WS01.06When is best to start treatment with tobramycin in people with chronic Pseudomonas aeruginosa infection? An emulated trial using the UK CF Registry

Abstract

Pseudomonas aeruginosa (Pa) is among the most common pathogen detected in people with cystic fibrosis (CF). Once infection is established, the management of chronic Pa infection requires long-term suppressive treatment. In the UK, colistimethate is recommend as a first line treatment for chronic Pa with tobramycin second line. Second line treatment is considered in those who are clinically deteriorating despite regular inhaled colistimethate, however little guidance exists as to what constitutes clinical deterioration.

Our aim is to determine at what level of lung function is best to add tobramycin in people with CF who are already taking colistimethate to treat chronic Pa.

Using data from the UK CF Registry, we include individuals who have been taking colistimethate for at least one year, have isolated Pa and are aged at least 6 years. We applied inverse-probability-of-treatment weighting to investigate when is best to add tobramycin in people who met our eligibility criteria. The goal was to optimise X, when the treatment decision rule is to “initiate long-term treatment with tobramycin when the percent forced expiratory volume in one second (FEV₁%) first falls under X”. We considered values of X ranging between 20 and 120 and the outcome of interest was FEV₁% after 1-, 2-, 3-, 4- and 5- years.

A preliminary analysis, based on UK CF Registry data from 2016-2018, found no evidence that expected outcomes changed for individuals who initiated tobramycin at different levels of lung function. We plan to present the results of an updated analysis which includes more recent data up to 2022.

Our study provides an illustrative example of optimising a dynamic treatment regime, i.e., a sequence of decision rules which dictate a person's subsequent treatment, conditional on their treatment and clinical history. This approach has the potential to improve personalised treatment decisions by informing clinical guidance on treatment regimes that change over time.