Open Access Highly Accessed Research article

Synovial phenotypes in rheumatoid arthritis correlate with response to biologic therapeutics

Glynn Dennis1, Cécile TJ Holweg2, Sarah K Kummerfeld3, David F Choy2, A Francesca Setiadi2, Jason A Hackney3, Peter M Haverty3, Houston Gilbert4, Wei Yu Lin1, Lauri Diehl5, S Fischer6, An Song6, David Musselman7, Micki Klearman7, Cem Gabay8, Arthur Kavanaugh9, Judith Endres10, David A Fox10, Flavius Martin111 and Michael J Townsend2*

  • * Corresponding author: Michael J Townsend townsem1@gene.com

  • † Equal contributors

Author Affiliations

1 Departments of Immunology Discovery, Genentech, South San Francisco, California, USA

2 ITGR Diagnostics Discovery, Genentech, South San Francisco, California, USA

3 Bioinformatics and Computational Biology, Genentech, South San Francisco, California, USA

4 Non-clinical Biostatistics, Genentech, South San Francisco, California, USA

5 Pathology, Genentech, South San Francisco, California, USA

6 Bioanalytical Sciences, Genentech, South San Francisco, California, USA

7 Product Development, Genentech, South San Francisco, California, USA

8 University Hospital of Geneva, Geneva, Switzerland

9 University of California San Diego, San Diego, California, USA

10 Rheumatic Disease Core Center and Division of Rheumatology, Department of Internal Medicine, University of Michigan Medical School, Ann Arbor, Michigan, USA

11 Current address: Inflammation Therapeutic Area, Amgen, 1201 Amgen Court West, Seattle, Washington, USA

For all author emails, please log on.

Arthritis Research & Therapy 2014, 16:R90  doi:10.1186/ar4555

Published: 30 April 2014

Additional files

Additional file 1:

Lists of the Database for Annotation, Visualization and Integrated Discovery (DAVID) biological processes genes represented within the upregulated genes in the synovial subgroups. For each gene, we report differential gene expression between each group and all other samples. We provide the t-statistic values (positive or negative) with associated P-values for each group. L = lymphoid, M = myeloid, X = low inflammatory, F = fibroid.

Format: ZIP Size: 126KB Download file

Open Data

Additional file 2:

Lists of the Database for Annotation, Visualization and Integrated Discovery (DAVID) biological process genes represented within the downregulated genes in the synovial subgroups. For each gene, we report differential gene expression between each group and all other samples. We provide the t-statistic values (positive or negative) with associated P-values for each group. L = lymphoid, M = myeloid, X = low inflammatory, F = fibroid.

Format: ZIP Size: 208KB Download file

Open Data

Additional file 3: Table S1:

List of genes utilized in gene set enrichment analyses.

Format: XLS Size: 153KB Download file

This file can be viewed with: Microsoft Excel Viewer

Open Data

Additional file 4: Figure S1:

Assessment of robustness of synovial gene expression heterogeneity. (A) Principal component analysis showing the first (x-axis) and second (y-axis) components of variation over approximately 7,000 probes and 49 patients using the prcomp R-function on quantile-normalized expression data. Each patient tissue is color-coded according to the groupings in Figure 1A, and grouping circles have been added for visual clarity. (B) Re-sampling analysis using partitioning around medoids (PAM) analysis of approximately 7,000 probes, 49 patients and 5 predefined clusters of tissue samples (k = 5). Heatmap colors represent the frequency with which a pair of samples are found in the same cluster, and are represented as a percentage of the total number of samplings in which the pair was observed. (C) Assessment of cluster robustness via determination of silhouette width of approximately 7,000 clustered probes from the 49 patients. Average silhouette widths for each of the five clusters are indicated.

Format: ZIP Size: 462KB Download file

Open Data

Additional file 5: Figure S2:

Assessment of overlap between biological process gene-sets utilized by the Database for Annotation, Visualization and Integrated Discovery (DAVID) pathway analysis tool for unregulated genes in each of the four synovial clusters defined in Figure 1A. The overlap of genes shared by gene sets are illustrated using a heatmap where each value represents the proportion of genes from the category on the y-axis that are in common with the corresponding gene set on the x axis (indicated by the color bar; 0 = 0%, 1 = 100%). The matrix is not symmetrical because the size of the gene sets is not constant.

Format: ZIP Size: 395KB Download file

Open Data

Additional file 6: Figure S3:

(A) Heatmap visualization of processes enriched in downregulated genes in each of the four synovial clusters defined in Figure 1A using the Database for Annotation, Visualization and Integrated Discovery (DAVID) pathway analysis tool. Colors refer to statistical significance of processes to each cluster. (B) Assessment of overlap between biological process gene sets utilized by the DAVID pathway analysis tool for downregulated genes in each of the four synovial clusters defined in Figure 1A. The overlap of genes shared by gene sets are illustrated using a heatmap where each value represents the proportion of genes from the category on the y-axis that are in common with the corresponding gene set on the x-axis (indicated by the color bar; 0 = 0%, 1 = 100%). The matrix is not symmetrical because the size of the gene sets is not constant.

