Imaging-guided prognostic score-based approach to assess the benefits of combotherapy versus monotherapy with immune checkpoint inhibitors in metastatic MSI-H colorectal cancer patients

BACKGROUND

This retrospective study determined survival responses to immune checkpoint inhibitors (ICIs), comparing mono- (mono) and combo-immunotherapy (combo) in patients with microsatellite instability-high (MSI-H) metastatic colorectal cancer (mCRC) by analyzing quantitative imaging data and clinical factors.

METHODS

One hundred fifty patients were included from two centers and divided into training (n = 105) and validation (n = 45) cohorts. Radiologists manually annotated chest-abdomen-pelvis computed tomography and calculated tumor burden. Progression-free survival (PFS) was assessed, and variables were selected through Recursive Feature Elimination. Cutoff values were determined using maximally selected rank statistics to binarize features, forming a risk score with hazard ratio-derived weights.

RESULTS

In total, 2258 lesions were annotated with excellent reproducibility. Key variables in the training cohort included: total tumor volume (cutoff: 73 cm3), lesion count (cutoff: 20), age (cutoff: 60) and the presence of peritoneal carcinomatosis. Their respective weights were 1.13, 0.96, 0.91, and 0.38, resulting in a risk score cutoff of 1.36. Low-score patients showed similar overall survival and PFS regardless of treatment, while those with a high-score had significantly worse survivals with mono vs combo (P = 0.004 and P = 0.0001). In the validation set, low-score patients exhibited no significant difference in overall survival and PFS with mono or combo. However, patients with a high-score had worse PFS with mono (P = 0.046).

CONCLUSIONS

A score based on total tumor volume, lesion count, the presence of peritoneal carcinomatosis, and age can guide MSI-H mCRC treatment decisions, allowing oncologists to identify suitable candidates for mono and combo ICI therapies.

Detection and severity quantification of pulmonary embolism with 3D CT data using an automated deep learning-based artificial solution

PURPOSE

The purpose of this study was to propose a deep learning-based approach to detect pulmonary embolism and quantify its severity using the Qanadli score and the right-to-left ventricle diameter (RV/LV) ratio on three-dimensional (3D) computed tomography pulmonary angiography (CTPA) examinations with limited annotations.

MATERIALS AND METHODS

Using a database of 3D CTPA examinations of 1268 patients with image-level annotations, and two other public datasets of CTPA examinations from 91 (CAD-PE) and 35 (FUME-PE) patients with pixel-level annotations, a pipeline consisting of: (i), detecting blood clots; (ii), performing PE-positive versus negative classification; (iii), estimating the Qanadli score; and (iv), predicting RV/LV diameter ratio was followed. The method was evaluated on a test set including 378 patients. The performance of PE classification and severity quantification was quantitatively assessed using an area under the curve (AUC) analysis for PE classification and a coefficient of determination (R²) for the Qanadli score and the RV/LV diameter ratio.

RESULTS

Quantitative evaluation led to an overall AUC of 0.870 (95% confidence interval [CI]: 0.850-0.900) for PE classification task on the training set and an AUC of 0.852 (95% CI: 0.810-0.890) on the test set. Regression analysis yielded R² value of 0.717 (95% CI: 0.668-0.760) and of 0.723 (95% CI: 0.668-0.766) for the Qanadli score and the RV/LV diameter ratio estimation, respectively on the test set.

CONCLUSION

This study shows the feasibility of utilizing AI-based assistance tools in detecting blood clots and estimating PE severity scores with 3D CTPA examinations. This is achieved by leveraging blood clots and cardiac segmentations. Further studies are needed to assess the effectiveness of these tools in clinical practice.

EFSUMB Technical Review - Update 2023: Dynamic Contrast-Enhanced Ultrasound (DCE-CEUS) for the Quantification of Tumor Perfusion

Dynamic contrast-enhanced ultrasound (DCE-US) is a technique to quantify tissue perfusion based on phase-specific enhancement after the injection of microbubble contrast agents for diagnostic ultrasound. The guidelines of the European Federation of Societies for Ultrasound in Medicine and Biology (EFSUMB) published in 2004 and updated in 2008, 2011, and 2020 focused on the use of contrast-enhanced ultrasound (CEUS), including essential technical requirements, training, investigational procedures and steps, guidance regarding image interpretation, established and recommended clinical indications, and safety considerations. However, the quantification of phase-specific enhancement patterns acquired with ultrasound contrast agents (UCAs) is not discussed here. The purpose of this EFSUMB Technical Review is to further establish a basis for the standardization of DCE-US focusing on treatment monitoring in oncology. It provides some recommendations and descriptions as to how to quantify dynamic ultrasound contrast enhancement, and technical explanations for the analysis of time-intensity curves (TICs). This update of the 2012 EFSUMB introduction to DCE-US includes clinical aspects for data collection, analysis, and interpretation that have emerged from recent studies. The current study not only aims to support future work in this research field but also to facilitate a transition to clinical routine use of DCE-US.

BR55 Ultrasound Molecular Imaging of Clear Cell Renal Cell Carcinoma Reflects Tumor Vascular Expression of VEGFR-2 in a Patient-Derived Xenograft Model

Standard imaging cannot reliably predict the nature of renal tumors. Among malignant renal tumors, clear cell renal cell carcinoma (ccRCC) is the most common histological subtype, in which the vascular endothelial growth factor 2 (VEGFR-2) is highly expressed in the vascular endothelium. BR55, a contrast agent for ultrasound imaging, consists of gas-core lipid microbubbles that specifically target and bind to the extracellular portion of the VEGFR-2. The specific information provided by ultrasound molecular imaging (USMI) using BR55 was compared with the vascular tumor expression of the VEGFR-2 by immunohistochemical (IHC) staining in a preclinical model of ccRCC. Patients' ccRCCs were orthotopically grafted onto Nod-Scid-Gamma (NSG) mice to generate patient-derived xenografts (PdX). Mice were divided into four groups to receive either vehicle or axitinib an amount of 2, 7.5 or 15 mg/kg twice daily. Perfusion parameters and the BR55 ultrasound contrast signal on PdX renal tumors were analyzed at D0, D1, D3, D7 and D11, and compared with IHC staining for the VEGFR-2 and CD34. Significant Pearson correlation coefficients were observed between the area under the curve (AUC) and the CD34 (0.84, p < 10-4), and between the VEGFR-2-specific signal obtained by USMI and IHC (0.72, p < 10-4). USMI with BR55 could provide instant, quantitative information on tumor VEGFR-2 expression to characterize renal masses non-invasively.

Detection and quantification of pulmonary embolism with artificial intelligence: The SFR 2022 artificial intelligence data challenge

PURPOSE

In 2022, the French Society of Radiology together with the French Society of Thoracic Imaging and CentraleSupelec organized their 13th data challenge. The aim was to aid in the diagnosis of pulmonary embolism, by identifying the presence of pulmonary embolism and by estimating the ratio between right and left ventricular (RV/LV) diameters, and an arterial obstruction index (Qanadli's score) using artificial intelligence.

MATERIALS AND METHODS

The data challenge was composed of three tasks: the detection of pulmonary embolism, the RV/LV diameter ratio, and Qanadli's score. Sixteen centers all over France participated in the inclusion of the cases. A health data hosting certified web platform was established to facilitate the inclusion process of the anonymized CT examinations in compliance with general data protection regulation. CT pulmonary angiography images were collected. Each center provided the CT examinations with their annotations. A randomization process was established to pool the scans from different centers. Each team was required to have at least a radiologist, a data scientist, and an engineer. Data were provided in three batches to the teams, two for training and one for evaluation. The evaluation of the results was determined to rank the participants on the three tasks.

RESULTS

A total of 1268 CT examinations were collected from the 16 centers following the inclusion criteria. The dataset was split into three batches of 310, 580 and 378 C T examinations provided to the participants respectively on September 5, 2022, October 7, 2022 and October 9, 2022. Seventy percent of the data from each center were used for training, and 30% for the evaluation. Seven teams with a total of 48 participants including data scientists, researchers, radiologists and engineering students were registered for participation. The metrics chosen for evaluation included areas under receiver operating characteristic curves, specificity and sensitivity for the classification task, and the coefficient of determination r2 for the regression tasks. The winning team achieved an overall score of 0.784.

CONCLUSION

This multicenter study suggests that the use of artificial intelligence for the diagnosis of pulmonary embolism is possible on real data. Moreover, providing quantitative measures is mandatory for the interpretability of the results, and is of great aid to the radiologists especially in emergency settings.

Synergic prognostic value of 3D CT scan subcutaneous fat and muscle masses for immunotherapy-treated cancer

BACKGROUND

Our aim was to explore the prognostic value of anthropometric parameters in a large population of patients treated with immunotherapy.

