Abstract
Background
Development of a screening assay for the clinical use of broadly neutralizing antibodies (bnAbs) is a priority for HIV therapy and cure initiatives.
Methods
We assessed the PhenoSense Monoclonal Antibody Assay (Labcorp-Monogram Biosciences), which is Clinical Laboratory Improvement Amendments (CLIA) validated and has been used prospectively and retrospectively in multiple recent bnAb clinical trials.
Results
When performed on plasma and longitudinal peripheral blood mononuclear cell samples (before and during antiretroviral therapy, respectively), as sourced from a recent clinical trial, the PhenoSense assay produced robust reproducibility, concordance across sample types, and expected ranges in the susceptibility measures of bnAbs in clinical development. When applied retrospectively to baseline samples from 3 recent studies, the PhenoSense assay correlated with published laboratory-based study evaluations, but baseline bnAb susceptibility was not consistently predictive of durable virus suppression. Assessment of assay feasibility in 4 recent clinical studies provides estimates of assay success rate and processing time.
Conclusions
The PhenoSense Monoclonal Antibody Assay provides reproducible bnAb susceptibility measurements across relevant sample types yet is not consistently predictive of virus suppression. Logistical and operational assay requirements can affect timely clinical trial conduct. These results inform bnAb studies in development.
Keywords: HIV, broadly neutralizing antibodies, clinical trials, screening assays
Development of a screening assay for broadly neutralizing antibody susceptibility is a priority for HIV therapy and curative studies. Here, we assess the PhenoSense Monoclonal Antibody Assay for reproducibility, feasibility, and predictive capacity.
Graphical Abstract
Graphical Abstract.
Preexisting resistance is a barrier to the efficacy of broadly neutralizing antibodies (bnAbs) for the treatment and cure of HIV-1 infection. Development of a screening assay to guide participant selection in clinical trials and ultimately clinical use is a high priority yet a challenging goal. To be useful, a screening assay must (1) use readily available clinical biospecimens, (2) retrieve or generate relevant HIV-1 envelope proteins (Envs), (3) provide a reliable estimate of the relative susceptibility of the collected Envs to specified bnAbs, and (4) be interpretable with validated cutoffs that predict clinical efficacy. To our knowledge, the only potential assay that is currently Clinical Laboratory Improvement Amendments (CLIA) validated and available for use in clinical trials is the PhenoSense Monoclonal Antibody Assay (PhenoSense mAb assay) from Labcorp-Monogram Biosciences. This phenotypic infectivity assay is capable of assessing the neutralization susceptibility of a swarm of pseudovirions bearing plasma- or peripheral blood mononuclear cell (PBMC)–derived HIV-1 Envs to specific bnAbs. This assay has been used prospectively and retrospectively in ongoing and recently completed studies [1–3]. In this report, we include data from 8 clinical trials that enrolled people with HIV (PWH) [1, 3–7]. We determine assay reproducibility and concordance across sample types, retrospectively apply the assay to published bnAb studies [4–6], and describe assay feasibility in recent studies. Together, results describe the current strengths and limitations of this screening methodology.
METHODS
The PhenoSense HIV mAb assay was developed by leveraging the PhenoSense assay platform, which was developed to evaluate antiretroviral drug susceptibility [8] and later adapted to evaluate entry inhibitors, neutralizing antibodies, and coreceptor tropism [9–11]. The measurement of bnAb activity is performed by generating HIV-1 pseudovirions that express participant virus-derived Envs. Pseudoviruses are prepared by cotransfecting HEK293 producer cells with an HIV-1 genomic luciferase reporter vector and an Env expression vector. Neutralizing antibody activity is measured via reduction of luciferase activity in U87 cells expressing CD4 and CXCR4/CCR5 following preincubation with specific bnAbs; neutralization is measured as a continuous outcome with an upper limit that is right censored to a predefined starting antibody concentration. The bnAb concentration conferring 50%, 80%, 90%, and 95% inhibition (IC50, IC80, IC90, and IC95) of pseudovirus infection can be reported. The residual infectivity at saturating bnAb concentrations, defined as maximal percentage inhibition (MPI), can also be measured. Here, we assess the distribution of the neutralization measurements (IC50, IC80, IC90, and IC95), correlations between bnAbs, and correlations with participant characteristics.
We include data from 8 clinical trials on PWH [1, 3–7] (Supplementary Table 1). These trials have different designs; thus, independent analyses were conducted for each. Trials with similar samples and data were combined to evaluate the associations between (1) clinical characteristics and (2) assay failure and susceptibility results. Wilcoxon rank sum and Kruskal-Wallis tests were used to assess associations between measures. The rank-based Spearman correlation test was used to assess correlations in neutralization measurements between time points, bnAbs, assays, and continuous clinical markers; partial Spearman correlations were used to adjust for other clinical markers. The Fisher exact test evaluated associations between categorical variables. Cohen κ was used to estimate the consistency of assay success between time points. For combined analyses of assay failure and susceptibility, generalized linear regression modeling was used; for results from PBMC samples, repeated measures assuming exchangeable correlation structure were used to include both time points for A5257. The combined analyses are univariate due to the variation in data availability across trials. Analyses were not adjusted for multiple comparisons due to their exploratory nature. All analyses were conducted with SAS version 9.4 (SAS Institute). All studies were approved by the University of Pennsylvania Institutional Review Board.