Format: ZIP Size: 461KB Download file

Open Data

Additional file 7: Figure S4:

B cell, M1 classically activated monocyte, and fibroid gene modules capture synovial tissue transcriptional heterogeneity in additional rheumatoid arthritis (RA) patient cohorts. (A) Scatter plot of the training cohort of 49 patient synovial samples projected in gene set space of the B cell (x-axis) and M1 monocyte (y-axis) biological modules. Samples are colored according to their cluster assignments in Figure 1 (red = lymphoid, purple = myeloid, green = fibroid, grey = low inflammatory). Filled circles indicate samples with histologic aggregates and empty circles indicate samples lacking aggregates. Scatter plot of the same 49 RA patients projected in gene set space of the B cell (x-axis) and M1 monocyte (y-axis) biological modules, and samples are also colored according to their respective fibroid gene set scores as indicated by the color bar. (C) Scatter plot of 33 previously unanalyzed patient samples from a parallel Michigan RA cohort projected in gene-set space of the B cell (x-axis) and M1 monocyte (y-axis) biological modules. Samples are colored according to their respective fibroid gene-set scores as indicated by the color bar. (D) Scatter plot of a publicly available cohort of 62 RA histologically characterized patients (GSE21537) projected in gene-set space of the B cell (x-axis) and M1 monocyte (y-axis) biological modules. Samples are colored according to their respective fibroid gene-set scores as indicated by the color bar.

Format: ZIP Size: 392KB Download file

Open Data

Additional file 8: Figure S5:

CD20 Immunohistochemistry (IHC) correlates with B cell gene-set score in a replication rheumatoid arthritis (RA) patient cohort. Representative CD20 IHC (brown staining) is shown for synovial samples with a high or low B cell gene-set score with low (A, B respectively) and high (C, D respectively) magnification. B cell gene-set scores were also plotted against CD20 IHC scores and the P-value for Spearman rank correlation coefficient is indicated (E).

Format: ZIP Size: 6MB Download file

Open Data

Additional file 9: Figure S6:

Association of pretreatment synovial gene-set scores with good versus poor European League Against Rheumatism (EULAR) response to anti-TNFα (infliximab) therapy at 16 weeks in the GSE21537 synovial expression dataset. Statistical significance for good compared with poor response for the level of each gene-set module was calculated based upon the t-statistic. Scaled gene-set scores for M2 alternatively activated monocytes (A) (P = 0.054), TNFα-stimulated fibroblast-like synoviocytes (B) (P = 0.08), and angiogenesis (C) (P = 0.02) marked with asterisk) are plotted against 16-week EULAR response.

Format: ZIP Size: 369KB Download file

Open Data

Additional file 10: Figure S7:

Receiver-operating-characteristic (ROC) curves to assess the ability of pretreatment synovial phenotypes, defined by scaled gene-set scores, to differentiate between good versus poor European League Against Rheumatism (EULAR) response to anti-TNFα (infliximab) therapy at 16 weeks in the GSE21537 synovial expression dataset. ROC curves were generated for the myeloid (A), lymphoid (B) and fibroid (C) phenotypes, and also for gene sets reflective of M1 classically-activated monocytes (D), B cells (E) and T cells (F). Area under the ROC curve (AUC) is indicated for each plot.

Format: ZIP Size: 326KB Download file

Open Data

Additional file 11: Figure S8:

Biomarker subpopulation treatment effect pattern plot (STEPP) analysis of the ADalimumab ACTemrA (ADACTA) trial. Assessment of individual biomarkers compared with treatment effect. One-dimensional STEPP analysis of week-24 American College of Rheumatology (ACR) 50 relative treatment effectiveness of adalimumab compared with tocilizumab for the serum markers soluble intercellular adhesion molecule 1 (sICAM1) (A) and C-X-C motif chemokine 13 (CXCL13) (B) respectively in the ADACTA trial. Week-24 ACR50 odds ratios are shown in solid blue and 95% CIs as accompanying dashed lines. The x-axes correspond to the subgroup of subjects whose baseline biomarker levels were within 20 percentiles below and above the indicated subpopulation median with actual values (pg/ml) in parentheses. The dotted horizontal line indicates equivalent relative treatment effect. (C) Two-dimensional STEPP analysis for sICAM1 and CXCL13. Each cell of the heatmap corresponds to a subgroup of subjects whose baseline biomarker levels were within 25 percentiles below and above the indicated subpopulation median as defined by each biomarker. Concentrations of each biomarker at the indicated percentage are in parentheses in plot margins. Heatmap colors indicate odds ratio (95% CI in brackets) from logistic regression corresponding to outcomes for adalimumab versus tocilizumab. Counts of subjects in each treatment arm for each subgroup are indicated as n = (tocilizumab)/(adalimumab).

Format: ZIP Size: 353KB Download file

Open Data

Additional file 12: Figure S9:

Receiver-operating-characteristic (ROC) curves to assess the ability of pretreatment C-X-C motif chemokine 13 (CXCL13) and soluble intercellular adhesion molecule 1 (sICAM1) to differentiate for clinical response in the ADalimumab ACTemrA (ADACTA) trial biomarker population. ROC curves were generated for sICAM1 versus achievement of an American College of Rheumatology (ACR)50 response at week 24 for adalimumab in all-comers (A), CXCL13-high (B), and CXCL13-low patient subsets (C), and for CXCL13 versus achievement of an ACR50 response at week 24 for tocilizumab in all-comers (D), sICAM1-high (E) and sICAM1-low patient subsets (F). Biomarker high and low designations were made using their respective medians as the cutoff. Area under the ROC curve (AUC) is indicated for each plot.

Format: ZIP Size: 387KB Download file

Open Data