Automatic Diagnosis of Hepatocellular Carcinoma and Metastases Based on Computed Tomography Images.

Liver cancer, a leading cause of cancer mortality, is often diagnosed by analyzing the grayscale variations in liver tissue across different computed tomography (CT) images. However, the intensity similarity can be strong, making it difficult for radiologists to visually identify hepatocellular carcinoma (HCC) and metastases. It is crucial for the management and prevention strategies to accurately differentiate between these two liver cancers. This study proposes an automated system using a convolutional neural network (CNN) to enhance diagnostic accuracy to detect HCC, metastasis, and healthy liver tissue. This system incorporates automatic segmentation and classification. The liver lesions segmentation model is implemented using residual attention U-Net. A 9-layer CNN classifier implements the lesions classification model. Its input is the combination of the results of the segmentation model with original images. The dataset included 300 patients, with 223 used to develop the segmentation model and 77 to test it. These 77 patients also served as inputs for the classification model, consisting of 20 HCC cases, 27 with metastasis, and 30 healthy. The system achieved a mean Dice score of 87.65 % in segmentation and a mean accuracy of 93.97 % in classification, both in the test phase. The proposed method is a preliminary study with great potential in helping radiologists diagnose liver cancers.

Neurological hospitalisations in childhood cancer survivors treated before 2001: findings from the French Childhood Cancer Survivor Study cohort.

PURPOSE: Childhood cancer survivors (CCS) have an increased risk of developing late chronic diseases, which can be influenced by the cancer type and its treatment. These chronic diseases can be severe and disabling, typically emerging years to decades after treatment. These deficits negatively impact quality of life, intelligence quotient, and memory. This study investigated how much the cancer type and treatment could affect the neurological hospitalisations in the French Childhood Cancer Survivors Study (FCCSS). METHODS: We included 5579 childhood cancer survivors (CCS), diagnosed with solid tumours or lymphoma between 1945 and 2000, treated before 2001 and below the age of 21 years at initial treatment. The follow-up period was from 2006 to 2018. Hospitalisation data were obtained by linkage with the National Health Data System. We calculated the relative hospitalisation rate (RHRs) and absolute excess rate (AERs). Multivariable analyses were conducted using a Generalized Linear Model (GLM) with a Poisson distribution to estimate the association between neurological hospitalisation and patient characteristics. The expected number of hospitalisations served as an offset to compare the risk for FCCSS survivors with that of the reference population. Risk estimates were reported as relative risk (RR) with 95% confidence intervals. RESULTS: The hospitalisation rate for CCS was 114.2 per 10,000 person-years (PY), compared to 48.4 in the reference population. The highest hospitalisation rates were observed for epilepsy (AER = 27.1 per 10000 PY, 95%CI: 23.5-31.2 and RHR = 5.1, 95%CI 4.4-5.7). In multivariable analyses, central nervous system (CNS) tumours survivors had the highest relative risk (RR) of hospitalisation (RR = 9.4, 95%CI: 6.7-13.1) followed by neuroblastoma survivors (RR = 2.5, 95%CI: 1.7-3.7). In the whole population, survivors who received radiation to the head and neck had a significantly higher risk of hospitalisation (RR = 3.9, 95%CI: 3.3-4.7) compared to those who did not receive radiotherapy. CONCLUSIONS: Head and neck irradiation was identified as a strong risk factor for hospitalisation. This underlines the importance of implementing specific neurologic surveillance programs for at-risk individuals.

Genetic Risk of Second Malignant Neoplasm after Childhood Cancer Treatment: A Systematic Review.

Second malignant neoplasm (SMN) is one of the most severe long-term risks for childhood cancer survivors (CCS), significantly impacting long-term patient survival. While radiotherapy and chemotherapy are known risk factors, the observed inter-individual variability suggests a genetic component contributing to the risk of SMN. This article aims to conduct a systematic review of genetic factors implicated in the SMN risk among CCS. Searches were performed in PubMed, Scopus, and Web of Sciences. Eighteen studies were included (eleven candidate gene studies, three genome-wide association studies, and four whole exome/genome sequencing studies). The included studies were based on different types of first cancers, investigated any or specific types of SMN, and focused mainly on genes involved in drug metabolism and DNA repair pathways. These differences in study design and methods used to characterize genetic variants limit the scope of the results and highlight the need for further extensive and standardized investigations. However, this review provides a valuable compilation of SMN risk-associated variants and genes, facilitating efficient replication and advancing our understanding of the genetic basis for this major risk for CCS.