RESULTS
Clinical Demonstration Study
To characterize the reproducibility of longitudinal PhenoSense mAb susceptibility measurements across distinct biospecimens, we sought a recently completed study that collected plasma and PBMCs before and during antiretroviral therapy (ART), respectively, and had defined clinical data. We identified A5257: a US-based phase 3 randomized clinical trial comparing 3 combination antiretroviral regimens for treatment of participants who were ART naive [7]. Among 295 participants who (1) achieved viral suppression by week 24, (2) remained suppressed through week 144, and (3) had samples available for testing, we randomly selected 69 participants for analysis. The PhenoSense assay was performed on pre-ART plasma samples and PBMCs collected following 1 and 3 years of ART. Each sample was tested for neutralization susceptibility to each of 7 bnAbs targeting different Env epitopes:
CD4 binding site (CD4bs): VRC01, VRC07.523LS, 3BNC117, N6
V3 glycan supersite: 10-1074
V2 apex: CAP256-VRC26.25
gp41 membrane proximal external region: 10E8
The A5257 study was conducted between 2008 and 2014; thus, stored samples were up to 9 years old when processed for these analyses. Of the 69 participants, 64 had at least 1 PhenoSense neutralization measurement. Baseline demographics for these 64 participants with evaluable data are shown in Table 1. Assay success rate is primarily dependent on the amplification of virus env sequences from participant samples and was successful in 59 (85%) pre-ART plasma samples, 61 (88%) year 1 PBMCs, and 63 (91%) year 3 PBMCs. There is strong evidence that intraparticipant assay success is consistent across time points (κ = 0.68–0.74, P < .001). For the 10 participants who had nonreportable results from pre-ART plasma samples, 8 (80%) also had at least 1 nonreportable result from year 1 or year 3 PBMCs.
Table 1.
Demographics of Study A5257 Participants
| Characteristic | No. (%) |
|---|---|
| No. of participants | 64 |
| Age, y | |
| Mean (SD) | 37 (10) |
| Median (IQR) | 37 (30–46) |
| Range | 20–61 |
| Sex at birth | |
| Female | 8 (12) |
| Male | 56 (88) |
| Race and ethnicity | |
| Black non-Hispanic | 17 (27) |
| Hispanic regardless of race | 9 (14) |
| Other | 4 (6) |
| White Non-Hispanic | 34 (53) |
| Pre-ART CD4 count, cells/mm3 | |
| Mean (SD) | 350 (198) |
| Median (IQR) | 350 (245–438) |
| Range | 9–1087 |
| Pre-ART HIV-1 RNA, copies/mL | |
| 1000–9999 | 11 (17) |
| 10 000–99 999 | 39 (61) |
| 100 000–500 000 | 12 (19) |
| >500 000 | 2 (3) |
Abbreviation: ART, antiretroviral therapy.
Each clinical sample was tested for susceptibility to inhibition by 7 bnAbs, with concentrations to inhibit virus infection by 50% to 95% (IC50 to IC95; Figure 1A, Supplementary Tables 2–4). Overall, IC50s across participant samples ranged 1000- to 10 000-fold for each bnAb, and within-participant pre-ART plasma and year 1 and 3 PBMC values were highly correlated (r = 0.62–1, P < .001 for all comparisons), as shown in Figure 1C for VRC01 and Supplementary Table 5 for all bnAbs. The PhenoSense inhibitory concentration results were generally concordant with published in vitro susceptibility of panels of Envs tested by the TZM.bl assay [12, 13]. Among the CD4bs bnAbs tested against pre-ART plasma, VRC01 was the least potent, with a median IC50 of 0.758 μg/mL. N6, VRC07523.LS, and 3BNC117 were more potent, with medians of 0.243, 0.135, and 0.103 μg/mL, respectively, and varying ranges of highly resistant samples. The most overt resistance was observed with 3BNC117 (10% with IC50 >50 μg/mL). Other bnAb classes performed as previously reported. The V3 glycan supersite targeting bnAb 10-1074 was generally the most potent bnAb, exhibiting a median IC50 of 0.047 μg/mL, with a higher proportion of overt resistance (19% with IC50 >50 μg/mL). 10E8, which targets the membrane proximal external region, was less potent (median IC50, 0.921 μg/mL) but broadly reactive, with detectable activity against 98% of samples. Finally, the clade B viruses of A5257 participants were largely resistant to CAP256-VRC26.25, which targets a V2 epitope observed more frequently in clade C viruses [14, 15].
Figure 1.
In vitro neutralizing susceptibility to tested bnAbs among A5257 participants. Inhibitory concentrations >50 μg/mL are plotted at 50 μg/mL and represent results above the upper assay limit. A, IC50 of pre-ART plasma. B, IC80 of pre-ART plasma. C, Correlations between samples for IC50 to VRC01. D, Correlations between IC50s of the 7 bnAbs based on pre-ART plasma samples. Correlations were highest between the 4 CD4 binding-site bnAbs (VRC01, VRC07523.LS, 3BNC117, N6) and between CD4 binding-site bnAbs and 10-1074. ART, antiretroviral therapy; bnAb, broadly neutralizing antibody; IC50, 50% inhibitory concentration; IC80, 80% inhibitory concentration; MPER, membrane proximal external region; nAb, neutralizing antibody; PBMC, peripheral blood mononuclear cell.
A cross-comparison of bnAb inhibition demonstrated strong correlations between IC50s within the CD4bs bnAb class, with Spearman correlations ranging from 0.71 to 0.86 (P < .001) for all comparisons (Figure 1D, Supplementary Tables 6A–C). There were significant yet weaker correlations between CD4bs bnAbs and 10-1074 (Spearman ρ = 0.20–0.40 for pre-ART, P = .122 for VRC01, and P < .05 for other CD4bs bnAbs). Similar trends were observed for IC80, IC90, and IC95 values (data not shown).