METHODS

We retrospectively included 623 patients with advanced non-small cell lung cancer (NSCLC) (n=318) or melanoma (n=305) treated by an immune-checkpoint-inhibitor having a pretreatment (thorax-)abdomen-pelvis CT scan. An external validation cohort of 55 patients with NSCLC was used. Anthropometric parameters were measured three-dimensionally (3D) by a deep learning software (Anthropometer3DNet) allowing an automatic multislice measurement of lean body mass, fat body mass (FBM), muscle body mass (MBM), visceral fat mass (VFM) and sub-cutaneous fat mass (SFM). Body mass index (BMI) and weight loss (WL) were also retrieved. Receiver operator characteristic (ROC) curve analysis was performed and overall survival was calculated using Kaplan-Meier (KM) curve and Cox regression analysis.

RESULTS

In the overall cohort, 1-year mortality rate was 0.496 (95% CI: 0.457 to 0.537) for 309 events and 5-year mortality rate was 0.196 (95% CI: 0.165 to 0.233) for 477 events. In the univariate Kaplan-Meier analysis, prognosis was worse (p<0.001) for patients with low SFM (<3.95 kg/m2), low FBM (<3.26 kg/m2), low VFM (<0.91 kg/m2), low MBM (<5.85 kg/m2) and low BMI (<24.97 kg/m2). The same parameters were significant in the Cox univariate analysis (p<0.001) and, in the multivariate stepwise Cox analysis, the significant parameters were MBM (p<0.0001), SFM (0.013) and WL (0.0003). In subanalyses according to the type of cancer, all body composition parameters were statistically significant for NSCLC in ROC, KM and Cox univariate analysis while, for melanoma, none of them, except MBM, was statistically significant. In multivariate Cox analysis, the significant parameters for NSCLC were MBM (HR=0.81, p=0.0002), SFM (HR=0.94, p=0.02) and WL (HR=1.06, p=0.004). For NSCLC, a KM analysis combining SFM and MBM was able to separate the population in three categories with the worse prognostic for the patients with both low SFM (<5.22 kg/m2) and MBM (<6.86 kg/m2) (p<0001). On the external validation cohort, combination of low SFM and low MBM was pejorative with 63% of mortality at 1 year versus 25% (p=0.0029).

CONCLUSIONS

3D measured low SFM and MBM are significant prognosis factors of NSCLC treated by immune checkpoint inhibitors and can be combined to improve the prognostic value.

New calibration setup for quantitative DCE-US imaging protocol: Toward standardization

BACKGROUND

The DCE-US (Dynamic Contrast-Enhanced Ultrasonography) imaging protocol predicts the vascular modifications compared with Response Evaluation Criteria in Solid Tumors (RECIST) based mainly on morphological changes. A quantitative biomarker has been validated through the DCE-US multi-centric study for early monitoring of the efficiency of anti-angiogenic cancer treatments. In this context, the question of transposing the use of this biomarker to other types of ultrasound scanners, probes and settings has arisen to maintain the follow-up of patients under anti-angiogenic treatments. As a consequence, radiologists encounter standardization issues between the different generations of ultrasound scanners to perform quantitative imaging protocols.

PURPOSE

The aim of this study was to develop a new calibration setup to transpose the DCE-US imaging protocol to the new generation of ultrasound scanners using both abdominal and linear probes.

METHODS

This calibration method has been designed to be easily reproducible and optimized, reducing the time required and cost incurred. It is based on an original set-up that includes using a concentration splitter to measure the variation of the harmonic signal intensity, obtained from the Area Under the time-intensity Curve (AUC) as a function of various contrast-agent concentrations. The splitter provided four different concentrations simultaneously ranging from 12.5% to 100% of the initial concentration of the SonoVue contrast agent (Bracco Imaging S.p.A., Milan, Italy), therefore, measuring four AUCs in a single injection. The plot of the AUC as a function of the four contrast agent concentrations represents the intensity variation of the harmonic signal: the slope being the calibration parameter. The standardization through this method implied that both generations of ultrasound scanners had to have the same slopes to be considered as calibrated. This method was tested on two ultrasound scanners from the same manufacturer (Aplio500, Aplioi900, Canon Medical Systems, Tokyo, Japan). The Aplio500 used the settings defined by the initial multicenter DCE-US study. The Mechanical Index (MI) and the Color Gain (CG) of the Aplioi900 have been adjusted to match those of the Aplio500. The reliability of the new setup was evaluated in terms of measurement repeatability, and reproducibility with the agreement between the measurements obtained once the two ultrasound scanners were calibrated.

RESULTS

The new setup provided excellent repeatability measurements with a value of 96.8%. Once the two ultrasound scanners have been calibrated for both types of probes, the reproducibility was excellent with the agreement between their respective quantitative measurement was at the lowest 95.4% and at the best 98.8%. The settings of the Aplioi900 (Canon Medical Systems) were adjusted to match those of the Aplio500 (Canon Medical Systems) and these validated settings were for the abdominal probe: MI = 0.13 and CG = 34 dB; and for the linear probe: MI = 0.10 and CG = 38 dB.

CONCLUSION

This new calibration setup provided reliable measurements and enabled the rapid transfer and the use of the DCE-US imaging protocol on new ultrasound scanners, thus permitting a continuation of the therapeutic evaluation of patients through quantitative imaging.

Outcomes of mandibular reconstruction using three-dimensional custom-made porous titanium prostheses

BACKGROUND

Our aim was to report the long-term outcomes of mandibular reconstruction using CAD-CAM-designed 3D-printed porous titanium implants in patients not amenable to a free vascularized fibula flap reconstruction.

METHODS

The implants were designed with ProPlan CMF® 2.2 software and manufactured with a Selective Laser Melting (SLM) "layer-by-layer" 3D-printing of pure porous titanium powder beds. Primary endpoints were implant exposure and implant removal calculated using Gray's tests. Secondary endpoints were predictive factors of implant exposure and implant removal, and rates of dental rehabilitation.

RESULTS

Thirty-six patients were operated between 2015 and 2017 and were included in this study. Reconstruction using a porous titanium 3D-printed implant was proposed due to medical contraindication for a fibula free flap (n = 13), due to the failure of a previous fibula free flap reconstruction (n = 7), or due to refusal of a fibula free flap reconstruction by the patient (n = 16). The medical indications for mandibular reconstruction were a primary tumor requiring mandibulectomy in nine patients, mandibular osteoradionecrosis requiring mandibulectomy in nineteen patients, and secondary reconstruction in eight patients. The 2-year rates of implant exposure and implant removal were 69.4% and 52.8%. Reconstruction of the symphysis was a high-risk exposure variable (OR 30; p = 0.0003). Only one patient underwent a successful dental rehabilitation.

CONCLUSION

The use of a porous titanium 3D- implant for mandibular reconstruction in head and neck cancer patients resulted in high rates of implant exposure and of implant removal, notably when symphysis involvement.

Better than RECIST and faster than iRECIST: defining the Immunotherapy Progression Decision score to better manage progressive tumors on immunotherapy

PURPOSE

The objective of the study is to propose the immunotherapy progression decision (iPD) score, a practical tool based on patient features that are available at the first evaluation of immunotherapy treatment, to help oncologists decide whether to continue the treatment or switch rapidly to another therapeutic line when facing a progressive disease patient at the 1st evaluation.

METHODS

This retrospective study included 107 patients with progressive disease at first evaluation according to RECIST 1.1. Clinical, radiological, and biological data at baseline and 1st evaluation were analyzed. An external validation set consisting of 31 patients with similar baseline characteristics was used for the validation of the score.

RESULTS

Variables were analyzed in a univariate study. The iPD score was constructed using only independent variables, each considered as a worsening factor for the survival of patients. The patients were stratified in three groups: Good Prognosis (GP), Poor Prognosis (PP) and Critical Prognosis (CP). Each group showed significantly different survivals (GP: 11.4, PP: 4.4, CP: 2.3 months median OS, p<0.001 log-rank-test). Moreover, the iPD score was able to detect the pseudo-progressors better than other scores. On the validation set, CP patients had significantly worse survival than PP and GP patients (p<0.05, log-rank-test).

CONCLUSION

The iPD score provides oncologists with a new evaluation, computable at first-progression, to decide if treatment should be continued (for the GP group), or immediately changed for the PP and CP groups. Further validation on larger cohorts is needed to prove its efficacy in clinical practice.

Synthetic MR image generation of macrotrabecular-massive hepatocellular carcinoma using generative adversarial networks

PURPOSE

The purpose of this study was to develop a method for generating synthetic MR images of macrotrabecular-massive hepatocellular carcinoma (MTM-HCC).

