Detection Of Brain Metastases Research Articles

Abstract 7019: Identifying cell free DNA methylation signatures in cerebrospinal fluids for the early detection of brain metastasis in non-small cell lung cancer

Abstract Brain metastasis (BM) is major complication for adult cancer patients. Up to 2% of all cancer patients develop BM. Non-small cell lung cancer (NSCLC) contributes up to 55% of all BM. Approximately 15-20% of NSCLC patients show metastases at diagnosis, and approximately half will develop BM during their clinical course. Even after treatments like surgery, radiation and immunotherapy, the prognosis is dismal with median overall survival less than 13 months. No standard screening or early diagnostic testing exists for BM now. There is an urgent need to identify genomic features that predispose towards brain metastasis - these biomarkers could be used for early detection and disease monitoring. The cfDNA methylation and fragmentation patterns provide potential epigenetic biomarkers for detecting cancer. These patterns are specific to cell/tissue origin and cancer types. DNA methylation refers to a chemical modification of a specific carbon with an added methyl group in the cytosine. Hypomethylation in the gene body and hypermethylation in the promoter region lead to cancer development. We hypothesize that there are specific cfDNA methylation and fragmentation signatures in CSF that reflect the epigenetic states of BMs, and cancer predisposition for brain metastasis. We applied a novel single molecule sequencing approach to study CSF cfDNA methylation of BM from NSCLC. Our cohort consisted of ten patients with NSCLC-BM (samples N=14) and eleven healthy controls (N=11). Briefly, extracted cfDNA was obtained from CSF samples. Multiplexed sequencing libraries were prepared following an optimized protocol, and sequenced on PromethION flow cell for 72hr by Oxford nanopore technologies. Data were processed with the onboard base and methylation calling, followed by our bioinformatic pipelines for cfDNA size and methylation analysis. Distinct methylation patterns were found on significant CpG sites (fdr-corrected p<0.01) between cancer patients and controls, in both gene-body and promoter region. Biological pathway analysis showed enrichment in essential functions involving PI3K, ERBB2, TNF-α signaling (p<0.03), all of which play a role in NSCLC. Enriched mono-nucleosome levels (p = 1.07e-03) and differentiated mono-/tri- nucleosome ratios (p = 1.34e-03) were found in cancer vs. controls. Overall, our results suggest that a combination of cfDNA methylation and fragmentation can detect NSCLC-BM. This study will benefit BM patients with early detection and better prognosis, and extend our understanding in brain metastases in NSCLC. Citation Format: Tianqi (Kiki) Chen, Billy T. Lau, Xiangqi Bai, Melanie Hayden Gephart, Hanlee P. Ji. Identifying cell free DNA methylation signatures in cerebrospinal fluids for the early detection of brain metastasis in non-small cell lung cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 7019.

Comparison between postcontrast thin-slice T1-weighted 2D spin echo and 3D T1-weighted SPACE sequences in the detection of brain metastases at 1.5 and 3 T

ObjectivesAccurate detection of metastatic brain lesions (MBL) is critical due to advances in radiosurgery. We compared the results of three readers in detecting MBL using T1-weighted 2D spin echo (SE) and sampling perfection with application-optimized contrasts using different flip angle evolution (SPACE) sequences with whole-brain coverage at both 1.5 T and 3 T.MethodsFifty-six patients evaluated for MBL were included and underwent a standard protocol (1.5 T, n = 37; 3 T, n = 19), including postcontrast T1-weighted SE and SPACE. The rating was performed by three raters in two sessions > six weeks apart. The true number of MBL was determined using all available imaging including follow-up. Intraclass correlations for intra-rater and inter-rater agreement were calculated. Signal intensity ratios (SIR; enhancing lesion, white matter) were determined on a subset of 46 MBL > 4 mm. A paired t-test was used to evaluate postcontrast sequence order and SIR. Reader accuracy was evaluated by the coefficient of determination.ResultsA total of 135 MBL were identified (mean/subject 2.41, SD 6.4). The intra-rater agreement was excellent for all 3 raters (ICC = 0.97–0.992), as was the inter-rater agreement (ICC = 0.995 SE, 0.99 SPACE). Subjective qualitative ratings were lower for SE images; however, signal intensity ratios were higher in SE sequences. Accuracy was high in all readers for both SE (R2 0.95–0.96) and SPACE (R2 0.91–0.96) sequences.ConclusionsAlthough SE sequences are superior to gradient echo sequences in the detection of small MBL, they have long acquisition times and frequent artifacts. We show that T1-weighted SPACE is not inferior to standard thin-slice SE sequences in the detection of MBL at both imaging fields.Critical relevance statementOur results show the suitability of 3D T1-weighted turbo spin echo (TSE) sequences (SPACE, CUBE, VISTA) in the detection of brain metastases at both 1.5 T and 3 T.Key points• Accurate detection of brain metastases is critical due to advances in radiosurgery.• T1-weighted SE sequences are superior to gradient echo in detecting small metastases.• T1-weighted 3D-TSE sequences may achieve high resolution and relative insensitivity to artifacts.• T1-weighted 3D-TSE sequences have been recommended in imaging brain metastases at 3 T.• We found T1-weighted 3D-TSE equivalent to thin-slice SE at 1.5 T and 3 T.Graphical

Multiomic analysis for minimal residual disease detection: Addressing challenges in stage II-III colon cancer from COSMOS-CRC-01.

