First Experience With Augmented Reality Guidance for Cerebral Embolic Protection During TAVR.

Background: Augmented reality (AR) guidance holds potential to improve transcatheter interventions by enabling visualization of and interaction with patient-specific 3-dimensional virtual content. Positioning of cerebral embolic protection devices (CEP) during transcatheter aortic valve replacement (TAVR) increases patient exposure to radiation and iodinated contrast, and increases procedure time. AR may enhance procedural guidance and facilitate a safer intervention. Objectives: The purpose of this study was to develop and test a novel AR guidance system with a custom user interface that displays virtual, patient-specific 3-dimensional anatomic models, and assess its intraprocedural impact during CEP placement in TAVR. Methods: Patients undergoing CEP during TAVR were prospectively enrolled and assigned to either AR guidance or control groups. Primary endpoints were contrast volume used prior to filter placement, times to filter placement, and fluoroscopy time. Postprocedure questionnaires were administered to assess intraprocedural physician experience with AR guidance. Results: A total of 24 patients presenting for TAVR were enrolled in the study (12 with AR guidance and 12 controls). AR guidance eliminated the need for aortic arch angiograms prior to device placement thus reducing contrast volume (0 mL vs 15 mL, P < 0.0001). There was no significant difference in the time required for filter placement or fluoroscopy time. Postprocedure questionnaires indicated that AR guidance increased confidence in wiring of the aortic arch and facilitated easier device placement. Conclusions: We developed a novel AR guidance system that eliminated the need for additional intraprocedural angiograms prior to device placement without any significant difference in time to intervention and offered a subjective improvement in performance of the intervention.

OptimICE-RD: sacituzumab govitecan + pembrolizumab vs pembrolizumab (± capecitabine) for residual triple-negative breast cancer.

Patients with early-stage triple-negative breast cancer (TNBC) with residual invasive disease after neoadjuvant therapy have a high risk of recurrence even with neoadjuvant and adjuvant treatment with pembrolizumab. Sacituzumab govitecan, a Trop-2-directed antibody-drug conjugate with a topoisomerase I inhibitor payload, improved progression-free survival (PFS) and overall survival (OS) versus chemotherapy in patients with pre-treated metastatic TNBC. Moreover, preclinical data suggest that topoisomerase I inhibitors may enhance the effects of immune checkpoint inhibitors through activation of the cGAS-STING pathway. Here we describe the international randomized phase III AFT-65/ASCENT-05/OptimICE-RD trial, which evaluates the efficacy and safety of sacituzumab govitecan plus pembrolizumab versus treatment of physician's choice (pembrolizumab ± capecitabine) among patients with early-stage TNBC with residual invasive disease after neoadjuvant therapy.Clinical Trial Registration: NCT05633654 (ClinicalTrials.gov)Other Study ID Number(s): Gilead Study ID: GS-US-595-6184Registration date: 1 December 2022Study start date: 12 December 2022Recruitment status: Recruiting.

AFT-65/ASCENT-05/OptimICE-RD is an ongoing clinical trial that is testing a new treatment combination for patients with stage II or III triple-negative breast cancer (TNBC). Stage II­III means the cancer is confined to the breast and/or nearby lymph nodes and can be surgically removed. However, there remains a risk that the cancer could recur after surgery. To reduce this risk, patients with stage II­III TNBC receive anti-cancer medication before and after surgery. For some patients, receipt of anti-cancer medication before surgery produces a pathologic complete response (pCR), meaning there is no observable cancer left behind at surgery. Patients with a pCR have a lower risk of recurrence than patients with residual disease.The AFT-65/ASCENT-05/OptimICE-RD trial includes people with stage II-III TNBC who have residual cancer after completing their course of pre-surgery anti-cancer medication. All participants have any remaining cancer in their breast and/or lymph nodes removed surgically, after which they are randomly assigned to receive one of two treatments. The experimental therapy consists of pembrolizumab along with a medication called sacituzumab govitecan, which kills cancer cells directly and may strengthen the anti-cancer immune response. Pembrolizumab strengthens the anti-cancer immune response, so the hypothesis of this trial is that the two medications will be more effective together. The control therapy consists of pembrolizumab, alone or in combination with a chemotherapy medication called capecitabine, which is the current standard of care. To study the effectiveness of each treatment, the researchers are following up with all participants to learn if and when their breast cancer returns.

