Inhibition of type 1 immunity with tofacitinib is associated with marked improvement in longstanding sarcoidosis

Abstract

Sarcoidosis is an idiopathic inflammatory disorder that is commonly treated with glucocorticoids. An imprecise understanding of the immunologic changes underlying sarcoidosis has limited therapeutic progress. Here in this open-label trial (NCT03910543), 10 patients with cutaneous sarcoidosis are treated with tofacitinib, a Janus kinase inhibitor. The primary outcome is the change in the cutaneous sarcoidosis activity and morphology instrument (CSAMI) activity score after 6 months of treatment. Secondary outcomes included change in internal organ involvement, molecular parameters, and safety. All patients experience improvement in their skin with 6 patients showing a complete response. Improvement in internal organ involvement is also observed. CD4⁺ T cell-derived IFN-γ is identified as a central cytokine mediator of macrophage activation in sarcoidosis. Additional type 1 cytokines produced by distinct cell types, including IL-6, IL-12, IL-15 and GM-CSF, also associate with pathogenesis. Suppression of the activity of these cytokines, especially IFN-γ, correlates with clinical improvement. Our results thus show that tofacitinib treatment is associated with improved sarcoidosis symptoms, and predominantly acts by inhibiting type 1 immunity.

Fig. 1. Tofacitinib treatment leads to improvement in cutaneous sarcoidosis.

a Left panel: baseline treatment regimens for each patient, HCQ: hydroxychloroquine (dose shown as mg twice daily), MTX: methotrexate (mg weekly), Pred: prednisone (mg daily), and Tofa: tofacitinib, Dur: duration of most recent therapeutic regimen; patients taking prednisone often had been on and off for significantly longer than indicated. *Treatment with prednisone or methotrexate was recommended but declined by patients due to prior adverse effects with these medications. Scad: maximum previous Scadding stage. b CSAMI activity scores and c total lesion glycolysis (TLG) over the study period. d Treatment regimens at the end of the study (6 months), tofacitinib (Tofa) dose shown as mg twice daily, blue: discontinued/reduced dose during study, **prednisone increased due to worsening pre-existing Achilles tendinopathy, not worsening of sarcoidosis. e Dermal papules/plaques and subcutaneous nodules of sarcoidosis before and after treatment. f Extensive involvement of the face before and after 6 months of treatment; scarring from the longstanding (>20 years) lesions persisted. g Sarcoid dactylitis before and after 6 months of treatment. h Lupus pernio presentation of sarcoidosis, also with annular plaque on the cheek, before and after treatment; significant scarring persisted in this patient. i Lupus pernio presentation of sarcoidosis before and after treatment. j Nail dystrophy related to sarcoidal inflammation in the nail matrix (demonstrated by matrix biopsy) resolved after 6 months of treatment. Post-inflammatory hyperpigmentation persisted. k Papules and plaques on the forearm of a patient before and after treatment. l Hematoxylin and eosin (H&E) stained and CD68 immunohistochemistry on skin biopsies from a representative complete responder (Pt 7, both from the back) and partial responder (Pt 5, both from the arm), scale bar: 150 μM. Similar results were seen in two additional partial responders in whom biopsies were performed. Source data are provided as a Data Source File.

Fig. 2. Tofacitinib treatment leads to improvement in pulmonary and myocardial sarcoidosis.

a, b PET (coronal) and PET-CT (axial) studies before and after 6 months of tofacitinib in patients with complete or near complete internal organ response, TLG: total lesional glycolysis. c PET (coronal) studies before and after 6 months of treatment in a patient with improvement in lymph node avidity below the threshold of quantification. d PET (coronal) studies before and after 6 months of treatment in a patient with a slight increase in PET avidity which was felt to be clinically insignificant. e Cardiac PET-CT (coronal) before and after 6 months of treatment in a patient myocardial involvement of the inferior intraventricular septum (arrow), CMA: cardiac metabolic activity. f Scatterplot comparing change in cutaneous sarcoidosis and extra-cutaneous sarcoidosis. Depicted as simple linear regression line with 95% confidence interval (shared area), Goodness of Fit represented by R squared. Cutaneous involvement shown as percent reduction in CSAMI during the study period. Extracutaneous involvement shown as percent reduction in TLG during the study period; for patients with increase in TLG during the study, worsening of 25% was arbitrarily assigned. Source data are provided as a Data Source File.

