Phase II multicenter trial combining nivolumab and radiosurgery for NSCLC and RCC brain metastases.

Background: Anti-PD-1 has activity in brain metastases (BM). This phase II open labeled non-randomized single arm trial examined the safety and efficacy of combining nivolumab with radiosurgery (SRS) in the treatment of patients with BM from non-small cell lung cancer (NSCLC) and renal cell carcinoma (RCC). Methods: This was a multicenter trial (NCT02978404) in which patients diagnosed with NSCLC or RCC, having ≤ 10 cc of un-irradiated BM and no prior immunotherapy were eligible. Nivolumab (240 mg or 480 mg IV) was administered for up to 2 years until progression. SRS (15-21 Gy) to all un-irradiated BM was delivered within 14 days after the first dose of nivolumab. The primary endpoint was intracranial progression free survival (iPFS). Results: Twenty-six patients (22 NSCLC and 4 RCC) were enrolled between August 2017 and January 2020. A median of 3 (1-9) BM were treated with SRS. Median follow-up was 16.0 months (0.43-25.9 months). Two patients developed nivolumab and SRS related grade 3 fatigue. One-year iPFS and OS were 45.2% (95% CI 29.3-69.6%) and 61.3% (95% CI 45.1-83.3%), respectively. Overall response (partial or complete) of SRS treated BM was attained in 14 out of the 20 patients with ≥1 evaluable follow-up MRI. Mean FACT-Br total scores were 90.2 at baseline and improved to 146.2 within 2-4 months (P = .0007). Conclusions: The adverse event profile and FACT-Br assessments suggested that SRS during nivolumab was well tolerated. Upfront SRS with the initiation of anti-PD-1 prolonged the 1-year iPFS and achieved high intracranial control. This combined approach merits validation randomized studies.

Assessment of Radiology Artificial Intelligence Software: A Validation and Evaluation Framework.

Artificial intelligence (AI) software in radiology is becoming increasingly prevalent and performance is improving rapidly with new applications for given use cases being developed continuously, oftentimes with development and validation occurring in parallel. Several guidelines have provided reporting standards for publications of AI-based research in medicine and radiology. Yet, there is an unmet need for recommendations on the assessment of AI software before adoption and after commercialization. As the radiology AI ecosystem continues to grow and mature, a formalization of system assessment and evaluation is paramount to ensure patient safety, relevance and support to clinical workflows, and optimal allocation of limited AI development and validation resources before broader implementation into clinical practice. To fulfil these needs, we provide a glossary for AI software types, use cases and roles within the clinical workflow; list healthcare needs, key performance indicators and required information about software prior to assessment; and lay out examples of software performance metrics per software category. This conceptual framework is intended to streamline communication with the AI software industry and provide healthcare decision makers and radiologists with tools to assess the potential use of these software. The proposed software evaluation framework lays the foundation for a radiologist-led prospective validation network of radiology AI software. Learning Points: The rapid expansion of AI applications in radiology requires standardization of AI software specification, classification, and evaluation. The Canadian Association of Radiologists' AI Tech & Apps Working Group Proposes an AI Specification document format and supports the implementation of a clinical expert evaluation process for Radiology AI software.

Should Magnetic Resonance Angiography Be Used for Screening of Intracranial Aneurysm in Adults with Sickle Cell Disease?

Magnetic resonance imaging (MRI) is used in patients with sickle cell disease (SCD) to detect silent cerebral infarcts. MR angiography (MRA) can identify arterial stenoses and intracranial aneurysms (ICANs) associated with SCD. In this study, we aimed to estimate the prevalence of ICANs in asymptomatic adult patients with SCD referred from the SCD clinic for routine screening by MRI/MRA using a 3T-MRI scanner. Findings were independently reviewed by two neuroradiologists. Between 2016 and 2020, 245 asymptomatic adults with SCD were stratified according to genotype (SS/S-ß0thalassemia and SC/Sß+). ICANs were found in 27 patients (11%; 0.95 CI: 8-16%). ICANs were more frequent in SS/S-ß0thalassemia patients (20/118 or 17%; 0.95 CI: 11-25%) than in SC/ßb+ patients (7/127 or 6%; 0.95 CI: 2-11%; p = 0.007). Individuals with SCD (particularly SS/S-ß0thalassemia) have a higher prevalence of ICANs than the general population. We believe that MRA should be considered in the current American Society of Hematology guidelines, which already contain a recommendation for MRI at least once in adult SCD patients. However, the clinical significance of preventive treatment of unruptured aneurysms remains controversial.

