Fully automated volumetric measurement of malignant pleural mesothelioma by deep learning AI: validation and comparison with modified RECIST response criteria.

BACKGROUND: In malignant pleural mesothelioma (MPM), complex tumour morphology results in inconsistent radiological response assessment. Promising volumetric methods require automation to be practical. We developed a fully automated Convolutional Neural Network (CNN) for this purpose, performed blinded validation and compared CNN and human response classification and survival prediction in patients treated with chemotherapy. METHODS: In a multicentre retrospective cohort study; 183 CT datasets were split into training and internal validation (123 datasets (80 fully annotated); 108 patients; 1 centre) and external validation (60 datasets (all fully annotated); 30 patients; 3 centres). Detailed manual annotations were used to train the CNN, which used two-dimensional U-Net architecture. CNN performance was evaluated using correlation, Bland-Altman and Dice agreement. Volumetric response/progression were defined as ≤30%/≥20% change and compared with modified Response Evaluation Criteria In Solid Tumours (mRECIST) by Cohen's kappa. Survival was assessed using Kaplan-Meier methodology. RESULTS: Human and artificial intelligence (AI) volumes were strongly correlated (validation set r=0.851, p<0.0001). Agreement was strong (validation set mean bias +31 cm3 (p=0.182), 95% limits 345 to +407 cm3). Infrequent AI segmentation errors (4/60 validation cases) were associated with fissural tumour, contralateral pleural thickening and adjacent atelectasis. Human and AI volumetric responses agreed in 20/30 (67%) validation cases κ=0.439 (0.178 to 0.700). AI and mRECIST agreed in 16/30 (55%) validation cases κ=0.284 (0.026 to 0.543). Higher baseline tumour volume was associated with shorter survival. CONCLUSION: We have developed and validated the first fully automated CNN for volumetric MPM segmentation. CNN performance may be further improved by enriching future training sets with morphologically challenging features. Volumetric response thresholds require further calibration in future studies.

Inter-observer variation in image interpretation and the prognostic importance of non-expansile lung in malignant pleural effusion.

BACKGROUND AND OBJECTIVE: Non-expansile lung (NEL) frequently complicates management of malignant pleural effusion (MPE) and is an important factor in clinical practice and trials. NEL is frequently diagnosed on a single radiographic observation, but neither the inter-observer agreement of this approach nor the prognostic importance of NEL in MPE has been reported. METHODS: A multicentre retrospective cohort study was performed in two UK pleural centres. NEL was defined as <50% pleural re-apposition on post-drainage radiographs by primary and secondary assessors at each site. Inter-observer agreement was assessed by Cohen's kappa (κ). Kaplan-Meier methodology and multivariate Cox models were used to assess the prognostic impact of NEL versus no NEL and 'complete NEL' versus 'complete expansion', based on a single assessor's results from each site. RESULTS: NEL was identified by the primary assessor in 33 of 97 (34%) in Cohort 1 and 15 of 86 (17%) in Cohort 2. Inter-observer agreement between assessors was only fair-to-moderate (Cohort 1 κ: 0.38 (95% CI: 0.21-0.55), Cohort 2 κ: 0.51 (95% CI: 0.30-0.72)). In both cohorts, NEL was associated with shorter median overall survival (Cohort 1: 188 vs 371 days, Cohort 2: 192 vs 412 days). This prognostic association was independent in Cohort 1 (hazard ratio (HR): 2.19, 95% CI: 1.31-3.66) but not in Cohort 2 (HR: 1.42, 95% CI: 0.71-2.87). Survival was inferior in both cohorts in cases of complete NEL versus complete expansion. CONCLUSION: Radiographic NEL is common but inter-observer agreement is only fair-to-moderate. NEL is associated with adverse survival. These data do not support the use of single radiographic assessments to classify NEL.

Time to Computerized Tomography Scan, Age, and Mortality in Acute Stroke.

