Abstract
Background
Romidepsin (RMD) is a histone deacetylase inhibitor reported to reverse HIV-1 latency. We sought to identify doses of RMD that were safe and induced HIV-1 expression.
Methods
Enrollees had HIV-1 RNA <40 copies/mL on antiretroviral therapy. Measurements included RMD levels, plasma viremia by single-copy HIV-1 RNA assay, HIV-1 DNA, cell-associated unspliced HIV-1 RNA (CA-RNA), acetylation of histone H3-lysine-9 (H3K9ac+), and phosphorylation of transcription factor P-TEFb. Wilcoxon tests were used for comparison.
Results
In the single-dose cohorts 1–3, 43 participants enrolled (36 participants 0.5, 2, 5 mg/m 2 RMD; 7 placebo) and 16 enrolled in the multidose cohort 4 (13 participants 5 mg/m 2 RMD; 3 placebo). One grade 3 event (neutropenia) was possibly treatment related. No significant changes in viremia were observed in cohorts 1–4 compared to placebo. In cohort 4, pharmacodynamic effects of RMD were reduced proportions of CD4+ T cells 24 hours after infusions 2–4 (median, −3.5% to −4.5%) vs placebo (median, 0.5% to 1%; P ≤ .022), and increased H3K9ac+ and phosphorylated P-TEFb in CD4 + T cells vs placebo (P ≤ .02).
Conclusions
RMD infusions were safe but did not increase plasma viremia or unspliced CA-RNA despite pharmacodynamic effects on CD4 + T cells.
Clinical Trials Registration
Keywords: HIV-1 latency, clinical trial, romidepsin, randomized, histone deacetylase inhibitor, HIV-1 expression
The histone deacetylase inhibitor, romidepsin, failed to induce HIV expression when administered to persons with HIV on suppressive ART. Expected pharmacodynamic effects of romidepsin were observed but did not lead to significant changes in plasma viremia or unspliced cell-associated RNA.
One approach to eliminate latent human immunodeficiency virus-1 (HIV-1) is to induce proviral expression. If HIV-1 expression kills infected cells or exposes them to immune clearance while antiretroviral therapy (ART) prevents infection of new cells, this may reduce the HIV-1 reservoir. Histone deacetylase inhibitors (HDACi) induce HIV-1 expression ex vivo by increasing histone acetylation and facilitating transcriptional activation of proviruses [1–3]. Several HDACi have been studied as a part of the “kick and kill” HIV cure strategy [4–12]. Romidepsin (RMD) is Food and Drug Administration-approved for the treatment of cutaneous T-cell lymphoma (CTCL) and is administered at a dose of 14 mg/m2 as 3 sequential doses given on days 1, 8, and 15 of a 28-day cycle. It is highly protein bound, undergoes extensive metabolism by the CYP3A4 pathway, and has a terminal half-life of about 3 hours after infusion. In clinical studies of RMD, participants experienced toxicities including nausea, vomiting, fatigue, and bone marrow suppression.
Studies using an in vitro T-cell model of HIV-1 latency showed that RMD was the most potent inducer of HIV-1 (half maximum effective concentration [EC50] = 4.5 nM) compared with other HDACi, including vorinostat and panobinostat [13]. In CD4+ T cells isolated from individuals with HIV-1 on suppressive ART, transient exposure to 40 nM RMD for 4 hours induced a mean 6-fold increase in intracellular HIV-1 RNA. RMD-induced intracellular HIV-1 RNA expression persisted for 48 hours and correlated with sustained inhibition of cell-associated HDAC activity. RMD also increased levels of HIV-1 RNA in virions from memory and resting CD4+ T-cell cultures [13]. The activation of HIV-1 expression was observed at RMD plasma concentrations well below what is used to treat CTCL.
Some of the first in vivo evidence of the effect of RMD on proviral expression in people with HIV-1 came from Søgaard et al [10], who administered 3 infusions of 5 mg/m2 RMD to 6 participants in a single-arm, nonrandomized study. The primary objectives of the current study were to assess the safety and tolerability of single (0.5, 2, and 5 mg/m2) and then multiple doses of RMD (5 mg/m2 × 4) versus placebo to determine the best dosing regimen for induction of HIV-1 expression [14].