MATERIALS AND METHODS

A set of abdominal MR images including fat-saturated T1-weighted images obtained during the arterial and portal venous phases of enhancement and T2-weighted images of 91 patients with MTM-HCC, and another set of MR abdominal images from 67 other patients were used. Synthetic images were obtained using a 3-step pipeline that consisted in: (i), generating a synthetic MTM-HCC tumor on a neutral background; (ii), randomly selecting a background among the 67 patients and a position inside the liver; and (iii), merging the generated tumor in the background at the specified location. Synthetic images were qualitatively evaluated by three radiologists and quantitatively assessed using a mix of 1-nearest neighbor classifier metric and Fréchet inception distance.

RESULTS

A set of 1000 triplets of synthetic MTM-HCC images with consistent contrasts were successfully generated. Evaluation of selected synthetic images by three radiologists showed that the method gave realistic, consistent and diversified images. Qualitative and quantitative evaluation led to an overall score of 0.64.

CONCLUSION

This study shows the feasibility of generating realistic synthetic MR images with very few training data, by leveraging the wide availability of liver backgrounds. Further studies are needed to assess the added value of those synthetic images for automatic diagnosis of MTM-HCC.

Generative adversarial networks (GAN)-based data augmentation of rare liver cancers: The SFR 2021 Artificial Intelligence Data Challenge

PURPOSE

The 2021 edition of the Artificial Intelligence Data Challenge was organized by the French Society of Radiology together with the Centre National d'Études Spatiales and CentraleSupélec with the aim to implement generative adversarial networks (GANs) techniques to provide 1000 magnetic resonance imaging (MRI) cases of macrotrabecular-massive (MTM) hepatocellular carcinoma (HCC), a rare and aggressive subtype of HCC, generated from a limited number of real cases from multiple French centers.

MATERIALS AND METHODS

A dedicated platform was used by the seven inclusion centers to securely upload their anonymized MRI examinations including all three cross-sectional images (one late arterial and one portal-venous phase T1-weighted images and one fat-saturated T2-weighted image) in compliance with general data protection regulation. The quality of the database was checked by experts and manual delineation of the lesions was performed by the expert radiologists involved in each center. Multidisciplinary teams competed between October 11^th, 2021 and February 13^th, 2022.

RESULTS

A total of 91 MTM-HCC datasets of three images each were collected from seven French academic centers. Six teams with a total of 28 individuals participated in this challenge. Each participating team was asked to generate one thousand 3-image cases. The qualitative evaluation was performed by three radiologists using the Likert scale on ten randomly selected cases generated by each participant. A quantitative evaluation was also performed using two metrics, the Frechet inception distance and a leave-one-out accuracy of a 1-Nearest Neighbor algorithm.

CONCLUSION

This data challenge demonstrates the ability of GANs techniques to generate a large number of images from a small sample of imaging examinations of a rare malignant tumor.

Quantification and reduction of cross-vendor variation in multicenter DWI MR imaging: results of the Cancer Core Europe imaging task force

OBJECTIVES

In the Cancer Core Europe Consortium (CCE), standardized biomarkers are required for therapy monitoring oncologic multicenter clinical trials. Multiparametric functional MRI and particularly diffusion-weighted MRI offer evident advantages for noninvasive characterization of tumor viability compared to CT and RECIST. A quantification of the inter- and intraindividual variation occurring in this setting using different hardware is missing. In this study, the MRI protocol including DWI was standardized and the residual variability of measurement parameters quantified.

METHODS

Phantom and volunteer measurements (single-shot T2w and DW-EPI) were performed at the seven CCE sites using the MR hardware produced by three different vendors. Repeated measurements were performed at the sites and across the sites including a traveling volunteer, comparing qualitative and quantitative ROI-based results including an explorative radiomics analysis.

RESULTS

For DWI/ADC phantom measurements using a central post-processing algorithm, the maximum deviation could be decreased to 2%. However, there is no significant difference compared to a decentralized ADC value calculation at the respective MRI devices. In volunteers, the measurement variation in 2 repeated scans did not exceed 11% for ADC and is below 20% for single-shot T2w in systematic liver ROIs. The measurement variation between sites amounted to 20% for ADC and < 25% for single-shot T2w. Explorative radiomics classification experiments yield better results for ADC than for single-shot T2w.

CONCLUSION

Harmonization of MR acquisition and post-processing parameters results in acceptable standard deviations for MR/DW imaging. MRI could be the tool in oncologic multicenter trials to overcome the limitations of RECIST-based response evaluation.

KEY POINTS

• Harmonizing acquisition parameters and post-processing homogenization, standardized protocols result in acceptable standard deviations for multicenter MR-DWI studies. • Total measurement variation does not to exceed 11% for ADC in repeated measurements in repeated MR acquisitions, and below 20% for an identical volunteer travelling between sites. • Radiomic classification experiments were able to identify stable features allowing for reliable discrimination of different physiological tissue samples, even when using heterogeneous imaging data.

Predicting immunotherapy outcomes in patients with MSI tumors using NLR and CT global tumor volume

BACKGROUND

Anti-PD-(L)1 treatment is indicated for patients with mismatch repair-deficient (MMRD) tumors, regardless of tumor origin. However, the response rate is highly heterogeneous across MMRD tumors. The objective of the study is to find a score that predicts anti-PD-(L)1 response in patients with MMRD tumors.

METHODS

Sixty-one patients with various origin of MMRD tumors and treated with anti-PD-(L)1 were retrospectively included in this study. An expert radiologist annotated all tumors present at the baseline and first evaluation CT-scans for all the patients by circumscribing them on their largest axial axis (single slice), allowing us to compute an approximation of their tumor volume. In total, 2120 lesions were annotated, which led to the computation of the total tumor volume for each patient. The RECIST sum of target lesions' diameters and neutrophile-to-lymphocyte (NLR) were also reported at both examinations. These parameters were determined at baseline and first evaluation and the variation between the first evaluation and baseline was calculated, to determine a comprehensive score for overall survival (OS) and progression-free survival (PFS).

RESULTS

Total tumor volume at baseline was found to be significantly correlated to the OS (p-value: 0.005) and to the PFS (p-value:<0.001). The variation of the RECIST sum of target lesions' diameters, total tumor volume and NLR were found to be significantly associated to the OS (p-values:<0.001, 0.006,<0.001 respectively) and to the PFS (<0.001,<0.001, 0.007 respectively). The concordance score combining total tumor volume and NLR variation was better at stratifying patients compared to the tumor volume or NLR taken individually according to the OS (pairwise log-rank test p-values: 0.033,<0.001, 0.002) and PFS (pairwise log-rank test p-values: 0.041,<0.001, 0.003).

CONCLUSION

Total tumor volume appears to be a prognostic biomarker of anti-PD-(L)1 response to immunotherapy in metastatic patients with MMRD tumors. Combining tumor volume and NLR with a simple concordance score stratifies patients well according to their survival and offers a good predictive measure of response to immunotherapy.

Influence of Different Measurement Methods of Arterial Input Function on Quantitative Dynamic Contrast-Enhanced MRI Parameters in Head and Neck Cancer

BACKGROUND

Head and neck cancer (HNC) is the sixth most prevalent cancer worldwide. Dynamic contrast-enhanced MRI (DCE-MRI) helps in diagnosis and prognosis. Quantitative DCE-MRI requires an arterial input function (AIF), which affects the values of pharmacokinetic parameters (PKP).

PURPOSE

To evaluate influence of four individual AIF measurement methods on quantitative DCE-MRI parameters values (K^trans , v_e , k_ep , and v_p ), for HNC and muscle.

STUDY TYPE

Prospective.

POPULATION

A total of 34 HNC patients (23 males, 11 females, age range 24-91) FIELD STRENGTH/SEQUENCE: A 3 T; 3D SPGR gradient echo sequence with partial saturation of inflowing spins.

ASSESSMENT

Four AIF methods were applied: automatic AIF (AIFa) with up to 50 voxels selected from the whole FOV, manual AIF (AIFm) with four voxels selected from the internal carotid artery, both conditions without (Mc-) or with (Mc+) motion correction. Comparison endpoints were peak AIF values, PKP values in tumor and muscle, and tumor/muscle PKP ratios.

STATISTICAL TESTS

Nonparametric Friedman test for multiple comparisons. Nonparametric Wilcoxon test, without and with Benjamini Hochberg correction, for pairwise comparison of AIF peak values and PKP values for tumor, muscle and tumor/muscle ratio, P value ≤ 0.05 was considered statistically significant.

