Citations

Adaptive radiotherapy (ART) workflows are increasingly adopted to achieve dose escalation and tissue sparing under dynamic anatomical conditions. However, recontouring and time constraints hinder the implementation of real-time ART workflows. Various auto-segmentation methods, including deformable image registration, atlas-based segmentation, and deep learning-based segmentation (DLS), have been developed to address these challenges. Despite the potential of DLS methods, clinical implementation remains difficult due to the need for large, high-quality datasets to ensure model generalizability. This study introduces an InterVision framework for segmentation. The InterVision framework can interpolate or create intermediate visuals between existing images to generate specific patient characteristics. The InterVision model is trained in two steps: (1) generating a general model using the dataset, and (2) tuning the general model using the dataset generated from the InterVision framework. The InterVision framework generates intermediate images between existing patient image slides using deformable vectors, effectively capturing unique patient characteristics. By creating a more comprehensive dataset that reflects these individual characteristics, the InterVision model demonstrates the ability to produce more accurate contours compared to general models. Models are evaluated using the volumetric dice similarity coefficient (VDSC) and the Hausdorff distance 95% (HD95%) for 18 structures in 20 test patients. As a result, the Dice score was 0.81 ± 0.05 for the general model, 0.82 ± 0.04 for the general fine-tuning model, and 0.85 ± 0.03 for the InterVision model. The Hausdorff distance was 3.06 ± 1.13 for the general model, 2.81 ± 0.77 for the general fine-tuning model, and 2.52 ± 0.50 for the InterVision model. The InterVision model showed the best performance compared to the general model. The InterVision framework presents a versatile approach adaptable to various tasks where prior information is accessible, such as in ART settings. This capability is particularly valuable for accurately predicting complex organs and targets that pose challenges for traditional deep learning algorithms.

Abstract Background In the TOPAZ-1, patients with biliary tract cancers (BTC) and recurrence within 6 months after surgery were excluded, even if this event is frequently observed in clinical practice. Our study aimed to assess if the efficacy of cisplatin-gemcitabine-durvalumab (CGD) in this population is comparable to that reported in the phase 3 trial. Methods The study cohort included patients with BTC who underwent surgery on the primary tumor, experienced disease recurrence occurring ≤6 months or >6 months after surgery or after the end of adjuvant therapy and started CGD. The primary objectives were overall survival (OS) and progression free survival (PFS). Results A total of 178 patients were enrolled. No significant differences were observed between early and late relapse groups in OS (23.4 months vs not reached; HR 1.26; 95% CI, 0.67-2.37; P = .45) and PFS [7.0 months vs 9.8 months; HR 1.3(95% CI, 0.9-2.1) P = .13]. Overall response rate and disease control rate (P = .33 and P = .62) were comparable between the 2 groups, as the overall safety profile. In addition, we compared survival outcomes between the selected population and a historical cohort of patients with BTC treated with cisplatin-gemcitabine (CG) and found that despite the absence of statistical significance, CGD showed an outcome trend compared with CG regardless of the time of recurrence after surgery or adjuvant chemotherapy [(CG ≤ 6 vs CGD ≤ 6 months: HR 0.59, 95%CI, 0.35-1.01, P = .05; HR 0.70; 95%CI, 0.46-1.06, P = .09, OS and PFS, respectively) and (CG > 6 vs. CGD > 6 months: HR 0.50; 95%CI, 0.29-0.88, P = 0.0165; HR 0.54; 95%CI, 0.35-0.84, P = .0068, OS and PFS, respectively)]. Conclusion Our analysis suggests that CGD retains its efficacy independently of the timing of relapse after surgery or completion of adjuvant treatment in patients with advanced BTC.

Abstract This post hoc analysis of the randomized, placebo‐controlled N‐MOmentum study (NCT02200770) of inebilizumab in neuromyelitis optica spectrum disorder (NMOSD) evaluated relationships between circulating B‐cell subsets and aquaporin‐4 immunoglobulin G (AQP4‐lgG) titers and attacks. Among participants receiving placebo, CD20 + and CD27 + B‐cell counts were modestly increased from the pre‐attack visit to attack; plasmablast/plasma cell gene signature was increased from baseline to the pre‐attack visit ( p = 0.016) and from baseline to attack ( p = 0.009). With inebilizumab treatment, B‐cell subset counts decreased and did not increase with attacks. No difference in change of AQP4‐IgG titers from baseline to time of attack was observed.

