Engineered to recognize and eliminate cancer cells, chimeric antigen receptor (CAR) T cells are life-saving therapeutics. Since 2017, the FDA has approved six different CAR T cell therapies to treat aggressive blood cancers (1). But as more patients began receiving these cancer-fighting cells, clinicians noticed that some of them experienced neurological symptoms such as confusion and difficulty speaking, a complication called immune effector cell-associated neurotoxicity syndrome (ICANS).

Scientists don’t know what causes ICANS. Patients diagnosed with ICANS following CAR T cell therapy showed increased inflammation in cells that line blood vessels and increased blood-brain-barrier permeability (2). Other studies also implicated CAR T cells in altering blood-brain-barrier permeability, which may contribute to neurotoxicity (3). And a recent case report in the New England Journal of Medicine noted that the reactivation of a latent human herpes virus 6 (HHV-6), a common herpes virus, contributed to a case of neurotoxicity in a CAR T cell patient (4). 

In a new preprint, scientists reported that CAR T cells can carry latent HHV-6 that reactivates in a subset of patients (5). They observed this HHV-6 viral reactivation in both FDA-approved CAR T cell therapies and in cell therapies currently in clinical trials.

“Viral reactivation was not necessarily something that we set out to study,” said Caleb Lareau, a postdoctoral fellow at Stanford University and coauthor of the preprint. “It was more trying to understand what the complications are that are seen recurrently in patients receiving cell therapies, and can we do really good molecular biology to better understand those?”

Lareau and his colleagues probed the Serratus database, a cloud computing resource that enables researchers to search for viral sequences in all publicly available sequencing data (6). While the Serratus team initially designed the database as a way to identify new viruses, they incidentally created a catalog of all known viruses as well.

Using Serratus, Lareau searched for families of viruses that hibernate and reactivate later. When they looked at data from human T cells, they noticed that significantly more HHV-6 RNA was expressed than any other viral RNA. While HHV-6A mostly resides in sub-Saharan Africa, HHV-6B is prevalent worldwide. In fact, by age three, 90% of all people have been infected by HHV-6B (7). Because the Serratus database contains data collected from labs all over the world over the course of decades, Lareau and his team were confident that their findings were not a technical artifact.

“They took a very pragmatic approach,” said Rayan Chikhi, a computational biologist at the Institut Pasteur and one of the Serratus developers. “The Serratus people were impressed by this preprint because of the way they creatively used our tool for finding new and unexpected results.”

Lareau and his team found published examples of HHV-6 reactivation in cultured CAR T cells (8-9), so they investigated whether HHV-6 could come from the CAR T cell therapy itself. When they isolated cells from healthy donors, they observed an increase in HHV-6 expression in some of the T cells after CAR T conversion and cell culture. HHV-6 normally infects T cells via the OX40 receptor on the T cell surface. When T cells get activated during CAR T cell manufacturing or in the body of a patient, they upregulate the amount of OX40 on their surface. The researchers hypothesized that HHV-6 could spread to new T cells via this increased OX40 availability.

Using single-cell RNA sequencing (sc-RNAseq), the researchers identified 0.1-0.3% of HHV-6 “super-expressing” cells within the population of T cells from three healthy donors. When they performed sc-RNAseq on two of the donor samples again at day 25 or day 27 of in vitro cell culture, 49% or 62% of T cells respectively were now HHV-6 super-expressors, indicating that HHV-6 reactivates in a subset of cells and spreads to other T cells in the population.

“The really missing piece was, does this actually occur in CAR T cells that are actually going into patients now?” Lareau asked.

He and his team took samples of both FDA-approved CAR T cell products and those in a clinical trial and used sc-RNAseq to screen for HHV-6 expression. There were no HHV-6 positive cells in the preinfusion CAR T cell products, but when the researchers screened blood samples taken from patients, they identified 13 HHV-6 positive cells, including eight super expressor cells. One patient had about one super expressing T cell out of 1000 one week after CAR T cell infusion and developed ICANS from day 9 to 14 postinfusion. There was, however, no direct evidence that HHV-6 caused the patient’s ICANS.

