Efficacy of Whole-Blood Model of Gadolinium-Based Contrast Agent Relaxivity in Predicting Vascular MR Signal Intensity In Vivo.

BACKGROUND: Previous in vitro studies have described sub-linear longitudinal and heightened transverse H2 O relaxivities of gadolinium-based contrast agents (GBCAs) in blood due to their extracellular nature. However, in vivo validation is lacking. PURPOSE: Validate theory describing blood behavior of R1 and R2 * in an animal model. STUDY TYPE: Prospective, animal. ANIMAL MODEL: Seven swine (54-65 kg). FIELD STRENGTH/SEQUENCE: 1.5 T; time-resolved 3D spoiled gradient-recalled echo (SPGR) and quantitative Look-Locker and multi-echo fast field echo sequences. ASSESSMENT: Seven swine were each injected three times with 0.1 mmol/kg intravenous doses of one of three GBCAs: gadoteridol, gadobutrol, and gadobenate dimeglumine. Injections were randomized for rate (1, 2, and 3 mL/s) and order, during which time-resolved aortic 3D SPGR imaging was performed concurrently with aortic blood sampling via an indwelling catheter. Time-varying [GBCA] was measured by mass spectrometry of sampled blood. Predicted signal intensity (SI) was determined from a model incorporating sub-linear R1 and R2 * effects (whole-blood model) and simpler models incorporating linear R1 , with and without R2 * effects. Predicted SIs were compared to measured aortic SI. STATISTICAL TESTS: Linear correlation (coefficient of determination, R2 ) and mean errors were compared across the SI prediction models. RESULTS: There was an excellent correlation between predicted and measured SI across all injections and swine when accounting for the non-linear dependence of R1 and high blood R2 * (regression slopes 0.91-1.04, R2 ≥ 0.91). Simplified models (linear R1 with and without R2 * effects) showed poorer correlation (slopes 0.67-0.85 and 0.54-0.64 respectively, both R2 ≥ 0.89) and higher averaged mean absolute and mean square errors (128.4 and 177.4 vs. 42.0, respectively, and 5506 and 11,419 vs. 699, respectively). DATA CONCLUSION: Incorporating sub-linear R1 and high first-pass R2 * effects in arterial blood models allows accurate SPGR SI prediction in an in vivo animal model, and might be utilized when modeling MR blood SI. LEVEL OF EVIDENCE: 1 TECHNICAL EFFICACY: Stage 1.

Relationship Between Simulated Gadolinium-Based Contrast Agent Injection Profile and Achievable Resolution Metrics in Contrast-Enhanced Magnetic Resonance Angiography.

BACKGROUND: Contrast bolus variation during contrast-enhanced magnetic resonance angiography (CE-MRA) acquisition may lead to vessel blurring. PURPOSE: To combine knowledge of how contrast signal intensity (SI) evolves for different injection strategies with anatomically familiar parametric computer models to measure and visually assess the effects of a wide range of variables on modeled CE-MRA, and in doing so develop contrast rate injection guidelines. STUDY TYPE: Computer modeling. PHANTOM: Digital three-dimensional phantom consisting of orthogonal "aorta," 7 mm diameter "renal arteries" (with 57% and 86% diameter stenoses), and 7 mm diameter "superior mesenteric artery" (with 57% diameter stenosis). FIELD STRENGTH/SEQUENCE: One millimeter in-plane resolution arterial CE-MRA imaging at 3 T. ASSESSMENT: "Background" (time invariant) and "vascular" (time varying) components of the phantom were each Fourier transformed into the spatial frequency domain, the latter modulated by the SI evolution of a contrast bolus of varying "plateau" lengths and "tail" heights. Data are presented as surface plots of stenosis measurement error and blurring vs. a reference-standard injection. STATISTICAL TESTS: Descriptive. RESULTS: Shorter plateau lengths and lower tail heights resulted in increased measured stenosis error and blurring vs. the reference standard. Under a 44-second acquisition, full width half maximum stenosis error of the 86% stenosis with 25% plateau length and 25% tail height is 24% as compared to that from the reference standard. As plateau length and tail height approach 100%, stenosis error and blurring approach a floor defined by the MR acquisition's limitations. DATA CONCLUSION: We propose that to achieve minimal degradation with CE-MRA, one can create a contrast bolus with either 60% plateau and 50% tail height or 80% plateau with any tail. These considerations may well prove to be of practical importance, possibly via manipulating the tail by means of multiphasic contrast injections. LEVEL OF EVIDENCE: 3 TECHNICAL EFFICACY: Stage 1.

