Tumor- and circulating-free DNA methylation identifies clinically relevant small cell lung cancer subtypes

Small cell lung cancer (SCLC) is an aggressive malignancy composed of distinct transcriptional subtypes, but implementing subtyping in the clinic has remained challenging, particularly due to limited tissue availability. Given the known epigenetic regulation of critical SCLC transcriptional programs, we hypothesized that subtype-specific patterns of DNA methylation could be detected in tumor or blood from SCLC patients. Using genomic-wide reduced-representation bisulfite sequencing (RRBS) in two cohorts totaling 179 SCLC patients and using machine learning approaches, we report a highly accurate DNA methylation-based classifier (SCLC-DMC) that can distinguish SCLC subtypes. We further adjust the classifier for circulating-free DNA (cfDNA) to subtype SCLC from plasma. Using the cfDNA classifier (cfDMC), we demonstrate that SCLC phenotypes can evolve during disease progression, highlighting the need for longitudinal tracking of SCLC during clinical treatment. These data establish that tumor and cfDNA methylation can be used to identify SCLC subtypes and might guide precision SCLC therapy.

Brief Report: Comprehensive Clinicogenomic Profiling of Small Cell Transformation From EGFR-Mutant NSCLC Informs Potential Therapeutic Targets

Introduction

NSCLC transformation to SCLC has been best characterized with EGFR-mutant NSCLC, with emerging case reports seen in ALK, RET, and KRAS-altered NSCLC. Previous reports revealed transformed SCLC from EGFR-mutant NSCLC portends very poor prognosis and lack effective treatment. Genomic analyses revealed TP53 and RB1 loss of function increase the risk of SCLC transformation. Little has been reported on the detailed clinicogenomic characteristics and potential therapeutic targets for this patient population.

Methods

In this study, we conducted a single-center retrospective analysis of clinical and genomic characteristics of patients with EGFR-mutant NSCLC transformed to SCLC. Demographic data, treatment course, and clinical molecular testing reports were extracted from electronic medical records. Kaplan-Meier analyses were used to estimate survival outcomes. Next generation sequencing-based assays was used to identify EGFR and co-occurring genetic alterations in tissue or plasma before and after SCLC transformation. Single-cell RNA sequencing (scRNA-seq) was performed on a patient-derived-xenograft model generated from a patient with EGFR-NSCLC transformed SCLC tumor.

Results

A total of 34 patients were identified in our study. Median age at initial diagnosis was 58, and median time to SCLC transformation was 24.2 months. 68% were female and 82% were never smokers. 79% of patients were diagnosed as stage IV disease, and over half had brain metastases at baseline. Median overall survival of the entire cohort was 38.3 months from initial diagnoses and 12.4 months from time of SCLC transformation. Most patients harbored EGFR exon19 deletions as opposed to exon21 L858R alteration. Continuing EGFR tyrosine kinase inhibitor post-transformation did not improve overall survival compared with those patients where tyrosine kinase inhibitor was stopped in our cohort. In the 20 paired pretransformed and post-transformed patient samples, statistically significant enrichment was seen with PIK3CA alterations (p = 0.04) post-transformation. Profiling of longitudinal liquid biopsy samples suggest emergence of SCLC genetic alterations before biopsy-proven SCLC, as shown by increasing variant allele frequency of TP53, RB1, PIK3CA alterations. ScRNA-seq revealed potential therapeutic targets including DLL3, CD276 (B7-H3) and PTK7 were widely expressed in transformed SCLC.

Conclusions

SCLC transformation is a potential treatment resistance mechanism in driver-mutant NSCLC. In our cohort of 34 EGFR-mutant NSCLC, poor prognosis was observed after SCLC transformation. Clinicogenomic analyses of paired and longitudinal samples identified genomic alterations emerging post-transformation and scRNA-seq reveal potential therapeutic targets in this population. Further studies are needed to rigorously validate biomarkers and therapeutic targets for this patient population.