“There's a clear connection to HHV-6 and cell therapies now,” said Lareau. “This should be at the forefront of clinicians minds where patients are starting to present with complications, that HHV-6 is tested for quite quickly.” Clinicians can treat HHV-6 reactivations with antivirals such as foscarnet.

Saar Gill, a CAR T cell researcher at the University of Pennsylvania and clinician who treats cancer patients with CAR T cells, said that the preprint is “super, really great science,” but, “the potential implications for CAR T cell therapy are, I think, a little bit overstated.”

Gill emphasized that ICANS typically resolves spontaneously, and its incidence is at most 30%, depending on the type of cancer being treated and CAR T cell therapy being used.

“If you put that number together with the very, very small incidence of HHV-6B detection, I think it has to tell you, at the very least, that it either explains a minority of cases, or that it … doesn't necessarily explain this neurotoxicity,” he said. “In patients who have an atypical neurotoxicity, for example, it fails to resolve for a long period of time or resolves but then recurs later, as in the case report in the New England Journal of Medicine, then yes, we should look” (4).

Lareau and his team recommend that scientists screen CAR T cell therapies for HHV-6 reactivation and that clinicians monitor blood samples from patients receiving CAR T cell therapy. But Gill explained that the types of techniques Lareau’s team used to screen these CAR T cell therapies, such as sc-RNAseq, are not routinely available in the clinic and that the recommendation is premature.

While the findings in this preprint do not alter Gill’s thoughts on treating patients with CAR T cell therapy, “what it will do is absolutely cement for us the need to think about alternative causes, including HHV-6 reactivation, in patients who have atypical neurotoxicity after CAR T cell treatment.”

Lareau agreed with that sentiment. He added, “The efficacy and the promise of cell therapies is still very evident… We encourage studies like this to better understand the safety and the potential space of what happens when you take cells out of the human, you manipulate them, and you put them back in. It just goes to show that strange things can happen, and you should be on alert.”

Moving forward, Lareau wants to figure out why HHV-6 reactivates in CAR T cells and hopes to find a molecular strategy to suppress that reactivation. He would also like to investigate methods to better detect the rare HHV-6 super-expressor cells and find ways to remove them from CAR T cell products during the manufacturing process.

“We absolutely hope that this only bolsters the field of cell therapy and that more patients can receive these,” said Lareau. “We hope that our research has really helped make these therapies safer as a function of understanding this.”

References

1. National Cancer Institute. CAR T Cells: Engineering Patients’ Immune Cells to Treat Their Cancers (2022). Available at: https://www.cancer.gov/about-cancer/treatment/research/car-t-cells
2. Gust, J. et al. Endothelial Activation and Blood–Brain Barrier Disruption in Neurotoxicity after Adoptive Immunotherapy with CD19 CAR-T Cells. Cancer Discov  7, 1404-1419 (2017).
3. Parker, K.R. et al. Single-Cell Analyses Identify Brain Mural Cells Expressing CD19 as Potential Off-Tumor Targets for CAR-T Immunotherapies. Cell  183, 126-142 (2020).
4. Spanjaart, A.M. et al. Confused about Confusion. N Engl J Med  386, 80-7 (2022).
5. Lareau, C.A. et al. Latent human herpesvirus 6 is reactivated in chimeric antigen receptor T cells. Preprint at: https://www.biorxiv.org/content/10.1101/2022.08.12.503683v2 
6. Edgar, R.C. et al. Petabase-scale sequence alignment catalyses viral discovery. Nature  602, 142-147 (2022). 
7. King O, Al Khalili Y. Herpes Virus Type 6. [Updated 2022 Jul 21]. In: StatPearls [Internet]. Treasure Island (FL): StatPearls Publishing (2022). Available from: https://www.ncbi.nlm.nih.gov/books/NBK540998/
8. Shytaj, I.L. et al. Alterations of redox and iron metabolism accompany the development of HIV latency. EMBO J  39, e102209 (2020).
9. LaMere, S. et al. Promoter H3K4 methylation dynamically reinforces activation-induced pathways in human CD4 T cells. Genes Immun  17, 283-297 (2016).