RESULTS

Peak AIF values differed significantly for all AIF methods, with mean AIFmMc+ peaks being up to 66.4% higher than those for AIFaMc+. Almost all PKP values were significantly higher for AIFa in both, tumor and muscle, up to 76% for mean K^trans values. Motion correction effect was smaller. Considering tumor/muscle parameter ratios, most differences were not significant (0.068 ≤ Wilcoxon P value ≤ 0.8).

DATA CONCLUSION

We observed important differences in PKP values when using either AIFa or AIFm, consequently choice of a standardized AIF method is mandatory for DCE-MRI on HNC. From the study findings, AIFm and inflow compensation are recommended. The use of the tumor/muscle PKP ratio should be of interest for multicenter studies.

EVIDENCE LEVEL

2 TECHNICAL EFFICACY: Stage 1.

Opportunities for improved clinical trial designs in acute respiratory distress syndrome

The acute respiratory distress syndrome (ARDS) is a common critical illness syndrome with high morbidity and mortality. There are no proven pharmacological therapies for ARDS. The current definition of ARDS is based on shared clinical characteristics but does not capture the heterogeneity in clinical risk factors, imaging characteristics, physiology, timing of onset and trajectory, and biology of the syndrome. There is increasing interest within the ARDS clinical trialist community to design clinical trials that reduce heterogeneity in the trial population. This effort must be balanced with ongoing work to craft an inclusive, global definition of ARDS, with important implications for trial design. Ultimately, the two aims-to design trials that are applicable to the diverse global ARDS population while also advancing opportunities to identify targetable traits-should coexist. In this Personal View, we recommend two primary strategies to improve future ARDS trials: the development of new methods to target treatable traits in clinical trial populations, and improvements in the representativeness of ARDS trials, with the inclusion of global populations. We emphasise that these two strategies are complementary. We also discuss how a proposed expansion of the definition of ARDS could affect the future of clinical trials.

An artificial intelligence model predicts the survival of solid tumour patients from imaging and clinical data

BACKGROUND

The need for developing new biomarkers is increasing with the emergence of many targeted therapies. Artificial Intelligence (AI) algorithms have shown great promise in the medical imaging field to build predictive models. We developed a prognostic model for solid tumour patients using AI on multimodal data.

PATIENTS AND METHODS

Our retrospective study included examinations of patients with seven different cancer types performed between 2003 and 2017 in 17 different hospitals. Radiologists annotated all metastases on baseline computed tomography (CT) and ultrasound (US) images. Imaging features were extracted using AI models and used along with the patients' and treatments' metadata. A Cox regression was fitted to predict prognosis. Performance was assessed on a left-out test set with 1000 bootstraps.

RESULTS

The model was built on 436 patients and tested on 196 patients (mean age 59, IQR: 51-6, 411 men out of 616 patients). On the whole, 1147 US images were annotated with lesions delineation, and 632 thorax-abdomen-pelvis CTs (total of 301,975 slices) were fully annotated with a total of 9516 lesions. The developed model reaches an average concordance index of 0.71 (0.67-0.76, 95% CI). Using the median predicted risk as a threshold value, the model is able to significantly (log-rank test P value < 0.001) isolate high-risk patients from low-risk patients (respective median OS of 11 and 31 months) with a hazard ratio of 3.5 (2.4-5.2, 95% CI).

CONCLUSION

AI was able to extract prognostic features from imaging data, and along with clinical data, allows an accurate stratification of patients' prognoses.

Abstract 1924: PULS-AI: A multimodal artificial intelligence model to predict survival of solid tumor patients treated with antiangiogenics

The need for developing new biomarkers is increasing with the emergence of many targeted therapies. In this study, we used artificial intelligence (AI) to develop a multimodal model (PULS-AI) predicting the survival of solid tumor patients treated with antiangiogenic treatments.

Our retrospective, multicentric study included 616 patients with 7 different cancer types: renal cell carcinoma, colorectal carcinoma, hepatocellular carcinoma, gastrointestinal carcinoma, melanoma, breast cancer, and sarcoma. A set of 196 patients was left out of the study for validation. Clinical data including patient, treatment, and cancer metadata were collected at baseline for all patients, as well as computed tomography (CT) and ultrasound (US) images. Radiologists annotated all metastases on the CT images and the visible tumor lesion on the US images. AI models were used to extract relevant features from the regions of interest on CT and US images. In addition, handcrafted features related to the tumor burden were extracted from the annotations of all lesions on CT such as the number of lesions and the tumor burden volume per organ (lungs, liver, skull, bone, other). Finally, a Cox regression model was fitted to the set of imaging features and clinical features.

The annotation process led to 1147 annotated US images with lesions delineation and 4564 reviewed CTs, of which 989 were selected and fully annotated with a total of 9516 annotated lesions.The developed model reaches an average concordance index of 0.71 (0.67-0.75, 95% CI). Using a risk threshold of 50%, PULS-AI model is able to significantly isolate (log-rank test P-value &lt; 0.001) high-risk patients from low-risk patients (respective median OS of 12 and 32 months) with a hazard ratio of 3.52 (2.35-5.28, 95% CI).

The results of this study show that AI algorithms are able to extract relevant information from radiology images and to aggregate data from multiple modalities to build powerful prognostic tools. Such tools may provide assistance to oncology clinicians in therapeutic decision-making.

Citation Format: Kathryn Schutte, Fabien Brulport, Sana Harguem-Zayani, Jean-Baptiste Schiratti, Ridouane Ghermi, Paul Jehanno, Alexandre Jaeger, Talal Alamri, Raphael Naccache, Leila Haddag-Miliani, Teresa Orsi, Jean-Philippe Lamarque, Isaline Hoferer, Littisha Lawrance, Baya Benatsou, Imad Bousaid, Mickael Azoulay, Antoine Verdon, François Bidault, Corinne Balleyguier, Victor Aubert, Etienne Bendjebbar, Charles Maussion, Nicolas Loiseau, Benoit Schmauch, Meriem Sefta, Gilles Wainrib, Thomas Clozel, Samy Ammari, Nathalie Lassau. PULS-AI: A multimodal artificial intelligence model to predict survival of solid tumor patients treated with antiangiogenics [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 1924.

Posterior Reversible Encephalopathy Syndrome Following Chemotherapy and Immune Checkpoint Inhibitor Combination in a Patient with Small-Cell Lung Cancer

Posterior reversible encephalopathy syndrome (PRES) is a rare neurological complication that occurs following a sudden blood pressure increase. We report the case of a 64-year-old patient presenting PRES several hours after the administration of a combination of chemotherapy and a checkpoint inhibitor (carboplatin-etoposide-atezolizumab) for small-cell lung cancer. He presented consciousness disorders associated with partial epileptic seizure secondarily generalized. His arterial blood pressure was elevated and brain imaging showed multiple bilateral subcortical parietal, temporal, occipital and cerebellar T2 high signals, predominantly in the posterior region. There were no abnormal T1 signals nor bleeding but a left apparent diffusion coefficient restriction was noted. On arterial spin labelling perfusion sequences, there was an increased perfusion within the left temporo-parieto-occipital, left thalamic and right cerebellar regions. Finally, the neurological symptoms completely regressed after several days of optimal antihypertensive and antiepileptic treatment. The clinical context and radiological features, as well as the progressive resolution of the neurological symptoms, were all in favor of PRES. PRES can occur after the administration of chemotherapy and/or immunotherapy. Prompt diagnosis is crucial through a spectrum of suspicious clinical and radiological characteristics that must be rapidly recognized to quickly anticipate the optimal therapeutic strategy and avoid unnecessary complications.

Prediction of Early Response to Immunotherapy: DCE-US as a New Biomarker

Simple Summary

Immune checkpoint inhibitors (ICI) have revolutionized cancer care. However, assessing the efficacy of these new molecules with targeted therapeutic responses may induce too much delay when using classical biomarkers derived from morphological imaging (CT). The objective of our study is to propose fast, cost-effective, convenient, and effective biomarkers using the perfusion parameters from dynamic contrast-enhanced ultrasound (DCE-US) for the evaluation of ICI early response. In a population of 63 patients with metastatic cancer eligible for immunotherapy, we demonstrate that a decrease of more than 45% in the area under the perfusion curve (AUC) between baseline and day 21 is significantly associated with better overall survival. Thus, AUC from DCE-US looks to be a promising new biomarker for the early evaluation of response to immunotherapy.

