Greenlee, Jeremy D W

Abstract Glioblastoma (GBM), isocitrate dehydrogenase wild-type (IDHwt) and central nervous system diffuse large B-cell lymphoma (CNS-DLBCL) are aggressive brain tumors with overlapping MRI features, yet distinct treatment approaches. Noninvasive tools are needed to aid in differential diagnosis. Deep learning on T1 postcontrast and T2-weighted MRI sequences were used to differentiate GBM and CNS-DLBCL. A three-stage temporal study design was utilized. Model development was performed on 146 patients with CNS-DLBCL and 146 age-matched, sex-matched, and MRI year–matched patients with GBM diagnosed at Mayo Clinic between 1998 and 2019. Models were tested on independent temporal test cohorts. Initial testing included 240 independent GBM diagnosed at Mayo Clinic between 1998 and 2019. The prospective test cohort included 37 patients with CNS-DLBCL and 256 patients with GBM diagnosed at Mayo Clinic after January 1, 2020, and 36 patients with CNS-DLBCL diagnosed at an external institution. Of the patients diagnosed at Mayo Clinic, 47% had MRIs generated from non-Mayo institutions. Two different model approaches were compared: (i) ensemble approach using area under the receiver operating characteristic curve (AUC) and cross-validation for model selection and (ii) loss approach minimizing cross-entropy loss and cross-validation to evaluate prediction performance. The AUCs on the prospective test cohort were 0.84 [95% confidence interval (CI), 0.78–0.90] and 0.83 (95% CI, 0.77–0.88) for the ensemble and loss approaches, respectively. Stability of ensemble prediction improved with the increasing number of models. Stratified AUC analysis demonstrated consistent performance across sex and age. We utilized a robust temporal study design and applied 2 different analytic approaches to develop a classification model. The findings confirm the feasibility of using MRI-based deep learning models to differentiate GBM from CNS-DLBCL. Significance: GBM, IDHwt and CNS-DLBCL are aggressive brain tumors with overlapping MRI features, yet distinct treatment approaches. Noninvasive tools are needed to aid in differential diagnosis. We developed MRI-based deep learning models to differentiate GBM, IDHwt from CNS-DLBCL using a rigorous three-stage temporal design that included prospective validation. The model AUC on a prospective cohort was 0.84.

Abstract Purpose: Pharmacologic ascorbate (P-AscH−) is hypothesized to be an iron (Fe)-dependent tumor-specific adjuvant to chemoradiation in treating glioblastoma (GBM). This study determined the efficacy of combining P-AscH− with radiation and temozolomide in a phase II clinical trial while simultaneously investigating a mechanism-based, noninvasive biomarker in T2* mapping to predict GBM response to P-AscH− in humans. Patients and Methods: The single-arm phase II clinical trial (NCT02344355) enrolled 55 subjects, with analysis performed 12 months following the completion of treatment. Overall survival (OS) and progression-free survival (PFS) were estimated with the Kaplan–Meier method and compared across patient subgroups with log-rank tests. Forty-nine of 55 subjects were evaluated using T2*-based MRI to assess its utility as an Fe-dependent biomarker. Results: Median OS was estimated to be 19.6 months [90% confidence interval (CI), 15.7–26.5 months], a statistically significant increase compared with historic control patients (14.6 months). Subjects with initial T2* relaxation < 50 ms were associated with a significant increase in PFS compared with T2*-high subjects (11.2 months vs. 5.7 months, P < 0.05) and a trend toward increased OS (26.5 months vs. 17.5 months). These results were validated in preclinical in vitro and in vivo model systems. Conclusions: P-AscH− combined with temozolomide and radiotherapy has the potential to significantly enhance GBM survival. T2*-based MRI assessment of tumor iron content is a prognostic biomarker for GBM clinical outcomes. See related commentary by Nabavizadeh and Bagley, p. 255

Abstract Dural membrane is an important anatomic structure that surrounds and protects the entire central nervous system. Physical properties of the dura have many pathophysiological and therapeutic implications in cranial surgery, especially skull base disorders. The aim of this study is to investigate variation in skull base dural thickness and correlation with different demographic parameters. At the time of autopsy, the petrous apex dura with the underlying bone of 20 cadavers was harvested. Dural thickness was independently measured by two pathologists at the thinnest and thickest segments in the specimen. Correlational analyses were then performed to compare dural thickness with gender, age, neck circumference, height, weight, and body mass index (BMI). Mean, minimum, and maximum skull base dural thickness in our study was 0.36, 0.27, and 0.46 mm, respectively. Age demonstrated a negative correlation with dural thickness with significantly thinner dura in the older subjects, p = 0.01. There was a trend toward thinner dura in females that approached statistical significance, p = 0.06. No strong correlation could be found with body weight, height, neck circumference, or BMI. Our findings show a considerable intersubject and intrasubject variability in skull base dural thickness. Some demographic parameters also seem to impact dural thickness. Additional histological studies are needed for better understanding of the pathophysiological mechanisms pertaining to the tensile properties of the dural membrane.