Non-contrast CT markers of intracerebral hematoma expansion: a reliability study.

OBJECTIVES: We evaluated whether clinicians agree in the detection of non-contrast CT markers of intracerebral hemorrhage (ICH) expansion. METHODS: From our local dataset, we randomly sampled 60 patients diagnosed with spontaneous ICH. Fifteen physicians and trainees (Stroke Neurology, Interventional and Diagnostic Neuroradiology) were trained to identify six density (Barras density, black hole, blend, hypodensity, fluid level, swirl) and three shape (Barras shape, island, satellite) expansion markers, using standardized definitions. Thirteen raters performed a second assessment. Inter- and intra-rater agreement were measured using Gwet's AC1, with a coefficient > 0.60 indicating substantial to almost perfect agreement. RESULTS: Almost perfect inter-rater agreement was observed for the swirl (0.85, 95% CI: 0.78-0.90) and fluid level (0.84, 95% CI: 0.76-0.90) markers, while the hypodensity (0.67, 95% CI: 0.56-0.76) and blend (0.62, 95% CI: 0.51-0.71) markers showed substantial agreement. Inter-rater agreement was otherwise moderate, and comparable between density and shape markers. Inter-rater agreement was lower for the three markers that require the rater to identify one specific axial slice (Barras density, Barras shape, island: 0.46, 95% CI: 0.40-0.52 versus others: 0.60, 95% CI: 0.56-0.63). Inter-observer agreement did not differ when stratified for raters' experience, hematoma location, volume, or anticoagulation status. Intra-rater agreement was substantial to almost perfect for all but the black hole marker. CONCLUSION: In a large sample of raters with different backgrounds and expertise levels, only four of nine non-contrast CT markers of ICH expansion showed substantial to almost perfect inter-rater agreement. KEY POINTS: • In a sample of 15 raters and 60 patients, only four of nine non-contrast CT markers of ICH expansion showed substantial to almost perfect inter-rater agreement (Gwet's AC1> 0.60). • Intra-rater agreement was substantial to almost perfect for eight of nine hematoma expansion markers. • Only the blend, fluid level, and swirl markers achieved substantial to almost perfect agreement across all three measures of reliability (inter-rater agreement, intra-rater agreement, agreement with the results of a reference reading).

Angiographic results of surgical or endovascular treatment of intracranial aneurysms: a systematic review and inter-observer reliability study.

PURPOSE: Results of surgical or endovascular treatment of intracranial aneurysms are often assessed using angiography. A reliable method to report results irrespective of treatment modality is needed to enable comparisons. Our goals were to systematically review existing classification systems, and to propose a 3-point classification applicable to both treatments and assess its reliability. METHODS: We conducted two systematic reviews on classification systems of angiographic results after clipping or coiling to select a simple 3-category scale that could apply to both treatments. We then circulated an electronic portfolio of angiograms of clipped (n=30) or coiled (n=30) aneurysms, and asked raters to evaluate the degree of occlusion using this scale. Raters were also asked to choose an appropriate follow-up management for each patient based on the degree of occlusion. Agreement was assessed using Krippendorff's α statistics (αK), and relationship between occlusion grade and clinical management was analyzed using Fisher's exact and Cramer's V tests. RESULTS: The systematic reviews found 70 different grading scales with heterogeneous reliability (kappa values from 0.12 to 1.00). The 60-patient portfolio was independently evaluated by 19 raters of diverse backgrounds (neurosurgery, radiology, and neurology) and experience. There was substantial agreement (αK=0.76, 95%CI, 0.67-0.83) between raters, regardless of background, experience, or treatment used. Intra-rater agreement ranged from moderate to almost perfect. A strong relationship was found between angiographic grades and management decisions (Cramer's V: 0.80±0.12). CONCLUSION: A simple 3-point scale demonstrated sufficient reliability to be used in reporting aneurysm treatments or in evaluating treatment results in comparative randomized trials.

Surgical or Endovascular Management of Middle Cerebral Artery Aneurysms: A Randomized Comparison.