180 Background: A considerable number of patients with stage II-III colon cancer experience disease recurrence following definitive treatment while others may receive unnecessary toxicity from adjuvant chemotherapy. Minimal residual disease (MRD) detection through ctDNA analysis can be evaluated without reliance on tissue which comes with logistic advantages and may be the only option for patients with limited tumor specimens. Methods: COSMOS-CRC-01 is an ongoing biomarker study that enrolled patients with resectable stage 0-III colorectal cancer between January 2020 and April 2021. Plasma samples and clinical data were collected 28 days post-operatively and every 3-6 months for up to 5 years alongside radiographic imaging. Samples were tested on the Guardant Reveal Infinity Oncology Platform for MRD detection. This analytically validated multiomic NGS assay interrogates >700 genes and ~15Mb of methylated regions. A bioinformatics algorithm trained for CRC detection classifies each sample as ctDNA detectable or undetectable based on a predefined statistical likelihood threshold and returns quantitative tumor fraction. The primary endpoint was time to recurrence (TTR) based on ctDNA detection status. Results: This subgroup analysis evaluated 801 post-surgical plasma samples (4mL each, median input 19ng, 96% passed QC) from 136 patients with R0 resected stage II-III colon cancer with a median follow-up of 28 months. ctDNA detection at Day 28 post-surgery and following adjuvant chemotherapy were both associated with significantly shorter TTR (p<0.0001; Table). Sensitivity with serial measurement was 80% (95% CI, 56.3-94.3%), with a median lead time of 5.3 months. Sensitivity was highest for detection of locoregional (2/2), brain (1/1), liver (7/7), or multiple (3/3) metastasis compared to isolated lung (2/4) or peritoneal (1/3) metastasis. Tumor fraction increased as expected in samples collected closer to the time of radiographic detection. Three recurred patients had KRAS G12C variants detected in ctDNA prior to recurrence, suggesting the potential of early intervention with targeted agents. Among 553 post-treatment samples from 114 non-recurred patients, sample specificity was 97.3% (95% CI 95.6-98.5%). Conclusions: We show a plasma-only genomic- and methylation-based assay to have sensitive and specific detection of MRD in stage II-III colon cancer. Serial measurement provides superior performance versus one-time measurement at Day 28 post-surgery and detects targetable variants prior to recurrence. [Table: see text]

Development of RLK-Unet: a clinically favorable deep learning algorithm for brain metastasis detection and treatment response assessment.

Previous deep learning (DL) algorithms for brain metastasis (BM) detection and segmentation have not been commonly used in clinics because they produce false-positive findings, require multiple sequences, and do not reflect physiological properties such as necrosis. The aim of this study was to develop a more clinically favorable DL algorithm (RLK-Unet) using a single sequence reflecting necrosis and apply it to automated treatment response assessment. A total of 128 patients with 1339 BMs, who underwent BM magnetic resonance imaging using the contrast-enhanced 3D T1 weighted (T1WI) turbo spin-echo black blood sequence, were included in the development of the DL algorithm. Fifty-eight patients with 629 BMs were assessed for treatment response. The detection sensitivity, precision, Dice similarity coefficient (DSC), and agreement of treatment response assessments between neuroradiologists and RLK-Unet were assessed. RLK-Unet demonstrated a sensitivity of 86.9% and a precision of 79.6% for BMs and had a DSC of 0.663. Segmentation performance was better in the subgroup with larger BMs (DSC, 0.843). The agreement in the response assessment for BMs between the radiologists and RLK-Unet was excellent (intraclass correlation, 0.84). RLK-Unet yielded accurate detection and segmentation of BM and could assist clinicians in treatment response assessment.

A Review of Recent Advances in the Molecular Mechanisms Underlying Brain Metastasis in Lung Cancer.

Brain metastasis from lung cancer is a prevalent mode of treatment failure associated with a poor prognosis. The incidence of brain metastasis has recently shown a dramatic increase. The early detection and risk stratification of lung cancer-related brain metastasis would be highly advantageous for patients. However, our current knowledge and comprehension of the underlying mechanisms driving brain metastasis in lung cancer pose significant challenges. This review summarizes the mechanisms underlying brain metastasis, focusing on the intricate interplay between lung cancer-derived tumor cells and the unique characteristics of the brain, recent advancements in the identification of driver genes, concomitant genes, epigenetic features, including miRNAs and long noncoding RNAs, as well as the molecular characterization of brain metastasis originating from other organs, which may further enhance risk stratification and facilitate precise treatment strategies.

