Extracellular Vesicle miR-122-5p as a Prognostic Biomarker in Pediatric Classical Hodgkin Lymphoma

Abstract

Currently, risk stratification for pediatric Hodgkin lymphoma is based on clinical factors such as stage, bulk, and systemic symptoms. Novel minimally invasive biomarkers could enhance both prognosis and treatment strategies. Therefore, the plasma extracellular vesicles' microRNA profile was characterized by small RNA sequencing in 36 classical Hodgkin lymphoma cases and these findings were confirmed in an extended cohort of 86 patients by RT-qPCR. It was found that the levels of miR-122-5p at diagnosis were significantly higher (p-value: 0.0002) in patients who relapsed compared to patients in remission. The 5-year event-free survival of cases with high and low levels of miR-122-5p was 65 ± 7% and 93 ± 4%, respectively. MiR-122-5p levels were significantly associated with clinical events in both univariate (p-value: 0.0009) and multivariate (p-value: 0.0037) analysis (hazard ratio 5.8). Target prediction analysis suggests an involvement in the polarization of immune cells. The phenotypic characterization of peripheral blood mononuclear cells in 12 patients showed significantly increased levels of CD4+ T-cells in cases with high miR-122-5p levels as compared to low levels (p-value: 0.048). Moreover, CCL17 (TARC) and IL-6 plasma levels at diagnosis were significantly higher as compared to healthy donors (p-value: ≤0.0001). MiR-122-5p could complement current prognostic assays to identify patients at high risk of relapse.

Keywords: Hodgkin lymphoma; T-cells; biomarkers; diagnostics; extracellular vesicles; immune escape; inflammation; liquid biopsy; miRNA; relapse.

Figure 1

cHL patients that later relapsed (REL) have a specific microRNA (miRNA) signature in plasma-derived small extracellular vesicles (EVs) at diagnosis. Heatmap representing differentially abundant miRNAs at diagnosis between REL and complete remission (CR) cHL patients with miR-122-5p is the most upregulated miRNA in the cHL cohort compared between REL and CR. Total RNA from plasma sEVs of 36 cHL patients (CR = 25, REL = 11) was sequenced.

Figure 2

RT-qPCR analysis of miR-122-5p in 86 pediatric cHL plasma samples, 27 pediatric healthy donor (HD) plasma samples, 15 primary tumor biopsies from pediatric cHL patients and 3 cHL cell lines. (A,B) Outliers (1 HD and 14 cHL) were removed by the ROUT (Q = 1%) method. At DIA miR-122-5p expression is not significantly different in plasma small extracellular vesicles (sEVs) of cHL patients compared to HDs. Unpaired student’s two-tailed t-test analysis of cHL vs. HD: not significant (p-value: 0.0836, ROC = 0.6149, 95% CI: 0.502–0.7277). (C,D) Outliers (14 CR and 3 REL) were removed by the ROUT (Q = 1%) method. At DIA miR-122-5p expression is significantly different in plasma sEVs of CR compared to relapsed (REL) patients. Unpaired student’s two-tailed t-test analysis of CR vs. REL: significant (p-value: 0.0002, ROC = 0.8065, 95% CI: 0.6495–0.9635). (E) In the cHL cell lines KM-H2, L-428, and L-540 (green) and primary tumor biopsies (red) at DIA the miR-122-5p is barely or not detectable. Data are expressed in triplicates as cycle threshold (CT) with CT 40 being the limit of detection.

Figure 3

miR-122-5p has prognostic potential based on the stratification into miR-122 low and miR-122 high according to the median expression values. (A) The 5-year event-free survival (EFS) according to patient stratification into miR-122 low and miR-122 high. The miR-122 low and miR-122 high cohorts contain 43 patients, respectively. In the miR-122 low cohort, 3 patients experienced an event with a 5-year EFS of 93%. The miR-122 high cohort experienced 15 events with one death with a 5-year EFS of 65%. (B) Dot plot representing EFS with a 5-year cut-off of 86 cHL patients (CR = 68, REL = 18) stratified into miR-122 high and miR-122 low in plasma small extracellular vesicles (sEVs) at DIA. Events are marked in red; death is marked in blue. The 5-year event-free survival was significantly different between the miR-122 high and miR-122 low groups (p-value: 0.0016, **). (C) ROC curves of miR-122 low and miR-122 high as a control (AUC 0.679, p-value 0.0043, 95% CI: 0.5649–0.7931).

Figure 4

Flow cytometric characterization of peripheral blood mononuclear cells (PBMCs) from 12 pediatric cHL patients stratified into miR-122 high and miR-122 low based on the median expression value in plasmatic small extracellular vesicles (sEVs). Data were analyzed with GraphPad Prism (Version 8.0.2., GraphPad Software, Boston, MA, USA) using a Mann–Whitney test (confidence interval 95%, * p ≤ 0.05). (A) Percentage of T-cells characterized by CD3 + CD45+, (p-value: 0.1061). (B) Percentage of B-cells characterized by CD19 + CD45+, (p-value: 0.048). (C) Percentage of myeloid cells characterized by CD33 + CD45+, (p-value: 0.1061). (D) Percentage of CD4+ T-cells characterized by CD3 + CD45 + CD4+, (p-value: 0.048). (E) Percentage of CD8+ T-cells characterized by CD3 + CD45 + CD8+, (p-value: 0.202). (F) Ratio of CD4+/CD8+ T-cells, (p-value: 0.3434).

Figure 5

Plasma CCL17 (TARC) and IL-6 levels measured by ELISA. Data were analyzed with GraphPad Prism (Version 8.0.2., GraphPad Software, Boston, MA, USA) using a Mann–Whitney test (confidence interval 95%, **** p ≤ 0.0001). (A) CCL17 plasma levels at day of diagnosis (DIA) in cHL are significantly higher than in healthy donors (HDs) (p-value: ≤0.0001, ****). (B) CCL17 plasma levels at follow-up in HDs and cHL are not significantly different (p-value: 0.1253, ns). (C) IL-6 plasma levels at DIA in cHL are significantly higher than in HDs (p-value: ≤0.0001, ****).