at previously treated lesions. Most LCT courses (67%) were without ongoing chemotherapy, while 25% entailed continuing the same regimen and 9% had planned treatment post-RT. There were 2 Grade 3 toxicities (pneumonitis and esophagitis) and no Grade 4-5 events.

CONCLUSION

With high local control rates and minimal toxicity, LCT can be a feasible strategy to defer systemic therapy escalation for oligometastatic and oligoprogressive thyroid cancer.

Long-Term Outcomes and Toxicity in Esophageal Cancer Patients after Neoadjuvant or Definitive Concurrent Chemotherapy with Proton Beam Therapy

PURPOSE/OBJECTIVE(S)

Proton-beam therapy (PT) is increasingly utilized over three dimensional-conformal radiation therapy (3D-CRT) and intensity modulated radiation therapy (IMRT) photon irradiation for the treatment of various malignancies due to better toxicity reduction. We investigated the long-term outcomes and toxicity in esophageal cancer (EC) patients treated with PT as part of their neoadjuvant concurrent chemoradiation followed by surgery (nCRT) or definitive concurrent chemoradiation (dCRT) treatment regimen.

MATERIALS/METHODS

All consecutively treated, American Joint Committee on Cancer 7th edition clinical stage I-IV EC patients from 2006 to 2022 were retrospectively analyzed. Standard RT dose for most patients was 50.4 Gy/28 fractions. nCRT patients had surgery within 4 months post-RT. Kaplan-Meier method was used to determine overall survival (OS), locoregional recurrence-free survival (LRRFS) and distant metastatic-free survival (DMFS). Acute and chronic RT-related toxicities were graded with Common Terminology Criteria for Adverse Events version 4.0.

RESULTS

There were 510 EC PT patients: 204 (40%) had nCRT and 306 (60%) had dCRT. Most lesions were located in the lower esophagus, of adenocarcinoma histology and treated with passive scatter PT. Overall median follow-up was 72 months. Median, 3- and 5-year OS for all patients were 43 months, 54.1% and 44.9%, respectively. Median LRRFS and DMFS were not reached. Esophagitis was the most common grade ≥3 (G3+) toxicity (59 patients; 28.9%, including a G4 and a G5 toxicity), followed by nausea (29 patients; 14.2%) and esophageal stricture (26 patients, 12.7%). With nCRT, median, 3- and 5-year OS were 80 months, 64.7% and 56.1%, respectively, while the median LRRFS and DMFS were not reached again. Their most common G3+ toxicity was esophagitis in 14 patients (6.9%) followed by nausea (8 patients; 3.9%). An nCRT patient developed G4 RT pneumonitis. Pathological complete response (pCR) was observed in 58 patients (28.4%). Surgery-related pulmonary, cardiac and gastrointestinal complications were reported in 38 (18.6%), 40 (19.6%) and 43 (21.1%) patients, respectively. dCRT patients had a median follow-up of 65 months, and median, 3- and 5-year OS of 32 months, 46.7% and 37.0%, respectively. Although the median LRRFS was not reached, the median DMFS was 74 months. The most observed dCRT G3+ toxicity was esophagitis (45 patients, 22.1%: including both G4 and G5 patients) and then esophageal stricture (23 patients, 11.3%). A dCRT patient developed G4 fistula.

CONCLUSION

To our knowledge, this is the largest single-institutional study on EC long-term outcomes and toxicity using PT. Our cohort reveals good outcomes and mostly mild CRT-related toxicities. Trimodality nCRT with protons demonstrates excellent outcomes relative to the CROSS trial (49.4 months) with identical pCR rate (29% in CROSS) and similar toxicity profile. nCRT with protons should be studied rigorously in the current randomized phase III trial NRG GI006.

Comorbidities and Their Impact on Treatment Tolerance and Outcome in Elderly NSCLC Patients Treated with Concurrent Chemoradiation Using Proton or Photon Followed by Immunotherapy

PURPOSE/OBJECTIVE(S)

To investigate comorbidities and their impact on treatment tolerance and survival for elderly NSCLC patients treated with concurrent chemoradiation using proton beam therapy (PBT) or intensity modulated radiation therapy (IMRT) followed by immunotherapy (IO).

MATERIALS/METHODS

This is a retrospective study in senior patients (≥ 65 years) with inoperable locally advanced- NSCLC (LA-NSCLC) who received concurrent chemoradiation (CRT) therapy followed by adjuvant IO. Comorbidities are listed in table 1 and scored according to hazard ratios (HRs) of death. Treatment tolerance and comorbidities were compared between modalities using Fisher's exact test. Association between comorbidities and treatment tolerance was analyzed using logistic regression; association between comorbidities and survival was analyzed by Cox proportional hazards regression.