An evaluation of radiographers’ extended practice in the detection of brain metastases on magnetic resonance images

IntroductionPatients who undergo magnetic resonance (MR) imaging to confirm or rule out metastatic brain disease are required to wait for image review by a radiologist before leaving the department at the institute where this study was carried out. The aim was to evaluate whether radiographers can review images and reduce waiting times in those patients without metastases. MethodsProspective observational study of MR radiographers (n = 11) was undertaken. Radiographers commented on images to confirm whether the images showed evidence of metastatic disease, pathology but no metastases, or no pathology. Responses were compared to the radiological report (reference standard). Online questionnaires determined the views and opinions of radiographers (n = 8) and consultant radiologists (n = 6) towards radiographers expanding their scope of practice to include the confirmation or exclusion of brain metastases. ResultsDespite a lack of formal training for image reviewing, overall level of agreement between the radiographer reviews and reference standard was 77.9 % (κ = 0.45). Pooled sensitivity and specificity were 88.6 % & 71.3 % respectively. Kendall's τ = −0.03 (bootstrap 95 % CI -0.73 to 0.61, p = 0.925). Positive predictive value (PPV) was 65.5 % (CI 59.2%–71.4 %) and negative predictive value (NPV) 91.1 % (CI 84.9%–94.9 %). Radiographers and radiologists surveyed demonstrated a willingness to engage with role expansion. ConclusionBased on our small study and interdisciplinary survey, local radiographers and radiologists agree, following a program of radiographer training, screening for brain metastases by radiographers could be implemented. Implications for practiceWith appropriate governance and training support, the introduction of formal radiographer screening for patients referred to exclude brain metastases could provide more efficient working practice.

Role of Contrast-Enhanced MRI in Diagnosing Brain Metastases in Patients With Lung Cancer

Introduction: Lung cancer is one of the most common cancers in the world. Approximately 40% of all metastases make up brain metastases that occur in 3% of patients at early stages of lung cancer. Early diagnosis of brain metastases directly affects patients’ survival and quality of life.Objective: To study the role of contrast-­enhanced magnetic resonance imaging (MRI) in detecting brain metastases at the outpatient stage in patients with newly diagnosed and histologically confirmed lung cancer and to compare the findings to that of computed tomography (CT).Materials and methods: Our study was conducted at the premises of the Scientific Research Institute – Ochapovsky Regional Clinical Hospital No. 1 (Krasnodar, Russian Federation) between November 2022 and January 2023. Seventy­five patients with histologically confirmed lung cancer underwent brain MRI on a Philips Panorama 1.0T MRI scanner and intravenous administration of Clariscan to detect metastatic lesions. As a diagnostic method, we used a “short” scanning protocol in T1­-WI (TR: 450, TE: 15, slice thickness 5 mm), T2­-WI (TR: 4533, TE: 100, slice thickness 5 mm) and DWI (TR: 3712, TE: 97, slice thickness 5 mm) modes in the axial plane, then the Clariscan contrast agent was injected intravenously at the rate of 0.2 ml per 1 kg of the patient's weight, after which the scan was repeated using 3D T1 (TR: 450, TE: 15, slice thickness 1 mm).Results: The average age of the patients was 60 years. The majority of the participants were male. Of all the patients, the majority had central lung cancer (57.3%). During contrast-­enhanced brain MRI, brain metastases that accumulated a contrast agent in 100% of cases were detected in 30 patients (40%) of the total number of patients. In 7% of the patients, there were non-­metastatic findings undetected by prior contrast­-enhanced CT, such as pontine cavernoma and venous malformations. After MRI for brain metastases, management strategies changed in 30 (40%) patients.Discussion: Contrast-­enhanced MRI is one of the most effective methods for diagnosing metastatic lesions. It allows to localize brain metastases and helps the clinician choose the most effective management strategy. The introduction of contrast­-enhanced brain MRI into routine medical practice could improve the accuracy and quality of brain metastasis detection in patients with lung cancer.

Detection of Brain Metastases by Contrast-Enhanced MRI: Comparison of Gadopiclenol and Gadobenate in a Mouse Model.

