Evaluation of GPT-4 ability to identify and generate patient instructions for actionable incidental radiology findings

OBJECTIVES

To evaluate the proficiency of a HIPAA-compliant version of GPT-4 in identifying actionable, incidental findings from unstructured radiology reports of Emergency Department patients. To assess appropriateness of artificial intelligence (AI)-generated, patient-facing summaries of these findings.

MATERIALS AND METHODS

Radiology reports extracted from the electronic health record of a large academic medical center were manually reviewed to identify non-emergent, incidental findings with high likelihood of requiring follow-up, further sub-stratified as "definitely actionable" (DA) or "possibly actionable-clinical correlation" (PA-CC). Instruction prompts to GPT-4 were developed and iteratively optimized using a validation set of 50 reports. The optimized prompt was then applied to a test set of 430 unseen reports. GPT-4 performance was primarily graded on accuracy identifying either DA or PA-CC findings, then secondarily for DA findings alone. Outputs were reviewed for hallucinations. AI-generated patient-facing summaries were assessed for appropriateness via Likert scale.

RESULTS

For the primary outcome (DA or PA-CC), GPT-4 achieved 99.3% recall, 73.6% precision, and 84.5% F-1. For the secondary outcome (DA only), GPT-4 demonstrated 95.2% recall, 77.3% precision, and 85.3% F-1. No findings were "hallucinated" outright. However, 2.8% of cases included generated text about recommendations that were inferred without specific reference. The majority of True Positive AI-generated summaries required no or minor revision.

CONCLUSION

GPT-4 demonstrates proficiency in detecting actionable, incidental findings after refined instruction prompting. AI-generated patient instructions were most often appropriate, but rarely included inferred recommendations. While this technology shows promise to augment diagnostics, active clinician oversight via "human-in-the-loop" workflows remains critical for clinical implementation.

Functional Connectivity Changes on Resting-State fMRI after Mild Traumatic Brain Injury: A Systematic Review

BACKGROUND

Mild traumatic brain injury is theorized to cause widespread functional changes to the brain. Resting-state fMRI may be able to measure functional connectivity changes after traumatic brain injury, but resting-state fMRI studies are heterogeneous, using numerous techniques to study ROIs across various resting-state networks.

PURPOSE

We systematically reviewed the literature to ascertain whether adult patients who have experienced mild traumatic brain injury show consistent functional connectivity changes on resting-state -fMRI, compared with healthy patients.

DATA SOURCES

We used 5 databases (PubMed, EMBASE, Cochrane Central, Scopus, Web of Science).

STUDY SELECTION

Five databases (PubMed, EMBASE, Cochrane Central, Scopus, and Web of Science) were searched for research published since 2010. Search strategies used keywords of "functional MR imaging" and "mild traumatic brain injury" as well as related terms. All results were screened at the abstract and title levels by 4 reviewers according to predefined inclusion and exclusion criteria. For full-text inclusion, each study was evaluated independently by 2 reviewers, with discordant screening settled by consensus.

DATA ANALYSIS

Data regarding article characteristics, cohort demographics, fMRI scan parameters, data analysis processing software, atlas used, data characteristics, and statistical analysis information were extracted.

DATA SYNTHESIS

Across 66 studies, 80 areas were analyzed 239 times for at least 1 time point, most commonly using independent component analysis. The most analyzed areas and networks were the whole brain, the default mode network, and the salience network. Reported functional connectivity changes varied, though there may be a slight trend toward decreased whole-brain functional connectivity within 1 month of traumatic brain injury and there may be differences based on the time since injury.

LIMITATIONS

Studies of military, sports-related traumatic brain injury, and pediatric patients were excluded. Due to the high number of relevant studies and data heterogeneity, we could not be as granular in the analysis as we would have liked.

CONCLUSIONS

Reported functional connectivity changes varied, even within the same region and network, at least partially reflecting differences in technical parameters, preprocessing software, and analysis methods as well as probable differences in individual injury. There is a need for novel rs-fMRI techniques that better capture subject-specific functional connectivity changes.

A cytomics-on-a-chip platform and diagnostic model stratifies risk for oral lichenoid conditions

OBJECTIVE

A small fraction of oral lichenoid conditions (OLC) have potential for malignant transformation. Distinguishing OLCs from other oral potentially malignant disorders (OPMDs) can help prevent unnecessary concern or testing, but accurate identification by nonexpert clinicians is challenging due to overlapping clinical features. In this study, the authors developed a 'cytomics-on-a-chip' tool and integrated predictive model for aiding the identification of OLCs.

STUDY DESIGN

All study subjects underwent both scalpel biopsy for histopathology and brush cytology. A predictive model and OLC Index comprising clinical, demographic, and cytologic features was generated to discriminate between subjects with lichenoid (OLC+) (N = 94) and nonlichenoid (OLC-) (N = 237) histologic features in a population with OPMDs.

RESULTS

The OLC Index discriminated OLC+ and OLC- subjects with area under the curve (AUC) of 0.76. Diagnostic accuracy of the OLC Index was not significantly different from expert clinician impressions, with AUC of 0.81 (P = .0704). Percent agreement was comparable across all raters, with 83.4% between expert clinicians and histopathology, 78.3% between OLC Index and expert clinician, and 77.3% between OLC Index and histopathology.

CONCLUSIONS

The cytomics-on-a-chip tool and integrated diagnostic model have the potential to facilitate both the triage and diagnosis of patients presenting with OPMDs and OLCs.

Evaluation of Socioeconomic Disparities in Follow-Up Completion for Incidental Pulmonary Nodules

PURPOSE

The aim of this study was to evaluate the association between census tract-level measures of social vulnerability and residential segregation and incidental pulmonary nodule (IPN) follow-up.

METHODS

This retrospective cohort study included patients with IPNs ≥6 mm in size or multiple subsolid or ground-glass IPNs <6 mm (with nonoptional follow-up recommendations) diagnosed between January 1, 2018, and December 30, 2019, at a large urban tertiary center and followed for ≥2 years. Geographic sociodemographic context was characterized by the 2018 Centers for Disease Control and Prevention Social Vulnerability Index (SVI) and the index of concentration at the extremes (ICE), categorized in quartiles. Multivariable binomial regression models were used, with a primary outcome of inappropriate IPN follow-up (late or no follow-up). Models were also stratified by nodule risk.

RESULTS

The study consisted of 2,492 patients (mean age, 65.6 ± 12.6 years; 1,361 women). Top-quartile SVI patients were more likely to have inappropriate follow-up (risk ratio [RR], 1.24; 95% confidence interval [CI], 1.12-1.36) compared with the bottom quartile; risk was also elevated in top-quartile SVI subcategories of socioeconomic status (RR, 1.23; 95% CI, 1.13-1.34), Minority status and language (RR, 1.24; 95% CI, 1.03-1.48), housing and transportation (RR, 1.13; 95% CI, 1.02-1.26), and ICE (RR, 1.20; 95% CI, 1.11-1.30). Furthermore, top-quartile ICE was associated with greater risk for inappropriate follow-up among high-risk versus lower risk IPNs (RR, 1.33 [95% CI, 1.18-1.50] versus 1.13 [95% CI, 1.02-1.25]), respectively; P for interaction = .017).

CONCLUSIONS

Local social vulnerability and residential segregation are associated with inappropriate IPN follow-up and may inform policy or interventions tailored for neighborhoods.

Chronic kidney disease and risk of kidney or urothelial malignancy: systematic review and meta-analysis

BACKGROUND

Chronic kidney disease (CKD) is highly prevalent, affecting approximately 11% of U.S. adults. Multiple studies have evaluated a potential association between CKD and urinary tract malignancies. Summary estimates of urinary tract malignancy risk in CKD patients with and without common co-existing conditions may guide clinical practice recommendations.

