First-in-human immunoPET imaging of HIV-1 infection using 89Zr-labeled VRC01 broadly neutralizing antibody

Abstract

A major obstacle to achieving long-term antiretroviral (ART) free remission or functional cure of HIV infection is the presence of persistently infected cells that establish a long-lived viral reservoir. HIV largely resides in anatomical regions that are inaccessible to routine sampling, however, and non-invasive methods to understand the longitudinal tissue-wide burden of HIV persistence are urgently needed. Positron emission tomography (PET) imaging is a promising strategy to identify and characterize the tissue-wide burden of HIV. Here, we assess the efficacy of using immunoPET imaging to characterize HIV reservoirs and identify anatomical foci of persistent viral transcriptional activity using a radiolabeled HIV Env-specific broadly neutralizing antibody, 89Zr-VRC01, in HIV-infected individuals with detectable viremia and on suppressive ART compared to uninfected controls (NCT03729752). We also assess the relationship between PET tracer uptake in tissues and timing of ART initiation and direct HIV protein expression in CD4 T cells obtained from lymph node biopsies. We observe significant increases in 89Zr-VRC01 uptake in various tissues (including lymph nodes and gut) in HIV-infected individuals with detectable viremia (N = 5) and on suppressive ART (N = 5) compared to uninfected controls (N = 5). Importantly, PET tracer uptake in inguinal lymph nodes in viremic and ART-suppressed participants significantly and positively correlates with HIV protein expression measured directly in tissue. Our strategy may allow non-invasive longitudinal characterization of residual HIV infection and lays the framework for the development of immunoPET imaging in a variety of other infectious diseases.

Conflict of interest statement

T.J.H. receives grant support from Gilead Biosciences, Merck and Bristol Myers Squibb. R.R.F. receives grant support from Fukushima SiC Applied Engineering. All other authors declare no competing interests.

© 2022. The Author(s).

Figures

Fig. 1. First-in-human immunoPET imaging of HIV persistence.

A schematic of PET-MR imaging workflow including tracer administration and imaging time points is shown in a. Imaging on day 6 was limited to a subset of participants to determine if meaningful information could be obtained over longer periods of time various tissues of interest. Representative 3-dimensional maximum intensity projection images (MR attenuation corrected, but without MR overlay) of 89Zr-VRC01 uptake 6 and 72 h in a viremic (V2) and uninfected control (C3) participant are shown (b).

Fig. 2. 89Zr-VRC01 uptake is preferentially increased in participants with HIV compared with uninfected controls.

a Representative axial PET-MR images with inset magnification of bilateral inguinal lymph nodes on imaging day 3 for a viremic (V), ART-suppressed (A) and uninfected control (C). b Axial image of persistent inguinal lymph node 89Zr-VRC01 uptake in inguinal lymph nodes on imaging day 0 (6 h post injection), and imaging day 3 (72 h post injection) (b). rSUVmax and rSUV mean values for c inguinal and d axillary lymph node uptake averaged across bilateral nodes at imaging day 0, day 1, and day 3. Participant A1 (green triangle) started ART approximately 3 months prior PET imaging (early treatment initiation). ART suppressed participants were imaged starting on day 3 as per study protocol. SUV standardized uptake value, rSUV blood pool-adjusted 89Zr-VRC01 standardized uptake value (tissue to blood pool SUV ratio); P values <0.05 and <0.01 represented by * and **, respectively from two-sided non-parametric analyses (Mann–Whitney tests for analyses involving two groups and Kruskal–Wallis tests for analyses involving three groups). Mean and standard error bars are shown. Control N = 4 on day 0 and day 1, viremic case N = 4 and 3 on day 0 and day 1, respectively (‡statistical analyses not performed on day 1 as one viremic participant missed this visit). Control and viremic and ART case N = 5 on day 3. Source data are provided as a Source Data file.