The aim of this study was to evaluate the capacity of gadopiclenol, a high-relaxivity gadolinium-based contrast agent to detect brain metastases in mice as a function of dose (0.08 mmol/kg or 0.1 mmol/kg) compared with gadobenate at 0.1 mmol/kg. Brain metastases were induced by ultrasound-guided intracardiac implantation of 1.10 5 MDA-MB-231Br cells in the left ventricle of 18 anesthetized Balb/c Nude nu/nu female mice. At day 28 ± 3 after cell injection, each mouse received 2 crossover intravenous injections at 24-hour intervals, randomly selected from 2 doses of gadopiclenol (0.08 mmol/kg or 0.1 mmol/kg) and gadobenate (0.1 mmol/kg) with n = 6 mice/group (3 groups). Brain magnetic resonance imaging sessions were performed at 4 weeks on a 2.35 T magnet with a 3-dimensional T1-weighted high-resolution gradient echo sequence, before and after each injection. Images were blindly and randomly analyzed to detect enhancing lesions. Contrast-to-noise ratio between the metastases and the surrounding healthy parenchyma was calculated, based on region-of-interest signal measurements. In 2 animals per group, an early time point was added to the protocol (day 22 ± 3) to evaluate the sensitivity of detection as a function of time. After the last imaging session, the presence and location of whole-brain metastases were confirmed by histology in 4 mice. After gadopiclenol, approximately twice as many metastases were detected compared with gadobenate, regardless of the dose. Contrast-to-noise ratios of the detected metastases were 2.3 and 3.3 times higher with gadopiclenol at 0.08 mmol/kg and 0.1 mmol/kg, respectively, compared with gadobenate at 0.1 mmol/kg ( P < 0.0001). Gadopiclenol at the dose of 0.1 mmol/kg resulted in a 1.4-fold higher contrast compared with gadopiclenol at 0.08 mmol/kg ( P < 0.02). In a subset of mice that were imaged 1 week earlier, 2 metastases were detected with gadopiclenol and not with gadobenate. The high-relaxivity macrocyclic gadolinium-based contrast agent gadopiclenol allowed higher diagnostic performance for detecting brain enhancing metastases in terms of contrast-to-noise ratio and number of detected metastases compared with gadobenate, at both equal (0.1 mmol/kg) dose and 20% lower Gd dose (0.08 mmol/kg). Tumor detection was higher after gadopiclenol at the dose of 0.1 mmol/kg compared with 0.08 mmol/kg.

Reducing false positives in deep learning-based brain metastasis detection by using both gradient-echo and spin-echo contrast-enhanced MRI: validation in a multi-center diagnostic cohort.

To develop a deep learning (DL) for detection of brain metastasis (BM) that incorporates both gradient- and turbo spin-echo contrast-enhanced MRI (dual-enhanced DL) and evaluate it in a clinical cohort in comparison with human readers and DL using gradient-echo-based imaging only (GRE DL). DL detection was developed using data from 200 patients with BM (training set) and tested in 62 (internal) and 48 (external) consecutive patients who underwent stereotactic radiosurgery and diagnostic dual-enhanced imaging (dual-enhanced DL) and later guide GRE imaging (GRE DL). The detection sensitivity and positive predictive value (PPV) were compared between two DLs. Two neuroradiologists independently analyzed BM and reference standards for BM were separately drawn by another neuroradiologist. The relative differences (RDs) from the reference standard BM numbers were compared between the DLs and neuroradiologists. Sensitivity was similar between GRE DL (93%, 95% confidence interval [CI]: 90-96%) and dual-enhanced DL (92% [89-94%]). The PPV of the dual-enhanced DL was higher (89% [86-92%], p < .001) than that of GRE DL (76%, [72-80%]). GRE DL significantly overestimated the number of metastases (false positives; RD: 0.05, 95% CI: 0.00-0.58) compared with neuroradiologists (RD: 0.00, 95% CI: - 0.28, 0.15, p < .001), whereas dual-enhanced DL (RD: 0.00, 95% CI: 0.00-0.15) did not show a statistically significant difference from neuroradiologists (RD: 0.00, 95% CI: - 0.20-0.10, p = .913). The dual-enhanced DL showed improved detection of BM and reduced overestimation compared with GRE DL, achieving similar performance to neuroradiologists. The use of deep learning-based brain metastasis detection with turbo spin-echo imaging reduces false positive detections, aiding in the guidance of stereotactic radiosurgery when gradient-echo imaging alone is employed. •Deep learning for brain metastasis detection improved by using both gradient- and turbo spin-echo contrast-enhanced MRI (dual-enhanced deep learning). •Dual-enhanced deep learning increased true positive detections and reduced overestimation. •Dual-enhanced deep learning achieved similar performance to neuroradiologists for brain metastasis counts.

An Automated Brain Metastasis Detection and Segmentation System from MRI with a Large Multi-Institutional Dataset

