Citations

Abstract WSD0922-Fu is an oral, CNS-penetrant, small molecule EGFR inhibitor which potently inhibits EGFR aberrations specific to non-small cell lung cancer (NSCLC) and glioblastoma (GBM). MC1914 (NCT04197934) is a first-in-human phase 1 trial in patients with EGFR aberrant recurrent GBM and NSCLC. After determining the maximum tolerated dose, adults with recurrent GBM were treated in two expansion cohorts: (a) non-surgical cohort (6/12 patients thus far): EGFRvIII mutation, first recurrence; food-effect study included (b) surgical cohort (6/6 patients), any EGFR mutation, any recurrence; window-of-opportunity study (WOO) included. Among these 12 patients, median age = 60 years; 50% female; 58% ECOG 0). All patients received prior radiation and temozolomide. After the WOO or food-effect study, patients were dosed at 120 mg BID (5 days on, 2 days off), and escalated to 160 mg BID (5 days on, 2 days off) based on tolerability. Common adverse events (>40%) are comparable to other EGFR inhibitors (acneiform rash, diarrhea, fatigue, xerosis, pruritis). Treatment-related grade 3+ adverse events were limited to lymphopenia (n = 1). WSD0922-Fu exposure in the fed state exceeded the fasted state (median plasma Cmax and AUC0-24h are 1.6x and 1.3x, respectively) with a delayed Tmax (6.0 vs 1.8 hours). For the WOO, time between last dose administered to tumor collection ranged between 4 - 13.5 hours. Median tumor WSD0922-Fu concentration was 0.76 µg/g; estimated median unbound tumor concentration was 24.9 ng/g and within expectations for pharmacodynamic effect. Median brain/plasma ratio = 0.59; median unbound brain/unbound plasma = 0.88. Enrollment to the non-surgical cohort is ongoing, so formal efficacy data remain immature. However, examples of clinical efficacy will be shown along with data on recurrence patterns. Phase 2 clinical trial development is in progress.

Abstract Background Experts' interpretations of medical images for lesion diagnosis may not always align with the underlying in vivo tissue pathology and, therefore, cannot be considered the definitive truth regarding malignancy or benignity. While current machine learning (ML) models in medical imaging can replicate expert interpretations, their results may also diverge from the actual ground truth. Purpose This study investigates various factors contributing to these discrepancies and proposes solutions. Methods The central idea of the proposed solution is to integrate prior knowledge into ML models to enhance the characterization of in vivo tissues. The incorporation of prior knowledge into decision‐making is task‐specific, tailored to the data acquired for that task. This central idea was tested on the diagnosis of lesions using low dose computed tomography (LdCT) for early cancer detection, particularly focusing on more challenging, ambiguous or indeterminate lesions (IDLs) as classified by experts. One key piece of prior knowledge involves CT x‐ray energy spectrum, where different energies interact with in vivo tissues within a lesion, producing variable but reproducible image contrasts that encapsulate biological information. Typically, CT imaging devices use only the high‐energy portion of this spectrum for data acquisition; however, this study considers the full spectrum for lesion diagnostics. Another critical aspect of prior knowledge includes the functional or dynamic properties of in vivo tissues, such as elasticity, which can indicate pathological conditions. Instead of relying solely on abstract image features as current ML models do, this study extracts these tissue pathological characteristics from the image contrast variations. Results The method was tested on LdCT images of four sets of IDLs, including pulmonary nodules and colorectal polyps, with pathological reports serving as the ground truth for malignancy or benignity. The method achieved an area under the receiver operating characteristic curve (AUC) of 0.98 ± 0.03, demonstrating a significant improvement over existing state‐of‐the‐art ML models, which typically have AUCs in the 0.70 range. Conclusion Leveraging prior knowledge in machine intelligence can enhance lesion diagnosis, resolve the ambiguity of IDLs interpreted by experts, and improve the effectiveness of LdCT screening for early‐stage cancers.

Abstract Objectives Vacuoles, E1 enzyme, X-linked, Autoinflammatory, Somatic (VEXAS) syndrome is an autoinflammatory disease caused by somatic mutation of UBA1 and affects multiple organ systems. Involvement of the kidneys is not well characterized. We aimed to investigate the incidence, risk factors and histopathologic features of acute kidney injury (AKI) in VEXAS syndrome. Methods Patients with genetically confirmed UBA1 mutation consistent with VEXAS were included. Charts were manually reviewed. Cox regression analysis was used to identify variables associated with time-to-first AKI event. For patients with a kidney biopsy, histopathologic findings were reviewed. Results Eighty-one patients were included, all white men, with a mean age of 66.3 ± 8.6 years. Median (IQR) follow up was 3.5 (2.1–5.2) years during which 20 (25%) developed AKI and 22% died. AKI relapsed in 90% of cases for a median of 6 times during the follow up period. Cumulative incidence estimates (95% CI) for AKI at 1, 3 and 5 years were 6.2% (0.80–11.3%), 16.7% (7.5–25.0%) and 27.9% (14.9–38.9%), respectively. Age and baseline C-reactive protein (CRP) were significantly associated with time-to-first AKI event. Six patients underwent a kidney biopsy. Findings included, plasma cell-rich interstitial nephritis (n = 3), neutrophilic-rich interstitial inflammation (n = 1), leukocytoclastic peritubular capillaritis (n = 1) and acute tubular injury (n = 1). AKI responded well to treatment with glucocorticoids but had relapse upon tapering. Conclusion AKI is an underrecognized feature of VEXAS occurring in 25% of patients in this cohort. Age at diagnosis and CRP were associated with time to first AKI event during follow up. Plasma cell-rich interstitial nephritis was the most common histopathologic finding.

AbstractFinding a feeding vessel with diameter greater than 0.9 mm during transarterial chemoembolization (TACE) for hepatocellular cancer (HCC) has been shown to predict tumor necrosis on subsequent pathology. However, whether this translates into a useful clinical predictor for post-chemoembolization response and survival is unknown. This study aimed to determine whether the presence of an arterial feeder on pre-TACE cross-sectional imaging is associated with treatment response and survival after TACE for unresectable HCC. Retrospective medical record search for all chemo-embolizations performed for HCC from 2015 to 2016 yielded 138 patients who underwent 275 TACE sessions spanning 2011 to 2017. Patients consisted of 98 males (71%) with mean age of 62 (range, 37–86). Each patient underwent an average of two TACE sessions (range, 1–11). Endpoints included target tumor response and overall response defined by the mRECIST criteria, as well as patient survival. Preprocedural MRI/CT was reviewed for the presence of arterial feeder. Multivariable logistic regressions and Cox proportional hazard regressions were used to assess the effects of arterial feeder presence on treatment response and survival, respectively, adjusting for other covariates. Overall response was seen in 69% of patients. Arterial feeder was present on preprocedural cross-sectional imaging for 28% of TACE sessions. Median survival was 26.5 months (interquartile range, 13.2–38.1). The presence of arterial feeder led to better target tumor response (OR = 11.9, p < 0.0001), overall response (OR = 9.3, p < 0.0001), and improved survival (HR = 0.55, p = 0.02). The presence of an arterial feeder on pre-TACE cross-sectional imaging is associated with target tumor response, overall response, and survival after TACE.