Fig. 3. scRNAseq of cutaneous sarcoidosis and control skin samples.

a UMAP projection of scRNA-seq data showing clustering of all cells, colored by cell type. b UMAP projection of scRNA-seq data in (a), colored by condition/library. c Histograms showing the contribution of each library/condition to each cluster. NK natural killer cell, L-endo lymphatic endothelium.

Fig. 4. IFN-γ produced by Th1 polarized CD4⁺ T cells activates macrophages in cutaneous sarcoidosis.

a UMAP projection of scRNA-seq data showing T cell clusters in sarcoidosis (shades of red) compared to healthy controls (grey). b Histograms showing contribution of each condition (sarcoidosis: shades of red, grey: controls) to each T cell cluster. c Volcano plot showing the most differentially expressed genes between CD4⁺ SAR-1 (clusters 2,7,12) versus CD4⁺ CTRL (clusters 0,5), corresponding to Fig. 4a. p value determined using Wilcon Rank-Sum test, two-tailed. d Heatmap showing expression of selected transcripts in CD4⁺FOXP3^- T cell clusters. e Histogram showing selected predicted upstream regulators of CD4⁺ SAR-1 clusters (2,7,12) versus CD4⁺ CTRL clusters (0, 5) as determined by IPA. Significance cutoff of p < 0.001 is shown by a dotted horizontal line and determined using Fisher exact test, right-tailed. f UMAP projection of scRNA-seq data showing myeloid cell clusters in sarcoidosis (shades of red) compared to healthy controls (grey). g Histograms showing contribution of each condition (sarcoidosis: shades of red, grey: controls) to each myeloid cluster. h Volcano plot showing the most differentially expressed genes between Mac SAR-1 (clusters 0,1,4,6,9,10) versus control myeloid (clusters 2,8,13). p value determined using Wilcon Rank-Sum test, two-tailed. i Heatmap showing expression of selected transcripts in myeloid clusters. j Histogram showing selected predicted upstream regulators in Mac SAR-1 and Mac SAR-2 (clusters 0,1,3,4,6,7,9,10,11) versus myeloid cells in controls (clusters 2,8,13) as determined by IPA. Significance cutoff of p < 0.001 is shown by a dotted horizontal line, as determined using Fisher exact test, right-tailed.

Fig. 5. Analysis of scRNAseq data of bronchoalveolar lavage (BAL).

Patients with sarcoidosis (n = 4) and healthy control patients (n = 10) were included. a UMAP projection of scRNA-seq data of all cells colored by condition (red: sarcoidosis libraries, grey: control libraries). b UMAP projection of scRNA-seq data of all cells colored by cell type. c UMAP projection of scRNAseq data of all cells showing relative CD3E expression. d UMAP projection of scRNAseq data of T cell clusters colored by condition (red: sarcoidosis libraries, grey: control libraries). e Histogram showing the contribution of each condition (sarcoidosis: red, grey: controls) to each T cell cluster. f Violin plots showing expression of select genes in each cluster of CD4⁺FOXP3^- cluster of T cells.