RESULTS

Total 218 patients were included with median age of 72 (range 65-86) years. Fifty-three (24.3%) were PBT and 165 (75.7%) were IMRT. Majority patients were white ethnics, had stage III adenocarcinoma disease and good performance score, and received RT dose 60-66 Gy. The median follow-up time for the whole group was 23 months. A 97% of patients had ≥1, and 62% had ≥4 comorbidities. The combined comorbidity scores ranged from 0 to 25. No significant difference in comorbidities between modalities except IMRT had more patients with COPD (36% vs. 51%, p = 0.047). Rates of hospitalization and ER visit during CRT+IO were 28% and 5%, respectively. Rates of IO discontinuation and interruption were 55% and 8%, respectively. Patients received >6 months IO had significantly lower risk of death (HR: 0.25, p<0.001). No significant difference in treatment tolerance between modalities. In multivariable analysis, atrial fibrillation, pulmonary disease, and depression were the comorbidities associated with hospitalization during CRT+IO (p<0.05); dementia was the comorbidity associated with IO interruption (p = 0.042); heart failure, COPD, asthma, osteoporosis and HIV were the comorbidities associated with poor OS (p<0.05). Combined comorbidity score was associated with OS with HR of 1.13 (p<0.001) with adjustment of race, GTV and induction chemotherapy.

CONCLUSION

Comorbidities were associated with hospitalization and IO interruption during the whole course of CRT and IO therapy, with increased IO discontinuation which negatively impacted survival. Evaluating and scoring comorbidities at diagnosis could be a useful method to predict risk of death before treatment start.

Comparing 90-Day Post-Operative Mortality after Neoadjuvant Proton-Based vs. Photon-Based Chemoradiotherapy for Esophageal Cancer

PURPOSE/OBJECTIVE(S)

Standard of treatment for locally advanced esophageal cancer consists of chemoradiotherapy (CRT) followed by surgery. Evidence suggests that proton beam therapy (PBT) results in lower toxicity and fewer post-operative complications compared to photon-based radiotherapy (RT). Mortality in the first 90 days after surgery is a rare event occurring in 2-8% of patients, with higher reported rates (of up to 17%) in older patients. This 90-day mortality (90DM) rate is an important measure of post-operative (non-oncologic) mortality as a proxy of quality of care. We hypothesize that PBT could reduce the incidence of 90DM compared to photon-based RT.

MATERIALS/METHODS

From a single-center retrospectively acquired database patients with esophageal cancer treated with neoadjuvant CRT and esophagectomy in 1998-2022 were selected. Univariable logistic regression analyses were used to study the associations of RT modality and other patient- and treatment-related characteristics with 90DM. Subsequently, 3 separate methods were applied to adjust for confounding bias. These included multivariable logistic regression, 1:1 nearest-neighbor propensity score matching (PSM), and inverse probability of treatment weighting (IPTW). Finally, stratified analyses for patient groups aged ≥67 vs. <67 years were performed.

RESULTS

A total of 894 eligible patients were included (PBT, n = 202; photon-based RT, n = 692). PBT patients had a significantly higher age, better performance score, and a higher number of comorbidities. The 90DM rate was 5 (2.5%) in the PBT group and 29 (4.2%) in the photon-based RT group (p = 0.262). Significant univariable predictors of 90DM included higher age and tumor location. After multivariable adjustment, PBT vs. photon therapy was not significantly associated with 90DM (OR 0.49, 95% CI 0.18-1.31). The 90DM rates in the PSM cohort (n = 181 vs. n = 181) were 2.8% for PBT and 3.3% for photon-based RT (p = 0.379). The 90DM rates in the IPTW cohort were 2.8% for PBT and 4.1% for photon-based RT (p = 0.427). In the full cohort, stratified analysis for age groups revealed that in patients aged ≥67 years, PBT was associated with a decreased risk of 90DM compared to photon-based RT (1.3% vs. 8.8%; p = 0.046), which was not the case in patients aged <67 years. In the PSM cohort, a comparable (but non-significant) difference was observed in favor of PBT in patients aged ≥67 years (i.e., 1.5% vs. 7.5%; p = 0.099). Within-group analyses in the original cohort demonstrated that a higher age significantly increased the risk of 90DM within the photon-based RT group (8.8% vs. 2.7% for age ≥67 vs. <67 years; p = 0.001), but not within the PBT group (1.3% vs. 3.2%; p = 0.398).

CONCLUSION

Post-operative 90DM after esophagectomy for cancer was not significantly different between PBT and photon-based neoadjuvant CRT. However, among older patients we observed a signal that PBT may reduce the risk of 90DM. Higher age increased the risk of 90DM in patients who underwent photon-based RT, but not in patients who underwent PBT.

Prospective Trial of Using Imaging to Predict Pathologic Response and Clinical Outcomes in Locally Advanced Esophageal Cancer

PURPOSE/OBJECTIVE(S)

Trimodality therapy with chemoradiation (CRT) followed by esophagectomy is the standard of care for locally advanced esophageal cancer. An unresolved question is whether pathologic complete response (pCR) can be assessed non-invasively for patients post-CRT. In this study, we assessed whether diffusion-weighted imaging (DWI) with MRI or PET can be used as predictors of pCR and other clinical outcomes after CRT.