Comparison of Histogram-Based Gaussian Analysis With and Without Noise Correction for the Characterization of Indeterminate Adrenal Nodules.

OBJECTIVE. The purpose of this study is to determine whether gaussian-based histogram analysis without and with noise correction can characterize indeterminate adrenal nodules (those with attenuation greater than 10 HU on unenhanced CT) as lipid-poor adenomas. MATERIALS AND METHODS. This retrospective study evaluated adrenal nodules larger than 1 cm on unenhanced CT using gaussian analysis without and with noise correction on intralesional ROIs. Two independent readers who were blinded to the final diagnoses evaluated the nodules. The final diagnosis for each nodule was determined on the basis of pathologic findings or accepted imaging criteria. Interreader agreement was assessed using the intraclass correlation coefficient. Algorithm performance was summarized using sensitivity, specificity, and the AUC. RESULTS. Ninety-four adrenal nodules in 85 patients were analyzed; 36 of these were metastases (34 of which were pathologically confirmed), and 58 were presumed adenomas. Interreader agreement was excellent for nodule size, mean attenuation, SD of attenuation, and the gaussian index. Noise-corrected gaussian analysis had significantly higher specificity (81.9% vs 55.6%; p < 0.001) and lower sensitivity (36.2% vs 56.9%; p < 0.001) for identifying adenomas than did the uncorrected gaussian analysis. The AUC of corrected gaussian analysis was 0.72, which is significantly greater than that of uncorrected gaussian analysis (0.51; p ≤ 0.001) and similar to that of mean attenuation (0.77). CONCLUSION. Noise correction is necessary when using a gaussian analysis characterization of indeterminate adrenal nodules on modern unenhanced CT examinations. This method may be able to discriminate between adenomas and nonadenomas.

Evaluation of four injection profiles for uniform contrast-enhanced signal intensity profiles in MR angiography.

BACKGROUND: Gadolinium concentration variation during acquisition of contrast-enhanced MR angiography (CE-MRA) may lead to artifacts. PURPOSE: To compare signal intensity (SI) profiles of four different contrast agent injection strategies during CE-MRA with the goal of minimizing SI variation during acquisition. STUDY TYPE: Prospective. SUBJECTS: Forty subjects randomized to receive one of four injection profiles of gadobenate dimeglumine (0.1 mmol/kg), either undiluted (0.5 M) or diluted to 40 ml total volume. Tested profiles: 1) nondiluted single-phase ("standard" NS; 1.6 ml/s), 2) diluted single-phase (DS; 1.6 ml/s), 3) diluted biphasic (DB; 9 ml @ 3.3 ml/s, 29 ml @ 1.4 ml/s), 4) patient-tailored protocol using linear prediction (DT). FIELD STRENGTH/SEQUENCE: Time-resolved SI measured at 3T with spoiled gradient echo sequences having analogous parameters to those of CE-MRA. ASSESSMENT: Plateau arrival time, rise time, duration, peak and tail SI, plateau quality (sum of squared residuals; SSR), average SI for each injection type derived were used. STATISTICAL TEST: Two-tailed t-test. RESULTS: Peak SI, arrival, and rise times were not significantly different between groups, excepting peak SI DB slightly > DS (P = 0.042). Duration of NS vs. the diluted groups was significantly shorter (all P < 0.0001), and DS duration was significantly shorter than that of DT and DB (NS 11.4 ± 3.5 vs. DS 22.9 ± 4.3, DB 25.4 ± 2.3, DT 28.3 ± 4.1 sec). Quality (SSR) of the 20-second plateau was significantly better for DS, DB, DT as compared with NS (all P < 0.001). DATA CONCLUSION: Three different strategies to power-inject diluted gadobenate dimeglumine targeting a 20-second plateau produced SI profiles with longer duration, more consistent plateau, and no significant loss in peak SI. Such injection profiles may provide more uniform SI during CE-MRA, potentially reducing blurring artifacts. LEVEL OF EVIDENCE: 2 Technical Efficacy: Stage 1 J. Magn. Reson. Imaging 2019;50:1808-1816.