Abstract

Purpose: The objective of our study is to propose fast, cost-effective, convenient, and effective biomarkers using the perfusion parameters from dynamic contrast-enhanced ultrasound (DCE-US) for the evaluation of immune checkpoint inhibitors (ICI) early response. Methods: The retrospective cohort used in this study included 63 patients with metastatic cancer eligible for immunotherapy. DCE-US was performed at baseline, day 8 (D8), and day 21 (D21) after treatment onset. A tumor perfusion curve was modeled on these three dates, and change in the seven perfusion parameters was measured between baseline, D8, and D21. These perfusion parameters were studied to show the impact of their variation on the overall survival (OS). Results: After the removal of missing or suboptimal DCE-US, the Baseline-D8, the Baseline-D21, and the D8-D21 groups included 37, 53, and 33 patients, respectively. A decrease of more than 45% in the area under the perfusion curve (AUC) between baseline and D21 was significantly associated with better OS (p = 0.0114). A decrease of any amount in the AUC between D8 and D21 was also significantly associated with better OS (p = 0.0370). Conclusion: AUC from DCE-US looks to be a promising new biomarker for fast, effective, and convenient immunotherapy response evaluation.

Can Deep Learning Replace Gadolinium in Neuro-Oncology?: A Reader Study

MATERIALS AND METHODS

This monocentric retrospective study leveraged 200 multiparametric brain MRIs acquired between November 2019 and February 2020 at Gustave Roussy Cancer Campus (Villejuif, France). A total of 145 patients were included: 107 formed the training sample (55 ± 14 years, 58 women) and 38 the separate test sample (62 ± 12 years, 22 women). Patients had glioma, brain metastases, meningioma, or no enhancing lesion. T1, T2-FLAIR, diffusion-weighted imaging, low-dose, and standard-dose postcontrast T1 sequences were acquired. A deep network was trained to process the precontrast and low-dose sequences to predict "virtual" surrogate images for contrast-enhanced T1. Once trained, the deep learning method was evaluated on the test sample. The discrepancies between the predicted virtual images and the standard-dose MRIs were qualitatively and quantitatively evaluated using both automated voxel-wise metrics and a reader study, where 2 radiologists graded image qualities and marked all visible enhancing lesions.

RESULTS

The automated analysis of the test brain MRIs computed a structural similarity index of 87.1% ± 4.8% between the predicted virtual sequences and the reference contrast-enhanced T1 MRIs, a peak signal-to-noise ratio of 31.6 ± 2.0 dB, and an area under the curve of 96.4% ± 3.1%. At Youden's operating point, the voxel-wise sensitivity (SE) and specificity were 96.4% and 94.8%, respectively. The reader study found that virtual images were preferred to standard-dose MRI in terms of image quality (P = 0.008). A total of 91 reference lesions were identified in the 38 test T1 sequences enhanced with full dose of contrast agent. On average across readers, the brain lesion SE of the virtual images was 83% for lesions larger than 10 mm (n = 42), and the associated false detection rate was 0.08 lesion/patient. The corresponding positive predictive value of detected lesions was 92%, and the F1 score was 88%. Lesion detection performance, however, dropped when smaller lesions were included: average SE was 67% for lesions larger than 5 mm (n = 74), and 56% with all lesions included regardless of their size. The false detection rate remained below 0.50 lesion/patient in all cases, and the positive predictive value remained above 73%. The composite F1 score was 63% at worst.

CONCLUSIONS

The proposed deep learning method for virtual contrast-enhanced T1 brain MRI prediction showed very high quantitative performance when evaluated with standard voxel-wise metrics. The reader study demonstrated that, for lesions larger than 10 mm, good detection performance could be maintained despite a 4-fold division in contrast agent usage, unveiling a promising avenue for reducing the gadolinium exposure of returning patients. Small lesions proved, however, difficult to handle for the deep network, showing that full-dose injections remain essential for accurate first-line diagnosis in neuro-oncology.

Prolonged SARS-CoV-2 RNA virus shedding and lymphopenia are hallmarks of COVID-19 in cancer patients with poor prognosis

Patients with cancer are at higher risk of severe coronavirus infectious disease 2019 (COVID-19), but the mechanisms underlying virus-host interactions during cancer therapies remain elusive. When comparing nasopharyngeal swabs from cancer and noncancer patients for RT-qPCR cycle thresholds measuring acute respiratory syndrome coronavirus-2 (SARS-CoV-2) in 1063 patients (58% with cancer), we found that malignant disease favors the magnitude and duration of viral RNA shedding concomitant with prolonged serum elevations of type 1 IFN that anticorrelated with anti-RBD IgG antibodies. Cancer patients with a prolonged SARS-CoV-2 RNA detection exhibited the typical immunopathology of severe COVID-19 at the early phase of infection including circulation of immature neutrophils, depletion of nonconventional monocytes, and a general lymphopenia that, however, was accompanied by a rise in plasmablasts, activated follicular T-helper cells, and non-naive Granzyme B+FasL+, EomeshighTCF-1high, PD-1+CD8+ Tc1 cells. Virus-induced lymphopenia worsened cancer-associated lymphocyte loss, and low lymphocyte counts correlated with chronic SARS-CoV-2 RNA shedding, COVID-19 severity, and a higher risk of cancer-related death in the first and second surge of the pandemic. Lymphocyte loss correlated with significant changes in metabolites from the polyamine and biliary salt pathways as well as increased blood DNA from Enterobacteriaceae and Micrococcaceae gut family members in long-term viral carriers. We surmise that cancer therapies may exacerbate the paradoxical association between lymphopenia and COVID-19-related immunopathology, and that the prevention of COVID-19-induced lymphocyte loss may reduce cancer-associated death.

New method for quantification of intratumoral heterogeneity: a feasibility study on Ktrans maps from preclinical DCE-MRI

OBJECT

To develop new imaging biomarkers of therapeutic efficacy through the quantification of intratumoral microvascular heterogeneity.

MATERIALS AND METHODS

The described method was a combination of non-supervised clustering and extraction of intratumoral complexity features (ICF): number of non-connected objects, volume fraction. It was applied to a set of 3D DCE-MRI K^trans maps acquired previously on tumor bearing mice prior to and on day 4 of anti-angiogenic treatment. Evolutions of ICF were compared to conventional summary statistics (CSS) and to heterogeneity related whole tumor texture features (TF) on treated (n = 9) and control (n = 6) mice.

RESULTS

Computed optimal number of clusters per tumor was 4. Several intratumoral features extracted from the clusters were able to monitor a therapy effect. Whereas no feature significantly changed for the control group, 6 features significantly changed for the treated group (4 ICF, 2 CSS). Among these, 5 also significantly differentiated the two groups (3 ICF, 2 CSS). TF failed in demonstrating differences within and between the two groups.

DISCUSSION

ICF are potential imaging biomarkers for anti-angiogenic therapy assessment. The presented method may be expected to have advantages with respect to texture analysis-based methods regarding interpretability of results and setup of standardized image analysis protocols.

A Predictive Clinical-Radiomics Nomogram for Survival Prediction of Glioblastoma Using MRI

Glioblastoma (GBM) is the most common and aggressive primary brain tumor in adult patients with a median survival of around one year. Prediction of survival outcomes in GBM patients could represent a huge step in treatment personalization. The objective of this study was to develop machine learning (ML) algorithms for survival prediction of GBM patient. We identified a radiomic signature on a training-set composed of data from the 2019 BraTS challenge (210 patients) from MRI retrieved at diagnosis. Then, using this signature along with the age of the patients for training classification models, we obtained on test-sets AUCs of 0.85, 0.74 and 0.58 (0.92, 0.88 and 0.75 on the training-sets) for survival at 9-, 12- and 15-months, respectively. This signature was then validated on an independent cohort of 116 GBM patients with confirmed disease relapse for the prediction of patients surviving less or more than the median OS of 22 months. Our model insured an AUC of 0.71 (0.65 on train). The Kaplan–Meier method showed significant OS difference between groups (log-rank p = 0.05). These results suggest that radiomic signatures may improve survival outcome predictions in GBM thus creating a solid clinical tool for tailoring therapy in this population.

Incorporating radiomics into clinical trials: expert consensus endorsed by the European Society of Radiology on considerations for data-driven compared to biologically driven quantitative biomarkers

Existing quantitative imaging biomarkers (QIBs) are associated with known biological tissue characteristics and follow a well-understood path of technical, biological and clinical validation before incorporation into clinical trials. In radiomics, novel data-driven processes extract numerous visually imperceptible statistical features from the imaging data with no a priori assumptions on their correlation with biological processes. The selection of relevant features (radiomic signature) and incorporation into clinical trials therefore requires additional considerations to ensure meaningful imaging endpoints. Also, the number of radiomic features tested means that power calculations would result in sample sizes impossible to achieve within clinical trials. This article examines how the process of standardising and validating data-driven imaging biomarkers differs from those based on biological associations. Radiomic signatures are best developed initially on datasets that represent diversity of acquisition protocols as well as diversity of disease and of normal findings, rather than within clinical trials with standardised and optimised protocols as this would risk the selection of radiomic features being linked to the imaging process rather than the pathology. Normalisation through discretisation and feature harmonisation are essential pre-processing steps. Biological correlation may be performed after the technical and clinical validity of a radiomic signature is established, but is not mandatory. Feature selection may be part of discovery within a radiomics-specific trial or represent exploratory endpoints within an established trial; a previously validated radiomic signature may even be used as a primary/secondary endpoint, particularly if associations are demonstrated with specific biological processes and pathways being targeted within clinical trials. KEY POINTS: • Data-driven processes like radiomics risk false discoveries due to high-dimensionality of the dataset compared to sample size, making adequate diversity of the data, cross-validation and external validation essential to mitigate the risks of spurious associations and overfitting. • Use of radiomic signatures within clinical trials requires multistep standardisation of image acquisition, image analysis and data mining processes. • Biological correlation may be established after clinical validation but is not mandatory.

