RESUMEN
There are large differences in premature mortality in the USA by race/ethnicity, education, rurality and social vulnerability index groups. Using existing concentration-response functions, published particulate matter (PM2.5) air pollution estimates, population estimates at the census tract level and county-level mortality data from the US National Vital Statistics System, we estimated the degree to which these mortality discrepancies can be attributed to differences in exposure and susceptibility to PM2.5. We show that differences in PM2.5-attributable mortality were consistently more pronounced by race/ethnicity than by education, rurality or social vulnerability index, with the Black American population having the highest proportion of deaths attributable to PM2.5 in all years from 1990 to 2016. Our model estimates that over half of the difference in age-adjusted all-cause mortality between the Black American and non-Hispanic white population was attributable to PM2.5 in the years 2000 to 2011. This difference decreased only marginally between 2000 and 2015, from 53.4% (95% confidence interval 51.2-55.9%) to 49.9% (95% confidence interval 47.8-52.2%), respectively. Our findings underscore the need for targeted air quality interventions to address environmental health disparities.
RESUMEN
There are large differences in premature mortality in the USA by racial/ethnic, education, rurality, and social vulnerability index groups. Using existing concentration-response functions, particulate matter (PM2.5) air pollution, population estimates at the tract level, and county-level mortality data, we estimated the degree to which these mortality discrepancies can be attributed to differences in exposure and susceptibility to PM2.5. We show that differences in mortality attributable to PM2.5 were consistently more pronounced between racial/ethnic groups than by education, rurality, or social vulnerability index, with the Black American population having by far the highest proportion of deaths attributable to PM2.5 in all years from 1990 to 2016. Over half of the difference in age-adjusted all-cause mortality between the Black American and non-Hispanic White population was attributable to PM2.5 in the years 2000 to 2011.
RESUMEN
Generative Artificial Intelligence is set to revolutionize healthcare delivery by transforming traditional patient care into a more personalized, efficient, and proactive process. Chatbots, serving as interactive conversational models, will probably drive this patient-centered transformation in healthcare. Through the provision of various services, including diagnosis, personalized lifestyle recommendations, dynamic scheduling of follow-ups, and mental health support, the objective is to substantially augment patient health outcomes, all the while mitigating the workload burden on healthcare providers. The life-critical nature of healthcare applications necessitates establishing a unified and comprehensive set of evaluation metrics for conversational models. Existing evaluation metrics proposed for various generic large language models (LLMs) demonstrate a lack of comprehension regarding medical and health concepts and their significance in promoting patients' well-being. Moreover, these metrics neglect pivotal user-centered aspects, including trust-building, ethics, personalization, empathy, user comprehension, and emotional support. The purpose of this paper is to explore state-of-the-art LLM-based evaluation metrics that are specifically applicable to the assessment of interactive conversational models in healthcare. Subsequently, we present a comprehensive set of evaluation metrics designed to thoroughly assess the performance of healthcare chatbots from an end-user perspective. These metrics encompass an evaluation of language processing abilities, impact on real-world clinical tasks, and effectiveness in user-interactive conversations. Finally, we engage in a discussion concerning the challenges associated with defining and implementing these metrics, with particular emphasis on confounding factors such as the target audience, evaluation methods, and prompt techniques involved in the evaluation process.
RESUMEN
BACKGROUND AND OBJECTIVE: Data on enteral tube feeding in head and neck cancer (HNC) patients undergoing chemoradiotherapy vary considerably between German institutions. This survey aims to investigate the management of feeding tubes in an interdisciplinary context across Germany. MATERIALS AND METHODS: Between December 2022 and May 2023, 70 participants (42 radiation oncologists, 12 medical oncologists, 14 head and neck surgeons, and 2 physicians covering several specialties) responded to our web-based survey. In addition to the type of institution (university hospital, private practice, etc.), their age, and professional experience (in years), participants were asked several questions on the indication and institutional policy for tube placement and management (prophylactic/reactive nasogastric or gastrostomy tube). All questions were mandatory single- or multiple-choice questions, while additional comments were possible by email. RESULTS: Most participants were employed at a university hospital (nâ¯= 52; 74.3%) and came from a radiation oncology background (nâ¯= 42; 60%). Fifty-four contributors (77.1%) reported that no nutritional risk screening prior to chemoradiotherapy was routinely performed, and 71.4% (nâ¯= 50) stated that no standardized protocol was used at the institution to set the indication for tube placement. Generally, policies and methods of tube feeding vary considerably between the individual institutions and specialties. However, the majority (nâ¯= 56, 80%) recommended a prophylactic percutaneous enteral gastrostomy (PEG) tube to their patients before chemoradiotherapy. Still, there was no consistent trend regarding the approach for reactive tube feeding. CONCLUSION: The policies and methods of tube feeding vary considerably between the individual institutions and specialties in Germany. In the era of individualized medicine, uniform protocols are difficult to establish. However, a baseline nutritional risk screening could simplify decision-making in clinical practice.
