Development of Screening Assays for Use of Broadly Neutralizing Antibodies in People Living with HIV

Abstract

Purpose of review:

Treatment with combinations of complementary broadly neutralizing antibodies (bnAbs) has increased the proportion of participants for whom bnAbs can maintain virus suppression upon cessation of antiretroviral therapy (ART). There remains, however, a population of trial participants who experience virus rebound despite high plasma concentrations of bnAbs. Thus, baseline resistance remains a critical barrier to the efficacy of bnAbs for use in the treatment and cure of HIV, and the development of a screening assay to guide bnAb selection is a high priority.

Recent findings:

There are two conceptual approaches to assess the putative rebound-competent HIV-1 reservoir for bnAb sensitivity: (i) to assess neutralization sensitivity of reactivated virus in outgrowth assays and (ii) sequence-based approaches that include a selection for intact genomes and assessment of known resistance mutations within the env gene. Currently the only phenotypic assay for bnAb screening that is CLIA certified and available for clinical trial use is Monogram Biosciences’ PhenoSense HIV Neutralizing Antibody Assay.

Summary:

Several new approaches for screening are currently under development and future screening methods must address three issues. First, complete sampling of the reservoir may be impossible, and determination of if and when partial sampling is relevant is needed. Second, multiple lines of evidence indicate that in vitro neutralization measures are at least one correlate of in vivo bnAb activity that should be included in screening, but more research is needed on interpretation of neutralization measures and measures of other antibody features. Third, the feasibility of screening assays must be a priority. A feasible, predictive bnAb screening assay will remain relevant until a time when bnAb combinations are substantially more broad and potent.

Introduction.

Advances in B cell biology and molecular virology have enabled the discovery, characterization, and commercial development of broadly neutralizing antibodies (bnAb) targeting a broad range of genetic variants of HIV-1. BnAbs represent a promising antiviral immunotherapy that can be incorporated in strategies for the treatment and/or cure of HIV-1 (1–5). Multiple bnAbs have been tested in clinical trials, demonstrating improved efficacy when they are used in combination (6–10). Like small molecule antivirals, bnAbs require combination therapy due to (i) the ability of HIV-1 to rapidly escape when replicating and/or (ii) pre-existing variants that are resistant enough to replicate in the presence of the infused bnAb. In all recent trials of combination bnAbs where participants undergo analytic treatment interruption (ATI), a fraction of participants experience rapid virus rebound. This rebound has been associated with minor resistant variants within the individuals’ virus reservoir identified in post-hoc analyses. Thus, development of a screening assay to identify individuals who would most benefit from bnAb therapy is a high priority.

HIV-1 diversity.

HIV-1 employs high mutation, recombination and replication rates to enable rapid escape from selection pressures. This evolutionary strategy makes HIV-1 one of the most genetically diverse human pathogens, with extensive variability between HIV subtypes globally and the potential to generate remarkable diversity within any individual person living with HIV (PLWH). HIV-1 envelope trimers (Env), the target of bnAbs, are similarly diverse and able to tolerate remarkable variability across key epitopes. The extensive diversity of Env sequence and conformation are a fundamental obstacle to the development of monoclonal bnAbs as agents for HIV-1 prevention or therapy.

Resistance to bnAbs in preclinical and clinical trials.

Preclinical studies in nonhuman primates (NHP) and early clinical trials have delineated the challenge of both baseline and emergent resistance to bnAbs. Early preclinical studies testing the in vivo antiviral activity of bnAbs in SHIV-infected rhesus macaques showed that bnAbs administered as single agents or in combination to viremic SHIV-infected RM could transiently suppress plasma viremia, but that escape mutants were readily detected upon viral rebound (11,12). As these NHP studies used a well-defined clonal SHIV infection, this de novo resistance could be distinguished from baseline resistance. In studies in viremic human participants administered infusions of single CD4 binding site (CD4bs) or V3 glycan bnAbs, the extensive heterogeneity of viruses within any individual complicates the distinction of pre-existing from newly emergent resistance mutations. Multiple studies demonstrated rapid selection for more resistant viral variants over just days to weeks of bnAb administration (13–16), which could have been selected from pre-existing minor resistant variants or be the result of rapid adaptation within a large, heterogeneous pool of replicating viruses. In human studies of single CD4bs-targeting bnAbs given to PLWH on suppressive ART in the context of ATI, evidence of both pre-existing resistance and the development of novel resistance has been demonstrated. Studies of VRC01 or 3BNC117 showed pre-existing resistance in viruses replicating prior to ART initiation and/or within proviral DNA reservoir elicited via outgrowth cultures (17–19). Emergent resistance was also implicated, as viruses became more resistant to administered bnAbs during the course of replication post-virus rebound and sequence changes consistent with recombination occurring pre-ART or post-rebound were detected (17–19). Rapidly emergent resistance remains a substantial challenge to bnAb therapy, thus maintenance of near complete virus suppression may be required for success in PLWH. Treatment with combinations of complementary bnAbs has increased the proportion of participants for whom bnAbs can maintain virus suppression upon switch from ART (7–9), but there remains a portion of trial participants who experience relatively rapid virus rebound despite high plasma concentrations of bnAbs. Thus, baseline resistance remains a critical barrier to the efficacy of bnAbs for use in the treatment and cure of HIV, and the development of a screening assay to guide bnAb selection is a high priority.

