Automatic gross tumor volume segmentation with failure detection for safe implementation in locally advanced cervical cancer.

Background and Purpose: Automatic segmentation methods have greatly changed the RadioTherapy (RT) workflow, but still need to be extended to target volumes. In this paper, Deep Learning (DL) models were compared for Gross Tumor Volume (GTV) segmentation in locally advanced cervical cancer, and a novel investigation into failure detection was introduced by utilizing radiomic features. Methods and materials: We trained eight DL models (UNet, VNet, SegResNet, SegResNetVAE) for 2D and 3D segmentation. Ensembling individually trained models during cross-validation generated the final segmentation. To detect failures, binary classifiers were trained using radiomic features extracted from segmented GTVs as inputs, aiming to classify contours based on whether their Dice Similarity Coefficient ( DSC ) < T and DSC ⩾ T . Two distinct cohorts of T2-Weighted (T2W) pre-RT MR images captured in 2D sequences were used: one retrospective cohort consisting of 115 LACC patients from 30 scanners, and the other prospective cohort, comprising 51 patients from 7 scanners, used for testing. Results: Segmentation by 2D-SegResNet achieved the best DSC, Surface DSC ( SDSC 3 mm ), and 95th Hausdorff Distance (95HD): DSC = 0.72 ± 0.16, SDSC 3 mm =0.66 ± 0.17, and 95HD = 14.6 ± 9.0 mm without missing segmentation ( M =0) on the test cohort. Failure detection could generate precision ( P = 0.88 ), recall ( R = 0.75 ), F1-score ( F = 0.81 ), and accuracy ( A = 0.86 ) using Logistic Regression (LR) classifier on the test cohort with a threshold T = 0.67 on DSC values. Conclusions: Our study revealed that segmentation accuracy varies slightly among different DL methods, with 2D networks outperforming 3D networks in 2D MRI sequences. Doctors found the time-saving aspect advantageous. The proposed failure detection could guide doctors in sensitive cases.

Radiomics to evaluate interlesion heterogeneity and to predict lesion response and patient outcomes using a validated signature of CD8 cells in advanced melanoma patients treated with anti-PD1 immunotherapy.

PURPOSE: While there is still a significant need to identify potential biomarkers that can predict which patients are most likely to respond to immunotherapy treatments, radiomic approaches have shown promising results. The objectives of this study were to evaluate whether a previously validated radiomics signature of CD8 T-cells could predict progressions at a lesion level and whether the spatial heterogeneity of this radiomics score could be used at a patient level to assess the clinical response and survival of melanoma patients. METHODS: Clinical data from patients with advanced melanoma treated in our center with immunotherapy were retrieved. Radiomic features were extracted and the CD8 radiomics signature was applied. A progressive lesion was defined by an increase in lesion size of 20% or more. Dispersion metrics of the radiomics signature were estimated to evaluate the impact of interlesion heterogeneity on patient's response. Fine-tuned cut-offs for predicting overall survival were evaluated after splitting data into training and test sets. RESULTS: A total of 136 patients were included in this study, with 1120 segmented lesions at baseline, and 1052 lesions at first evaluation. A low CD8 radiomics score at baseline was associated with a significantly higher risk of lesion progression (AUC=0.55, p=0.0091), especially for lesions larger than >1 mL (AUC=0.59 overall, p=0.0035, with AUC=0.75, p=0.002 for subcutaneous lesions, AUC=0.68, p=0.01, for liver lesions and AUC=0.62, p=0.03 for nodes). The least infiltrated lesion according to the radiomics score of CD8 T-cells was positively associated with overall survival (training set HR=0.31, p=0.00062, test set HR=0.28, p=0.016), which remained significant in a multivariate analysis including clinical and biological variables. CONCLUSIONS: These results confirm the predictive value at a lesion level of the biologically inspired CD8 radiomics score in melanoma patients treated with anti-PD1-based immunotherapy and may be interesting to assess the disease spatial heterogeneity to evaluate the patient prognosis with potential clinical implication such as tumor selection for focal ablative therapies.

Prognostic of recurrence and survival in poorly differentiated thyroid cancer.

