NLR stability predicts response to immune checkpoint inhibitors in advanced hepatocellular carcinoma

Abstract

A high baseline NLR is associated with a poor prognosis of immunotherapy in patients with advanced HCC. As anti-tumour immune activation takes time, early dynamic changes in NLR may serve as a biomarker for predicting immunotherapy response. We conducted a retrospective study in which we enrolled 209 patients with aHCC who received ICIs (training cohort: N = 121, validation cohort: N = 88). In the training cohort, we categorized the patients based on the early changes in their NLR. Specifically, we defined patients as NLR Stable-Responder, NLR Responder and NLR Non-Responder. We compared the outcomes of these three patient groups using survival analysis. Additionally, we shortened the observation period to 6 weeks and validated the findings in the validation cohort. In the training cohort, early dynamic changes in NLR (HR 0.14, 95%CI 0.03-0.65, p = 0.012, HR 0.19, 95%CI 0.07-0.54, p = 0.002; HR 0.21, 95%CI 0.10-0.42, p < 0.001, HR 0.40, 95%CI 0.23-0.69, p = 0.001), PD-L1 < 1% (HR 5.36, 95%CI 1.12-25.66, p = 0.036; HR 2.98, 95%CI 1.51-5.91, p = 0.002) and MVI (HR 3.52, 95%CI 1.28-9.69, p = 0.015; HR 1.99, 95%CI 1.14-3.47, p = 0.015) were identified as independent predictors of OS and PFS. In the validation cohort, when the observation period was reduced to 6 weeks, early NLR changes still have predictive value. Early dynamic changes in NLR may be an easily defined, cost-effective, non-invasive biomarker to predict aHCC response to ICIs.

Keywords: Immunotherapy; NLR stability; OS; PD-L1; aHCC.

Figure 1

Data collected on the diagnosis of unresectable advanced hepatocellular carcinoma.

Figure 2

Kaplan–Meier survival curves showing OS (A) and PFS (B) after initiation of ICIs stratified by NLR kinetics groups in the training cohort.

Figure 3

Response distribution based on both mRECIST and RECIST1.1 in different NLR kinetics groups (A), as well as baseline NLR (B), PD-L1 expression (C), and MVI (D).

Figure 4

Kaplan–Meier survival curves showing OS (A) and PFS (B) after initiation of ICIs stratified by baseline NLR in the training cohort. Kaplan–Meier survival curves showing OS (C) and PFS (D) after initiation of ICIs stratified by PD-L1 expression in the training cohort. Kaplan–Meier survival curves showing OS (E) and PFS (F) after initiation of ICIs stratified by MVI in the training cohort.

Figure 5

Longitudinal NLR changes from baseline after initiation of immunotherapy in the three NLR kinetics subgroups in the training cohort. (A) For NLR Stable-Responders, (B) for NLR Responders, and (C) for NLR Non-Responders: the NLR changes of each individual patient are shown in the left panel; the percentage change in median NLR from baseline is shown in the right panel. The dashed lines indicate the thresholds for NLR Stable-Responder and NLR Responder subgroups. (D) Integration of median change in baseline NLR of the three NLR kinetics subgroups in the immune monitoring of immunotherapy aHCC training cohort and (E) conceptual representation of three NLR kinetics subgroups.