Radiation Doses Received by Major Organs at Risk in Children and Young Adolescents Treated for Cancer with External Beam Radiation Therapy: A Large-scale Study from 12 European Countries.

PURPOSE: Childhood cancer survivors, in particular those treated with radiation therapy, are at high risk of long-term iatrogenic events. The prediction of risk of such events is mainly based on the knowledge of the radiation dose received to healthy organs and tissues during treatment of childhood cancer diagnosed decades ago. We aimed to set up a standardized organ dose table to help former patients and clinicians in charge of long-term follow-up clinics. METHODS AND MATERIALS: We performed whole body dosimetric reconstruction for 2646 patients from 12 European countries treated between 1941 and 2006 (median, 1976). Most plannings were 2- or 3-dimensional. A total of 46% of patients were treated using Cobalt 60, and 41%, using a linear accelerator. The median prescribed dose was 27.2 Gy (IQ1-IQ3, 17.6-40.0 Gy). A patient-specific voxel-based anthropomorphic phantom with more than 200 anatomic structures or substructures delineated as a surrogate of each subject's anatomy was used. The radiation therapy was simulated with a treatment planning system based on available treatment information. The radiation dose received by any organ of the body was estimated by extending the treatment planning system dose calculation to the whole body, by type and localization of childhood cancer. RESULTS: The integral dose and normal tissue doses to most of the 23 considered organs increased between the 1950s and 1970s and decreased or plateaued thereafter. Whatever the organ considered, the type of childhood cancer explained most of the variability in organ dose. The country of treatment explained only a small part of the variability. CONCLUSIONS: The detailed dose estimates provide very useful information for former patients or clinicians who have only limited knowledge about radiation therapy protocols or techniques, but who know the type and site of childhood cancer, sex, age, and year of treatment. This will allow better prediction of the long-term risk of iatrogenic events and better referral to long-term follow-up clinics.

Deep-Learning for Rapid Estimation of the Out-of-Field Dose in External Beam Photon Radiation Therapy - A Proof of Concept.

PURPOSE: The dose deposited outside of the treatment field during external photon beam radiation therapy treatment, also known as out-of-field dose, is the subject of extensive study as it may be associated with a higher risk of developing a second cancer and could have deleterious effects on the immune system that compromise the efficiency of combined radio-immunotherapy treatments. Out-of-field dose estimation tools developed today in research, including Monte Carlo simulations and analytical methods, are not suited to the requirements of clinical implementation because of their lack of versatility and their cumbersome application. We propose a proof of concept based on deep learning for out-of-field dose map estimation that addresses these limitations. METHODS AND MATERIALS: For this purpose, a 3D U-Net, considering as inputs the in-field dose, as computed by the treatment planning system, and the patient's anatomy, was trained to predict out-of-field dose maps. The cohort used for learning and performance evaluation included 3151 pediatric patients from the FCCSS database, treated in 5 clinical centers, whose whole-body dose maps were previously estimated with an empirical analytical method. The test set, composed of 433 patients, was split into 5 subdata sets, each containing patients treated with devices unseen during the training phase. Root mean square deviation evaluated only on nonzero voxels located in the out-of-field areas was computed as performance metric. RESULTS: Root mean square deviations of 0.28 and 0.41 cGy/Gy were obtained for the training and validation data sets, respectively. Values of 0.27, 0.26, 0.28, 0.30, and 0.45 cGy/Gy were achieved for the 6 MV linear accelerator, 16 MV linear accelerator, Alcyon cobalt irradiator, Mobiletron cobalt irradiator, and betatron device test sets, respectively. CONCLUSIONS: This proof-of-concept approach using a convolutional neural network has demonstrated unprecedented generalizability for this task, although it remains limited, and brings us closer to an implementation compatible with clinical routine.

Very long-term outcomes of pediatric patients treated for optic pathway gliomas: A longitudinal cohort study.