MATERIALS/METHODS

Patients were enrolled on a single-arm institutional trial (PA13-0380) assessing the role of imaging in predicting outcomes in potentially resectable esophageal patients undergoing trimodality therapy. All patients received neoadjuvant CRT, and 29 patients had subsequent surgery. DWI MRI and PET scans were obtained at baseline, 2 weeks after the start of CRT (interim) and 4 to 6 weeks after completion of CRT (follow up). Apparent diffusion coefficients (ADCs) were calculated based on DWI images. Circulating tumor DNA was obtained for 27 patients post-radiation using CAPP-Seq. Mann-Whitney tests compared imaging changes associated with pCR. Discrimination of pCR by imaging changes was quantified by received operating characteristics. Youden's index was applied to select optimal thresholds. Kaplan-Meier analysis was performed to assess differences in overall survival (OS) and progression-free survival (PFS) by changes in DWI, PET, and ctDNA parameters.

RESULTS

Our cohort of 60 patients had a median follow up of 42.7 months, age of 65.4 yrs, and ECOG of 1 at completion of CRT. 90% were male, 58% had a history of smoking, and 85% were white. 83% had adenocarcinoma with the rest squamous cell carcinoma. Stages of the patients ranged from IIA to IIIB. All had moderately (47%) or poorly (53%) differentiated disease. All received 41.4-50.4 Gy in 1.8 Gy fractions with the majority receiving 50.4 Gy (95%). 29 patients underwent surgery after CRT of which 8 (27.6%) had pCR. Mean ΔADC from baseline to mid-treatment was most associated with pCR (AUC = 0.98, p<0.001) for patients undergoing surgery. Max ΔADC from baseline to first follow-up was most associated with OS (p = 0.002) and PFS (p<0.001) for the whole cohort. 27 patients had ctDNA analyzed after RT with the presence of ctDNA significantly associated with worse OS (HR = 0.12, p = 0.05) and PFS (HR = 0.10, p = 0.002). Combining ctDNA and max ΔADC generated a model that was more predictive of OS and PFS than either alone. We found that neither the PET parameters of TLG or SUV max at baseline or changes in these parameters from baseline to mid-treatment or first follow-up were as predictive as DWI.

CONCLUSION

We show that changes in DWI is associated with pCR, OS, and PFS in resectable esophageal cancer patients undergoing CRT. DWI was more predictive than PET and a model combining DWI and ctDNA was the most predictive of clinical outcomes. This study shows the significant promise of using DWI in potentially guiding treatment decisions in esophageal cancer patients and will require validation in a larger cohort.

Towards coordinated and robust real-time control: a decentralized approach for combined sewer overflow and urban flooding reduction based on multi-agent reinforcement learning

The real-time control (RTC) of urban drainage systems can make full use of the capabilities of existing infrastructures to mitigate combined sewer overflow (CSO) and urban flooding. Despite the benefits of RTC, it may encounter potential risks and failures, which need further consideration to enhance its robustness. Besides failures of hardware components such as sensors and actuators, the RTC performance is also sensitive to communication failures between the devices that are spatially distributed in a catchment-scale system. This paper proposes a decentralized control strategy based on multi-agent reinforcement learning to enhance communication robustness and coordinate the decentralized control agents through centralized training. To investigate different control structures, a centralized and a fully decentralized strategy are also developed based on reinforcement learning (RL) for comparison. A benchmark drainage model and a real-world drainage model are formulated as two cases, and the control agents are trained to control the orifices or pumps for CSO or flooding mitigation in each case. The three RL strategies reduce the CSO volume by 5.62-9.30% compared with a static baseline in historical rainfalls of the benchmark case and reduce the CSO and flooding volume by 14.39-21.36% compared with currently-used rule-based control in synthetic rainfalls of the real-world case. Benefitting from centralized training, the decentralized agents can achieve similar performance to the centralized agent. The decentralized control also enhances the communication robustness with smaller performance loss than the centralized control when observation communication fails, and provides a robust backup at the local level to limit the uncertainties when action commands from the centralized agent are lost. The results and findings indicate that multi-agent RL contributes to a coordinated and robust solution for RTC of urban drainage systems.

Radiation Therapy Decision Making Process and Operations for COVID-19 Positive Patients

Purpose/Objective(s)

A challenging clinical dilemma during the COVID-19 pandemic is management of cancer patients who test positive for COVID. Given the need to balance the risk of disease progression with the risk of transmission to other patients and staff, radiation therapy for these patients requires careful consideration and modification of standard workflows. It is also critical to develop processes to mitigate radiation treatment interruption, which can affect patient outcomes. The objective of this study was to report the clinical operations and outcomes for COVID positive patients receiving radiation therapy during the pandemic at a tertiary cancer center including 2 network locations.

Materials/Methods

During March 2020 to March 2022, the Radiation Oncology COVID committee (RO COVID) developed an integrated process to triage patients, provide treatment recommendations, and implement infection control procedures to safely deliver radiation therapy to COVID positive patients. Policies were created for each center with multidisciplinary input from infectious disease, radiation oncology, radiation therapy, and nursing. All COVID positive patients were presented to the RO COVID group and evaluated for clinical urgency, benefit with radiation, and life expectancy. If deemed necessary, a limited planned break or hypofractionated regimen was recommended to minimize staff exposure. We conducted a retrospective review of COVID positive patients with different primary malignancies treated through the COVID positive pathway.

