PD-L1 expression on circulating tumor cells can be a predictive biomarker to PD-1 inhibitors combined with radiotherapy and antiangiogenic therapy in advanced hepatocellular carcinoma

Abstract

Aim: A programmed death 1 (PD-1) inhibitor coupled with radiotherapy and antiangiogenic therapy is a potential therapeutic strategy for advanced hepatocellular carcinoma (HCC). We aimed to determine if circulating tumor cells (CTCs) positive for programmed death-ligand 1 (PD-L1) could be employed as a predictive biomarker in HCC patients receiving triple therapy.

Methods: In this study, HCC patients received a PD-1 inhibitor in combination with intensity-modulated radiotherapy (IMRT) and antiangiogenic therapy. Following IMRT, the PD-1 inhibitor was administrated once every 3 weeks, while the antiangiogenic drug was given once a day. Treatment was continued until the disease progressed. Two mL of peripheral blood was collected at baseline, 1 month, and 3 months after treatment for CTC enrichment using the CytoSorter^® system with a CytoSorter™ CTC PD-L1 Kit (Watson Biotech., China).

Result: A total of 47 HCC patients receiving the triple therapy were enrolled in this study. Patients with < 2 PD-L1⁺ CTCs at baseline had a higher objective response rate (ORR) and longer overall survival (OS) than those with ≥ 2 PD-L1⁺ CTCs (56.5% vs. 16.7%, p = 0.007; not reach vs. 10.8 months, p = 0.001, respectively). The count of PD-L1⁺ CTCs was found to be an independent predictive biomarker of OS. Furthermore, the objective response was more likely to be achieved in patients with a dynamic decrease in PD-L1⁺ CTC counts at 1 month after treatment.

Conclusions: Our study demonstrated that PD-L1⁺ CTCs could be a predictive biomarker for HCC patients receiving PD-1 inhibitors in combination with IMRT and antiangiogenic therapy.

Keywords: antiangiogenic therapy; circulating tumor cells; hepatocellular carcinoma; programmed death 1 inhibitor; programmed death-ligand 1; radiotherapy.

Figure 1

Detection of CTCs and PD-L1⁺ CTCs. The counts of CTCs (blue bars) and PD-L1⁺ CTCs (red bars) per patient detected are presented at baseline. Nonresponders and responders were divided into two groups. PD-L1, programmed death-ligand 1; CTCs, circulating tumor cells.

Figure 2

Comparison of the number of CTCs (A) and PD-L1⁺ CTCs (B) between the nonresponse group and response group. (C) ROC curve was adopted to investigate the predictive value of PD-L1⁺ CTC counts. When applying a cut-off of 2 PD-L1⁺ CTCs, a specificity of 65.5%, a sensitivity of 76.5%, and an AUC of 0.710 were observed. R, responder; CTCs, circulating tumor cells; PD-L1, programmed death-ligand 1; ROC, receiver operating characteristic; CI, confidence interval; AUC, area under the curve.

Figure 3

Kaplan–Meier plots: (A) progression-free survival and (B) overall survival based on PD-L1 expression on CTCs at baseline. HR, hazard ratio; CI, confidence interval; CTCs, circulating tumor cells; PD-L1, programmed death-ligand 1.

Figure 4

Multivariate Cox regression analysis of progression-free survival and overall survival. HR, hazard ratio; CI, confidence interval; AFP, alpha fetoprotein; PD-L1, programmed death-ligand 1; CTCs, circulating tumor cells.

Figure 5

Changes in total CTC counts, PD-L1⁺ CTC counts and proportion of PD-L1⁺ CTC counts at baseline (T0), 1 months (T1), and 3 months (T2) after the beginning of triple therapy. (A) CTC counts in total enrolled patients. (B) PD-L1⁺ CTC counts in total enrolled patients. (C) PD-L1⁺ CTC counts in responders. (D) PD-L1⁺ CTC counts in nonresponders. PD-L1, programmed death-ligand 1; CTCs, circulating tumor cells.