BACKGROUND: Optic pathway gliomas (OPGs) represent 5% of childhood brain tumors. Successive relapses lead to multiple treatments exposing to late complications. METHODS: We included patients treated at Gustave Roussy (GR) between January 1980 and December 2015 for OPG, before 18 years old and alive at 5 years from diagnosis. Mortality and physical health conditions data were extracted from medical data files and updated, thanks to the GR long-term follow-up program and French national mortality registry for patients included in the French Childhood Cancer Survivor Study. RESULTS: We included 182 5-year OPG-childhood survivors in the analysis (sex ratio M/F 0.8, 35% with neurofibromatosis type 1 [NF1]). With a median follow-up of 17.2 years (range = 5-41), we registered 82 relapses, 9 second malignancies, and 15 deaths as first events after 5 years, resulting in 20-year conditional overall survival (C-OS) and late events-free survival of 79.9% (95% confidence interval [CI] = 71-86) and 43.5% (95% CI = 36-51), respectively. Radiotherapy exposure in NF1 patients (hazard ratio [HR] = 6, 95% CI = 1.7-21.2) and hypothalamic involvement (HR = 3.2, 95% CI = 1.4-7.3) were significantly associated with C-OS in multivariable analyses. Ninety-five percent of 5-year OPG survivors suffered from any health condition, especially visual acuity "<1/10" (n = 109), pituitary deficiency (n = 106), and neurocognitive impairment (n = 89). NF1 (HR 2.1) was associated with precocious puberty. With a median time post-diagnosis of 4.2 years, 33 cerebrovascular events were observed in 21 patients. CONCLUSIONS: Late relapses, second malignancies, and cerebrovascular diseases are severe late events resulting in premature mortality. Morbidity is high and needs after-cancer care to improve quality of life. Risk factors could be considered to better stratify long-term follow-up.

Radiomics-Based Classification of Tumor and Healthy Liver on Computed Tomography Images.

Liver malignancies, particularly hepatocellular carcinoma and metastasis, stand as prominent contributors to cancer mortality. Much of the data from abdominal computed tomography images remain underused by radiologists. This study explores the application of machine learning in differentiating tumor tissue from healthy liver tissue using radiomics features. Preoperative contrast-enhanced images of 94 patients were used. A total of 1686 features classified as first-order, second-order, higher-order, and shape statistics were extracted from the regions of interest of each patient's imaging data. Then, the variance threshold, the selection of statistically significant variables using the Student's t-test, and lasso regression were used for feature selection. Six classifiers were used to identify tumor and non-tumor liver tissue, including random forest, support vector machines, naive Bayes, adaptive boosting, extreme gradient boosting, and logistic regression. Grid search was used as a hyperparameter tuning technique, and a 10-fold cross-validation procedure was applied. The area under the receiver operating curve (AUROC) assessed the performance. The AUROC scores varied from 0.5929 to 0.9268, with naive Bayes achieving the best score. The radiomics features extracted were classified with a good score, and the radiomics signature enabled a prognostic biomarker for hepatic tumor screening.

Risk of subsequent gliomas and meningiomas among 69,460 5-year survivors of childhood and adolescent cancer in Europe: the PanCareSurFup study.

BACKGROUND: Childhood cancer survivors are at risk of subsequent gliomas and meningiomas, but the risks beyond age 40 years are uncertain. We quantified these risks in the largest ever cohort. METHODS: Using data from 69,460 5-year childhood cancer survivors (diagnosed 1940-2008), across Europe, standardized incidence ratios (SIRs) and cumulative incidence were calculated. RESULTS: In total, 279 glioma and 761 meningioma were identified. CNS tumour (SIR: 16.2, 95% CI: 13.7, 19.2) and leukaemia (SIR: 11.2, 95% CI: 8.8, 14.2) survivors were at greatest risk of glioma. The SIR for CNS tumour survivors was still 4.3-fold after age 50 (95% CI: 1.9, 9.6), and for leukaemia survivors still 10.2-fold after age 40 (95% CI: 4.9, 21.4). Following cranial radiotherapy (CRT), the cumulative incidence of a glioma in CNS tumour survivors was 2.7%, 3.7% and 5.0% by ages 40, 50 and 60, respectively, whilst for leukaemia this was 1.2% and 1.7% by ages 40 and 50. The cumulative incidence of a meningioma after CRT in CNS tumour survivors doubled from 5.9% to 12.5% between ages 40 and 60, and in leukaemia survivors increased from 5.8% to 10.2% between ages 40 and 50. DISCUSSION: Clinicians following up survivors should be aware that the substantial risks of meningioma and glioma following CRT are sustained beyond age 40 and be vigilant for symptoms.