Increased Incidence of Pseudoaneurysm Bleeding With Lumen-Apposing Metal Stents Compared to Double-Pigtail Plastic Stents in Patients With Peripancreatic Fluid Collections.

BACKGROUND & AIMS: There have been few studies that compared the effects of lumen-apposing metal stents (LAMS) and double-pigtail plastic stents (DPS) in patients with peripancreatic fluid collections from pancreatitis. We aimed to compare technical and clinical success and adverse events in patients who received LAMS vs DPS for pancreatic pseudocysts and walled-off necrosis. METHODS: We performed a retrospective study of endoscopic ultrasound-mediated drainage in 149 patients (65% male; mean age, 47 y) with pancreatic pseudocysts or walled-off necrosis (97 received LAMS and 152 received DPS), from January 2011 through September 2016 at a single center. We collected data on patient characteristics, outcomes, hospitalizations, and imaging findings. Technical success was defined as LAMS insertion or a minimum of 2 DPS. Clinical success was defined as resolution of pancreatic pseudocysts or walled-off necrosis based on imaging results. The primary outcome was resolution of peripancreatic fluid collection with reduced abdominal pain or obstructive signs or symptoms. Secondary outcomes included the identification and management of adverse events, number of additional procedures required to resolve fluid collection, and the recurrence of fluid collection. RESULTS: Patients who received LAMS had larger peripancreatic fluid collections than patients who received DPS prior to intervention (P = .001), and underwent an average 1.7 interventions vs 1.9 interventions for patients who received DPS (P = .93). Technical success was achieved for 90 patients with LAMS (92.8%) vs 137 patients with DPS (90.1%) (odds ratio [OR] for success with DPS, 0.82; 95% CI, 0.33-2.0; P = .67). Despite larger fluid collections in the LAMS group, there was no significant difference in proportions of patients with clinical success following placement of LAMS (82 of 84 patients, 97.6%) vs DPS (118 of 122 patients, 96.7%) (OR for clinical success with DPS, 0.73; 95% CI, 0.13-4.0; P = .71). Adverse events developed in 24 patients who received LAMS (24.7%) vs 27 patients who received DPS (17.8%) (OR for an adverse event in a patient receiving a DPS, 0.82; 95% CI, 0.33-2.0; P = .67). However, patients with LAMS had a higher risk of pseudoaneurysm bleeding than patients with DPS (OR, 10.0; 95% CI, 1.19-84.6; P = .009). CONCLUSIONS: In a retrospective study of patients undergoing drainage of pancreatic pseudocysts or walled-off necrosis, we found LAMS and DPS to have comparable rates of technical and clinical success and adverse events. Drainage of walled-off necrosis or pancreatic pseudocysts using DPS was associated with fewer bleeding events overall, including pseudoaneurysm bleeding, but bleeding risk with LAMS should be weighed against the trend of higher actionable perforation and infection rates with DPS.

Effect of injection rate on contrast-enhanced MR angiography image quality: Modulation transfer function analysis.