Laboratory methods to screen for bnAb resistance.

The first approach for post-hoc analysis of bnAb resistance in clinical trials of viremic PLWH was single genome sequencing (SGS) of plasma HIV-1 env genes (20,21). After cloning, Envs were tested for antibody sensitivity using the validated TZM.bl neutralization assay (6,10,13–15). These studies identified pre-existing resistance and mapped pathways of virus resistance developed after bnAb infusion by linking sequence and phenotype. This approach, however, requires sequencing circulating plasma viremia, and so is not applicable to the majority of clinical trials in PLWH on suppressive ART for whom no plasma virus sampling exists. The question of how characterize viral bnAb sensitivity in PLWH on ART remains unanswered, as how best to sample the totality of the persistent virus reservoir (22) as well as which assays to use on the sampled virus populations (23,24) remain active areas of investigation. Notably, virus reservoirs are substantially smaller and less diverse in individuals starting ART during acute and early HIV (25–28). Screening assays in early ART initiators with homogeneous virus reservoirs are likely to be far more predictive of clinically relevant bnAb resistance, a finding that has been borne out in recent trials in this population (9,25)

In general, there are two current conceptual approaches to sample the reservoir – (i) to study reactivated virus in outgrowth assays and (ii) to use sequence-based approaches to evaluate bnAb epitopes of inferred replication competent genomes. In quantitative co-culture assays for viral outgrowth, resting CD4 T cells from participants are plated in serial dilutions with target cells and stimulated (25). Viral replication is monitored for reactivation, and these culture supernatants can be used in the TZM.bl assay to screen for bnAb sensitivity. The sensitivity of this approach is defined by the number and origin of the cells sampled, and the ability of cells to be reactivated ex vivo(30). This approach was used to pre-screen viremic participants in clinical trials with a single antibody (14,15) as well as a combination of 2 bnAbs (6), but there were always a small minority of participants who did not respond with the expected drop in virus load after bnAb infusion. In each case, post-hoc analyses with SGS and Env cloning revealed pre-existing resistance. In a subsequent ATI trial, bulk culture outgrowths were used for pre-screening prior to 3BNC117 monotherapy using a cutoff of IC50 ≤ 2.0 μg/ml, yet for some participants, rebound with a resistant virus was rapid. This study found no correlation between the bulk outgrowth virus neutralization sensitivity and time to rebound (18). However, a VRC01 ATI trial in acutely treated participants that did not pre-screen found that IC80 did correlate with time to rebound (25). Using the cutoff of IC50 ≤ 2.0 μg/ml in bulk outgrowth assays as a pre-screening tool for a combination 3BNC117 and 10–1074 ATI trial led to an increased frequency of durable virus suppression, but similarly failed to detect minor resistant variants in a few participants (7). Post-hoc analysis employing the quantitative and qualitative viral outgrowth assay (Q2VOA), which involves limiting dilution virus outgrowth and sequence confirmation (23), again identified minor resistant variants within the reservoir of early rebounding participants. A second combination bnAb ATI trial that did not pre-screen participants found that testing outgrowth virus for neutralization sensitivity did in fact identify baseline resistant variants that were associated with early virus rebound (9). Thus, post-hoc analyses reveal that in vitro neutralization assays of reservoir variants generally correlate with time to rebound, but this correlation may be lost when a pre-screening assay is used as an entry criterion. Further, correlations appear more robust when screening early treated participants. Together, trial results suggest that pre-screening may exclude participants who harbor dominant resistant virus while missing those who harbor minor variants.

Another approach for analyzing the putative replication competent reservoir is Q4PCR, where near full-length (NFL) PCR is performed on genomic DNA for one round followed by a multiplex quantitative PCR (qPCR) with probes targeting four conserved genomic sites (31). Only samples with a putative positive signal for intact genomes are fully sequenced and assessed for bNAb resistance mutations. To date the Q4PCR approach has been used for post-hoc analysis of a combination bnAb ATI clinical trial, suggesting that participants with 10–1074-associated resistance mutations may have rebounded earlier than those without, but this was not statistically significant (8). Sequence predictions were, however, only used for 10–1074 (a V3 glycan bnAb with well-characterized escape mutations), as sensitivity to 3BNC117 (a CD4bs bnAb with many context-dependent escape mutations) is more complex (14,15,32). The challenge of predicting CD4bs bnAb resistance highlights how sequence-based inference of antibody susceptibility is in the early stage of development, requiring continued study of resistance pathways to each clinically relevant bnAb across diverse HIV-1 Envs (33). These sequence-based approaches are potentially higher-throughput than outgrowth methods, but are labor and cost-intensive at present.