Burden of Illness and Treatment Patterns Among Patients With von Willebrand Disease in US Clinical Practice.

In this retrospective cohort study, data from an integrated US healthcare system containing both electronic medical record data and linked claims data (from 01/2004 to 12/2020) were used to evaluate the clinical burden, treatment patterns, and healthcare resource use (HRU) in patients with von Willebrand disease (VWD). Two patient cohorts were analyzed: the overall VWD population (n = 396) and a subset of these patients (n = 75) who were considered potentially eligible for prophylaxis treatment with von Willebrand factor (VWF) based on a history of severe and frequent bleeding. HRU (hospitalizations, outpatient visits, and emergency department visits) were measured in patients with linked claims data (n = 110, overall VWD patients; n = 23 potentially VWF-prophylaxis-eligible VWD patients). In general, patients with VWD experienced a substantial burden of bleeding events, comorbidities, and HRU. Patients with VWD who were considered potentially eligible for prophylaxis owing to severe and frequent bleeds suffered from a higher clinical burden and HRU than the overall VWD population, and thus may benefit from VWF prophylactic treatment. The findings from this study could help improve clinical outcomes and manage HRU for patients with VWD.

Outcomes in Patients With von Willebrand Disease Receiving Recombinant von Willebrand Factor on Demand and in Surgical Settings: Chart Review.

This European observational chart review assessed the efficacy/safety of recombinant von Willebrand factor (rVWF) for on-demand treatment of spontaneous/traumatic bleeds and prevention and/or treatment of surgery-related bleeding in adults with von Willebrand disease (VWD). Patients (n = 91) were enrolled at first rVWF administration (index). Data were collected for the 12 months before index and until death, loss to follow-up, or end of study (3-12 months after index). Fifteen patients reported an rVWF-treated spontaneous/traumatic bleed at index. Bleed resolution was obtained for 14 patients (unknown status, n = 1), and investigators assessed treatment satisfaction for 13 rVWF prescriptions (2 moderate, 5 good, and 6 excellent). rVWF was used to prevent/treat surgery-related bleeds at index in 76 patients. Bleed resolution was achieved in 25/58 rVWF-treated surgeries; bleed resolution was not applicable for 33 surgeries. In both groups, there were no reports of treatment-emergent adverse events after initiating rVWF, including hypersensitivity reactions, thrombotic events, and VWF inhibitor development. rVWF was shown to be effective for the on-demand treatment of spontaneous/traumatic bleeds, and for the prevention and treatment of surgical bleeds in this real-world VWD population.

Microsimulation to compare activity-related bleed risks between pharmacokinetic-guided rurioctocog alfa pegol prophylaxis and emicizumab.

BACKGROUND: Owing to a lack of clinical study data, computational modeling was used to estimate activity-related bleed risk during prophylaxis with either rurioctocog alfa pegol or emicizumab. RESEARCH DESIGN AND METHODS: A pharmacokinetic (PK)-based computational model was used to estimate factor VIII (FVIII) levels for individual patients in the PROPEL study who were treated with PK-guided rurioctocog alfa pegol prophylaxis targeting FVIII trough levels of 1-3% or 8-12%. Emicizumab was assumed to have 20% FVIII equivalency. Six hypothetical patient activity profiles were created using the National Hemophilia Foundation's physical activity risk ratings scale. For each profile and treatment regimen combination, the total bleed risk over a 2-week period was estimated relative to a patient who was participating in a low-risk activity with 0% FVIII, and the overall relative bleed risks were compared. RESULTS: For all evaluated activity profiles, rurioctocog alfa pegol prophylaxis targeting either 1-3% or 8-12% FVIII trough levels was associated with a lower estimated bleed risk compared with emicizumab prophylaxis with assumed 20% FVIII equivalency (P < 0.001). CONCLUSION: Although this model does not reflect actual patient outcomes, it suggests that PK-guided rurioctocog alfa pegol prophylaxis may reduce the bleed risk during physical activities.

