Frequent Overexpression of HER3 in Brain Metastases from Breast and Lung Cancer.

PURPOSE: HER3 belongs to a family of receptor tyrosine kinases with oncogenic properties and is targeted by a variety of novel anticancer agents. There is a huge unmet medical need for systemic treatment options in patients with brain metastases (BM). Therefore, we aimed to investigate HER3 expression in BM of breast (BCa) and non-small cell lung cancer (NSCLC) as the basis for future clinical trial design. EXPERIMENTAL DESIGN: We analyzed 180 BM samples of breast cancer or NSCLC and 47 corresponding NSCLC extracranial tissue. IHC was performed to evaluate protein expression of HER3, and immune cells based on CD3, CD8, and CD68. To identify dysregulated pathways based on differential DNA methylation patterns, we used Infinium MethylationEPIC microarrays. RESULTS: A total of 99/132 (75.0%) of BCa-BM and 35/48 (72.9%) of NSCLC-BM presented with HER3 expression. Among breast cancer, HER2-positive and HER2-low BM showed significantly higher rates of HER3 coexpression than HER2-negative BM (87.1%/85.7% vs. 61.0%, P = 0.004). Among NSCLC, HER3 was more abundantly expressed in BM than in matched extracranial samples (72.9% vs. 41.3%, P = 0.003). No correlation of HER3 expression and intratumoral immune cell density was observed. HER3 expression did not correlate with overall survival from BM diagnosis. Methylation signatures differed according to HER3 status in BCa-BM samples. Pathway analysis revealed subtype-specific differences, such as TrkB and Wnt signaling pathways dysregulated in HER2-positive and triple-negative breast cancer BM, respectively. CONCLUSIONS: HER3 is highly abundant in BM of breast cancer and NSCLC. Given the promising results of antibody-drug conjugates in extracranial disease, BM-specific trials that target HER3 are warranted. See related commentary by Kabraji and Lin, p. 2961.

Prophylactic treatment with oral azithromycin in cancer patients during the COVID-19 pandemic (OnCoVID): a randomized, single-blinded, placebo-controlled phase 2 trial.

BACKGROUND: Patients with cancer are at high risk for severe courses of COVID-19. Based on (pre-)clinical data suggesting a potential protective effect due to the immunomodulating properties of azithromycin, we have initiated a prospective randomized trial. METHODS: This randomized, single-center, single-blinded, placebo-controlled phase 2 trial included adult patients with cancer undergoing systemic treatment. Patients were 1:1 randomized to oral azithromycin (1500 mg once weekly for 8 weeks) or placebo. The primary endpoint was the cumulative number of SARS-CoV-2 infections 12 weeks after treatment initiation. RESULTS: In total, 523 patients were screened, 68 patients were randomized, and 63 patients received at least one dose of the study drug. Due to low acceptance and a lack of SARS-CoV-2 infections in the study cohort, the study was prematurely closed. With no reported grade III-IV possibly treatment-related adverse events, azithromycin was generally well tolerated. Overall survival (OS) rates after 12 months were 83.5% and 70.3% in the azithromycin and placebo group, respectively (p = 0.37). Non-SARS-CoV-2 infections occurred in 4/32 (12.5%) in the azithromycin and 3/31 (9.7%) in the placebo group (p = 1). No emergence of azithromycin-resistant S. aureus strains could be observed. According to treatment group, longitudinal alterations in systemic inflammatory parameters were detected for neutrophil/lymphocyte and leukocyte/lymphocyte ratios. CONCLUSION: Although efficacy could not be assessed due to premature closure and low incidence of SARS-CoV-2 infections, azithromycin was associated with a favorable side effect profile in patients with cancer. As other prophylactic treatments are limited, SARS-CoV-2 vaccination remains a high priority in oncological patients. CLINICALTRIALS: gov registration number and date (dd/mm/yyyy): NCT04369365, 30/04/2020.

Trastuzumab deruxtecan in HER2-positive breast cancer with brain metastases: a single-arm, phase 2 trial.

Trastuzumab deruxtecan is an antibody-drug conjugate with high extracranial activity in human epidermal growth factor receptor 2 (HER2)-positive metastatic breast cancer. We conducted the prospective, open-label, single-arm, phase 2 TUXEDO-1 trial. We enrolled patients aged ≥18 years with HER2-positive breast cancer and newly diagnosed untreated brain metastases or brain metastases progressing after previous local therapy, previous exposure to trastuzumab and pertuzumab and no indication for immediate local therapy. Patients received trastuzumab deruxtecan intravenously at the standard dose of 5.4 mg per kg bodyweight once every 3 weeks. The primary endpoint was intracranial response rate measured according to the response assessment in neuro-oncology brain metastases criteria. A Simon two-stage design was used to compare a null hypothesis of <26% response rate against an alternative of 61%. Fifteen patients were enrolled in the intention-to-treat population of patients who received at least one dose of study drug. Two patients (13.3%) had a complete intracranial response, nine (60%) had a partial intracranial response and three (20%) had stable disease as the best intracranial response, with a best overall intracranial response rate of 73.3% (95% confidential interval 48.1-89.1%), thus meeting the predefined primary outcome. No new safety signals were observed and global quality-of-life and cognitive functioning were maintained over the treatment duration. In the TUXEDO-1 trial (NCT04752059, EudraCT 2020-000981-41), trastuzumab deruxtecan showed a high intracranial response rate in patients with active brain metastases from HER2-positive breast cancer and should be considered as a treatment option in this setting.