METHODS
Study Design and Participants
A5315 was a phase 1/2, double-blinded, randomized, placebo-controlled, dose-escalation study in 4 cohorts of up to 15 participants per cohort conducted at 10 US-based AIDS clinical trials group (ACTG) sites. Eligible participants were randomized 4:1 to receive pharmaceutical grade RMD or placebo (0.9% saline) and included persons with chronic HIV-1 infection, at least 18 years of age, with CD4+ T-cell counts ≥300 cells/mm3, suppressed viremia (<50 copies/mL), and no HIV-1 RNA blips on a raltegravir (RAL)-, dolutegravir (DTG)-, or efavirenz (EFV)-based regimen; or, for cohort 4, on a RAL- or DTG-based regimen. Female participants of childbearing potential agreed to use contraception. Persons with a history of corrected QT interval (QTc) prolongation, class III or IV heart failure, or a family history of prolonged QTc syndrome were excluded. Persons requiring treatment with medications eliminated by or that induce CYP3A4 were not permitted to enroll, including those receiving boosted protease inhibitors. Immunizations were not permitted. All participants provided written informed consent for their participation in the study prior to any study procedures.
Randomization and Masking
A computer algorithm with permuted blocks assignment was used at the ACTG Statistical and Data Management Center to randomly assign participants (4:1) to receive RMD or placebo. Participants and site personnel were masked to treatment allocation. Statisticians were masked to randomization details and only unmasked for routine monitoring, generating dose escalation reports, and conducting primary analysis. All other study personnel were masked to randomization details and treatment allocation.
Procedures
RMD dosage (Celgene) was calculated based on body surface area in square meters (m2). In the single-dose cohorts, participants received 0.5 mg/m2, 2 mg/m2, or 5 mg/m2 RMD in 0.9% saline or placebo 0.9% saline. Samples were obtained for pharmacokinetic analysis prior to and following RMD infusion to measure antiretroviral (RAL, DTG, and EFV) and RMD plasma concentrations. Participants had plasma HIV-1 RNA assessed by single-copy assay (SCA) [14] at screening, preentry, entry, hours 6, 12, 24, and 48 following the start of the infusion, and days 7, 14, and 28 postinfusion. Plasma HIV-1 RNA was measured by the Abbott M2000 standard assay in real time on day 7 after each infusion.
Once safety was established for the 5 mg/m2 dose, enrollment proceeded into the fourth cohort in which participants received a total of 20 mg/m2 RMD (over 4 sequential doses every other week) or placebo 0.9% saline at days 0, 14, 28, and 42 (± 2 days). Pharmacokinetics samples were obtained before and after the third and fourth RMD infusions. Plasma HIV-1 RNA levels were measured by SCA at preentry, entry, preinfusion, and at 24 and 72 hours postinfusion, and by Abbott M2000 in real time at day 7 after each infusion. RMD or placebo was infused intravenously over 4 hours. An electrocardiogram was performed prior to and following RMD administration. ART adherence assessments and pregnancy tests were required at each study visit.
Outcomes
The prespecified primary safety outcome was occurrence of grade 3+ adverse events probably, possibly, or definitely related to study treatment. The primary efficacy outcomes were: changes in plasma HIV-1 RNA from baseline as measured by SCA at 24 and 48 hours (average) after RMD or placebo in cohorts 1–3, and before and 24 hours after each dose in cohort 4; and changes in HIV-1 DNA and CA-RNA in resting CD4+ T cells from baseline to 24 hours in cohorts 1–3 or in peripheral blood mononuclear cells (PBMC) before and 24 hours after each dose in cohort 4. Key secondary outcomes included changes in percent CD4+ T cells, cellular proliferation (Ki67) and immune activation (CD38+HLA-DR+), histone H3-lysine-9 (H3K9) acetylation (K9ac), and P-TEFb phosphorylation at serine 175 (pS175), all by flow cytometry.
Statistical Analysis
The sample size for each cohort (n = 12 RMD and n = 3 placebo) provided a reasonably high probability of dose escalation when the true event rates were acceptable. For the primary efficacy objectives, the overall sample size for cohorts 1–3 provided 80% power to detect a 0.42 log10 increase in HIV-1 RNA by SCA and a 0.38 log10 copies/million cells difference in CA-RNA using a 2-sided Wilcoxon rank-sum test at 5% level for comparisons between pooled RMD dose cohorts and pooled placebo. Cohort 4 had 80% power to detect a difference of 0.79 log10 copies/million cells in CA-RNA between multidose RMD and placebo.
Primary comparisons were conducted between the pooled RMD dose cohort and pooled placebo in cohorts 1–3, and RMD and placebo in cohort 4 using Wilcoxon rank-sum tests for continuous outcomes and Fisher’s exact test for binary outcomes. Secondary comparisons were carried out between each RMD dose cohort and pooled placebo for cohorts 1–3. Jonckheere-Terpstra test was used to assess dose-dependent responses among 3 single-dose cohorts. SCA measures below the detection limit were imputed with half of the detection limit and CA-RNA and HIV-1 DNA measures below the detection limit were imputed with 1 copy/million cells. SCA, CA-RNA, and HIV-1 DNA measures were log10 transformed. No adjustments were made for multiple comparison and all analyses were performed using SAS version 9.4.