Asunto(s)
Quimioradioterapia , Nutrición Enteral , Gastrostomía , Neoplasias de Cabeza y Cuello , Intubación Gastrointestinal , Humanos , Alemania , Neoplasias de Cabeza y Cuello/terapia , Masculino , Encuestas y Cuestionarios , Pautas de la Práctica en Medicina/estadística & datos numéricos , Femenino , Encuestas de Atención de la Salud , Comunicación Interdisciplinaria , Persona de Mediana Edad , Oncólogos de RadiaciónRESUMEN
In this study, we develop a 3D beta variational autoencoder (beta-VAE) to advance lung cancer imaging analysis, countering the constraints of conventional radiomics methods. The autoencoder extracts information from public lung computed tomography (CT) datasets without additional labels. It reconstructs 3D lung nodule images with high quality (structural similarity: 0.774, peak signal-to-noise ratio: 26.1, and mean-squared error: 0.0008). The model effectively encodes lesion sizes in its latent embeddings, with a significant correlation with lesion size found after applying uniform manifold approximation and projection (UMAP) for dimensionality reduction. Additionally, the beta-VAE can synthesize new lesions of varying sizes by manipulating the latent features. The model can predict multiple clinical endpoints, including pathological N stage or KRAS mutation status, on the Stanford radiogenomics lung cancer dataset. Comparisons with other methods show that the beta-VAE performs equally well in these tasks, suggesting its potential as a pretrained model for predicting patient outcomes in medical imaging.
Asunto(s)
Procesamiento de Imagen Asistido por Computador , Neoplasias Pulmonares , Humanos , Neoplasias Pulmonares/diagnóstico por imagen , Mutación , Proyección , RadiómicaRESUMEN
During the SARS-CoV-2 pandemic, the German healthcare system faced challenges of efficiently allocating testing resources. To address this, we developed an open-source personalized recommendation system (PRS) called "CovApp". The PRS utilized a questionnaire to estimate the risk of infection, provided personalized recommendations such as testing, self-isolation, or quarantine, and featured QR code data transmission to electronic health records. The PRS served up to 2.5 million monthly users and received 67,000 backlinks from 1800 domains. We clinically evaluated the PRS at the SARS-CoV-2 testing facility at Charité and observed a 21.7% increase in patient throughput per hour and a 22.5% increase in patients per day. Patients using the PRS were twice as likely to belong to the High Risk group eligible for testing (18.6% vs. 8.9%, p < 0.0001), indicating successful compliance with CovApp's recommendations. CovApp served as a digital bridge between the population and medical staff and significantly improved testing efficiency. As an open-source platform, CovApp can be readily customized to address emerging public health crises. Further, given the EHR interface, the app is of great utility for other applications in clinical settings.
