A Single Dose Study of Pembrolizumab in People With HIV Infection Who are Immunologic Nonresponders

Abstract

Background

Some people with HIV (PWH) have a poor immunologic response to antiretroviral therapy (ART). Such immunologic nonresponders are at increased risk for HIV and non-HIV related complications. Immune exhaustion may contribute to poor immune reconstitution, and blockade of PD-1, an immune checkpoint molecule, may thus be beneficial.

Method

We undertook a phase 1, 3:1 randomized, placebo-controlled trial of a single 200 mg dose of pembrolizumab in PWH with CD4+ T-cell counts between 100 and 350 cells/mm³, who had been on ART and were virally suppressed for at least 12 months. The primary endpoint was the frequency of either Grade 3 or higher adverse events or Grade 2 or higher autoimmune events requiring corticosteroid therapy. Additional endpoints included levels of PD-1 expression, and changes in immune and virologic parameters.

Results

Of the 7 enrolled participants, 6 received pembrolizumab and 1 received placebo. The only grade 3 event, ophthalmic zoster, developed in the placebo recipient. There were no autoimmune events requiring corticosteroid therapy. CD4+ and CD8+ T-cell counts were stable over time, though both showed increased CD38 and HLA-DR expression. PD-1 expression declined from a baseline of 60.4% (SD, 7.4%) to a nadir of 0.8%–23.5% for CD4+ T cells, and from 45.2% (SD, 9.4%) to 0.1%–23.2% for CD8+ T cells. There were no significant changes in viral killing, plasma viral loads or proviral DNA levels.

Conclusions

A single dose of pembrolizumab may be safely administered to PWH who are immunologic nonresponders.

A placebo-controlled trial in 7 persons with HIV who had virologic control but poor immune reconstitution found that a single dose of pembrolizumab appeared safe but did not result in increased CD4+ T-cell numbers or changes in HIV killing or DNA levels.

A subset of people with HIV-1 (PWH) with a virologic response to antiretroviral therapy (ART), with plasma viral loads below detection limits of commercial assays for more than 1 year, nevertheless have a CD4+ T-cell count of <300-350 cells/mm³ [1–4]. Such immunologic nonresponders (INRs) are at increased risk for both HIV related and non-HIV related complications compared with immunologic responders, though they are at lower risk than patients with uncontrolled viremia [5–9]. Changing or intensifying ART regimens has not resulted in increased CD4+ T-cell counts [10]. Thus, novel approaches are needed to facilitate management of this patient population.

Cell-mediated immunity appears critical to controlling HIV infection [11–13]. Both CD4+ and CD8+ T-cells play important roles in controlling infections by many intracellular pathogens, including viruses such as HIV [12, 14, 15]. However, HIV avoids T-cell control through effects on the function of both CD4+ and CD8+ T cells, and by a high level of genetic diversity. One potential contributor to this limited efficacy is the development of immune exhaustion of HIV-specific CD8+ T cells. Immune exhaustion is mediated in part by upregulation of surface-expressed inhibitory molecules such as programmed cell death protein 1 (PD-1, also known as CD279), a member of the B7-CD28 family of co-signaling molecules [16, 17]. PD-1 can be expressed on CD4+ and CD8+ T cells, natural killer cells, B cells, and monocytes after activation. When PD-1 is engaged by one of its ligands, PD-L1, in conjunction with T-cell receptor activation, it down-modulates T-cell proliferation and effector functions. During normal activity, this serves to dampen potentially harmful excessive immune responses, including autoimmunity. However, as initially demonstrated in animal models, during chronic viral infection (eg, lymphocytic choriomeningitis virus), persistently high levels of PD-1 expression are associated with poor control of infection, and blockade of the PD-1/PD-L1 interaction resulted in improved T-cell function and better viral control [18–22]. In PWH, expression of lower levels of PD-1 on HIV-specific CD8+ T cells is associated with lower levels of virus and PD-1 blockade in vitro causes a modest increase in the proliferation of CD8+ T cells. However, the cause and effect relationship between immunologic control of HIV-1 and PD-1 expression on HIV-specific cells is uncertain given it is an activation marker, reduced by ART and increased by antigen exposure [23]. In INRs, PD-1 expression has been reported to be increased on CD4+ but not CD8+ T cells [24, 25]. Thus it remains possible that PD-1 blockade could mediate positive effects in INRs either by increasing proliferation and recovery of CD4+ T cells or increasing the function of CD8+ T cells.

