See article by Davids et al., p. 4371

Venetoclax, an oral BCL-2 inhibitor, is approved for treatment of patients with previously-treated chronic lymphocytic leukemia (CLL). With increased use of venetoclax in clinical practice, Davids and colleagues report on a comprehensive safety analysis of venetoclax monotherapy at the approved 400 mg daily dose in 350 patients enrolled in three phase-I or -II studies. The most common toxicities of venetoclax in relapsed/refractory CLL were mild gastrointestinal symptoms and transient, uncomplicated neutropenia. Tumor lysis syndrome at initiation was observed in some of the initial patients, but this risk was subsequently mitigated by risk-adapted prophylaxis, monitoring and stepwise dose ramp-up to 400 mg/day. Long-term continuous therapy with venetoclax was generally well-tolerated in patients with relapsed/refractory CLL when initiated with the current treatment algorithm. This study will help to inform the optimal management of CLL patients receiving venetoclax.

See article by Dickler et al., p. 4380

Taselisib is a potent and selective PI3K inhibitor with greater efficacy against mutant versus wild-type phosphatidylinositol-4,5bisphosphate 3-kinase, catalytic subunit-alpha (PIK3CA). Inhibition of PI3K signaling in HR-positive breast cancer results in upregulation of estrogen receptor (ER)-dependent function. The combination of PI3K inhibition and ER inhibition with fulvestrant produces marked tumor regression in breast cancer models, providing a rationale for dual PI3K and ER inhibition in breast cancer. In this phase II trial, Dickler and colleagues evaluated the safety and efficacy of oral taselisib plus intramuscular injections of fulvestrant in women with advanced HR-positive breast cancer. The adverse event profile was typical of that previously reported with PI3K inhibitors, and objective responses were obtained in women with PIK3CA-mutated and -mutation-not-detected tumors. On this basis, taselisib plus fulvestrant is being evaluated in a phase III study.

See article by Hu et al., p. 4437

Plasma cell-free DNA (cfDNA) analysis has emerged as a powerful noninvasive tool for cancer genotyping and detection. Hu and colleagues report on a surprising source of tumor-plasma discordance and false-positive plasma genotyping – mutant DNA derived from peripheral blood cells (PBC), known as clonal hematopoiesis (CH). Through paired tumor, cfDNA, and PBC analysis the investigators identify that most JAK2 mutations, some TP53 mutations, and rare KRAS mutations in cfDNA can be traced to CH. Thus, mutations detected in cfDNA may not always reflect tumor genotype, impacting how cfDNA genomics evolves further as a tool for cancer genotyping and detection.

See article by Lu et al., p. 4429

Machine learning based on the radiomics of multimodal magnetic resonance images (MRI) provides an alternative for noninvasive assessment of glioma subtypes in line with 2016 WHO classification. A three-level machine-learning model was proposed to stratify molecular subtypes of gliomas. The accuracies of classifications for the IDH and 1p/19q status of gliomas were between 87.7% and 96.1%. The complete classification of five molecular subtypes solely based on the MRI radiomics achieved an 81.8% accuracy, and a higher accuracy of 89.2% could be achieved if the histology diagnosis is available. The proposed approach can add value in the molecular diagnosis of gliomas by clinical MRI.