BACKGROUND: Time to computerized tomography (CT) is important to institute appropriate and timely hyperacute management in stroke. We aimed to evaluate mortality outcomes in relation to age and time to CT scan. METHODS: We used routinely collected data in 8 National Health Service trusts in East of England between September 2008 and April 2011. Stroke cases were prospectively identified and confirmed. Odds ratios (ORs) for unadjusted and adjusted models for age categories (<65, 65-74, 75-84, and ≥85 years) as well as time to CT categories (<90 minutes, ≥90 to <180 minutes, ≥180 minutes to 24 hours, and >24 hours) and in-hospital and early (<7 days) mortality outcomes were calculated. RESULTS: Of the 7693 patients (mean age 76.1 years, 50% male) included, 1151 (16%) died as inpatients and 336 (4%) died within 7 days. Older patients and those admitted from care home had a significantly longer time from admission until CT (P < .001). Patients who had earlier CT scans were admitted to stroke units more frequently (P < .001) but had higher in-patient (P < .001) and 7-day mortality (P < .001). Whereas older age was associated with increased odds of mortality outcomes, longer time to CT was associated with significantly reduced mortality within 7 days (corresponding ORs for the above time periods were 1.00, .61 [95% confidence interval {CI}: .39-.95], .39 [.24-.64], and .16 [.08-.33]) and in-hospital mortality (ORs 1.00, .86 [.64-1.15], .57 [.42-.78] and .71 [.52-.98]). CONCLUSIONS: Older age was associated with a significantly longer time to CT. However, using CT scan time as a benchmarking tool in stroke may have inherent limitations and does not appear to be a suitable quality marker.

A simple 5-point scoring system, NaURSE (Na+, urea, respiratory rate and shock index in the elderly), predicts in-hospital mortality in oldest old.

BACKGROUND: the mortality is high in acutely ill oldest old patients. Understanding the prognostic factors which influence mortality will help clinicians make appropriate management decisions. METHODS: we analysed prospective mortality audit data (November 2008 to January 2009) to identify variables associated with in-patient mortality in oldest old. We selected those with P < 0.10 from univariate analysis and determined at which cut-point they served as the strongest predictor of mortality. Using these cut-off points, we constructed multivariate logistic regression models. A 5-point score was derived from cut-off points which were significantly associated with mortality tested in a smaller independent re-audit sample conducted in October 2011. RESULTS: a total of 405 patients (mean 93.5 ± 2.7 years) were included in the study. The mean length of stay was 18.5 ± 42.4 days and 13.8% died as in-patients. Variables (cut-off values) found to be significantly associated with in-patient mortality were admission sodium (>145 mmol/l), urea (≥14 mmol/l), respiratory rate (>20/min) and shock index (>1.0): creating a 5-point score (NaURSE: NaURS in the Elderly). The crude mortality rates were 9.5, 19.9, 34.4, 66.7, and 100% for scores 0, 1, 2, 3 and 4, respectively. Using the cut-off point of ≥2, the NaURSE score has a specificity of 87% (83.1-90.3) and sensitivity of 39% (28.5-50.0), with an AUC value of 0.69 (0.63-0.76). An external independent validation study (n = 121) showed similar results. CONCLUSIONS: the NaURSE score may be particularly useful in identifying oldest old who are likely to die in that admission to guide appropriate care.

Body composition and lung cancer-associated cachexia in TRACERx.

Cancer-associated cachexia (CAC) is a major contributor to morbidity and mortality in individuals with non-small cell lung cancer. Key features of CAC include alterations in body composition and body weight. Here, we explore the association between body composition and body weight with survival and delineate potential biological processes and mediators that contribute to the development of CAC. Computed tomography-based body composition analysis of 651 individuals in the TRACERx (TRAcking non-small cell lung Cancer Evolution through therapy (Rx)) study suggested that individuals in the bottom 20th percentile of the distribution of skeletal muscle or adipose tissue area at the time of lung cancer diagnosis, had significantly shorter lung cancer-specific survival and overall survival. This finding was validated in 420 individuals in the independent Boston Lung Cancer Study. Individuals classified as having developed CAC according to one or more features at relapse encompassing loss of adipose or muscle tissue, or body mass index-adjusted weight loss were found to have distinct tumor genomic and transcriptomic profiles compared with individuals who did not develop such features. Primary non-small cell lung cancers from individuals who developed CAC were characterized by enrichment of inflammatory signaling and epithelial-mesenchymal transitional pathways, and differentially expressed genes upregulated in these tumors included cancer-testis antigen MAGEA6 and matrix metalloproteinases, such as ADAMTS3. In an exploratory proteomic analysis of circulating putative mediators of cachexia performed in a subset of 110 individuals from TRACERx, a significant association between circulating GDF15 and loss of body weight, skeletal muscle and adipose tissue was identified at relapse, supporting the potential therapeutic relevance of targeting GDF15 in the management of CAC.