METHODS

Four electronic databases were searched for original cohort studies evaluating the association between CKD and urinary tract cancers (kidney cancer and urothelial carcinoma) through May 25, 2023, in persons with at least moderate CKD and no dialysis or kidney transplantation. Quality assessment was performed for studies meeting inclusion criteria using the Newcastle-Ottawa Scale. Meta-analysis with a random-effects model was performed for unadjusted incidence rate ratios (IRR) as well as adjusted hazard ratios (aHR) for confounding conditions (diabetes, hypertension, and/or tobacco use), shown to have association with kidney cancer and urothelial carcinoma. Sub-analysis was conducted for estimates associated with CKD stages separately.

RESULTS

Six cohort studies with 8 617 563 persons were included. Overall, methodological quality of the studies was good. CKD was associated with both higher unadjusted incidence and adjusted hazard of kidney cancer (IRR, 3.36; 95% confidence interval [CI], 2.32-4.88; aHR, 2.04; 95% CI, 1.77-2.36) and urothelial cancer (IRR, 3.96; 95% CI, 2.44-6.40; aHR, 1.40; 95% CI, 1.22-1.68) compared with persons without CKD. Examining incident urinary tract cancers by CKD severity, risks were elevated in stage 3 CKD (kidney aHR, 1.89; 95% CI, 1.56-2.30; urothelial carcinoma aHR, 1.40; 95% CI, 1.18-1.65) as well as in stages 4/5 CKD (kidney cancer aHR, 2.30; 95% CI, 2.00-2.66, UC aHR, 1.24; 95% CI, 1.04-1.49).

CONCLUSIONS

Even moderate CKD is associated with elevated risk of kidney cancer and UC. Providers should consider these elevated risks when managing individuals with CKD, particularly when considering evaluation for the presence and etiology of hematuria.

How to Critically Appraise and Interpret Systematic Reviews and Meta-Analyses of Diagnostic Accuracy: A User Guide

Systematic reviews of diagnostic accuracy studies can provide the best available evidence to inform decisions regarding the use of a diagnostic test. In this guide, the authors provide a practical approach for clinicians to appraise diagnostic accuracy systematic reviews and apply their results to patient care. The first step is to identify an appropriate systematic review with a research question matching the clinical scenario. The user should evaluate the rigor of the review methods to evaluate its credibility (Did the review use clearly defined eligibility criteria, a comprehensive search strategy, structured data collection, risk of bias and applicability appraisal, and appropriate meta-analysis methods?). If the review is credible, the next step is to decide whether the diagnostic performance is adequate for clinical use (Do sensitivity and specificity estimates exceed the threshold that makes them useful in clinical practice? Are these estimates sufficiently precise? Is variability in the estimates of diagnostic accuracy across studies explained?). Diagnostic accuracy systematic reviews that are judged to be credible and provide diagnostic accuracy estimates with sufficient certainty and relevance are the most useful to inform patient care. This review discusses comparative, noncomparative, and emerging approaches to systematic reviews of diagnostic accuracy using a clinical scenario and examples based on recent publications.

White Paper: Best Practices in the Communication and Management of Actionable Incidental Findings in Emergency Department Imaging

PURPOSE

Actionable incidental findings (AIFs) are common in radiologic imaging. Imaging is commonly performed in emergency department (ED) visits, and AIFs are frequently encountered, but the ED presents unique challenges for communication and follow-up of these findings. The authors formed a multidisciplinary panel to seek consensus regarding best practices in the reporting, communication, and follow-up of AIFs on ED imaging tests.

METHODS

A 15-member panel was formed, nominated by the ACR and American College of Emergency Physicians, to represent radiologists, emergency physicians, patients, and those involved in health care systems and quality. A modified Delphi process was used to identify areas of best practice and seek consensus. The panel identified four areas: (1) report elements and structure, (2) communication of findings with patients, (3) communication of findings with clinicians, and (4) follow-up and tracking systems. A survey was constructed to seek consensus and was anonymously administered in two rounds, with a priori agreement requiring at least 80% consensus. Discussion occurred after the first round, with readministration of questions where consensus was not initially achieved.

RESULTS

Consensus was reached in the four areas identified. There was particularly strong consensus that AIFs represent a system-level issue, with need for approaches that do not depend on individual clinicians or patients to ensure communication and completion of recommended follow-up.

CONCLUSIONS

This multidisciplinary collaboration represents consensus results on best practices regarding the reporting and communication of AIFs in the ED setting.

How to Perform Economic Evaluation in Implementation Studies: Imaging-Specific Considerations and Comparison of Financial Models

Economic evaluation for implementation science merits unique considerations for a local context, including the main audience of local decision makers. This local context is in contrast with traditional methods for developing coverage policy for medical tests and interventions, which typically emphasize benefits and costs more broadly, for society. Regardless of the strength of evidence backing the efficacy or effectiveness of a clinical intervention, local context is paramount when implementing evidence-based practices. Understanding the costs throughout the processes of implementing a program will inform the decision of whether to plan for and adopt the program, how to sustain the program, and whether to scale up widely. To guide economic evaluation for implementation of evidence-based imaging practices, we describe approaches that consider local stakeholders' needs and connect these with outcomes of cost and clinical utility. Illustrative examples of implementation strategies and economic evaluation are explored in areas of cancer screening and care delivery.

Improving breast cancer diagnostics with deep learning for MRI

Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) has a high sensitivity in detecting breast cancer but often leads to unnecessary biopsies and patient workup. We used a deep learning (DL) system to improve the overall accuracy of breast cancer diagnosis and personalize management of patients undergoing DCE-MRI. On the internal test set (n = 3936 exams), our system achieved an area under the receiver operating characteristic curve (AUROC) of 0.92 (95% CI: 0.92 to 0.93). In a retrospective reader study, there was no statistically significant difference (P = 0.19) between five board-certified breast radiologists and the DL system (mean ΔAUROC, +0.04 in favor of the DL system). Radiologists' performance improved when their predictions were averaged with DL's predictions [mean ΔAUPRC (area under the precision-recall curve), +0.07]. We demonstrated the generalizability of the DL system using multiple datasets from Poland and the United States. An additional reader study on a Polish dataset showed that the DL system was as robust to distribution shift as radiologists. In subgroup analysis, we observed consistent results across different cancer subtypes and patient demographics. Using decision curve analysis, we showed that the DL system can reduce unnecessary biopsies in the range of clinically relevant risk thresholds. This would lead to avoiding biopsies yielding benign results in up to 20% of all patients with BI-RADS category 4 lesions. Last, we performed an error analysis, investigating situations where DL predictions were mostly incorrect. This exploratory work creates a foundation for deployment and prospective analysis of DL-based models for breast MRI.

Does histologic subtype impact overall survival in observed T1a kidney cancers compared with competing risks? Implications for biopsy as a risk stratification tool

OBJECTIVES

We sought to assess if adding a biopsy proven histologic subtype to a model that predicts overall survival that includes variables representing competing risks in observed, biopsy proven, T1a renal cell carcinomas, enhances the model's performance.

METHODS

The National Cancer Database was assessed (years 2004-2015) for patients with observed T1a renal cell carcinoma who had undergone renal mass biopsy. Kaplan-Meier curves were utilized to estimate overall survival stratified by histologic subtype. We utilized C-index from a Cox proportional hazards model to evaluate the impact of adding histologic subtypes to a model to predict overall survival for each stage.

RESULTS

Of 132 958 T1a renal masses identified, 1614 had biopsy proven histology and were managed non-operatively. Of those, 61% were clear cell, 33% papillary, and 6% chromophobe. Adjusted Kaplan-Meier curves demonstrated a difference in overall survival between histologic subtypes (P = 0.010) with greater median overall survival for patients with chromophobe (85.1 months, hazard rate 0.45, P = 0.005) compared to clear cell (64.8 months, reference group). Adding histology to a model with competing risks alone did not substantially improve model performance (C-index 0.65 vs 0.64 respectively).