The prognosis of poorly differentiated thyroid carcinomas (PDTC) defined by the Turin criteria is variable. The aim of this study on 51 PDTC patients was to determine clinical, histological and molecular prognostic factors associated with recurrence in patients with localized disease at initial treatment and with overall survival in patients with distant metastases. Of 40 patients for whom next-generation sequencing (NGS) by ThyroSeq v3 was able to be performed on historical samples, we identified high-risk molecular signature (TERT, TP53 mutations) in 24 (60%) cases, intermediate risk signature in 9 (22.5%) cases and low-risk signature in 7 (17.5%) cases. Potentially actionable mutations were identified in 10% of cases. After a median follow-up of 57.5 months, recurrence occurred in 11 (39%) of the 28 patients with localized disease. The American Thyroid Association (ATA) high risk of relapse, high mitotic count, high molecular risk signature and CD163 expression were associated with recurrence (P = 0.009, 0.01, 0.049, 0.03 respectively). After a median follow-up of 49.5 months, thyroid cancer-related death occurred in 53% of the patients with distant metastases. There was no significant prognostic factor associated with death in univariate analysis. However, none of the patients with intermediate ATA risk of recurrence and none of the patients with low-risk molecular signature died from the disease. In addition, high molecular-risk signature was associated with the presence of synchronous or metachronous distant metastasis (P = 0.007) and with poor overall survival (P = 0.01). In conclusion, ATA risk of relapse and high mitotic count was associated with higher rate of recurrence in localized PDTC. High molecular-risk signature was associated with the presence of distant metastasis and poor overall survival. Further studies are needed to determine if molecular testing adds to ATA risk stratification or response to therapy in predicting outcomes.

Imaging approaches and radiomics: toward a new era of ultraprecision radioimmunotherapy?

Strong rationale and a growing number of preclinical and clinical studies support combining radiotherapy and immunotherapy to improve patient outcomes. However, several critical questions remain, such as the identification of patients who will benefit from immunotherapy and the identification of the best modalities of treatment to optimize patient response. Imaging biomarkers and radiomics have recently emerged as promising tools for the non-invasive assessment of the whole disease of the patient, allowing comprehensive analysis of the tumor microenvironment, the spatial heterogeneity of the disease and its temporal changes. This review presents the potential applications of medical imaging and the challenges to address, in order to help clinicians choose the optimal modalities of both radiotherapy and immunotherapy, to predict patient's outcomes and to assess response to these promising combinations.

18F-FDG PET/CT for early response assessment of jaw osteoradionecrosis after the PENTOCLO protocol: A promising imaging modality.

INTRODUCTION: The Pentoxifylline, Tocopherol and Clodronate protocol (PENTOCLO) showed promising results for jaw osteoradionecrosis (ORN) management. However, the clinical and radiological improvements are often delayed, leading to unwanted long-term treatment, with potential loss of opportunity for more radical surgical treatments. Our objective was to assess the diagnosis performance of 18F-FDG PET/CT to early predict ORN response to the PENTOCLO protocol. MATERIALS AND METHODS: All patients from our center who were treated with the PENTOCLO protocol and with a 18F-FDG PET/CT performed at diagnosis and three months after the end of antibiotherapy were retrospectively included. The PENTOCLO protocol was always combined with prior appropriate antibiotherapy for six weeks. The healing endpoint was divided into healing, stability or worsening, according to the combination of clinical and radiological assessments at the date of last follow-up. For each patient, the difference between the maximal standardized uptake value (ΔSUVmax) of the ORN lesion at three months and baseline were computed. Diagnostic performance of 18F-FDG PET/CT was evaluated by sensitivity, specificity and the area under the receiver operating characteristic curve (ROC-AUC) of ΔSUVmax. RESULTS: 24 patients were included with an average follow-up of 29.3 months. The healing, stability and worsening rate were 25%, 62.5% and 12.5% respectively. The AUC for discriminating worsening vs stability or healing was 0.92 (IC95 [0.81-1.00]). A ΔSUVmax greater than or equal to 0 was predictive of a worsening with a sensitivity and specificity of 84 and 66% respectively. CONCLUSION: 18F-FDG PET/CT imaging could be useful for early prediction of PENTOCLO treatment resistance with appropriate antibiotherapy.

Radiation dose reduction with the wide-volume scan mode for interstitial lung diseases.