When assessed for predictors of bnAb susceptibility, PhenoSense measures did not significantly correlate with any tested participant demographics, pre-ART HIV-1 plasma viral load, or ART regimen (Supplementary Table 7).
Retrospective Assessments of Clinical Trials Evaluating VRC01, 3BNC117, and 10-1074 in the Setting of Analytic Treatment Interruption
A5340
A5340 was a phase 1 trial testing VRC01 in 14 PWH with ART suppression, of whom 9 participants had available baseline PBMCs and 5 also had pre-ART plasma. Participants were administered multiple doses of VRC01, and ART was interrupted 1 week after the first bnAb infusion. Administration of VRC01 was associated with a clinically modest delay in time to rebound (median, 4 weeks; range, 2–11) as compared with historical controls. Single-genome sequencing–derived Env pseudoviruses from pre-ART and rebound plasma tested in the TZM.bl assay demonstrated prevalent preexisting and rebound Env resistance to VRC01.
Demographics for the 9 studied participants are shown in Table 2. In the 5 participants who had available pre-ART plasma samples, the PhenoSense results from entry PBMCs and published pre-ART plasma single-genome sequencing–derived Env pseudoviruses in the TZM.bl assay were highly correlated (Spearman ρ = 0.9, P = .04; Figure 2A). Among 9 participants with entry PBMCs, PhenoSense IC50s to VRC01 did not significantly correlate with time to rebound (Spearman ρ = −0.35, P = .37). It is notable that the 2 participants who maintained suppression at the 8-week primary end point had greater susceptibility. Furthermore, the respective IC50 and IC80 cutoffs of >0.5 and >1 μg/mL were associated with rapid rebound (<8 weeks; all within 2–5 weeks) vs delayed (>8 weeks; P = .028, Fisher exact test), suggesting a possible relationship between a threshold of baseline VRC01 susceptibility and duration of virus suppression.
Table 2.
Demographics of Study A5340, MCA-866, and MCA-906 Participants
| Study, No. (%) or Median (IQR) | |||
|---|---|---|---|
| Characteristic | A5340 | MCA-866 | MCA-906 |
| No. of participants | 9 | 13 | 10 |
| Age, y | 43 (36–47) | 39 (30–50) | 38 (30–43) |
| Sex at birth | |||
| Female | 0 (0) | 1 (8) | 1 (10) |
| Male | 9 (100) | 12 (92) | 9 (90) |
| Race and ethnicity categorized | |||
| Black Non-Hispanic | 3 (33) | 6 (46) | 4 (40) |
| Hispanic regardless of race | 2 (22) | 3 (23) | 4 (40) |
| Other | 0 (0) | 1 (8) | 1 (10) |
| White non-Hispanic | 4 (44) | 3 (23) | 1 (10) |
| CD4 count, cells/mm3 | |||
| Entry | 874 (599–1026) | 693 (560–896) | 663 (583–736) |
| Reported nadir | 343 (290–425) | 372 (300–400) | 475 (350–600) |
Figure 2.
Correlation between susceptibility measures and clinical outcomes: A, A5340; B, MCA-866; C, MCA-906. ATI, analytic treatment interruption; IC50, 50% inhibitory concentration; IC90, 90% inhibitory concentration; nAb, neutralizing antibody; PBMC, peripheral blood mononuclear cell; SGS, single-genome sequencing.
MCA-866 and MCA-906
MCA-866 (NCT02588586) was an exploratory phase 2 study evaluating repeated doses of the CD4bs bnAb 3BNC117 in PWH with ART suppression [5]. Participants were administered up to 4 infusions at study weeks 0, 12, 24, and 27, and ART was discontinued 2 days after the week 24 infusion. After analytical treatment interruption (ATI), virus rebound occurred within a median 5.5 weeks (range, 2–17), which was statistically more delayed than in historical controls. This study conducted enrollment without prescreening for bnAb susceptibility, although Q2VOA [5], a virus outgrowth and sequencing assay, was performed retrospectively. As previously reported [5], in this small cohort, the 3BNC117 IC50 of the most resistant Q2VOA outgrowth virus for each participant correlated with the time to rebound (R = −0.66; as derived from the presented R2 = 0.44, P = .04, n = 10).
From the phase 1b study MCA-906 (NCT02825797), we assessed available samples from 10 participants who received infusions of 3BNC117 and 10-1074 at study weeks 0, 3, and 6, and ART was interrupted 2 days after the first dose [4]. For MCA-906, virus isolates from baseline PBMCs were used to prospectively screen for resistance to the tested bnAbs. Participants with IC50 >2 μg/mL for 3BNC117 or 10-1074 were excluded from the study (52% of those screened were excluded). As in MCA-866, Q2VOA was performed retrospectively on baseline PBMC samples.
PBMCs collected at study entry from 13 participants of MCA-866 and 10 participants of MCA-906 were evaluated for 3BNC117 and 10-1074 susceptibility by the PhenoSense mAb assay. Baseline demographics of these participants are shown in Table 2. No demographic characteristics correlated with bnAb susceptibility (Supplementary Table 8). For the 23 participants (10 of whom were enrolled after baseline screening), median IC50 values for 3BNC117 and 10-1074 were 0.04 and 0.016 μg/mL, respectively, which were statistically more susceptible than year 1 PBMCs from A5257 participants (P < .0001).
As seen in A5257 analyses, strong correlations were observed between CD4bs bnAb susceptibilities (r = 0.71–0.92, P < .001) and between CD4bs bnAbs and 10-1074 susceptibilities (r = 0.45–0.54, P < .05; Supplementary Table 9). Baseline susceptibility determined in the PhenoSense mAb assay did not correlate with time to rebound for either the MCA-866 participants (median time to rebound, 4 weeks; P ≥ .34) or MCA-906 participants (median time to rebound, 19.5 weeks; P ≥ .44; Figure 2).