Descending mechanism by which medial prefrontal cortex endocannabinoid signaling controls the development of neuropathic pain and neuronal activity of dorsal root ganglion

ABSTRACT

Although regulation of nociceptive processes in the dorsal horn by deep brain structures has long been established, the role of cortical networks in pain regulation is minimally explored. The medial prefrontal cortex (mPFC) is a key brain area in pain processing that receives ascending nociceptive input and exerts top-down control of pain sensation. We have shown critical changes in mPFC synaptic function during neuropathic pain, controlled by endocannabinoid (eCB) signaling. This study tests whether mPFC eCB signaling modulates neuropathic pain through descending control. Intra-mPFC injection of cannabinoid receptor type 1 (CB1R) agonist WIN-55,212-2 (WIN) in the chronic phase transiently alleviates the pain-like behaviors in spared nerve injury (SNI) rats. By contrast, intra-mPFC injection of CB1R antagonist AM4113 in the early phase of neuropathic pain reduces the development of pain-like behaviors in the chronic phase. Spared nerve injury reduced the mechanical threshold to induce action potential firing of dorsal horn wide-dynamic-range neurons, but this was reversed in rats by WIN in the chronic phase of SNI and by mPFC injection of AM4113 in the early phase of SNI. Elevated dorsal root ganglion neuronal activity after injury was also diminished in rats by mPFC injection of AM4113, potentially by reducing antidromic activity and subsequent neuronal inflammation. These findings suggest that depending on the phase of the pain condition, both blocking and activating CB1 receptors in the mPFC can regulate descending control of pain and affect both dorsal horn neurons and peripheral sensory neurons, contributing to changes in pain sensitivity.

Beneficial effects of pulmonary embolism response team establishment in patients with pulmonary embolism in a developing country: a single-center experience

Objective

The impact of establishing a pulmonary embolism response team (PERT) in patients with pulmonary embolism (PE) has been proven in many developed countries. However, the efficacy of a PERT largely depends on expertise and infrastructure. This study explored the benefit of establishing a PERT in developing countries with limited healthcare resources by comparing the outcomes of patients with acute PE before and after PERT establishment at University Medical Center Ho Chi Minh City in Vietnam.

Methods

We conducted a single-center observational study from January 1, 2019, to August 1, 2021. All patients with PE confirmed on computed tomography were included. Patients admitted before PERT establishment were treated by cardiologists alone, while those hospitalized after PERT establishment were managed by the PERT.

Results

A total of 130 patients were included (pre-PERT estab-lishment: 51 patients; post-PERT establishment: 79 patients). The demographic characteristics, severity of PE, and clinical and laboratory findings were similar between the two groups. The post-PERT establishment group had a lower incidence rate of major and clinically relevant nonmajor bleeding (11.3% vs. 31.4%, p = 0.005) and required more interventional therapies (16.5% vs. 3.9%, p = 0.046) than did the pre-PERT establishment group. The in-hospital mortality rate decreased in the post-PERT establishment group compared with that in the pre-PERT establishment group (8.9% vs. 21.6%, p = 0.041).

Conclusions

Involvement of the PERT in PE management was associated with improved outcomes of patients with PE, including reduced bleeding and mortality rates in a resource-constrained hospital.

Induced neural phase precession through exogeneous electric fields

The gradual shifting of preferred neural spiking relative to local field potentials (LFPs), known as phase precession, plays a prominent role in neural coding. Correlations between the phase precession and behavior have been observed throughout various brain regions. As such, phase precession is suggested to be a global neural mechanism that promotes local neuroplasticity. However, causal evidence and neuroplastic mechanisms of phase precession are lacking so far. Here we show a causal link between LFP dynamics and phase precession. In three experiments, we modulated LFPs in humans, a non-human primate, and computational models using alternating current stimulation. We show that continuous stimulation of motor cortex oscillations in humans lead to a gradual phase shift of maximal corticospinal excitability by ~90°. Further, exogenous alternating current stimulation induced phase precession in a subset of entrained neurons (~30%) in the non-human primate. Multiscale modeling of realistic neural circuits suggests that alternating current stimulation-induced phase precession is driven by NMDA-mediated synaptic plasticity. Altogether, the three experiments provide mechanistic and causal evidence for phase precession as a global neocortical process. Alternating current-induced phase precession and consequently synaptic plasticity is crucial for the development of novel therapeutic neuromodulation methods.

Clinical evaluation of hysterectomy for the treatment of invasive mole in Southern Vietnam

OBJECTIVE

This study aimed to determine the rate of salvage chemotherapy and review associated factors in invasive mole patients treated by primary or delayed hysterectomy.

PATIENTS AND METHODS

This study was carried out at the Tu Du Hospital, where a total of 189 patients were diagnosed with invasive mole based on histologic examination by hysterectomy between 01/2016 to 12/2020. We used the life table method to estimate the cumulative rate. We applied the Cox proportional hazard model to determine the factors associated with the need for salvage chemotherapy.

