Blind Spot

Corresponding Author

Immune checkpoint inhibitors (ICIs) are a class of monoclonal antibody (mAb)–based anticancer drugs that reinforce T cell–mediated immune responses by blocking co-inhibitory T cell receptors (“check points”), thereby overcoming tumor-induced immune evasion. The Food and Drug Administration–approved ICIs—ipilimumab (anti-CTLA-4 [anti-cytotoxic T lymphocyte-associated antigen 4] mAb); nivolumab and pembrolizumab (anti-PD-1 [anti-programmed cell death protein 1] mAbs); and atezolizumab, avelumab, and durvalumab (anti-PD-L1 [anti-programmed cell death protein ligand 1] mAbs)—have been shown to be efficacious in the treatment of several malignancies. In this issue of JACC: CardioOncology, Poels et al. (1) investigated whether ICIs increase atherosclerosis. This hypothesis is based on the premise that, by overriding critical inhibitory regulations, ICIs lead to excessive inflammation, cytokine release, and immune cell infiltration affecting several organs, consequently resulting in immune-related adverse effects (2). Cardiovascular toxicities associated with ICIs, initially underestimated due to a lack of systematic monitoring and heterogenous clinical presentations, have now gained attention due to their fulminant progression and fatal consequences (3). In a retrospective 8-center international registry study, Mahmood et al. (4), reported 1.14% estimated prevalence of ICI-induced myocarditis with a median onset time of 34 days in patients with a mean age of 65 ± 13 years (4). Over a median follow-up of 102 days, 16 of 35 myocarditis cases experienced a major adverse cardiovascular event. A review of Vigibase, the World Health Organization’s database of individual case safety reports, revealed 46% fatality in 101 cases of ICI-induced severe myocarditis in patients with malignancies, mostly melanoma and lung cancer (5). In a multicenter case study series, Heinzerling et al. (6) documented autoimmune myocarditis, cardiomyopathy, heart failure, myocardial fibrosis, and cardiac arrest in 8 cases of ICI-related cardiotoxicity in melanoma patients.

Unexpectedly, plaque inflammation in pembrolizumab and nivolumab or ipilimumab–treated stage IV melanoma patients using 2-deoxy-2-[fluorine-18]fluoro- D-glucose (¹⁸F-FDG) positron emission tomography (PET) integrated with computed tomography showed no difference in the aortic, carotid, spinal, or splenic signal. ¹⁸F-FDG uptake mostly measures myeloid cell content, because their glucose consumption results in visible deoxyglucose uptake. To address the mechanism, the authors applied ¹⁸F-FDG PET to Apoe^–/– mice treated with anti-CTLA-4 and anti-PD-1 mAbs for 4 weeks. They report no signal difference in bone marrow and a slight signal increase in the spleen, but this was not mirrored by accumulation of Ly6C⁺ inflammatory monocytes, which increased slightly in the bone marrow but not in the spleen. However, the authors found a small but consistent and significant increase in CD3⁺ T cells, effector (CD44⁺) CD4⁺ T cells, and FoxP3⁺ T regulatory cells. The increase in effector CD4 T cells was mirrored by a concomitant decrease in naïve CD4 T cells. Histological analysis of CD3⁺ cells in nonhematopoietic organs revealed marked T cell infiltration in the lungs and heart but not in the colon.

Next, the authors focused on aortic arch and root lesions in mice. The aortic arch showed no difference in plaque area, but mice receiving ICIs showed more pathologic intimal thickening according to the Virmani classification. Hematoxylin and eosin staining revealed more necrotic core in the mice treated with anti-CTLA-4 and anti-PD-1, more leukocyte (CD45⁺), T cell (CD3+) and CD8 T cell infiltration, more apoptotic cells measured by TUNEL (terminal deoxynucleotidyl transferase dUTP nick end labeling) staining, and more immunofluorescence signal for the adhesion molecules ICAM (intercellular adhesion molecule)-1 and VCAM (vascular cell adhesion molecule)-1. Interestingly, Mac-3 staining, a macrophage marker, was decreased. Similar data were obtained in the aortic root.

This study is important because it begins to unravel the mechanism by which ICI treatment can increase atherosclerosis. It is clinically important that there is no signal in ¹⁸F-FDG PET, a commonly used imaging modality that sees myeloid cells but not T cells. FDG PET was used as a readout in the negative GLACIER (Study to Evaluate the Safety, Tolerability, and Activity of Intravenous MLDL1278A in Patients on Standard-of-Care Therapy for Stable Atherosclerotic Cardiovascular Disease) atherosclerosis vaccination trial (7). The present study shows that ¹⁸F-FDG-PET is not a suitable readout for ICI-induced plaque inflammation. The authors largely reproduced this in mice.

Aortas and lymphoid organs from mice are easily accessible and can be analyzed by flow cytometry after tissue harvesting and dissociation. Mice treated with anti-CTLA-4 and anti-PD-1 show not only proinflammatory changes (increased CD3⁺, CD4⁺, CD8⁺ effector T cells), but also more anti-inflammatory FoxP3⁺ regulatory T cells. The net effect is not more plaque, but rather a less favorable, more complex plaque phenotype with more pathologic intimal thickening, more necrotic core (known to be associated with plaque instability), and more adhesion molecules.

A major limitation of this study is the stark difference in the pathology between patients and mice. First, on the one hand, the melanoma patients did not have a history of cardiovascular disease, and only half of them had known cardiovascular risk factors. The Apoe^–/– and Ldlr^–/– mouse models, on the other hand, did not have any malignancy but had existing atherosclerosis. Second, the length of ICI treatment was significantly longer in mice (4 to 5 weeks of a ∼2-year life span) than in humans (6 weeks of ∼70-year life span). Third, the authors did not measure the levels of atherosclerosis-associated inflammatory markers such as high sensitivity C-reactive protein, serum matrix metalloproteinases, and cytokines.

Future studies are likely to provide deeper insights into the immune cell changes induced by ICI therapy. Multiparametric analysis of the infiltrated T cells by mass cytometry or single-cell RNA sequencing is likely to reveal more phenotypic heterogeneity of intraplaque activated immune cells. 5′ single-cell RNA sequencing can also be used to reconstruct the T cell receptor α and β chains, thus determining the clonality of the infiltrating T cells. In mice and humans, some CD4 T cells are specific for epitopes in apolipoprotein B, a major atherosclerosis antigen (8,9), which can be assessed by tetramers or restimulation assays. A proposed mechanism for ICI-mediated autoimmune myocarditis in cancer patients is the clonal expansion of cross-reactive T cells that recognize a common antigen shared by the tumor and the cardiac myocytes (10), but this hypothesis remains to be tested rigorously. Determining the antigen specificity, functional properties and clonal identity of the infiltrating T cells will help unravel the molecular pathogenesis of ICI-induced aggravation of subclinical atherosclerosis.

For now, this study shows that ¹⁸F-FDG-PET is not a suitable method to follow plaque inflammation in response to ICI. This means that better imaging techniques or other biomarkers are needed to monitor the response to ICI treatment, including possible exacerbated atherosclerosis.

Author Disclosures

Dr. Ley is co-founder of Atherovax. Dr. Roy has reported that they have no relationships relevant to the contents of this paper to disclose.