Three artificial intelligence data challenges based on CT and ultrasound

PURPOSE

The 2020 edition of these Data Challenges was organized by the French Society of Radiology (SFR), from September 28 to September 30, 2020. The goals were to propose innovative artificial intelligence solutions for the current relevant problems in radiology and to build a large database of multimodal medical images of ultrasound and computed tomography (CT) on these subjects from several French radiology centers.

MATERIALS AND METHODS

This year the attempt was to create data challenge objectives in line with the clinical routine of radiologists, with less preprocessing of data and annotation, leaving a large part of the preprocessing task to the participating teams. The objectives were proposed by the different organizations depending on their core areas of expertise. A dedicated platform was used to upload the medical image data, to automatically anonymize the uploaded data.

RESULTS

Three challenges were proposed including classification of benign or malignant breast nodules on ultrasound examinations, detection and contouring of pathological neck lymph nodes from cervical CT examinations and classification of calcium score on coronary calcifications from thoracic CT examinations. A total of 2076 medical examinations were included in the database for the three challenges, in three months, by 18 different centers, of which 12% were excluded. The 39 participants were divided into six multidisciplinary teams among which the coronary calcification score challenge was solved with a concordance index > 95%, and the other two with scores of 67% (breast nodule classification) and 63% (neck lymph node calcifications).

Automatic coronary artery calcium scoring from unenhanced-ECG-gated CT using deep learning

PURPOSE

The purpose of this study was to develop and evaluate an algorithm that can automatically estimate the amount of coronary artery calcium (CAC) from unenhanced electrocardiography (ECG)-gated computed tomography (CT) cardiac volume acquisitions by using convolutional neural networks (CNN).

MATERIALS AND METHODS

The method used a set of five CNN with three-dimensional (3D) U-Net architecture trained on a database of 783 CT examinations to detect and segment coronary artery calcifications in a 3D volume. The Agatston score, the conventional CAC scoring, was then computed slice by slice from the resulting segmentation mask and compared to the ground truth manually estimated by radiologists. The quality of the estimation was assessed with the concordance index (C-index) on CAC risk category on a separate testing set of 98 independent CT examinations.

RESULTS

The final model yielded a C-index of 0.951 on the testing set. The remaining errors of the method were mainly observed on small-size and/or low-density calcifications, or calcifications located near the mitral valve or ring.

CONCLUSION

The deep learning-based method proposed here to compute automatically the CAC score from unenhanced-ECG-gated cardiac CT is fast, robust and yields accuracy similar to those of other artificial intelligence methods, which could improve workflow efficiency, eliminating the time spent on manually selecting coronary calcifications to compute the Agatston score.

Automatic cervical lymphadenopathy segmentation from CT data using deep learning

PURPOSE

The purpose of this study was to develop a fast and automatic algorithm to detect and segment lymphadenopathy from head and neck computed tomography (CT) examination.

MATERIALS AND METHODS

An ensemble of three convolutional neural networks (CNNs) based on a U-Net architecture were trained to segment the lymphadenopathies in a fully supervised framework. The resulting predictions were assessed using the Dice similarity coefficient (DSC) on examinations presenting one or more adenopathies. On examinations without adenopathies, the score was given by the formula M/(M+A) where M was the mean adenopathy volume per patient and A the volume segmented by the algorithm. The networks were trained on 117 annotated CT acquisitions.

RESULTS

The test set included 150 additional CT acquisitions unseen during the training. The performance on the test set yielded a mean score of 0.63.

CONCLUSION

Despite limited available data and partial annotations, our CNN based approach achieved promising results in the task of cervical lymphadenopathy segmentation. It has the potential to bring precise quantification to the clinical workflow and to assist the clinician in the detection task.

Comorbidity and frailty assessment in renal cell carcinoma patients

PURPOSE

Renal cell carcinoma (RCC) incidence has considerably increased during the last decades without any real impact on age-standardized mortality. It questions the relevance of aggressive treatments carrying potential side effects. Conservative management should be considered for frail patients. Comorbidity and frailty assessment in RCC patients is paramount before engaging a treatment.

METHODS

Narrative, non-systematic review based on PubMed and EMBASE search with the terms "renal neoplasm", "elderly, frail", "comorbidities", "active surveillance", "metastatic". The selection was restricted to articles written in English.

RESULTS

Comorbidity and frailty assessment go along with the cancer-specific aggressivity and intervention risks assessment. In localized disease, several standardized algorithms offer patient health evaluation to define how suitable the patient would be for curative treatment. The pre-operative American Society of Anesthesiologists and the age-adjusted Charlson's scores are the most widely used. At the metastatic stage, drug combinations based on immunotherapies and targeted therapies improved cancer outcomes at the price of significant toxicities. Frail patients are not always suitable for such strategies. Commonly used scores like the International Metastatic RCC Database Consortium or Memorial Sloan Kettering Cancer Center integrate features to define patients' risk groups, more specifically the Karnofsky Performance Score is an easy way to document the frailty.

CONCLUSIONS

Comorbidity and frailty have to be assessed at any stage of the RCC disease based on a standardized scoring system to define the most suitable treatment strategy ranging from surveillance to aggressive treatment.

Integrating deep learning CT-scan model, biological and clinical variables to predict severity of COVID-19 patients

The SARS-COV-2 pandemic has put pressure on intensive care units, so that identifying predictors of disease severity is a priority. We collect 58 clinical and biological variables, and chest CT scan data, from 1003 coronavirus-infected patients from two French hospitals. We train a deep learning model based on CT scans to predict severity. We then construct the multimodal AI-severity score that includes 5 clinical and biological variables (age, sex, oxygenation, urea, platelet) in addition to the deep learning model. We show that neural network analysis of CT-scans brings unique prognosis information, although it is correlated with other markers of severity (oxygenation, LDH, and CRP) explaining the measurable but limited 0.03 increase of AUC obtained when adding CT-scan information to clinical variables. Here, we show that when comparing AI-severity with 11 existing severity scores, we find significantly improved prognosis performance; AI-severity can therefore rapidly become a reference scoring approach.

Clarification of Definitions of Hyperprogressive Disease During Immunotherapy for Non-Small Cell Lung Cancer

Importance

Hyperprogressive disease (HPD) is an aggressive pattern of progression reported for patients treated with programmed cell death 1 (PD-1)/programmed cell death 1 ligand (PD-L1) inhibitors as a single agent in several studies. However, the use of different definitions of HPD introduces the risk of describing different tumoral behaviors.

Objective

To assess the accuracy of each HPD definition to identify the frequency of HPD and the association with poorer outcomes of immune-checkpoint inhibitor (ICI) treatment in patients with advanced non-small cell lung cancer (NSCLC) and to provide an optimized and homogenized definition based on all previous criteria for identifying HPD.

Design, Setting, and Participants

This retrospective cohort study included 406 patients with advanced NSCLC treated with PD-1/PD-L1 inhibitors from November 1, 2012, to April 5, 2017, in 8 French institutions. Measurable lesions were defined using the Response Evaluation Criteria in Solid Tumors (RECIST) 1.1 criteria on at least 2 computed tomographic scans before the initiation of ICI therapy and 1 computed tomographic scan during treatment. Data were analyzed from November 1, 2012, to August 1, 2019.

Exposures

Advanced NSCLC and treatment with PD-1/PD-L1 inhibitors.

Main Outcomes and Measures

Association of the definition with the related incidence and the HPD subset constitution and the association between each HPD definition and overall survival. All dynamic indexes used in the previous proposed definitions, such as the tumor growth rate (TGR) or tumor growth kinetics (TGK), were calculated before and during treatment.