PURPOSE: Contrast-enhanced (CE)-MRA optimization involves interactions of sequence duration, bolus timing, contrast recirculation, and both R1 relaxivity and R2*-related reduction of signal. Prior data suggest superior image quality with slower gadolinium injection rates than typically used. METHODS: A computer-based model of CE-MRA was developed, with contrast injection, physiologic, and image acquisition parameters varied over a wide gamut. Gadolinium concentration was derived using Verhoeven's model with recirculation, R1 and R2* calculated at each time point, and modulation transfer curves used to determine injection rates, resulting in optimal resolution and image contrast for renal and carotid artery CE-MRA. Validation was via a vessel stenosis phantom and example patients who underwent carotid CE-MRA with low effective injection rates. RESULTS: Optimal resolution for renal and carotid CE-MRA is achieved with injection rates between 0.5 to 0.9 mL/s and 0.2 to 0.3 mL/s, respectively, dependent on contrast volume. Optimal image contrast requires slightly faster injection rates. Expected signal-to-noise ratio varies with both contrast volume and cardiac output. Simulated vessel phantom and clinical carotid CE-MRA exams at an effective contrast injection rate of 0.4 to 0.5 mL/s demonstrate increased resolution. CONCLUSION: Optimal image resolution is achieved at intuitively low, effective injection rates (0.2-0.9 mL/s, dependent on imaging parameters and contrast injection volume). Magn Reson Med 78:357-369, 2017. © 2016 International Society for Magnetic Resonance in Medicine.

Thyroid Cancer Polygenic Risk Score Improves Classification of Thyroid Nodules as Benign or Malignant.

CONTEXT: Thyroid nodule ultrasound-based risk stratification schemas rely on the presence of high-risk sonographic features. However, some malignant thyroid nodules have benign appearance on thyroid ultrasound. New methods for thyroid nodule risk assessment are needed. OBJECTIVE: We investigated polygenic risk score (PRS) accounting for inherited thyroid cancer risk combined with ultrasound-based analysis for improved thyroid nodule risk assessment. METHODS: The convolutional neural network classifier was trained on thyroid ultrasound still images and cine clips from 621 thyroid nodules. Phenome-wide association study (PheWAS) and PRS PheWAS were used to optimize PRS for distinguishing benign and malignant nodules. PRS was evaluated in 73 346 participants in the Colorado Center for Personalized Medicine Biobank. RESULTS: When the deep learning model output was combined with thyroid cancer PRS and genetic ancestry estimates, the area under the receiver operating characteristic curve (AUROC) of the benign vs malignant thyroid nodule classifier increased from 0.83 to 0.89 (DeLong, P value = .007). The combined deep learning and genetic classifier achieved a clinically relevant sensitivity of 0.95, 95% CI [0.88-0.99], specificity of 0.63 [0.55-0.70], and positive and negative predictive values of 0.47 [0.41-0.58] and 0.97 [0.92-0.99], respectively. AUROC improvement was consistent in European ancestry-stratified analysis (0.83 and 0.87 for deep learning and deep learning combined with PRS classifiers, respectively). Elevated PRS was associated with a greater risk of thyroid cancer structural disease recurrence (ordinal logistic regression, P value = .002). CONCLUSION: Augmenting ultrasound-based risk assessment with PRS improves diagnostic accuracy.

Splenic trauma: pictorial review of contrast-enhanced CT findings.

Splenic trauma is a common entity of high clinical significance that may have a variety of imaging manifestations. This article reviews splenic anatomy and function, mechanisms of splenic trauma, and the classification and grading of splenic trauma, with a particular focus on contrast-enhanced CT. Heterogeneous early enhancement of the spleen and technical suggestions to minimize this artifact are discussed. Mechanisms of trauma and their classification under the American Association for the Surgery of Trauma are discussed and illustrated. The merits of an alternative classification system by Marmery et al. are also considered. We conclude that knowledge of the mechanisms, classifications, and grades of splenic trauma can assist the clinician in clinical decision making and promote improved patient outcomes.