OBJECTIVES: The wide-volume mode, available on wide-area detector row CTs, has the advantage of reducing exposure time and radiation dose. It is infrequently used for lung diseases. The purpose of this study is to compare image quality and radiation dose of wide-volume chest CT to those of standard helical CT in the setting of interstitial lung diseases. METHODS: Retrospective monocentric study including 50 consecutive patients referred for follow-up or screening of interstitial lung diseases, requiring prone scan, acquired with the wide-volume mode, in addition to the routine supine scan, acquired with the helical mode. The optimal collimation in wide-volume mode (320 × 0.5mm or 240 × 0.5mm) was chosen according to the length of the thorax. Wide-volume acquisitions were compared to helical acquisitions for radiation dose (CTDIvol, DLP) and image quality, including analysis of normal structures, lesions, overall image quality, and artifacts (Wilcoxon signed-rank test). RESULTS: Median CTDIvol and DLP with wide volumes (3.1 mGy and 94.6 mGy·cm) were significantly reduced (p < 0.0001) as compared to helical mode (3.7mGy and 122.1 mGy·cm), leading to a median 21% and 32% relative reduction of CTDIvol and DLP, respectively. Image noise and quality were not significantly different between the two modes. Misalignment artifact at the junction of two volumes was occasionally seen in the wide-volume scans and, when present, did not impair the diagnostic quality in the majority of cases. CONCLUSIONS: Wide-volume mode allows 32% radiation dose reduction compared to the standard helical mode and could be used routinely for diagnosis and follow-up of interstitial lung diseases. KEY POINTS: • Retrospective monocentric study showed that wide-volume scan mode reduces radiation dose by 32% in comparison to helical mode for chest CT in the setting of interstitial lung diseases. • Mild misalignment may be observed at the junction between volumes with the wide-volume mode, without decrease of image quality in the majority of cases and without impairing diagnostic quality. • Wide-volume mode could be used routinely for the diagnosis and follow-up of interstitial lung diseases.

Reinventing radiation therapy with machine learning and imaging bio-markers (radiomics): State-of-the-art, challenges and perspectives.

Radiation therapy is a pivotal cancer treatment that has significantly progressed over the last decade due to numerous technological breakthroughs. Imaging is now playing a critical role on deployment of the clinical workflow, both for treatment planning and treatment delivery. Machine-learning analysis of predefined features extracted from medical images, i.e. radiomics, has emerged as a promising clinical tool for a wide range of clinical problems addressing drug development, clinical diagnosis, treatment selection and implementation as well as prognosis. Radiomics denotes a paradigm shift redefining medical images as a quantitative asset for data-driven precision medicine. The adoption of machine-learning in a clinical setting and in particular of radiomics features requires the selection of robust, representative and clinically interpretable biomarkers that are properly evaluated on a representative clinical data set. To be clinically relevant, radiomics must not only improve patients' management with great accuracy but also be reproducible and generalizable. Hence, this review explores the existing literature and exposes its potential technical caveats, such as the lack of quality control, standardization, sufficient sample size, type of data collection, and external validation. Based upon the analysis of 165 original research studies based on PET, CT-scan, and MRI, this review provides an overview of new concepts, and hypotheses generating findings that should be validated. In particular, it describes evolving research trends to enhance several clinical tasks such as prognostication, treatment planning, response assessment, prediction of recurrence/relapse, and prediction of toxicity. Perspectives regarding the implementation of an AI-based radiotherapy workflow are presented.

Insights into the variability of nasal potential difference, a biomarker of CFTR activity.

BACKGROUND: Nasal potential difference (NPD) is used to evaluate CFTR function in vivo. We aimed to evaluate the intrasubject and intersubject variability of NPD measurements. METHODS: We reviewed NPD tracings of 116 patients with CF enrolled in the placebo arm of a multicenter study. Patients carried at least one nonsense mutation and underwent repeated NPD tests every 16 weeks. NPD parameters included basal potential difference (basal PD), inhibition of sodium absorption by amiloride (Δ Amiloride), chloride (Cl-) transport in response to a Cl--free solution (Δ Low Cl-), isoproterenol (Δ Isoproterenol), the sum of Δ Low Cl- and Δ Isoproterenol (Δ Low Cl--Isoproterenol) and ATP (Δ ATP). RESULTS: Basal PD and Δ Amiloride displayed the highest variabilities, mainly stemming from intercenter and intrasubject effect. Δ Low Cl-, Δ Isoproterenol and Δ Low Cl--Isoproterenol demonstrated a large intrasubject variability but a smaller intersubject variability. The intrasubject measurement variability for Δ Low Cl--Isoproterenol, was within ± 7.2 mV with 95% probability. It was greater in patients reporting ongoing pulmonary exacerbations. CONCLUSIONS: The large intercenter variability of basal PD and Δ Amiloride highlights the operator-dependent aspect of these measurements. A difference greater than 7.2 mV in Δ Low Cl--Isoproterenol in a given patient on CFTR modulator can be attributed, with 95% probability, to a treatment effect rather than to the variability inherent in the measurement.

Longer survival in patients with metastatic uterine leiomyosarcoma treated with trabectedin: A case report.

Trabectedin (ET-743) is a marine alkaloid isolated from the Caribbean tunicate Ecteinascidia turbinata, with a chemical structure characterized by three fused tetrahydroisoquinoline rings. In the present case report, two patients with advanced and metastatic uterine leiomyosarcomas (ULMS) with significant progression-free survival (PFS) and overall survival (OS) administered Trabectedin as second and third line treatment are reported. The first case received third line Trabectedin with a PFS of 24 months and an OS of 35 months. The second case received second line Trabectedin with a PFS of 24 months and an OS of 30 months. In addition, a good safety record was obtained in the long-term administration of Trabectedin (more so in case 1 than case 2), with a good quality of life.

