Background/Purpose:

The majority of published models for prediction of PH use cross-sectional data, while studies exploring the use of repeated measurements of lung function tests (PFTs) are lacking.  Gas transfer factor (DLCO) and forced vital capacity (FVC) are among the most powerful predictors of development of pulmonary hypertension (PH) in patients with systemic sclerosis (SSc).

We examine the changes in FVC and DLCO over time in a large unselected cohort of SSc patients and to explore their association with development of PH.

Methods:

We identified subjects with SSc with follow-up data for over 10 years and PFTs performed on at least two occasions. Linear mixed models were used to look for associations between repeated measurements of FVC and DLCO and development of PH over the follow-up period.

Results:

In this study we included 371 SSc subjects – 13% of those were male, 36% had diffuse cutaneous (dc)SSc. The cohort demonstrated antibody frequencies similar to other previously published studies – 27% had anti-centromere antibody (ACA), 20% had anti-topoisomerase I antibody, 8% had anti-RNA polymerase antibodies (ARA) and 6% had anti-U3RNP antibodies. Of the whole cohort, 15% had developed PH at the end of follow-up and its cumulative incidence was 4% at 5 years and 13% at 10 years. PFTs were performed on average every 15 months, between 2 and 20 times (median 6 times).

The extrapolated average FVC and DLCO of the cohort at the time of disease onset (baseline) were 86% and 68% respectively. Overall, FVC tended to increase over time by an average of 0.5% per year (SD 1.29) while DLCO decreased by an average of 1% per year (SD 1.3).

Without adjusting for other factors, patients who developed PH had 8.5% lower baseline FVC, compared to those who did not develop PH (p=0.008). Over time, FVC decreased by 0.4 % per year in PH while in non-PH subjects it increased by 0.6% on average (p<0.001). As expected, the rate of change in FVC was influenced by a number of other clinical characteristics (Table 1.). Even after adjustment, baseline FVC was on average 7.1% lower in PH subjects compared to non-PH (p=0.014). While in non-PH patients FVC increased by about 1% per year, in those who developed PH, the increase was <0.1% (p=0.003).

Similarly, unadjusted analysis demonstrated that PH subjects had on average 12.6% lower baseline DLCO, compared to non-PH (p<0.001). In PH patients, DLCO decreased by 2.6% per year, while in non-PH, the yearly decrease was 0.8% (p<0.001). When adjusting for other covariates, PH was still associated with a substantially lower baseline DLCO (11.2% difference, p<0.001) and this decreased by 1.9% per year compared to 0.2% (p<0.001) in non-PH patients, when keeping all other covariates constant (Table 1).

Conclusion:

Our findings confirm the particular value of serial measurements of lung function in the identification of patients at risk of PH. Subjects who develop PH have lower baseline FVC and DLCO levels with increased rates of DLCO reduction over time.

« Back to 2015 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/modelling-of-longitudinal-changes-in-lung-function-in-patients-with-systemic-sclerosis-and-their-association-with-development-of-pulmonary-hypertension/