Fig. 6. IFN-γ signaling is a hallmark of sarcoidosis.

a IPA analysis showing predicted upstream regulators in cutaneous sarcoidosis dataset A (bulk RNAseq from this study) and cutaneous sarcoidosis dataset B (Judson et al); IPA content version: 51963813). Each data set consisted of a series of cutaneous sarcoidosis biopsies and normal control skin. b Box and whisker plots showing expression of selected genes in pulmonary sarcoidosis with Scadding stage 1 (n = 24), stage 2/3 (n = 40), stage 4 (n = 12) relative to healthy controls (C) (n = 6). Box plots indicate median (middle line), 25th, 75th percentile (box), and 5th and 95th percentile (whiskers), p values calculated using unpaired t tests, NS: not significant, ND: not determined (abundance below threshold of detection), FPKM: fragments per kilobase of exon per million mapped fragments. See also Supplementary Fig. 5. c Box and whisker plots showing expression of selected genes in various populations of FACS purified myeloid cells from BAL of healthy controls (n = 9) (Cont) vs sarcoidosis (n = 8) (Sar). Box plots indicate median (middle line), 25th, 75th percentile (box), and 5th and 95th percentile (whiskers), p values calculated using unpaired t tests, NS: not significant, cDC1 and 2: classical dendritic cell types 1 and 2, AM: alveolar macrophage, CM: classical monocyte, IM: intermediate monocyte. (See also Supplementary Fig. 6). Source data are provided as a Data Source File.

Fig. 7. Type 1 immunity is the predominant immune polarization in sarcoidosis.

a Photomicrographs showing representative RNA in situ hybridization staining patterns for selected markers (red chromogen) with hematoxylin counterstain (blue), scale bar: 50 μM, higher power inset. b Histograms showing quantification of RNA in situ hybridization staining for selected markers in control skin (n = 10), control lung (n = 5), cutaneous sarcoidosis (n = 10) and pulmonary sarcoidosis (n = 10) tissue. Data are presented as mean + /− 95% confidence interval, p values calculated using unpaired t tests, NS: not significant, Pos: positive control, Psor: psoriasis biopsies (n = 20), Atopic D: Atopic dermatitis biopsies (n = 26). Source data are provided as a Data Source File.

Fig. 8. Cell type specific receptor-ligand analysis and summary in sarcoidosis.

a Pie chart showing relative proportions of IL15 and IL6 producing cells in sarcoidosis scRNA-seq data from skin (see also Supplementary Fig. 7). b Dot plot of cellphone DB receptor-ligand interaction analysis for select receptor-ligand pairs, p value as determined using Cellphone DB script. c Summary of central cytokine and chemokine signals in sarcoidosis revealed by scRNA-seq and other experiments.

Fig. 9. Tofacitinib reduces cytokine and chemokine levels and markers of macrophage activation in plasma.

a Heatmap showing expression of selected transcripts from bulk RNAseq of skin; sample labels are summarized in Supplementary Table 5. b Volcano plot showing relative abundance of proteins in plasma from sarcoidosis patients in this trial compared to healthy controls, p values determined using t-tests, two-tailed. c Relative abundance of plasma proteins in sarcoidosis patients compared to controls (x-axis) compared to mRNA levels in skin of sarcoidosis patients (as determined by bulk-RNA-seq) compared to healthy controls (y-axis). Simple regression line with 95% confidence interval (shaded area) shown, goodness of fit calculated with R squared. d Relative abundance of proteins in plasma of sarcoidosis patients compared to healthy controls (x-axis) plotted against relative abundance of proteins in plasma of sarcoidosis patients at baseline (pre-tx) compared to after 6 months of tofacitinib (post-tx) (y-axis). The most consistently and highly differentially abundant proteins were used to create the sarcoidosis plasma signature (SPS) shown in the shaded box. e Heatmap showing plasma levels of SPS proteins in trial participants (before and after tofacitinib) and healthy controls, grouped by response pattern. Best responders include patients with a complete response in the skin (CSAMI 0 on treatment) and a complete or near complete response in other organs (98% or greater reduction in TLG). Partial responders include all other patients. f Clinical photographs of scalp and scrotal (biopsy-proven) sarcoidosis. Shown at baseline, after 6 months of tofacitinib at 5 mg twice daily, and after 12 months of tofacitinib 10 mg in the morning and 5 mg at night. g PET scans of the patient in panel f before tofacitinib, after 6 months at 5 mg twice daily, and after 14 months at 10 mg in the morning and 5 mg at night, TLG: total lesion glycolysis. h SPS levels corresponding to timepoints in (panel f).