Gene Therapy to Cure HIV: Where to from Here?

A variety of approaches are being tested to cure HIV. At the heart of each is the goal of perturbing and ultimately removing the viral reservoir—defined as all potentially pathogenic virus—that persists despite suppressive antiretroviral therapy (ART), or at the least controlling the reservoir such that ART can be removed without adverse health consequences.

In this regard, several approaches have been tested, but have found to be insufficiently effective. ART, when initiated during the earliest stages of infection—even before peak viremia—fails to prevent the establishment of HIV reservoirs, ¹ and supplementing clinically typical ART with additional classes of ART (mega-ART) appears to exert little effect on the persistence of the reservoir, ² although there are some hints that it can reduce ongoing cryptic viral replication. ³ An approach known as “shock and kill” aims to reverse viral latency by inducing viral production, thereby resulting in detection and destruction of productively infected cells by the immune system or other immune interventions. So far, putative latency reversing agents (LRAs) such as the histone deacetylase inhibitors vorinostat, panobinostat, and romidepsin—thought to increase the access of HIV to transcription factors—have failed to detectably induce the production of virions when tested in the clinic, and no significant decrease in measures of reservoir persistence has been observed. ⁴

It is also unclear whether the immune system that has been compromised by HIV infection would be capable of clearing infected cells revealed by LRAs. Brief interruptions in ART, in the form of structured treatment interruptions (STIs), have been studied as a means to refresh the immune response, ⁵ but STIs have resulted in clinical harm, including increased risk of opportunistic infections, cardiovascular, renal, and hepatic disease, and even death. ⁶ To address the issue of killing persistently infected cells, many approaches designed to boost the existing immune response have been designed, and some tested clinically. Therapeutic vaccines have so far shown disappointing immunogenicity, ⁷ and passive immunity interventions such as adoptive transfer of ex vivo manipulated immune-competent cells, ⁸ chimeric antigen receptor (CAR) T, ^9,10 or the administration of broadly neutralizing antibodies ¹¹ are currently being tested, although it appears likely that these and other interventions would need to be implemented in combinations to prevent viral resistance, or personalized to most effectively address the particular strains of virus that persist in an individual's reservoir. Immune checkpoint blockers, currently approved for the treatment of a variety of cancers, have some potential to reverse immune exhaustion observed in HIV and are currently under clinical evaluation. ¹²

Despite the range of approaches under investigation, none has resulted in the elimination of the persistent viral reservoir with the exception of the allogeneic transplantation of CCR5delta32 stem cells that resulted in the Berlin patient cure. ¹³ The Berlin patient had been living with HIV for ∼10 years and was treated successfully with ART for about 4 years, until he developed acute myeloid leukemia (AML) refractory to chemotherapy. He received an allogeneic stem-cell transplant from a donor who was homozygous for the CCR5delta32 allele, but his AML relapsed a little under a year later. A second transplant from the same donor was administered, and despite comprehensive efforts to identify persistent HIV, ¹⁴ he remains free of AML and HIV 8 years later, despite discontinuation of ART.

This sole case of HIV cure strongly suggests that gene therapy may be a promising route to cure HIV. Gene therapy involves the disruption, modification, or addition of genetic material (commonly DNA, but sometimes RNA) to achieve a therapeutic goal. Several powerful gene-editing tools exist, such as nucleases, recombinases, and RNAi, that can potently edit viral or host genes. HIV infection presents a wide range of potential targets, including host genes such as CCR5^15,16 or the viral genome itself. ¹⁷ Other strategies might involve the addition of genes encoding proteins that mimic ART to protect edited cells, ¹⁶ optimized versions of restriction factors, or methods to bypass a defective and exhausted immune response through CAR-edited cells. ^9,10 The availability of tools and targets suggests that designing a gene therapy intervention to cure HIV is arguably a question of technology rather than discovery.