Imaging in pleural mesothelioma: A review of the 15th International Conference of the International Mesothelioma Interest Group.

Imaging of mesothelioma plays a role in all aspects of patient management, including disease detection, staging, evaluation of treatment options, response assessment, pre-surgical evaluation, and surveillance. Imaging in this disease impacts a wide range of disciplines throughout the healthcare enterprise. Researchers and clinician-scientists are developing state-of-the-art techniques to extract more of the information contained within these medical images and to utilize it for more sophisticated tasks; moreover, image-acquisition technology is advancing the inherent capabilities of these images. This paper summarizes the imaging-based topics presented orally at the 2021 International Conference of the International Mesothelioma Interest Group (iMig), which was held virtually from May 7-9, 2021. These topics include an update on the mesothelioma staging system, novel molecular targets to guide therapy in mesothelioma, special considerations and potential pitfalls in imaging mesothelioma in the immunotherapy setting, tumor measurement strategies and their correlation with patient survival, tumor volume measurement in MRI and CT, CT-based texture analysis for differentiation of histologic subtype, diffusion-weighted MRI for the assessment of biphasic mesothelioma, and the prognostic significance of skeletal muscle loss with chemotherapy.

The Detection of Lymphatic Invasion in Colorectal Polyp Cancer Using D2-40 Immunohistochemistry and Its Association With Prognosis.

Introduction The aim of this study was to compare the detection of lymphatic invasion using haematoxylin and eosin (H&E) staining versus D2-40 immunostaining on specimens from a retrospective cohort of patients with colorectal polyp cancer and to investigate the association of lymphatic invasion, detected by either method, with survival. Methods Specimens from patients with pathologically diagnosed colorectal polyp cancer were selected from the Greater Glasgow and Clyde Bowel Cancer Screening Registry for D2-40 immunohistochemistry staining. Clinicopathological information was retrieved from patient electronic records including analysis of pathology reports to determine if a lymphatic invasion was detected using H&E staining. Results Over 100 patients were included in this study with a median age at polypectomy of 66 years (range 50-76). All patients were followed up for a minimum of four years and five patients died due to colorectal cancer. The lymphatic invasion was detected in 8% of cases by H&E staining and 23% of cases with D2-40 immunostaining. Only D2-40-detected lymphatic invasion showed a statistically significant relationship with colorectal cancer-specific mortality using univariate analysis (p=0.01). Survival analysis performed separately by Cox regression demonstrated that lymphatic invasion detected by D2-40 immunostaining was associated with worse disease-specific survival (hazard ratio [HR] 14.07, 95% CI 1.57-125.97, p=0.018). Conclusion This study shows that D2-40 immunostaining can improve the detection of lymphatic invasion in colorectal polyp cancer when compared to H&E staining. In addition, the lymphatic invasion detected by D2-40 immunostaining significantly associates with survival allowing it to be used as a prognostic indicator in colorectal polyp cancer.

Cancer cachexia in thoracic malignancy: a narrative review.

PURPOSE OF REVIEW: Thoracic malignancies are amongst the most lethal of all cancers. Cancer cachexia lacks unanimously accepted diagnostic criteria, and therefore is referenced to as a conceptual framework whereby cancer cachexia is 'an ongoing loss of skeletal muscle mass (termed sarcopenia), with or without loss of fat mass that cannot be reversed by conventional nutritional support and leads to progressive functional impairment'. This review summarises the current evidence base in this field, including imaging techniques currently used to define sarcopenia, inflammatory and metabolic changes associated with the syndrome and ongoing research into potential treatment strategies. RECENT FINDINGS: Sarcopenia is a key component of the cancer cachexia syndrome. It is common in patients with both early-stage and advanced NSCLC. Patients with sarcopenia have more treatment-related side effects and poorer overall survival compared with nonsarcopenic patients. SUMMARY: Early identification of cancer cachexia may facilitate stratification of patients most-at-risk and initiation of emerging anticachexia treatments. If these are proven to be effective, this strategy has the potential to improve tolerance to anti-cancer therapies, improving the quality of life, and perhaps the survival, of patients with thoracic malignancies.