Emerging and Evolving Concepts in Cancer Immunotherapy Imaging.

Criteria based on measurements of lesion diameter at CT have guided treatment with historical therapies due to the strong association between tumor size and survival. Clinical experience with immune checkpoint modulators shows that editing immune system function can be effective in various solid tumors. Equally, novel immune-related phenomena accompany this novel therapeutic paradigm. These effects of immunotherapy challenge the association of tumor size with response or progression and include risks and adverse events that present new demands for imaging to guide treatment decisions. Emerging and evolving approaches to immunotherapy highlight further key issues for imaging evaluation, such as dissociated response following local administration of immune checkpoint modulators, pseudoprogression due to immune infiltration in the tumor environment, and premature death due to hyperprogression. Research that may offer tools for radiologists to meet these challenges is reviewed. Different modalities are discussed, including immuno-PET, as well as new applications of CT, MRI, and fluorodeoxyglucose PET, such as radiomics and imaging of hematopoietic tissues or anthropometric characteristics. Multilevel integration of imaging and other biomarkers may improve clinical guidance for immunotherapies and provide theranostic opportunities.

Disease Burden in Patients with von Willebrand Disease Potentially Eligible for Prophylaxis: Post Hoc Analysis of a European Cross-Sectional Study.

Recent international guidelines conditionally recommend von Willebrand factor (VWF) prophylaxis for von Willebrand disease (VWD) patients with a history of severe/frequent bleeds. This post hoc analysis of the Cost of VWD Across Europe, a Socioeconomic Study (CVESS; conducted in 2018), assessed patient characteristics and disease burden in patients aged >1 year with congenital VWD not receiving but potentially eligible for prophylaxis based on severe/frequent bleeds, and those receiving prophylaxis in the previous 12 months. Data were collected using medical records and a patient questionnaire. Patients considered potentially prophylaxis-eligible (n = 102) experienced more bleeds than patients receiving prophylaxis (n = 229) and were more likely to be admitted to the hospital due to bleeding events in the prior 12 months. Quality of life and work productivity were similar between the two groups. Logistic regression analysis showed that the prophylaxis-eligible group was more likely to have poor joint function and moderate chronic pain than the prophylaxis group. This retrospective study suggests that 1/7 patients not receiving VWF prophylaxis had a higher disease burden than patients receiving prophylaxis and would potentially benefit from prophylaxis.

Comparison of Real-World Dose and Consumption for Two Extended Half-Life Recombinant Factor VIII Products for the Treatment of Hemophilia A in the United States.

Background: US patients with hemophilia A can receive prophylaxis with extended half-life recombinant factor VIII (rFVIII) products, including efmoroctocog alfa (fragment crystallizable fusion protein) and rurioctocog alfa pegol (antihemophilic factor [recombinant], PEGylated). Objective: To evaluate dosing patterns and weekly consumption of extended half-life rFVIII products in the United States. Methods: We performed a retrospective analysis using the US Specialty Pharmacy Database (2015-2018). Included patients had a diagnosis of hemophilia A, ≥2 consecutive monthly claims for efmoroctocog alfa or rurioctocog alfa pegol for prophylaxis, and weight data. Outcome measures included weekly dosing frequency and dispensed weekly dose. Results: The analysis included 774 patients (efmoroctocog alfa, 506; rurioctocog alfa pegol, 268). Mean (SD) age was 24.2 (15.8) and 26.3 (14.9) years for patients receiving efmoroctocog alfa and rurioctocog alfa pegol, respectively; mean (SD) weight was 68.4 (36.8) and 79.8 (37.7) kg, respectively. The most frequent efmoroctocog alfa regimen was twice weekly (45.7%), followed by every 4 days (20.6%), every 3 days (9.1%), and 3 times per week (7.5%). The most frequent rurioctocog alfa pegol regimen was twice weekly (72.4%), followed by 3 times per week (8.7%), every 4 days (7.6%), and every 3 days (5.5%). The proportion of efmoroctocog alfa twice-weekly dispensing records increased from 31.5% to 50.9%, and every 4 days dispensing records decreased from 31.3% to 14.5% (2015-2018). The proportion of rurioctocog alfa pegol dispensing records remained broadly stable (2016-2018). Overall, mean (SD; median) weekly prophylactic dose was 105.4 (77.9; 92.6) IU/kg with efmoroctocog alfa, and 96.8 (41.9; 90.9) IU/kg with rurioctocog alfa pegol. Conclusion: In this database study, the most frequently observed dosing frequency was twice weekly for patients receiving efmoroctocog alfa or rurioctocog alfa pegol. The observed mean weekly consumption was slightly higher, and variation was greater, in patients receiving efmoroctocog alfa versus rurioctocog alfa pegol.