CONCLUSIONS

Incorporation of histologic subtype into a risk stratification model to determine prognostic overall survival did not improve modeling of overall survival compared with variables representing competing risks in patients with T1a renal cell carcinoma managed with observation. These results suggest that performing renal mass biopsy in order to obtain tumor histology may have limited utility. Future studies should further investigate the overall utility of renal mass biopsy for observed T1a kidney cancers.

18F-Fluciclovine versus PSMA PET Imaging in Primary Tumor Detection during Initial Staging of High-Risk Prostate Cancer: A Systematic Review and Meta-Analysis

Purpose Fluorine 18 (¹⁸F)-fluciclovine and prostate-specific membrane antigen (PSMA) tracers are commonly used for localizing biochemical recurrence of prostate cancer, but their accuracy in primary tumor detection in the initial staging of high-risk prostate cancer has not been established. Materials and Methods A systematic review was performed of the electronic databases for original studies published between 2012 and 2020. Included studies were those in which ¹⁸F-fluciclovine or PSMA PET was used for initial staging of patients with high-risk prostate cancer. The diagnostic performance data were collected for primary tumor with histopathologic results as reference standard. The Quality Assessment of Diagnostic Accuracy Studies-2 tool was used for quality appraisal. A random-effects model was used to summarize the effect sizes and to evaluate the difference between two groups. Results Overall, 28 studies met the eligibility criteria, and 17 were included in the meta-analysis (¹⁸F-fluciclovine = 4, PSMA = 13). Of these 17 studies, 12 (70%) were judged to have high risk of bias in one of the evaluated domains, and nine studies were deemed to have applicability concerns. The pooled sensitivity, specificity, and diagnostic odds ratio for ¹⁸F-fluciclovine versus PSMA were 85% (95% CI: 73%, 92%) versus 84% (95% CI: 77%, 89%) (P = .78), 77% (95% CI: 60%, 88%) versus 83% (95% CI: 76%, 89%) (P = .40), and 18.88 (95% CI: 5.01, 71.20) versus 29.37 (95% CI: 13.35, 64.60) (P = .57), respectively, with no significant difference in diagnostic test accuracy. Conclusion ¹⁸F-fluciclovine and PSMA PET demonstrated no statistically significant difference in diagnostic accuracy in primary tumor detection during initial staging of high-risk prostate cancer. Keywords: PET, Prostate, Molecular Imaging-Cancer, Staging Supplemental material is available for this article. © RSNA, 2022.

Outcomes of Incidental Lung Nodules With Structured Recommendations and Electronic Tracking

OBJECTIVE

To evaluate the impact of structured recommendations on follow-up completion for incidental lung nodules (ILNs).

METHODS

Patients with ILNs before and after implementation of structured Fleischner recommendations and electronic tracking were sampled randomly. The cohorts were compared for imaging follow-up. Multivariable logistic regression was used to assess appropriate follow-up and loss to follow-up, with independent variables including use of structured recommendations or tracking, age, gender, race, ethnicity, setting of the index test (inpatient, outpatient, emergency department), smoking history, and nodule features.

RESULTS

In all, 1,301 patients met final inclusion criteria, including 255 patients before and 1,046 patients after structured recommendations or tracking. Baseline differences were found in the pre- and postintervention groups, with smaller ILNs and younger age after implementing structured recommendations. Comparing pre- versus postintervention outcomes, 40.0% (100 of 250) versus 29.5% (309 of 1,046) of patients had no follow-up despite Fleischner indications for imaging (P = .002), and among the remaining patients, 56.6% (82 of 145) versus 75.0% (553 of 737) followed up on time (P < .001). Delayed follow-up was more frequent before intervention. Differences postintervention were mostly accounted for by nodules ≤ 8 mm in the outpatient setting (P < .001). In multivariable analysis, younger age, White race, outpatient setting, and larger nodule size showed significant association with appropriate follow-up completion (P < .015), but structured recommendations did not. Similar results applied for loss to follow-up.

DISCUSSION

Consistent use of structured reporting is likely key to mitigate selection bias when benchmarking rates of appropriate follow-up of ILN. Emergency department patients and inpatients are at high risk of missed or delayed follow-up despite structured recommendations.

ACR Appropriateness Criteria® Staging and Follow-up of Primary Vaginal Cancer

Primary vaginal cancer is rare, comprising 1% to 2% of gynecologic malignancies and 20% of all malignancies involving the vagina. More frequently, the vagina is involved secondarily by direct invasion from malignancies originating in adjacent organs or by metastases from other pelvic or extrapelvic primary malignancies. Data on the use of imaging in vaginal cancer are sparse. Insights are derived from the study of imaging in cervical cancer and have reasonable generalizability to vaginal cancer due to similar tumor biology. Given the trend toward definitive chemoradiation for both cancers in all but early stage lesions, principles of postchemoradiation tumor response evaluation are largely analogous. Accordingly, many of the recommendations outlined here are informed by principles translated from the literature on cervical cancer. For pretreatment assessment of local tumor burden and in the case of recurrent vaginal cancer, MRI is the preferred imaging modality. PET/CT has demonstrated utility for the detection of nodal metastatic and unexpected distant metastatic disease. The American College of Radiology Appropriateness Criteria are evidence-based guidelines for specific clinical conditions that are reviewed annually by a multidisciplinary expert panel. The guideline development and revision include an extensive analysis of current medical literature from peer reviewed journals and the application of well-established methodologies (RAND/UCLA Appropriateness Method and Grading of Recommendations Assessment, Development, and Evaluation or GRADE) to rate the appropriateness of imaging and treatment procedures for specific clinical scenarios. In those instances where evidence is lacking or equivocal, expert opinion may supplement the available evidence to recommend imaging or treatment.

Active Surveillance Strategies for Low-Grade Prostate Cancer: Comparative Benefits and Cost-effectiveness

Background Active surveillance (AS) is the recommended treatment option for low-risk prostate cancer (PC). Surveillance varies in MRI, frequency of follow-up, and the Prostate Imaging Reporting and Data System (PI-RADS) score that would repeat biopsy. Purpose To compare the effectiveness and cost-effectiveness of AS strategies for low-risk PC with versus without MRI. Materials and Methods This study developed a mathematical model to evaluate the cost-effectiveness of surveillance strategies in a simulation of men with a diagnosis of low-risk PC. The following strategies were compared: watchful waiting, prostate-specific antigen (PSA) and annual biopsy without MRI, and PSA testing and MRI with varied PI-RADS thresholds for biopsy. MRI strategies differed regarding scheduling and use of PI-RADS score of at least 3, or a PI-RADS score of at least 4 to indicate the need for biopsy. Life-years, quality-adjusted life-years (QALYs), and incremental cost-effectiveness ratios were calculated by using microsimulation. Sensitivity analysis was used to assess the impact of varying parameter values on results. Results For the base case of 60-year-old men, all strategies incorporating prostate MRI extended QALYs and life-years compared with watchful waiting and non-MRI strategies. Annual MRI strategies yielded 16.19 QALYs, annual biopsy with no MRI yielded 16.14 QALYs, and watchful waiting yielded 15.94 QALYs. Annual MRI with PI-RADS score of at least 3 or of at least 4 as the biopsy threshold and annual MRI with biopsy even after MRI with negative findings offered similar QALYs and the same unadjusted life expectancy: 23.05 life-years. However, a PI-RADS score of at least 4 yielded 42% fewer lifetime biopsies. With a cost-effectiveness threshold of $100 000 per QALY, annual MRI with biopsy for lesions with PI-RADS scores of 4 or greater was most cost-effective (incremental cost-effectiveness ratio, $67 221 per QALY). Age, treatment type, risk of initial grade misclassification, and quality-of-life impact of procedural complications affected results. Conclusion The use of active surveillance (AS) with biopsy decisions guided by findings from annual MRI reduces the number of biopsies while preserving life expectancy and quality of life. Biopsy in lesions with PI-RADS scores of 4 or greater is likely the most cost-effective AS strategy for men with low-risk prostate cancer who are younger than 70 years. © RSNA, 2021 Online supplemental material is available for this article. An earlier incorrect version appeared online. This article was corrected on July 13, 2021.