Feasibility in Recent Clinical Trials
To understand the impact of PhenoSense mAb testing on study feasibility, we describe the experience of incorporating this assay into screening procedures in 4 recently completed bnAb trials: the UCSF-amfAR study, conducted at the University of California, San Francisco; the eCLEAR and TITAN studies, conducted at Aarhus University; and the A5357 study, conducted at multiple ACTG sites (Advancing Clinical Therapeutics Globally). Notably, the inhibitory concentration levels for the 50%, 75%, and 90% threshold cutoffs to tested bnAbs (Supplementary Table 10) were shared with the study teams and the enrollment criteria of these trials were informed by the A5257 analytic data.
The UCSF-amfAR study (NCT04357821) is a single-arm study that enrolled 10 PWH with suppressive ART who received combination immunotherapy, including bnAbs 10-1074 and VRC07523.LS [16]. As a part of eligibility determination, potential participants (n = 27) were screened with the PhenoSense mAb assay by small batches of recently collected and stored PBMCs. Enrollment was excluded according to decreased bnAb susceptibility, defined as IC50 above the 90th percentile of bnAb measurements for at least 1 of the antibodies (10-1074, 50 μg/mL; VRC07523.LS, 0.3834 μg/mL) or IC50 above the 75th percentile cutoff for both antibodies (10-1074, 0.1386 μg/mL; VRC07523.LS, 0.2024 μg/mL). In cases in which a result could not be generated, participants were enrolled.
Demographics of the 27 screened participants are shown in Table 3. PhenoSense mAb assay results were generated for 20 (74%) on the first attempt; 1 of 7 samples was successfully tested upon repeating the assay (78%). The median time to assay result, defined as the date of specimen receipt to the date of result, was 4 weeks, with all results completed within 7 weeks.
Table 3.
Demographics of Study USCF-amfAR, TITAN, eCLEAR, and A5357 Participants
| Study, No. (%) and Median (IQR) | ||||
|---|---|---|---|---|
| Characteristic | UCSF-amfAR | TITAN | eCLEAR | A5357 |
| No. of participants | 27 | 116 | 59 | 137 |
| Age, y | 44 (34–53) | 50 (42–56) | 36 (28–47) | 52 (45–59) |
| Sex at birth | ||||
| Female | 0 (0) | 16 (14) | 5 (8) | 28 (22) |
| Male | 27 (100) | 100 (86) | 54 (92) | 102 (78) |
| Race/ethnicity categorized | ||||
| Black Non-Hispanic | 0 (0) | 10 (9) | 3 (5) | 52 (38) |
| Hispanic regardless of race | 7 (26) | 0 (0) | 0 (0) | 21 (15) |
| Other | 4 (15) | 10 (9) | 8 (13) | 5 (4) |
| White Non-Hispanic | 16 (59) | 96 (83) | 48 (81) | 59 (43) |
| Reported nadir CD4 count, cells/mm3 | 448 (357–548) | 305 (189–427) | ||
eCLEAR and TITAN Studies
The TITAN study (NCT03837756) tested combination immunotherapy with the Toll-like receptor 9 agonist lefitolimod and 2 bnAbs, 3BNC117 and 10-1074, in the context of an ATI in PWH with ART suppression [1]. The PhenoSense mAb assay was conducted with PBMCs collected from 116 participants at screening visits, requiring IC90 <1.5 μg/mL for 3BNC117 and IC90 <2 μg/mL for 10-1074, as well as MPI ≥98% for both bnAbs, to enter the study; participants without results were included. Of the 116 screened participants, 86 (74%) had successful assay results. The median time to assay completion was 5.7 weeks, with all results obtained by 15 weeks. Despite prescreening, 3 participants in the bnAb group rebounded within 8 weeks of ATI, and sequencing of env showed resistance to 10-1074. In addition, 2 participants with resistance to 3BNC117 based on PhenoSense mAb testing remained suppressed throughout the 25 weeks of ATI.
The eCLEAR study (NCT03041012) tested the latency reversal agent romidepsin and the bnAb 3BNC117 in 59 PWH at the time of initiating ART [3]. PhenoSense mAb testing was not required for enrollment but was performed retrospectively on pre-ART plasma virus (n = 44) or PBMCs (n = 15). The PhenoSense mAb results were considered in primary trial analyses [2, 3] as described, with the effects of the bnAb therapy shown to be more substantial in participants with baseline susceptibility to 3BNC117 (defined as IC90 <1.5 μg/mL and MPI >98%). Of the 59 participants, 46 (78%) were successfully reported according to the first sample. A subset of assay failures was retested with a second sample, of which only 2 of 9 were successful (81% overall success rate). The median time to results was 10.3 weeks, with all results completed within 18.3 weeks. Since PhenoSense mAb testing was performed retrospectively and in batches, the turn around time is not reflective of prospective testing.
ACTG A5357
A5357 (NCT03739996) is a phase 2, single-arm, open-label switch study testing the combination of the injectable antiretroviral cabotegravir plus the bnAb VRC07523.LS in PWH with ART suppression. Eligibility criteria for A5357 included PhenoSense mAb testing of PBMCs at a screening visit with IC50 ≤0.25 μg/mL and an MPI >98% for VRC07523-LS. Of 137 participants with PBMC samples submitted for the PhenoSense mAb assay, 111 (81%) had successful results with the first sample and an additional 4 of 9 upon repeat testing. In total, 115 (84%) participants obtained assay results, with a median turn around time of 4.3 weeks.