Among the most important technological barriers is delivery. If gene therapy is ever to be scalable, methods to deliver gene therapy in vivo will almost certainly be required. One key challenge is the inefficiency of vector uptake, which is lower still in resting cells. ¹⁵ Another is the ability to target the cells to be edited. Whether targeting host or viral genes, the cells hosting the DNA or RNA of interest are scattered throughout the body, in stark contrast to many clinical conditions in which in vivo gene therapy has been successfully applied. ¹⁷

It may be tempting to speculate that a gene-editing tool of sufficient specificity for the HIV genome might be delivered safely to every cell in the body, which bypasses the need to target already-infected cells and simultaneously protects susceptible cells. However, participants at a recent amfAR-convened meeting calculated that, based on the inefficiency of adeno-associated virus (AAV) vector uptake (roughly 1 in 220 enter target cells) and transgene expression (which occurs at best in 1 in 15 cells), the cost to manufacture sufficient product to cure a single person would be approximately $1 billion. The scale of the human body and the number of cells it contains mean that untargeted approaches that appear promising in mice may not translate to the clinic. Targeting might also be improved by detargeting—regardless of delivery method, most product is metabolized in the liver. Developing a means to maintain the gene-editing tools in the circulation by avoiding the liver and other potential sinks should be a top priority.

Gene therapy approaches in HIV currently focus on gene-editing cells outside the body (i.e., ex vivo cell transduction) followed by transplanting the cells into the subject (i.e., adoptive cell transfer). Challenges include discovering means to increase transduction efficiency, establishing and maintaining engraftment, as well as optimizing graft-versus-host disease (GVHD). Although severe GVHD can have obvious deleterious consequences, it is unclear whether it may help clear persistently HIV-infected cells, or how one might maintain a “Goldilocks” GVHD response, neither too much to be dangerous nor too little to clear infected cells. Engraftment is also a safety challenge—conditioning regimens in traditional cell transplant settings creates space for newly transplanted cells through blood cell ablation. The safety of such conditioning regimens in otherwise healthy, ART-suppressed individuals are currently being evaluated. ¹⁶

Given that gene modification is currently incompletely efficient—clinical trials in which CCR5 was modified ex vivo and bulk cells were transplanted report about one-third biallelic disruption in T cells and lower rates in hematopoietic stem cells ¹⁵ —it is important to determine whether transduction of 100% of target (transplant) cells is required. Disruption of CCR5 will protect target cells from HIV infection only if the gene editing is biallelic, but must every cell in an adoptive transfer setting have such complete genome editing? Preliminary evidence suggests that HIV itself may exert selective pressure on the transplanted cells, eliminating those that are not protected and allowing the expansion of those that are. ¹⁶ Even in post-transplant clinical trial settings in which CCR5-modified cells comprise a small minority, tantalizing hints of clinical benefit have been reported, including an increase in CD4⁺ T cells, a decrease in HIV DNA, and protection of gene-modified cells from infection-mediated death. ¹⁸ Which kind and degree of clinical benefit resulting from partial transplant efficacy justify the risks associated with adoptive transfer of gene-modified cells remains to be determined. In particular, given that HIV of other tropisms (e.g., CXCR4) can also infect cells—even in individuals homozygous for the CCR5delta32 gene ¹⁹ —care will need to be taken that an approach based solely on deleting CCR5 will not result in the selection of viruses that may be even more pathogenic.

Experience has taught that single-target approaches in HIV are likely to result in resistance. In this regard, it is fortunate that there are several targets to choose from, and combination gene-editing approaches are currently being tested in pre-clinical and clinical settings. ^9,16 However, editing multiple gene targets will foreseeably complicate the regulatory approval process, where concerns regarding the safety of any (even single) gene modification are foremost.

Of all approaches currently being investigated to cure HIV, there is probably the most data support for the efficacy of gene therapy. Depending on how it is deployed, it may have several advantages, including its specificity, permanence (with the potential to protect the patient from subsequent infection), one-time application, and ability to bypass the defective extant immune response. Proof-of-principle exists that a gene therapy approach can cure HIV, and optimization of transduction efficiency and in vivo persistence might result from improvements in current methodologies. However, the feasibility of the approach is still a major hurdle. The timeline, cost, and complexity of testing gene therapy in the clinic are formidable. Where gene editing has been submaximal, ART may need to be released to enable the selection of gene-modified cells, or to test whether a cure has taken place, although the latter potential safety challenge is shared by all cure approaches. Given the nature of gene therapy, exhaustive pre-clinical testing will be required before clinical trials can proceed. Although safety testing in uninfected animals is understandable and cost saving, efficacy testing must be conducted under conditions that mimic the clinical situation, in other words, during maximal, chronic ART suppression in infected animals. Even after successful clinical trials, there will be significant concerns regarding the practicality of curing millions of people, the majority of whom are in sub-Saharan Africa, using gene therapy.

The success of the Berlin patient and the results of several clinical trials in which various gene therapy interventions have been evaluated suggest that despite the challenges, gene therapy is a promising avenue in the quest for an HIV cure.