RESULTS

At 12-month follow-up, 47 patients had required salvage chemotherapy. The incidence was 24.87% (95% CI: 18.88-31.66). Applying the multivariate model, prophylactic chemotherapy (HR = 2.75, 95% Cl: 1.20-6.30) and two weeks postoperative hCG value greater than 1,900 mIU/mL (HR = 4.30, 95% Cl: 2.08-8.87) increased the risk of requiring salvage chemotherapy. Postoperative chemotherapy decreased the risk of requiring salvage chemotherapy (HR = 0.43, 95% Cl: 0.22-0.83).

CONCLUSIONS

Hysterectomy can be considered safe and effective in treating invasive mole patients. Although patients were treated by hysterectomy, 24.87% of patients needed salvage chemotherapy to achieve remission. This study affirms the malignant nature of invasive mole, a subtype of gestational trophoblastic neoplasia (GTN). It is not purely a local invasion of molar villi. Postoperative chemotherapy plays an essential role in reducing the risk of requiring salvage chemotherapy.

A novel statistical method for decontaminating T-cell receptor sequencing data

The T-cell receptor (TCR) repertoire is highly diverse among the population and plays an essential role in initiating multiple immune processes. TCR sequencing (TCR-seq) has been developed to profile the T cell repertoire. Similar to other high-throughput experiments, contamination can happen during several steps of TCR-seq, including sample collection, preparation and sequencing. Such contamination creates artifacts in the data, leading to inaccurate or even biased results. Most existing methods assume 'clean' TCR-seq data as the starting point with no ability to handle data contamination. Here, we develop a novel statistical model to systematically detect and remove contamination in TCR-seq data. We summarize the observed contamination into two sources, pairwise and cross-cohort. For both sources, we provide visualizations and summary statistics to help users assess the severity of the contamination. Incorporating prior information from 14 existing TCR-seq datasets with minimum contamination, we develop a straightforward Bayesian model to statistically identify contaminated samples. We further provide strategies for removing the impacted sequences to allow for downstream analysis, thus avoiding any need to repeat experiments. Our proposed model shows robustness in contamination detection compared with a few off-the-shelf detection methods in simulation studies. We illustrate the use of our proposed method on two TCR-seq datasets generated locally.

Intensity- and frequency-specific effects of transcranial alternating current stimulation are explained by network dynamics

Transcranial alternating current stimulation (tACS) can be used to non-invasively entrain neural activity, and thereby cause changes in local neural oscillatory power. Despite an increased use in cognitive and clinical neuroscience, the fundamental mechanisms of tACS are still not fully understood. Here, we develop a computational neuronal network model of two-compartment pyramidal neurons and inhibitory interneurons which mimic the local cortical circuits. We model tACS with electric field strengths that are achievable in human applications. We then simulate intrinsic network activity and measure neural entrainment to investigate how tACS modulates ongoing endogenous oscillations. First, we show that intensity-specific effects of tACS are non-linear. At low intensities (<0.3 mV/mm), tACS desynchronizes neural firing relative to the endogenous oscillations. At higher intensities (>0.3 mV/mm), neurons are entrained to the exogenous electric field. We then further explore the stimulation parameter space and find that entrainment of ongoing cortical oscillations also depends on frequency by following an Arnold tongue. Moreover, neuronal networks can amplify the tACS induced entrainment via excitation-inhibition balance. Our model shows that pyramidal neurons are directly entrained by the exogenous electric field and drive the inhibitory neurons. Our findings can thus provide a mechanistic framework for understanding the intensity- and frequency- specific effects of oscillating electric fields on neuronal networks. This is crucial for rational parameters selection for tACS in cognitive studies and clinical applications.

Uncovering hypercoagulation status using rotational thromboelastometry in patients with sepsis presented with hypocoagulation based on conventional coagulation tests: an observational study

OBJECTIVE

Conventional coagulation tests (CCTs) cannot identify hypercoagulation, despite being common in patients with sepsis. Moreover, CCTs overdiagnose hypocoagulation, which increases unnecessary blood transfusion. Therefore, we aimed to use rotational thromboelastometry (ROTEM) to classify the coagulation status of patients with sepsis with abnormal CCTs and to identify the main coagulation components that affect coagulation status.