Cumulative Absolute Risk of Subsequent Colorectal Cancer After Abdominopelvic Radiotherapy Among Childhood Cancer Survivors: A PanCareSurFup Study.

PURPOSE: Childhood cancer survivors are at the risk of developing subsequent colorectal cancers (CRCs), but the absolute risks by treatment modality are uncertain. We quantified the absolute risks by radiotherapy treatment characteristics using clinically accessible data from a Pan-European wide case-control study nested within a large cohort of childhood cancer survivors: the PanCareSurFup Study. METHODS: Odds ratios (ORs) from a case-control study comprising 143 CRC cases and 143 controls nested within a cohort of 69,460 survivors were calculated. These, together with standardized incidence ratios for CRC for this cohort and European general population CRC incidence rates and survivors' mortality rates, were used to estimate cumulative absolute risks (CARs) by attained age for different categories of radiation to the abdominopelvic area. RESULTS: Overall, survivors treated with abdominopelvic radiotherapy treatment (ART) were three times more likely to develop a subsequent CRC than those who did not receive ART (OR, 3.1 [95% CI, 1.4 to 6.6]). For male survivors treated with ART, the CAR was 0.27% (95% CI, 0.17 to 0.59) by age 40 years, 1.08% (95% CI, 0.69 to 2.34) by age 50 years (0.27% expected in the general population), and 3.7% (95% CI, 2.36 to 7.80) by age 60 years (0.95% expected). For female survivors treated with ART, the CAR was 0.29% (95% CI, 0.18 to 0.62) by age 40 years, 1.03% (95% CI, 0.65 to 2.22) by age 50 years (0.27% expected), and 3.0% (95% CI, 1.91 to 6.37) by age 60 years (0.82% expected). CONCLUSION: We demonstrated that by age 40 years survivors of childhood cancer treated with ART already have a similar risk of CRC as those age 50 years in the general population for whom population-based CRC screening begins in many countries. This information should be used in the development of survivorship guidelines for the risk stratification of survivors concerning CRC risk.

Excess healthcare expenditure in adults treated for solid cancer in childhood: a cohort study in France.

BACKGROUND: Due to late effects, childhood cancer survivors (CCS) are more likely to have multiple chronic conditions than the general population. However, little is known about the economic burden of care of CCS in the long term. OBJECTIVES: To estimate excess healthcare expenditure for long-term CCS in France compared to the general population and to investigate the associated factors. METHODS: We included 5353 5-year solid CCS diagnosed before the age of 21 years before 2000 from the French CCS cohort and obtained a random reference sample from the general population for each CCS, matched on age, gender and region of residence. We used the French national health data system to estimate annual healthcare expenditure between 2011 and 2018 for CCS and the reference sample, and computed the excess as the net difference between CCS expenditure and the median expenditure of the reference sample. We used repeated-measures linear models to estimate associations between excess healthcare expenditure and CCS characteristics. RESULTS: Annual mean (95% CI) excess healthcare expenditure was €3920 (3539; 4301), mainly for hospitalization (39.6%) and pharmacy expenses (17%). Higher excess was significantly associated with having been treated before the 1990s and having survived a central nervous system tumor, whereas lower excess was associated with CCS who had not received treatment with radiotherapy. CONCLUSIONS: Of the variables that influence excess healthcare expenditure, a lever for action is the type of treatment administered. Future research should focus on addressing the long-term cost-effectiveness of new approaches, especially those related to radiotherapy.

Investigation of common genetic risk factors between thyroid traits and breast cancer.