Results

Among the 406 patients with NSCLC included in the analysis (259 male [63.8%]; median age at start of ICI treatment, 64 [range, 30-91] years), the different definitions resulted in incidences of the HPD phenomenon varying from 5.4% (n = 22; definition based on a progression pace >2-fold and a time to treatment failure of <2 months) to 18.5% (n = 75; definition based on the TGR ratio). The concordance between these different definitions (using the Jaccard similarity index) varied from 33.3% to 69.3%. For every definition, HPD was associated with poorer survival (range of median overall survival, 3.4 [95% CI, 1.9-8.4] to 6.0 [95% CI, 3.7-9.4] months). The difference between TGR before and during therapy (ΔTGR) was the most correlated with poor overall survival with an initial plateau for a larger number of patients and a slower increase, and it had the highest ability to distinguish patients with HPD from those with progressive disease not classified as HPD. In addition, an optimal threshold of ΔTGR of greater than 100 was identified for this distinction.

Conclusions and Relevance

The findings of this retrospective cohort study of patients with NSCLC suggest that the previous 5 definitions of HPD were not associated with the same tumor behavior. A new definition, based on ΔTGR of greater than 100, appeared to be associated with the characteristics expected with HPD (increase of the tumor kinetics and poor survival). Additional studies on larger groups of patients are necessary to confirm the accuracy and validate this proposed definition.

Influence of Magnetic Field Strength on Magnetic Resonance Imaging Radiomics Features in Brain Imaging, an In Vitro and In Vivo Study

BACKGROUND

The development and clinical adoption of quantitative imaging biomarkers (radiomics) has established the need for the identification of parameters altering radiomics reproducibility. The aim of this study was to assess the impact of magnetic field strength on magnetic resonance imaging (MRI) radiomics features in neuroradiology clinical practice.

METHODS

T1 3D SPGR sequence was acquired on two phantoms and 10 healthy volunteers with two clinical MR devices from the same manufacturer using two different magnetic fields (1.5 and 3T). Phantoms varied in terms of gadolinium concentrations and textural heterogeneity. 27 regions of interest were segmented (phantom: 21, volunteers: 6) using the LIFEX software. 34 features were analyzed.

RESULTS

In the phantom dataset, 10 (67%) out of 15 radiomics features were significantly different when measured at 1.5T or 3T (student's t-test, p < 0.05). Gray levels resampling, and pixel size also influence part of texture features. These findings were validated in healthy volunteers.

CONCLUSIONS

According to daily used protocols for clinical examinations, radiomic features extracted on 1.5T should not be used interchangeably with 3T when evaluating texture features. Such confounding factor should be adjusted when adapting the results of a study to a different platform, or when designing a multicentric trial.

Artificial intelligence solution to classify pulmonary nodules on CT

PURPOSE

The purpose of this study was to create an algorithm to detect and classify pulmonary nodules in two categories based on their volume greater than 100 mm³ or not, using machine learning and deep learning techniques.

MATERIALS AND METHOD

The dataset used to train the model was provided by the organization team of the SFR (French Radiological Society) Data Challenge 2019. An asynchronous and parallel 3-stages pipeline was developed to process all the data (a data "pre-processing" stage; a "nodule detection" stage; a "classifier" stage). Lung segmentation was achieved using 3D U-NET algorithm; nodule detection was done using 3D Retina-UNET and classifier stage with a support vector machine algorithm on selected features. Performances were assessed using area under receiver operating characteristics curve (AUROC).

RESULTS

The pipeline showed good performance for pathological nodule detection and patient diagnosis. With the preparation dataset, an AUROC of 0.9058 (95% confidence interval [CI]: 0.8746-0.9362) was obtained, 87% yielding accuracy (95% CI: 84.83%-91.03%) for the "nodule detection" stage, corresponding to 86% specificity (95% CI: 82%-92%) and 89% sensitivity (95% CI: 84.83%-91.03%).

CONCLUSION

A fully functional pipeline using 3D U-NET, 3D Retina-UNET and classifier stage with a support vector machine algorithm was developed, resulting in high capabilities for pulmonary nodule classification.

Methodological Study to Investigate the Potential of Ultrasound-Based Elastography and Texture as Biomarkers to Monitor Liver Tumors

AIMS AND OBJECTIVES

In order to evaluate the responses of hepatic lesions to treatment in terms of tissue stiffness and heterogeneity, this work investigated the robustness of 2D shear-wave elastography (2D SWE) stiffness measurements and texture analyses in vitro and in vivo in terms of repeatability and variability.

METHODS AND MATERIALS

A multioperator (n = 5) study was performed with an ultrasonic elastography device on two sets of phantoms. For the first set of phantoms, 10 measurements for each of the eight inclusions were performed by each observer, whereas the second set of phantoms was used to evaluate the influence of depth on the stiffness measurements. Variability of the stiffness measurements was evaluated in vivo on 10 healthy livers, with 10 measurements for each hepatic segment. Texture analyses were performed in B-mode, obtaining elastography images for every hepatic segment.

RESULTS

Stiffness measurements were influenced by depth, particularly when exceeding 7 cm. In vivo measurements demonstrated that measurements of segments I, VII, and VIII were less reliable, mainly due to their deeper locations. The protocols used were more flexible in terms of acquisition setup and probe placement than those currently used with Fibroscan®. For texture analysis on the B-mode images, 12 features showed low variability regardless of the evaluated hepatic segment. On elastogram, only two features showed low variability, but not in every segment.

CONCLUSION

We demonstrated the robustness of two methodologies for the quantification of liver stiffness and heterogeneity. Further clinical studies should evaluate whether these techniques can assess tumor responses to treatment and, therefore, have the potential to be used as imaging biomarkers.

Guidelines and Good Clinical Practice Recommendations for Contrast Enhanced Ultrasound (CEUS) in the Liver - Update 2020 - WFUMB in Cooperation with EFSUMB, AFSUMB, AIUM, and FLAUS

The present, updated document describes the fourth iteration of recommendations for the hepatic use of contrast enhanced ultrasound (CEUS), first initiated in 2004 by the European Federation of Societies for Ultrasound in Medicine and Biology (EFSUMB). The previous updated editions of the guidelines reflected changes in the available contrast agents and updated the guidelines not only for hepatic but also for non-hepatic applications.The 2012 guideline requires updating as previously the differences of the contrast agents were not precisely described and the differences in contrast phases as well as handling were not clearly indicated. In addition, more evidence has been published for all contrast agents. The update also reflects the most recent developments in contrast agents, including the United States Food and Drug Administration (FDA) approval as well as the extensive Asian experience, to produce a truly international perspective.These guidelines and recommendations provide general advice on the use of ultrasound contrast agents (UCA) and are intended to create standard protocols for the use and administration of UCA in liver applications on an international basis to improve the management of patients.

Guidelines and Good Clinical Practice Recommendations for Contrast-Enhanced Ultrasound (CEUS) in the Liver-Update 2020 WFUMB in Cooperation with EFSUMB, AFSUMB, AIUM, and FLAUS

The present, updated document describes the fourth iteration of recommendations for the hepatic use of contrast-enhanced ultrasound, first initiated in 2004 by the European Federation of Societies for Ultrasound in Medicine and Biology. The previous updated editions of the guidelines reflected changes in the available contrast agents and updated the guidelines not only for hepatic but also for non-hepatic applications. The 2012 guideline requires updating as, previously, the differences in the contrast agents were not precisely described and the differences in contrast phases as well as handling were not clearly indicated. In addition, more evidence has been published for all contrast agents. The update also reflects the most recent developments in contrast agents, including U.S. Food and Drug Administration approval and the extensive Asian experience, to produce a truly international perspective. These guidelines and recommendations provide general advice on the use of ultrasound contrast agents (UCAs) and are intended to create standard protocols for the use and administration of UCAs in liver applications on an international basis to improve the management of patients.

Carbonic Anhydrase IX in Renal Cell Carcinoma, Implications for Disease Management

Carbonic Anhydrase IX (CAIX) is a well-described enzyme in renal cell carcinoma, with its expression being regulated by the hypoxia-inducible factor 1 alpha, it is known for interfering with hypoxia processes. Renal carcinoma encompasses a broad spectrum of histological entities and is also described as a heterogeneous malignant tumor. Recently, various combinations of checkpoint inhibitors and targeted therapies have been validated to manage this disease. Reliable markers to confirm the diagnosis, estimate the prognosis, predict or monitor the treatment response are required. Molecular imaging developments allow a comprehensive analysis of the tumor, overcoming the spatial heterogeneity issue. CAIX, being highly expressed at the tumor cell surfaces of clear cell renal carcinoma, also represents a potential treatment target. In this manuscript we reviewed the current knowledge from the literature on the pathophysiological interactions between renal cell carcinoma and CAIX, the role of CAIX as a marker for diagnosis, prognosis, treatment monitoring and molecular imaging, and the potential target for therapeutic strategies.