Results

A total of 68 COVID positive patients were treated with the COVID positive pathway (HN 15, Breast 9, CNS 8, GU 8, GYN, 7, Thoracic 6, GI 5, HEME 5, PED 3, SARC 2). The median age was 57.1 years (IQR 45.8-63.4) and 47% were female. There were 39 patients (57%) who were asymptomatic and were tested for routine pre-radiation screening or due to concerns of COVID exposure. Twenty-three (34%) patients were treated with palliative intent and 8 (12%) were treated for an emergent indication (i.e., spinal cord compression, bleeding). Thirteen (19%) patients were receiving radiation treatment, had a treatment break (7-21 days), and then resumed their radiation course. All treatments were successfully completed without known nosocomial spread of COVID to staff or other patients. Among this heterogenous group of patients, 58 (85%) were alive with a median follow up of 2 months (IQR 0.5-7.5). COVID infection may have contributed to 3 out of 10 deaths (4% of total cohort). The remaining deaths were due to progression of disease or other non-COVID causes.

Conclusion

In this study, COVID positive patients were safely treated with radiation therapy through a comprehensive decision making and clinical operations pathway taking into account evolving COVID guidelines for three different variant surges. Although limited in follow up, patient outcomes are promising with few COVID-related deaths and low overall mortality rates, even with hypofractionated regimens.

Cardiovascular disease risk prediction via machine learning using mental health data

Background

Robust and accurate risk prediction models are much needed in cardiovascular disease. It is well-known that mental health is associated with the risk of developing cardiovascular disease. It is unknown whether mental health markers can enhance existing risk prediction models for cardiovascular disease.

Purpose

The main purpose of this study was to assess capability of mental health factors along with traditional risk factors to be used in cardiovascular predictive machine learning models, and to develop a combined machine learning approach using both traditional risk and psychological factors in 375,145 participants of the UK Biobank.

Methods

A comprehensive Pearson correlation analysis is carried out on UK Biobank data. Subsequently, an ensemble model containing decision tree, random forest, XGBoost, support vector machine (SVM), and deep neural network (DNN) classification approaches was built to predict cardiovascular diseases (CVD) in UK Biobank participants. The model was first trained using traditional cardiovascular risk factors, and subsequently trained using a combination of cardiovascular risk and psychological factors.

Results

The correlation analysis revealed that there is a correlation between CVD and mental health factors suggesting the potential of mental health application for machine learning models. Our ensemble machine learning model was able to predict CVD with an accuracy of 73.49% using CVD risk factors alone. However, by combining psychological factors with CVD risk factors in the training data, an improved accuracy of 95.70% was achieved. The accuracy and robustness of ensemble machine learning model outperformed any of five constituent learning algorithms alone.

Conclusions

Our results suggest that mental health assessment data along with traditional risk factors provides a powerful, safe and affordable machine learning model enrichment that can be used for state-of-the-art prediction of CVD.

Funding Acknowledgement

Type of funding sources: None.

AB1043 IDENTIFICATION OF UNUSUAL TOPHI ANTERIOR TO THE THYROID CARTILAGE USING DIFFERENT IMAGING METHODS

Background

Inadequately treated gout patients could develop tophi. However, tophi could appear as atypical mass, leading to difficulty in differential diagnosis.

Objectives

The objectives of the study were to describe an unusual case with tophi anterior to the Thyroid Cartilage using different imaging methods.

Methods

Clinical information of the case was presented. Ultrasound, dual energy Computed Tomography (DECT) and CT scan were compared to describe the tophi.

Results

A 66-year-old man with a 30-year history of gout was admitted. He had received urate-lowering drugs intermittently over the past few years. Physical examination revealed multiple tophi on the hands, feet, and other auricular surfaces. Moreover, there was a subcutaneous mass above the thyroid cartilage (Figure 1A). Ultrasound revealed hyperechoic nodule anterior to the thyroid cartilage, 11*7cm in size, with uneven internal echo, no blood flow signal inside, and a close relationship between the barrier of the nodule and the thyroid cartilage (Figure 1B). DECT of the neck confirmed that the mass anterior to the thyroid cartilage were tophi (Figure 1C-D). A plain radiographic evaluation demonstrated soft-tissue swellings and para-articular erosions involving hands (Figure 1E-F). The serum creatinine and uric acid levels were elevated as follows, creatinine, 562 μmol per liter (normal range, 31.8 to 116); and uric acid, 621 μmol per liter (normal range, 90 to 420).

Figure 1.

Tophi detected by different imaging methods. A.A picture of the mass anterior to the thyroid cartilage; B. A mass anterior to the thyroid cartilage detected by ultrasound; C. The mass anterior to the thyroid cartilage detected by dual energy CT; D. The mass anterior to the thyroid cartilage detected by CT scan; E. a picture of tophi in the hands; F. a plain radiograph of tophi in the hands.

Conclusion

DECT has more advantages over ultrasound and CT for detection of atypical tophi in gout patients.

Disclosure of Interests

Yutong Jiang Grant/research support from: Scientific and Technological Planning Project of Guangzhou City [grant number 202102020150], Xinyu Wu: None declared, Zetao Liao: None declared, Jun Qi: None declared, Jieruo Gu: None declared

POS0014 EFFICACY AND SAFETY OF JAK INHIBITORS IN REFRACTORY OR INITIAL TREATMENT OF ADULT ONSET STILL’S DISEASE

Background

Inappropriate activation of pro-inflammatory cytokines such as interleukin (IL)-1, IL-6 or IL-18, is a pathogenic cornerstone in adult onset Still’s disease (AOSD). Beyond therapies targeting IL-1 and IL-6, Janus kinases (JAK) inhibitors have been suggested to be efficient in refractory AOSD patients [1].