Two randomized trials have treated PWH with antibodies that interfere with PD-1/PD-L1 signaling [26, 27]. Both enrolled PWH with CD4+ T-cell counts >350/mm³, and both were stopped early. Overall, no major immunologic changes were seen compared with placebo, although transient immunologic changes were seen in a minority of participants.

The current study was undertaken to evaluate the safety in INRs of a single dose of pembrolizumab (Keytruda™), a humanized monoclonal antibody that binds with high affinity to PD-1. In addition, the study examined the extent and duration of blockage of PD-1 on CD4+ and CD8+ T cells, as well as changes in various viral and other immune parameters.

METHODS

Study Design

The study was a phase 1 randomized, double-blind, placebo-controlled trial in adult PWH (≥18 years) who were randomized 3:1 using blocked randomization to receive either a single 200 mg dose of pembrolizumab or placebo via IV infusion. The 200 mg dose is FDA-approved for treatment of multiple cancers, and has been previously utilized in PWH [28]. Participants were followed at the NIH weekly for 4 weeks, then every 2–12 weeks through 48 weeks; follow-up via telephone calls were continued through 96 weeks. Blood for clinical and research evaluations was collected at each in-person visit, and an optional leukapheresis was performed at baseline, week 3–4, and week 12. The original study was designed to enroll 20 participants (15 pembrolizumab and 5 placebo). However, due to recruitment difficulties especially during and after the COVID-19 pandemic, the study was closed after enrolling 7 participants.

Eligible participants were INRs with a recent CD4+ T-cell count >100 and ≤350 cells/mm³, who had been on an ART regimen for at least 12 months (stable for at least 4 weeks) and were virally suppressed (plasma HIV RNA <40 copies/mL) for at least 12 months; a single viral load >40 copies/mL and <500 copies/mL was permitted if subsequent values were <40 copies/mL. Exclusion criteria included pregnancy or breast-feeding; latent or active tuberculosis or a history of tuberculosis; osteoporosis; diabetes mellitus; active hepatitis B or C; opportunistic infection requiring maintenance therapy; active autoimmune disease or a history of autoimmune disease requiring systemic treatment; malignancy requiring systemic therapy; receipt of recent corticosteroid medications or other immunotherapeutics; recent or planned vaccination; history of or active noninfectious pneumonitis.

The primary endpoint was safety, as described in detail under the statistics section, below. The secondary endpoint was duration of decline in PD-1 expression, and exploratory endpoints included changes in immunologic and virologic parameters.

Pembrolizumab was provided by Merck Sharp & Dohme Corp. (Kenilworth, NJ) under a Cooperative Research and Development Agreement with the National Institutes of Health Clinical Center.

Flow Cytometry

Flow cytometry staining was performed at each visit on fresh EDTA preserved whole blood using 2 panels with the following antibodies: panel 1, CD3, CD4, CD8, CD27, CD45RA, CD14, CD19, CD152 (CTLA4), CD197 (CCR7), CD279 (PD-1), CD274 (PDL-1), and CD160; panel 2, CD3, CD4, CD8, CD27, CD45RA, CD14, CD19, TIGIT, CD223 (LAG3), CD366 (TIM-3), and CD244 (2B4). The source and specific clones for the antibodies are listed in Supplementary Table 1. Samples were run on a BD LSR Fortessa analyzer (BD Biosciences). Binding of clone EH12.1, which was used to detect PD-1, is blocked by pembrolizumab and thus can be used to evaluate pembrolizumab binding [29].

Additional immunophenotyping was performed by spectral flow cytometry to identify major T- and B-cell populations. Detailed methods are provided in the supplemental information.