Abstract Background Sphingolipids play a crucial role in cellular functions and are essential components of cell membranes, signaling molecules, and lipid metabolism. In particular, ceramide is a key intermediate in sphingolipid metabolism and defects in ceramide metabolism can lead to various inborn errors of metabolism, making ceramides important targets for clinical screening and diagnosis. Detecting altered concentration patterns of sphingolipids is desirable for distinguishing related inborn errors of metabolism for diagnosis and treatment monitoring. Methods We developed a liquid chromatography-tandem mass spectrometry method with a pathway-oriented approach to focus on sphingolipids involved in ceramide metabolism. A total of 47 sphingolipids bearing different head groups and side chains were targeted. Precision/reproducibility, linearity, and spike recovery extraction efficiency tests were performed on plasma and serum samples from confirmed cases of sphingolipidosis. Results Linearity of the method showed the coefficient of determination (r2) for all standards to be >0.99 with a slope of 1.00 ± 0.01. Intra- and interday reproducibility of standards spiked into plasma and serum revealed a coefficient of variation <20%. Spike and recovery assessment showed recovery values of 80%–120% for all standards. Altered levels of sphingolipids from patients with hereditary sensory and autonomic neuropathy caused by pathogenic variants in SPTLC2 and hypomyelinating leukodystrophy related to variants in DEGS1 were detected, in agreement with trends reported in earlier studies confirming the utility of this pathway-centric method. Conclusions This method can serve as a useful tool to simultaneously monitor sphingolipids, enabling screening and diagnosis of inborn errors of ceramide metabolism.

AbstractBackgroundAdaptive radiotherapy (ART) workflows have been increasingly adopted to achieve dose escalation and tissue sparing under shifting anatomic conditions, but the necessity of recontouring and the associated time burden hinders a real‐time or online ART workflow. In response to this challenge, approaches to auto‐segmentation involving deformable image registration, atlas‐based segmentation, and deep learning‐based segmentation (DLS) have been developed. Despite the particular promise shown by DLS methods, implementing these approaches in a clinical setting remains a challenge, namely due to the difficulty of curating a data set of sufficient size and quality so as to achieve generalizability in a trained model.PurposeTo address this challenge, we have developed an intentional deep overfit learning (IDOL) framework tailored to the auto‐segmentation task. However, certain limitations were identified, particularly the insufficiency of the personalized dataset to effectively overfit the model. In this study, we introduce a personalized hyperspace learning (PHL)‐IDOL segmentation framework capable of generating datasets that induce the model to overfit specific patient characteristics for medical image segmentation.MethodsThe PHL‐IDOL model is trained in two stages. In the first, a conventional, general model is trained with a diverse set of patient data (n = 100 patients) consisting of CT images and clinical contours. Following this, the general model is tuned with a data set consisting of two components: (a) selection of a subset of the patient data (m < n) using the similarity metrics (mean square error (MSE), peak signal‐to‐noise ratio (PSNR), structural similarity index (SSIM), and the universal quality image index (UQI) values); (b) adjust the CT and the clinical contours using a deformed vector generated from the reference patient and the selected patients using (a). After training, the general model, the continual model, the conventional IDOL model, and the proposed PHL‐IDOL model were evaluated using the volumetric dice similarity coefficient (VDSC) and the Hausdorff distance 95% (HD95%) computed for 18 structures in 20 test patients.ResultsImplementing the PHL‐IDOL framework resulted in improved segmentation performance for each patient. The Dice scores increased from 0.810.05 with the general model, 0.83 for the continual model, 0.83 for the conventional IDOL model to an average of 0.87 with the PHL‐IDOL model. Similarly, the Hausdorff distance decreased from 3.06 with the general model, 2.84 for the continual model, 2.79 for the conventional IDOL model and 2.36 for the PHL‐IDOL model. All the standard deviations were decreased by nearly half of the values comparing the general model and the PHL‐IDOL model.ConclusionThe PHL‐IDOL framework applied to the auto‐segmentation task achieves improved performance compared to the general DLS approach, demonstrating the promise of leveraging patient‐specific prior information in a task central to online ART workflows.

Background and Aims: We compared the safety and efficacy of bintrafusp alfa (BA) in combination with gemcitabine+cisplatin (GemCis), to those of GemCis alone, in patients with biliary tract cancer. Approach and Results: This randomized, double-blind, placebo-controlled, adaptive design phase 2/3 trial (NCT04066491) included adults who are treatment-naive with locally advanced/metastatic biliary tract cancer. Patients (N = 297) were randomized to receive an IV infusion of BA (2400 mg once/3 wk) plus GemCis (gemcitabine 1000 mg/m 2 +cisplatin 25 mg/m 2 on days 1 and 8/3 wk; 8 cycles) (BA group, n = 148) or placebo+GemCis (placebo group, n = 149). The primary end point was overall survival (OS). For adaptation analysis (phase 2-phase 3; data cutoff: May 20, 2021), efficacy was assessed in the first 150 patients who were antibiotic-naive when 80 progression-free survival events had occurred and ≥ 19 weeks of follow-up had been completed (BA, n = 73; placebo, n = 77). Median OS (95% CI) for the BA (11.5 mo [9.3–not estimable]) and placebo (11.5 mo [10.0–not estimable]) groups was comparable (hazard ration 1.23 [95% CI 0.66–2.28]; p = 0.7394); OS data maturity was 27.2% (41 events/151 patients). The most common grade ≥3 treatment-related adverse event was anemia (BA, 26.0%; placebo, 22.8%). Bleeding adverse events were reported more frequently in the BA group (28.8%) versus the placebo group (7.4%). Deaths within 60 days of the first dose were reported in 7.5% and 1.3% of patients in the BA and placebo groups, respectively. Conclusions: BA+GemCis did not provide a clinically meaningful benefit compared with GemCis alone as first-line treatment for biliary tract cancer, and the study was discontinued early (terminated: August 20, 2021).