A study monitoring committee, appointed by the ACTG, monitored the study, and approved dose escalation decisions.
RESULTS
The study population was 59 adults with HIV-1 and CD4+ T-cell counts >300/mm3 with suppressed viremia (<50 copies/mL), on a RAL-, DTG-, or EFV-based antiretroviral regimen. In the single-dose cohorts 1–3, 40 (93%) of the 43 participants were male; the median age was 51 years, ranging from 25 to 72 years (Table 1). Twenty-seven (63%) were white and 14 (33%) were black. Median body mass index was 28 kg/m2. Over half (n = 25, 58%) were on an EFV-based ART regimen, 13 (30%) on RAL-based regimen, and 5 (12%) on DTG-based regimen. Median screening CD4+ T-cell count was 667 cells/mm3 and plasma HIV-1 RNA level by SCA was 1.5 copies/mL. Median duration on ART from first undetectable HIV-1 RNA level to study entry was 7.2 (range, 1.2–24) years and none had experienced prior virologic failure. Thirty-six participants were randomized to receive single-dose RMD; 7 received placebo. All participants in the single-dose cohorts completed the RMD/placebo infusion (Figure 1).
Table 1.
Baseline Characteristics, Cohorts 1–4
| Cohorts 1–3, Single Dose | Cohort 4, Multidose | |||||
|---|---|---|---|---|---|---|
| Baseline Characteristic | RMD 0.5 mg/m2 (n = 12) | RMD 2 mg/m2 (n = 12) | RMD 5 mg/m2 (n = 12) | Placebo (n = 7) | RMD 5 mg/m2 (n = 13) | Placebo (n = 3) |
| Male sex, No. (%) | 11 (92) | 11 (92) | 11 (92) | 7 (100) | 9 (69) | 2 (67) |
| Age, y, median (Q1, Q3) |
50 (28.5, 53.5) | 51.5 (44.0, 53.5) | 50.5 (36.5, 58.0) | 51 (44, 56) | 56 (48, 61) | 45 (34, 53) |
| Race/ethnicity, No. (%) | ||||||
| White, non-Hispanic | 8 (67) | 8 (67) | 6 (50) | 5 (71) | 6 (46) | 2 (67) |
| Black, non-Hispanic | 4 (33) | 3 (25) | 5 (42) | 2 (29) | 6 (46) | 1 (33) |
| Hispanic | 0 | 0 | 1 (8) | 0 | 1 (8) | 0 |
| Asian | 0 | 1 (8) | 0 | 0 | 0 | 0 |
| Duration of VL < LOQ, y, median (Q1, Q3) | 7.1 (4.7, 8.0) | 6.5 (3.6, 10.4) | 8.3 (4.6, 10.8) | 7.9 (4.3, 17.2) | 7 (5, 15) | 2 (1, 3) |
| ART regimen, No. (%) | ||||||
| RAL based | 4 (33) | 5 (42) | 0 | 4 (57) | 4 (31) | 0 |
| DTG based | 0 | 2 (17) | 3 (25) | 0 | 9 (69) | 3 (100) |
| EFV based | 8 (67) | 5 (42) | 9 (75) | 3 (43) | 0 | 0 |
| Screening CD4+, cells/mm3, median (Q1, Q3) |
673 (551, 966) | 773 (516, 1092) | 623 (531, 797) | 708 (510, 891) | 676 (535, 871) | 1172 (550, 1277) |
| SCA < LOQ, No. (%) | NA | NA | NA | NA | 5 (38) | 1 (33) |
| SCA > LOQ, copies/mL, median (Q1, Q3) |
2.2 (1.6, 4.3) | 0.9 (0.7, 1.5) | 1.3 (0.5, 5.1) | 2.9 (1.5, 3.7) | 2.7 (1.2, 8.3) | 0.6 (0.4, 0.8) |
SCA LOQ = 0.4 copies/mL. SCA was the screening measurement for cohorts 1–3 and preentry for cohort 4.
Abbreviations: ART, antiretroviral therapy; DTG, dolutegravir; EFV, efavirenz; LOQ, limit of quantification (50 copies/mL); NA, not applicable; RAL, raltegravir; RMD, romidepsin; SCA, single-copy assay; VL, viral load (plasma HIV-1 RNA).
Figure 1.