Artificial intelligence and radiomics: fundamentals, applications, and challenges in immunotherapy.

Immunotherapy offers the potential for durable clinical benefit but calls into question the association between tumor size and outcome that currently forms the basis for imaging-guided treatment. Artificial intelligence (AI) and radiomics allow for discovery of novel patterns in medical images that can increase radiology's role in management of patients with cancer, although methodological issues in the literature limit its clinical application. Using keywords related to immunotherapy and radiomics, we performed a literature review of MEDLINE, CENTRAL, and Embase from database inception through February 2022. We removed all duplicates, non-English language reports, abstracts, reviews, editorials, perspectives, case reports, book chapters, and non-relevant studies. From the remaining articles, the following information was extracted: publication information, sample size, primary tumor site, imaging modality, primary and secondary study objectives, data collection strategy (retrospective vs prospective, single center vs multicenter), radiomic signature validation strategy, signature performance, and metrics for calculation of a Radiomics Quality Score (RQS). We identified 351 studies, of which 87 were unique reports relevant to our research question. The median (IQR) of cohort sizes was 101 (57-180). Primary stated goals for radiomics model development were prognostication (n=29, 33.3%), treatment response prediction (n=24, 27.6%), and characterization of tumor phenotype (n=14, 16.1%) or immune environment (n=13, 14.9%). Most studies were retrospective (n=75, 86.2%) and recruited patients from a single center (n=57, 65.5%). For studies with available information on model testing, most (n=54, 65.9%) used a validation set or better. Performance metrics were generally highest for radiomics signatures predicting treatment response or tumor phenotype, as opposed to immune environment and overall prognosis. Out of a possible maximum of 36 points, the median (IQR) of RQS was 12 (10-16). While a rapidly increasing number of promising results offer proof of concept that AI and radiomics could drive precision medicine approaches for a wide range of indications, standardizing the data collection as well as optimizing the methodological quality and rigor are necessary before these results can be translated into clinical practice.

The Impact of Pharmacokinetic-Guided Prophylaxis on Clinical Outcomes and Healthcare Resource Utilization in Hemophilia A Patients: Real-World Evidence from the CHESS II Study.

