Dampened Inflammatory Signalling and Myeloid-Derived Suppressor-Like Cell Accumulation Reduces Circulating Monocytic HLA-DR Density and May Associate With Malignancy Risk in Long-Term Renal Transplant Recipients

Abstract

Background: Malignancy is a major cause of morbidity and mortality in transplant recipients. Identification of those at highest risk could facilitate pre-emptive intervention such as reduction of immunosuppression. Reduced circulating monocytic HLA-DR density is a marker of immune depression in the general population and associates with poorer outcome in critical illness. It has recently been used as a safety marker in adoptive cell therapy trials in renal transplantation. Despite its potential as a marker of dampened immune responses, factors that impact upon monocytic HLA-DR density and the long-term clinical sequelae of this have not been assessed in transplant recipients.

Methods: A cohort study of stable long-term renal transplant recipients was undertaken. Serial circulating monocytic HLA-DR density and other leucocyte populations were quantified by flow cytometry. Gene expression of monocytes was performed using the Nanostring nCounter platform, and 13-plex cytokine bead array used to quantify serum concentrations. The primary outcome was malignancy development during one-year follow-up. Risk of malignancy was calculated by univariate and multivariate proportionate hazards modelling with and without adjustment for competing risks.

Results: Monocytic HLA-DR density was stable in long-term renal transplant recipients (n=135) and similar to non-immunosuppressed controls (n=29), though was suppressed in recipients receiving prednisolone. Decreased mHLA-DRd was associated with accumulation of CD14+CD11b+CD33+HLA-DRlo monocytic myeloid-derived suppressor-like cells. Pathway analysis revealed downregulation of pathways relating to cytokine and chemokine signalling in monocytes with low HLA-DR density; however serum concentrations of major cytokines did not differ between these groups. There was an independent increase in malignancy risk during follow-up with decreased HLA-DR density.

Conclusions: Dampened chemokine and cytokine signalling drives a stable reduction in monocytic HLA-DR density in long-term transplant recipients and associates with subsequent malignancy risk. This may function as a novel marker of excess immunosuppression. Further study is needed to understand the mechanism behind this association.

Keywords: HLA-DR; cytokines; gene expression; immunosuppression; kidney transplantation; long term; malignancy; monocyte.

Figure 1

Distribution of HLA-DR binding density in transplant recipients, stratified by gender. The broken line indicates the cut-off for diagnosis of immunodepression as described in (25).

Figure 2

Forest plot of discrete variables at recruitment and their univariable influence upon mHLA-DRd in renal transplant recipients. Densities are recorded as mean and 95% confidence interval, with p values assessed by two-tailed independent t-test or analysis of variance with post-hoc Tukey testing. ‘ns’ not significant; * p<0.05; **p<0.01; ***p<0.001 across all groups (ANOVA or t-test, indicated on right, with post-hoc testing, where appropriate, indicated on left with the vertical line indicating comparison groups). ^#p<0.10.

Figure 3

Monocytic HLA-DR density shows stability over time. mHLA-DRd was calculated at enrolment and repeated a mean of 238 (A) and 385 (B) days later. Spearman’s test was used to assess goodness-of-fit (r²) and significance.

Figure 4

mHLA-DRd correlates inversely with CD19^-CD3^-CD56^-CD14^+/intCD11b⁺CD33⁺HLA-DR^lo monocyte myeloid-derived suppressor-like cell accumulation. (A) Example of gating strategy to delineate monocyte subpopulations. The plots on the far left were derived from live (7AAD-negative), singlet cells. MDS like-cells were derived from a Boolean gate, and were HLA-DRlo, CD11b+CD33+ and of classical, non-classical or intermediate monocyte morphology. Correlation between mHLA-DRd and (B) proportion of mMDSC-like cells within the monocyte population and (C) absolute number of mMDSC-like cells. Spearman’s test was used to assess goodness-of-fit (r²) and significance.

Figure 5

Monocytes from RTR with low mHLA-DRd exhibit suppression of gene sets relating to inflammatory response and chemotaxis. (A) Pathways with significantly altered enrichment in RTR exhibiting low mHLA-DRd, using the Gene Ontology knowledge base. Pathways are listed in order of descending GeneRatio. (B) Differential expression of core genes in ‘Chemokine receptor binding’ and ‘Inflammatory response’ Gene Ontology sets.

Figure 6

Decreased mHLA-DRd is associated with subsequent malignancy development. (A) Receiver-operator characteristic (ROC) curve for prediction of malignancy in the year following mHLA-DRd quantification. Predictive performance is given [area under curve (AUC) and 95% confidence interval] as well as the optimal cut-off and corresponding sensitivity and specificity. (B) Kaplan-Meier curve demonstrating cumulative incidence of malignancy stratified by mHLA-DRd. (unadjusted and adjusted hazard ratios and P values are given in Table 3 ; log-rank test for difference between curves is provided in-plot).