Creating the Black Box: A Primer on Convolutional Neural Network Use in Image Interpretation.

Convolutional neural networks have been shown to demonstrate high diagnostic performance in radiologic image interpretation tasks ranging from recognition of acute stroke on computed tomography to identification of tuberculosis on plain radiographs. To a radiologist not immersed in computer science jargon, it may seem that this inscrutable black box is best treated warily, at arm's length. In this work, we illustrate how a radiologist without a deep background in computer science may be able to set up a state-of-the-art convolutional neural network for image interpretation tasks through transfer learning. This technique is relatively simple to implement, has been shown to demonstrate equivalent performance to neural networks specifically trained on medical image data, and offers a chance for the interested-but-intimidated radiologist to deep her toe in the water without becoming overwhelmed.

Development of an Adrenocortical Cancer Humanized Mouse Model to Characterize Anti-PD1 Effects on Tumor Microenvironment.

CONTEXT: Although the development of immune checkpoint inhibitors has transformed treatment strategies of several human malignancies, research models to study immunotherapy in adrenocortical carcinoma (ACC) are lacking. OBJECTIVE: To explore the effect of anti-PD1 immunotherapy on the alteration of the immune milieu in ACC in a newly generated preclinical model and correlate with the response of the matched patient. DESIGN, SETTING, AND INTERVENTION: To characterize the CU-ACC2-M2B patient-derived xenograft in a humanized mouse model, evaluate the effect of a PD-1 inhibitor therapy, and compare it with the CU-ACC2 patient with metastatic disease. RESULTS: Characterization of the CU-ACC2-humanized cord blood-BALB/c-Rag2nullIl2rγnullSirpaNOD model confirmed ACC origin and match with the original human tumor. Treatment of the mice with pembrolizumab demonstrated significant tumor growth inhibition (60%) compared with controls, which correlated with increased tumor infiltrating lymphocyte activity, with an increase of human CD8+ T cells (P < 0.05), HLA-DR+ T cells (P < 0.05) as well as Granzyme B+ CD8+ T cells (<0.001). In parallel, treatment of the CU-ACC2 patient, who had progressive disease, demonstrated a partial response with 79% to 100% reduction in the size of target lesions, and no new sites of metastasis. Pretreatment analysis of the patient's metastatic liver lesion demonstrated abundant intratumoral CD8+ T cells by immunohistochemistry. CONCLUSIONS: Our study reports the first humanized ACC patient-derived xenograft mouse model, which may be useful to define mechanisms and biomarkers of response and resistance to immune-based therapies, to ultimately provide more personalized care for patients with ACC.

Cultivating Meaningful Work in Healthcare: A Paradigm and Practice.

The socioeconomic dynamics of the current healthcare system can challenge physicians to remain engaged with meaningful work, which is problematic given its role in mitigating burnout. In Man's Search For Meaning, Victor Frankl proposes that meaningful work is a prerequisite for a meaningful life. This article provides a road map for radiologists by applying key guiding principles outlined by Frankl as they intersect with current evidence in the field of physician burnout.

Strategies for Increasing Radiologist Efficiency.

Radiologists experience higher than average rates of burnout. Contributing factors include job tedium, a lack of a sense of autonomy, and decreasing compensation in the context of practices closely monitoring relative value unit output. Increasing radiologist efficiency may be a strategy for mitigating burnout, as freeing up time may allow radiologists to focus their attention on the aspects of their job that they enjoy and find most meaningful. In this work we propose several strategies of distraction minimization and time optimization that radiologists may employ to increase their efficiency.

Risk Threshold Algorithm for Thyroid Nodule Management Demonstrates Increased Specificity and Diagnostic Accuracy as Compared With American College of Radiology Thyroid Imaging, Reporting and Data System; Society of Radiologists in Ultrasound; and American Thyroid Association Management Guidelines.