Asunto(s)
COVID-19 , SARS-CoV-2 , Humanos , COVID-19/epidemiología , Registros Electrónicos de Salud , Prueba de COVID-19 , Atención a la Salud , InternetRESUMEN
PURPOSE: To investigate the influence of radiation dose to the swallowing muscles on the nutritional status in patients with head and neck cancer undergoing primary or adjuvant (chemo)radiotherapy (C)RT. METHODS: Between 2018 and 2020, 61 patients were prospectively randomized into the so-called HEADNUT trial (head and neck cancer patients undergoing nutritional intervention). Follow-up was continued until 2022. Contouring of the swallowing apparatus included the superior (scm), middle (mcm), and inferior constrictor muscle (icm), the cricopharyngeal muscle (cphm), and the esophageal inlet. Nutritional status was assessed by bioelectrical impedance analysis (BIA) at the beginning and the end of radiotherapy. The posttherapeutic nutritional status was evaluated by the BIA-derived fat-free mass index (FFMI; kg/m2). Malnutrition was assumed at FFMI values of <â¯15 (women) and <â¯17 (men) kg/m2. To find differences between dosimetric parameters in well- and malnourished patients, Mann-Whitney U test was used. To model the association between malnutrition and its potentially influencing variables, several logistic regression models were built. RESULTS: The following parameters differed between well- and malnourished patients at the end of therapy: icm Dmean, V40Gy (%), V50Gy (%), and V60Gy (%), and sphm V40Gy (%). After entering these parameters into a multivariable logistic regression model (dosimetric model), icm Dmean (bâ¯= -0.12; Exp(b)â¯= 0.88; 95% CI: 0.78-1.0; pâ¯= 0.06) and icm V40Gy (%; bâ¯= 0.06; Exp(b)â¯= 1.07; 95% CI: 1-1.13; pâ¯= 0.04) proved to be independent dosimetric predictors of malnutrition. We only determined the cut-off value for predicting malnutrition for icm V40Gy (%) since it was the only parameter which met pâ¯< 0.05. The optimal cut-off value for the predictor V40Gy (%) based on the Youden Index was 85.6%. Another logistic regression model (dosimetric-clinical model) consisted of icm V40 (%) and the clinical parameters tumor localization, malnutrition before RT, gender, and combined chemotherapy. It was confirmed that both icm V40% (bâ¯= -1.9; Exp(b)â¯= -2.7; 95% CI: 0.01-0.8; pâ¯= 0.03) and malnutrition at baseline (bâ¯= -1.9; Exp(b)â¯= 4.4; 95% CI: 8.4-816.6; pâ¯= 0.0002) were independent predictors of subsequent malnutrition the end of RT. CONCLUSION: Establishment of a normal nutritional status before the start of RT and adherence to dose constraints for the swallowing apparatus may prevent malnutrition in head and neck cancer patients at the end of therapy. Specifically, we suggest an icm V40Gy (%) of more than 86% to be predictive for nutritional complications.
Asunto(s)
Neoplasias de Cabeza y Cuello , Desnutrición , Masculino , Humanos , Femenino , Estado Nutricional , Estudios Prospectivos , Deglución , Neoplasias de Cabeza y Cuello/radioterapia , Desnutrición/diagnóstico , Desnutrición/etiologíaRESUMEN
Undetected infection and delayed isolation of infected individuals are key factors driving the monkeypox virus (now termed mpox virus or MPXV) outbreak. To enable earlier detection of MPXV infection, we developed an image-based deep convolutional neural network (named MPXV-CNN) for the identification of the characteristic skin lesions caused by MPXV. We assembled a dataset of 139,198 skin lesion images, split into training/validation and testing cohorts, comprising non-MPXV images (n = 138,522) from eight dermatological repositories and MPXV images (n = 676) from the scientific literature, news articles, social media and a prospective cohort of the Stanford University Medical Center (n = 63 images from 12 patients, all male). In the validation and testing cohorts, the sensitivity of the MPXV-CNN was 0.83 and 0.91, the specificity was 0.965 and 0.898 and the area under the curve was 0.967 and 0.966, respectively. In the prospective cohort, the sensitivity was 0.89. The classification performance of the MPXV-CNN was robust across various skin tones and body regions. To facilitate the usage of the algorithm, we developed a web-based app by which the MPXV-CNN can be accessed for patient guidance. The capability of the MPXV-CNN for identifying MPXV lesions has the potential to aid in MPXV outbreak mitigation.
Asunto(s)
Aprendizaje Profundo , Mpox , Humanos , Masculino , Estudios Prospectivos , Monkeypox virus , AlgoritmosRESUMEN
Malnutrition negatively impacts quality of life (QoL) in patients with head and neck cancer (HNC). This is the first prospective study to assess the impact of malnutrition (defined by the bioelectrical impedance analysis (BIA)-derived fat-free mass index) on QoL in patients with HNC undergoing (chemo)radiotherapy. Between October 2018 and October 2020, 58 HNC patients prospectively completed the QoL-questionnaires EORTC-QLQ-C30 and EORTC-QLQ-H&N35 at the beginning (tb) and at the end of (chemo)radiotherapy (te) as well as during follow-up (tf). At these time points, nutritional risk assessment (MUST, NRS-2002, Nutriscore), BIA measurement and laboratory testing was performed by a permanent study team. Differences between malnourished (n = 14) and well-nourished patients (n = 44) were observed in UICC classification (P < 0.001) and HPV status (P = 0.03). Well-nourished patients showed higher baseline hemoglobin (P = 0.025) and albumin (P = 0.005), but lower c-reactive protein levels (P < 0.001). At tb, mostly malnourished patients presented with worse QoL. Multivariable analysis showed that MUST, NRS-2002, HPV status, and UICC classification were related to QoL. Nutritional status has a crucial impact on QoL. The nutritional screening protocols MUST and NRS-2002 are suitable for identifying patients at risk and predicting QoL in patients with HNC undergoing (chemo)radiotherapy.