Breast cancer (BC) risk is suspected to be linked to thyroid disorders, however observational studies exploring the association between BC and thyroid disorders gave conflicting results. We proposed an alternative approach by investigating the shared genetic risk factors between BC and several thyroid traits. We report a positive genetic correlation between BC and thyroxine (FT4) levels (corr = 0.13, p-value = 2.0 × 10-4) and a negative genetic correlation between BC and thyroid-stimulating hormone (TSH) levels (corr = -0.09, p-value = 0.03). These associations are more striking when restricting the analysis to estrogen receptor-positive BC. Moreover, the polygenic risk scores (PRS) for FT4 and hyperthyroidism are positively associated to BC risk (OR = 1.07, 95%CI: 1.00-1.13, p-value = 2.8 × 10-2 and OR = 1.04, 95%CI: 1.00-1.08, p-value = 3.8 × 10-2, respectively), while the PRS for TSH is inversely associated to BC risk (OR = 0.93, 95%CI: 0.89-0.97, p-value = 2.0 × 10-3). Using the PLACO method, we detected 49 loci associated to both BC and thyroid traits (p-value < 5 × 10-8), in the vicinity of 130 genes. An additional colocalization and gene-set enrichment analyses showed a convincing causal role for a known pleiotropic locus at 2q35 and revealed an additional one at 8q22.1 associated to both BC and thyroid cancer. We also found two new pleiotropic loci at 14q32.33 and 17q21.31 that were associated to both TSH levels and BC risk. Enrichment analyses and evidence of regulatory signals also highlighted brain tissues and immune system as candidates for obtaining associations between BC and TSH levels. Overall, our study sheds light on the complex interplay between BC and thyroid traits and provides evidence of shared genetic risk between those conditions.

Subsequent female breast cancer risk associated with anthracycline chemotherapy for childhood cancer.

Anthracycline-based chemotherapy is associated with increased subsequent breast cancer (SBC) risk in female childhood cancer survivors, but the current evidence is insufficient to support early breast cancer screening recommendations for survivors treated with anthracyclines. In this study, we pooled individual patient data of 17,903 survivors from six well-established studies, of whom 782 (4.4%) developed a SBC, and analyzed dose-dependent effects of individual anthracycline agents on developing SBC and interactions with chest radiotherapy. A dose-dependent increased SBC risk was seen for doxorubicin (hazard ratio (HR) per 100 mg m-2: 1.24, 95% confidence interval (CI): 1.18-1.31), with more than twofold increased risk for survivors treated with ≥200 mg m-2 cumulative doxorubicin dose versus no doxorubicin (HR: 2.50 for 200-299 mg m-2, HR: 2.33 for 300-399 mg m-2 and HR: 2.78 for ≥400 mg m-2). For daunorubicin, the associations were not statistically significant. Epirubicin was associated with increased SBC risk (yes/no, HR: 3.25, 95% CI: 1.59-6.63). For patients treated with or without chest irradiation, HRs per 100 mg m-2 of doxorubicin were 1.11 (95% CI: 1.02-1.21) and 1.26 (95% CI: 1.17-1.36), respectively. Our findings support that early initiation of SBC surveillance may be reasonable for survivors who received ≥200 mg m-2 cumulative doxorubicin dose and should be considered in SBC surveillance guidelines for survivors and future treatment protocols.

Radiation exposure and leukaemia risk among cohorts of persons exposed to low and moderate doses of external ionising radiation in childhood.

BACKGROUND: Many high-dose groups demonstrate increased leukaemia risks, with risk greatest following childhood exposure; risks at low/moderate doses are less clear. METHODS: We conducted a pooled analysis of the major radiation-associated leukaemias (acute myeloid leukaemia (AML) with/without the inclusion of myelodysplastic syndrome (MDS), chronic myeloid leukaemia (CML), acute lymphoblastic leukaemia (ALL)) in ten childhood-exposed groups, including Japanese atomic bomb survivors, four therapeutically irradiated and five diagnostically exposed cohorts, a mixture of incidence and mortality data. Relative/absolute risk Poisson regression models were fitted. RESULTS: Of 365 cases/deaths of leukaemias excluding chronic lymphocytic leukaemia, there were 272 AML/CML/ALL among 310,905 persons (7,641,362 person-years), with mean active bone marrow (ABM) dose of 0.11 Gy (range 0-5.95). We estimated significant (P < 0.005) linear excess relative risks/Gy (ERR/Gy) for: AML (n = 140) = 1.48 (95% CI 0.59-2.85), CML (n = 61) = 1.77 (95% CI 0.38-4.50), and ALL (n = 71) = 6.65 (95% CI 2.79-14.83). There is upward curvature in the dose response for ALL and AML over the full dose range, although at lower doses (<0.5 Gy) curvature for ALL is downwards. DISCUSSION: We found increased ERR/Gy for all major types of radiation-associated leukaemia after childhood exposure to ABM doses that were predominantly (for 99%) <1 Gy, and consistent with our prior analysis focusing on <100 mGy.