Early Postoperative Treatment versus Initial Observation in CNS WHO Grade 2 and 3 Oligodendroglioma: Clinical Outcomes and DNA Methylation Patterns.

PURPOSE: The treatment of oligodendroglioma consists of tumor resection and radiochemotherapy. The timing of radiochemotherapy remains unclear, and predictive biomarkers are limited. EXPERIMENTAL DESIGN: Adult patients diagnosed with isocitrate dehydrogenase (IDH)-mutated, 1p/19q-codeleted CNS WHO grade 2 and 3 oligodendroglioma at the Medical University of Vienna and the Kepler University Hospital Linz (Austria) in 1992 to 2019 were included. Progression-free (PFS) and overall survival (OS) between early postoperative treatment and initial observation were compared using propensity score-weighted Cox regression models. DNA methylation analysis of tumor tissue was performed using Illumina MethylationEPIC 850k microarrays. RESULTS: One hundred thirty-one out of 201 (65.2%) patients with CNS WHO grade 2 and 70 of 201 (34.8%) with grade 3 oligodendroglioma were identified. Eighty-three of 201 (41.3%) patients underwent early postoperative treatment, of whom 56 of 83 (67.5%) received radiochemotherapy, 15 of 84 (18.1%) radiotherapy (RT) only and 12 of 83 (14.5%) chemotherapy only. Temozolomide-based treatment was administered to 64 of 68 (94.1%) patients, whereas RT + procarbazine, lomustine (CCNU), and vincristine (PCV) were applied in 2 of 69 (3.5%) patients. Early treatment was not associated with PFS [adjusted hazard ratio (HR) 0.74; 95% CI, 0.33-1.65, P = 0.459] or OS (adjusted HR: 2.07; 95% CI, 0.52-8.21, P = 0.302) improvement. Unsupervised clustering analysis of DNA methylation profiles from patients receiving early treatment revealed two methylation clusters correlating with PFS, whereas no association of clustering with O6-methylguanine methyltransferase (MGMT) promoter methylation, CNS WHO grade, extent of resection, and treating center could be observed. CONCLUSIONS: In this retrospective study, early postoperative treatment was not associated with improved PFS/OS in oligodendroglioma. The potentially predictive value of whole-genome methylation profiling should be validated in prospective trials.

Uptake and Discontinuation of Integrase Inhibitors (INSTIs) in a Large Cohort Setting.

BACKGROUND: Despite increased integrase strand transfer inhibitor (INSTI) use, limited large-scale, real-life data exists on INSTI uptake and discontinuation. SETTING: International multicohort collaboration. METHODS: RESPOND participants starting dolutegravir (DTG), elvitegravir (EVG), or raltegravir (RAL) after January 1, 2012 were included. Predictors of INSTI used were assessed using multinomial logistic regression. Kaplan-Meier and Cox proportional hazards models describe time to and factors associated with discontinuation. RESULTS: Overall, 9702 persons were included; 5051 (52.1%) starting DTG, 1933 (19.9%) EVG, and 2718 (28.0%) RAL. The likelihood of starting RAL or EVG vs DTG decreased over time and was higher in Eastern and Southern Europe compared with Western Europe. At 6 months after initiation, 8.9% (95% confidence interval: 8.3% to 9.5%) had discontinued the INSTI (6.4% DTG, 7.4% EVG, and 14.0% RAL). The main reason for discontinuation was toxicity (44.2% DTG, 42.5% EVG, 17.3% RAL). Nervous system toxicity accounted for a higher proportion of toxicity discontinuations on DTG (31.8% DTG, 23.4% EVG, 6.6% RAL). Overall, treatment simplification was highest on RAL (2.7% DTG, 1.6% EVG, and 19.8% RAL). Factors associated with a higher discontinuation risk included increasing year of INSTI initiation, female gender, hepatitis C coinfection, and previous non-AIDS-defining malignancies. Individuals in Southern and Eastern Europe were less likely to discontinue. Similar results were seen for discontinuations after 6 months. CONCLUSIONS: Uptake of DTG vs EVG or RAL increased over time. Discontinuation within 6 months was mainly due to toxicity; nervous system toxicity was highest on DTG. Discontinuation was highest on RAL, mainly because of treatment simplification.