The frequency of HIV-specific CD4+ and CD8+ T cells was assessed by immunostaining for functional markers as previously described [30]. Cells were stained with Zombie NIR (Biolegend #423106), and antibodies to cell-surface markers CD3, CD4, and CD8 (Supplementary Table 2). Antibodies to the following functional markers were utilized: for CD4+ T cells, IFNɣ, TNFα, CD40L, and IL-2, and for CD8+ T cells, IFNɣ, TNFα, MIP-1β, and CD107a (Supplementary Table 2). Data were acquired on a Cytek Aurora spectral cytometer using the SpectroFlo Software and analyzed using FlowJo version 10.10.0.

Viral Assays

HIV plasma viral load was measured by the RealTime HIV-1 Assay (Abbott Laboratories; lower limit of quantification, 40 copies/mL). The frequency of CD4+ T-cells carrying HIV DNA was quantified using droplet digital PCR (QIAcuity Digital PCR System, Qiagen, Germantown, MD) and primers and probe specific for the 5′ long terminal repeat region as previously described [30]. The frequency of CD4+ T-cells carrying intact HIV proviruses was determined by a modified Intact Proviral DNA Assay (IPDA) as previously described [31, 32]. For both assays, HIV DNA copy numbers were normalized per 1 × 10⁶ CD4+ T cells.

Cytotoxicity Assay

A cytotoxicity assay was used to examine killing of CD4+ T-cells superinfected with HIV_SF162, as reported previously [33]. Detailed methods are provided in the supplemental information, and the antibodies utilized are listed in Supplementary Table 3.

Plasma Biomarkers

Plasma biomarker levels were measured using the ELLA platform (ProteinSimple) per the manufacturer's instructions, for the following biomarkers: PD-L1 (B7-H1), CD25 (IL-2Rα), IL-6, Fas (TNFRSF6/CD95), Fas Ligand (TNFSF6), Granzyme B, Perforin, CRP, D-Dimer, IFN-α2, CCL5 (RANTES), CCL3 (MIP-1α), and CXCL10 (IP-10) (Bio-Techne, Minneapolis, MN).

Statistics

The primary endpoint was the frequency of either Grade 3 or higher adverse events or Grade 2 or higher autoimmune events requiring corticosteroid therapy that were possibly, probably, or definitely related to pembrolizumab. The original study design focused on estimating the probability of AEs in the pembrolizumab arm rather than on power to compare arms, though the placebo arm was planned to be used informally to interpret AEs in the pembrolizumab arm. A sample size of 15 pembrolizumab-administered participants provided a 91.3% chance of observing at least 1 AE under the assumption of single participant AE rate of 15%. Given the premature study closure and the randomization of only 1 participant to the placebo arm, the primary analyses focused on comparing baseline values to later time points for the 6 pembrolizumab recipients.

For flow cytometry parameters, mixed-effects modeling was used initially to determine differences over time compared with baseline, which if significant was followed by T-test analysis of individual time points versus baseline. For biomarkers and viral assays, the Wilcoxon matched-pairs signed rank test was used to compare individual time points to baseline values.

Patient Consent Statement

The study was approved by the NIAID/NIH Institutional Review Boards, and all participants provided written informed consent. The study was reviewed periodically by the intramural Data and Safety Monitoring Board (DSMB) of NIAID. The protocol was conducted under an Investigational New Drug Application—approved by the US Food and Drug Administration. The trial was registered at ClinicalTrials.gov, with identifier NCT03367754.