Table 4 summarizes PhenoSense results for baseline samples from UCSF-amfAR, TITAN, eCLEAR, and A5357. Of note, 51% of participants in eCLEAR and 30% in TITAN harbored non–B subtype viruses. According to trial-specific thresholds, between 38% and 83% of screened individuals met bnAb susceptibility eligibility criteria.
Table 4.
PhenoSense Results for Baseline Samples of UCSF-amfAR, TITAN, eCLEAR, and A5357 Studies
| Study, No. (%) or Median (IQR) | ||||
|---|---|---|---|---|
| Parameter | UCSF-amfAR | TITAN | eCLEAR | A5357 |
| Samples tested | 27 | 116 | 59 | 137 |
| Time to results, wk | 4 (2–4) | 5.7 (4.8–9.0) | 10.3a (8.3–14.7) | 4.3 (3.4–5.0) |
| Samples with results | ||||
| Initial testing | 20 (74) | 86 (74) | 46 (78) | 111 (81) |
| After retesting subset without resultsb | 21 (78) | No retesting | 48 (81) | 115 (84) |
| Participants meeting bnAb eligibility criteria or susceptibility definition among those with results | ||||
| Initial testing | 7 (35) | 39 (45) | 26 (57) | 92 (83) |
| After retesting subset without results | 8 (38) | No retesting | 27 (56) | 96 (83) |
Abbreviation: bnAb, broadly neutralizing antibody.
aPhenoSense assay testing for eCLEAR was performed retrospectively. Assay turn around time is not representative of prospective testing to enable trial enrollment.
bThe numbers of participants who were retested among the initial assay failures are 1 (14%) for UCSF-amfAR, 9 (69%) for eCLEAR, and 9 (35%) for A5357.
Combined Analyses of Assay Success Rates and Susceptibilities
To assess the performance of the PhenoSense mAb assay across multiple studies, we analyzed assay success rates and associated clinical factors from studies with similar sample types. We compiled the assay success rates for the first tested PBMC sample in A5257 (year 1), A5340, TITAN, UCSF-amfAR, and A5357. Across the 356 PBMC samples, there was an 80.1% assay success rate (Table 5). In a comparison of participant characteristics between those with assay success and failure, higher pre-ART viral load (odds ratio, 1.31 per log10 copies/mL higher; P = .0094) and longer time since HIV diagnosis (odds ratio, 1.86 per 10 years longer; P = .0031) were associated with assay success (Supplementary Table 11). Assessing studies that tested pre-ART plasma (A5257 and eCLEAR, n = 128) revealed an 82% assay success rate. Using the combined PBMC and plasma sample data sets, we found no clinically meaningful associations between participant characteristics and bnAb susceptibility (Supplementary Table 12).
Table 5.
Pretreatment Characteristics by Assay Failure or Success From Peripheral Blood Mononuclear Cells Only, Combining A5257 Year 1, A5340, UCSF-amfAR, TITAN, and A5357 Studies
| Characteristic | Assay Failure | Assay Success | Total |
|---|---|---|---|
| Participants, No. (%) | 71 (19.9) | 285 (80.1) | 356 |
| Pre-ART HIV-1 RNA (log10), copies/mL | |||
| No. | 45 | 172 | 217 |
| Median (IQR) | 3.52 (1.28–4.99) | 4.38 (2.55–5.03) | 4.31 (1.59–5.02) |
| Range | 1.28–7.00 | 1.28–7.38 | 1.28–7.38 |
| Nadir CD4, cells/mm3 | |||
| No. | 39 | 195 | 234 |
| Median (IQR) | 331 (239–477) | 335 (210–448) | 333 (217–448) |
| Range | 79–1017 | 5–983 | 5–1017 |
| Years undergoing ART | |||
| No. | 45 | 173 | 218 |
| Median (IQR) | 7 (2–12) | 5 (1–11) | 5 (1–11) |
| Range | 1–26 | 1–28 | 1–28 |
| Years since HIV diagnosis | |||
| No. | 42 | 206 | 248 |
| Median (IQR) | 10.5 (5–16) | 12 (8–20) | 12 (8–19) |
| Range | 1–26 | 0–39 | 0–39 |
| Years from HIV diagnosis to ART initiation | |||
| No. | 38 | 103 | 141 |
| Median (IQR) | 0 (0–2) | 0 (0–3) | 0 (0–2) |
| Range | 0–8 | 0–21 | 0–21 |
Abbreviation: ART, antiretroviral therapy.
DISCUSSION
Clinical testing of bnAbs for HIV therapy and cure has shown promise [1, 3, 4, 17, 18], with bnAb efficacy generally correlating with neutralization susceptibility [2, 3, 5, 6, 19]. Thus, an accurate and high-throughput assay for prospective bnAb susceptibility screening remains of high value [20]. Here, the CLIA-validated PhenoSense mAb assay demonstrated robust reproducibility across biospecimen types and clinical trial designs and correlated well with published research assays; however, it was inconsistently predictive of clinical outcomes and added time to study enrollment processes.
When tested on PBMC and plasma samples from A5257, the PhenoSense mAb assay proved robust with >88% assay success, exhibiting tight correlations between plasma and cell sample types and across 2 years of suppressive ART. The PhenoSense mAb assay had a marginally lower success rate in other clinical trials with varied designs, with a combined success rate of approximately 80%. Higher viral load at ART initiation was associated with assay success, but other markers of larger reservoirs were not (eg, nadir CD4 count, time from HIV diagnosis to ART initiation). The consistent intraparticipant assay success rates across sample types and the observation that repeat attempts were rarely successful across multiple studies indicate that participant-specific factors (eg, sequence composition of virus populations) may challenge this and other screening approaches. For implementation purposes, an expected assay failure rate of approximately 20% should be incorporated into study designs, and investigators should consider the limited value of repeat testing.