Comparison of Non-Tumoral Portal Vein Thrombosis Management in Cirrhotic Patients: TIPS Versus Anticoagulation Versus No Treatment

BACKGROUND

There is a lack of consensus in optimal management of portal vein thrombosis (PVT) in patients with cirrhosis. The purpose of this study is to compare the safety and thrombosis burden change for cirrhotic patients with non-tumoral PVT managed by transjugular intrahepatic portosystemic shunt (TIPS) only, anticoagulation only, or no treatment.

METHODS

This single-center retrospective study evaluated 52 patients with cirrhosis and non-tumoral PVT managed by TIPS only (14), anticoagulation only (11), or no treatment (27). The demographic, clinical, and imaging data for patients were collected. The portomesenteric thrombosis burden and liver function tests at early follow-up (6-9 months) and late follow-up (9-16 months) were compared to the baseline. Adverse events including bleeding and encephalopathy were recorded.

RESULTS

The overall portomesenteric thrombosis burden improved in eight (72%) TIPS patients, three (27%) anticoagulated patients, and two (10%) untreated patients at early follow-up (p = 0.001) and in seven (78%) TIPS patients, two (29%) anticoagulated patients, and three (17%) untreated patients in late follow-up (p = 0.007). No bleeding complications attributable to anticoagulation were observed.

CONCLUSION

TIPS decreased portomesenteric thrombus burden compared to anticoagulation or no treatment for cirrhotic patients with PVT. Both TIPS and anticoagulation were safe therapies.

ACR Appropriateness Criteria® Staging and Follow-up of Vulvar Cancer

Vulvar cancer is an uncommon gynecologic tumor and one of several human papillomavirus-associated malignancies. Squamous cell carcinoma is the most prevalent histologic subtype of vulvar cancer, accounting for the majority of cases. Imaging plays an important role in managing vulvar cancer. At initial diagnosis, imaging is useful to assess the size and extent of primary tumor and to evaluate the status of inguinofemoral lymph nodes. If recurrent disease is suspected, imaging is essential to demonstrate local extent of tumor and to identify lymph node and distant metastases. In this publication, we summarize the recent literature and describe the panel's recommendations about the appropriate use of imaging for various phases of patient management including initial staging, surveillance, and restaging of vulvar cancer. The American College of Radiology Appropriateness Criteria are evidence-based guidelines for specific clinical conditions that are reviewed annually by a multidisciplinary expert panel. The guideline development and revision include an extensive analysis of current medical literature from peer reviewed journals and the application of well-established methodologies (RAND/UCLA Appropriateness Method and Grading of Recommendations Assessment, Development, and Evaluation or GRADE) to rate the appropriateness of imaging and treatment procedures for specific clinical scenarios. In those instances where evidence is lacking or equivocal, expert opinion may supplement the available evidence to recommend imaging or treatment.

When does nonadherence indicate a deviation from patient-centered care?

Patient-centered care, defined as "providing care that is respectful of, and responsive to, individual patient preferences, needs and values, and ensuring that patient values guide all clinical decisions," is advocated by clinicians and professional organizations and is part of a composite criterion for augmented reimbursement for various health care settings, including patient-centered medical homes. Despite general agreement that patient-centered care is a good idea and worthy of incentivization, patient-centered care is difficult to assess accurately, scalably, and feasibly. In this commentary, we suggest that assessment of patient-centered care may be improved by identifying circumstances that indicate its probable absence-in particular, by flagging probable discordance between a patient's preferences and their treatment care plan. One potential marker of this discordance is persistent lack of control of a comorbid condition that is easily controllable by existing therapies and where existing therapies are sufficiently diverse to be compatible with a wide range of patient preferences (eg, stage 1 hypertension, type 2 diabetes with glycated hemoglobin < 8.5%). We outline how this approach may be tested, validated, and harmonized with existing quality improvement activities.

Overall Survival of Biopsy-confirmed T1B and T2A Kidney Cancers Managed With Observation: Prognostic Value of Tumor Histology

INTRODUCTION

The natural history of T1b (4-7 cm) or T2a (> 7-10 cm) kidney cancers managed with observation is not well-understood. The aim of our study was to determine if the addition of histologic subtype to a predictive model of overall survival (OS) that includes covariates for competing risks in observed, biopsy-proven, T1b and T2a renal cell carcinomas (RCCs) improves the model's performance.

MATERIALS AND METHODS

We queried the National Cancer Database for patients with biopsy-proven stage T1b or T2a RCC and managed nonoperatively between 2004 and 2015. OS was estimated by Kaplan-Meier curves based on histologic subtype. The concordance index (c-index) from a Cox proportional hazards model was used to estimate the extent to which histologic subtypes predict survival for each stage when included in a model along with competing risks of age, gender, race/ethnicity, insurance status, area-level socioeconomic indicators, Charlson-Deyo index, and tumor grade.

RESULTS

A total of 937 patients (754 with T1b and 185 with T2a) with biopsy-proven RCC were identified. Kaplan-Meier analysis suggested differences in OS by histologic subtype where sarcomatoid, followed by clear cell, papillary, and chromophobe, had the highest mortality risk at 1, 3, and 5 years. However, there was marginal improvement in the multivariable model of OS using competing risks and histology (c-index, 0.64 and 0.697) compared with competing risks alone (c-index, 0.631 and 0.671) for T1b and T2a RCCs, respectively.

CONCLUSIONS

In patients with T1b or T2a RCC managed with observation, incorporation of histologic subtype into a risk-stratification model to determine prognostic OS did not improve modeling of OS compared with variables representing competing risks. Histologic subtype of observed T1b and T2a RCC appears to have prognostic OS value when not considering competing risks. These findings may impact the usefulness of renal biopsy to inform decision-making when managing patients with T1b and T2a renal tumors with observation.

Point-of-care characterization and risk-based management of oral lesions in primary dental clinics: A simulation model

OBJECTIVES

Oral potentially malignant disorders (OPMDs) encompass histologically benign, dysplastic, and cancerous lesions that are often indistinguishable by appearance and inconsistently managed. We assessed the potential impact of test-and-treat pathways enabled by a point-of-care test for OPMD characterization.

MATERIALS AND METHODS

We constructed a decision-analytic model to compare life expectancy of test-treat strategies for 60-year-old patients with OPMDs in the primary dental setting, based on a trial for a point-of-care cytopathology tool (POCOCT). Eight strategies of OPMD detection and evaluation were compared, involving deferred evaluation (no further characterization), prompt OPMD characterization using POCOCT measurements, or the commonly recommended usual care strategy of routine referral for scalpel biopsy. POCOCT pathways differed in threshold for additional intervention, including surgery for any dysplasia or malignancy, or for only moderate or severe dysplasia or cancer. Strategies with initial referral for biopsy also reflected varied treatment thresholds in current practice between surgery and surveillance of mild dysplasia. Sensitivity analysis was performed to assess the impact of variation in parameter values on model results.