Asunto(s)
Neoplasias de Cabeza y Cuello , Desnutrición , Infecciones por Papillomavirus , Neoplasias de Cabeza y Cuello/radioterapia , Humanos , Desnutrición/etiología , Evaluación Nutricional , Estado Nutricional , Estudios Prospectivos , Calidad de Vida , Encuestas y CuestionariosRESUMEN
INTRODUCTION: The success of intensification and personalisation of the curative treatment of non-small cell lung cancer (NSCLC) is strongly associated with the precision in radiotherapy. Here, we evaluate the impact of radiotherapy protocol adherence in a prospective multicentre trial. METHODS: In the open-label, randomised, controlled PET-Plan trial, patients with inoperable NSCLC were randomized at a 1:1 ratio regarding the target volume delineation informed by 1F-FDG PET and CT plus elective nodal irradiation (arm A) or target volumes informed by PET alone (arm B) and received iso-toxically dose-escalated concurrent chemoradiation. The prospectively organised quality assurance program (RTQA) included individual case review by predefined criteria. For evaluation, protocol adherence was scored as per protocol (pP), with minor (miD), intermediate (inD) and major (maD) deviations. In order to exclude biases through patients who discontinued treatment, patients who received ≥60 Gy were additionally analysed. RESULTS: Between 05/2009-11/2016, 205 patients were randomized, 204 patients started treatment according to protocol of which 31 (15%) patients had maD. Patients with maD had an inferior overall survival (OS) (HR 2.9, 95% CI 1.8-4.4, p < 0.0001) and a higher risk of loco-regional progression (HR 5.7, 95% CI 2.7-11.1, p < 0.0001). These results were significant also in the subgroup of patients receiving ≥ 60 Gy. Patients with maD concerning normal tissue delineation and/or dose constraints had a worse OS (p = 0.006) although no higher incidence of grade ≥ 3 toxicities. CONCLUSIONS: Non-adherence to the radiotherapy protocol was associated with an inferior OS and loco-regional control. These results underline the importance of RTQA.
Asunto(s)
Carcinoma de Pulmón de Células no Pequeñas , Neoplasias Pulmonares , Carcinoma de Pulmón de Células no Pequeñas/terapia , Quimioradioterapia , Humanos , Neoplasias Pulmonares/terapia , Tomografía de Emisión de Positrones , Estudios ProspectivosRESUMEN
Over the last months, cases of SARS-CoV-2 surged repeatedly in many countries but could often be controlled with nonpharmaceutical interventions including social distancing. We analyzed deidentified Global Positioning System (GPS) tracking data from 1.15 to 1.4 million cell phones in Germany per day between March and November 2020 to identify encounters between individuals and statistically evaluate contact behavior. Using graph sampling theory, we estimated the contact index (CX), a metric for number and heterogeneity of contacts. We found that CX, and not the total number of contacts, is an accurate predictor for the effective reproduction number R derived from case numbers. A high correlation between CX and R recorded more than 2 wk later allows assessment of social behavior well before changes in case numbers become detectable. By construction, the CX quantifies the role of superspreading and permits assigning risks to specific contact behavior. We provide a critical CX value beyond which R is expected to rise above 1 and propose to use that value to leverage the social-distancing interventions for the coming months.
Asunto(s)
COVID-19/transmisión , COVID-19/virología , Teléfono Celular , Trazado de Contacto , SARS-CoV-2/fisiología , COVID-19/epidemiología , Alemania/epidemiología , HumanosRESUMEN
(1) Background: The optimal chemotherapy (CHT) regimen for concurrent chemoradiation (cCRT) is not well defined. In this secondary analysis of the international randomized PET-Plan trial, we evaluate the efficacy of different CHT. (2) Methods: Patients with inoperable NSCLC were randomized at a 1:1 ratio regarding the target volume definition and received isotoxically dose-escalated cCRT using cisplatin 80 mg/m2 (day 1, 22) and vinorelbin 15 mg/m2 (day 1, 8, 22, 29) (P1) or cisplatin 20 mg/m2 (day 1-5, 29-33) and vinorelbin 12.5 mg/m2 (day 1, 8, 15, 29, 36, 43) (P2) or carboplatin AUC1 (day 1-5, 29-33) and vinorelbin 12.5 mg/m2 (day 1, 8, 15, 29, 36, 43) (P3) or other CHT at the treating physician's discretion. (3) Results: Between 05/2009 and 11/2016, 205 patients were randomized and 172 included in the per-protocol analysis. Patients treated in P1 or P2 had a better overall survival (OS) compared to P3 (p = 0.015, p = 0.01, respectively). Patients treated with carboplatin had a worse OS compared to cisplatin (HR 1.78, p = 0.03), but the difference did not remain significant after adjusting for age, ECOG, cardiac function creatinine and completeness of CHT. (4) Conclusions: Carboplatin doublets show no significant difference compared to cisplatin, after adjusting for possibly relevant factors, probably due to existing selection bias.
