Th17 Lymphocytes In Human Gvhd Target Tissues Are Associated With Disease Severity and Refractoriness To Systemic Glucocorticoid Therapy

Experimental animal data has shown that Th17 cells are pathogenic in acute graft vs. host disease (GVHD) together with Th1, however the relevance of Th17 to human GVHD remains uncertain. We determined whether Th17 cells are present in human GVHD target tissues (skin, GI, liver). Furthermore, we examined whether sirolimus (rapamycin) suppressed Th17 infiltration in GVHD target tissues. Finally, we determined whether infiltrating Th17 cells were associated with response to systemic glucocorticoid therapy for established acute GVHD.

We quantified Th17, Th1, and Treg in target organ biopsies among patients treated on a randomized trial of sirolimus(SIR)/tacrolimus(TAC) vs. methotrexate(MTX)/TAC for initial GVHD prophylaxis. Acute GVHD was graded according to consensus criteria, and pathologic grade was scored by a Pathologist blinded to study assignment. Controls were contemporaneous HCT recipients who had no pathologic evidence of GVHD on diagnostic biopsies. Tissue microarrays were constructed, and stained with conjugated antibodies (RORy, Tbet, FoxP3, CD4) for immunohistochemistry analyses. RORy was utilized to identify Th17, Tbet for Th1, FoxP3 for Treg, and CD4 for total CD4 lymphocytes. Stained slides were scanned, and the absolute number of positive cells per TMA core (area 1.13mm²) for each marker (both individual marker and marker per total CD4) was quantitatively scored through image analysis (TMA module of TissueStudio v3.0 software platform from Definiens). A subset of 10 random tissue cores was selected for co-registration of CD4 and RORy staining on contiguous 4μm sections. We examined the association of pathologic grade, clinical grade, and group (SIR, MTX, control) with Th subset numbers with ANOVA, and therapeutic response to primary GVHD therapy with Th subset numbers using logistic regression analysis.

From 48 patients (25 SIR/TAC, 23 MTX/TAC), 110 tissue biopsies were included (duodenum 30, gastric 27, rectum 34, liver 3, skin 16), all with pathologic and clinical evidence of acute GVHD. Median days from HCT to GVHD biopsy (SIR 21, MTX 27, p = 0.6), and median time between biopsy and initiation of topical (p=0.17) or systemic (p=0.55) therapy did not differ between SIR and MTX groups. From 18 controls, 39 biopsies were included (duodenum 10, gastric 13, rectum 11, liver 0, skin 5). Th17 were detected in human GVHD tissues. Of the tested contiguous sections, a median of 98% of RORy+ cells were double (CD4 and RORy) positive. Among GVHD cases, we detected a significant positive association between pathologic grade and tissue Th17 (p=0.03) and Th17/CD4 (p=0.02) on ANOVA, adjusted for organ site of biopsy. While overall clinical grade was not associated with tissue Th subset numbers, sub-group analysis of GI cases demonstrated significant association between GI stage and both Th17/CD4 (p=0.004) and Treg/CD4 (p=0.016), adjusted for GI organ site. Tissue Th subset numbers differed according to group: Control subjects had greater Th17, Th1, and Treg than the GVHD patients, and there were no significant differences among control patients according to ultimate clinical diagnosis. There were significantly lower Th17 among SIR (median 4) vs. MTX (median 9.5) patient biopsies (p=0.016). Th17 numbers remained significantly lower in the SIR group (p=0.04) in ANOVA after adjustment for clinical and pathologic grade. Refractoriness (no resolution of acute GVHD by 28 days after ≥ 1mg/kg prednisone) to systemic therapy was significantly associated with greater tissue (refractory median 27, responsive median 5) Th17 (OR 6.6, 95% CI 1.6-27, p=0.008), and clinical grade (p=0.019). No association was observed between Th1, Treg, or total CD4 and response to GVHD therapy.

Th17 lymphocytes are present in human GVHD target organs, and are associated with disease severity and resistance to treatment with systemic glucocorticoids. Sirolimus and other targeted interventions to inhibit or deplete Th17 may be a valuable adjunct to approaches that inhibit Th1 in GVHD prevention and control.

No relevant conflicts of interest to declares.