Determinants of COPD Stage Progression and Regression: a Markov Transition Model Analysis of The COPDGene Cohort

RationaleChronic obstructive pulmonary disease (COPD) is a leading cause of death but with variable progression. ObjectiveEstimate factors influencing transition rates between PRISm and GOLD stages. MethodsUsing a Markov multistate model, transition rates between GOLD-0, PRISm and GOLD-1, GOLD-2 and GOLD 3-4 were estimated for 5,728 US adult ever cigarette users from the COPDGene cohort over 10-years. We calculated one and five-year transition probabilities for progressive and regressive transitions and estimated the mean sojourn time for severity states. Main ResultsGOLD-1 and PRISm individuals spent the least time in any single stage (GOLD-1: 6 years; PRISm: 7 years). PRISm and GOLD-1 individuals were equally likely to transition to GOLD-2 vs. GOLD-0 (PRISm: HR 1.09, 95% confidence interval [CI] 0.90-1.33, GOLD-1 (HR 1.15, 95%CI 0.93-1.42) per five-year period, but rarely transition between PRISm and GOLD-1. Individuals at GOLD-0 were equally likely to progress to GOLD-1 or PRISm (HR 1.11, 95%CI 0.93-1.31) but the transient time for this stage was the longest of any GOLD stage (16 years, 95%CI 15.2-17.3). GOLD-2 was the most likely stage to progress (HR 2.4, 95%CI 1.9-3.02) to GOLD 3-4 vs. regress to GOLD-1. For GOLD-2 individuals, current smoking status (HR 0.84, 95%CI 0.67-1.06) or intensity (HR 0.84, 95%CI 0.54-1.29) was not associated with disease progression. ConclusionsGOLD-1 and PRISm are the most transient stages equally likely to regress to GOLD-0 or progress to GOLD-2 and may benefit from smoking cessation interventions. GOLD-2 individuals are the most likely to progress and may benefit most from targeted disease interventions.