OBJECTIVE: There are few randomized data comparing clipping and coiling for middle cerebral artery (MCA) aneurysms. We analyzed results from patients with MCA aneurysms enrolled in the CURES (Collaborative UnRuptured Endovascular vs. Surgery) and ISAT-2 (International Subarachnoid Aneurysm Trial II) randomized trials. METHODS: Both trials are investigator-led parallel-group 1:1 randomized studies. CURES includes patients with 3-mm to 25-mm unruptured intracranial aneurysms (UIAs), and ISAT-2 includes patients with ruptured aneurysms (RA) for whom uncertainty remains after ISAT. The primary outcome measure of CURES is treatment failure: 1) failure to treat the aneurysm, 2) intracranial hemorrhage during follow-up, or 3) residual aneurysm at 1 year. The primary outcome of ISAT-2 is death or dependency (modified Rankin Scale score >2) at 1 year. One-year angiographic outcomes are systematically recorded. RESULTS: There were 100 unruptured and 71 ruptured MCA aneurysms. In CURES, 90 patients with UIA have been treated and 10 await treatment. Surgical and endovascular management of unruptured MCA aneurysms led to treatment failure in 3/42 (7%; 95% confidence interval [CI], 0.02-0.19) for clipping and 13/48 (27%; 95% CI, 0.17-0.41) for coiling (P = 0.025). All 71 patients with RA have been treated. In ISAT-2, patients with ruptured MCA aneurysms managed surgically had died or were dependent (modified Rankin Scale score >2) in 7/38 (18%; 95% CI, 0.09-0.33) cases, and 8/33 (24%; 95% CI, 0.13-0.41) for endovascular. One-year imaging results were available in 80 patients with UIA and 62 with RA. Complete aneurysm occlusion was found in 30/40 (75%; 95% CI, 0.60-0.86) patients with UIA allocated clipping, and 14/40 (35%; 95% CI, 0.22-0.50) patients with UIA allocated coiling. Complete aneurysm occlusion was found in 24/34 (71%; 95% CI, 0.54-0.83) patients with RA allocated clipping, and 15/28 (54%; 95% CI, 0.36-0.70) patients with RA allocated coiling. CONCLUSIONS: Randomized data from 2 trials show that better efficacy may be obtained with surgical management of patients with MCA aneurysms.

Artificial Intelligence Applications for Workflow, Process Optimization and Predictive Analytics.

There is great potential for artificial intelligence (AI) applications, especially machine learning and natural language processing, in medical imaging. Much attention has been garnered by the image analysis tasks for diagnostic decision support and precision medicine, but there are many other potential applications of AI in radiology and have potential to enhance all levels of the radiology workflow and practice, including workflow optimization and support for interpretation tasks, quality and safety, and operational efficiency. This article reviews the important potential applications of informatics and AI related to process improvement and operations in the radiology department.

Diagnostic Reliability of Leptomeningeal Disease Using Magnetic Resonance Imaging.

Brain metastases are seen in 20%-50% of patients with metastatic solid tumors. On the other hand, leptomeningeal disease (LMD) occurs more rarely. The gold standard for the diagnosis of LMD is serial cerebrospinal fluid (CSF) analyses, although in daily practice, the diagnosis of LMD is often made by neuroimaging. Leptomeningeal metastases (LM) have been a relative contra-indication to radiosurgery. It can be noted that focal LMD can be difficult to distinguish from a superficially located/cortical-based brain metastasis which is not a contra-indication for radiosurgery. Hence, justifying the need of a reliable diagnosis method. The goal of this study was to determine the inter-observer reliability of contrast-enhanced magnetic resonance imaging (gdMRI) in the differentiation of focal cortical-based metastases from leptomeningeal spread. This is a retrospective review of a prospectively collected database of patients with brain metastases. A total of 42 cases with superficial lesions were selected for review. Additionally, eight control cases demonstrating deep and/or white-matter based lesions were included in the study. Three neuroradiologists and three radiation oncologists were asked to review each study and score the presence of LM. Inter-observer agreement was calculated using group-derived agreement coefficients (Gwet's AC1 and Gwet's AC2). Pair-wise inter-observer agreement coefficients never reached substantial values for trichotomized outcomes (LMD, non-LMD or indeterminate) but did reach a substantial value in a minority of cases for dichotomised outcomes (LMD or non-LMD). The control subgroup analysis revealed substantial agreement between most pairs for both trichotomized and dichotomised outcomes. We observed low inter-observer agreement amongst specialists for the diagnosis of focal LMD by gdMRI. Neuroimaging should not be relied upon to make treatment decisions, notably to deny patients radiosurgery.

