Disease monitoring using lung function trajectory in lymphangioleiomyomatosis: assessment in two national cohorts.

Abstract

In lymphangioleiomyomatosis, airflow obstruction and impairment of gas transfer progress at variable rates and serial lung function is recommended for disease monitoring. As these measurements are variable, recognising subjects needing treatment can be difficult. We used two prospective national cohorts to study change over time and variation in FEV₁ to inform clinical decision making. Clinical and lung function data for 141 UK and 148 American subjects were studied. Multilevel mixed effects modelling, route mean square analysis of errors and Bland-Altman analysis were used to analyse variability in lung function over time. At baseline assessment, DL_CO was reduced to a greater degree than FEV₁. In untreated patients, FEV₁ and DL_CO declined at proportionately similar rates independent of initial lung function. In mechanistic target of rapamycin (mTOR) inhibitor treated patients, FEV₁ stabilised but DL_CO continued to decline. FEV₁/DL_CO per cent predicted ratio was 1.37 (0.43) at baseline and increased to 1.41 (0.50) after 42 (24) months (p=0.0002). At least five measurements were required before >70% of individuals had estimates of rate of FEV₁ loss within 50 mL/year and DL_CO loss within 0.1 mmol/min/kPa/year of the final values. While FEV₁ and DL_CO fall proportionately in most, in early disease and during mTOR inhibitor treatment, DL_CO should also be monitored as it may fall independent of FEV₁. Since at least five observations over many months are required to make confident estimates of FEV₁ and DL_CO trajectories, new strategies are needed to measure disease activity and target early treatment appropriately.