Developments of automated systems for brain metastasis (BM) detection and segmentation from MRI for assisting early detection and stereotactic radiosurgery (SRS) have been reported but most based upon relatively small datasets from single institutes. This work aims to develop and evaluate a system using a large multi-institutional dataset, and to improve both identification of small/subtle BMs and segmentation accuracy of large BMs. A 3D U-Net system was trained and evaluated to detect and segment intraparenchymal BMs with a size > 2mm using 1856 MRI volumes from 1791 patients treated with SRS from seven institutions (1539 volumes for training, 183 for validation, and 134 for testing). All patients had 3D post-Gd T1w MRI scans pre-SRS. Gross tumor volumes (GTVs) of BMs for SRS were curated by each institute first. Then, additional efforts were spent to create GTVs for the untreated and/or uncontoured BMs, including central reviews by two radiologists, to improve accuracy of ground truth. The training dataset was augmented with synthetic BMs of 3773 MRIs using a 3D generative pipeline. Our system consists of two U-Nets with one using small 3D patches dedicated for detecting small BMs and another using large 3D patches for segmenting large BMs, and a random-forest based fusion module for combining the two network outputs. The first U-Net was trained with 3D patches containing at least one BM < 0.1 cm3. For detection performance, we measured BM-level sensitivity and case-level false-positive (FP) rate. For segmentation performance, we measured BM-level Dice similarity coefficient (DSC) and 95-percentile Hausdorff distance (HD95). We also stratified performances based upon BM sizes. For 739 BMs in the 134 testing cases, the overall lesion-level sensitivity was 0.870 with an average case-level FP of 1.34±1.92 (95% CI: 1.02-1.67). The sensitivity was >0.969 for the BMs >0.1 cm3, but dropped to 0.755 for the BMs < 0.1 cm3 (Table 1). The average DSC and HD95 for all detected BMs were 0.786 and 1.35mm. The worse performance for BMs > 20 cm3 was caused by a case with 83 cm3 GTV and artifacts in the MRI volume. We achieved excellent detection sensitivity and segmentation accuracy for BMs > 0.1 cm3, and promising performance for small BMs (<0.1cm3) with a controlled FP rate using a large multi-institutional dataset. Clinical utility for assisting early detection and SRS planning will be investigated. Table 1: Per-lesion detection and segmentation performance stratified by individual BM size. N is the number of BMs in each category.

Deep learning-based detection and quantification of brain metastases on black-blood imaging can provide treatment suggestions: a clinical cohort study

ObjectivesWe aimed to evaluate whether deep learning–based detection and quantification of brain metastasis (BM) may suggest treatment options for patients with BMs.MethodsThe deep learning system (DLS) for detection and quantification of BM was developed in 193 patients and applied to 112 patients that were newly detected on black-blood contrast-enhanced T1-weighted imaging. Patients were assigned to one of 3 treatment suggestion groups according to the European Association of Neuro-Oncology (EANO)-European Society for Medical Oncology (ESMO) recommendations using number and volume of the BMs detected by the DLS: short-term imaging follow-up without treatment (group A), surgery or stereotactic radiosurgery (limited BM, group B), or whole-brain radiotherapy or systemic chemotherapy (extensive BM, group C). The concordance between the DLS-based groups and clinical decisions was analyzed with or without consideration of targeted agents. The performance of distinguishing high-risk (B + C) was calculated.ResultsAmong 112 patients (mean age 64.3 years, 63 men), group C had the largest number and volume of BM, followed by group B (4.4 and 851.6 mm3) and A (1.5 and 15.5 mm3). The DLS-based groups were concordant with the actual clinical decisions, with an accuracy of 76.8% (86 of 112). Modified accuracy considering targeted agents was 81.3% (91 of 112). The DLS showed 95% (82/86) sensitivity and 81% (21/26) specificity for distinguishing the high risk.ConclusionDLS-based detection and quantification of BM have the potential to be helpful in the determination of treatment options for both low- and high-risk groups of limited and extensive BMs.Clinical relevance statementFor patients with newly diagnosed brain metastasis, deep learning–based detection and quantification may be used in clinical settings where prompt and accurate treatment decisions are required, which can lead to better patient outcomes.Key Points• Deep learning–based brain metastasis detection and quantification showed excellent agreement with ground-truth classifications.• By setting an algorithm to suggest treatment based on the number and volume of brain metastases detected by the deep learning system, the concordance was 81.3%.• When dividing patients into low- and high-risk groups, the sensitivity for detecting the latter was 95%.

Diagnostic value of [ 68 Ga]Ga-Pentixafor versus [ 18 F]FDG PET/CTs in non-small cell lung cancer: a head-to-head comparative study.

In this study, we aimed to compare the diagnostic value of [ 68 Ga]Ga-Pentixafor and [ 18 F]FDG PET/CT in the evaluation of non-small cell lung cancer (NSCLC) patients. Patients with pathology-proven NSCLC were prospectively included. Patients underwent [ 18 F]FDG and [ 68 Ga]Ga-Pentixafor PET/CT within 1 week. All suspicious lesions were interpreted as benign or malignant, and the corresponding PET/CT semi-quantitative parameters were recorded. A two-sided P -value <0.05 was considered significant. Twelve consecutive NSCLC patients (mean age: 60 ± 7) were included. All patients underwent both [ 18 F]FDG and [ 68 Ga]Ga-Pentixafor PET/CT scans with a median interval of 2 days. Overall, 73 abnormal lesions were detected, from which 58 (79%) were concordant between [ 18 F]FDG and [ 68 Ga]Ga-Pentixafor PET/CT. All primary tumors were clearly detectable in both scans visually. Also, [ 68 Ga]Ga-Pentixafor PET/CT demonstrated rather comparable results with [ 18 F]FDG PET/CT scan in detecting metastatic lesions. However, malignant lesions demonstrated significantly higher SUVmax and SUVmean in [ 18 F]FDG PET/CT ( P -values <0.05). Regarding the advantages, [ 68 Ga]Ga-Pentixafor depicted two brain metastases that were missed by [ 18 F]FDG PET/CT. Also, a highly suspicious lesion for recurrence on [ 18 F]FDG PET/CT scan was correctly classified as benign by subsequent [ 68 Ga]Ga-Pentixafor PET/CT. [ 68 Ga]Ga-Pentixafor PET/CT was concordant with [ 18 F]FDG PET/CT in detecting primary NSCLC tumors and could visualize the majority of metastatic lesions. Moreover, this modality was found to be potentially helpful in excluding tumoural lesions when the [ 18 F]FDG PET/CT was equivocal, as well as in detecting brain metastasis where [ 18 F]FDG PET/CT suffers from poor sensitivity. However, the count statistics were significantly lower.