Residual Cerebral Aneurysms After Microsurgical Clipping: A New Scale, an Agreement Study, and a Systematic Review of the Literature.

OBJECTIVE: The surgical repair of a cerebral aneurysm does not always lead to complete occlusion. A standardized repeatable method of reporting results of surgical clipping is desirable. Our purpose was to systematically review methods of classifying aneurysm remnants, provide a new scale with precise definitions of categories, and perform an agreement study to assess the variability in adjudicating remnants after aneurysm clipping. METHODS: A systematic review was performed to identify ways to report angiographic results of surgical clipping between 1963 and 2017. Postclipping angiographic results of 43 patients were also independently evaluated by 10 raters of various experience and backgrounds using a new 4-category scale. Agreement between responses were analyzed using κ statistics. RESULTS: The systematic review yielded 63 articles with 37 different nomenclatures using 2-6 categories. The reliability of judging the presence of an aneurysm remnant on catheter angiography was studied only twice, with only 2 raters each time, with contradictory results. Interobserver agreement using the new 4-category scale was moderate (κ = 0.52; 95% confidence interval, 0.43-0.62) for all observers, but improved to substantial (κ = 0.62; 95% confidence interval, 0.47-0.76) when results were dichotomized (grade 0/1 vs. 2/3). CONCLUSIONS: Various classification schemes to evaluate angiographic results after surgical clipping exist in the literature, but they lack standardization. Adjudication using fewer, better defined categories may yield more reliable agreement.

Cervical Internal Carotid Occlusion versus Pseudo-occlusion at CT Angiography in the Context of Acute Stroke: An Accuracy, Interobserver, and Intraobserver Agreement Study.

Purpose To evaluate the diagnostic accuracy and reliability of computed tomographic (CT) angiography to distinguish true cervical internal carotid artery (ICA) occlusion from pseudo-occlusion (defined as an isolated intracranial thrombus that impedes ascending blood flow) in the context of acute stroke. Materials and Methods This was a retrospective study of patients who underwent thrombectomy with preprocedural CT angiography that helps to demonstrate a lack of attenuation in the cervical ICA on the symptomatic side (24 men and 13 women; mean age, 63 years; age range, 30-86 years). Seven readers, including five neuroradiologists and two interventional neuroradiology fellows, independently reviewed the CT angiography images to assess whether there was true cervical ICA occlusion. Their results were compared with digital subtraction angiography (DSA) as the reference standard. Sensitivity and specificity for detecting true occlusion as well as accuracy and diagnostic odds ratio were calculated, with inter- and intraobserver κ statistics. Results Cervical ICA pseudo-occlusion occurred in 12 of 37 patients (32.4%) with nonattenuation of the cervical ICA on the symptomatic side. Interobserver agreement coefficients did not reach the substantial value of 0.61 for either pairs or groups of readers. The cohort's average sensitivity and specificity was 68% (95% confidence interval [CI]: 59%, 76%) and 75% (95% CI: 71%, 80%), respectively, with a diagnostic odds ratio of 8 (95% CI: 3, 18) and only fair interobserver agreement (κ = 0.32; 95% CI: 0.16, 0.47). Conclusion In the context of acute ischemic stroke with ipsilateral ICA nonattenuation at single-phase CT angiography, even specialized radiologists may not reliably distinguish true cervical occlusion from pseudo-occlusion. © RSNA, 2017 Online supplemental material is available for this article.

Surgical clipping or endovascular coiling for unruptured intracranial aneurysms: a pragmatic randomised trial.