RESULTS

Study Population

Seven participants were enrolled between August 2018 and January 2020 (Table 1). One received placebo, and 6 received pembrolizumab. Six of the seven were male, and the average age was 53 years (SD, 8.7 years). Four were white, two were black, and one was Asian. The time since diagnosis of HIV infection in the pembrolizumab recipients ranged from 4 to 33 years, with a mean of 18.2 years (SD, 13.2 years), while in the placebo recipient it was 11 years. All participants were virally suppressed below the level of detection of 40 copies/mL. Baseline CD4+ T-cell counts ranged from 120 to 320 cells/mm³ (mean of 2 values) with a mean of 236 cells/mm³ (SD, 76 cells/mm³) in the pembrolizumab group, and 114 cells/mm³ in the placebo recipient. Baseline CD8+ T-cell counts ranged from 252 to 644 cells/mm³ with a mean of 490 cells/mm³ (SD, 159 cells/mm³) in the pembrolizumab group, and 395 cells/mm³ in the placebo recipient. Despite all participants having persistently low CD4+ T-cell counts prior to enrollment, most had no history of opportunistic infections (OIs). The participant who received placebo had the most extensive history of OIs, with previously treated disseminated histoplasmosis, esophageal candidiasis, and CMV; however, all of these conditions had resolved at least 2 years prior to the study. One participant in the pembrolizumab group had thrush in the past, with no recurrence.

Table 1.

Baseline Characteristics of Study Participants

	Pembrolizumab (N = 6)	Placebo (N = 1)
Age (y, mean ± SD)	52.3 (±9.2)	58
Gender
Male	5	1
Female	1
Race
White	4
Black	1	1
Asian	1
Ethnicity
Hispanic or Latino	1
Years since diagnosis of HIV infection (mean ± SD)	18.2 ± 13.2	11
Baseline CD4+ T-cell count (cells/mm3, mean ± SD)a	236 ± 76	114
Baseline CD8+ T-cell count (cells/mm3, mean ± SD)a	490 ± 159	395
OI history	Thrush (N = 1)	Disseminated histoplasmosis; esophageal candidiasis; CMV (N = 1)

^aBaseline counts represent the mean of 2 values for each participant.

Safety Results

One grade 3 adverse event was reported during the study: at week 5, the placebo recipient developed varicella-zoster virus reactivation of the ophthalmic branch of the left trigeminal nerve, with associated keratitis and mild anterior uveitis. The participant was closely followed by ophthalmology and was treated with valacyclovir and oral and topical prednisone, with gradual resolution of symptoms.

No pembrolizumab-related immune events requiring prednisone therapy occurred during the study. Endocrine testing was routinely performed to screen for endocrine autoimmune events. One participant in the pembrolizumab group developed an elevated ACTH level at week 2 (56.7 pg/mL, normal range, 5–26 pg/mL) that was normal (15.6 pg/mL) 1 week later, and remained within normal limits for the duration of the study. A second pembrolizumab participant developed an elevated ACTH level at week 9 (60 pg/mL) that was normal (17.4 pg/mL) 1 week later, and remained within normal limits for the duration of the study. The same participant had a mildly elevated TSH at week 16 (4.34 µIU/mL, normal range, 0.27–4.2 µIU/mL) with a normal thyroxine level; 1 week later, the TSH was normal (2.37 µIU/mL). At week 20, the same participant developed 2 small areas (1–2 cm) of alopecia on his scalp, similar to what had developed at the time of his HIV diagnosis; he reported regrowth beginning at week 48. Due to travel restrictions related to the COVID-9 pandemic, the participant could not be evaluated at the NIH after week 16, but was followed by his primary physician and telephone call protocol visits.

Flow Cytometry

CD4+ and CD8+ T-cell numbers remained stable in most participants during the study (Figure 1). Participant 3 had a significant increase in CD4+ T-cell numbers (P = .034) based on simple linear regression analysis. PD-1 expression as detected by monoclonal antibody EH12.1 was elevated at baseline in both CD4+ T cells (mean, 59.1%; SD, 7.6%) and CD8+ T cells (mean, 48.6%; SD, 12.5%); expression decreased in all pembrolizumab recipients to nadir levels at weeks 1–3 of 0.8%–23.5% for CD4+ T cells, and 0.1%–23.2% for CD8 cells (P < .0001 for both by mixed-effects modeling; Figure 2), while remaining stable in the placebo recipient. No correlation was seen between nadir PD-1 expression levels and either baseline CD4+ or CD8+ T-cell numbers, or the baseline level of expression of PD-1 on either cell population. None of the other immune checkpoint markers, including PD-L1, changed significantly following administration of pembrolizumab.