Objectives

To assess the efficacy and safety of JAK inhibitors in the treatment of AOSD patient refractory to, or with initial treatment.

Methods

This retrospective study was based on our single center of the department of rheumatology and immunology. The data was collected from the patients’ medical records using a standardized questionnaire and analyzed at different time points. The response to JAK inhibitors was categorized as complete remission (CR), partial remission (PR) or failure (F) [2].

Results

7 patients were recruited (Table 1), including 4 refractory patients and 3 initial patients. Mean age at JAK inhibitor treatment start was 27.5 years for refractory patients and 35 years for initial patients; and mean disease duration was 66.5 months and 1 month respectively. All patients had fever and polyarthritis, 5 patents had rash. In addition, pulmonary hypertension, abnormal liver function tests, abdominal pain, and heart failure were also observed in our patients. Response to corticosteroids, conventional synthetic or biological Disease Modifying Anti-Rheumatic Drugs (DMARDs) had been considered inadequate in 4 refractory patients. Tofacitinib was added in the initial treatment for 3 patients for high disease activity. In total, baricitinib was used in 2 patients and tofacitinib in 5 patients. Steroids were concurrently used in 6 patients, MTX in three, SASP and NSAIDs in one. At a mean follow up of 3.8 months, complete remission was observed in one patient (with tofacitinib), partial remission was in 5 patients (4 patients with tofacitinib and one with baricitinib), and failure in one (patients with baricitinib). At the last visit, steroids could be decreased but not stopped in those 6 patients. Tolerance of JAK inhibitors was excellent, none infectious disease or other severe side effect were observed.

Table 1.

Characteristics of the AOSD patients

No.SexAge (year)Disease Duration (Month)Main symptomsTreatments before JAKionsetJAKiSteroids at onsetConconmitant treatmentSteroids at the end of F-UF-U(month)Outcome1M2629Fever, polyarthritisPred+MTXTofacitinib0MTX04PR2M259Fever, polyarthritis, rash, pulmonary hypertensionPred+MTXBaricitinib16MTX+Pred124F3M3812Fever, polyarthritis, rashPred+MTX+CsA+NSAIDsTofacitinib24Pred123PR4M21216Fever, polyarthritisPred+MTX+SASP+NSAIDs+TNFiBaricitinib4Pred+MTX+SASP+NSAIDs43PR5F331Fever, polyarthritis, rash, abnormal liver function testsPredTofacitinib40Pred25CR6F471Fever, polyarthritis, rash, abdominal painPredTofacitinib36Pred45PR7F251Fever, polyarthritis, rash, heart failurePredTofacitinib40Pred123PR

Pred: prednisone; MTX: Methotrexate; SASP: salicylazosulfapyridine; CsA: ciclosporin A; NSAIDs: Non-Steroidal Antiinflammatory Drugs; TNFi: Tumor necrosis factor inhibitor. CR: complete remission; PR: partial remission; F: failure.

Conclusion

JAK inhibitors treatment may be helpful for some patients with refractory AOSD, or patients with severe disease activity at initial treatment. Different treatment responses were observed in these short series of cases, which might be due to the phenotype of patients. However, the scale of patients in our study was too low to draw a conclusion. Further study and additional information are needed to evaluate more precisely the risk-benefit ratio of this treatment, and a possible difference in efficacy among the different groups of patients or JAK inhibitors.

References

[1]Aosd T, Table E. Ann Rheum Dis 2020;79:842–4.

[2]Vercruysse F, Barnetche T, Lazaro E, et al. Arthritis Res Ther 2019;21:1–11. doi:10.1186/s13075-019-1838-6.

Figure 1.

Changes in laboratory test (A) WBC; (B) CRP; (C) ESR;(D) Ferritin. M0: baseline; M1: month 1; M3: month 3; M5: month 5.

Acknowledgements

None

Disclosure of Interests

None declared

AB1520 EFFECTIVE TREATMENT OF TOCILIZUMAB IN PATIENTS WITH REFRACTORY ADULT-ONSET STILL’S DISEASE

Background

Pro-inflammatory cytokines such as interleukin 6 (IL-6) are involved in the pathogenesis of adult-onset Still’s disease (AOSD). Anti-IL-6 agents such as tocilizumab have been tried to treat AOSD successfully.

Objectives

To access the efficacy of tocilizumab in the treatment of AOSD patient refractory to,or with initial treatment.

Methods

We reviewed three cases with refractory AOSD treated with tocilizumab. All patients fulfill the Cush criteria for AOSD. All patients performed blood culture, auto-antibodies test and bone marrow test to exclude infectious diseases, other rheumatic diseases and tumors. All patients received broad-spectrum antibiotics and had no response. All patients received glucocorticoid therapy and at least one kind of anti-cytokine therapies but didn’t have full response. Then the three patients received intravenous tocilizumab treatment of 8mg/kg every 2 weeks or 4 weeks.