RESUMEN
OBJECTIVE: In Berlin, the first public severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) testing site started 1 day after the first case in the city occurred. We describe epidemiological and clinical characteristics and aim at identifying risk factors for SARS-CoV-2 detection during the first 6 weeks of operation. METHODS: Testing followed national recommendations, but was also based on the physician's discretion. We related patient characteristics to SARS-CoV-2 test positivity for exploratory analyses using a cross-sectional, observational study design. RESULTS: Between 3 March and 13 April 2020, 5179 individuals attended the site (median age 34 years; interquartile range 26-47 years). The median time since disease onset was 4 days (interquartile range 2-7 days). Among 4333 persons tested, 333 (7.7%) were positive. Test positivity increased up to 10.3% (96/929) during the first 3 weeks and then declined, paralleling Germany's lock-down and the course of the epidemic in Berlin. Strict adherence to testing guidelines resulted in 10.4% (262/2530) test positivity, compared with 3.9% (71/1803) among individuals tested for other indications. A nightclub was a transmission hotspot; 27.7% (26/94) of one night's visitors were found positive. Smell and/or taste dysfunction indicated coronavirus disease 2019 (COVID-19) with 85.6% specificity (95% CI 82.1%-88.1%). Four per cent (14/333) of those infected were asymptomatic. Risk factors for detection of SARS-CoV-2 infection were recent contact with a positive case (second week after contact, OR 3.42; 95% CI 2.48-4.71), travel to regions of high pandemic activity (e.g. Austria, OR 4.16; 95% CI 2.48-6.99), recent onset of symptoms (second week, OR 3.61; 95% CI 1.87-6.98) and an impaired sense of smell/taste (4.08; 95% CI 2.36-7.03). CONCLUSIONS: In this young population, early-onset presentation of COVID-19 resembled flu-like symptoms, except for smell and/or taste dysfunction. Risk factors for SARS-CoV-2 detection were return from regions with high incidence and contact with confirmed SARS-CoV-2 cases, particularly when tests were administered within the first 2 weeks after contact and/or onset of symptoms.
Asunto(s)
Prueba de COVID-19/estadística & datos numéricos , COVID-19/epidemiología , Portador Sano/epidemiología , Adulto , Berlin/epidemiología , COVID-19/diagnóstico , COVID-19/fisiopatología , Prueba de COVID-19/métodos , Portador Sano/diagnóstico , Portador Sano/virología , Estudios Transversales , Femenino , Humanos , Incidencia , Masculino , Tamizaje Masivo/estadística & datos numéricos , Persona de Mediana Edad , Trastornos del Olfato/epidemiología , Trastornos del Olfato/virología , Pandemias/estadística & datos numéricos , Factores de Riesgo , Sensibilidad y Especificidad , Trastornos del Gusto/epidemiología , Trastornos del Gusto/virologíaRESUMEN
BACKGROUND: With increasingly precise radiotherapy and advanced medical imaging, the concept of radiotherapy target volume planning might be redefined with the aim of improving outcomes. We aimed to investigate whether target volume reduction is feasible and effective compared with conventional planning in the context of radical chemoradiotherapy for patients with locally advanced non-small-cell lung cancer. METHODS: We did a multicentre, open-label, randomised, controlled trial (PET-Plan; ARO-2009-09) in 24 centres in Austria, Germany, and Switzerland. Previously untreated patients (aged older than 18 years) with inoperable locally advanced non-small-cell lung cancer suitable for chemoradiotherapy and an Eastern Cooperative Oncology Group performance status of less than 3 were included. Undergoing 18F-fluorodeoxyglucose (18F-FDG) PET and CT for treatment planning, patients were randomly assigned (1:1) using a random number generator and block sizes between four and six to target volume delineation informed by 18F-FDG PET and CT plus elective nodal irradiation (conventional target group) or target volumes informed by PET alone (18F-FDG PET-based target group). Randomisation was stratified by centre and Union for International Cancer Control stage. In both groups, dose-escalated radiotherapy (60-74 Gy, 2 Gy per fraction) was planned to the respective target volumes and applied with concurrent platinum-based chemotherapy. The primary endpoint was time to locoregional progression from randomisation with the objective to test non-inferiority of 18F-FDG PET-based planning with a prespecified hazard ratio (HR) margin of 1·25. The per-protocol set was included in the primary analysis. The safety set included all patients receiving any study-specific treatment. Patients and study staff were not masked to treatment assignment. This study is registered with ClinicalTrials.gov, NCT00697333. FINDINGS: From May 13, 2009, to Dec 5, 2016, 205 of 311 recruited patients were randomly assigned to the conventional target group (n=99) or the 18F-FDG PET-based target group (n=106; the intention-to-treat set), and 172 patients were treated per protocol (84 patients in the conventional target group and 88 in the 18F-FDG PET-based target group). At a median follow-up of 29 months (IQR 9-54), the risk of locoregional progression in the 18F-FDG PET-based target group was non-inferior to, and in fact lower than, that in the conventional target group in the per-protocol set (14% [95% CI 5-21] vs 29% [17-38] at 1 year; HR 0·57 [95% CI 0·30-1·06]). The risk of locoregional progression in the 18F-FDG PET-based target group was also non-inferior to that in the conventional target group in the intention-to-treat set (17% [95% CI 9-24] vs 30% [20-39] at 1 year; HR 0·64 [95% CI 0·37-1·10]). The most common acute grade 3 or worse toxicity was oesophagitis or dysphagia (16 [16%] of 99 patients in the conventional target group vs 17 [16%] of 105 patients in the 18F-FDG PET-based target group); the most common late toxicities were lung-related (12 [12%] vs 11 [10%]). 20 deaths potentially related to study treatment were reported (seven vs 13). INTERPRETATION: 18F-FDG PET-based planning could potentially improve local control and does not seem to increase toxicity in patients with chemoradiotherapy-treated locally advanced non-small-cell lung cancer. Imaging-based target volume reduction in this setting is, therefore, feasible, and could potentially be considered standard of care. The procedures established might also support imaging-based target volume reduction concepts for other tumours. FUNDING: German Cancer Aid (Deutsche Krebshilfe).
Asunto(s)
Carcinoma de Pulmón de Células no Pequeñas/diagnóstico por imagen , Carcinoma de Pulmón de Células no Pequeñas/terapia , Quimioradioterapia , Fluorodesoxiglucosa F18 , Neoplasias Pulmonares/diagnóstico por imagen , Neoplasias Pulmonares/terapia , Tomografía Computarizada por Tomografía de Emisión de Positrones , Radiofármacos , Anciano , Carcinoma de Pulmón de Células no Pequeñas/patología , Femenino , Humanos , Neoplasias Pulmonares/patología , Masculino , Persona de Mediana Edad , Estadificación de NeoplasiasRESUMEN
Lung cancer is the most common fatal malignancy in adults worldwide, and non-small cell lung cancer (NSCLC) accounts for 85% of lung cancer diagnoses. Computed tomography (CT) is routinely used in clinical practice to determine lung cancer treatment and assess prognosis. Here, we developed LungNet, a shallow convolutional neural network for predicting outcomes of NSCLC patients. We trained and evaluated LungNet on four independent cohorts of NSCLC patients from four medical centers: Stanford Hospital (n = 129), H. Lee Moffitt Cancer Center and Research Institute (n = 185), MAASTRO Clinic (n = 311) and Charité - Universitätsmedizin (n=84). We show that outcomes from LungNet are predictive of overall survival in all four independent survival cohorts as measured by concordance indices of 0.62, 0.62, 0.62 and 0.58 on cohorts 1, 2, 3, and 4, respectively. Further, the survival model can be used, via transfer learning, for classifying benign vs malignant nodules on the Lung Image Database Consortium (n = 1010), with improved performance (AUC=0.85) versus training from scratch (AUC=0.82). LungNet can be used as a noninvasive predictor for prognosis in NSCLC patients and can facilitate interpretation of CT images for lung cancer stratification and prognostication.
