Establishing a whole blood CD4+ T cell immunity measurement to predict response to anti-PD-1

Abstract

Background: Biomarkers that can accurately predict the efficacy of immune checkpoint inhibitors (ICIs) against programmed death 1 (PD-1) ligand in cancer immunotherapy are urgently needed. We have previously reported a novel formula that predicts the response to treatment with second-line nivolumab with high sensitivity and specificity in patients with non-small cell lung cancer (NSCLC) previously treated with chemotherapy. The formula was based on the percentages of CD62L^lowCD4⁺ T cells (effector T cells; %Teff) and CD4⁺CD25⁺FOXP3⁺ T cells (regulatory T cells; %Treg) in the peripheral blood before treatment estimated using the peripheral blood mononuclear cell (PBMC) method. Here, we investigated the applicability of the formula (K-index) to predict the response to treatment with another ICI to expand its clinical applicability. Furthermore, we developed a simpler assay method based on whole blood (WB) samples to overcome the limitations of the PBMC method, such as technical difficulties, in obtaining the K-index.

Methods: The K-index was evaluated using the PBMC method in 59 patients with NSCLC who received first-line pembrolizumab treatment. We also assessed the K-index using the WB method and estimated the correlation between the measurements obtained using both methods in 76 patients with lung cancer.

Results: This formula consistently predicted the response to first-line pembrolizumab therapy in patients with NSCLC. The WB method correlated well with the PBMC method to obtain %Teff, %Treg, and the formula value. The WB method showed high repeatability (coefficient of variation, < 10%). The data obtained using WB samples collected in tubes containing either heparin or EDTA-2K and stored at room temperature (18-24 °C) for one day after blood sampling did not differ. Additionally, the performance of the WB method was consistent in different flow cytometry instruments.

Conclusions: The K-index successfully predicted the response to first-line therapy with pembrolizumab, as reported earlier for the second-line therapy with nivolumab in patients with NSCLC. The WB method established in this study can replace the cumbersome PBMC method in obtaining the K-index. Overall, this study suggests that the K-index can predict the response to anti-PD-1 therapy in various cancers, including NSCLC.

Keywords: Biomarkers; Effector CD4+ T cells; Flow cytometry; Non-small cell lung cancer; PD-1 blockade therapy.

Fig. 1

Correlation of the K-index with response to initial pembrolizumab therapy in patients with NSCLC. (a) PFS curves of patients receiving the first-line treatment. The blue line indicates K-index ≥192, and the red line indicates a K-index < 192. (b) OS curves of patients receiving first-line treatment. Statistical significance was assessed using the log-rank test. NSCLC, non-small cell lung cancer; OS, overall survival; mOS, median OS; PFS, progression-free survival; mPFS, median PFS

Fig. 2

Schematic representation of the PBMC and WB methods. WB, whole blood; FCM, flow cytometry; PB, peripheral blood; PBMC, peripheral blood mononuclear cell; RBC, red blood cell; WBC, white blood cell

Fig. 3

Gating strategy for FCM data analysis. a, f Singlet cells gated using FSC-A/FSC-H dot plot. b, g Lymphocytes gated using FSC-A/SSC-A dot plot. c, h CD4⁺T cells gated using CD4⁺/SSC-A dot plot. d, i Cells with low CD62L expression are selected as CD4⁺Teff cells. e, j Cells expressing both CD25 and FOXP3 are selected as CD4⁺Treg cells. FCM, flow cytometry; Teff, effector T cells; Treg, regulatory T cells

Fig. 4

The correlation between measurements using the PBMC and WB methods. a–c Blood samples were collected from 53 patients with lung cancer; each sample was divided into two and measured using PBMC and WB methods. Regression lines and correlation coefficients were calculated for %Teff (a), %Treg (b), and the K-index (c). The gray line represents the regression line, the light gray line represents the 95% confidence interval, and the dotted line represents the 95% prediction interval. PBMC, peripheral blood mononuclear cells; Teff, effector T cells; Tregs, regulatory T cells

Fig. 5

Effect of different anticoagulants during blood sampling in the WB method. a–c Blood samples were collected from 50 patients with lung cancer. Each sample was collected in heparin-containing and EDTA-2K-contained blood collection tubes. The correlation of the data obtained by both blood sampling methods is shown as %Teff (a), %Treg (b), and K-index (c). Regression lines and correlation coefficients were calculated for %Teff, %Treg, and the K-index. The gray line represents the regression line, the light gray line represents the 95% confidence interval, and the dotted line represents the 95% prediction interval. WB, whole blood; Teff, effector T cells; Treg, regulatory T cells

Fig. 6

Storage stability of blood samples in the WB method. a–c Blood samples collected into heparin-containing blood collection tubes from 20 patients with lung cancer patients were measured by the WB method a day after blood collection and compared with the values measured in the same manner within 4 h after blood collection. Regression lines and correlation coefficients were calculated for %Teff (a), %Treg (b), and K-index (c). Twenty-two samples collected from healthy individuals were measured using the WB method a day after blood collection and compared with the values measured in the same manner within 4 h after blood collection. The gray line represents the regression line, the light gray line represents the 95% confidence interval, and the dotted line represents the 95% prediction interval. WB, whole blood; Teff, effector T cells; Treg, regulatory T cells

Fig. 7

Effect of different FCM measuring devices on measured values in the WB method. a–c Blood samples were collected from 24 pretreated patients with lung cancer using the WB method and measured simultaneously using both XF-1600 and FACS Canto II. The correlation of both measurements was analyzed, and regression lines and correlation coefficients were calculated for %Teff (a), %Treg (b), and the K-index (c). The gray line represents the regression line, the light gray line represents the 95% confidence interval, and the dotted line represents the 95% prediction interval. WB, whole blood; FCM, flow cytometry; Teff, effector T cells; Treg, regulatory T cells