Citations

e16550 Background: Improved computational power and modern algorithms have generated significant interest in radiomics for cancer diagnosis and staging. Here we assess the performance of deep learning (DL) models as a means for feature extraction in combination with supervised machine learning (ML) algorithms for accurate staging and chemotherapy response assessment of bladder cancer. Methods: Deidentified grayscale CT images from bladder cancer patients scheduled to undergo radical cystectomy were included in this retrospective study. These images were manually annotated with two regional masks (normal region and cancer region). Five DL models- namely, AlexNet, GoogleNet, InceptionV3, ResNet-50, and XceptionNet pre-trained on the ImageNet dataset, a public dataset, were then fine-tuned on our bladder CT scan data to extract features. Through feature selection process, the subset of the features was used to build ML classifiers for classification. The classification was performed using five different ML classifiers, namely k-Nearest Neighbor (KNN), Naïve-Bayes (NB), Support Vector Machine (SVM), Linear Discriminant Analysis (LDA), and Decision Tree (DT). The classification task was performed with 10-fold cross-validation, and each of the experiments contained a different but not mutually exclusive subset of samples. The evaluation metrics include accuracy, sensitivity, specificity, precision, and F1-score. Results: A total of 200 deidentified grayscale CT images of 100 patients with histologically proven bladder cancer, were included in this study. For experiment (1) normal vs. cancer, the LDA classifier on XceptionNet based features provides the best performance with an accuracy of 86.07%, sensitivity of 96.75%, specificity of 69.65%, precision of 83.07%, and F1-score of 89.39%. For experiment (2) non-muscle invasive Cancer (NMIBC) vs. muscle invasive bladder cancer (MIBC), the LDA classifier on XceptionNet based features provided the best performance with an accuracy of 79.72%, sensitivity of 66.62%, specificity of 87.39%, precision of 75.58%, and F1-score of 70.81%. For experiment (3) T0 lesion vs. MIBC, the LDA classifier on XceptionNet based features provides the best performance with an accuracy of 74.96%, sensitivity of 80.51%, specificity of 70.22%, precision of 69.78%, and F1-score of 74.73%. Conclusions: Our proposed model has shown good results in differentiating normal vs cancer and promising performance in differentiating T0 vs MIBC after chemotherapy treatment. We are expanding our dataset to further improve the performance in differentiating T0 vs MIBC. In addition, we will investigate the applicability of GAN for data augmentation to address data limit. We believe the hybrid DL and ML framework may facilitates radiologists' decisions and clinical decision-making in patients with bladder cancer.

Introduction: Bortezomib has been reported to favourably impact the outcomes of t(4;14) and del(17p) in multiple myeloma (MM), but its impact on gain 1q (+1q) is unknown. Methods: To address this, 250 patients treated with bortezomib-based induction were analysed. All myeloma samples had fluorescence in situ hybridization (FISH) performed on CD138-sorted bone marrow aspirate, and plasma cells were analysed using DNA probes specific for the following chromosomal aberrations: del(13q14), del(17p), t(14;16), t(4;14), and +1q. Presence of +1q was defined as the presence of at least three copies of 1q21 at the cut off level of 20% of bone marrow plasma cells. Results: +1q identified in 167 (66.8%) and associated with t(4;14) and high lactate dehydrogenase (LDH). +1q was not associated with response rate but shorter event-free survival (EFS) (median EFS 35 vs 55 months, p = 0.05) and overall survival (OS) (median OS 74 vs 168 months, p = 0.00025). Copy number and clone size did not impact survival. Multivariate analysis showed +1q was an independent adverse factor for OS together with International Staging System (ISS)3, high LDH, del(17p) and t(4;14). When a risk score of 1 was assigned to each independent adverse factor, OS was shortened incrementally by a risk score from 0 to 4. Post-relapse/progression survival was inferior in those with +1q (median 60 vs 118 months, p = 0.000316). Autologous stem cell transplantation (ASCT) improved OS for those with +1q (median OS 96 vs 49 months, p = 0.000069). Conclusion: +1q is an adverse factor for OS in MM uniformly treated with bortezomib-based induction but was partially mitigated by ASCT. A risk scoring system comprising +1q, LDH, high-risk FISH, and ISS is a potential tool for risk stratification in MM.