Prediction of the molecular status in non-small cell lung cancer based on metastatic pattern: A free webtool powered by artificial intelligence

9535

Background: Molecular characterization of metastatic lung adenocarcinomas is mandatory but might be hampered by the quantity of tissue, restricted access to molecular platforms or limited economical resources. Our aim was to develop a tool supported by the hypothesis that radiological patterns of pts could help predict the rate of positivity of the most common oncogenic drivers. Methods: We defined an algorithm based on a molecularly defined cohort of 656 pts with stage IV lung adenocarcinoma. Two radiologists centrally reviewed the baseline imaging. Clinical data were retrospectively collected. There were 135 EGFR mutations, 81 ALK fusions, 47 BRAF mutations, 141 KRAS mutations, and 146 pan-negative tumors for these 4 oncogenic drivers. Univariate correlation analyses were performed to define an algorithm predicting the molecular testing positivity based on the metastatic pattern. Subsequently, an online tool was developed. This study was approved by our institutional review board. Results: Metastatic patterns correlated with the genomic drivers when compared to the pan-negative group. In the EGFR group, pleural metastases were more frequent (32% vs. 20%; p = 0.021), whereas adrenal and node metastases less frequent (6% vs.23%; p < 0.001 and 11% vs. 23% respectively; p = 0.011). In the ALK group, there were more brain and lung metastases (respectively 42% vs. 29%; p = 0.043 and 37% vs. 24% respectively; p = 0.037). In the BRAF group, pleural and pericardial metastases were more common (47% vs. 20%; p < 0.001 and 11% vs. 3% respectively; p = 0.04) and bone metastases less common (21% vs. 42%; p = 0.011). Lymphangitis was more frequent in EGFR, ALK and BRAF groups (6%, 7% and 15% vs. 1%; p = 0,016, p = 0,009 and p < 0,001 respectively). A free online access to the algorithm is now available after registration at http//tactic-ct.fr. Physicians enter age, sex, smoking status and the sites of metastases at diagnosis (present/absent/unknown). A mutation score is calculated, reflecting the % of chance to find an oncogenic driver. On the website, contributors can also enter new cases and an artificial intelligence will refine the algorithm and expand the number of oncogenic drivers. Conclusions: Our free access tool allows establishing a hierarchy in the molecular testing based on simple clinical and radiological information. Continual learning from new cases entered in the database will increase the sensitivity of the tool. This tool might save time, tumor tissue, economical resources and accelerate access to personalized treatment.

Association of metastatic pattern and molecular status in stage IV non-small cell lung cancer adenocarcinoma

OBJECTIVES

The aim of our study was to investigate the association between driver oncogene alterations and metastatic patterns on imaging assessment, in a large cohort of metastatic lung adenocarcinoma patients.

METHODS

From January 2010 to May 2017, 550 patients with stage IV lung adenocarcinoma with molecular analysis were studied retrospectively including 135 EGFR-mutated, 81 ALK-rearrangement, 47 BRAF-mutated, 141 KRAS-mutated, and 146 negative tumors for these 4 mutations (4N). After review of the complete imaging report by two radiologists (junior and senior) to identify metastatic sites, univariate correlation analyzes were performed.

RESULTS

We found differences in metastatic tropism depending on the molecular alteration type when compared with the non-mutated 4N group: in the EGFR group, pleural metastases were more frequent (32% versus 20%; p = 0.021), and adrenal and node metastases less common (6% versus 23%; p < 0.001 and 11% versus 23%; p = 0.011). In the ALK group, there were more brain and lung metastases (respectively 42% versus 29%; p = 0.043 and 37% versus 24%; p = 0.037). In the BRAF group, pleural and pericardial metastases were more common (respectively 47% versus 20%; p < 0.001 and 11% versus 3%; p = 0.04) and bone metastases were rarer (21% versus 42%; p = 0.011). Lymphangitis was more frequent in EGFR, ALK, and BRAF groups (respectively 6%, 7%, and 15% versus 1%); p = 0.016; p = 0.009; and p < 0.001.

CONCLUSION

The application of these correlations between molecular status and metastatic tropism in clinical practice may lead to earlier and more accurate identification of patients for targeted therapy.

KEY POINTS

• Bone and brain metastasis are the most common organs involved in lung adenocarcinoma but the relative incidence of each metastatic site depends on the molecular alteration. • EGFR-mutated tumors preferentially spread to the pleura and less commonly to adrenals, ALK-rearrangement tumors usually spread to the brain and the lungs, whereas BRAF-mutated tumors are unlikely to spread to bones and have a serous (pericardial ad pleural) tropism. • These correlations could help in the clinical management of patients with metastatic lung adenocarcinoma.

Advanced Ultrasound Imaging for Patients in Oncology: DCE-US

Neovascularization is a key stage in the growth of malignancies beyond 2-3 mm³. This neoangiogenesis is an important target for novel anticancer treatments [1], and many new antiangiogenesis or antivascular treatments aim at destroying or limiting the growth of tumor vessels [2].

Pravastatin Reverses Established Radiation-Induced Cutaneous and Subcutaneous Fibrosis in Patients With Head and Neck Cancer: Results of the Biology-Driven Phase 2 Clinical Trial Pravacur

PURPOSE

The "PRAVACUR" phase 2 trial (NCT01268202) assessed the efficacy of pravastatin as an antifibrotic agent in patients with established cutaneous and subcutaneous radiation-induced fibrosis (RIF) after head and neck squamous cell carcinoma (HNSCC) radiation therapy and/or radiochemotherapy.

METHODS AND MATERIALS

The main inclusion criteria were: NSCC in remission, grade ≥2 cutaneous and subcutaneous neck RIF (National Cancer Institute Common Terminology Criteria for Adverse Events, version 4.0), and no current treatment with statins or fibrates. Patients received pravastatin 40 mg/d for 12 months. The primary endpoint was reduction of RIF thickness by more than 30% at 12 months, as measured by cutaneous high-frequency ultrasonography. Secondary endpoints included RIF severity reduction, pravastatin tolerance, and quality of life.

RESULTS

Sixty patients with grade 2 (n = 37), grade 3 (n = 22), or grade 4 (n = 1) RIF were enrolled from February 2011 to April 2016. The mean interval between RIF diagnosis and pravastatin initiation was 17.1 months. Pravastatin was stopped before 11 months of treatment in 18 patients (because of grade ≥2 adverse events related to pravastatin in 8 patients [13%]). In the 40 patients in whom pravastatin efficacy was assessed by high-frequency ultrasonography at baseline and at 12 months of treatment, a reduction of RIF thickness ≥30% was observed in 15 of 42 patients (35.7%; 95% confidence interval, 21.6%-52.0%). At the 12-month clinical evaluation, RIF severity was decreased in 50% of patients (n = 21; 95% confidence interval, 34.2%-65.8%), and the patients' self-perception, mood state, and social functioning were significantly improved. Pravastatin was well tolerated, with a very low occurrence of grade 3 toxicities (myalgia, n = 1) and grade 2 toxicities (myalgia/arthralgia or esophagitis, n = 3).

CONCLUSIONS

This phase 2 prospective study supports the notion of radioinduced fibrosis reversibility. It showed that pravastatin (40 mg/d for 12 months) is an efficient antifibrotic agent in patients with grade ≥2 cutaneous and subcutaneous fibrosis after HNSCC radiation therapy.

New Ultrasound Techniques Challenge the Diagnosis of Sinusoidal Obstruction Syndrome

Sinusoidal obstruction syndrome, also known as veno-occlusive disease (SOS/VOD), is a potentially life-threatening complication that can develop after hematopoietic cell transplantation. Clinically, SOS/VOD is characterized by hepatomegaly, right upper quadrant pain, jaundice and ascites, most often occurring within the first 3 wk after hematopoietic cell transplantation. Early therapeutic intervention is pivotal for survival in SOS/VOD. Thus, a rapid and reliable diagnosis has to be made. Diagnosis of SOS/VOD is based on clinical criteria, such as the Seattle, Baltimore or recently issued European Society for Blood and Marrow Transplantation criteria, to which hemodynamic and/or ultrasound evidence of SOS were added for the first time. However, to rule out major differential diagnoses and to verify the diagnosis, a reliable imaging method is needed. Ultrasound techniques have been proposed in SOS/VOD. Nevertheless, the sensitivity and specificity of transabdominal ultrasound and Doppler techniques need to be improved. Innovative ultrasound methods such as a combination of Doppler ultrasound with shear wave elastography and contrast-enhanced ultrasound techniques should be evaluated for diagnosis and follow-up of SOS/VOD. The goals of this review are to discuss currently available ultrasound techniques and to identify areas for future studies in SOS/VOD.