Human immunodeficiency virus-infected young people in Australia: data from the Australian HIV Observational Database.

BACKGROUND: Individuals aged 13-24 years undergo vast physical, cognitive, social and psychological changes. Australian data regarding clinical outcomes of those diagnosed with HIV in this age are sparse. AIM: We aimed to describe demographic factors, virologic and clinical outcomes of individuals aged 13-24 years diagnosed with human immunodeficiency virus (HIV). METHODS: Patients diagnosed with HIV after 1997 in the Australian HIV Observational Database were divided into young adults, diagnosed at age <25 years (n = 223), and older adults (n = 1957). Demographic and clinical factors were compared between groups. RESULTS: Young adults had a median age at diagnosis of 22 years (inter quartile range (IQR) 20-24) and median age at treatment initiation of 24 years (IQR 22-26). They were more likely to be female than the older cohort (21.1 vs 10.8%; P < 0.001). Men who have sex with men was the most common exposure category in both groups. CD4 count at diagnosis was significantly higher in younger than older adults (median 460 vs 400 cells/mm3 , P = 0.006), whereas HIV viral load at diagnosis was lower (35 400 vs 61 659 copies/mL, P = 0.011). The rate of loss to follow up (LTFU) was higher in young adults (8.0 vs 4.3 per 100PY, P < 0.001). Young adults were more likely to have a treatment interruption compared to older adults (5.3 vs 4.0 per 100PY, P = 0.039). Rates of treatment switch, time to treatment change, and CD4 and viral load responses to treatment were similar between groups. CONCLUSIONS: Young adults were diagnosed with HIV at higher CD4 counts and lower viral loads than their older counterparts. LTFU and treatment interruption were more common highlighting the need for extra efforts directed towards retention in care and education regarding the risks of treatment interruptions.

Firth's logistic regression with rare events: accurate effect estimates and predictions?

Firth's logistic regression has become a standard approach for the analysis of binary outcomes with small samples. Whereas it reduces the bias in maximum likelihood estimates of coefficients, bias towards one-half is introduced in the predicted probabilities. The stronger the imbalance of the outcome, the more severe is the bias in the predicted probabilities. We propose two simple modifications of Firth's logistic regression resulting in unbiased predicted probabilities. The first corrects the predicted probabilities by a post hoc adjustment of the intercept. The other is based on an alternative formulation of Firth's penalization as an iterative data augmentation procedure. Our suggested modification consists in introducing an indicator variable that distinguishes between original and pseudo-observations in the augmented data. In a comprehensive simulation study, these approaches are compared with other attempts to improve predictions based on Firth's penalization and to other published penalization strategies intended for routine use. For instance, we consider a recently suggested compromise between maximum likelihood and Firth's logistic regression. Simulation results are scrutinized with regard to prediction and effect estimation. We find that both our suggested methods do not only give unbiased predicted probabilities but also improve the accuracy conditional on explanatory variables compared with Firth's penalization. While one method results in effect estimates identical to those of Firth's penalization, the other introduces some bias, but this is compensated by a decrease in the mean squared error. Finally, all methods considered are illustrated and compared for a study on arterial closure devices in minimally invasive cardiac surgery. Copyright © 2017 John Wiley & Sons, Ltd.

Bias-reduced and separation-proof conditional logistic regression with small or sparse data sets.

Conditional logistic regression is used for the analysis of binary outcomes when subjects are stratified into several subsets, e.g. matched pairs or blocks. Log odds ratio estimates are usually found by maximizing the conditional likelihood. This approach eliminates all strata-specific parameters by conditioning on the number of events within each stratum. However, in the analyses of both an animal experiment and a lung cancer case-control study, conditional maximum likelihood (CML) resulted in infinite odds ratio estimates and monotone likelihood. Estimation can be improved by using Cytel Inc.'s well-known LogXact software, which provides a median unbiased estimate and exact or mid-p confidence intervals. Here, we suggest and outline point and interval estimation based on maximization of a penalized conditional likelihood in the spirit of Firth's (Biometrika 1993; 80:27-38) bias correction method (CFL). We present comparative analyses of both studies, demonstrating some advantages of CFL over competitors. We report on a small-sample simulation study where CFL log odds ratio estimates were almost unbiased, whereas LogXact estimates showed some bias and CML estimates exhibited serious bias. Confidence intervals and tests based on the penalized conditional likelihood had close-to-nominal coverage rates and yielded highest power among all methods compared, respectively. Therefore, we propose CFL as an attractive solution to the stratified analysis of binary data, irrespective of the occurrence of monotone likelihood. A SAS program implementing CFL is available at: http://www.muw.ac.at/msi/biometrie/programs.