Figure 1.

CD4+ (top) and CD8+ (bottom) T-cell numbers over time for the 7 participants (Pt.) enrolled in the study. Pt. 1 (black with open circles) received placebo, while the remaining participants received pembrolizumab at week 0, indicated by the vertical line. The solid black line indicates the mean counts over time for the pembrolizumab recipients.

Figure 2.

PD-1 expression, as measured by clone EH12.1, in CD4+ (top) and CD8+ (bottom) T-cells over time for the 7 participants (Pt.) enrolled in the study. Pt. 1 (black with open circles) received placebo, while the remaining participants received pembrolizumab at week 0. The solid black line indicates the mean counts over time for the pembrolizumab recipients. By mixed-effects modeling, both CD4+ (P < .0001) and CD8+ (P < .0001) T-cells showed significant changes; a paired T test was then used to determine P values for individual weeks compared with week 0: ^‡P < .05; ^†P < .01; *P < .001.

Pembrolizumab led to activation of both CD4+ and CD8+ T cells. By mixed-effects modeling, both CD4+DR+CD38+ (P < .0001) and CD8+DR+CD38+ (P < .0001) cells showed significant increases that peaked at weeks 3–4, and returned to baseline by ∼week 24 (Figure 3). Single positive cells (DR+ or CD38+) showed similar, statistically significant, patterns (P = .01 to <.0001).

Figure 3.

Expression of both HLA-DR and CD38 by CD4+ (squares) and CD8+ (triangles) over time following pembrolizumab administration. Pembrolizumab, which was administered at week 0, indicated by the vertical black line, led to activation of both CD4+ and CD8+ T cells. Values represent the mean ± standard error of the mean (SEM) for the 6 participants who received pembrolizumab. By mixed-effects modeling, both CD4+DR+CD38+ (P < .0001) and CD8+DR+D38+ (P < .0001) showed significant changes; a paired T test was then used to determine P values for individual weeks compared with week 0: ^‡P < .05; ^†P < .01; *P < .001.

Spectral flow cytometry was used to characterize in greater detail both T- and B-cell populations in the 5 participants (Pt. 3–7) who received pembrolizumab and underwent apheresis. Of the 30 T-cell clusters identified, one (cluster 16) showed a significant change over time (P = .043 by mixed-effects model analysis), with enrichment at the 3–4 week time point (Figure 4A, B). Cluster 16 represents transitional/effector memory cytotoxic CD8+ T cells that express high levels of activation (CD38 and HLA-DR) and activation/exhaustion (TIGIT and 2B4) markers. In addition, the subset of CD8+ T cells that expressed both CD38 and HLA-DR increased significantly at weeks 3–4, and were declining by week 12, consistent with the standard flow data (Figure 3).

Figure 4.

Characterization of T- (A, B) and B-cell (C–E) populations by high dimensional spectral flow cytometry. Results represent data from the 5 participants who received pembrolizumab and were able to undergo apheresis. (A) Uniform Manifold Approximation and Projection (UMAP) visualization of 30 T-cell clusters identified by FlowSOM clustering, changes in the cluster intensities over time, and results for individual participants over time for cluster 16. Of the 30 clusters identified, by mixed-modeling analysis only cluster 16 changed significantly over time (P = .043), though subsequent T-test analysis showed only a trend toward an increase at weeks 3–4. The red circles highlight cluster 16 in the UMAPs for the 3 time points. (B) Heat map representing the levels of expression of the cell-surface markers used for T-cell phenotyping among the 30 clusters. Cluster 16 (arrow) represents transitional/effector memory cytotoxic CD8+ T cells that express high levels of activation (CD38 and HLA-DR) and exhaustion (TIGIT and 2B4) markers. (C) UMAP visualization of 20 B-cell clusters identified by FlowSOM clustering using data collected from all participants at all time points, and the changes in cluster intensities over time. The red circles highlight cluster 02 (C-02) in the UMAPs for the 3 time points. (D) Heat map representing the levels of expression of the cell-surface markers used for B-cell phenotyping among the 20 clusters. Cluster 02 (arrow) represents resting memory B cells (CD27+ and IgG+). (E) Results for individual participants over time for cluster 02, the only cluster that showed a significant temporal change (a decrease), as wells as for frequencies of B-cells expressing CD27, IgG, or PD-L1, respectively, analyzed by manual gating. PB, plasmablasts.