BACKGROUND: Unruptured intracranial aneurysms (UIAs) are increasingly diagnosed and are commonly treated using endovascular treatment or microsurgical clipping. The safety and efficacy of treatments have not been compared in a randomised trial. How to treat patients with UIAs suitable for both options remains unknown. METHODS: We randomly allocated clipping or coiling to patients with one or more 3-25 mm UIAs judged treatable both ways. The primary outcome was treatment failure, defined as: initial failure of aneurysm treatment, intracranial haemorrhage or residual aneurysm on 1-year imaging. Secondary outcomes included neurological deficits following treatment, hospitalisation >5 days, overall morbidity and mortality and angiographic results at 1 year. RESULTS: The trial was designed to include 260 patients. An analysis was performed for slow accrual: 136 patients were enrolled from 2010 through 2016 and 134 patients were treated. The 1-year primary outcome, available for 104 patients, was reached in 5/48 (10.4% (4.5%-22.2%)) patients allocated surgical clipping, and 10/56 (17.9% (10.0%-29.8%)) patients allocated endovascular coiling (OR: 0.54 (0.13-1.90), p=0.40). Morbidity and mortality (modified Rankin Scale>2) at 1 year occurred in 2/48 (4.2% (1.2%-14.0%)) and 2/56 (3.6% (1.0%-12.1%)) patients allocated clipping and coiling, respectively. New neurological deficits (15/65 vs 6/69; OR: 3.12 (1.05-10.57), p=0.031), and hospitalisations beyond 5 days (30/65 vs 6/69; OR: 8.85 (3.22-28.59), p=0.0001) were more frequent after clipping. CONCLUSION: Surgical clipping or endovascular coiling of UIAs did not show differences in morbidity at 1 year. Trial continuation and additional randomised evidence will be necessary to establish the supposed superior efficacy of clipping.

Endovascular thrombectomy and medical therapy versus medical therapy alone in acute stroke: A randomized care trial.

BACKGROUND: Until recently, the benefits of endovascular treatment in stroke were not proven. Care trials have been designed to simultaneously offer yet-to-be validated interventions and verify treatment outcomes. Our aim was to implement a care trial for patients with acute ischemic stroke. METHODS: The study was offered to all patients considered for endovascular management of acute ischemic stroke in one Canadian hospital. Inclusion criteria were broad: onset of symptoms≤5h or at any time in the presence of clinical-imaging mismatch and suspected or demonstrated proximal large vessel occlusion. Exclusion criteria were few: established infarction or hemorrhagic transformation of the target symptomatic territory and poor 3-month prognosis. The primary outcome was mRS≤2 at 3 months. Patients were randomly allocated to standard care or standard care plus endovascular treatment. ClinicalTrials.gov: Identifier NCT02157532. RESULTS: Seventy-seven patients were recruited in 19 months (March 2013-October 2014) at a single center. Randomized allocation was interrupted when other trials showed the benefits of endovascular therapy. At 3 months, 20 of 40 patients (50.0%; 95% CI: 35%-65%) in the intervention group had reached the primary outcome, compared to 14 of 37 patients (37.8%; 95% CI: 24%-54%) in the control group (P=0.36). Eleven patients in the intervention group died within 3 months compared to 9 patients in the standard care group. CONCLUSION: A care trial was implemented to offer verifiable care to acute stroke patients. This approach offers a promising means to manage clinical dilemmas and guide uncertain practices.

Early CT changes in patients admitted for thrombectomy: Intrarater and interrater agreement.

OBJECTIVE: To systematically review the literature and assess agreement on the Alberta Stroke Program Early CT Score (ASPECTS) among clinicians involved in the management of thrombectomy candidates. METHODS: Studies assessing agreement using ASPECTS published from 2000 to 2015 were reviewed. Fifteen raters reviewed and scored the anonymized CT scans of 30 patients recruited in a local thrombectomy trial during 2 independent sessions, in order to study intrarater and interrater agreement. Agreement was measured using intraclass correlation coefficients (ICCs) and Fleiss kappa statistics for ASPECTS and dichotomized ASPECTS at various cutoff values. RESULTS: The review yielded 30 articles reporting 40 measures of agreement. Populations, methods, analyses, and results were heterogeneous (slight to excellent agreement), precluding a meta-analysis. When analyzed as a categorical variable, intrarater agreement was slight to moderate (κ = 0.042-0.469); it reached a substantial level (κ > 0.6) in 11/15 raters when the score was dichotomized (0-5 vs 6-10). The interrater ICCs varied between 0.672 and 0.811, but agreement was slight to moderate (κ = 0.129-0.315). Even in the best of cases, when ASPECTS was dichotomized as 0-5 vs 6-10, interrater agreement did not reach a substantial level (κ = 0.561), which translates into at least 5 of 15 raters not giving the same dichotomized verdict in 15% of patients. CONCLUSIONS: In patients considered for thrombectomy, there may be insufficient agreement between clinicians for ASPECTS to be reliably used as a criterion for treatment decisions.