CONSORT flow diagram of randomized, dose-escalation trial of romidepsin (RMD) vs placebo (0.9% saline). Fifty-nine study participants were randomized to receive RMD or placebo out of 137 screened for eligibility. Allocation of RMD to placebo followed a 4:1 ratio in each dose cohort. Most participants completed study follow-up, and those who received placebo were grouped together across cohorts for analysis.
The multidose cohort 4 (5 mg/m2 × 4) enrolled 16 participants (Table 1 and Figure 1). Eleven (69%) participants were male with a median age of 54 (range, 26–72) years. Eight (50%) were white, 7 (44%) were black, and 1 (6%) was Hispanic. Four (25%) were on an RAL-based regimen and 12 (75%) on a DTG-based regimen. Median duration on ART with undetectable HIV-1 RNA level at study entry was 6 (range, 1–17) years. No participant experienced prior virologic failure while on ART. Median screening CD4+ T-cell count was 699 (range, 411–1385) cells/mm3. Preentry SCA was above the limit of detection (<0.4 copies/mL) for 10 participants with a median of 1.6 (range, 0.4–22.4) copies/mL. Thirteen were randomized to receive RMD and 3 to placebo. Fourteen of the 16 enrolled participants completed all 4 infusions, but 2 from the RMD arm went off study prematurely. One did not meet the requirements for subsequent infusions and was replaced per protocol; another discontinued the study very late in the enrollment phase and was not replaced.
RMD was generally safe and well tolerated at all doses (Table 2). No grade ≥3 adverse events were reported in cohorts 1–3. Seven participants receiving RMD reported 8 grade 2 events that were possibly treatment related: fatigue (n = 3), increases in the corrected QT interval (n = 2), hypophosphatemia (n = 2), and neutropenia (n = 1). Prolongation of the QT interval resolved within 7 days in 1 participant and within 15 days in the other participant. One participant from the single-dose 5 mg/m2 cohort 3 experienced a grade 2 neutropenia (962 cells/mm3) judged possibly related to RMD. Two participants randomized to the placebo arm in cohorts 1–3 experienced events judged possibly related to study treatment. One participant reported fatigue (grade 2) and diarrhea (grade 1); another had asymptomatic bradycardia (grade 1).
Table 2.
Adverse Events Possibly, Probably, or Definitely Related to Romidepsin
| No. of Adverse Events | ||||
|---|---|---|---|---|
| Adverse event | Grade 1 | Grade 2 | Grade 3 | Total No. Participants |
| Cohorts 1–3, single dose, 0.5, 2, 5 mg/m2 | ||||
| Neutropenia | 1 | 1 | ||
| Fatigue | 1 | 3 | 4 | |
| Headache | 2 | 2 | ||
| Normal QTc, increased from baseline | 1 | 1 | ||
| Increased QTc interval | 1 | 1 | ||
| Hypophosphatemia | 2 | 2 | ||
| Difficulty urinating | 1 | 1 | ||
| Cohort 4, multidose, 5 mg/m2 | ||||
| Nausea | 4 | 2 | ||
| Neutropenia | 2 | 3 | 1 | 4 |
| Fatigue | 1 | 1 | ||
| Headache | 1 | 1 | ||
| Blurred vision | 1 | 1 | ||
In the multidose cohort 4, there was 1 grade 3 event (neutropenia). The neutrophil count was 744 cells/mm3 1 day after the first infusion and was deemed possibly related to study treatment. Four additional participants in cohort 4 had 1 or more grade 2 events judged possibly, probably, or definitely related to RMD. One participant reported blurred vision at week 7 and at the same visit also experienced grade 2 neutropenia (970 cells/mm3). A second participant experienced grade 2 nausea with the first infusion. A third developed grade 2 neutropenia (927 cells/mm3) 1 week following the first dose of RMD that persisted through week 2 (951 cells/mm3, grade 2). A fourth reported grade 2 fatigue and nausea following the first infusion, grade 2 headache and nausea with the third infusion, and grade 2 nausea associated with the fourth infusion. No participants randomized to the placebo in cohort 4 experienced an adverse event judged possibly, probably, or definitely related to study treatment.
We measured RMD and ARV levels prior to and following infusion(s) in cohorts 1–4 (Table 3 and Supplementary Tables 1 and 2). Dose-dependent increases in RMD exposures were observed. RMD levels declined quickly, consistent with the published terminal half-life of approximately 3 hours [15]. In multidose cohort 4, median RMD levels were 69 (interquartile range [IQR], 64–164) ng/mL and 134 (IQR, 84–211) ng/mL at hour 4 after the third and fourth infusions, respectively. The RMD exposures observed were expected from prior pharmacokinetic modeling and in range of those associated with ex vivo activation of HIV-1 expression in cells from donors with HIV-1 on ART [13]. ARV concentrations were measured prior to and following RMD dosing, across all cohorts, and no consistent changes were observed compared with placebo (Supplementary Table 1).