The diagnostic performance of 18 fluorine-fluorodeoxyglucose PET for brain metastases: a systematic review and meta-analysis.

Brain metastases may manifest as hypermetabolism or hypometabolism compared with normal brain activity on 18 fluorine-fluorodeoxyglucose PET ( 18 F-FDG PET). We aim to undertake a systematic review and meta-analysis to determine the diagnostic accuracy of FDG PET for detecting brain metastases from different extracranial primary cancers. PubMed and EMBASE were searched systematically. Study selection and quality assessment were performed independently by two authors. Meta-analysis was performed using a bivariate random-effects model. Subgroup analysis and meta-regression would be performed if heterogeneity was found. A total of 2227 patients from 11 studies were included in the review and analysis. Using the bivariate random-effects model, summary patient-based sensitivity and specificity for all 11 studies were estimated to be 0.440 [95% confidence interval (CI)] 0.295-0.597) and 0.997 (95% CI, 0.977-1.000). In view of significant between-study heterogeneity ( I2  = 74.0% for sensitivity and I2  = 67.3% for specificity), subgroup analyses were performed. Meta-regression showed significantly higher patient-based summary sensitivity for the three better-quality studies (a total of 1037 patients) with satisfactory index test (counting both hypermetabolism and hypometabolism as positive index test) and satisfactory reference standards (other imaging and clinical follow-up) compared with other included studies [0.735 (95% CI, 0.601-0.836) vs 0.304 (95% CI, 0.223-0.400), P value = 0.000]. Our systematic review and meta-analysis showed that FDG PET has overall limited sensitivity and excellent specificity in the detection of brain metastases from extracranial primary cancers. Importantly, subgroup analyses showed that the sensitivity can be significantly improved by raising awareness of asymmetrical hypometabolism. Further studies are warranted to assess the benefits of including the brain in FDG PET studies for all or certain groups of oncological patients.

Deep Learning Accelerated Image Reconstruction of Fluid-Attenuated Inversion Recovery Sequence in Brain Imaging: Reduction of Acquisition Time and Improvement of Image Quality

Fluid-attenuated inversion recovery (FLAIR) imaging is playing an increasingly significant role in the detection of brain metastases with a concomitant increase in the number of magnetic resonance imaging (MRI) examinations. Therefore, the purpose of this study was to investigate the impact on image quality and diagnostic confidence of an innovative deep learning-based accelerated FLAIR (FLAIRDLR) sequence of the brain compared to conventional (standard) FLAIR (FLAIRS) imaging. Seventy consecutive patients with staging cerebral MRIs were retrospectively enrolled in this single-center study. The FLAIRDLR was conducted using the same MRI acquisition parameters as the FLAIRS sequence, except for a higher acceleration factor for parallel imaging (from 2 to 4), which resulted in a shorter acquisition time of 1:39minute instead of 2:40minutes (-38%). Two specialized neuroradiologists evaluated the imaging datasets using a Likert scale that ranged from 1 to 4, with 4 indicating the best score for the following parameters: sharpness, lesion demarcation, artifacts, overall image quality, and diagnostic confidence. Additionally, the image preference of the readers and the interreader agreement were assessed. The average age of the patients was 63±11years. FLAIRDLR exhibited significantly less image noise than FLAIRS, with P-values of<.001 and<.05, respectively. The sharpness of the images and the ability to detect lesions were rated higher in FLAIRDLR, with a median score of 4 compared to a median score of 3 in FLAIRS (P-values of<.001 for both readers). In terms of overall image quality, FLAIRDLR was rated superior to FLAIRS, with a median score of 4 vs 3 (P-values of<.001 for both readers). Both readers preferred FLAIRDLR in 68/70 cases. The feasibility of deep learning FLAIR brain imaging was shown with additional 38% reduction in examination time compared to standard FLAIR imaging. Furthermore, this technique has shown improvement in image quality, noise reduction, and lesion demarcation.

Identification of incidental brain tumors in prostate cancer patients via PSMA PET/CT.