Doctors Are Inconsistent in Estimating Survival after CPR and Are Not Using Such Predictions Consistently in Determining DNACPR Decisions.

Background: It is unclear whether doctors base their resuscitation decisions solely on their perceived outcome. Through the use of theoretical scenarios, we aimed to examine the 'do not attempt cardiopulmonary resuscitation' (DNACPR) decision-making. Methods: A questionnaire survey was sent to consultants and specialty trainees across two Norfolk (UK) hospitals during December 2013. The survey included demographic questions and six clinical scenarios with varying prognosis. Participants were asked if they would resuscitate the patient or not. Identical scenarios were then shown in a different order and doctors were asked to quantify patients' estimated chance of survival. Results: A total of 137 individuals (mean age 41 years (SD 7.9%)) responded. The response rate was 69%. Approximately 60% were consultants. We found considerable variation in clinician estimates of median chance of survival. In three out of six of our scenarios, the survival estimated varied from <1% to 95%. There was a statistically significant difference identified in the estimated median survival between those clinicians who would or would not resuscitate in four of the six scenarios presented. Conclusion: This study has highlighted the wide variation between clinicians in their estimates of likely survival and little concordance between clinicians over their resuscitation decisions. The diversity in clinician decision-making should be explored further.

Pre-EDIT: A Randomized Feasibility Trial of Elastance-Directed Intrapleural Catheter or Talc Pleurodesis in Malignant Pleural Effusion.

BACKGROUND: Talc slurry pleurodesis (TSP) prevents recurrence of symptomatic malignant pleural effusion (MPE) in 71% to 78% patients. Nonexpansile lung (NEL) frequently accounts for TSP failure but is often occult predrainage, impairing selection of patients. NEL is associated with high pleural elastance (PEL), but technical limitations have hampered the development of PEL as a predictive NEL marker. We performed a single-center, randomized, controlled, open-label feasibility trial of EDIT (elastance-directed indwelling pleural catheter or TSP) management, using a novel digital manometer and a new definition of high PEL. METHODS: Patients with symptomatic MPE were randomized 1:1 between EDIT and standard care (TSP). EDIT involved PEL assessment during large-volume thoracentesis; patients with high PEL (maximum PEL sustained over 250 mL [MaxPEL250] ≥ 14.5 cm H2O/L) were allocated to immediately receive an indwelling pleural catheter; the remainder underwent immediate drain placement for TSP. The primary outcome measure was recruitment feasibility, defined a priori as 30 patients over 12 months. Secondary outcomes included safety, technical reliability, and the aspiration volume required to detect high PEL. The accuracy of the PEL definition for NEL was analyzed post hoc. RESULTS: Thirty-one patients were randomized (one allocation failure) over 12 months. PEL assessment (mean duration, 33 minutes) was successful in 13 of 15 patients (87%). No directly attributable serious adverse events occurred. High PEL was detected in seven of 13 patients (54%), associated with 100% sensitivity and 67% specificity for NEL, and was first detected at a median volume of 325 mL (range, 250-800 mL). CONCLUSIONS: A phase 3 trial testing the effect of EDIT management on symptomatic MPE recurrence following TSP is feasible. TRIAL REGISTRY: ClinicalTrials.gov; No.: NCT03319186; URL: www.clinicaltrials.gov.

Pre-EDIT: protocol for a randomised feasibility trial of elastance-directed intrapleural catheter or talc pleurodesis (EDIT) in malignant pleural effusion.