PATIENTS AND METHODS

This study was part of an observational study to investigate ROTEM use in 161 patients with sepsis with the Sepsis-3 criteria. They underwent concurrent CCTs and ROTEM assessments within 24 hours of Intensive Care Unit admission at the University Medical Center, Ho Chi Minh City, from June 2020 to December 2021. This study only extracted data from patients with sepsis with abnormal CCTs, including activated partial thromboplastin time ratio, international normalized ratio (INR), platelet count, and fibrinogen concentration.

RESULTS

A total of 158 patients with sepsis with abnormal CCTs had a median age of 69, and 48.7% were women. Of 34 patients with INR ≥1.6, ROTEM identified 11.8% with hypercoagulation and 20.6% with normal coagulation. Of 29 patients with platelet counts <100 (103/mm3), ROTEM identified 3.5% with hypercoagulation and 24.1% with normal coagulation. In the ROTEM-based hypercoagulability group, an increase in maximum clot firmness was observed in 95.1% of cases; also, this group had significantly higher plasma fibrinogen concentrations than other groups (p<0.005).

CONCLUSIONS

ROTEM can reveal hypercoagulability in patients with sepsis with hypocoagulation based on CCTs. Hyperfibrinogenemia causes hypercoagulation in patients with sepsis.

Abstract 1398: Clinical and transcriptomic analysis demonstrates improved survival and unique gene expression signatures among SCLC arising in patients with minimal tobacco use

Small-cell lung cancer (SCLC) represents 15% of all lung cancers with roughly 30,000 new cases in the U.S. annually. SCLC remains the deadliest histologic subtype with a median survival of only 12.3 months among those with extensive-stage (ES-SCLC) disease as demonstrated in the IMPower133 trial. Recent research by our group (Gay et al, 2021) has demonstrated that SCLC is defined by four transcriptionally defined subtypes, characterized by the predominant expression of the three transcription factors ASCL1 (SCLC-A), NEUROD1 (SCLC-N), POU2F3 (SCLC-P), and a fourth, inflamed subtype (SCLC-I). While SCLC is commonly associated with a history of heavy smoking, there is a paucity of information regarding tumors arising in patients with minimal tobacco use. This represents an unmet need given an increasing appreciation for low pack-year and never-smoker SCLC with some studies, like the CAPSTONE-1 trial, demonstrating an incidence of never-smoking SCLC &gt;20%. To address this disparity, we profiled a cohort of 113 SCLC patients with ≤10 pack-year smoking history (~12.9% of patients with SCLC seen over that time period) treated at the University of Texas MD Anderson Cancer Center. Clinical outcomes were analyzed, as was transcriptomic analysis in a subset of patients. The overall survival of limited-stage (LS-SCLC) and ES-SCLC in the low-pack year cohort was 26.9 and 16.5 months respectively (P=0.029); superior to historic medians of roughly 17.0 and 12.3 months respectively. Interestingly, TP53 and RB1 mutations were significantly less common in the low-pack year cohort compared to patients with higher smoking burdens as determined by George et al. who evaluated 110 SCLC samples using whole-genome sequencing. Clinically obtained mutational analysis of our cohort (N=49) demonstrated a TP53 mutational incidence of 51.0% vs 98.2% from George et al. (P&lt;0.001). Similarly, our RB1 mutational incidence was 26.5% vs 90.9% from George et al. (P&lt;0.001). Samples lacking TP53-RB1 coexisting mutations frequently demonstrated abnormalities in alternative DNA repair genes including STK11, POLE, PALB2, MUTYH, MSH2, MSH6, MLH1, MDM2, BRCA2, and ARID1A. Given these findings, we performed whole-transcriptome profiling on our low pack-year SCLC samples to determine the SCLC-subtypes and identify unique gene signatures which may drive oncogenesis via mechanisms distinct from loss of RB1 and TP53. These data demonstrate that SCLC arising from patients with minimal smoking histories confers a more favorable prognosis and harbors unique RNA signatures with potential therapeutic implications. We anticipate these results will shift the current clinical practice toward routine evaluation of non-RB1 or TP53-mediated drivers of oncogenesis among low pack-year patients with SCLC and promote further work in identifying novel therapies for this population.

Citation Format: Kyle Concannon, Simon Heeke, Moushumi Sahu, Ximing Tang, Koji Sasaki, Sonia Patel, Maria Gabriela Raso, Hai Tran, Carl Gay, Lauren Byers. Clinical and transcriptomic analysis demonstrates improved survival and unique gene expression signatures among SCLC arising in patients with minimal tobacco use [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 1398.