Dose reference levels and clinical determinants in stroke neuroradiology interventions.

OBJECTIVES: To establish dose reference levels (RLs) for stroke interventions while carefully analysing the impact of clinical and technical parameters on patient exposure. METHODS: The study retrospectively analysed data from 377 stroke patients prospectively collected between 15 October 2015 and 30 March 2017 at a single, level-3 stroke centre equipped with Philips Allura Clarity systems. Local dose RLs were first derived as the 75th percentile of the dose area product (DAP), cumulative air kerma (Ka,r), fluoroscopy time (FT) and the number of images (NI). Univariate and multivariate negative binomial regressions were considered for the statistical analysis to investigate the dose variability with clinical and technical parameters such as patient's age and sex, occlusion removal technique, number of passages, single-plane or biplane equipment, etc. RESULTS: Local stroke dose RLs were derived in terms of total DAP (162 Gy cm2), Ka,r (854 mGy), FT (42 min) and NI (559). Gender (relative dose multiplier (RDM) 1.31; 95% CI 1.12-1.45), number of passages (RDM 1.22 per passage; 95% CI 1.10-1.22) and procedure success (RDM 0.52, 95% CI 0.55-0.80) proved to be key parameters affecting patient dose. Meanwhile the statistical analysis did not find any difference in relative dose received by patients owing to age, baseline NIHSS score, occlusion removal technique, posterior circulation, support of an anaesthesiologist or use of biplane equipment. CONCLUSIONS: Stroke dose RLs introduced in this work promote the optimisation of patient doses. Male gender, number of passages and success of recanalisation are independent key parameters affecting patient dose. KEY POINTS: • Stroke dose RLs derived in terms of total DAP (162 Gy cm 2 ), K a,r (854 mGy), FT (42 min) and NI (559) will help optimise the radiation safety of patients treated with mechanical thrombectomy. • Male gender (relative dose multiplier 1.31; 95% CI 1.12-1.45), number of passages (RDM 1.22 per passage; 95% CI 1.10-1.22) and success of recanalisation TICI score > 2b (RDM 0.52, 95% CI 0.55-0.80) are independent key parameters affecting patient dose. • Stent retriever or aspiration technique showed no significant difference in terms of the dose delivered to the patient; neither technique should be favoured for dosimetric reasons provided that there is no difference regarding clinical outcomes.

Rapid Contour-based Segmentation for 18F-FDG PET Imaging of Lung Tumors by Using ITK-SNAP: Comparison to Expert-based Segmentation.

Purpose To assess the performance of the ITK-SNAP software for fluorodeoxyglucose (FDG) positron emission tomography (PET) segmentation of complex-shaped lung tumors compared with an optimized, expert-based manual reference standard. Materials and Methods Seventy-six FDG PET images of thoracic lesions were retrospectively segmented by using ITK-SNAP software. Each tumor was manually segmented by six raters to generate an optimized reference standard by using the simultaneous truth and performance level estimate algorithm. Four raters segmented 76 FDG PET images of lung tumors twice by using ITK-SNAP active contour algorithm. Accuracy of ITK-SNAP procedure was assessed by using Dice coefficient and Hausdorff metric. Interrater and intrarater reliability were estimated by using intraclass correlation coefficients of output volumes. Finally, the ITK-SNAP procedure was compared with currently recommended PET tumor delineation methods on the basis of thresholding at 41% volume of interest (VOI; VOI41) and 50% VOI (VOI50) of the tumor's maximal metabolism intensity. Results Accuracy estimates for the ITK-SNAP procedure indicated a Dice coefficient of 0.83 (95% confidence interval: 0.77, 0.89) and a Hausdorff distance of 12.6 mm (95% confidence interval: 9.82, 15.32). Interrater reliability was an intraclass correlation coefficient of 0.94 (95% confidence interval: 0.91, 0.96). The intrarater reliabilities were intraclass correlation coefficients above 0.97. Finally, VOI41 and VOI50 accuracy metrics were as follows: Dice coefficient, 0.48 (95% confidence interval: 0.44, 0.51) and 0.34 (95% confidence interval: 0.30, 0.38), respectively, and Hausdorff distance, 25.6 mm (95% confidence interval: 21.7, 31.4) and 31.3 mm (95% confidence interval: 26.8, 38.4), respectively. Conclusion ITK-SNAP is accurate and reliable for active-contour-based segmentation of heterogeneous thoracic PET tumors. ITK-SNAP surpassed the recommended PET methods compared with ground truth manual segmentation.