RESUMEN
PURPOSE OR OBJECTIVE: A homogeneity index (HI) measures the uniformity of a dose distribution within a given target volume. Traditional HIs only use a limited number of dose-volume histogram data-points for calculation. A voxel-based homogeneity index (VHI) is proposed which utilizes the entire information of the three-dimensional dose distribution. We compared the VHI with existing HIs and analyzed if VHI results were associated with treatment outcomes in patients who underwent therapeutic WBRT. MATERIAL AND METHODS: The VHI analyzes deviations from the prescribed dose in each voxel of the target volume. We retrospectively analyzed WBRT treatment plans. Overall survival (OS), CNS progression-free-survival (CNS PFS) and hazard rates were compared for tertile-split levels of the VHI using the Kaplan-Meier methods and multivariable Cox-regression analysis. RESULTS: WBRT treatment plans (nâ¯=â¯770) were used for HIs comparison. OS and CNS PFS were assessed for 430 patients. The VHI showed a higher sensitivity for dose inhomogeneities. Lower OS and CNS PFS were observed for higher levels of VHIUnderdosage, particularly in patients with good performance status (KPSâ¯>70%) (OS: Log-rank Pâ¯=â¯.007, HRâ¯=â¯1.37 95%CI [1.09, 1.72]). CONCLUSION: Higher sensitivity and feasibility to assess treatment plan quality using the VHI were demonstrated. First clinical implications were found in terms of compromised OS/CNS PFS for WBRT with radiation underdosage.
Asunto(s)
Neoplasias Encefálicas/radioterapia , Irradiación Craneana/métodos , Adulto , Anciano , Anciano de 80 o más Años , Encéfalo/efectos de la radiación , Neoplasias Encefálicas/mortalidad , Neoplasias Encefálicas/secundario , Irradiación Craneana/mortalidad , Métodos Epidemiológicos , Femenino , Humanos , Masculino , Persona de Mediana Edad , Dosificación Radioterapéutica , Resultado del TratamientoRESUMEN
PURPOSE: Lung cancer remains the leading cause of cancer-related mortality worldwide. Stage III non-small cell lung cancer (NSCLC) includes heterogeneous presentation of the disease including lymph node involvement and large tumour volumes with infiltration of the mediastinum, heart or spine. In the treatment of stage III NSCLC an interdisciplinary approach including radiotherapy is considered standard of care with acceptable toxicity and improved clinical outcome concerning local control. Furthermore, gross tumour volume (GTV) changes during definitive radiotherapy would allow for adaptive replanning which offers normal tissue sparing and dose escalation. METHODS: A literature review was conducted to describe the predictive value of GTV changes during definitive radiotherapy especially focussing on overall survival. The literature search was conducted in a two-step review process using PubMed®/Medline® with the key words "stage III non-small cell lung cancer" and "radiotherapy" and "tumour volume" and "prognostic factors". RESULTS: After final consideration 17, 14 and 9 studies with a total of 2516, 784 and 639 patients on predictive impact of GTV, GTV changes and its impact on overall survival, respectively, for definitive radiotherapy for stage III NSCLC were included in this review. Initial GTV is an important prognostic factor for overall survival in several studies, but the time of evaluation and the value of histology need to be further investigated. GTV changes during RT differ widely, optimal timing for re-evaluation of GTV and their predictive value for prognosis needs to be clarified. The prognostic value of GTV changes is unclear due to varying study qualities, re-evaluation time and conflicting results. CONCLUSION: The main findings were that the clinical impact of GTV changes during definitive radiotherapy is still unclear due to heterogeneous study designs with varying quality. Several potential confounding variables were found and need to be considered for future studies to evaluate GTV changes during definitive radiotherapy with respect to treatment outcome.