Results

The first patient, a 21-year-old woman, performed positron emission tomography (PET-CT) and lymphoglandula pathology in addition to routine tests. She received high dose methylprednisolone (500mg per day for 3 days and followed by 80mg per day), gamma globulin injection (20g per day for 3 days) and baricitinib 4mg per day for 12 days, but had no response to the treatment. Then she received tocilizumab of 8mg/kg every 2 weeks and stopped baricitinib. And the symptoms and blood tests improved gradually, and the methylprednisolone dose reduced to 16mg per day at the last follow-up.

The second patient is a 52-year-old man, and performed bone marrow cytology and PET-CT to excluded hematological diseases. He received methylprednisolone 80mg per day, adalimumab and tofacitinib treatment. But the patients still got recurrent fever, high ESR, CRP and serum ferritin. Then he stopped adalimumab and tofacitinib, and received tocilizumab of 8mg/kg every 2 weeks and reduced to 8mg/kg every 4 weeks because of economic factors. The patient did not develop fever and the inflammatory indexes such as ESR/CRP gradually decreased to normal range. And methylprednisolone dose reduced to 32mg nowadays.

The third patient is a 30-year-old woman and has recurrent AOSD for 15 years. She had tried glucocorticoids, methotrexate, iguratimod, baritinib and entanercept successively. Yet she still had recurrent arthritis on hand and knee, and elevated ESR/CRP/serum ferritin. Additionally she suffered femoral head necrosis because of excessive doses of glucocorticoid. The patient received tocilizumab 8mg/kg every 4 weeks, then joint symptoms and inflammatory indicators improved significantly. The methylprednisolone dose was also successfully reduced to 4mg/d.

The Figure 1 showed the main course of disease evolution.

Figure 1.

Conclusion

Tocilizumab may be an effective candidate in refractory AOSD despite no response to other treatments.

Disclosure of Interests

None declared

Evaluation of the pollutant interactions between different overlying water and sediment in simulated urban sewer system by excitation-emission matrix fluorescence spectroscopy

The water quality in the sewer systems can be significantly influenced by the interaction between sediment and overlying water, which are still many doubts about the impact of pollutants transformation, degradation sequence, and reaction time. In this study, the exchanging processes between sewer sediment and four different overlying waters were evaluated in simulated urban sewer systems (dark and anaerobic environments). Dissolved organic matter (DOM) was used as an indicator to reflect the mitigation and exchange processes of pollutants. Excitation-emission matrix (EEM) fluorescence spectroscopy coupled with parallel factor analysis (PARAFAC) as an effective method for deciphering DOM properties was applied. There are three findings: (1) Three main processes (biological degradation, desorption, and adsorption) happened in the simulated sewer systems, in which the predominant pathway in the interaction process is biological degradation though consuming amino acid components. (2) The characteristics of overlying water could induce significant changes in sediment signatures; the amino acid-like components are more susceptible to degradation, and the humic-like compositions are more readily absorbed by sediments. (3) The reaction time is another significant factor (14 days was the turning point of the processes). This study unravels the transformation processes in sediment and different overlying waters, which provides the theoretical foundation for urban sewer efficient management and operation.

Quality of life in first episode psychosis: a cluster analytic approach

PURPOSE

Quality of life is increasingly recognised as an important outcome for young people with first episode psychosis (FEP). The first aim was to determine whether distinct homogenous subgroups of young people with FEP could be delineated based on profiles on quality of life domains (Physical Health, Psychological, Social relationships and Environmental). The second aim was to examine the discriminant validity of these subgroups with respect to demographic, functioning and clinical features of FEP.

METHOD

Quality of life, demographic characteristics, clinical characteristics, cognition and functioning were assessed in 145 people with FEP. Cluster analysis using Ward's methods and Squared Euclidean Distance with a k-means verification were employed to identify subgroups with homogenous quality of life profiles. The clusters were externally validated using multinomial logistic regressions.

RESULTS

Three distinct quality of life profiles were identified: one with good quality of life across all domains (30%), one with poor quality of life particularly in Psychological and Social relationships domains (28%), and one 'intermediate' group with comparatively low Psychological quality of life (42%). Depression, semantic verbal fluency, social inclusion and social/occupational functioning showed associations with group membership.

CONCLUSION

Our results suggest the potential of maintaining relatively good quality of life despite the experience of FEP. Future research on interventions to improve quality of life may consider the potential of addressing depression, social inclusion and social/occupational functioning.

Revealing the characteristics of dissolved organic matter in urban runoff at three typical regions via optical indices and molecular composition

Dissolved organic matter (DOM) plays a major role in ecological systems and influences the fate and transportation of many pollutants. Despite the significance of DOM, understanding of how environmental and anthropogenic factors influence its composition and characteristics is limited, especially in urban stormwater runoff. In this article, the chemical properties (pollutant loads, molecular weight, aromaticity, sources, and molecular composition) of DOM in stormwater extracted from three typical end-members (traffic, residential, and campus regions) were characterized by UV-visible (UV-vis) spectroscopy, excitation-emission matrix spectroscopy combined with parallel factor analysis (EEM-PARAFAC), and ultra-performance liquid chromatography quadrupole time-of-flight mass spectrometry (UPLC-Q-TOF-MS). There are three findings: (1) The basic properties of DOM in stormwater runoff varied obviously from three urban fields, and the effect of initial flush was also apparent. (2) The DOM in residential areas mainly came from autochthonous sources, while allochthonous sources primarily contributed to the DOM in traffic and campus areas. However, it was mainly composed of terrestrial humic-like components with CHO and CHON element composition and HULO and aliphatic formulas. (3) The parameters characterizing DOM were primarily related to terrestrial source and aromaticity, but their correlations varied. Through the combination of optical methods and UPLC-Q-TOF spectrometry, the optical and molecular characteristics of rainwater are effectively revealed, which may provide a solid foundation for the classification management of stormwater runoff in different urban regions.