Background: Using a pharmacokinetic (PK)-guided approach to personalize the dose and frequency of prophylactic treatment can help achieve and maintain targeted factor VIII (FVIII) trough levels in patients with hemophilia A. Objective: Investigate clinical and healthcare resource use outcomes in patients with hemophilia A treated with or without PK-guided prophylaxis using data from the Cost of Haemophilia in Europe: A Socioeconomic Survey (CHESS) II database. Methods: CHESS II was a cross-sectional, retrospective, burden-of-illness study incorporating data from eight European countries. Patients were eligible for this analysis if they were male, ≥18 years of age, and diagnosed with congenital hemophilia A of any severity. The clinical endpoints included annualized bleeding rate (ABR), presence and number of problem/target joints, and occurrence of joint surgeries. Healthcare resource utilization endpoints included the number of hematologist consultations and bleed-related hospitalizations or emergency department admissions. Data from November 2018 to October 2020 were included and were stratified according to treatment regimen and use of PK-guided dosing. Results: Altogether, 281 patients on prophylaxis had available FVIII trough level data. Mean (SD) age was 35.7 (13.8) years. A specific FVIII trough level was targeted in 120 (42.7%) patients and 47 (39.2%) received PK-guided dosing. Patients receiving PK-guided dosing had a mean (SD) ABR of 2.8 (2.1) and target joint number of 0.5 (0.7), compared with 3.9 (2.7) and 0.9 (1.4), respectively, for patients receiving non-PK-guided treatment. The mean (SD) number of hematologist consultations was 7.1 (5.3) for patients receiving PK-guided dosing versus 10.7 (5.7) for those who were not. A higher proportion of patients in the non-PK-guided group required hospitalization during their lifetime compared with the PK-guided group. Conclusion: This analysis of real-world data suggests that PK-guided dosing for prophylaxis has a beneficial impact on clinical and healthcare resource utilization outcomes in patients with hemophilia A.

Real-world study of rurioctocog alfa pegol and emicizumab in US clinical practice among patients with hemophilia A.

BACKGROUND: FVIII replacement is standard treatment for hemophilia A without inhibitors, but non-factor therapies, such as emicizumab, are changing the treatment landscape. We explore the ramifications of switching treatment. METHODS: Pharmacy database data (July 2017-May 2020) from patients with hemophilia A without inhibitors who switched to rurioctocog alfa pegol or emicizumab prophylaxis after ≥6 months' prophylaxis with another FVIII product were analyzed for total mean weekly consumption, dosing frequency, product wastage, and ABR. RESULTS: Post-switch mean weekly consumption of prophylactic rurioctocog alfa pegol and emicizumab were 6224 IU/kg and 109 mg, respectively. Dosing schedules for emicizumab were primarily weekly (48.2%) and every 2 weeks (40.0%). Most patients in the rurioctocog alfa pegol cohort received treatment twice-weekly (83.3%). Mean product wastage of emicizumab (8.4%) was significantly higher versus rurioctocog alfa pegol (-0.3%; P < 0.001). Mean annualized emicizumab and rurioctocog alfa pegol wastage during prophylaxis was 330.82 mg and -974.80 IU, respectively. ABR change was not significantly different (P = 0.527) for patients switching to emicizumab (-1.05) or rurioctocog alfa pegol (-1.53). CONCLUSIONS: Bleed rates were similar for patients receiving prophylaxis with emicizumab or rurioctocog alfa pegol after switching from prophylaxis with another FVIII. However, wastage associated with dispensing inaccuracies was greater with emicizumab.

Recent advances in therapeutic options for rare hemostatic disorders: selected poster extracts of recent research in hemophilia A, congenital hemophilia with inhibitors, von Willebrand disease, and thrombotic thrombocytopenic purpura presented at the 29th congress of the International Society on Thrombosis and Haemostasis (ISTH 2021, Jul 17-21; virtual congress).

Hemophilia, von Willebrand disease (VWD), and thrombotic thrombocytopenic purpura (TTP) are rare diseases affecting normal hemostasis. Although they differ in their pathogenesis and clinical manifestation, if left undiagnosed and untreated, all these conditions can result in severe long-term consequences and can be potentially life-threatening. This article summarizes a poster series funded by Takeda and presented virtually at the 29th annual congress of the International Society on Thrombosis and Haemostasis (ISTH) in 2021: Data from real-world evidence highlight the importance of joint health and personalized prophylaxis to prevent bleeding for patients with hemophilia, the need to further raise disease awareness in support of timely diagnosis and access to treatment in general practice settings for patients with VWD, and describe the clinical burden for patients with TTP and the importance to advance treatment options for these patients.

Assessing Pathologic Response in Resected Lung Cancers: Current Standards, Proposal for a Novel Pathologic Response Calculator Tool, and Challenges in Practice.