We studied diagnostic performance of an algorithm guiding thyroid nodule management using a malignancy risk model as compared with extant management guidelines. Single-institution, retrospective study was performed with sequential cases from pathology registry from 2012 to 2015. Seventy-eight patients were enrolled, with benign and malignant groups defined by aspiration results. Risk Threshold Algorithm determined management based off of a logistic regression model and a risk threshold. American College of Radiology Thyroid Imaging, Reporting and Data System (ACR TI-RADS), Society of Radiologists in Ultrasound (SRU), and American Thyroid Association (ATA) guidelines were used in comparison. Sensitivity, specificity, positive/negative predictive values, receiver operating characteristic (ROC) values were derived, with significance assessed via McNemar and permutation tests. Forty-four benign nodules and 40 papillary thyroid carcinomas were included. Risk Threshold Algorithm area under the ROC curve was 0.80 versus 0.59 (ACR TI-RADS), 0.49 (SRU), and 0.44 (ATA); all areas under the ROC curve differences were statistically significant. Risk Threshold Algorithm demonstrates sensitivity, specificity, positive predictive value, and negative predictive values of 63%, 91%, 86%, and 73% at the risk threshold maximizing diagnostic performance, compared with 85%, 39%, 56%, and 74% (ACR TI-RADS); 85%, 18%, 50%, and 57% (SRU); and 89%, 11%, 50%, and 83% (ATA). Sensitivity and specificity were significantly different between all groups except SRU versus TI-RADS. The Risk Threshold Algorithm, based on a malignancy risk model, demonstrates increased overall diagnostic accuracy as compared with ACR TI-RADS, SRU, and ATA management guidelines. Through eliminating unnecessary biopsy, patient anxiety, and morbidity can be reduced.

Non-invasive diagnostic criteria of hepatocellular carcinoma: Comparison of diagnostic accuracy of updated LI-RADS with clinical practice guidelines of OPTN-UNOS, AASLD, NCCN, EASL-EORTC, and KLSCG-NCC.

PURPOSE: To retrospectively compare the diagnostic performance of different noninvasive diagnostic criteria of HCC including LI-RADS, OPTN-UNOS, AASLD, NCCN, EASL-EORTC, KLCSG-NCC. MATERIALS AND METHODS: We reviewed the medical records of 3,491 pathologically examined liver lesions from January-2011 to January-2015 in our institution. 195 lesions in 133 patients (M:F = 100:33) with chronic hepatitis B/C and/or cirrhosis for any etiology were finally included in our study, with 98 lesions ≥ 2 cm, 72 lesions between 1-2 cm, and 25 lesions < 1 cm. The main comparison was made with the largest nodules of each patient (n = 133). The lesions were retrospectively evaluated for the diagnosis of HCC on DCE-CT or MR using different noninvasive diagnostic criteria including LI-RADS, OPTN-UNOS, AASLD, NCCN, EASL-EORTC, and KLCSG-NCC. With pathological evaluation serving as a gold-standard, sensitivity, specificity, PPV and NPV as well as accuracy of the diagnostic criteria were calculated. RESULTS: There was no statistically significant differences in diagnostic accuracy among noninvasive diagnostic criteria. For 133 lesions of the largest lesion per patient, the overall accuracy was highest with LI-RADS criteria (89.3%) and the overall sensitivity was highest with LI-RADS, AASLD, NCCN criteria (all 89.5%). For 1-2 cm lesions, sensitivity decreased for all criteria in the following order: EASL-EORTC (59.1%), KLCSG-NCC (58.3%), LI-RADS, AASLD, NCCN (all 56.5%), and OPTN-UNOS (22.7%) criteria. OPTN-UNOS had the highest specificity in cirrhotic livers, 91.7%. CONCLUSIONS: The current noninvasive diagnostic criteria of HCC have no statistically significant difference in diagnostic accuracy. Overall, LI-RADS had the highest sensitivity and accuracy among the guidelines. OPTN had the highest specificity for cirrhotic livers.