PhenoSense mAb assay results exhibited concordance with published bnAb susceptibility, as assessed by other laboratory-based assays used to characterize individual bnAbs against global panels of pseudoviruses [21], thus supporting PhenoSense mAb screening [22, 23]. While the combined clinical cohorts evaluated here afforded an opportunity to assess for clinical or demographic predictors of baseline bnAb susceptibility, we did not find evidence for clinically relevant correlations. These data suggest that bnAb resistance is unlikely to be predicted without direct sampling and testing of archived viruses.
When applied retrospectively to published studies, the PhenoSense mAb assay was unable to consistently predict the duration of virus suppression. Given the assay's reproducibility and correlation with laboratory-based assays, we posit 2 possible explanations for the disconnect with clinical outcomes. First, the poor predictive capacity may be due to incomplete sampling of diverse HIV reservoirs. In PWH with chronic ART initiation, the archived reservoir is substantial, highly diverse, and spread throughout the body [24–27], and viruses that reactivate at ATI are often minor variants from within that heterogenous virus population [5, 28, 29]. Indeed, no readily available biospecimen allows for complete sampling of minor populations of provirus or access to tissue reservoirs [25]. Notably, baseline PhenoSense measures did correlate with immunomodulatory bnAb activity in the eCLEAR study [2, 3], where a substantial proportion (47%) of participants were sampled during early HIV, when virus diversity is far more limited. Alternatively, virus susceptibility to bnAbs may be just one of many factors that determine bnAb clinical efficacy in most studies. Indeed, many factors outside of neutralization susceptibility have been suggested to inform bnAb clinical activity and propensity for rebound viremia at ATI, including contributors to virus reactivation (eg, inflammatory biomarkers [30–34], reservoir size and activity [35–37], virus reactivation rates [38–42], bnAb pharmacokinetics [43, 44] and activity within tissues [45, 46], and the effects of host immune pressures [28, 47, 48]). In addition, the trials studied to date overwhelmingly enrolled men from the global north; as trial diversity expands, assay performance will need to be reassessed in women and expanded virus subtypes.
Experience in recent studies indicates that regardless of clinical utility, implementation of the PhenoSense assay can affect study feasibility. The ∼20% assay failure rates and median assay turn around times of 4 to 6 weeks with the exploratory status of the PhenoSense mAb should be considered in future study protocols. We note that testing for some of these studies occurred during the COVID-19 pandemic, which influenced assay processing times; turn around time has improved postpandemic. In addition, factors outside the assay itself, including infrequent batch scheduling, inadequate sample volume, incomplete specimen data, and shipping or handling problems, contributed to assay success rates and turn around time.
As bnAbs become more realistic options in HIV therapy and cure, baseline resistance remains a primary concern, and the need to understand the utility of, and clinical cutoffs for, available screening assays is paramount. Here, we show that the PhenoSense assay, the only currently CLIA-validated assay available for use in clinical trials, is reproducible and correlates with bnAb susceptibility as determined by other laboratory assays. The clinical application of the assay has been heterogeneous to date, with distinct exploratory cutoffs and different study designs. The field has not yet defined validated clinical cutoffs for specific bnAbs in any therapeutic context, and our retrospective analyses of the assay similarly show inconsistent predictive capacity. Together, the strengths, limitations, and feasibility considerations presented here should aid clinical investigators in designing future bnAb efficacy studies.
Supplementary Data
Supplementary materials are available at The Journal of Infectious Diseases online (http://jid.oxfordjournals.org/). Supplementary materials consist of data provided by the author that are published to benefit the reader. The posted materials are not copyedited. The contents of all supplementary data are the sole responsibility of the authors. Questions or messages regarding errors should be addressed to the author.
Supplementary Material
Contributor Information
Marie Høst Pahus, Department of Infectious Diseases, Aarhus University Hospital, Denmark.
Yu Zheng, Chan School of Public Health, Harvard University, Boston, Massachusetts.
Maxine Olefsky, Chan School of Public Health, Harvard University, Boston, Massachusetts.
Jesper Damsgaard Gunst, Department of Infectious Diseases, Aarhus University Hospital, Denmark.
Pablo Tebas, Division of Infectious Disease, Department of Medicine, University of Pennsylvania, Philadelphia.
Babafemi Taiwo, Division of Infectious Disease, Department of Medicine, Northwestern University, Chicago, Illinois.
Ole S Søgaard, Department of Infectious Diseases, Aarhus University Hospital, Denmark.
Michael J Peluso, Division of HIV, Infectious Diseases and Global Medicine, Department of Medicine, University of California San Francisco.
Yolanda Lie, Labcorp-Monogram Biosciences, South San Francisco, California.
Jacqueline D Reeves, Labcorp-Monogram Biosciences, South San Francisco, California.
Christos J Petropoulos, Labcorp-Monogram Biosciences, South San Francisco, California.
Marina Caskey, Department of Clinical Investigation, Rockefeller University, New York City, New York.
Katharine J Bar, Division of Infectious Disease, Department of Medicine, University of Pennsylvania, Philadelphia.
Notes
Acknowledgments. We thank Kristi Strommen, Elizabeth Anton, Tim Persyn, and the Labcorp-Monogram Biosciences Clinical Reference Lab and colleagues for PhenoSense mAb assay development, testing, and project management for these studies.