RESULTS

Requisite referral for scalpel biopsy offered the highest life expectancy of 20.92 life-years compared with deferred evaluation (+0.30 life-years), though this outcome was driven by baseline assumptions of limited patient adherence to surveillance using POCOCT. POCOCT characterization and surveillance offered only 0.02 life-years less than the most biopsy-intensive strategy, while resulting in 27% fewer biopsies. When the probability of adherence to surveillance and confirmatory biopsy was ≥ 0.88, or when metastasis rates were lower than reported, POCOCT characterization extended life-years (+0.04 life-years) than prompt specialist referral.

CONCLUSION

Risk-based OPMD management through point-of-care cytology may offer a reasonable alternative to routine referral for specialist evaluation and scalpel biopsy, with far fewer biopsies. In patients who adhere to surveillance protocols, POCOCT surveillance may extend life expectancy beyond biopsy and follow up visual-tactile inspection.

Nuclear F-actin Cytology in Oral Epithelial Dysplasia and Oral Squamous Cell Carcinoma

Oral cavity cancer has a low 5-y survival rate, but outcomes improve when the disease is detected early. Cytology is a less invasive method to assess oral potentially malignant disorders relative to the gold-standard scalpel biopsy and histopathology. In this report, we aimed to determine the utility of cytological signatures, including nuclear F-actin cell phenotypes, for classifying the entire spectrum of oral epithelial dysplasia and oral squamous cell carcinoma. We enrolled subjects with oral potentially malignant disorders, subjects with previously diagnosed malignant lesions, and healthy volunteers without lesions and obtained brush cytology specimens and matched scalpel biopsies from 486 subjects. Histopathological assessment of the scalpel biopsy specimens classified lesions into 6 categories. Brush cytology specimens were analyzed by machine learning classifiers trained to identify relevant cytological features. Multimodal diagnostic models were developed using cytology results, lesion characteristics, and risk factors. Squamous cells with nuclear F-actin staining were associated with early disease (i.e., lower proportions in benign lesions than in more severe lesions), whereas small round parabasal-like cells and leukocytes were associated with late disease (i.e., higher proportions in severe dysplasia and carcinoma than in less severe lesions). Lesions with the impression of oral lichen planus were unlikely to be either dysplastic or malignant. Cytological features substantially improved upon lesion appearance and risk factors in predicting squamous cell carcinoma. Diagnostic models accurately discriminated early and late disease with AUCs (95% CI) of 0.82 (0.77 to 0.87) and 0.93 (0.88 to 0.97), respectively. The cytological features identified here have the potential to improve screening and surveillance of the entire spectrum of oral potentially malignant disorders in multiple care settings.

Planning considerations prior to laryngectomy for a patient infected with severe acute respiratory syndrome coronavirus-2 pre-operatively

BACKGROUND

Coronavirus disease 2019 was declared a pandemic on 11th March 2020. All non-urgent surgical procedures have been postponed indefinitely. The British Association of Head and Neck Oncology state that only those with treatable head and neck cancer unsuitable for alternative treatment should undergo surgery. This paper details our management of a patient who tested positive for severe acute respiratory syndrome coronavirus-2 days before curative surgery for laryngeal cancer.

CASE REPORT

By following British Association of Head and Neck Oncology guidance, a 49-year-old male scheduled for total laryngectomy and bilateral neck dissection for a T3 transglottic squamous cell cancer was pre-operatively identified as an asymptomatic carrier of severe acute respiratory syndrome coronavirus-2. Following 14-day isolation and laboratory proven viral clearance, he underwent successful major surgery. He was managed throughout the peri- and post-operative phases without complications or adverse effects on staff.

CONCLUSION

With careful planning, previous coronavirus disease 2019 positive status should not prevent an individual from undergoing successful total laryngectomy and bilateral neck dissection in a safe and timely manner during the pandemic.

ACR Appropriateness Criteria® Pretreatment Evaluation and Follow-Up of Endometrial Cancer

To date, there is little consensus on the role of pelvic imaging in assessing local disease extent during initial staging in patients with endometrial carcinoma, with practices differing widely across centers. However, when pretreatment assessment of local tumor extent is indicated, MRI is the preferred imaging modality. Preoperative imaging of endometrial carcinoma can define the extent of disease and indicate the need for subspecialist referral in the presence of deep myometrial invasion, cervical extension, or suspected lymphadenopathy. If distant metastatic disease is clinically suspected, preoperative assessment with cross-sectional imaging or PET/CT may be performed. However, most patients with low-grade disease are at low risk of lymph node and distant metastases. Thus, this group may not require a routine pretreatment evaluation for distant metastases. Recurrence rates in patients with endometrial carcinoma are infrequent. Therefore, radiologic evaluation is typically used only to investigate suspicion of recurrent disease due to symptoms or physical examination and not for routine surveillance after treatment. The American College of Radiology Appropriateness Criteria are evidence-based guidelines for specific clinical conditions that are reviewed annually by a multidisciplinary expert panel. The guideline development and revision include an extensive analysis of current medical literature from peer reviewed journals and the application of well-established methodologies (RAND/UCLA Appropriateness Method and Grading of Recommendations Assessment, Development, and Evaluation or GRADE) to rate the appropriateness of imaging and treatment procedures for specific clinical scenarios. In those instances where evidence is lacking or equivocal, expert opinion may supplement the available evidence to recommend imaging or treatment.

ACR Appropriateness Criteria® Gestational Trophoblastic Disease

Gestational trophoblastic disease (GTD), a rare complication of pregnancy, includes both benign and malignant forms, the latter collectively referred to as gestational trophoblastic neoplasia (GTN). When metastatic, the lungs are the most common site of initial spread. Beta-human chorionic gonadotropin, elaborated to some extent by all forms of GTD, is useful in facilitating disease detection, diagnosis, monitoring treatment response, and follow-up. Imaging evaluation depends on whether GTD manifests in one of its benign forms or whether it has progressed to GTN. Transabdominal and transvaginal ultrasound with duplex Doppler evaluation of the pelvis are usually appropriate diagnostic procedures in either of these circumstances, and in posttreatment surveillance. The appropriateness of more extensive imaging remains dependent on a diagnosis of GTN and on other factors. The use of imaging to assess complications, typically hemorrhagic, should be guided by the location of clinical signs and symptoms. The American College of Radiology Appropriateness Criteria are evidence-based guidelines for specific clinical conditions that are reviewed annually by a multidisciplinary expert panel. The guideline development and revision include an extensive analysis of current medical literature from peer reviewed journals and the application of well-established methodologies (RAND/UCLA Appropriateness Method and Grading of Recommendations Assessment, Development, and Evaluation or GRADE) to rate the appropriateness of imaging and treatment procedures for specific clinical scenarios. In those instances where evidence is lacking or equivocal, expert opinion may supplement the available evidence to recommend imaging or treatment.

Managing COVID-19 With a Clinical Decision Support Tool in a Community Health Network: Algorithm Development and Validation

BACKGROUND

The coronavirus disease (COVID-19) pandemic has resulted in significant morbidity and mortality; large numbers of patients require intensive care, which is placing strain on health care systems worldwide. There is an urgent need for a COVID-19 disease severity assessment that can assist in patient triage and resource allocation for patients at risk for severe disease.

OBJECTIVE

The goal of this study was to develop, validate, and scale a clinical decision support system and mobile app to assist in COVID-19 severity assessment, management, and care.

METHODS

Model training data from 701 patients with COVID-19 were collected across practices within the Family Health Centers network at New York University Langone Health. A two-tiered model was developed. Tier 1 uses easily available, nonlaboratory data to help determine whether biomarker-based testing and/or hospitalization is necessary. Tier 2 predicts the probability of mortality using biomarker measurements (C-reactive protein, procalcitonin, D-dimer) and age. Both the Tier 1 and Tier 2 models were validated using two external datasets from hospitals in Wuhan, China, comprising 160 and 375 patients, respectively.