Table 3.
RMD Pharmacokinetics, Cohort 4, Multidose (5 mg/m2)
| Measurement | ≤BQL, No. (%) | >BQL, No. (%) | RMD Concentration, Median, ng/mL (Q1, Q3) |
|---|---|---|---|
| Preinfusion 3, n = 11 | 10 (91) | 1 (9) | 207 (207, 207) |
| 4 h postinfusion 3, n = 9 | 0 | 9 (100) | 69 (64, 164) |
| Preinfusion 4, n = 11 | 11 (100) | 0 | 0 |
| 4 h postinfusion 4, n = 8 | 0 | 8 (100) | 134 (84, 211) |
Abbreviations: BQL, below quantifiable limit of 0.54 ng/mL; RMD, romidepsin.
Changes in plasma HIV-1 RNA were determined by SCA as the primary measure of HIV-1 expression. In cohorts 1–3, HIV-1 RNA by SCA was measured at screening, preentry, entry, hours 6, 12, 24, and 48, and days 7, 14, and 28. No consistent or significant changes in viremia were found compared to placebo at any time point (all P > .05; Supplementary Figure 1). CA-RNA and HIV-1 DNA were measured at preentry, hour 24, and day 14 postinfusion. No significant difference for CA-RNA or HIV-1 DNA levels were observed across cohorts 1–3 compared to placebo (all P > .05) (Supplementary Figures 2 and 3).
In the multidose cohort 4, no significant differences in the plasma HIV-1 RNA levels were found between the baseline and postinfusion time points, nor between pre- and postinfusion time points, except at the hour 72 time point following the second infusion. At 72 hours, an increase in viremia was noted in the placebo arm but not in the RMD arm (P = .029) (Figure 2A). All participants had HIV-1 RNA levels <40 copies/mL at all postinfusion time points by Abbott M2000, except 1 whose HIV-1 RNA levels rose to 282 copies/mL 5 weeks following the fourth infusion, and then declined to <40 copies/mL 5 weeks later. The participant did not report missed doses of ART before or following the fourth infusion. In the RMD arm, there were no changes in the proportion of “target not detected” compared to <40 copies/mL results of the Abbott M2000 plasma HIV-1 RNA assay. CA-RNA and HIV-1 DNA per million CD4+ T cells were also measured prior to each infusion, 24 hours after each infusion, and 72 hours following the second infusion. No significant differences were found between pre- and postinfusion time points (all P > .05; Figure 2B and 2C).
Figure 2.
Virology outcomes for cohort 4, RMD 5 mg/m2 × 4 infusions. A, Levels of HIV-1 RNA in 5 mL of plasma were measured by single-copy assay targeting the 3′ region of pol with a lower limit of detection of 0.4 copies/mL [14]. CA HIV-1 RNA (B) and HIV-1 DNA (C) were measured in PBMC by qPCR [18] and normalized per 1 million CD4+ T cells. In all panels, solid red lines represent the median and interquartile range (Q1, Q3) of 13 RMD-treated participants, and dotted blue lines represent placebo results from 10 participants combined across all cohorts. Median indicators are slightly offset for illustration purposes only. Vertical gray background lines indicate infusion time points. Abbreviations: CA, cell associated; HIV-1, human immunodeficiency virus; PBMC, peripheral blood mononuclear cell; postinf, postinfusion; qPCR, quantitative polymerase chain reaction; RMD, romidepsin.
The failure of RMD to impact HIV-1 expression led to careful assessments of its pharmacodynamic effects. RMD has been shown to decrease the percentage of CD4+ T cells in the peripheral blood of individuals with CTCL [16]. In cohort 4, significant decreases in percent CD4+ T cells were observed at 24 hours after infusions 2, 3, and 4 (median −3.5% to −4.5% vs +0.5% to +1% in placebo recipients, all P ≤ .022; Figure 3A). T-cell apoptosis was assessed by measuring annexin V and 7-aminoactinomycin D in CD4+ and CD8+ T cells. No differences were detected in apoptotic markers between the RMD and placebo groups (all P > .05), suggesting the decrease in percent CD4+ T cells was not due to increased apoptosis.
Figure 3.