Disclaimer. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Financial support. This work was supported by the National Institute of Allergy and Infectious Diseases of the National Institutes of Health under awards UM1 AI068634, UM1 AI068636, and UM1 AI106701; the Statistical and Data Management Center of the Advancing Clinical Therapeutics Globally under National Institute of Allergy and Infectious Diseases grant UM1 AI068634; the REACH Delaney (UM1 AI164565, 5U01AI169385, ERC CFAR P30AI124414); the Lundbeck Foundation (grants R313-2019-790 and R381-2021-1405); the BEAT-HIV Delaney (UM1AI164570, R01AI162646, U01AI169767, P01AI131338); and Penn Center for AIDS Research (P30AI045008).
References
- 1. Gunst JD, Højen JF, Pahus MH, et al. Impact of a TLR9 agonist and broadly neutralizing antibodies on HIV-1 persistence: the randomized phase 2a TITAN trial. Nat Med 2023; 29:2547–58. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Rosás-Umbert M, Gunst JD, Pahus MH, et al. Administration of broadly neutralizing anti-HIV-1 antibodies at ART initiation maintains long-term CD8(+) T cell immunity. Nat Commun 2022; 13:6473. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Gunst JD, Pahus MH, Rosás-Umbert M, et al. Early intervention with 3BNC117 and romidepsin at antiretroviral treatment initiation in people with HIV-1: a phase 1b/2a, randomized trial. Nat Med 2022; 28:2424–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Mendoza P, Gruell H, Nogueira L, et al. Combination therapy with anti-HIV-1 antibodies maintains viral suppression. Nature 2018; 561:479–84. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Cohen YZ, Lorenzi JCC, Krassnig L, et al. Relationship between latent and rebound viruses in a clinical trial of anti-HIV-1 antibody 3BNC117. J Exp Med 2018; 215:2311–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Bar KJ, Sneller MC, Harrison LJ, et al. Effect of HIV antibody VRC01 on viral rebound after treatment interruption. N Engl J Med 2016; 375:2037–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Lennox JL, Landovitz RJ, Ribaudo HJ, et al. Efficacy and tolerability of 3 nonnucleoside reverse transcriptase inhibitor-sparing antiretroviral regimens for treatment-naive volunteers infected with HIV-1: a randomized, controlled equivalence trial. Ann Intern Med 2014; 161:461–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Petropoulos CJ, Parkin NT, Limoli KL, et al. A novel phenotypic drug susceptibility assay for human immunodeficiency virus type 1. Antimicrob Agents Chemother 2000; 44:920–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Havlir DV, Hellmann NS, Petropoulos CJ, et al. Drug susceptibility in HIV infection after viral rebound in patients receiving indinavir-containing regimens. JAMA 2000; 283:229–34. [DOI] [PubMed] [Google Scholar]
- 10. Martinez-Picado J, DePasquale MP, Kartsonis N, et al. Antiretroviral resistance during successful therapy of HIV type 1 infection. Proc Natl Acad Sci U S A 2000; 97:10948–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Genebat M, Ruiz-Mateos E, Leon JA, et al. Correlation between the Trofile test and virological response to a short-term maraviroc exposure in HIV-infected patients. J Antimicrob Chemother 2009; 64:845–9. [DOI] [PubMed] [Google Scholar]
- 12. Sarzotti-Kelsoe M, Bailer RT, Turk E, et al. Optimization and validation of the TZM-bl assay for standardized assessments of neutralizing antibodies against HIV-1. J Immunol Methods 2014; 409:131–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Gama L, Koup RA. New-generation high-potency and designer antibodies: role in HIV-1 treatment. Annu Rev Med 2018; 69:409–19. [DOI] [PubMed] [Google Scholar]
- 14. Doria-Rose NA, Bhiman JN, Roark RS, et al. New member of the V1V2-directed CAP256-VRC26 lineage that shows increased breadth and exceptional potency. J Virol 2016; 90:76–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Moore PL, Gray ES, Sheward D, et al. Potent and broad neutralization of HIV-1 subtype C by plasma antibodies targeting a quaternary epitope including residues in the V2 loop. J Virol 2011; 85:3128–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Peluso MJ, Deitchman A, Magombedze G, et al. Rebound dynamics following immunotherapy with an HIV vaccine, TLR9 agonist, and bNAbs. In: Conference on Retroviruses and Opportunistic Infections. Seattle, WA: CROI, 2023.
- 17. Sneller MC, Blazkova J, Justement JS, et al. Combination anti-HIV antibodies provide sustained virological suppression. Nature 2022; 606:375–81. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Gaebler C, Nogueira L, Stoffel E, et al. Prolonged viral suppression with anti-HIV-1 antibody therapy. Nature 2022; 606:368–74. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Corey L, Gilbert PB, Juraska M, et al. Two randomized trials of neutralizing antibodies to prevent HIV-1 acquisition. N Engl J Med 2021; 384:1003–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Lynch RM, Bar KJ. Development of screening assays for use of broadly neutralizing antibodies in people with HIV. Curr Opin HIV AIDS 2023; 18:171–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Wagh K, Bhattacharya T, Williamson C, et al. Optimal combinations of broadly neutralizing antibodies for prevention and treatment of HIV-1 clade C infection. PLoS Pathog 2016; 12:e1005520. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Tebas P, Lynn K, Azzoni L, et al. Susceptibility to 3BNC117 and 10-1074 in ART suppressed chronically infected persons. AIDS 2023; 37:1203–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Tebas P, Azzoni L, Papasavvas E, et al. BEAT2 primary trial outcomes: PEGIFN-a2b, 3BNC117 and 10-1074 in chronic HIV infection. In: Conference on Retroviruses and Opportunistic Infections. Seattle, WA: CROI, 2023.