IL6 Mediates Suppression of T- and NK-cell Function in EMT-associated TKI-resistant EGFR-mutant NSCLC

PURPOSE

Patients with advanced non-small cell lung cancer (NSCLC) harboring activating EGFR mutations are initially responsive to tyrosine kinase inhibitors (TKI). However, therapeutic resistance eventually emerges, often via secondary EGFR mutations or EGFR-independent mechanisms such as epithelial-to-mesenchymal transition. Treatment options after EGFR-TKI resistance are limited as anti-PD-1/PD-L1 inhibitors typically display minimal benefit. Given that IL6 is associated with worse outcomes in patients with NSCLC, we investigate whether IL6 in part contributes to this immunosuppressed phenotype.

EXPERIMENTAL DESIGN

We utilized a syngeneic genetically engineered mouse model (GEMM) of EGFR-mutant NSCLC to investigate the effects of IL6 on the tumor microenvironment and the combined efficacy of IL6 inhibition and anti-PD-1 therapy. Corresponding in vitro studies used EGFR-mutant human cell lines and clinical specimens.

RESULTS

We identified that EGFR-mutant tumors which have oncogene-independent acquired resistance to EGFR-TKIs were more mesenchymal and had markedly enhanced IL6 secretion. In EGFR-mutant GEMMs, IL6 depletion enhanced activation of infiltrating natural killer (NK)- and T-cell subpopulations and decreased immunosuppressive regulatory T and Th17 cell populations. Inhibition of IL6 increased NK- and T cell-mediated killing of human osimertinib-resistant EGFR-mutant NSCLC tumor cells in cell culture. IL6 blockade sensitized EGFR-mutant GEMM tumors to PD-1 inhibitors through an increase in tumor-infiltrating IFNγ+ CD8+ T cells.

CONCLUSIONS

These data indicate that IL6 is upregulated in EGFR-mutant NSCLC tumors with acquired EGFR-TKI resistance and suppressed T- and NK-cell function. IL6 blockade enhanced antitumor immunity and efficacy of anti-PD-1 therapy warranting future clinical combinatorial investigations.

Sinus arrest in familial hypokalemic periodic paralysis caused by SCN4A mutation: a case report

BACKGROUND

Primary hypokalemic periodic paralysis (HypoPP), a rare skeletal muscle channelopathy resulting in episodic muscle weakness or paralysis under hypokalemic conditions, is caused by autosomal-dominant genetic mutations. HypoPP limits physical activity, and cardiac arrhythmias during paralytic attacks have been reported. We describe a rare familial HypoPP case complicated by sinus arrest and syncope requiring urgent temporary pacemaker implantation.

CASE REPORT

A 27-year-old Vietnamese man with a family history of periodic paralysis presented with his third attack of muscle weakness triggered by intense football training the previous day. Clinical and laboratory features justified a HypoPP diagnosis. During intravenous potassium replacement, the patient experienced syncopal sinus arrest requiring urgent temporary pacemaker implantation. The patient gradually improved, responding favorably to oral potassium supplements. Genetic testing revealed an Arg1132Gln mutation in the sodium ion channel (SCN4A, chromosome 17: 63947091). At discharge, the patient received expert consultation regarding nonpharmacological preventive strategies, including avoidance of vigorous exercise and carbohydrate-rich diet.

CONCLUSIONS

No evidence has established a relationship between hypokalemia and sinus arrest, and no specific treatment exists for familial HypoPP due to SCN4A mutation. Clinician awareness of this rare condition will promote appropriate diagnostic approaches and management strategies for acute paralytic attacks. Treatment should be tailored according to HypoPP phenotypes and genotypes.

CD70 is a therapeutic target upregulated in EMT-associated EGFR tyrosine kinase inhibitor resistance

Effective therapeutic strategies are needed for non-small cell lung cancer (NSCLC) patients with epidermal growth factor receptor (EGFR) mutations that acquire resistance to EGFR tyrosine kinase inhibitors (TKIs) mediated by epithelial-to-mesenchymal transition (EMT). We investigate cell surface proteins that could be targeted by antibody-based or adoptive cell therapy approaches and identify CD70 as being highly upregulated in EMT-associated resistance. Moreover, CD70 upregulation is an early event in the evolution of resistance and occurs in drug-tolerant persister cells (DTPCs). CD70 promotes cell survival and invasiveness, and stimulation of CD70 triggers signal transduction pathways known to be re-activated with acquired TKI resistance. Anti-CD70 antibody drug conjugates (ADCs) and CD70-targeting chimeric antigen receptor (CAR) T cell and CAR NK cells show potent activity against EGFR TKI-resistant cells and DTPCs. These results identify CD70 as a therapeutic target for EGFR mutant tumors with acquired EGFR TKI resistance that merits clinical investigation.