[An optimized protocol of meniscus cell extraction for single-cell RNA sequencing]

OBJECTIVE

To optimize the protocol of meniscus cell extraction to enhance the efficiency of cell suspension preparation and maintain a high cell viability for single-cell RNA sequencing.

METHODS

We compared the efficiency of the routine cell extraction methods (short-time digestion and long-time digestion) and the optimized protocol for obtaining meniscus cell suspensions by evaluating the cell number obtained and the cell viability. Single-cell RNA sequencing datasets were analyzed to evaluate the stability of the cell suspension prepared using the optimized protocol. The reliability of the optimized protocol was assessed by comparing the single-cell RNA sequencing dataset obtained by the optimized protocol with published single-cell RNA sequencing datasets of the meniscus.

RESULTS

The optimized protocol harvested a greater number of cells (over 1×10⁵) than the routine protocols. The cell suspension prepared with the optimized protocol showed a cell viability higher than 80%, the highest among the 3 methods. Analysis of single-cell RNA sequencing datasets showed that the ratio of the mitochondrial genes was below 20% in over 80% of the cells. CD34⁺ cells, MCAM⁺ cells and COL1A1⁺ cells were identified in the datasets. Comparison with the publish datasets showed that the optimized protocol was capable of harvesting COL3A1⁺, COL1A1⁺, MYLK⁺, BMP2⁺, CD93⁺ and CDK1⁺ cells.

CONCLUSION

Single-cell suspension prepared from the meniscus can be stably obtained using the optimized protocol for single-cell RNA sequencing using the 10× Genomics platform.

Prediction of spinal anesthesia-induced hypotension during elective cesarean section: a systematic review of prospective observational studies

BACKGROUND

Spinal anesthesia is the standard for elective cesarean section but spinal anesthesia-induced hypotension remains an important problem. Accurate prediction of hypotension could enhance clinical decision-making, alter management, and facilitate early intervention. We performed a systematic review of predictors of spinal anesthesia-induced hypotension and their predictive value during cesarean section.

METHODS

PubMed, Embase, Cochrane Library, Google Scholar and Web of Science databases were searched for prospective observational studies assessing the diagnostic accuracy of predictors of spinal anesthesia-induced hypotension in elective cesarean section. The quality of studies was assessed and predictors were grouped in domains based on the type of predictor.

RESULTS

Thirty-eight studies (n=3086 patients) were included. In most studies, patients received 500-1000 mL crystalloid preload or 500-2000 mL crystalloid coload. Vasopressors for post-spinal hypotension were boluses of ephedrine 5-15 mg and/or phenylephrine 25-100 µg in most studies. The hypotension rate varied from 29% to 80% based on the definition. For analysis, >30 predictors were classified into seven domains: demographic characteristics, baseline hemodynamic variables, baseline sympathovagal balance, postural stress testing, peripheral perfusion indices, blood volume and fluid responsiveness indices, and genetic polymorphism.

CONCLUSIONS

Environmental and individual factors increased outcome variability, which restricted the value of the autonomic nervous system and peripheral perfusion indices for prediction of spinal anesthesia-induced hypotension. Supine stress tests may reflect parturients' cardiovascular tolerance during hemodynamic fluctuations and may optimize the predictive value of static state predictors. Future research for predicting spinal anesthesia-induced hypotension should focus on composite and dynamic parameters during the supine stress tests.

Retrospective Planning Study of Patients with Superior Sulcus Tumours Comparing Pencil Beam Scanning Protons to Volumetric-Modulated Arc Therapy

AIMS

Twenty per cent of patients with non-small cell lung cancer present with stage III locally advanced disease. Precision radiotherapy with pencil beam scanning (PBS) protons may improve outcomes. However, stage III is a heterogeneous group and accounting for complex tumour motion is challenging. As yet, it remains unclear as to whom will benefit. In our retrospective planning study, we explored if patients with superior sulcus tumours (SSTs) are a select cohort who might benefit from this treatment.

MATERIALS AND METHODS

Patients with SSTs treated with radical radiotherapy using four-dimensional planning computed tomography between 2010 and 2015 were identified. Tumour motion was assessed and excluded if greater than 5 mm. Photon volumetric-modulated arc therapy (VMAT) and PBS proton single-field optimisation plans, with and without inhomogeneity corrections, were generated retrospectively. Robustness analysis was assessed for VMAT and PBS plans involving: (i) 5 mm geometric uncertainty, with an additional 3.5% range uncertainty for proton plans; (ii) verification plans at maximal inhalation and exhalation. Comparative dosimetric and robustness analyses were carried out.