- 24. Siliciano JD, Siliciano RF. In vivo dynamics of the latent reservoir for HIV-1: new insights and implications for cure. Annu Rev Pathol 2022; 17:271–94. [DOI] [PubMed] [Google Scholar]
- 25. Abdel-Mohsen M, Richman D, Siliciano RF, et al. Recommendations for measuring HIV reservoir size in cure-directed clinical trials. Nat Med 2020; 26:1339–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Falcinelli SD, Ceriani C, Margolis DM, Archin NM. New frontiers in measuring and characterizing the HIV reservoir. Front Microbiol 2019; 10:2878. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Estes JD, Kityo C, Ssali F, et al. Defining total-body AIDS-virus burden with implications for curative strategies. Nat Med 2017; 23:1271–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Salantes DB, Zheng Y, Mampe F, et al. HIV-1 latent reservoir size and diversity are stable following brief treatment interruption. J Clin Invest 2018; 128:3102–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Lu CL, Pai JA, Nogueira L, et al. Relationship between intact HIV-1 proviruses in circulating CD4(+) T cells and rebound viruses emerging during treatment interruption. Proc Natl Acad Sci U S A 2018; 115:E11341–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Bartsch YC, Loos C, Rossignol E, et al. Viral rebound kinetics correlate with distinct HIV antibody features. mBio 2021; 12:e00170-21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Offersen R, Yu WH, Scully EP, et al. HIV antibody Fc N-linked glycosylation is associated with viral rebound. Cell Rep 2020; 33:108502. [DOI] [PubMed] [Google Scholar]
- 32. Giron LB, Papasavvas E, Yin X, et al. Phospholipid metabolism is associated with time to HIV rebound upon treatment interruption. mBio 2021; 12:e03444-20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Giron LB, Papasavvas E, Azzoni L, et al. Plasma and antibody glycomic biomarkers of time to HIV rebound and viral setpoint. AIDS 2020; 34:681–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Aid M, Colarusso A, Walker-Sperling V, Barouch DH. Peripheral blood biomarkers predict viral rebound following antiretroviral therapy discontinuation in SIV-infected, early ART-treated rhesus macaques. Cell Rep Med 2023; 4:101122. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Li JZ, Aga E, Bosch RJ, et al. Time to viral rebound after interruption of modern antiretroviral therapies. Clin Infect Dis 2022; 74:865–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Li JZ, Etemad B, Ahmed H, et al. The size of the expressed HIV reservoir predicts timing of viral rebound after treatment interruption. AIDS 2016; 30:343–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Kearney MF, Wiegand A, Shao W, et al. Origin of rebound plasma HIV includes cells with identical proviruses that are transcriptionally active before stopping of antiretroviral therapy. J Virol 2016; 90:1369–76. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. van Dorp CH, Conway JM, Barouch DH, Whitney JB, Perelson AS. Models of SIV rebound after treatment interruption that involve multiple reactivation events. PLoS Comput Biol 2020; 16:e1008241. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Hill AL, Rosenbloom DI, Goldstein E, et al. Real-time predictions of reservoir size and rebound time during antiretroviral therapy interruption trials for HIV. PLoS Pathog 2016; 12:e1005535. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Fennessey CM, Pinkevych M, Immonen TT, et al. Genetically-barcoded SIV facilitates enumeration of rebound variants and estimation of reactivation rates in nonhuman primates following interruption of suppressive antiretroviral therapy. PLoS Pathog 2017; 13:e1006359. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Pinkevych M, Kent SJ, Tolstrup M, et al. Modeling of experimental data supports HIV reactivation from latency after treatment interruption on average once every 5–8 days. PLoS Pathog 2016; 12:e1005740. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. Pinkevych M, Cromer D, Tolstrup M, et al. HIV reactivation from latency after treatment interruption occurs on average every 5–8 days—implications for HIV remission. PLoS Pathog 2015; 11:e1005000. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43. Gaudinski MR, Coates EE, Houser KV, et al. Safety and pharmacokinetics of the Fc-modified HIV-1 human monoclonal antibody VRC01LS: a phase 1 open-label clinical trial in healthy adults. PLoS Med 2018; 15:e1002493. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44. Mayer KH, Seaton KE, Huang Y, et al. Safety, pharmacokinetics, and immunological activities of multiple intravenous or subcutaneous doses of an anti-HIV monoclonal antibody, VRC01, administered to HIV-uninfected adults: results of a phase 1 randomized trial. PLoS Med 2017; 14:e1002435. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45. Reh L, Magnus C, Schanz M, et al. Capacity of broadly neutralizing antibodies to inhibit HIV-1 cell-cell transmission is strain- and epitope-dependent. PLoS Pathog 2015; 11:e1004966. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46. Abela IA, Berlinger L, Schanz M, et al. Cell-cell transmission enables HIV-1 to evade inhibition by potent CD4bs directed antibodies. PLoS Pathog 2012; 8:e1002634. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47. Gondim MVP, Sherrill-Mix S, Bibollet-Ruche F, et al. Heightened resistance to host type 1 interferons characterizes HIV-1 at transmission and after antiretroviral therapy interruption. Sci Transl Med 2021; 13:eabd8179. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48. Bertagnolli LN, Varriale J, Sweet S, et al. Autologous IgG antibodies block outgrowth of a substantial but variable fraction of viruses in the latent reservoir for HIV-1. Proc Natl Acad Sci U S A 2020; 117:32066–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.