AbstractObjectiveSummarize the current evidence on efficacy and tolerability of vagus nerve stimulation (VNS), responsive neurostimulation (RNS), and deep brain stimulation (DBS) through a systematic review and meta‐analysis.MethodsWe followed the Preferred Reporting Items of Systematic reviews and Meta‐Analyses reporting standards and searched Ovid Medline, Ovid Embase, and the Cochrane Central Register of Controlled Trials. We included published randomized controlled trials (RCTs) and their corresponding open‐label extension studies, as well as prospective case series, with ≥20 participants (excluding studies limited to children). Our primary outcome was the mean (or median, when unavailable) percentage decrease in frequency, as compared to baseline, of all epileptic seizures at last follow‐up. Secondary outcomes included the proportion of treatment responders and proportion with seizure freedom.ResultsWe identified 30 eligible studies, six of which were RCTs. At long‐term follow‐up (mean 1.3 years), five observational studies for VNS reported a pooled mean percentage decrease in seizure frequency of 34.7% (95% confidence interval [CI]: −5.1, 74.5). In the open‐label extension studies for RNS, the median seizure reduction was 53%, 66%, and 75% at 2, 5, and 9 years of follow‐up, respectively. For DBS, the median reduction was 56%, 65%, and 75% at 2, 5, and 7 years, respectively. The proportion of individuals with seizure freedom at last follow‐up increased significantly over time for DBS and RNS, whereas a positive trend was observed for VNS. Quality of life was improved in all modalities. The most common complications included hoarseness, and cough and throat pain for VNS and implant site pain, headache, and dysesthesia for DBS and RNS.SignificanceNeurostimulation modalities are an effective treatment option for drug‐resistant epilepsy, with improving outcomes over time and few major complications. Seizure‐reduction rates among the three therapies were similar during the initial blinded phase. Recent long‐term follow‐up studies are encouraging for RNS and DBS but are lacking for VNS.

TPS486 Background: Hepatocellular carcinoma (HCC) is a major cause of cancer-related death in the U.S. HCC surveillance is associated with improved early detection and receipt of curative treatment. However, current surveillance strategies suffer from low adherence and suboptimal sensitivity. The multi-target HCC blood test (mt-HBT) is a novel early detection assay that provides a qualitative result based on a multiparametric algorithm that incorporates information from methylation markers ( HOXA1, TSPYL5, and B3GALT6), AFP, and patient sex. Here, we describe the protocol for a prospective, multicenter clinical trial designed to investigate the performance of the mt-HBT in patients at risk for HCC. Methods: ALTernative to U/S (ALTUS) is a longitudinal, multicenter, prospective study with a targeted enrollment of 3,000 participants. The main inclusion and exclusion criteria are described in the table. Participants will receive HCC surveillance imaging (either U/S or CT/MRI) and blood collection for the mt-HBT. Those with positive U/S are anticipated to receive standard-of-care imaging follow-up with a multiphase CT or MRI, and those with a negative U/S will receive a study CT/MRI within 30 days of blood collection. Imaging interpretation is based on the CT/MRI LI-RADS (LR) v2018 categories. The primary endpoints are the overall sensitivity and specificity of the mt-HBT to detect HCC, defined as pathology-confirmed HCC or hepatocellular cholangiocarcinoma (cHCC-CCA), LR-4/5, or LR-TIV/M plus definite/probable HCC or cHCC-CCA. The secondary endpoints are mt-HBT early-stage sensitivity, and performance in a) participants with LR-5 or biopsy-confirmed HCC and b) participants confirmed non-HCC. Participants not diagnosed with a liver malignancy at visit one will continue to receive HCC surveillance, returning for a second visit ̃6 months from first blood collection. Blood will be collected after the second visit for the mt-HBT. Data on treatment will be collected for those diagnosed with HCC. Study participants and providers will be blinded to the mt-HBT results. Participants are currently enrolling across the U.S., with 60 targeted sites and an anticipated 12-month enrollment window. Clinical trial information: NCT05064553. [Table: see text]

