Patient-Level Evaluation of Components of the American College of Rheumatology Combined Response Index in Systemic Sclerosis (CRISS) Using Patient-Reported Anchors

Vivek Nagaraja¹, John Powers², Celia J. F. Lin³, Benjamin Brennan⁴, Veronica J. Berrocal⁵ and Dinesh Khanna¹, ¹Division of Rheumatology, Department of Internal Medicine, University of Michigan Scleroderma Program, University of Michigan, Ann Arbor, MI, ²University of Maryland School of Medicine, Rockville, MD, ³Genentech, Inc., South San Francisco, CA, ⁴School of Public Health, University of Michigan, Ann Arbor, MI, ⁵Div of Rheumatology, University of Michigan, Ann Arbor, MI

Meeting: 2018 ACR/ARHP Annual Meeting

Keywords: Health Assessment Questionnaire, Outcome measures, patient health, patient outcomes and systemic sclerosis

Session Information

Date: Tuesday, October 23, 2018

Title: Systemic Sclerosis and Related Disorders – Clinical Poster III

Session Type: ACR Poster Session C

Session Time: 9:00AM-11:00AM

Background/Purpose: Treatment benefit is demonstrated by evidence that interventions have positive impacts on how patients feel, function, and/or survive (FDA Guidance, 21CFR314.510). The ACR CRISS, a composite endpoint for trials in systemic sclerosis (SSc), uses a weighted score combining outcome assessments that directly measure how patients feel and function [HAQ-DI and patient (PGA) global assessment] with outcome assessments that are indirect measures of patient symptoms/function [FVC, modified Rodnan skin score (mRSS), and physician global assessments (MDGA)]. Understanding the relationship and magnitude of effects on these indirect assessments would provide confidence that each component of CRISS would reliably predict an effect on direct measures of patient benefit. Our objective was to provide data to support evaluation of CRISS using patient-reported outcome (PRO) anchors – HAQ-DI and PGA.

Methods: We evaluated 2 cohorts: an early diffuse cutaneous SSc (dcSSc) cohort that was used for development of ACR CRISS (1) and a phase 2 trial of TCZ vs. placebo in dcSSc [faSScinate trial (2)]. Using the early dcSSc cohort, we assessed the effect size (ES) at the patient-level for non-PRO variables (mRSS, MDGA, and FVC%) in those we defined as “responders” who met minimal clinically important differences (MCID) estimates for HAQ-DI (defined as an improvement of ≥ 0.22) and PGA improvement of ≥ 1.0 (range 0-10). We interpreted ES using Cohen’s criteria: < 0.20 = negligible, 0.20-0.49 = small, 0.50-0.79 = medium, >0.80 = large (3). We assessed whether ES in patients who met the responder criteria in HAQ-DI and PGA in the faSScinate trial was associated with larger improvements in the ACR CRISS scores at week 24 and 48.

Results: In the dcSSc cohort, the ES was of greater magnitude for responders vs. non-responders (Table), except for HAQ-DI and FVC% when using PGA as an anchor. In the faSScinate trial, statistically significant improvements in the median ACR CRISS scores were seen in those who attained MCID vs. patients who did not (Table).

Conclusion: In a dcSSc cohort, generally patients who achieved MCID in HAQ-DI and PGA are associated with larger magnitude of improvement in ACR CRISS non-PRO variables. Limitations of the analysis are: 1. HAQ-DI and PGA are part of the ACR CRISS score, and 2. Assessing the relationships in an observational cohort. Ongoing trials should confirm the relationships between non-PRO variables (mRSS, MDGA, and FVC%) vs. PRO anchors.

References:

Khanna, D. Arthritis & Rheumatology
Khanna D. Lancet. 2016
Psychol Bull.1992

Role of the study sponsor: F Hoffmann-La Roche Ltd funded the study and was involved in writing the abstract.

Table: Change in the CRISS variables in improvers and non-improvers based on HAQ-DI and PGA MCID estimates

	Patients with HAQ-DI ≥ 0.22 (MCID) Responders	Patients with HAQ-DI <0.22 Non-responders	P value	Patients with PGA ≥ 1 (MCID)	Patients with PGA<1	P value
Early diffuse SSc cohort
MRSS, ES	– 0.70, N= 27	– 0.30, N= 84	0.06	– 0.65, N= 37	– 0.25, N=74	0.03
FVC%, ES	0.20, N= 24	– 0.07, N= 87	0.002	0.11, N= 32	– 0.07, N=79	0.04
PGA, ES	– 0.29, N= 28	0.03, N= 69	0.13	NA	NA
MDGA, ES	– 0.43, N= 23	– 0.06, N= 72	0.30	– 0.31, N=31	-0.06, N=64	0.40
HAQ-DI, ES	NA	NA		– 0.14, N=37	0.10, N=62	0.16
Data from faSScinate trial
CRISS Score at 24 week, median	0.381, N=18	0.002, N=49	0.045	0.229, N=21	0.018, N=46	0.02
CRISS Score at 48 week, median	0.947, N=15	0.011, N=43	<0.001	0.705, N=22	0.018, N=36	0.01

Disclosure: V. Nagaraja, None; J. Powers, Corbus Pharmaceuticals Holdings, 5; C. J. F. Lin, Genentech, Inc., 1, 3; B. Brennan, None; V. J. Berrocal, None; D. Khanna, Eicos Sciences, 1,Pfizer, Inc., 2,Horizon, 2,BMS, 2,Actelion, 5,Bayer, 5,Bayer, 2,Corbus, 5,Cytori, 5,EMD Serono, 5,Genentech, Inc., 5,Sanofi-Aventis, 5,GSK, 5,Boehringer Ingelheim, 5,Civi BioPharma, 3.

To cite this abstract in AMA style:

Nagaraja V, Powers J, Lin CJF, Brennan B, Berrocal VJ, Khanna D. Patient-Level Evaluation of Components of the American College of Rheumatology Combined Response Index in Systemic Sclerosis (CRISS) Using Patient-Reported Anchors [abstract]. Arthritis Rheumatol. 2018; 70 (suppl 9). https://acrabstracts.org/abstract/patient-level-evaluation-of-components-of-the-american-college-of-rheumatology-combined-response-index-in-systemic-sclerosis-criss-using-patient-reported-anchors/. Accessed .

« Back to 2018 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/patient-level-evaluation-of-components-of-the-american-college-of-rheumatology-combined-response-index-in-systemic-sclerosis-criss-using-patient-reported-anchors/