Brain metastases are rare in patients with prostate cancer and portend poor outcome. Prostate-specific membrane antigen positron emission tomography (PSMA PET)/CT scans including the brain have identified incidental tumors. We sought to identify the incidental brain tumor detection rate of PSMA PET/CT performed at initial diagnosis or in the setting of biochemical recurrence. An institutional database was queried for patients who underwent 68Ga-PSMA-11 or 18F-DCFPyL (18F-piflufolastat) PET/CT imaging at an NCI-designated Comprehensive Cancer Center from 1/2018 to 12/2022. Imaging reports and clinical courses were reviewed to identify brain lesions and describe clinical and pathologic features. Two-thousand seven hundred and sixty-three patients underwent 3363 PSMA PET/CT scans in the absence of neurologic symptoms. Forty-four brain lesions were identified, including 33 PSMA-avid lesions: 10 intraparenchymal metastases (30%), 4 dural-based metastases (12%), 16 meningiomas (48%), 2 pituitary macroadenomas (6%), and 1 epidermal inclusion cyst (3%) (incidences of 0.36, 0.14, 0.58, 0.07, and 0.04%). The mean parenchymal metastasis diameter and mean SUVmax were 1.99cm (95%CI:1.25-2.73) and 4.49 (95%CI:2.41-6.57), respectively. At the time of parenchymal brain metastasis detection, 57% of patients had no concurrent extracranial disease, 14% had localized prostate disease only, and 29% had extracranial metastases. Seven of 8 patients with parenchymal brain metastases remain alive at a median 8.8 months follow-up. Prostate cancer brain metastases are rare, especially in the absence of widespread metastatic disease. Nevertheless, incidentally detected brain foci of PSMA uptake may represent previously unknown prostate cancer metastases, even in small lesions and in the absence of systemic disease.

Brain metastasis of hepatocellular carcinoma- single center experience

Background: Although the most common intracranial neoplasm in the adult population is metastatic tumors, brain metastasis from hepatocellular carcnoma (HCC) are very rare. The aim of this study is to analyze patients with advanced HCC, in order to determine the incidence of brain metastasis and evaluate the clinicopathologic properties. Methods: The records of HCC patients treated in our university between 2011 and 2019 were reviewed retrospectively. Patient characteristics, symptoms, laboratory data, treatment modalities, and survival after both the diagnosis of HCC and detection of brain metastasis were recorded. Results: Of the 119 hepatocellular carcinoma patients, 34 had metastasis, 8 of which were to the brain. The median time elapsed between the diagnosis of HCC and brain metastasis was 14.6 months and the median overall survival after the detection of brain metastasis was 1.6 months. In 34 patients with metastasis, median survival was 26.2 months for those without brain metastasis, whereas it was 15.8 months for those with brain metastasis (P = 0.460). The survival times after brain metastasis were 11.6 and 3.9 months for the two patients treated with regorafenib and sorafenib after the detection of brain metastasis, respectively. Conclusion: In this study, it was found that patients who were clinically eligible to receive tyrosine kinase inhibitors survived longer after the detection of brain metastasis. Our study shows that multidisciplinary evaluation of these patients is vital for treatment guidance, and survival outcomes can be improved with the advancements in surgical and radiotherapy techniques even in patients with poor prognosis.

Construction and evaluation of a gated high-resolution neural network for automatic brain metastasis detection and segmentation.

To construct and evaluate a gated high-resolution convolutional neural network for detecting and segmenting brain metastasis (BM). This retrospective study included craniocerebral MRI scans of 1392 patients with 14,542 BMs and 200 patients with no BM between January 2012 and April 2022. A primary dataset including 1000 cases with 11,686 BMs was employed to construct the model, while an independent dataset including 100 cases with 1069 BMs from other hospitals was used to examine the generalizability. The potential of the model for clinical use was also evaluated by comparing its performance in BM detection and segmentation to that of radiologists, and comparing radiologists' lesion detecting performances with and without model assistance. Our model yielded a recall of 0.88, a dice similarity coefficient (DSC) of 0.90, a positive predictive value (PPV) of 0.93 and a false positives per patient (FP) of 1.01 in the test set, and a recall of 0.85, a DSC of 0.89, a PPV of 0.93, and a FP of 1.07 in dataset from other hospitals. With the model's assistance, the BM detection rates of 4 radiologists improved significantly, ranging from 5.2 to 15.1% (all p < 0.001), and also for detecting small BMs with diameter ≤ 5 mm (ranging from 7.2 to 27.0%, all p < 0.001). The proposed model enables accurate BM detection and segmentation with higher sensitivity and less time consumption, showing the potential to augment radiologists' performance in detecting BM. This study offers a promising computer-aided tool to assist the brain metastasis detection and segmentation in routine clinical practice for cancer patients. • The GHR-CNN could accurately detect and segment BM on contrast-enhanced 3D-T1W images. • The GHR-CNN improved the BM detection rate of radiologists, including the detection of small lesions. • The GHR-CNN enabled automated segmentation of BM in a very short time.