Surveillance after childhood cancer: are survivors with an increased risk for cardiomyopathy regularly followed-up?

BACKGROUND: We aimed to study adherence to cardiac screening in long-term childhood cancer survivors (CCS) at high risk of cardiomyopathy. METHODS: This study involved 976 5-year CCS at high risk for cardiomyopathy from the French Childhood Cancer Survivor Study. Determinants of adherence to recommended surveillance were studied using multivariable logistic regression models. Association of attendance to a long-term follow-up (LTFU) visit with completion of an echocardiogram was estimated using a Cox regression model. RESULTS: Among participants, 32% had an echocardiogram within the 5 previous years. Males (adjusted RR [aRR] 0.71, 95% CI 0.58-0.86), survivors aged 36-49 (aRR 0.79, 95% CI 0.64-0.98), Neuroblastoma (aRR 0.53, 95% CI 0.30-0.91) and CNS tumour survivors (aRR 0.43, 95% CI 0.21-0.89) were less likely to adhere to recommended surveillance. Attendance to an LTFU visit was associated with completion of an echocardiogram in patients who were not previously adherent to recommendations (HR 8.20, 95% CI 5.64-11.93). CONCLUSIONS: The majority of long-term survivors at high risk of cardiomyopathy did not adhere to the recommended surveillance. Attendance to an LTFU visit greatly enhanced the completion of echocardiograms, but further interventions need to be developed to reach more survivors.

Predictive factors of long-term follow-up attendance in very long-term childhood cancer survivors.

BACKGROUND: Long-term follow-up (LTFU) clinics have been developed but only some childhood cancer survivors (CCS) attend long-term follow-up (LTFU). OBJECTIVE: To identify factors that influence LTFU attendance. METHODS: Five-year CCS treated for a solid tumor or lymphoma in Gustave Roussy before 2000, included in the FCCSS cohort (French Childhood Cancer Survivor Study), aged >18 years and alive at the date of the LTFU Clinic opening (January 2012) were invited to a LTFU visit. Factors associated with attendance at the LTFU clinic between 2012 and 2020 were estimated using logistic regression analyses. Analyses included different types of factors: clinical (tumor characteristics, cancer treatments, late effects), medical (medical expenses were used as a proxy of survivor's health status), social (deprivation index based on census-tract data relating to income, educational level, proportion of blue-collar workers, and unemployed people living in the area of residence), and spatial (distance to the LTFU clinic). RESULTS: Among 2341 CCS contacted (55% males, mean age at study, 45 years; SD ± 10 years; mean age at diagnosis, 6 years; SD ± 5 years), 779 (33%) attended at least one LTFU visit. Initial cancer-related factors associated with LTFU visit attendance were: treatment with both radiotherapy and chemotherapy (odds ratio [OR], 4.02; 95% CI, 2.11-7.70), bone sarcoma (OR, 2.43; 95% CI, 1.56-3.78), central nervous system primitive tumor (OR, 1.65; 95% CI, 1.02-2.67), and autologous hematopoietic cell transplant (OR, 2.07; 95% CI, 1.34-3.20). Late effects (OR, 1.70; 95% CI, 1.31-2.20), highest medical expenses (OR, 1.65; 95% CI, 1.22-2.22), living in the most advantaged area (OR vs. the most deprived area = 1.60; 95% CI, 1.15-2.22), and shorter distance from LTFU care center (<12 miles) also increased attendance. CONCLUSIONS: Patients who are apparently healthy as well as socially disadvantaged and living far away from the center are less likely to attend LTFU care. PLAIN LANGUAGE SUMMARY: Among 2341 adult childhood cancer survivors contacted between 2012 and 2020, 33% attended at least one long-term follow-up visit. Clinical factors related to attendance were multimodal treatment of first cancer (combining chemotherapy and radiotherapy), stem cell transplant, type of diagnosis (bone tumor and central nervous system primitive tumor), late effects (at least one disease among second malignancy, heart disease, or stroke), and highest medical expenses. In addition, the study identified social and spatial inequalities related to attendance, with independent negative effects of distance and social deprivation on attendance, even though the medical costs related to the long-term follow-up examinations are covered by the French social security system.

Radiation-induced lung injury after breast cancer treatment: incidence in the CANTO-RT cohort and associated clinical and dosimetric risk factors.