RESULTS

All biomarkers were measured at significantly higher levels in patients who died vs those who were not hospitalized or discharged (P<.001). The Tier 1 and Tier 2 internal validations had areas under the curve (AUCs) of 0.79 (95% CI 0.74-0.84) and 0.95 (95% CI 0.92-0.98), respectively. The Tier 1 and Tier 2 external validations had AUCs of 0.79 (95% CI 0.74-0.84) and 0.97 (95% CI 0.95-0.99), respectively.

CONCLUSIONS

Our results demonstrate the validity of the clinical decision support system and mobile app, which are now ready to assist health care providers in making evidence-based decisions when managing COVID-19 patient care. The deployment of these new capabilities has potential for immediate impact in community clinics and sites, where application of these tools could lead to improvements in patient outcomes and cost containment.

Clinical decision support tool and rapid point-of-care platform for determining disease severity in patients with COVID-19

SARS-CoV-2 is the virus that causes coronavirus disease (COVID-19) which has reached pandemic levels resulting in significant morbidity and mortality affecting every inhabited continent. The large number of patients requiring intensive care threatens to overwhelm healthcare systems globally. Likewise, there is a compelling need for a COVID-19 disease severity test to prioritize care and resources for patients at elevated risk of mortality. Here, an integrated point-of-care COVID-19 Severity Score and clinical decision support system is presented using biomarker measurements of C-reactive protein (CRP), N-terminus pro B type natriuretic peptide (NT-proBNP), myoglobin (MYO), D-dimer, procalcitonin (PCT), creatine kinase-myocardial band (CK-MB), and cardiac troponin I (cTnI). The COVID-19 Severity Score combines multiplex biomarker measurements and risk factors in a statistical learning algorithm to predict mortality. The COVID-19 Severity Score was trained and evaluated using data from 160 hospitalized COVID-19 patients from Wuhan, China. Our analysis finds that COVID-19 Severity Scores were significantly higher for the group that died versus the group that was discharged with median (interquartile range) scores of 59 (40-83) and 9 (6-17), respectively, and area under the curve of 0.94 (95% CI 0.89-0.99). Although this analysis represents patients with cardiac comorbidities (hypertension), the inclusion of biomarkers from other pathophysiologies implicated in COVID-19 (e.g., D-dimer for thrombotic events, CRP for infection or inflammation, and PCT for bacterial co-infection and sepsis) may improve future predictions for a more general population. These promising initial models pave the way for a point-of-care COVID-19 Severity Score system to impact patient care after further validation with externally collected clinical data. Clinical decision support tools for COVID-19 have strong potential to empower healthcare providers to save lives by prioritizing critical care in patients at high risk for adverse outcomes.

Imaging for Metastatic Renal Cell Carcinoma

Patients with renal cell carcinoma may develop metastases after radical nephrectomy, and therefore monitoring with imaging for recurrent or metastatic disease is critical. Imaging varies with specific suspected site of disease. Computed tomography/MRI of the abdomen and pelvis are mainstay modalities. Osseous and central nervous system imaging is reserved for symptomatic patients. Radiologic reporting is evolving to reflect effects of systemic therapy on lesion morphology. Nuclear medicine studies compliment routine imaging as newer agents are evaluated for more accurate tumor staging. Imaging research aims to fill gaps in treatment selection and monitoring of treatment response in metastatic renal cell carcinoma.

Clinical Decision Support Tool and Rapid Point-of-Care Platform for Determining Disease Severity in Patients with COVID-19

SARS-CoV-2 is the virus that causes coronavirus disease (COVID-19) which has reached pandemic levels resulting in significant morbidity and mortality affecting every inhabited continent. The large number of patients requiring intensive care threatens to overwhelm healthcare systems globally. Likewise, there is a compelling need for a COVID-19 disease severity test to prioritize care and resources for patients at elevated risk of mortality. Here, an integrated point-of-care COVID-19 Severity Score and clinical decision support system is presented using biomarker measurements of C-reactive protein (CRP), N-terminus pro B type natriuretic peptide (NT-proBNP), myoglobin (MYO), D-dimer, procalcitonin (PCT), creatine kinase-myocardial band (CK-MB), and cardiac troponin I (cTnI). The COVID-19 Severity Score combines multiplex biomarker measurements and risk factors in a statistical learning algorithm to predict mortality. The COVID-19 Severity Score was trained and evaluated using data from 160 hospitalized COVID-19 patients from Wuhan, China. Our analysis finds that COVID-19 Severity Scores were significantly higher for the group that died versus the group that was discharged with median (interquartile range) scores of 59 (40-83) and 9 (6-17), respectively, and area under the curve of 0.94 (95% CI 0.89-0.99). These promising initial models pave the way for a point-of-care COVID-19 Severity Score system to impact patient care after further validation with externally collected clinical data. Clinical decision support tools for COVID-19 have strong potential to empower healthcare providers to save lives by prioritizing critical care in patients at high risk for adverse outcomes.

Prognostic Value of Histologic Subtype and Treatment Modality for T1a Kidney Cancers

INTRODUCTION

To evaluate overall survival (OS) of T1a kidney cancers stratified by histologic subtype and curative treatment including partial nephrectomy (PN), percutaneous ablation (PA), and radical nephrectomy (RN).

MATERIALS AND METHODS

We queried the National Cancer Data Base (2004-2015) for patients with T1a kidney cancers who were treated surgically. OS was estimated by Kaplan-Meier curves based on histologic subtype and management. Cox proportional regression models were used to determine whether histologic subtypes and management procedure predicted OS.

RESULTS

46,014 T1a kidney cancers met inclusion criteria. Kaplan Meier curves demonstrated differences in OS by treatment for clear cell, papillary, chromophobe, and cystic histologic subtypes (all p < 0.001), but no differences for sarcomatoid (p = 0.110) or collecting duct (p = 0.392) were observed. Adjusted Cox regression showed worse OS for PA than PN among patients with clear cell (HR 1.58, 95%CI [1.44-1.73], papillary RCC (1.53 [1.34-1.75]), and chromophobe RCC (2.19 [1.64-2.91]). OS was worse for RN than PN for clear cell (HR 1.38 [1.28-1.50]) papillary (1.34 [1.16-1.56]) and chromophobe RCC (1.92 [1.43-2.58]). Predictive models using Cox proportional hazards incorporating histology and surgical procedure alone were limited (c-index 0.63) while adding demographics demonstrated fair predictive power for OS (c-index 0.73).

CONCLUSIONS

In patients with pathologic T1a RCC, patterns of OS differed by surgery and histologic subtype. Patients receiving PN appears to have better prognosis than both PA and RN. However, the incorporation of histologic subtype and treatment modality into a risk stratification model to predict OS had limited utility compared with variables representing competing risks.

Point-of-care oral cytology tool for the screening and assessment of potentially malignant oral lesions

BACKGROUND

The effective detection and monitoring of potentially malignant oral lesions (PMOL) are critical to identifying early-stage cancer and improving outcomes. In the current study, the authors described cytopathology tools, including machine learning algorithms, clinical algorithms, and test reports developed to assist pathologists and clinicians with PMOL evaluation.

METHODS

Data were acquired from a multisite clinical validation study of 999 subjects with PMOLs and oral squamous cell carcinoma (OSCC) using a cytology-on-a-chip approach. A machine learning model was trained to recognize and quantify the distributions of 4 cell phenotypes. A least absolute shrinkage and selection operator (lasso) logistic regression model was trained to distinguish PMOLs and cancer across a spectrum of histopathologic diagnoses ranging from benign, to increasing grades of oral epithelial dysplasia (OED), to OSCC using demographics, lesion characteristics, and cell phenotypes. Cytopathology software was developed to assist pathologists in reviewing brush cytology test results, including high-content cell analyses, data visualization tools, and results reporting.