Among the 20 B-cell clusters identified, only 1, cluster 02, changed over time, with a significant decrease between baseline and week 12 (Figure 4C–E). This cluster represents resting memory B cells and the decline is driven by IgG-class-switched and CD27-expressing B cells. Of note, PD-L1 expression on B cells increased at weeks 3–4, and then declined by week 12.

The frequency of HIV-specific and CMV-specific polyfunctional CD4+ and CD8+ T cells did not change significantly over time for the 5 participants undergoing apheresis, though individual participants did show fluctuations (Supplementary Figure 1).

Plasma Viremia, HIV Reservoirs, and Cytotoxicity Activities

Plasma HIV viral loads remained <40 copies/mL at all time points in 5 participants. In 1 pembrolizumab recipient (Pt. 2), the week 1 viral load was 1359; by week 2, it had returned to <40 copies/mL. In the placebo recipient, the week 48 VL was 277 copies/mL likely due to running out of ART medications. It returned to <40 copies/mL once ART was reinitiated.

The level of total HIV DNA remained unchanged over time. Similarly, the levels of intact as well as 3′ and 5′ defective proviral DNA were largely unchanged at the week 3–4 and week 12 time points (Figure 5A). No pembrolizumab recipient had a >0.5 log change in any of these parameters.

Figure 5.

Viral and cytotoxicity assays. (A) Changes in total, intact, 5′ defective, and 3′ defective HIV DNA for all participants following pembrolizumab administration. Pembrolizumab was administered at day 0. Results for the placebo recipient (Pt. 1) are shown in the black lines with open circles. By the Wilcoxon matched-pairs signed rank test, there were no significant changes seen following pembrolizumab administration compared with baseline. (B) Changes in cytotoxicity at week 3–4 in the 5 pembrolizumab recipients who had sufficient cells available to perform the assay (left panel). Overall, by the Wilcoxon matched-pairs signed rank test, there were no significant changes seen at weeks 3–4 compared with baseline. All values were below the levels seen in prior studies in long-term nonprogressors/elite controllers (LTNP/EC) and similar to the values seen in progressors (right panel) [33].

The cytotoxicity assay could be performed at baseline and weeks 3–4 on the 5 pembrolizumab recipients who underwent apheresis. Again, there were no consistent changes seen, although participant 4 showed an increase in CD4+ T-cell elimination from 4.6% to 26.3%, and participant 7 showed an increase from 3.0% to 9.3% (Figure 5B). Levels remained in the range seen in prior studies for progressors rather than long-term nonprogressors/elite controllers [33].

Plasma Biomarkers

Of the plasma biomarkers evaluated by ELLA, only CD25 (IL-2Rα) showed a statistically significant increase following pembrolizumab administration, although other biomarkers such as CXCL-10, perforin, granzyme B, and CRP, showed nonsignificant increases (Supplementary Table 4).

DISCUSSION

The current study has suggested, based on treatment of a small group of PWH who have had an inadequate immunologic response to ART that a single dose of pembrolizumab is safe in this population. The study also found that PD-1 blockade was maximal at 1–3 weeks post infusion, with a subsequent increase in PD-1 expression to baseline levels by 12–24 weeks. Overall, there were no changes seen in CD4+ or CD8+ T-cell numbers, or in viral DNA or HIV-specific cytotoxicity, although there was individual variability in these parameters.