Prediction of model generalizability for unseen data: Methodology and case study in brain metastases detection in T1-Weighted contrast-enhanced 3D MRI

Background and purposeA medical AI system's generalizability describes the continuity of its performance acquired from varying geographic, historical, and methodologic settings. Previous literature on this topic has mostly focused on “how” to achieve high generalizability (e.g., via larger datasets, transfer learning, data augmentation, model regularization schemes), with limited success. Instead, we aim to understand “when” the generalizability is achieved: Our study presents a medical AI system that could estimate its generalizability status for unseen data on-the-fly. Materials and methodsWe introduce a latent space mapping (LSM) approach utilizing Fréchet distance loss to force the underlying training data distribution into a multivariate normal distribution. During the deployment, a given test data's LSM distribution is processed to detect its deviation from the forced distribution; hence, the AI system could predict its generalizability status for any previously unseen data set. If low model generalizability is detected, then the user is informed by a warning message integrated into a sample deployment workflow. While the approach is applicable for most classification deep neural networks (DNNs), we demonstrate its application to a brain metastases (BM) detector for T1-weighted contrast-enhanced (T1c) 3D MRI. The BM detection model was trained using 175 T1c studies acquired internally (from the authors' institution) and tested using (1) 42 internally acquired exams and (2) 72 externally acquired exams from the publicly distributed Brain Mets dataset provided by the Stanford University School of Medicine. Generalizability scores, false positive (FP) rates, and sensitivities of the BM detector were computed for the test datasets. Results and conclusionThe model predicted its generalizability to be low for 31% of the testing data (i.e., two of the internally and 33 of the externally acquired exams), where it produced (1) ∼13.5 false positives (FPs) at 76.1% BM detection sensitivity for the low and (2) ∼10.5 FPs at 89.2% BM detection sensitivity for the high generalizability groups respectively. These results suggest that the proposed formulation enables a model to predict its generalizability for unseen data.

Melanoma Brain Metastases: A Systematic Review of Opportunities for Earlier Detection, Diagnosis, and Treatment

Introduction: Melanoma continues to represent the most serious skin cancer worldwide. However, few attempts have been made to connect the body of research on advanced melanoma. In the present review, we report on strides made in the diagnosis and treatment of intracranial metastatic melanoma. Methods: Relevant Cochrane reviews and randomized-controlled trials published by November 2022 were systematically retrieved from the Cochrane Library, EMBASE, and PubMed databases (N = 27). Search and screening methods adhered to the 2020 revision of the Preferred Reporting Items for Systematic reviews and Meta-Analyses guidelines. Results: Although the research surrounding the earlier detection of melanoma brain metastasis is scarce, several studies have highlighted specific markers associated with MBM. Such factors include elevated BRAFV600 mutant ctDNA, high LDH concentration, and high IGF-1R. The approach to treating MBM is moving away from surgery and toward nonsurgical management, namely, a combination of stereotactic radiosurgery (SRS) and immunotherapeutic agents. There is an abundance of emerging research seeking to identify and improve both novel and established treatment options and diagnostic approaches for MBM, however, more research is still needed to maximize the clinical efficacy, especially for new immunotherapeutics. Conclusions: Early detection is optimal for the efficacy of treatment and MBM prognosis. Current treatment utilizes chemotherapies and targeted therapies. Emerging approaches emphasize biomarkers and joint treatments. Further exploration toward preliminary identification, the timing of therapies, and methods to ameliorate adverse treatment effects are needed to advance MBM patient care.

Treatment Outcome of Gamma Knife Radiosurgery for Brain Metastasis from Thyroid Cancer: Favorable Local Control but Poor Survival

Brain metastasis from thyroid cancer (TCBM) is extremely rare; thus, despite a good treatment outcome for thyroid cancer, TCBM has shown poor clinical outcomes. Considering the short survival and poor general condition of patients with TCBM, stereotactic radiosurgery may be preferred to achieve local control. A total of 25 patients with TCBM who underwent Gamma Knife radiosurgery (GKS) were initially included in this study; however, 3 patients were excluded because of a lack of data. There were 7 men (31.8%) and 15 women (68.2%) and the mean age was 63.7 years. The most common type of thyroid cancer histology was papillary carcinoma. Fourteen patients (63.6%) harbored single brain metastatic tumor and 8 (36.3%) had multiple brain metastatic tumors. The mean duration from thyroid cancer diagnosis to detection of brain metastasis was 7.7 years (range, 0-23 years). The median dose of radiation of GKS was 22 Gy (range, 18-25 Gy). There was no radiation-induced complication after GKS. The median overall survival (OS) was 15 months and the 1-year OS of patients with TCBM was 63%, the 2-year OS was 38%, and the 5-year OS was 28%. The 6-month progression-free survival (PFS) for local recurrence of TCBM was 90.4%, the 1-year PFS was 84%, and the 3-year PFS was 84%. GKS showed favorable local control for TCBM. However, the rate of distant brain metastasis was high and median survival of patients with TCBM was only 15 months.