Analgesic dorsal root ganglion field stimulation blocks both afferent and efferent spontaneous activity in sensory neurons of rats with monosodium iodoacetate-induced osteoarthritis

OBJECTIVES

Chronic joint pain is common in patients with osteoarthritis (OA). Non-steroidal anti-inflammatory drugs and opioids are used to relieve OA pain, but they are often inadequately effective. Dorsal root ganglion field stimulation (GFS) is a clinically used neuromodulation approach, although it is not commonly employed for patients with OA pain. GFS showed analgesic effectiveness in our previous study using the monosodium iodoacetate (MIA) - induced OA rat pain model. This study was to evaluate the mechanism of GFS analgesia in this model.

METHODS

After osteoarthritis was induced by intra-articular injection of MIA, pain behavioral tests were performed. Effects of GFS on the spontaneous activity (SA) were tested with in vivo single-unit recordings from teased fiber saphenous nerve, sural nerve, and dorsal root.

RESULTS

Two weeks after intra-articular MIA injection, rats developed pain-like behaviors. In vivo single unit recordings from bundles teased from the saphenous nerve and third lumbar (L3) dorsal root of MIA-OA rats showed a higher incidence of SA than those from saline-injected control rats. GFS at the L3 level blocked L3 dorsal root SA. MIA-OA reduced the punctate mechanical force threshold for inducing AP firing in bundles teased from the L4 dorsal root, which reversed to normal with GFS. After MIA-OA, there was increased retrograde SA (dorsal root reflex), which can be blocked by GFS.

CONCLUSIONS

These results indicate that GFS produces analgesia in MIA-OA rats at least in part by producing blockade of afferent inputs, possibly also by blocking efferent activity from the dorsal horn.

P02.11.B An hypothesis generating study of MRI-Derived Radiomics on tumor and microenvironment tissue heterogeneity to guide post-operative management of glioblastoma: toward personalized radiation treatment volume delineation

Background

The glioblastoma’s bad prognosis is primarily due to intra-tumor heterogeneity, demonstrated from several studies that collected molecular biology, cytogenetic data and more recently radiomic features for a better prognostic stratification.The GLIFA project (GLIoblastoma Feature Analysis) is a multicentric project planned to investigated the role of radiomic analysis in GBM management, to verify if radiomic features in the tissue around the resection cavity which may guide the radiation target volume delineation.

Material and Methods

We retrospectively analyze from three centers radiomic features extracted from 90 patients with total or near total resection, who completed the standard adjuvant treatment and for whom we had post-operative images available for features extraction.The Manual segmentation was performed on post gadolinium T1w MRI sequence by 2 radiation oncologist reviewed by a neuroradiologist, both with at least 10 years of experience. The Region of interest (ROI) considered for the analysis were: the surgical cavity +/- post-surgical residual mass (CTV_cavity); the CTV a margin of 1.5 cm added to CTV_cavity and the volume resulting from subtracting the CTV_cavity from the CTV was defined as CTV_Ring. Radiomic analysis and modelling were conducted in RStudio. Z-score normalization was applied to each radiomic feature. A radiomic model was generated using the 226 features extracted from the Ring to perform a binary classification and predict the PFS at 6 months (statistical, morphological and textural features). A 3-fold cross-validation repeated five times was implemented for internal validation of the model.

Results

Two-hundred and seventy ROIs were contoured. The proposed radiomic model was given by the best fitting logistic regression model, and included the following 3 features: F_cm_merged.contrast, F_cm_merged.info.corr.2, F_rlm_merged.rlnu. A good agreement between model predicted probabilities and observed outcome probabilities was obtained (p-value of 0.49 by Hosmer and Lemeshow statistical test). The ROC curve of the model reported an AUC of 0.78 (95% CI: 0.68 - 0.88).

Conclusion

This is the first hypothesis-generating study who applies a radiomic analysis focusing on healthy tissue ring around the surgical cavity on post-operative MRI. This study provides a preliminary model for a decision support tool for a customization of the radiation target volume in GBM patients in order to achieve a margin reduction strategy.