SIOP Ependymoma I: Final results, long-term follow-up, and molecular analysis of the trial cohort-A BIOMECA Consortium Study

BACKGROUND

SIOP Ependymoma I was a non-randomised trial assessing event free and overall survival (EFS/OS) of non-metastatic intracranial ependymoma in children aged 3-21 years treated with a staged management strategy. A further aim was to assess the response rate (RR) of subtotally resected (STR) ependymoma to vincristine, etoposide, and cyclophosphamide (VEC). We report final results with 12-year follow-up and post hoc analyses of recently described biomarkers.

METHODS

Seventy-four participants were eligible. Children with gross total resection (GTR) received radiotherapy, whilst those with STR received VEC before radiotherapy. DNA methylation, 1q, hTERT, ReLA, Tenascin-C, H3K27me3, and pAKT status were evaluated.

RESULTS

Five- and ten-year EFS was 49.5% and 46.7%, OS was 69.3% and 60.5%. GTR was achieved in 33/74 (44.6%) and associated with improved EFS (P = .003, HR = 2.6, 95% confidence interval (CI) 1.4-5.1). Grade 3 tumours were associated with worse OS (P = .005, HR = 2.8, 95%CI 1.3-5.8). 1q gain and hTERT expression were associated with poorer EFS (P = .003, HR = 2.70, 95%CI 1.49-6.10 and P = .014, HR = 5.8, 95%CI 1.2-28) and H3K27me3 loss with worse OS (P = .003, HR = 4.6, 95%CI 1.5-13.2). Methylation profiles showed expected patterns. 12 participants with STR did not receive chemotherapy; a protocol violation. However, best chemotherapy RR was 65.5% (19/29, 95%CI 45.7-82.1), exceeding the prespecified 45%.

CONCLUSIONS

Participants with totally resected ependymoma had the best outcomes. RR of STR to VEC exceeded the pre-specified efficacy criterion. However, cases of inaccurate stratification highlighted the need for rapid central review. 1q gain, H3K27me3 loss, and hTERT expression were all associated with poorer survival outcomes.

The impact of the COVID-19 pandemic on antimicrobial prescribing at a specialist paediatric hospital: an observational study

BACKGROUND

The COVID-19 pandemic has severely impacted healthcare delivery and there are growing concerns that the pandemic will accelerate antimicrobial resistance.

OBJECTIVES

To evaluate the impact of the COVID-19 pandemic on antibiotic prescribing in a tertiary paediatric hospital in London, UK.

METHODS

Data on patient characteristics and antimicrobial administration for inpatients treated between 29 April 2019 and Sunday 28 March 2021 were extracted from the electronic health record (EHR). Interrupted time series analysis was used to evaluate antibiotic days of therapy (DOT) and the proportion of prescribed antibiotics from the WHO 'Access' class.

RESULTS

A total of 23 292 inpatient admissions were included. Prior to the pandemic there were an average 262 admissions per week compared with 212 during the pandemic period. Patient demographics were similar in the two periods but there was a shift in the specialities that patients had been admitted to. During the pandemic, there was a crude increase in antibiotic DOTs, from 801 weekly DOT before the pandemic to 846. The proportion of Access antibiotics decreased from 44% to 42%. However, after controlling for changes in patient characteristics, there was no evidence for the pandemic having an impact on antibiotic prescribing.

CONCLUSIONS

The patient population in a specialist children's hospital was affected by the COVID-19 pandemic, but after adjusting for these changes there was no evidence that antibiotic prescribing was significantly affected by the pandemic. This highlights both the value of routine, high-quality EHR data and importance of appropriate statistical methods that can adjust for underlying changes to populations when evaluating impacts of the pandemic on healthcare.

Nothing about us without us: involving patient collaborators for machine learning applications in rheumatology

Novel machine learning methods open the door to advances in rheumatology through application to complex, high-dimensional data, otherwise difficult to analyse. Results from such efforts could provide better classification of disease, decision support for therapy selection, and automated interpretation of clinical images. Nevertheless, such data-driven approaches could potentially model noise, or miss true clinical phenomena. One proposed solution to ensure clinically meaningful machine learning models is to involve primary stakeholders in their development and interpretation. Including patient and health care professionals' input and priorities, in combination with statistical fit measures, allows for any resulting models to be well fit, meaningful, and fit for practice in the wider rheumatological community. Here we describe outputs from workshops that involved healthcare professionals, and young people from the Your Rheum Young Person's Advisory Group, in the development of complex machine learning models. These were developed to better describe trajectory of early juvenile idiopathic arthritis disease, as part of the CLUSTER consortium. We further provide key instructions for reproducibility of this process.Involving people living with, and managing, a disease investigated using machine learning techniques, is feasible, impactful and empowering for all those involved.