Pharmacodynamic outcomes for cohort 4, RMD 5 mg/m2 × 4 infusions. A, Median (Q1, Q3) percent of peripheral CD4+ T cells by flow cytometry in RMD and combined placebo groups. B, Median (Q1, Q3) percent of CD4+ T cells staining positive by flow cytometry for intracellular histone 3 lysine 9 (K9ac+). C, Activated P-TEFb as a percent of CD4+ T cells staining positive by flow for phosphorylated serine 175 (pS175) on cyclin-dependent kinase 9 (CDK9). Solid red lines represent the median and interquartile range (Q1, Q3) of 13 RMD treated participants, and dotted blue lines represent placebo results from 10 participants combined across all cohorts. Median indicators are slightly offset for illustration purposes only. Vertical gray background lines indicate infusion time points. Abbreviations: postinf, postinfusion; RMD, romidepsin.
Acetylation of histones changes the structure of chromatin and permits DNA binding proteins to interact with sites that activate gene transcription. RMD has been shown to produce a dose-dependent increase in histone acetylation and P-TEFb activation ex vivo. Phosphorylation of CDK9 kinase at the serine 175 position is a marker of activated P-TEFb in CD4+ T cells, which can enhance HIV-1 transcription [17]. Histone H3K9-acetylation in CD4+ T cells following a single dose of RMD showed significant dose responses at hours 12, 24, and 48, and day 28, with the greatest increases observed in cohort 3 (all P < .05; Supplementary Table 3). In the multidose cohort 4, significant increases in histone H3K9 acetylation and P-TEFb activation in CD4+ T cells were observed at all postinfusion time points compared to baseline (all P ≤ .02; Figure 2).
In the multidose cohort 4, we found no evidence that RMD activated CD4+ or CD8+ T cells (increases in percent CD25, CD69, or CD38+HLA-DR+; Supplementary Table 4) or altered cell cycling (Ki67; Supplementary Table 4) (all P > .05). No significant changes in interleukin-6 (IL-6) levels were noted from baseline to infusions 1 and 4, 24 hours postinfusions 1 and 4, and 10 weeks postinfusion 4 (data not shown). Significant decreases in high sensitivity C-reactive protein (hsCRP) compared to baseline were found 24 hours postinfusion 1 and 4 (data not shown).
DISCUSSION
Despite achieving levels of RMD shown to induce HIV-1 expression ex vivo, no significant effects of RMD infusions were observed on plasma HIV-1 RNA, CA-RNA, or HIV-1 DNA following single doses (0.5, 2, or 5 mg/m2) or 4 sequential doses totaling of 20 mg/m2. By contrast, there were reproducible, infusion-related pharmacodynamic effects of RMD, including increased histone acetylation and P-TEFb activation in CD4+ T cells. Administration of single and multiple doses of RMD were generally safe and well tolerated.
The lack of virologic effects differs from an earlier study of RMD (REDUC A), which reported activation of HIV-1 expression in persons on suppressive ART [10]. There are important differences between A5315 and the REDUC A trial. A5315 was a randomized, placebo-controlled, dose-escalation study, compared to the nonrandomized, single-arm, fixed dose (5 mg/m2) REDUC A study that assessed within-participant changes without parallel controls. A5315 assessed single, ascending doses, then 4 doses of RMD at 5 mg/m2 administered every 14 days, whereas REDUC A assessed 3 once-weekly doses of 5 mg/m2. REDUC A reported increases in plasma viremia after infusions of 5 mg/m2 RMD in 5 of 6 participants with standard HIV-1 RNA assays. Three of the 6 REDUC A participants were receiving a CYP3A4 inhibitor-boosted protease inhibitor (n = 2) or integrase inhibitor (n = 1) in contrast to A5315 participants, who were receiving only nonboosted regimens because of concern about toxicity from excessive RMD exposure. It is unknown if a boosting effect of RMD levels may have played a role in increases in HIV-1 RNA levels in the REDUC A study. RMD levels were not reported from the REDUC A trial.
Discrepancies in virologic outcomes between the studies may be partially explained by sample processing procedures and the use of the Roche COBAS AmpliPrep/TaqMan HIV-1 test in the REDUC A study to quantify plasma HIV-1 RNA. This assay has a lower limit of quantification of 20 copies/mL and targets gag and the long terminal repeat (LTR) regions of the HIV-1 genome. In A5315, we used an improved SCA assay that targets highly conserved sequences in the 3′ region of pol and has single-copy sensitivity [14]. A5315 plasma was double-spun and SCA performed with no reverse transcriptase controls to exclude HIV-1 DNA contamination, which can be an issue with the TaqMan assay that does not differentiate between RNA and DNA. The use of SCA in A5315 to detect changes in viremia reduced the likelihood of false-negative and false-positive results. It should be noted the REDUC A study team performed tat/rev-induced limiting dilution assays and viral outgrowth assays that showed no effect of RMD treatment.
