Multiplex serum protein analysis reveals potential mechanisms and markers of response to hyperimmune caprine serum in systemic sclerosis

Abstract

Background: Hyperimmune caprine serum (HICS) is a novel biological therapy with potential benefit for skin in established diffuse cutaneous systemic sclerosis. Here we report multiplex protein analysis of blood samples from a placebo-controlled phase II clinical trial and explore mechanisms of action and markers of response.

Methods: Patients were treated with HICS (n = 10) or placebo (n = 10) over 26 weeks, with follow-up open-label treatment to 52 weeks in 14 patients. Serum or plasma samples at baseline, 26 and 52 weeks were analysed using multiplex or individual immunoassays for 41 proteins. Patterns of change were analysed by clustering using Netwalker 1.0, Pearson coefficient and significance analysis of microarrays (SAM) correction.

Results: Cluster analysis, SAM multiplex testing and paired comparison of individual analytes identified proteins that were upregulated or downregulated during treatment with HICS. There was upregulation of the hypothalamo-pituitary-adrenal axis after HICS treatment evidenced by increases in α-MSH and ACTH in cases treated with HICS. Interestingly, significant increase in PIIINP was associated with HICS treatment and improved MRSS suggesting that this may be a marker of extracellular matrix turnover. Other relevant factors reduced in HICS-treated patients compared with controls, although not reaching statistical significance included COMP, CCL2, IL6, TIMP2, Fractalkine and TGFβ1 levels.

Conclusions: Our results suggest mechanisms of action for HICS, including upregulation of α-MSH, that has been shown to be anti-fibrotic in preclinical models, and possible markers to be included in future trials targeting skin in diffuse cutaneous systemic sclerosis.

Trial registration: Eudract, No. 2007-003122-24. ClinTrials.gov, No. NCT00769028 . Registered 7 October 2008.

Keywords: Biomarker; Clinical trial; Goat serum; Melanocortin; Scleroderma.

Figures

Fig. 1

Change in modified Rodnan skin score after 26 weeks of treatment with hyperimmune caprine serum compared with placebo or no treatment. In the extended clinical trial dataset patients received 26 weeks of therapy with HICS or placebo. Seven subjects that had received placebo were then treated with HICS. For all subjects there was a 52-week visit and so MRSS between 26 and 52 weeks was available for placebo-treated cases, providing an extended dataset of 17 subjects receiving HICS and 13 with no active treatment. Treatment blind for the first 26 weeks was maintained to week 52. There was statistically significant difference between these two treatment groups. These data extended the dataset from the placebo-controlled phase of the study that had demonstrated a trend of improvement for MRSS and for responder frequency defined by improvement of at least four skin score units and 20% of baseline MRSS. The placebo-controlled phase has been previously described [11]. HICS hyperimmune caprine serum, MRSS modified Rodnan skin score

Fig. 2

Unsupervised hierarchical cluster analysis for multiplex serum proteins at baseline and 26 weeks during the placebo-controlled trial of HICS. Unsupervised cluster analysis was undertaken to identify any subgroups within the study cohort at baseline based upon the serum levels of multiple protein analytes as described in text. The same analysis was repeated for serum samples after 26 weeks of treatment with HICS or placebo. The randomly assigned treatment allocation is shown for each subject. Data for baseline samples are shown in panel a. After 26 weeks of treatment there were clear changes in the patterns of protein analytes that were spread between the two treatment arms as shown in panel b. HICS hyperimmune caprine serum

Fig. 3

Supervised analysis for change in serum proteins according to treatment or responder status in the 26-week placebo-controlled trial of HICS. a Cluster analysis was undertaken for serum analytes after allocation to HICS or placebo during the 26-week parallel-group controlled phase of the clinical trial to identify any subgroups within the study cohort at baseline based upon the serum levels of multiple protein analytes as described in text. A pattern of upregulation or downregulation of specific proteins is seen more frequently in cases treated with HICS than controls. b The same analysis was repeated with subjects allocated to responder or non-responder categories based upon improvement in MRSS by at least four skin score units and 20% of baseline MRSS. Of the responders 5/6 were in the HICS-treated arm of the study. In this way groups of proteins that are clustered for similar patterns of change in response to HICS or associated with clinically meaningful improvement in MRSS

Fig. 4

Baseline MRSS correlation and change during treatment with HICS for selected proteins. Individual patient data is shown for key serum proteins identified as significantly increasing during the 26-week treatment phase of the placebo-controlled trial. The panels compare changes in the HICS-treated or control extended dataset over 26 weeks that extend the data included in Table 1 for the placebo-controlled phase of the study. Adjacent plots show the change in the extended 52-week dataset for cases that moved from placebo to HICS at 26 weeks. These findings confirm the clear pattern of upregulation for PIIINP (a), αMSH (b), and ACTH (c). There is a trend for increase in bFGF (d) whilst the mean levels for COMP (e), TGFβ1 (f), and MCP1 (g) shows a fall over 26 weeks with HICS treatment. Changes in these key analytes also demonstrate the heterogeneity of the groups consistent with the heat maps shown above, with some subjects being obvious outliers and only some cases showing increase in serum levels after moving from placebo to HICS treatment at 26 weeks. HICS hyperimmune caprine serum