Purpose: Radiation-induced lung injury (RILI) is strongly associated with various clinical conditions and dosimetric parameters. Former studies have led to reducing radiotherapy (RT) doses to the lung and have favored the discontinuation of tamoxifen during RT. However, the monocentric design and variability of dosimetric parameters chosen have limited further improvement. The aim of our study was to assess the incidence of RILI in current practice and to determine clinical and dosimetric risk factors associated with RILI occurrence. Material and methods: Data from 3 out of the 10 top recruiting centers in CANTO-RT, a subset of the CANTO prospective longitudinal cohort (NCT01993498), were retrospectively analyzed for RILI occurrence. This cohort, which recruited invasive cT0-3 cN0-3 M0 breast cancer patients from 2012 to 2018, prospectively recorded the occurrence of adverse events by questionnaires and medical visits at the end of, and up to 60 months after treatment. RILI adverse events were defined in all patients by the association of clinical symptoms and compatible medical imaging. Results: RILI was found in 38/1565 (2.4%) patients. Grade II RILI represented 15/38 events (39%) and grade III or IV 2/38 events (6%). There were no grade V events. The most frequently used technique for treatment was 3D conformational RT (96%). In univariable analyses, we confirmed the association of RILI occurrence with pulmonary medical history, absence of cardiovascular disease medical history, high pT and pN, chemotherapy use, nodal RT. All dosimetric parameters were highly correlated and had close predictive value. In the multivariable analysis adjusted for chemotherapy use and nodal involvement, pulmonary medical history (OR=3.05, p<0.01) and high V30 Gy (OR=1.06, p=0.04) remained statistically significant risk factors for RILI occurrence. V30 Gy >15% was significantly associated with RILI occurrence in a multivariable analysis (OR=3.07, p=0.03). Conclusion: Our study confirms the pulmonary safety of breast 3D RT in CANTO-RT. Further analyses with modern radiation therapy techniques such as IMRT are needed. Our results argue in favor of a dose constraint to the ipsilateral lung using V30 Gy not exceeding 15%, especially in patients presenting pulmonary medical history. Pulmonary disease records should be taken into account for RT planning.

Individual Participant Data Network Meta-Analysis of Neoadjuvant Chemotherapy or Chemoradiotherapy in Esophageal or Gastroesophageal Junction Carcinoma.

PURPOSE: The optimal neoadjuvant treatment for resectable carcinoma of the thoracic esophagus (TE) or gastroesophageal junction (GEJ) remains a matter of debate. We performed an individual participant data (IPD) network meta-analysis (NMA) of randomized controlled trials (RCTs) to study the effect of chemotherapy or chemoradiotherapy, with a focus on tumor location and histology subgroups. PATIENTS AND METHODS: All, published or unpublished, RCTs closed to accrual before December 31, 2015 and having compared at least two of the following strategies were eligible: upfront surgery (S), chemotherapy followed by surgery (CS), and chemoradiotherapy followed by surgery (CRS). All analyses were conducted on IPD obtained from investigators. The primary end point was overall survival (OS). The IPD-NMA was analyzed by a one-step mixed-effect Cox model adjusted for age, sex, tumor location, and histology. The NMA was registered in PROSPERO (CRD42018107158). RESULTS: IPD were obtained for 26 of 35 RCTs (4,985 of 5,807 patients) corresponding to 12 comparisons for CS-S, 12 for CRS-S, and four for CRS-CS. CS and CRS led to increased OS when compared with S with hazard ratio (HR) = 0.86 (0.75 to 0.99), P = .03 and HR = 0.77 (0.68 to 0.87), P < .001 respectively. The NMA comparison of CRS versus CS for OS gave a HR of 0.90 (0.74 to 1.09), P = .27 (consistency P = .26, heterogeneity P = .0038). For CS versus S, a larger effect on OS was observed for GEJ versus TE tumors (P = .036). For the CRS versus S and CRS versus CS, a larger effect on OS was observed for women (P = .003, .012, respectively). CONCLUSION: Neoadjuvant chemotherapy and chemoradiotherapy were consistently better than S alone across histology, but with some variation in the magnitude of treatment effect by sex for CRS and tumor location for CS. A strong OS difference between CS and CRS was not identified.