RESULTS

Cell phenotypes were determined accurately through an automated cytological assay and machine learning approach (99.3% accuracy). Significant differences in cell phenotype distributions across diagnostic categories were found in 3 phenotypes (type 1 ["mature squamous"], type 2 ["small round"], and type 3 ["leukocytes"]). The clinical algorithms resulted in acceptable performance characteristics (area under the curve of 0.81 for benign vs mild dysplasia and 0.95 for benign vs malignancy).

CONCLUSIONS

These new cytopathology tools represent a practical solution for rapid PMOL assessment, with the potential to facilitate screening and longitudinal monitoring in primary, secondary, and tertiary clinical care settings.

Enhancing communication in radiology using a hybrid computer-human based system

INTRODUCTION

Communication and physician burn out are major issues within Radiology. This study is designed to determine the utilization and cost benefit of a hybrid computer/human communication tool to aid in relay of clinically important imaging findings.

MATERIAL AND METHODS

Analysis of the total number of tickets, (requests for assistance) placed, the type of ticket and the turn-around time was performed. Cost analysis of a hybrid computer/human communication tool over a one-year period was based on human costs as a multiple of the time to close the ticket. Additionally, we surveyed a cohort of radiologists to determine their use of and satisfaction with this system.

RESULTS

14,911 tickets were placed in the 6-month period, of which 11,401 (76.4%) were requests to "Get the Referring clinician on the phone." The mean time to resolution (TTR) of these tickets was 35.3 (±17.4) minutes. Ninety percent (72/80) of radiologists reported being able to interpret a new imaging study instead of waiting to communicate results for the earlier study, compared to 50% previously. 87.5% of radiologists reported being able to read more cases after this system was introduced. The cost analysis showed a cost savings of up to $101.12 per ticket based on the length of time that the ticket took to close and the total number of placed tickets.

CONCLUSIONS

A computer/human communication tool can be translated to significant time savings and potentially increasing productivity of radiologists. Additionally, the system may have a cost savings by freeing the radiologist from tracking down referring clinicians prior to communicating findings.

Rescue of PB-resistant neonatal seizures with single-dose of small-molecule TrkB antagonist show long-term benefits

A recently characterized CD-1 mouse model of phenobarbital (PB)-resistant neonatal ischemic-seizures (i.e.; unilateral carotid ligation) was shown to be associated with age-dependent (P7 vs. P10) acute seizure severity and PB-efficacy (i.e.; PB-resistant vs. PB-responsive). ANA12, a novel small-molecule TrkB antagonist, rescued the PB-resistance at P7 in a dose-dependent manner and prevented the post-ischemic downregulation of KCC2, the chief Cl^- extruder in neurons. The long-term consequences of this novel rescue-intervention with ANA12 + PB in P7 and P10 ligated pups was investigated and compared to the standard first-line protocol of PB-alone loading dose. The mice underwent neurobehavioral testing, 24 h video-EEG-EMG monitoring, and immunohistochemistry in ipsi- and contralateral cortices as adults following the neonatal interventions. ANA12 + PB rescued the emergence of hyperactivity in post-ischemic P7, but not in P10 pups as adults. ANA12 + PB administration at neither P7 nor P10 significantly altered 24 h macro-sleep architecture in adults when compared to PB-alone. Behavioral state-dependent gamma (35-50 Hz) power homeostasis showed the most significant between-group differences that were age-dependent. ANA12 + PB treatment, but not PB-alone, rescued the loss of gamma power homeostasis present in P7 ligate-control but absent in P10 ligate group, highlighting the age-dependence. In contrast, PB-alone treatment, but not ANA12+PB, significantly reduced the elevated delta-AUC observed in P10 ligate-controls, when PB is efficacious by itself. These results indicate that the rescue of acute PB-resistant neonatal seizures using a novel intervention positively modulates the long-term outcomes at P7 when the seizures are refractory.

Natural Language Processing for Identification of Incidental Pulmonary Nodules in Radiology Reports

PURPOSE

To develop natural language processing (NLP) to identify incidental lung nodules (ILNs) in radiology reports for assessment of management recommendations.

METHODS AND MATERIALS

We searched the electronic health records for patients who underwent chest CT during 2014 and 2017, before and after implementation of a department-wide dictation macro of the Fleischner Society recommendations. We randomly selected 950 unstructured chest CT reports and reviewed manually for ILNs. An NLP tool was trained and validated against the manually reviewed set, for the task of automated detection of ILNs with exclusion of previously known or definitively benign nodules. For ILNs found in the training and validation sets, we assessed whether reported management recommendations agreed with Fleischner Society guidelines. The guideline concordance of management recommendations was compared between 2014 and 2017.

RESULTS

The NLP tool identified ILNs with sensitivity and specificity of 91.1% and 82.2%, respectively, in the validation set. Positive and negative predictive values were 59.7% and 97.0%. In reports of ILNs in the training and validation sets before versus after introduction of a Fleischner reporting macro, there was no difference in the proportion of reports with ILNs (108 of 500 [21.6%] versus 101 of 450 [22.4%]; P = .8), or in the proportion of reports with ILNs containing follow-up recommendations (75 of 108 [69.4%] versus 80 of 101 [79.2%]; P = .2]. Rates of recommendation guideline concordance were not significantly different before and after implementation of the standardized macro (52 of 75 [69.3%] versus 60 of 80 [75.0%]; P = .43).

CONCLUSION

NLP reliably automates identification of ILNs in unstructured reports, pertinent to quality improvement efforts for ILN management.

ACR Appropriateness Criteria® Staging and Follow-Up of Ovarian Cancer

In the management of epithelial ovarian cancers, imaging is used for cancer detection and staging, both before and after initial treatment. The decision of whether to pursue initial cytoreductive surgery for ovarian cancer depends in part on accurate staging. Contrast-enhanced CT of the abdomen and pelvis (and chest where indicated) is the current imaging modality of choice for the initial staging evaluation of ovarian cancer. Fluorine-18-2-fluoro-2-deoxy-d-glucose PET/CT and MRI may be appropriate for problem-solving purposes, particularly when lesions are present on CT but considered indeterminate. In patients who achieve remission, clinical suspicion for relapse after treatment prompts imaging evaluation for recurrence. Contrast-enhanced CT is the modality of choice to assess the extent of recurrent disease, and fluorine-18-2-fluoro-2-deoxy-d-glucose PET/CT is also usually appropriate, as small metastatic foci may be identified. If imaging or clinical examination confirms a recurrence, the extent of disease and timing of disease recurrence then determines the choice of treatments, including surgery, chemotherapy, and radiation therapy. The American College of Radiology Appropriateness Criteria are evidence-based guidelines for specific clinical conditions that are reviewed annually by a multidisciplinary expert panel. The guideline development and revision include an extensive analysis of current medical literature from peer reviewed journals and the application of well-established methodologies (RAND/UCLA Appropriateness Method and Grading of Recommendations Assessment, Development, and Evaluation or GRADE) to rate the appropriateness of imaging and treatment procedures for specific clinical scenarios. In those instances where evidence is lacking or equivocal, expert opinion may supplement the available evidence to recommend imaging or treatment.