RESULTS

Ten patients were suitable. The mean clinical target volume D95 was 98.1% ± 0.4 (97.5-98.8) and 98.4% ± 0.2 (98.1-98.9) for PBS and VMAT plans, respectively. All normal tissue tolerances were achieved. The same four PBS and VMAT plans failed robustness assessment. Inhomogeneity corrections minimally impacted proton plan robustness and made it worse in one case. The most important factor affecting target coverage and robustness was the clinical target volume entering the spinal canal. Proton plans significantly reduced the mean lung dose (by 21.9%), lung V5, V10, V20 (by 47.9%, 36.4%, 12.1%, respectively), mean heart dose (by 21.4%) and thoracic vertebra dose (by 29.2%) (P < 0.05).

CONCLUSIONS

In this planning study, robust PBS plans were achievable in carefully selected patients. Considerable dose reductions to the lung, heart and thoracic vertebra were possible without compromising target coverage. Sparing these lymphopenia-related organs may be particularly important in this era of immunotherapy.

Statistical comparison between SARIMA and ANN's performance for surface water quality time series prediction

The performance comparison studies of the autoregressive integrated moving average model (ARIMA) and the artificial neural network (ANN) were mostly carried out between the selected model structures through trial-and-error, strongly influenced by model structure uncertainty. This research aims to make up for this inadequacy. First, a surface water quality prediction case study including eight monitoring sites in China was introduced. Second, the ARIMA and ANN's performance was compared statistically between 6912 Seasonal ARIMA (SARIMA) and 110,592 feedforward ANN with different model structures, based on the mean square error (MSE) distributions depicted by boxplots. In a statistical view, the ANN models obtained a significantly lower median value and a more concentrated distribution of validation MSEs, which indicated lighter overfitting and better generalization ability. Furthermore, the optimal SARIMA models' performance is inferior to even the median of the ANN models in the case study. In contrast with the previous comparisons among selected models, the statistical comparison in this study shows lower uncertainty.

The Relationship between Pre-Admission Waiting Time and the Surgical Outcomes after Hip Fracture Operation in the Elderly

OBJECTIVES

To investigate the association between pre-admission waiting time and postoperative complications, length of stay (LOS), and costs during hospitalization.

METHOD

This was a single-center, observational study. The subjects of this study were elderly hip fracture patients who were admitted to the Department of Orthopedics, West China Hospital, Sichuan University, from December 2010 to June 14, 2017, and that underwent internal fixation or joint replacement surgery. The pre-admission waiting time was treated as a categorical variable according to median and first quartile values. Outcomes included postoperative complications (pneumonia and other complications [urinary tract infection, heart failure, non-A-grade healing]), LOS, and costs during hospitalization. LOS and costs during hospitalization were converted into binary outcomes based upon median values. Binary logistic regression analyses were used to analyze correlations between pre-admission waiting time and patient outcomes.

RESULTS

A total of 889 patients 60 years of age and older were enrolled in this study, of whom 65.47% were females and 34.53% were males. The proportion of patients with pre-admission waiting times less than 8 h, 8 - 24 h, and ≥ 24 h were 24.3%, 17.32%, and 58.38%, respectively. Postoperative pneumonia and other complications affected 12.04% and 6.30% of patients, respectively. Relative to patients with the pre-admission waiting times of less than 8 h, those with longer pre-admission waiting times exhibited a higher risk of postoperative pneumonia (8 - 24 h: OR = 2.72,95% CI: 1.29-5.74, p = 0.009; ≥ 24 h: OR = 2.76,95% CI: 1.48-5.14, p = 0.001). Patients with the pre-admission waiting time ≥ 24 h also exhibited a higher risk of the other complications (OR = 2.55, 95% CI: 1.53-4.26, p <0.001), a longer LOS (OR = 1.43, 95% CI:1.02-2.01, p = 0.036), and higher costs during hospitalization (OR = 1.51, 95% CI:1.05 - 2.17, p = 0.026) relative to patients with a waiting time less than 8 hours.

CONCLUSION

Pre-admission waiting time was associated with postoperative complications, LOS, and hospitalization costs among older Chinese patients undergoing surgery to treat hip fractures.

Hippocampal adult-born granule cells drive network activity in a mouse model of chronic temporal lobe epilepsy

Temporal lobe epilepsy (TLE) is characterized by recurrent seizures driven by synchronous neuronal activity. The reorganization of the dentate gyrus (DG) in TLE may create pathological conduction pathways for synchronous discharges in the temporal lobe, though critical microcircuit-level detail is missing from this pathophysiological intuition. In particular, the relative contribution of adult-born (abGC) and mature (mGC) granule cells to epileptiform network events remains unknown. We assess dynamics of abGCs and mGCs during interictal epileptiform discharges (IEDs) in mice with TLE as well as sharp-wave ripples (SPW-Rs) in healthy mice, and find that abGCs and mGCs are desynchronized and differentially recruited by IEDs compared to SPW-Rs. We introduce a neural topic model to explain these observations, and find that epileptic DG networks organize into disjoint, cell-type specific pathological ensembles in which abGCs play an outsized role. Our results characterize identified GC subpopulation dynamics in TLE, and reveal a specific contribution of abGCs to IEDs.