Early Development of Atherosclerotic Plaques in the Coronary Arteries after Radiotherapy for Breast Cancer (BACCARAT Study).

Background-Radiotherapy (RT) for breast cancer (BC) can lead to an increased risk of coronary artery disease several years after RT. The aim of this study was to evaluate the development of overall, non-calcified and calcified atherosclerotic plaques over 2 years after BC for RT and associations with cardiac exposure. Methods-The study included 101 left- or right-sided BC patients treated with RT without chemotherapy. A coronary CT angiography was performed before and 2 years after RT. Plaque development thorough the entire coronary network was defined as an increased number of plaques. Cardiac exposure was quantified with mean doses to the heart, left ventricle, and coronary arteries. Logistic regression models were used to assess association with doses. Results-At inclusion, 37% of patients had plaques, increasing to 42% two years after RT. Overall plaque development was observed in seven patients: five with calcified plaque development and four with non-calcified plaque development. The risk of overall plaque development was significantly associated with doses to the Left Main and Circumflex coronary arteries (OR at 1 Gy = 2.32, p = 0.03 and OR at 1 Gy = 2.27, p = 0.03, respectively). Specific analyses for calcified and non-calcified plaque development showed similar results. Conclusion-Our study suggests an association between coronary arteries exposure and the risk of developing both calcified and non-calcified atherosclerotic plaques over 2 years after BC RT. Trial registration number: NCT02605512.

Dosiomics-Based Prediction of Radiation-Induced Valvulopathy after Childhood Cancer.

Valvular Heart Disease (VHD) is a known late complication of radiotherapy for childhood cancer (CC), and identifying high-risk survivors correctly remains a challenge. This paper focuses on the distribution of the radiation dose absorbed by heart tissues. We propose that a dosiomics signature could provide insight into the spatial characteristics of the heart dose associated with a VHD, beyond the already-established risk induced by high doses. We analyzed data from the 7670 survivors of the French Childhood Cancer Survivors' Study (FCCSS), 3902 of whom were treated with radiotherapy. In all, 63 (1.6%) survivors that had been treated with radiotherapy experienced a VHD, and 57 of them had heterogeneous heart doses. From the heart-dose distribution of each survivor, we extracted 93 first-order and spatial dosiomics features. We trained random forest algorithms adapted for imbalanced classification and evaluated their predictive performance compared to the performance of standard mean heart dose (MHD)-based models. Sensitivity analyses were also conducted for sub-populations of survivors with spatially heterogeneous heart doses. Our results suggest that MHD and dosiomics-based models performed equally well globally in our cohort and that, when considering the sub-population having received a spatially heterogeneous dose distribution, the predictive capability of the models is significantly improved by the use of the dosiomics features. If these findings are further validated, the dosiomics signature may be incorporated into machine learning algorithms for radiation-induced VHD risk assessment and, in turn, into the personalized refinement of follow-up guidelines.

Genetic factors for differentiated thyroid cancer in French Polynesia: new candidate loci.

Background: Populations of French Polynesia (FP), where France performed atmospheric tests between 1966 and 1974, experience a high incidence of differentiated thyroid cancer (DTC). However, up to now, no sufficiently large study of DTC genetic factors in this population has been performed to reach definitive conclusion. This research aimed to analyze the genetic factors of DTC risk among the native FP populations. Methods: We analyzed more than 300 000 single nucleotide polymorphisms (SNPs) genotyped in 283 DTC cases and 418 matched controls born in FP, most being younger than 15 years old at the time of the first nuclear tests. We analyzed the genetic profile of our cohort to identify population subgroups. We then completed a genome-wide analysis study on the whole population. Results: We identified a specific genetic structure in the FP population reflecting admixture from Asian and European populations. We identified three regions associated with increased DTC risk at 6q24.3, 10p12.2, and 17q21.32. The lead SNPs at these loci showed respective p-values of 1.66 × 10-7, 2.39 × 10-7, and 7.19 × 10-7 and corresponding odds ratios of 2.02, 1.89, and 2.37. Conclusion: Our study results suggest a role of the loci 6q24.3, 10p12.2 and 17q21.32 in DTC risk. However, a whole genome sequencing approach would be better suited to characterize these factors than genotyping with microarray chip designed for the Caucasian population. Moreover, the functional impact of these three new loci needs to be further explored and validated.