While the study was small as it was stopped early, it still provides valuable information given the limited data currently available describing the safety and immune effects of PD-1 inhibitors in PWH, especially INRs. There were no major adverse autoimmune events, a primary concern with pembrolizumab therapy in cancer patients as well as PWH [34]. Transient increases in ACTH and thyroxine levels were seen in 2 participants, but normalized without intervention by the following week, and 1 participant developed small patches of alopecia that gradually resolved without therapy. Our experience is in contrast to that of 2 other studies in PWH without malignancies. In ACTG A5326, 1 participant who received BMS 936559 (an anti-PD-L1 monoclonal antibody) developed hypophysitis, manifested as hypoadrenalism and hypogonadism, approximately 9 months after dosing [27]. In ACTG A5370, 2 participants who received cemiplimab, an anti-PD-1 monoclonal antibody, developed potential autoimmune events, including thyroiditis and hepatitis, resulting in DSMB-recommended discontinuation of the study [26]. In both studies, enrollment criteria included a CD4+ count ≥350 cells/mm³, while for the current study eligibility included a CD4 + count ≤350 cells/mm³. It is possible that the lower level of immune reconstitution may have led to a lower risk of autoimmune adverse events, though alternatively the small number of participants, or incomplete PD-1 blockade may account for this [35].

While pembrolizumab clearly blocked PD-1 expression, the nadir level of PD-1 expression was in general substantially higher than in ACTG A5370, in which expression levels in all 4 cemiplimab recipients declined to ∼1%–2% for both CD4+ and CD8+ T cells, whereas in our study only 1 (for CD4+) and 2 (for CD8+) of 6 pembrolizumab recipients had nadir levels below 5%, with nadir levels >10% in the other 4, levels which were inadequate to increase IL-2 production (as a marker of increased function) in vitro [36]. This may be related in part to the higher baseline PD-1 expression levels seen in the current study, with a mean of 60.4% (SD, ±7.4%) for CD4+ and 45.2% (SD, ±9.4%) for CD8+ T cells, compared with ∼4% and 9%, respectively, in ACTG A5370, and 6.9% and 12.2%, respectively, in a cross-sectional study of virologically suppressed PWH [24]. Presumably more complete and sustained PD-1 blockade is needed for maximal immunologic effects, which may require higher or multiple doses.

Despite this incomplete blockade, pembrolizumab administration was associated with immunologic perturbations, as evidenced by transient increases in activated (CD38+DR+) CD4+ and CD8+ T cells. Activation of immune cells was supported by the increase in plasma levels of CD25 as well as trends toward an increase in other biomarkers. Activation among T cells may also explain the transient increase of PD-L1 expression on B cells [37]. Conversely, B-cell dependence on PD-1:PD-L1 signaling may explain the slight yet significant decrease in memory B-cells observed [38]. Overall, these changes potentially represent blockade of PD-1 mediated immune checkpoint regulation. There was no evidence, however, of increased anti-HIV activity, as measured by the cytotoxicity assay or cell-associated HIV DNA, which includes cells with intact as well as defective HIV genomes. Given that this was a single dose study, and the standard dose utilized was insufficient to result in a high level of PD-1 blockade (<1%–2%), higher doses or multiple doses of pembrolizumab would be needed to determine if greater and sustained inhibition of PD-1 checkpoint activities impacts HIV clearance. It is also possible that examining anti-HIV activity at earlier time points, eg, week 1, might have demonstrated such activity, but we chose 3–4 weeks as being more clinically meaningful [27, 39].

The study was limited by the difficulties in participant enrollment, related in part to the COVID-19 pandemic, which resulted in early closure of the study and the randomization of only 1 participant to placebo, which resulted in an inability to compare placebo to pembrolizumab recipients. The administration of a single dose limited the duration of PD-1 blockade, but the study was designed primarily as a safety study as a prelude to a multidosing study if a single dose was found to be safe.

In summary, a single dose of 200 mg of pembrolizumab administered to this small cohort of INRs was safe and well-tolerated, and resulted in immune activation of both CD4+ and CD8+ T cells, but with no evidence of improved anti-HIV activity. Subsequent studies focusing on sustained blockade of PD-1, likely to higher levels than were achieved in the current study, will help determine if pembrolizumab can play a role in improving immune responses to HIV in INRs.

Supplementary Data

Supplementary materials are available at Open Forum Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.