It may be possible that our sampling frequency, especially in the multidose cohort 4, was too infrequent to capture early, transient changes in HIV-1 transcription or viremia after RMD treatment. The REDUC A study reported changes in CA-RNA 30 minutes after the end of each 4-hour RMD infusion (4.5 hours after the infusion start). After the first dose, the increase in CA-RNA persisted for at least 24 hours but sampling after the second and third doses occurred only at 4.5 hours and 72 hours after infusion, thus the duration of the effects cannot be discerned. In the single-dose cohorts 1–3 of A5315, we sampled immediately before dosing, then 6, 12, 24, and 48 hours following the infusions, which should have captured increases in HIV-1 transcription; however, no significant effects on CA-RNA or viremia were observed. Due to blood volume restrictions in cohort 4, we were only able to measure CA-RNA 24 and 72 hours postinfusion. It is possible we missed some early transient effect of RMD treatment on HIV-1 RNA induction. However, the possibility that RMD significantly affected HIV-1 transcription and plasma viremia, which returned to baseline levels within 24 hours, seems unlikely and is contrary to both the sustained virologic effects of RMD observed ex vivo [13] and the prolonged pharmacodynamic effects observed in vivo (discussed below). If the effect of RMD is very short lived, it is unlikely to have an appreciable impact on the HIV-1 reservoir.
Although HIV-1 levels were not affected, significant and reproducible increases in histone acetylation and P-TEFb phosphorylation were observed following RMD infusion(s), as were decreases in the percent CD4+ T cells, all consistent with the known pharmacodynamic effects of RMD. The 5 mg/m2 dose used in this study represents 36% of the RMD dose approved to treat T-cell lymphoma. Despite this modest dose of RMD, significant increases in histone acetylation were found at 24 hours following the first 3 infusions, extending to 72 hours following the second infusion, indicating a sustained pharmacodynamic effect. Higher doses of RMD might produce greater pharmacodynamic effects but are likely to increase toxicity. As noted above, significant decreases in circulating CD4+ T cells were observed after multiple infusions of RMD.
Most studies of latency reversal agents, including vorinostat, panobinostat, and RMD (REDUC B), have been single-arm studies enrolling 5–20 participants who received 1 or more doses at varying dosing intervals. A5315 was a randomized, dose-escalating trial with a parallel placebo control group that enrolled a diverse population and involved nearly 60 participants. The differences in RMD activity observed in trials to date suggest that including a parallel control group, rather than self-controls, is an important design feature for future studies of latency reversal agents. None of the studies of HDACi as single latency reversal agents have shown a reduction in the inducible HIV-1 reservoir or delay in viral rebound, strongly suggesting that other therapeutic modalities will be required to achieve a cure for HIV-1 infection.
Supplementary Data
Supplementary materials are available at The Journal of 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.
Notes
Acknowledgments. We thank the study participants.
Author contributions. D. M., J. W. M., L. Z., R. G., C. G., M. P., R. M., B. M., and J. H. designed the study. D. M., J. W. M., L. Z., R. G., C. G., M. P., R. M., B. M., E. A., and J. D. conducted the study. J. W. M., J. C., B. M., C. D., J. K., E. A., and J. D. supervised and/or conducted the laboratory assays and experiments. L. Z. and E. A. did the statistical analysis. D. M., J. W. M., J. C., L. Z., R. G., M. P., C. G., E. A., and E. A. drafted the manuscript.
Disclaimer. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute of Allergy and Infectious Diseases or the National Institutes of Health.
Financial support. This work was supported by the National Institute of Allergy and Infectious Diseases, National Institutes of Health (grant numbers UM1 AI068634, UM1 AI068636, UM1 AI106701, UM1 AI069494, UM1 AI069412, UM1 AI069481, UM1 AI069511, UM1 AI069452, UM1 AI069423, UM1 AI069424, UM1 AI069432, and UM1 AI069534); and Gilead Sciences, Inc. Celgene provided romidepsin for cohort 4.
Potential conflict of interests. J. H. was employed by Gilead Sciences. R. G. has served on a scientific advisory board for Merck. J. W. M. serves or has served as a consultant for Gilead Sciences, Merck, and Accelevir Diagnostics, and owns share options in Co-Crystal Pharmaceuticals, Inc., Infectious Diseases Connect, and Abound Bio, Inc. All other authors report no potential conflicts of interest. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.
Presented in part: Conference on Retrovirus Infections and Opportunistic Infections, Boston, MA, 4–7 March 2018; and Conference on Retrovirus Infections and Opportunistic Infections, Seattle, WA, 4–7 March 2019.