Evaluation of diagnostic accuracy: multidetector CT image noise correction improves specificity of a Gaussian model-based algorithm used for characterization of incidental adrenal nodules.

OBJECTIVES: To investigate whether the histogram analysis method of characterizing adrenal nodules as adenomas is affected by increased noise with modern CT technique, and if an extension that allows for noise correction will improve diagnostic performance. MATERIALS AND METHODS: This is a HIPAA-compliant, IRB-approved retrospective study performed on 58 total patients. The first group of 29 patients had 33 adrenal lesions that were pathology-proven non-adenomas. The second group had 29 patients with 33 pathology-proven or presumed adenomas based on established imaging criteria. The nodules were evaluated using the histogram method, mean attenuation method, and a Gaussian model-based algorithm without (uncorrected Gaussian algorithm) and with correction (corrected Gaussian algorithm) for image noise. Sensitivity, specificity, and accuracy for identifying adenoma were derived. RESULTS: There were no significant differences in identifying adenoma from non-adenoma when using the histogram analysis method and the uncorrected Gaussian algorithm, both of which had low specificities of 42.4% and 47.0%, respectively (p = 0.30). Adding noise correction to the Gaussian algorithm resulted in a statistically significant increase in specificity relative to the histogram method (86.4% vs. 42.4%, p < 0.001). The corrected Gaussian algorithm improved sensitivity compared to the mean attenuation method (71.2% vs. 54.5%, p < 0.001), but had lower specificity (86.4% vs. 100%, p < 0.001), and similar overall accuracy (78.8% vs. 77.3%, p = 0.74). CONCLUSION: With modern low-dose CT technique, the specificity scores of the histogram method for discrimination of adrenal adenomas and non-adenomas are lower than with previous higher dose scans. The specificity and accuracy of a histogram-equivalent method can be increased mathematically through image noise correction, and the corrected Gaussian algorithm has improved sensitivity to the mean attenuation with similar accuracy albeit with lower specificity. Although this suggests limited utility for histogram analysis in adrenal nodule characterization, our study demonstrates the potential mathematical application for other noise-dependent CT characterization methods.

Response to Immunotherapy in Combination With Mitotane in Patients With Metastatic Adrenocortical Cancer.

Adrenocortical carcinoma (ACC) is a rare orphan disease with a dismal prognosis. Surgery remains the first-line treatment, but most patients eventually develop metastatic disease. Mitotane is often used with chemotherapy with modest success. Little information is available concerning the efficacy of immunotherapy in combination with mitotane. We conducted a retrospective review of our initial six patients with metastatic ACC, for whom mitotane alone or with chemotherapy failed, and who were subsequently treated with a combination of pembrolizumab and mitotane, between July 2016 and March 2019. Imaging was analyzed per Response Evaluation Criteria in Solid Tumours 1.1 criteria. Two patients had a partial response and four patients had stable disease (8 to 19 months). One patient had grade 3 hepatitis and pembrolizumab was discontinued after 8 months. She died with disease progression 16 months after initiating pembrolizumab. One patient developed brain metastasis after 19 months of treatment and was transitioned to hospice. One patient had focal pneumonitis after 18 months of treatment, and pembrolizumab was discontinued. Three remaining patients continue pembrolizumab plus mitotane at the time of this writing. The current standard of care for ACC is a combination of etoposide, doxorubicin, cisplatin, and mitotane with an overall survival of 14.8 months. All six patients lived for at least 16 months after starting pembrolizumab added to mitotane therapy. The therapy appeared to be effective in both microsatellite instability-high and microsatellite stable tumors, suggesting some synergistic effect with mitotane. Combined immunotherapy and mitotane should be considered in future clinical trials in patients with ACC.

Diagnostic Performance of SRU and ATA Thyroid Nodule Classification Algorithms as Tested With a 1 Million Virtual Thyroid Nodule Model.