Personalized Treatment for Small Renal Tumors: Decision Analysis of Competing Causes of Mortality

Purpose To compare the effectiveness of personalized treatment for small (≤4 cm) renal tumors versus routine partial nephrectomy (PN), accounting for various competing causes of mortality. Materials and Methods A state-transition microsimulation model was constructed to compare life expectancy of management strategies for small renal tumors by using 1 000 000 simulations in the following ways: routine PN or personalized treatment involving percutaneous ablation for risk factors for worsening chronic kidney disease (CKD), and otherwise PN; biopsy, with triage of renal cell carcinoma (RCC) to PN or ablation depending on risk factors for worsening CKD; active surveillance for growth; and active surveillance when MRI findings are indicative of papillary RCC. Transition probabilities were incorporated from the literature. Effects of parameter variability were assessed in sensitivity analysis. Results In patients of all ages with normal renal function, routine PN yielded the longest life expectancy (eg, 0.67 years in 65-year-old men with nephrometry score [NS] of 4). Otherwise, personalized strategies extended life expectancy versus routine PN: in CKD stages 2 or 3a, moderate or high NS, and no comorbidities, MRI guidance for active surveillance extended life expectancy (eg, 2.60 years for MRI vs PN in CKD 3a, NS 10); and with Charlson comorbidity index of 1 or more, biopsy or active surveillance for growth extended life expectancy (eg, 2.70 years for surveillance for growth in CKD 3a, NS 10). CKD 3b was most effectively managed by using MRI to help predict papillary RCC for surveillance. Conclusion For patients with chronic kidney disease and small renal tumors, personalized treatment selection likely extends life expectancy. © RSNA, 2019 Online supplemental material is available for this article.

Measuring the value of MRI: Comparative effectiveness & outcomes research

Magnetic resonance imaging (MRI) now provides diagnostic assessment for numerous clinical indications, including lesion detection, characterization, functional assessment, and response to treatment. To maximize the potential to improve health through the use of MRI, it is critical to investigate the impact of MRI on outcomes, and to compare the effectiveness of MRI with existing standard diagnostic approaches. Outcomes of MRI can include survival but also intermediate steps such as potential reduction in unnecessary therapy, shorter time to the appropriate therapy, or shorter periods of hospital admission. To understand the effectiveness of an imaging test's sensitivity and specificity, the results' consequences are weighed, reflecting the disease type, severity, and treatment effects. In some instances, other modalities may be faster, more readily available, or less costly than MRI but additional disease-related information or better accuracy may translate to greater population level benefit. For health policy decisions and clinical guidelines, studies of comparative outcomes can lend depth to the strength of the evidence, the specific benefits vs. harms of using one test over another, and the most effective use of the test in terms of target population. Cost effectiveness then allows for a direct comparison of approaches in terms of the cost for the projected gain in life expectancy and/or quality adjusted life expectancy. Expanding the literature on improved efficiency, accessibility, clinical effectiveness, and cost effectiveness will support the directive for better quality and value in healthcare. Level of Evidence 5 Technical Efficacy Stage 5 J. Magn. Reson. Imaging 2019.

Utility of diffusion-weighted MR imaging in the diagnosis of placenta accreta spectrum abnormality

PURPOSE

The aim of this study was to evaluate the utility of added DWI sequences as an adjunct to traditional MR imaging in the evaluation of abnormal placentation in patients with suspicion for placenta accreta spectrum abnormality or morbidly adherent placenta (MAP).

MATERIALS AND METHODS

The study was approved by local ethics committee. The subjects included pregnant women with prenatal MRI performed between July 2013 to July 2015. All imaging was performed on a Philips 1.5T MR scanner using pelvic phased-array coil. Only T2-weighted and diffusion-weighted imaging (DWI) series were compiled for review. Two randomized imaging sets were created: set 1 included T2-weighted series only (T2W); set 2 included T2W with DWI series together (T2W + DWI). Three radiologists, blinded to history and pathology, reviewed the imaging, with 2 weeks of time between the two image sets. Sensitivity, specificity, and overall accuracy for MAP were calculated and compared between T2W only and T2W + DWI reads. Associations between imaging findings and invasion on pathology were tested using the Chi-squared test. Confidence scores, inter-reader agreement, and systematic differences were documented.

RESULTS

A total of 17 pregnant women were included in the study. 8 cases were pathologically diagnosed with MAP. There were no significant differences in the diagnostic accuracy between T2W and T2W + DWI in the diagnosis of MAP in terms of overall accuracy (62.7% for T2W vs. 68.6% for T2W + DWI, p = 0.68), sensitivity (70.8% for T2W vs. 95.8% for T2W + DWI, p = 0.12), and specificity (55.6% for T2W vs. 44.4% for T2W + DWI, p = 0.49). There was no significant difference in the diagnostic confidence between the review of T2W images alone and the T2W + DWI review (mean 7.3 ± 1.8 for T2W vs. 7.5 ± 1.8 for T2W + DWI, p = 0.37).

CONCLUSION

With the current imaging technique, addition of DWI sequence to the traditional T2W images cannot be shown to significantly increase the accuracy or reader confidence for diagnosis of placenta accreta spectrum abnormality. However, DWI does improve identification of abnormalities in the placental-myometrial interface.

Use of Breast Cancer Screening and Its Association with Later Use of Preventive Services among Medicare Beneficiaries

Purpose To retrospectively assess whether there is an association between screening mammography and the use of a variety of preventive services in women who are enrolled in Medicare. Materials and Methods U.S. Medicare claims from 2010 to 2014 Research Identifiable Files were reviewed to retrospectively identify a group of women who underwent screening mammography and a control group without screening mammography in 2012. The screened group was divided into positive versus negative results at screening, and the positive subgroup was divided into false-positive and true-positive findings. Multivariate logistic regression models and inverse probability of treatment weighting were used to examine the relationship between screening status and the probabilities of undergoing Papanicolaou test, bone mass measurement, or influenza vaccination in the following 2 years. Results The cohort consisted of 555 705 patients, of whom 185 625 (33.4%) underwent mammography. After adjusting for patient demographics, comorbidities, geographic covariates, and baseline preventive care, women who underwent index screening mammography (with either positive or negative results) were more likely than unscreened women to later undergo Papanicolaou test (odds ratio [OR], 1.49; 95% confidence interval: 1.40, 1.58), bone mass measurement (OR, 1.70; 95% confidence interval: 1.63, 1.78), and influenza vaccine (OR, 1.45; 95% confidence interval: 1.37, 1.53). In women who had not undergone these preventive measures in the 2 years before screening mammography, use of these three services after false-positive findings at screening was no different than after true-negative findings at screening. Conclusion In beneficiaries of U.S. Medicare, use of screening mammography was associated with higher likelihood of adherence to other preventive guidelines, without a negative association between false-positive results and cervical cancer screening.

Sleep dysfunction following neonatal ischemic seizures are differential by neonatal age of insult as determined by qEEG in a mouse model

Neonatal seizures associated with hypoxic-ischemic encephalopathy (HIE) pose a challenge in their acute clinical management and are often followed by long-term neurological consequences. We used a newly characterized CD-1 mouse model of neonatal ischemic seizures associated with age-dependent (P7 vs. P10) seizure severity and phenobarbital efficacy (i.e.; PB-resistant vs. PB-efficacious respectively) following unilateral carotid ligation. The long-term consequences following untreated neonatal seizures in P7 vs. P10 ligated pups were investigated using neurobehavioral testing, 24 h v- quantitative EEG -EMG (qEEG, qEMG), and western blot analyses in adult mice. Significant hyperactivity emerged in a small sub-set of mice in both age-groups associated with a failure to habituate during open-field (OF) testing. 24 h continuous qEEGs detected significantly altered sleep architecture due to long-wake cycles in both age-groups. Delta power (0.5-4 Hz) quantification during slow-wave-sleep (SWS) revealed significant SWS compensation in P10 ligates following periods of increased sleep pressure which the P7 ligate group failed to show. Theta/beta ratios deemed as negative correlation markers of attentional control were significantly higher only in the P10 ligates. These results indicate that neonatal age-dependent differences in the characteristics of ischemic neonatal seizures in CD-1 pups differentially modulate long-term outcomes, when evaluated with v-qEEG/EMG as adults.