Low Dose Whole Lung Radiation Therapy for Patients With COVID-19 and Respiratory Compromise

Overview

Low doses of radiation in the form of chest X-rays have been used to treat people with pneumonia. This treatment was found to be effective by reducing inflammation and with minimal side effects. However, it was an expensive treatment and was eventually replaced with less costly treatments such as antibiotics. Radiation has also been shown in some animal experiments to reduce some types of inflammation. Some patients diagnosed with COVID-19 pneumonia will experience worsening disease, which can become very serious, requiring the use of a ventilator. This is caused by inflammation in the lung from the virus and the immune system. For this study, the x-ray given is called radiation therapy. Radiation therapy uses high-energy X-ray beams from a large machine to target the lungs and reduce inflammation. Usually, it is given at much higher doses to treat cancers. The purpose of this study is to find out if adding a single treatment of low-dose x-rays to the lungs might reduce the amount of inflammation in the lungs from a COVID-19 infection, which could help a patient to breathe without use of a ventilator.

The primary outcome is the mortality rate 30 days after the ICU-based mechanical ventilation initiation.Based on current data available, the mortality rate for ventilated patients is assumed to be 80% in the current design. An interim futility analysis will be conducted after 16 evaluable patients have received the ultralow dose-whole lung radiation therapy (ULD-WLRT). If at least 3 patients survive for at least 30 days, we will enroll additional 8 patients (total of 24 patients). Otherwise, the trial will stop for further evaluation. Due to the limited data currently available in local institutions about a 30 day mortality rate , we will retrospectively evaluate the mortality rate of the ventilated patients without the ULD-WLRT in our institution when data is available. Time to event secondary objectives (e.g. overall survival, time to discharge) analyses will be performed using Kaplan-Meier survival analysis, with a competing risk model (leaving the study because of death), including effects for demographic/clinical characteristics in the model. Proportional endpoints (such as % patients off ventilator) will be calculated along with the 95% Clopper-Pearson exact confidence interval. Pre/post measurements will be evaluated using linear mixed models for repeated measures (with proper data transformation as needed). Association between demographic/clinical characteristics and other secondary objectives (size of ground glass opacities (GGO)/opacification, for example) will be accomplished with generalized linear models.