Assessment of treatment outcome using ASAS40 with etanercept in non-radiographic axSpA by baseline CRP – a re-analysis of trial data

Lead Investigator: Philip Robinson, University of Queensland
Title of Proposal Research: Assessment of treatment outcome using ASAS40 with etanercept in non-radiographic axSpA by baseline CRP – a re-analysis of trial data
Vivli Data Request: 5073
Funding Source: None
Potential Conflicts of Interest: In the last 2 years I have provided consulting to Pfizer on one occasion, in the form of one 1 hour teleconference, for which I was paid an honorarium. Prior to that I gave my paid primary care physician talk for Pfizer in April 2017. This project, whilst investigating a Pfizer produced drug is not connected in any way to Pfizer, and is a study I have initiated with a colleague to better understand the response predictors of etanercept in nr-axSpA. I am also doing this analysis across all biologics trialled in nr-axSpA.

Summary of the Proposed Research:

Non-radiographic axial spondyloarthritis (nr-axSpA) is an inflammatory arthritis affecting the spine and sacroiliac joints. Effective therapies are available, including etanercept, which is part of the anti-tumour necrosis factor (TNF) class. Previous work has demonstrated that predictors of response to anti-TNF agents include a raised C-reactive protein (CRP) (1-4). However, the level of raised CRP that predicts a response to etanercept has not been determined. This project is significant because often patients with nr-axSpA have low positive CRP levels and the level at which patients will respond to anti-TNFs like etanercept is not known. This work has the potential to provide significant guidance to practising clinicians on when it is best to commence agents like etanercept with the expectation that they will respond. Therefore, this work can provide guidance to clinicians on when etanercept use would be of limited value and when it would be of good value to the patient (and secondarily to the payer).

Statistical Analysis Plan:

Outcome measures – Primary = (1) ASAS40 (defined below); Secondary = (2) ASAS20, (3) ASDAS (Ankylosing Spondylitis Disease Activity Score) < 1.3 (ie. ASDAS inactive disease), (4) Bath ankylosing spondylitis disease activity index (BASDAI)-50 and (4) The change from baseline to week 12 in inflammation of the spine based on the Spondyloarthritis Research Consortium of Canada (SPARCC) MRI score.
The ASAS response is already calculated as an outcome measure for the study. We will divide the groups by their CRP levels and assess the proportion who reach the ASAS40 response. We will also construct a curve of baseline CRP vs ASAS40 response.
The descriptive analysis will be divided into two parts:
Descriptive 1: Simply ascertaining the ASAS40 response rates by different categories of baseline CRP
Category 1: CRP <5 mg/L Category 2: CRP 5-10mg/L Category 3: > 5mg/L, Category 4: >10mg/L
Descriptive 2: Plotting as a continuous measure the ASAS40 response vs baseline CRP.
Analytical 1: Logistic regression will examine the relationship of continuous CRP to outcome measures that are bivariate, Linear regression will examine the relationship between continuous CRP and continuous outcome measures (eg. change in MRI SIJ SPARCC score).
All variables were defined in the original study. The ASAS response criteria are used to assess improvement in axial spondyloarthritis in clinical trials. Each of four domains is scored by the patient on a visual analog scale ranging from 0 to 10. The four domains are as follows:
1. Patient global assessment of disease activity for the past week
2. Patient assessment of back over the past week
3. Function (BASFI)
4. Inflammation (severity and duration of morning stiffness)
An ASAS40 response is defined as an improvement of at least 40% and an absolute improvement of at least 1 unit (on a 0-10 scale) in at least three of four domains, with no worsening of the remaining domain. In each CRP group, there will be a proportion who reach the ASAS40 response, this is the main outcome measure. ASAS partial remission, BASDAI50, Change in SPARCC MRI SI score and change in BASDAI are measured and defined in the original study and I will be simply examining their relationship to dichotomous CRP groups and CRP as a continuous measure.

ASAS40 responses are defined as a composite (see above). Once each patient has had their ASAS response calculated (see above), then each CRP group will have a proportion that reach an ASAS40 response, this is the primary outcome measure. The process will be completed for each outcome measure, ie. in each CRP subgroup the outcome measures will be examined, and descriptive analyses generated to try to make a comparison between different CRP subgroups. The study was not designed or powered to look at these subgroups so making formal statistical comparisons may be of limited value but logistic regression analysis will be used to examine for the relationship of CRP to these outcome measures.

Requested Studies:

Sponsor: Pfizer, Inc.
Study ID: NCT01258738

Public Disclosure:

Tam, H.K.J., Nash, P. and Robinson, P.C. (2021), The Effect of Etanercept in Nonradiographic Axial Spondyloarthritis by Stratified C-Reactive Protein Levels. ACR Open Rheumatology.