Fig. 3. 89Zr-VRC01 uptake is preferentially increased in gut tissue in participants with HIV compared with uninfected volunteers up to three days following injection.

a Representative coronal PET-MR images showing increased gut tracer uptake in participants with HIV on or off ART. First, gating was performed on coronal images across multiple bowel loops excluding obvious intraluminal stool signal which likely represented HIV non-specific 89Zr-VRC01 excretion (b). rSUV values for sigmoid/descending colon wall (c) and anorectal tissue (d) were obtained by gating exclusively on multiple sections of bowel wall to exclude intraluminal signal. SUV standardized uptake value; rSUV blood pool-adjusted 89Zr-VRC01 standardized uptake value (tissue to blood pool SUV ratio); P values <0.05 and <0.01 represented by * and **, respectively from two-sided non-parametric analyses (Mann–Whitney tests for analyses involving two groups and Kruskal–Wallis tests for analyses involving three groups). Mean and standard error bars are shown. Control N = 4 on day 0 and day 1, viremic case N = 4 and 3 on day 0 and day 1, respectively (‡statistical analyses not performed on day 1 as one viremic participant missed this visit). Control and viremic and ART case N = 5 on day 3. Source data are provided as a Source Data file.

Fig. 4. 89Zr-VRC01 uptake is preferentially increased in iliac and femoral bone marrow in participants with HIV compared with uninfected volunteers 3 days following injection.

a Representative axial PET-MR images showing tracer uptake in axial marrow in representative viremic, ART suppressed and uninfected participants. rSUVmax and rSUV mean values for b iliac and c femoral shaft bone marrow 89Zr-VRC01 uptake from 3-dimensional gating at imaging day 0, day 1, and day 3. Minimal uptake was observed in osseus (femoral shaft) bone at any time point in all cohorts (d). SUV standardized uptake value; rSUV blood pool-adjusted 89Zr-VRC01 standardized uptake value (tissue to blood pool SUV ratio); P values <0.05 and <0.01 represented by * and **, respectively from two-sided non-parametric analyses (Mann–Whitney tests for analyses involving two groups and Kruskal–Wallis tests for analyses involving three groups). Mean and standard error bars are shown. Control N = 4 on day 0 and day 1, viremic case N = 4 and 3 on day 0 and day 1, respectively (‡statistical analyses not performed on day 1 as one viremic participant missed this visit). Control and viremic and ART case N = 5 on day 3. Source data are provided as a Source Data file.

Fig. 5. 89Zr-VRC01 uptake in various tissues.

rSUVmax values for a nasal turbinates, b testes, c adipose tissue (right flank), d adductor muscles (thigh), e liver and f spleen from 3-dimensional gating at imaging day 0, day 1, and day 3 are shown. Representative coronal PET-MR images showing tracer uptake liver as shown in g. 89Zr-VRC01 uptake was higher in viremic and ART-suppressed participants compared with uninfected controls on imaging Day 3. In contrast, tracer uptake in the liver was higher in uninfected controls (D0) may represents decreased lymphoid tissue and bone marrow distribution compared with participants with HIV. SUV standardized uptake value; rSUV blood pool-adjusted 89Zr-VRC01 standardized uptake value (tissue to blood pool SUV ratio); P values <0.05 and <0.01 represented by * and **, respectively from two-sided non-parametric analyses (Mann–Whitney tests for analyses involving two groups and Kruskal–Wallis tests for analyses involving three groups). Mean and standard error bars are shown. Control N = 4 on day 0 and day 1, viremic case N = 4 and 3 on day 0 and day 1, respectively (‡statistical analyses not performed on day 1 as one viremic participant missed this visit). Control and viremic and ART case N = 5 on day 3. Source data are provided as a Source Data file.

Fig. 6. Associations between time on ART, HIV p24 protein expression in lymph node CD4 T cells obtained by tissue biopsy, and tracer uptake in various anatomical regions.

Correlations between 89Zr-VRC01 rSUVmean in lymph node and gut tissues on imaging day 3 compared with days on ART for viral suppressed participants are shown in a. b Correlations between 89Zr-VRC01 rSUVmax/mean and the frequency of blood and inguinal lymph node HIV Gag P24+ CD4+ T cells as measured by flow cytometry. r and P values represent results from Spearman correlation analyses. Overall, inverse associations between tracer uptake and ART duration and a significant positive correlation between inguinal lymph node 89Zr-VRC01 uptake and direct measures of HIV protein expression in CD4 T cells from inguinal lymph node sampling were observed. Source data are provided as a Source Data file.