The efficacy of neoadjuvant treatment for NSCLC can be pathologically assessed in resected tissue. Major pathologic response (MPR) and pathologic complete response (pCR), defined as less than or equal to 10% and 0% viable tumor cells, respectively, are increasingly being used in NSCLC clinical trials to establish them as surrogate end points for efficacy to shorten time to outcome. Nevertheless, sampling and MPR calculation methods vary between studies. The International Association for the Study of Lung Cancer recently published detailed recommendations for pathologic assessment of NSCLC after neoadjuvant treatment, with methodology being critical. To increase methodological rigor further, we developed a novel MPR calculator tool (MPRCT) for standardized, comprehensive collection of percentages of viable tumor, necrosis, and stroma in the tumor bed. In addition, tumor width and length in the tumor bed are measured and unweighted and weighted MPR averages are calculated, the latter to account for the varying proportions of tumor beds on slides. We propose sampling the entire visible tumor bed for tumors having pCR regardless of size, 100% of tumors less than or equal to 3 cm in diameter, and at least 50% of tumors more than 3 cm. We describe the uses of this tool, including potential formal analyses of MPRCT data to determine the optimum sampling strategy that balances sensitivity against excessive use of resources. Solutions to challenging scenarios in pathologic assessment are proposed. This MPRCT will facilitate standardized, systematic, comprehensive collection of pathologic response data with a standardized methodology to validate studies designed to establish MPR and pCR as surrogate end points of neoadjuvant treatment efficacy.

Deep learning prediction of axillary lymph node status using ultrasound images.

OBJECTIVE: To investigate the ability of our convolutional neural network (CNN) to predict axillary lymph node metastasis using primary breast cancer ultrasound (US) images. METHODS: In this IRB-approved study, 338 US images (two orthogonal images) from 169 patients from 1/2014-12/2016 were used. Suspicious lymph nodes were seen on US and patients subsequently underwent core-biopsy. 64 patients had metastatic lymph nodes. A custom CNN was utilized on 248 US images from 124 patients in the training dataset and tested on 90 US images from 45 patients. The CNN was implemented entirely of 3 × 3 convolutional kernels and linear layers. The 9 convolutional kernels consisted of 6 residual layers, totaling 12 convolutional layers. Feature maps were down-sampled using strided convolutions. Dropout with a 0.5 keep probability and L2 normalization was utilized. Training was implemented by using the Adam optimizer and a final SoftMax score threshold of 0.5 from the average of raw logits from each pixel was used for two class classification (metastasis or not). RESULTS: Our CNN achieved an AUC of 0.72 (SD ± 0.08) in predicting axillary lymph node metastasis from US images in the testing dataset. The model had an accuracy of 72.6% (SD ± 8.4) with a sensitivity and specificity of 65.5% (SD ± 28.6) and 78.9% (SD ± 15.1) respectively. Our algorithm is available to be shared for research use. (https://github.com/stmutasa/MetUS). CONCLUSION: It's feasible to predict axillary lymph node metastasis from US images using a deep learning technique. This can potentially aid nodal staging in patients with breast cancer.

Predicting death or recurrence of portal hypertension symptoms after TIPS procedures.

