Surrogate Measures of Extent of Interstitial Lung Disease As Measured By Quantitative Radiographic Analysis in Patients with Systemic Sclerosis

Elizabeth Volkmann¹, Donald Tashkin², Chi-hong Tseng¹, Kim Hyun³, Jonathan Goldin³, Philip J. Clements⁴, Daniel E. Furst¹, Dinesh Khanna⁵, Eric Kleerup¹, Michael Roth¹ and Robert Elashoff⁶, ¹Medicine, University of California, Los Angeles, David Geffen School of Medicine, Los Angeles, CA, ²Medicine, University of California at Los Angeles, Los Angeles, CA, ³Radiology, University of California, Los Angeles, David Geffen School of Medicine, Los Angeles, CA, ⁴University of California, Los Angeles, Department of Medicine, Los Angeles, CA, ⁵University of Michigan Health System, Ann Arbor, MI, ⁶Biomathematics, University of California, Los Angeles, Los Angeles, CA

Meeting: 2014 ACR/ARHP Annual Meeting

Keywords: Imaging, interstitial lung disease and systemic sclerosis

Session Information

Title: Systemic Sclerosis, Fibrosing Syndromes and Raynaud's - Clinical Aspects and Therapeutics III: Updates in Predictors and Outcomes in Systemic Sclerosis

Session Type: Abstract Submissions (ACR)

Background/Purpose: Extent of systemic sclerosis (SSc)-related interstitial lung disease (ILD) predicts disease course, mortality and treatment response. While quantitative analyses of total extent of ILD (QILD) are more sensitive and reproducible than visual assessments of SSc-ILD, these analyses are not widely available. This study evaluates the relationship between disease parameters and QILD scores to identify potential surrogate measures of QILD.

Methods : Using baseline data from the Scleroderma Lung Study I (SLS I) (N=158), multivariate regression analyses were performed using the best subset selection method to identify 1 to 5 variable-models that best predict QILD scores in both whole lung (WL) and the zone of maximal involvement (ZM). These models were subsequently validated using baseline data from SLS II (N=142). SLS I&II did not include patients with clinically significant pulmonary hypertension (PH).

Results: Diffusing capacity for carbon monoxide (DLCO) was the single best predictor of QILD in the WL and ZM in all of the best subset models (Tables 1, 2). Adding other disease parameters to the models did not substantially improve model performance. Forced vital capacity (FVC) did not predict QILD scores in any of the models.

Conclusion: In the absence of PH, DLCO provides the best overall estimate of HRCT-measured QILD in patients from 2 large SSc cohorts. FVC, which is commonly used to monitor disease course in SSc-ILD, may not be the best surrogate measure of extent of QILD.

Table 1. Multivariate regression analyses of the best 1 to 5 variable models that predict extent of quantitative interstitial lung disease (QILD) in the zone of maximal involvement (ZM).

Variable	Estimate	SE	p value	Adjusted r² for SLS I	Correlation SLS I	Correlation SLS II
From best 1- variable model
DLCO% predicted	-0.87	0.13	<0.0001	0.29*	0.55*	0.44*
From best 2-variable model
Diffuse disease	6.70	3.23	0.04	0.32*	0.57*	0.39*
DLCO% predicted	-0.91	0.13	<0.0001	0.32*	0.57*	0.39*
From best 3-variable model
Diffuse disease	6.72	3.21	0.039	0.33*	0.59*	0.44*
DLCO% predicted	-0.75	0.16	<0.0001
TLC% predicted	-0.24	0.16	0.13
From best 4-variable model
Diffuse disease	6.22	3.23	0.057	0.33*	0.60*	0.46*
DLCO% predicted	-0.67	0.18	0.0003
TLC% predicted	-0.28	0.16	0.085
Breathing VAS	0.077	0.066	0.24
From best 5-variable model
Diffuse disease	10.61	4.51	0.021	0.33*	0.60*	0.46*
DLCO% predicted	-0.76	0.17	<0.0001
TLC% predicted	-0.24	0.16	0.14
mRSS^à	-0.28	0.21	0.18
Disease duration^¤	-0.90	0.73	0.22

* p<0.0001.

Visual analog scale for breathing (Range of scores 0-100).

à Modified Rodnan Skin Score (Range of scores 0-51).

¤ Disease duration = Number of years since diagnosis of first non-Raynaud’s symptom to randomization.

Table 2. Multivariate regression analyses of the best 1 to 5 variable models that predict extent of quantitative interstitial lung disease (QILD) in the whole lung (WL).

Variable	Estimate	SE	p value	Adjusted r² for SLS I	Correlation SLS I	Correlation SLS II
From best 1- variable model
DLCO% predicted	-0.59	0.12	<0.0001	0.19*	0.44*	0.37*
From best 2-variable model
DLCO% predicted	-0.49	0.12	0.0001	0.24*	0.50*	0.38*
Breathing VAS	0.15	0.057	0.0081	0.24*	0.50*	0.38*
From best 3-variable model
DLCO% predicted	-0.51	0.12	<0.0001	0.25*	0.52*	0.37*
Breathing VAS	0.15	0.056	0.0067
Age (years)	-0.20	0.12	0.086
From best 4-variable model
DLCO% predicted	-0.54	0.12	<0.0001	0.25*	0.53*	0.38*
Breathing VAS	0.14	0.057	0.010
Age (years)	-0.20	0.12	0.088
Diffuse disease	2.89	2.86	0.32
From best 5-variable model
DLCO% predicted	-0.56	0.12	<0.0001	0.26*	0.54*	0.40*
Breathing VAS	0.15	0.057	0.010
Age (years)	-0.21	0.12	0.085
Diffuse disease	6.29	4.01	0.12
mRSS^à	-0.22	0.18	0.23

* p<0.0001

Visual analog scale for breathing (Range of scores 0-100).

à Modified Rodnan Skin Score (Range of scores 0-51).

Disclosure:

E. Volkmann,
None;

D. Tashkin,
None;

C. H. Tseng,
None;

K. Hyun,
None;

J. Goldin,
None;

P. J. Clements,
None;

D. E. Furst,
None;

D. Khanna,
None;

E. Kleerup,
None;

M. Roth,
None;

R. Elashoff,
None.

« Back to 2014 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/surrogate-measures-of-extent-of-interstitial-lung-disease-as-measured-by-quantitative-radiographic-analysis-in-patients-with-systemic-sclerosis/