Availability of data. The authors confirm that all data underlying the findings are fully available without restriction.
References
- 1.Margolis DM. Histone deacetylase inhibitors and HIV latency. Curr Opin HIV AIDS 2011; 6:25–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Pérez M, de Vinuesa AG, Sanchez-Duffhues G, et al. Bryostatin-1 synergizes with histone deacetylase inhibitors to reactivate HIV-1 from latency. Curr HIV Res 2010; 8:418–29. [DOI] [PubMed] [Google Scholar]
- 3.Van Lint C, Emiliani S, Ott M, Verdin E. Transcriptional activation and chromatin remodeling of the HIV-1 promoter in response to histone acetylation. EMBO J 1996; 15:1112–20. [PMC free article] [PubMed] [Google Scholar]
- 4.Lehrman G, Hogue IB, Palmer S, et al. Depletion of latent HIV-1 infection in vivo: a proof-of-concept study. Lancet 2005; 366:549–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Archin NM, Espeseth A, Parker D, Cheema M, Hazuda D, Margolis DM. Expression of latent HIV induced by the potent HDAC inhibitor suberoylanilide hydroxamic acid. AIDS Res Hum Retroviruses 2009; 25:207–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Archin NM, Keedy KS, Espeseth A, Dang H, Hazuda DJ, Margolis DM. Expression of latent human immunodeficiency type 1 is induced by novel and selective histone deacetylase inhibitors. AIDS 2009; 23:1799–806. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Edelstein LC, Micheva-Viteva S, Phelan BD, Dougherty JP. Short communication: activation of latent HIV type 1 gene expression by suberoylanilide hydroxamic acid (SAHA), an HDAC inhibitor approved for use to treat cutaneous T cell lymphoma. AIDS Res Hum Retroviruses 2009; 25:883–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Archin NM, Liberty AL, Kashuba AD, et al. Administration of vorinostat disrupts HIV-1 latency in patients on antiretroviral therapy. Nature 2012; 487:482–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Rasmussen TA, Tolstrup M, Brinkmann CR, et al. Panobinostat, a histone deacetylase inhibitor, for latent-virus reactivation in HIV-infected patients on suppressive antiretroviral therapy: a phase 1/2, single group, clinical trial. Lancet HIV 2014; 1:e13–21. [DOI] [PubMed] [Google Scholar]
- 10.Søgaard OS, Graversen ME, Leth S, et al. The depsipeptide romidepsin reverses HIV-1 latency in vivo. PLoS Pathog 2015; 11:e1005142. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Winckelmann A, Barton K, Hiener B, et al. Romidepsin-induced HIV-1 viremia during effective antiretroviral therapy contains identical viral sequences with few deleterious mutations. AIDS 2017; 31:771–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Leth S, Schleimann MH, Nissen SK, et al. Combined effect of Vacc-4x, recombinant human granulocyte macrophage colony-stimulating factor vaccination, and romidepsin on the HIV-1 reservoir (REDUC): a single-arm, phase 1B/2A trial. Lancet HIV 2016; 3:e463–72. [DOI] [PubMed] [Google Scholar]
- 13.Wei DG, Chiang V, Fyne E, et al. Histone deacetylase inhibitor romidepsin induces HIV expression in CD4 T cells from patients on suppressive antiretroviral therapy at concentrations achieved by clinical dosing. PLoS Pathog 2014; 10:e1004071. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Cillo AR, Vagratian D, Bedison MA, et al. Improved single-copy assays for quantification of persistent HIV-1 viremia in patients on suppressive antiretroviral therapy. J Clin Microbiol 2014; 52:3944–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Woo S, Gardner ER, Chen X, et al. Population pharmacokinetics of romidepsin in patients with cutaneous T-cell lymphoma and relapsed peripheral T-cell lymphoma. Clin Cancer Res 2009; 15:1496–503. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Kelly-Sell MJ, Kim YH, Straus S, et al. The histone deacetylase inhibitor, romidepsin, suppresses cellular immune functions of cutaneous T-cell lymphoma patients. Am J Hematol 2012; 87:354–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Mbonye UR, Gokulrangan G, Datt M, et al. Phosphorylation of CDK9 at Ser175 enhances HIV transcription and is a marker of activated P-TEFb in CD4+ T lymphocytes. PLoS Pathog 2013; 9:e1003338. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Hong F, Aga E, Cillo AR, et al. Novel assays for measurement of total cell-associated HIV-1 DNA and RNA. J Clin Microbiol 2016; 54:902–11. [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.