Initial Myeloid Cell Status Is Associated with Clinical Outcomes of Renal Cell Carcinoma

Abstract

The therapeutic outcome of immune checkpoint inhibition (ICI) can be improved through combination treatments with ICI therapy. Myeloid-derived suppressor cells (MDSCs) strongly suppress tumor immunity. MDSCs are a heterogeneous cell population, originating from the unusual differentiation of neutrophils/monocytes induced by environmental factors such as inflammation. The myeloid cell population consists of an indistinguishable mixture of various types of MDSCs and activated neutrophils/monocytes. In this study, we investigated whether the clinical outcomes of ICI therapy could be predicted by estimating the status of the myeloid cells, including MDSCs. Several MDSC indexes, such as glycosylphosphatidylinositol-anchored 80 kD protein (GPI-80), CD16, and latency-associated peptide-1 (LAP-1; transforming growth factor-β1 precursor), were analyzed via flow cytometry using peripheral blood derived from patients with advanced renal cell carcinoma (n = 51) immediately before and during the therapy. Elevated CD16 and LAP-1 expressions after the first treatment were associated with a poor response to ICI therapy. Immediately before ICI therapy, GPI-80 expression in neutrophils was significantly higher in patients with a complete response than in those with disease progression. This is the first study to demonstrate a relationship between the status of the myeloid cells during the initial phase of ICI therapy and clinical outcomes.

Keywords: CD16; glycosylphosphatidylinositol-anchored 80 kD protein; immune checkpoint inhibition; latency-associated peptide-1; myeloid-derived suppressor cells; renal cell carcinoma.

Figure 1

Collection and analysis of peripheral blood during ICI therapy. (a) Schedule of blood sample collection immediately before and during ICI therapy. Blood was collected immediately before each ICI therapy session, and clinical outcomes were evaluated using RECIST version 1.1 after the end of treatment. (b) Representative analysis of neutrophilic and monocytic cell populations. Whole blood cells were stained with CD33, CD16, GPI-80, and LAP-1, as described previously [10]. The cell populations of CD33^hi monocytic cell type or CD16^hi neutrophilic cell type were gated.

Figure 2

Variable patterns of myeloid cell parameters (CD16, LAP-1, and GPI-80) before and during ICI therapy. The patients treated with ICI therapy as the 1st line were separated into four categories as shown in Table 2: complete response (CR; n = 3), partial response (PR; n = 6), stable disease (SD; n = 5), and progressive disease (PD; n = 5). The changing of the mean of fluorescence intensity (MFI) of CD16 (a) and LAP-1 (b) in CD33^hi monocytic cells or the MFI of GPI-80 (c) in CD16^hi neutrophilic cells were measured by flow cytometry using blood samples collected from the patients during ICI therapy.

Figure 3

Increased expression of CD16 and LAP-1 in CD33^hi monocytic cells in immunoresponse-non-dominant patients after the 1st treatment. In immunoresponse-dominant patients ((a); n = 18) and immunoresponse-non-dominant patients ((b); n = 26), the expression levels (MFI) of CD16 and LAP-1 in CD33^hi monocytic cells were measured using 1st and 2nd blood samples with flow cytometry. During the chemotherapies, 2nd blood samples were collected 2–3 weeks from the 1st blood sample collection, the same as for the ICI therapies. Statistical significances were calculated using the Wilcoxon rank test (two-tailed). Unpaired samples were removed from the data.

Figure 4

Associations of GPI-80 expression levels in CD16^hi neutrophilic cells immediately before ICI therapy with antitumor immune responses induced by ICI therapy. Both GPI-80 coefficient variation (CV; (a)) and GPI-80 MFI (b) in CD16^hi neutrophilic cells in blood samples were measured using flow cytometry. The blood samples were collected from healthy donors (HD represented by closed square; n = 11) or were the 1st blood samples (immediately before ICI therapy) of patients with RCC who received ICI therapy as the 1st line. The patients consisted of complete response (CR represented by closed circle; n = 3), partial response (PR represented by closed triangle; n = 6), stable disease (SD represented by open rhombus; n = 5), and progressive disease (PD represented by open circle; n = 5). The statistical significance was calculated using non-parametric ANOVA (Kruskal–Wallis) with a post-hoc test using Dunne, comparing CR vs. PR, SD, or PD. Each bar in figure is the mean of the data; ns—not significant, * p < 0.05.

Figure 5

No relation of proinflammatory factors in plasma with clinical outcome. Plasma collected from 1st blood sample (before treatment) from patients with complete response (CR) and progressive disease (PD) with the same samples as Figure 4. (a) Heat map of cytokine profile. Cytokines in the plasma were measured by LEGENDplex™ Human Inflammation Panel 1 and then unsupervised clustering was analyzed using Z-score distribution with Prism software (CR, n = 3; PD, n = 5). (b) Comparison of HMGB1 concentrations between patients with CR and PD. HMGB1 in the plasma were measured by ELISA, and data were analyzed using two-tailed unpaired Student’s t-tests. Each datum is presented as a dot with the mean (bar) ± standard deviation (CR, n = 3; PD, n = 4; ns, not significant).