BACKGROUND: Accurate prediction of portal hypertension recurrence after transjugular intrahepatic portosystemic shunt (TIPS) placement will improve clinical decision-making. PURPOSE: To evaluate if perioperative variables could predict disease-free survival (DFS) in cirrhotic patients with portal hypertension (PHT) treated with TIPS. MATERIALS AND METHODS: We recruited 206 cirrhotic patients with PHT treated with TIPS, randomly assigned to training (n = 138) and validation (n = 68) sets. We recorded 7 epidemiological, 4 clinical, and 9 radiological variables. TIPS-distal end positioning (TIPS-DEP) measured the distance between the distal end of the stent and the hepatocaval junction on contrast-enhanced CT scans. In the training set, the signature was defined as the random forest for survival algorithm achieving the lowest error rate for the prediction of DFS which was landmarked 4 weeks after the TIPS procedure. In the training set, a simple to use scoring system was derived from variables selected by the signature. The primary endpoint was to assess if TIPS-DEP was associated with DFS. The secondary endpoint was to validate the scoring system in the validation set. RESULTS: Overall, patients with TIPS-DEP ≥ 6 mm (n = 49) had a median DFS of 24.5 months vs. 72.8 months otherwise (n = 157, p = 0.004). In the training set, the scoring system was calculated by adding age ≥ 60 years old, Child-Pugh B or C, and TIPS-DEP ≥ 6 mm (1 point each) since the signature showed high DFS probability at 6.5 months post-landmark in patients that did not meet these criteria: 86%, 80%, and 78%, respectively. The hazard ratio [95 CI] between patients determined to be low-risk (< 2 points) and high-risk (≥ 2 points) was 2.30 [1.35-3.93] (p = 0.002) in the training set and 2.01 [0.94-4.32] (p = 0.072) in the validation set. CONCLUSION: TIPS-DEP is an actionable radiological biomarker which can be combined with age and Child-Pugh score to predict death or PHT symptom recurrence after TIPS procedure. KEY POINTS: • TIPS-DEP measurement was the third most important but only actionable variable for predicting DFS. • TIPS-DEP < 6 mm was associated with a DFS probability of 78% at 6.5 months post-landmark. • A simple scoring system calculated using age, Child-Pugh score, and TIPS-DEP predicted DFS after TIPS.

Effect of CT image acquisition parameters on diagnostic performance of radiomics in predicting malignancy of pulmonary nodules of different sizes.

OBJECTIVES: To investigate the effect of CT image acquisition parameters on the performance of radiomics in classifying benign and malignant pulmonary nodules (PNs) with respect to nodule size. METHODS: We retrospectively collected CT images of 696 patients with PNs from March 2015 to March 2018. PNs were grouped by nodule diameter: T1a (diameter ≤ 1.0 cm), T1b (1.0 cm < diameter ≤ 2.0 cm), and T1c (2.0 cm < diameter ≤ 3.0 cm). CT images were divided into four settings according to slice-thickness-convolution-kernels: setting 1 (slice thickness/reconstruction type: 1.25 mm sharp), setting 2 (5 mm sharp), setting 3 (5 mm smooth), and random setting. We created twelve groups from two interacting conditions. Each PN was segmented and had 1160 radiomics features extracted. Non-redundant features with high predictive ability in training were selected to build a distinct model under each of the twelve subsets. RESULTS: The performance (AUCs) on predicting PN malignancy were as follows: T1a group: 0.84, 0.64, 0.68, and 0.68; T1b group: 0.68, 0.74, 0.76, and 0.70; T1c group: 0.66, 0.64, 0.63, and 0.70, for the setting 1, setting 2, setting 3, and random setting, respectively. In the T1a group, the AUC of radiomics model in setting 1 was statistically significantly higher than all others; In the T1b group, AUCs of radiomics models in setting 3 were statistically significantly higher than some; and in the T1c group, there were no statistically significant differences among models. CONCLUSIONS: For PNs less than 1 cm, CT image acquisition parameters have a significant influence on diagnostic performance of radiomics in predicting malignancy, and a model created using images reconstructed with thin section and a sharp kernel algorithm achieved the best performance. For PNs larger than 1 cm, CT reconstruction parameters did not affect diagnostic performance substantially. KEY POINTS: • CT image acquisition parameters have a significant influence on the diagnostic performance of radiomics in pulmonary nodules less than 1 cm. • In pulmonary nodules less than 1 cm, a radiomics model created by using images reconstructed with thin section and a sharp kernel algorithm achieved the best diagnostic performance. • For PNs larger than 1 cm, CT image acquisition parameters do not affect diagnostic performance substantially.

Weakly Supervised Deep Learning Approach to Breast MRI Assessment.