PURPOSE: The Society of Radiologists in Ultrasound (SRU 2005) and American Thyroid Association (ATA 2009 and ATA 2015) have published algorithms regarding thyroid nodule management. Kwak et al. and other groups have described models that estimate thyroid nodules' malignancy risk. The aim of our study is to use Kwak's model to evaluate the tradeoffs of both sensitivity and specificity of SRU 2005, ATA 2009 and ATA 2015 management algorithms. MATERIALS AND METHODS: 1,000,000 thyroid nodules were modeled in MATLAB. Ultrasound characteristics were modeled after published data. Malignancy risk was estimated per Kwak's model and assigned as a binary variable. All nodules were then assessed using the published management algorithms. With the malignancy variable as condition positivity and algorithms' recommendation for FNA as test positivity, diagnostic performance was calculated. RESULTS: Modeled nodule characteristics mimic those of Kwak et al. 12.8% nodules were assigned as malignant (malignancy risk range of 2.0-98%). FNA was recommended for 41% of nodules by SRU 2005, 66% by ATA 2009, and 82% by ATA 2015. Sensitivity and specificity is significantly different (< 0.0001): 49% and 60% for SRU; 81% and 36% for ATA 2009; and 95% and 20% for ATA 2015. CONCLUSION: SRU 2005, ATA 2009 and ATA 2015 algorithms are used routinely in clinical practice to determine whether thyroid nodule biopsy is indicated. We demonstrate significant differences in these algorithms' diagnostic performance, which result in a compromise between sensitivity and specificity.

Launchpad for Onboarding New Faculty Into Academic Life.

We developed a faculty professional development seminar series in order to facilitate the integration of our numerous new faculty into academics. The changing nature of the healthcare system, increasing clinical and administrative responsibility, and lack of access to senior mentors can hinder junior faculty productivity and possibly increase attrition. Given that no ready-made resources existed to address these issues we established a Professional Development Committee, developed a curriculum that covers relevant topics including promotion, mentorship, conflict management and feedback, and effective presentation of scientific data, and instituted changes iteratively based upon feedback. We used surveys from successive years of this seminar series to assess effectiveness, and our data demonstrate that our Professional Development Seminar Series was valued by its participants and that individual lectures improved from year to year. While it is too early to determine whether our efforts will lead to long-term changes in promotion success or faculty retention, our initial data are promising.

Characterization of Adrenal Adenoma by Gaussian Model-Based Algorithm.

We confirmed that computed tomography (CT) attenuation values of pixels in an adrenal nodule approximate a Gaussian distribution. Building on this and the previously described histogram analysis method, we created an algorithm that uses mean and standard deviation to estimate the percentage of negative attenuation pixels in an adrenal nodule, thereby allowing differentiation of adenomas and nonadenomas. The institutional review board approved both components of this study in which we developed and then validated our criteria. In the first, we retrospectively assessed CT attenuation values of adrenal nodules for normality using a 2-sample Kolmogorov-Smirnov test. In the second, we evaluated a separate cohort of patients with adrenal nodules using both the conventional 10HU unit mean attenuation method and our Gaussian model-based algorithm. We compared the sensitivities of the 2 methods using McNemar's test. A total of 183 of 185 observations (98.9%) demonstrated a Gaussian distribution in adrenal nodule pixel attenuation values. The sensitivity and specificity of our Gaussian model-based algorithm for identifying adrenal adenoma were 86.1% and 83.3%, respectively. The sensitivity and specificity of the mean attenuation method were 53.2% and 94.4%, respectively. The sensitivities of the 2 methods were significantly different (P value < 0.001). In conclusion, the CT attenuation values within an adrenal nodule follow a Gaussian distribution. Our Gaussian model-based algorithm can characterize adrenal adenomas with higher sensitivity than the conventional mean attenuation method. The use of our algorithm, which does not require additional postprocessing, may increase workflow efficiency and reduce unnecessary workup of benign nodules.