INTRODUCTION: Non-expansile lung (NEL) is a common cause of talc pleurodesis (TP) failure in malignant pleural effusion (MPE), but is often occult prior to drainage. Reliable detection of NEL would allow patients to be allocated between intrapleural catheter (IPC) and TP. High pleural elastance (PEL) has been associated with NEL in observational studies. Pre-EDIT is a randomised feasibility trial of elastance-directed IPC or TP (EDIT) management using a novel, purpose-built digital pleural manometer (Rocket Medical, UK). METHODS AND ANALYSIS: Consecutive patients with MPE without prior evidence of NEL or preference for IPC will be randomised 1:1 between EDIT management and standard care (an attempt at TP). The primary objective is to determine whether sufficient numbers of patients (defined as 30 within 12 months (or 15 over 6 months)) can be recruited and randomised to justify a subsequent phase III trial testing the efficacy of EDIT management. Secondary objectives include safety, technical feasibility and validation of study design elements, including the definition of PEL using 4D pleural MRI before and after fluid aspiration. EDIT involves PEL assessment during a large volume pleural fluid aspiration, followed by an attempt at TP or placement of an IPC within 24 hours. Patients will be allocated to IPC if the rolling average PEL sustained over at least 250 mL fluid aspirated (PEL250) is ≥ 14.5 cm H2O/L. ETHICS AND DISSEMINATION: Pre-EDIT was approved by the West of Scotland Regional Ethics Committee on 8 March 2017 (Ref: 17/WS/0042). Results will be presented at scientific meetings and published in peer-reviewed journals. TRIAL REGISTRATION NUMBER: NCT03319186; Pre-results.

Survival prediction in mesothelioma using a scalable Lasso regression model: instructions for use and initial performance using clinical predictors.

INTRODUCTION: Accurate prognostication is difficult in malignant pleural mesothelioma (MPM). We developed a set of robust computational models to quantify the prognostic value of routinely available clinical data, which form the basis of published MPM prognostic models. METHODS: Data regarding 269 patients with MPM were allocated to balanced training (n=169) and validation sets (n=100). Prognostic signatures (minimal length best performing multivariate trained models) were generated by least absolute shrinkage and selection operator regression for overall survival (OS), OS <6 months and OS <12 months. OS prediction was quantified using Somers DXY statistic, which varies from 0 to 1, with increasing concordance between observed and predicted outcomes. 6-month survival and 12-month survival were described by area under the curve (AUC) scores. RESULTS: Median OS was 270 (IQR 140-450) days. The primary OS model assigned high weights to four predictors: age, performance status, white cell count and serum albumin, and after cross-validation performed significantly better than would be expected by chance (mean DXY0.332 (±0.019)). However, validation set DXY was only 0.221 (0.0935-0.346), equating to a 22% improvement in survival prediction than would be expected by chance. The 6-month and 12-month OS signatures included the same four predictors, in addition to epithelioid histology plus platelets and epithelioid histology plus C-reactive protein (mean AUC 0.758 (±0.022) and 0.737 (±0.012), respectively). The <6-month OS model demonstrated 74% sensitivity and 68% specificity. The <12-month OS model demonstrated 63% sensitivity and 79% specificity. Model content and performance were generally comparable with previous studies. CONCLUSIONS: The prognostic value of the basic clinical information contained in these, and previously published models, is fundamentally of limited value in accurately predicting MPM prognosis. The methods described are suitable for expansion using emerging predictors, including tumour genomics and volumetric staging.

The relationship between total anticholinergic burden (ACB) and early in-patient hospital mortality and length of stay in the oldest old aged 90 years and over admitted with an acute illness.

The use of prescription drugs in older people is high and many commonly prescribed drugs have anticholinergic effects. We examined the relationship between ACB on mortality and in-patient length of stay in the oldest old hospitalised population. This was a retrospective analysis of prospective audit using hospital audit data from acute medical admissions in three hospitals in England and Scotland. Baseline use of possible or definite anticholinergics was determined according to the Anticholinergic Cognitive Burden Scale. The main outcome measures were decline in-hospital mortality, early in-hospital mortality at 3- and 7-days and in-patient length of stay. A total of 419 patients (including 65 patients with known dementia) were included [median age=92.9, inter-quartile range (IQR) 91.4-95.1 years]. 256 (61.1%) were taking anticholinergic medications. Younger age, greater number of pre-morbid conditions, ischemic heart disease, number of medications, higher urea and creatinine levels were significantly associated with higher total ACB burden on univariate regression analysis. There were no significant differences observed in terms of in-patient mortality, in-patient hospital mortality within 3- and 7-days and likelihood of prolonged length of hospital stay between ACB categories. Compared to those without cardiovascular disease, patients with cardiovascular disease showed similar outcome regardless of ACB load (either =0 or >0 ACB). We found no association between ACB and early (within 3- and 7-days) and in-patient mortality and hospital length of stay outcomes in this cohort of oldest old in the acute medical admission setting.