RATIONALE AND OBJECTIVES: To evaluate a weakly supervised deep learning approach to breast Magnetic Resonance Imaging (MRI) assessment without pixel level segmentation in order to improve the specificity of breast MRI lesion classification. MATERIALS AND METHODS: In this IRB approved study, the dataset consisted of 278,685 image slices from 438 patients. The weakly supervised network was based on the Resnet-101 architecture. Training was implemented using the Adam optimizer and a final SoftMax score threshold of 0.5 was used for two class classification (malignant or benign). 278,685 image slices were combined into 92,895 3-channel images. 79,871 (85%) images were used for training and validation while 13,024 (15%) images were separated for testing. Of the testing dataset, 11,498 (88%) were benign and 1531 (12%) were malignant. Model performance was assessed. RESULTS: The weakly supervised network achieved an AUC of 0.92 (SD ± 0.03) in distinguishing malignant from benign images. The model had an accuracy of 94.2% (SD ± 3.4) with a sensitivity and specificity of 74.4% (SD ± 8.5) and 95.3% (SD ± 3.3) respectively. CONCLUSION: It is feasible to use a weakly supervised deep learning approach to assess breast MRI images without the need for pixel-by-pixel segmentation yielding a high degree of specificity in lesion classification.

Convolutional Neural Network Addresses the Confounding Impact of CT Reconstruction Kernels on Radiomics Studies.

Achieving high feature reproducibility while preserving biological information is one of the main challenges for the generalizability of current radiomics studies. Non-clinical imaging variables, such as reconstruction kernels, have shown to significantly impact radiomics features. In this study, we retrain an open-source convolutional neural network (CNN) to harmonize computerized tomography (CT) images with various reconstruction kernels to improve feature reproducibility and radiomic model performance using epidermal growth factor receptor (EGFR) mutation prediction in lung cancer as a paradigm. In the training phase, the CNN was retrained and tested on 32 lung cancer patients' CT images between two different groups of reconstruction kernels (smooth and sharp). In the validation phase, the retrained CNN was validated on an external cohort of 223 lung cancer patients' CT images acquired using different CT scanners and kernels. The results showed that the retrained CNN could be successfully applied to external datasets with different CT scanner parameters, and harmonization of reconstruction kernels from sharp to smooth could significantly improve the performance of radiomics model in predicting EGFR mutation status in lung cancer. In conclusion, the CNN based method showed great potential in improving feature reproducibility and generalizability by harmonizing medical images with heterogeneous reconstruction kernels.

Important Surgical and Clinical End Points in Neoadjuvant Immunotherapy Trials in Resectable NSCLC.

Neoadjuvant immunotherapy may improve outcomes in patients with resectable NSCLC and is being evaluated in phase 2 and 3 studies. Nevertheless, preoperative treatment postpones resection; the potential for increased surgical complexity and greater intra- and postoperative morbidity and mortality is an additional consideration. In studies primarily designed to evaluate efficacy, the impact of neoadjuvant immunotherapy on surgery is based on parameters that are poorly defined and reported differently between studies. Defining and reporting common end points among trials would improve understanding and facilitate cross-comparison of different immunotherapy regimens and may facilitate wider adoption of induction therapies by surgeons and oncologists. We propose several surgical end points and related metrics for neoadjuvant immunotherapy in resectable NSCLC. These include the periods from screening to treatment initiation and from last neoadjuvant dose to surgery; reporting of the allowable window for surgery to preclude masking delays caused by induction treatment-related toxicity; complete resection (R0) rate; preoperative downstaging; a standardized list of immune-related adverse events and associated delay to surgery; preoperative attrition; postoperative attrition before adjuvant therapy; and postoperative 30- and 90-day mortality and morbidity rates. Intraoperative end points (blood loss, duration, and type of surgery) and our proposed system of grading complexity based on lymphadenopathy and fibrosis would allow quantitation of technical difficulty and quality of oncologic resection. In conclusion, the standardization, reporting, and prospective inclusion of these end points in study protocols would provide a comparative overview of the impact of different neoadjuvant immunotherapy regimens on surgery and ultimately clinical oncologic outcomes in resectable NSCLC.