Extracellular vesicles for CAR T-cell therapy immunomonitoring Open Access

In this issue of Blood Advances, Lanuti et al¹ demonstrate a correlation between decreasing circulating CD19.chimeric antigen receptor–positive (CAR⁺) extracellular vesicle (EV) concentrations during the first 3 weeks after CD19.CAR T-cell (CD19.CAR T) infusion and failure of CD19.CAR T therapy. This led the authors to propose the use of the CD19.CAR⁺ EV levels as a biomarker to predict the efficacy of CD19.CAR T therapy. This study is remarkable in that it highlights, to our knowledge, for the first time, the importance of CD19.CAR⁺ EVs measured in the CAR T products vs in the peripheral blood of patients after infusion.

The mechanism underlying the ability of CD19.CAR⁺ EV levels to predict immunotherapy success remain unknown. Lanuti et al have suggested that EVs might have direct functional cytotoxic activity. EVs present during the first-month after CD19.CAR T infusion probably originates initially from vesicles present in the therapeutic product itself and then from the activation of CAR T cells in vivo. As suggested by the authors, this hypothesis is based on the carriage of proapoptotic cargo by EVs derived from CD19.CAR⁺ T cells.

However, the role of CD19.CAR⁺ T cell–derived EVs in CAR T products extend beyond simple prediction of treatment efficacy. Indeed, CAR⁺ EVs may also be involved in neurotoxicity-related adverse effects.² Immune effector cell-associated neurotoxicity syndrome (ICANS) is one of the most frequent adverse effects occurring after CAR T infusion. High levels of CAR⁺ EVs may be correlated with the onset and pathogenesis of ICANS via interactions with neural cells.² 

Based on these findings, we posit that these EVs might express other molecules on their surface, in addition to CAR, with possible immunomodulatory effects. Indeed, despite the considerable diversity of EVs, they are always derived from the budding of the plasma membrane and interact with their environment.³ In the context of immunoregulation, EVs derived from immune system cells have powerful remodeling effects on the immune system. It is, therefore reasonable to assume that EVs present in CAR T products may indeed affect the success of immunotherapy.

The spectrum of immunotherapies and immunomodulations is probably infinite. EVs are known to interact with conventional CD4⁺ T lymphocytes, CD4⁺ T folicular helper cells, IL17⁺ CD4⁺ T helper cells, CD4⁺ T regulatory cells (Treg), CD8 T lymphocytes, monocytes, B lymphocytes, and dendritic cells.^3-9 It is regrettable that Lanuti et al did not use other assays to evaluate immunoregulation. Furthermore, the cellular origin of EVs, their numbers, and the molecules they express on their surface drive immune interactions and immunomodulation in a nonrandom manner.⁸ Similar variations almost certainly occur in the EVs detected in CAR T products. For autologous CAR T products, of course, patient phenotypes vary at the time of apheresis collection and laboratory production methods are variable. It is also regretful that Lanuti et al restricted their analysis to CAR⁺ EV. However, it would be reasonable to assume that other EVs are present (ie, EVs not expressing CAR,² and not necessarily derived from lymphocytes), and that the spectrum of immunomodulation is broad.

The proteins identified by proteomics in the study by Lanuti et al suggest potentially strong expression of highly “immunoactive” molecules. Based on published findings in the EV field, we hypothesize that immunoregulatory molecules, for example, may be strongly expressed in vivo, with probable impacts on Tregs. Indeed, some EVs have inhibitory functions, preventing the plastic development of a proinflammatory phenotype in Tregs, thereby inhibiting interleukin-17 (IL-17) secretion by macrophages by reducing the release of tumor necrosis factor-α and IL-10 by activated macrophages.^5,6 CD39⁺ EVs can also affect CD4⁺ T lymphocytes by inhibiting their proliferation.⁹ The question as to whether these EVs are present in CAR T products needs to be addressed and may provide additional insight into the findings of Lanuti et al.

We also know that EVs have many other immunoactive properties beyond the regulation of immune responses,^4,7,8 and that they also reflect chronic immune activation in infectious diseases and may lead to comorbid conditions.¹⁰ These findings are important because Lanuti et al also showed that CAR⁺ EVs persist for at least 24 months in plasma, probably because CAR T cells also persist, and this is, of course, an important indicator that the effects of immunotherapy persist for much longer than you might think.

However, we do not know the functional and immunoregulatory role of EVs present after 24 months. We know that EVs exert various immune functions directly, including the presentation of antigens,³ and that the cytotoxic function alone may persist in vesicles derived from CAR T budding. In addition to this direct action, EVs may transfer this phenotype or enhance it in other cells.

In summary, although the results presented by Lanuti et al are preliminary, they raise interesting questions regarding CD19.CAR⁺ T cell–derived EVs. However, these results were obtained from only 22 patients treated with CD19.CAR T products; the data clearly indicate that this field of research will progress rapidly given the important properties of EVs. Therefore, we assert that these findings should be incorporated into the development of future immune monitoring protocols for patients treated with CD19.CAR T.

Conflict-of-interest disclosure: The authors declare no competing financial interests.