A case of idiopathic osteoporosis in a 12-year-old girl has been reported. Quantitative microradiographic studies of a biopsy specimen of iliac crest showed normal bone formation and an almost threefold increase in the level of bone resorption. The patient was observed without treatment, and a spontaneous clinical remission with roentgenographic improvement occurred. A second iliac crest biopsy specimen, obtained 15 months after the initial biopsy at a time when roentgenographic improvement was apparent, showed a return toward normal of the level of bone resorption. Metabolic balance studies conducted at this time showed that the patient was approximately in calcium balance. Because the majority of the previously reported cases have also exhibited spontaneous improvement, it is suggested that initially observation alone is indicated.

Observations on two mongoloid children, one having primary hypothyroidism and the other having primary hyperthyroidism, have been reported. Primary hypothyroidism in the mongoloid child, in each of the three cases reported to date, has been associated with uterine bleeding attributed to hormonal overlap in the pituitary feedback mechanism. The significance of this phenomenon is not clean. A total of sixteen cases of mongolism with hyperthyroidism have been described. There is no evidence that altered thyroid function plays a role in the clinical manifestations of mongolism, and significant alteration in thyroid function is uncommon among such patients. Both of the carrier protein systems responded to the stresses of the superimposed disease processes in these mongoloid children as they do in normal children. Specifically, the values for the erythrocytic uptake of 1-I131-triiodothyronine and thyroxine and the triiodothynonine binding capacities (Case 2 only) were all in the expected ranges for both thyroid derangements involved. Similarly, in Case 2, where an iron-deficiency anemia was also present, the increase in the iron binding capacity and presumably in the level of transferrin was on the order of that to be expected for the degree of iron deficiency. The response to stress of both carrier protein systems in Down's Syndrome suggests that protein synthesis of transferrin and TBG is normal.

In 12 children with hyperadrenocorticism caused by tumor of the adrenal cortex the most common clinical picture was that of a mixture of adrenogenital and Cushing's syndromes. All tumors were surgically removed; adrenal steroids were used in the preparation of 11 patients for operation. One patient died with extensive metastasis 5 months after operation, while 11 were alive and well 18 months to 29 years after operation. Data from the 12 cases in this series and from those in the literature suggest that the likelihood of metastasis may be correlated with the relation of tumor weight to body weight.

Normal values for the measurement of thyroidal function using the erythrocytic uptake of I131-labeled triiodothyronine and the thyroxine-binding capacity of the inter-alpha globulin were established. Paired maternal and cord blood samples collected at the time of delivery were studied with these methods. The erythrocytic uptake of labeled hormone was increased in cord blood as compared to maternal blood. Cord blood apparently binds exogenous triiodothyronine in a different manner than it does exogenous thyroxine. Whether this is a qualitative or quantitative difference was not shown in this study.

Observations on two girls in whom an unusual type of chronic renal insufficiency developed many months after the onset of nephrotic syndrome are reported. Each patient became free of edema in spite of persistent massive proteinuria. Growth was retarded and rickets and attacks of tetany developed. The chemical disturbances of the blood were characterized by hypocalcemia, hypokalemia, azotemia and metabolic acidosis. Hyposthenuria, proteinuria, amino-aciduria, and minimal erythrocyturia, cylindruria and glycosuria were present. Healing of the rickets and cessation of attacks of tetany followed the administration of vitamin D and calcium salts. Prednisone was administered to one patient and thereafter proteinuria decreased and renal tubular function improved. Both girls are relatively asymptomatic 11 and 9 years after the onset of nephrotic syndrome, although they are rather small and still have evidence of renal disease. It is possible that cells of the renal tubules have been damaged as a result of prolonged massive proteinuria.