Asunto(s)
Carcinoma de Pulmón de Células no Pequeñas/patología , Carcinoma de Pulmón de Células no Pequeñas/radioterapia , Neoplasias Pulmonares/patología , Neoplasias Pulmonares/radioterapia , Carga Tumoral/efectos de la radiación , Terapia Combinada , Humanos , Comunicación Interdisciplinaria , Colaboración Intersectorial , Metástasis Linfática/patología , Metástasis Linfática/radioterapia , Invasividad Neoplásica/patología , Estadificación de Neoplasias , PronósticoRESUMEN
BACKGROUND: Postoperative adjuvant radiation therapy (ART) in T3 and R1 prostate cancer as well as salvage radiation therapy (SRT) in case of postoperative biochemical failure (BF) are established treatments. Dose-intensified postoperative radiation therapy (RT) schemes have shown superior biochemical control accompanied by increased toxicity rates. In our study we evaluate a novel risk adapted dose-intensified postoperative RT scheme. METHODS: A consecutive series of prostate cancer patients receiving postoperative RT after radical prostatectomy using helical Tomotherapy between 04/2012 and 04/2015 was analyzed retrospectively. RT was administered using a simultaneous integrated boost (SIB) to the area at risk (37 fractions of 1.9 Gy, total dose: 70.3 Gy) being defined based on histopathological findings (T3/R1 region) and in few cases according to additional diagnostic imaging. The whole prostate bed was treated with a dose of 66.6 Gy (37 fractions of 1.8 Gy). Primary endpoints were acute and late genitourinary (GU) and gastrointestinal (GI) toxicities. Secondary endpoints included patient reported outcome as assessed by the International Prostate Symptom Score (IPSS), the International Consultation on Incontinence questionnaire (ICIQ) and prostate cancer specific Quality of Life questionnaire QLQ-PR25, as well as rates of BF. RESULTS: A total of 69 patients were analyzed. Sixteen patients underwent ART and 53 patients SRT, respectively. The median follow-up was 20 months (range, 8-41 months). Seven (10.1%) and four (5.8%) patients experienced acute grade 2 GU and GI toxicity. Two patients (2.9%) had late grade 2 GU toxicity, whereas no late grade 2 GI nor any grade 3 acute or late GU or GI events were observed. When compared to the baseline IPSS scores (p = 1.0) and ICIQ scores (p = 0.87) were not significantly different at the end of follow-up. Patient reported Quality of life (QoL) showed also no significant difference. A total of seven patients (10.1%) experienced a biochemical recurrence with the 2-year biochemical progression-free survival (bPFS) being 91%. CONCLUSIONS: Postoperative RT for prostate cancer patients with a risk adapted dose-intensified SIB using helical tomotherapy is feasible and associated with favorable acute and late GU and GI toxicity rates, no significant change of IPSS-, ICIQ scores and patient reported QoL and results in promising bPFS rates.
Asunto(s)
Neoplasias de la Próstata/radioterapia , Planificación de la Radioterapia Asistida por Computador/métodos , Radioterapia Adyuvante/métodos , Radioterapia de Intensidad Modulada/métodos , Anciano , Terapia Combinada/efectos adversos , Terapia Combinada/métodos , Supervivencia sin Enfermedad , Humanos , Masculino , Persona de Mediana Edad , Prostatectomía , Neoplasias de la Próstata/cirugía , Calidad de Vida , Dosificación Radioterapéutica , Radioterapia Adyuvante/efectos adversos , Radioterapia de Intensidad Modulada/efectos adversos , Estudios Retrospectivos , Terapia Recuperativa/efectos adversos , Terapia Recuperativa/métodosRESUMEN
OBJECTIVE: Current constraints aim to minimize the risk of radiation myelitis by the use of restrictive maximal spinal cord doses, commonly 50 Gy. However, several studies suggested that a dose-volume effect could exist. Based on these observations, we evaluated patients receiving potentially excessive doses to the spinal cord within minimal volumes. PATIENTS AND METHODS: Patients receiving radiotherapy between June 2010 and May 2015 using the NovalisTM (Varian, Palo Alto, CA, USA; Brainlab, Heimstetten, Germany) radiosurgery system were retrospectively analyzed. A total of 56 patients with 62 treated lesions that had been prescribed radiation doses close to the spinal cord potentially higher than the common 50 Gy 2Gy equivalent-dose (EQD2) constraint were selected for further analysis. Of these patients, 26 with 31 lesions had no history of previous irradiation, while 30 patients with 31 lesions had been previously irradiated within the treatment field. RESULTS: According to different dose evaluation approaches (spinal canal, spinal cord contour), 16 and 10 out of 31 primary irradiated lesions infringed constraints. For the 16 lesions violating spinal canal doses, the maximum doses ranged from 50.5 to 61.9 Gy EQD2. Reirradiated lesions had an average and median cumulative dose of 70.5 and 69 Gy, respectively. Dose drop-off was steep in both groups. Median overall survival was 17 months. No radiation myelitis or radiomorphological alterations were observed during follow-up. CONCLUSION: This study adds to the increasing body of evidence indicating that excessive spinal cord doses within a minimal volume, especially in a reirradiation setting with topographically distinct high-point doses, may be given to patients after careful evaluation of treatment- and tumor-associated risks.