Subdural Metastasis of Prostate Cancer.

Dural metastasis from prostate cancer is rare and may mimic a subdural hematoma (SDH). Preoperatively diagnosis may be difficult and only reveal its presence during surgery. We present such a case and review the literature to identify common characteristics. A 65-year-old man presented with headache, confusion, and progressive right upper limb weakness. Past history included a prostate adenocarcinoma with bone metastasis 3 years earlier. Head computed tomography (CT) scan without contrast revealed a multinodular bilateral hyperdense extra-axial lesion interpreted as acute SDH. At surgery planned for SDH drainage no blood was found; instead there was an en plaque subdural yellowish tumor. Histopathologic examination was consistent with metastatic adenocarcinoma of the prostate. We found 11 cases reported as dural metastasis of prostate cancer mimicking SDH. Surgery was performed on nine cases with no suspicion of dural metastasis. On preoperative nonenhanced CT scan images, three types of image patterns can be described: a nodule in SDH, multinodular metastasis surrounded by SDH, and large en plaque subdural tumor. The latter group consists of those cases where no blood but rather an en plaque subdural tumor was found at surgery. Even though rare, dural metastasis should be considered among the differential diagnoses in a patient known for prostate cancer.

Evolving role of MRI in optimizing the treatment of multiple sclerosis: Canadian Consensus recommendations.

BACKGROUND: Magnetic resonance imaging (MRI) is increasingly important for the early detection of suboptimal responders to disease-modifying therapy for relapsing-remitting multiple sclerosis. Treatment response criteria are becoming more stringent with the use of composite measures, such as no evidence of disease activity (NEDA), which combines clinical and radiological measures, and NEDA-4, which includes the evaluation of brain atrophy. METHODS: The Canadian MRI Working Group of neurologists and radiologists convened to discuss the use of brain and spinal cord imaging in the assessment of relapsing-remitting multiple sclerosis patients during the treatment course. RESULTS: Nine key recommendations were developed based on published sources and expert opinion. Recommendations addressed image acquisition, use of gadolinium, MRI requisitioning by clinicians, and reporting of lesions and brain atrophy by radiologists. Routine MRI follow-ups are recommended beginning at three to six months after treatment initiation, at six to 12 months after the reference scan, and annually thereafter. The interval between scans may be altered according to clinical circumstances. CONCLUSIONS: The Canadian recommendations update the 2006 Consortium of MS Centers Consensus revised guidelines to assist physicians in their management of MS patients and to aid in treatment decision making.

Value of 3.0 T MR imaging in refractory partial epilepsy and negative 1.5 T MRI.

BACKGROUND: High-field 3.0 T MR scanners provide an improved signal-to-noise ratio which can be translated in higher image resolution, possibly allowing critical detection of subtle epileptogenic lesions missed on standard-field 1.0-1.5 T MRIs. In this study, the authors explore the potential value of re-imaging at 3.0 T patients with refractory partial epilepsy and negative 1.5 T MRI. METHODS: We retrospectively identified all patients with refractory partial epilepsy candidate for surgery who had undergone a 3.0 T MR study after a negative 1.5 T MR study. High-field 3.0 T MRIs were reviewed qualitatively by neuroradiologists experienced in interpreting epilepsy studies with access to clinical information. Relevance and impact on clinical management were assessed by an epileptologist. RESULTS: Between November 2006 and August 2009, 36 patients with refractory partial epilepsy candidate for surgery underwent 3.0 T MR study after a 1.5 T MR study failed to disclose a relevant epileptogenic lesion. A potential lesion was found only in two patients (5.6%, 95% CI: 1.5-18.1%). Both were found to have hippocampal atrophy congruent with other presurgical localization techniques which resulted in omission of an invasive EEG study and direct passage to surgery. CONCLUSIONS: The frequency of detection of a new lesion by re-imaging at 3.0 T patients with refractory partial epilepsy candidate for surgery was found to be low, but seems to offer the potential of a significant clinical impact for selected patients. This finding needs to be validated in a prospective controlled study.