Both outgrowth and sequencing approaches have drawbacks. Both require large volumes of cells, mandating large blood draw or leukapheresis, which are burdensome for study participants and/or restrict the sites that can perform studies. These approaches often take weeks to obtain results, rendering them logistically challenging for pre-screening purposes. Methodologically, bulk outgrowths underestimate the true reservoir, as they sample a fraction of circulating cells and no tissue-based reservoirs (34), while also preferentially identifying cells that are easily induced in vitro. Sequencing based approaches also suffer from under-sampling, and may further underestimate the reservoir due to PCR inefficiency and primer/probe mismatching. Fundamentally, the level of sampling needed to accurately assess the rebound-competent virus in PLWH who started ART during chronic infection has not been established for either approach. These drawbacks may explain the remaining discrepancies between clinical outcomes and in depth post-hoc analyses. Overall, however, results from advanced laboratory assays demonstrate that virus populations can be characterized from blood cells at a sensitivity that identifies resistant variants in the majority of participants who fail to achieve durable virus suppression from bnAb administration. These findings provide optimism that a sensitive screening assay for bnAb resistance could be developed, ideally with the capacity to predict clinical activity and greater feasibility.

The PhenoSense Assay.

The first bnAb screening assay to be CLIA certified and available for clinical trial use is the PhenoSense HIV Neutralizing Antibody Assay from Monogram Biosciences. This assay is based on the Trofile assay system, which generates a swarm of pseudotyped Env clones from plasma virus or cellular provirus to determine the concentrations of bnAbs that limit entry on receptive cells. The Trofile assay has been used extensively for early antibody studies (35–39) and as a commercial assay for antiretroviral sensitivity (40–42).

The PhenoSense assay was validated for clinical bnAb use in a collaboration between Monogram and the AIDS Clinical Trials Group (ACTG), which paved the way for use of this assay as a screening assay in clinical trials (43,44). In a demonstration study of pre-ART plasma and on-ART peripheral blood mononuclear cells (PBMC) from 69 participants of a previously conducted study of antiretroviral therapy (45), the assay successfully generated a resistance measure in 88% of samples, with values that were similar between plasma and PBMC sample types (43,44). Across seven bnAbs targeting CD4bs, V3, V2, and MPER epitopes, PhenoSense sensitivity measures were similar to published measures of the bnAbs to large panels of virus isolates (46). These results suggest the PhenoSense assay accomplished some goals of screening assays: it generates consistent Env pseudovirus swarms across available biospecimens (pre-ART plasma virus and longitudinal PBMC provirus) and provides an output measure that is comparable to validated assays commonly used in bnAb clinical development (21,23,29,31).

To begin to understand its predictive capacity, the PhenoSense assay has been applied post-hoc to small studies of bnAbs administered to PLWH with chronic ART initiation at ATI. As a part of the ACTG validation study, the PhenoSense DNA assay was used on PBMCs collected at screening visits for studies of either VRC01, 3BNC117, or combinations of the two (7,17,47). In a subset of participants of the study of VRC01 at ATI, pre-trial PhenoSense DNA neutralization measures tightly correlated with published pre-trial assessments of baseline VRC01 sensitivity in pre-ART plasma virus. When assessed for predictive capacity, however, baseline PhenoSense DNA neutralization measures to VRC01 did not correlate with time to rebound. Baseline PhenoSense DNA neutralization measures of sensitivity to bnAbs 3BNC117 and/or 10–1074 also did not predict time to rebound in the two additional ATI studies. Finally, PhenoSense DNA measures at baseline in a study of multiple doses of 3BNC117 + 10–1074 did not correlate with successful maintenance of virus suppression (8). Though these analyses are post-hoc, on a limited number of participants, and in heterogenous clinical trial settings, they suggest that the PhenoSense may not be able to consistently predict the efficacy of bnAbs to durably suppress viremia in PLWH started on ART in chronic infection, and that additional factors should be considered.

