Fig. 3. Complementary effects of ACAT inhibition and PD-1 blockade.

a–c Ex vivo CTB staining of PBMC from patients with CHB to identify GM1-enriched microdomains. a Representative flow cytometry plot and summary data of %CTBhi of PD-1^± (n = 46) and 2B4^± (n = 15) CD8⁺ T cells. b %CTBhi of exhausted (PD-1⁺ HLA-DR⁻) vs activated (PD-1⁺ HLA-DR⁺) CD8⁺ T cells (n = 46) and of activated (CD38⁺) vs non-activated (CD38⁻) CD8⁺ T cells (n = 25). c Representative flow cytometry plot and summary data of %CTBhi of CD28^± CD8⁺ T cells and CD28^± PD-1^± subsets (n = 39). d Ex vivo %CTBhi CD8⁺ T cells and %CD28⁺ CD8⁺ T cells according to response to ACAT inhibition (n = 23). Response defined as increased or de novo HBV-specific IFNγ production. e, f PBMC from patients with CHB stimulated with HBc OLP in the presence of DMSO (grey), ACAT inhibition (Avasimibe, black), PD-L1/PD-L2 blockade +DMSO (PD-1 block, white) or a combination of ACAT inhibition and PD-L1/PD-L2 blockade (combo, red) for 8d. e Representative flow cytometry plots and summary data of patients showing IFNγ production in any of the four conditions (n = 26). Boxes below the histogram indicate treatment strategy resulting in the highest IFNγ production in the respective patients (grey: DMSO; white: PD-1 block; black: ACAT inhibition; red: combo). f Fold change of IFNγ with PD-1 block or combo normalised to control without peptide stimulation. g Effector function of PBMC from a patient with HBV-related HCC 2 weeks after start of in vivo anti-PD-1 immunotherapy. PBMC stimulated with HBV OLP ±ACAT inhibition (Avasimibe or DMSO for 8d). P values determined by Wilcoxon matched-pairs signed-rank test (a–c (left), d–f) or Friedman test with Dunn’s multiple comparisons test (c (right)).