After dissemination of the early validation data, several trial investigators included the PhenoSense assay as a screening or analysis tool. One of the first studies to incorporate this assay was the eCLEAR study, which tested 3BNC117 and/or the LRA romedepsin at ART initiation (48,49). In eCLEAR, PhenoSense measures were not used as an entry criterion, but rather to distinguish between bnAb-sensitive and resistant participants in trial analyses. Notably, this distinction enabled the investigators to interpret that the bnAb activity in sensitive hosts drove enhanced clearance of virus-infected cells and augmented HIV-specific CD8 T cell responses (48). Ongoing or recently completed studies that incorporate the PhenoSense assay as a mandatory eligibility criterion include: A5357 (testing VRC07523.LS and injectable cabotegravir (50)); BEAT2 (testing 3BNC117, 10–1074 and type I interferon (51,52)); TITAN (testing 3BNC117, 10–1074 and the Toll Like Receptor Agonist lefitolimod (53)); and the SCOPE study (testing 10–1074 and VRC07523.LS with other immunotherapies; NCT04357821). These trials employed a range of neutralization cutoffs to hone study populations toward individuals with susceptible virus reservoirs. Primary results from these trials are expected soon, and with them a more thorough assessment of if and how prospective use of this assay is feasible or predictive of clinical activity.

Future Directions.

While the evaluation of the PhenoSense assay is ongoing across multiple studies, several new approaches are currently under development. These include outgrowth assays in the presence of the bnAbs for rapid selection, ultrasensitive binding assays that could rapidly infer neutralization sensitivity, and additional sequence-based prediction methods (54–58). Advances in these technologies will allow the field to build on current methods available only as laboratory assays or prototype clinical screening assays.

Conclusions:

The data generated over the past decade of early bnAb clinical trials suggest a path forward for bnAb screening assay development and highlight several common themes.

First, complete sampling of the reservoir may be impossible, but partial sampling may be helpful. With the notable exception of studies in participants with early ART initiation (9,25), post-hoc analyses have consistently identified minor variants that were not apparent with pre-screening. Even with foreseeable advances, the limits imposed by using readily available biospecimens remain concrete. Blood cells from PLWH on ART will not allow complete sampling of minor populations of circulating provirus, nor tissue reservoirs. Thus, the field may need to accept that the best possible screening assay will likely miss clinically relevant minor variants in individuals with diverse viral reservoirs, and may have greater power in participants with early ART initiation or similarly homogeneous reservoirs. Lest the perfect become the enemy of the good, the field should consider how an incompletely sensitive assay may be best applied.

Second, multiple lines of evidence indicate that in vitro neutralization measures are at least one correlate of bnAb activity, but more research is needed. Extensive analyses of the Antibody Mediated Protection (AMP) studies (59) of VRC01 determined a correlate for protection, the PT80, which identifies a bnAb concentration above the participants’ viruses IC80 neutralization titer that is efficacious in preventing HIV acquisition (60). Investigators estimated a need for bnAb concentration 200-fold higher than the IC80 of circulating viruses would confer higher than 90% efficacy in protection. Given the distinct clinical contexts, similarly rigorous evaluations of bnAb activity are needed for therapeutic and curative approaches, including for suppression of viremia, maintenance of suppression at ATI, clearance of latently infected reservoir cells, and induction of a vaccinal effect to boost HIV-specific immunity. We must not only define the requisite neutralization potency for these activities and translate these into reasonable “cutoffs” for in vitro screening assays, but also define the required antibody effector functions, half-life, and biodistribution. For all of these antibody functions, sequence-based inference holds promise to dramatically shorten evaluation timelines and reduce costs. But this promise is dependent on continued study of the distinct pathways of resistance for all bnAb classes across diverse virus clades for neutralization inferences, as well as identifying the sequence determinants of other relevant antibody properties.

Third, the feasibility of screening assays needs to be a priority. Current clinical trials of cure interventions are complex and demanding of participants and clinical sites. Adding weeks and dollars to the screening process and excluding potential participants without confidence in the predictive capacity of screening assay is harmful to the HIV cure agenda.

Finally, time is of the essence. A predictive screening assay would be of great value now, with the range of bnAbs available currently and in the near future (4,5). As the field develops increasingly broad and potent agents, one could optimistically envision a scenario in which baseline resistance is far less frequent and the role for a screening assay is diminished.

Key Points:

• Combination bnAbs are currently able to suppress rebound viremia in most, but not all, trial participants.

• Laboratory-based approaches can detect minor resistant variants associated with virus breakthrough during bnAb studies and provide the framework from which to develop a more feasible screening assay.

• Sampling the reservoir sufficiently to detect all resistant variants may not be possible, and screening methods may have greater power in participants with early ART initiation whose reservoirs are more homogeneous.

• We need a greater understanding of how to define clinically relevant “cutoffs” of neutralization resistance, as well as the role of effector function and antibody pharmacokinetics.

• Feasibility in terms of samples required, time, labor and cost must be factored into the design of an optimal screening assay.