(5R)-5-hydroxytriptolide for HIV immunological non-responders receiving ART: a randomized, double-blinded, placebo-controlled phase II study

Summary

Background

Therapeutic approaches to HIV-suppressed immunological non-responders (INRs) remain unsettled. We previously reported efficacy of Chinese herbal Tripterygium wilfordii Hook F in INRs. Its derivative (5R)-5-hydroxytriptolide (LLDT-8) on CD4 T cell recovery was assessed.

Methods

The phase II, double-blind, randomized, placebo-controlled trial was conducted in adults patients with long-term suppressed HIV infection and suboptimal CD4 recovery, at nine hospitals in China. The patients were 1:1:1 assigned to receive oral LLDT-8 0.5 mg or 1 mg daily, or placebo combined with antiretroviral therapy for 48 weeks. All study staff and participants were masked. The primary endpoints include change of CD4 T cell counts and inflammatory markers at week 48. This study is registered on ClinicalTrials.gov (NCT04084444) and Chinese Clinical Trial Register (CTR20191397).

Findings

A total of 149 patients were enrolled from Aug 30, 2019 and randomly allocated to receiving LLDT-8 0.5 mg daily (LT8, n = 51), 1 mg daily (HT8, n = 46), or placebo (PL, n = 52). The median baseline CD4 count was 248 cells/mm³, comparable among three groups. LLDT-8 was well-tolerated in all participants. At 48 weeks, change of CD4 counts was 49 cells/mm³ in LT8 group (95% confidence interval [CI]: 30, 68), 63 cells/mm³ in HT8 group (95% CI: 41, 85), compared to 32 cells/mm³ in placebo group (95% CI: 13, 51). LLDT-8 1 mg daily significantly increased CD4 count compared to placebo (p = 0.036), especially in participants over 45 years. The mean change of serum interferon-γ-induced protein 10 was −72.1 mg/L (95% CI −97.7, −46.5) in HT8 group at 48 weeks, markedly decreased compared to −22.8 mg/L (95% CI −47.1, 1.5, p = 0.007) in placebo group. Treatment-emergent adverse events (TEAEs) were reported in 41 of 46 (89.1%) participants in HT8 group, 43 of 51 (84.3%) in LT8, and 42 of 52 (80.7%) in PL group. No drug-related SAEs were reported.

Interpretation

LLDT-8 enhanced CD4 recovery and alleviated inflammation in long-term suppressed INRs, providing them a potential therapeutic option.

Fundings

Chinese Academy of Medical Sciences Innovation Fund for Medical Sciences, Shanghai Pharmaceuticals Holding Co., Ltd., and the National key technologies R&D program for the 13th five-year plan.

Research in context.

Evidence before this study

Around 20–30% of treated patients with HIV fail to achieve optimal immune reconstitution and remain at greater risk of morbidity and mortality. We searched PubMed, up to November 25, 2022, for published clinical randomized trials among HIV-infected immunological non-responders (INRs). The search terms used were (“HIV” or “human immunodeficiency virus”) AND (“INR” or “immunological non-responder” or “suboptimal CD4 recovery”) AND (“clinical trial” or “randomized controlled trial” or “randomised controlled trial”), and produced 17 results. We identified several published clinical trials in HIV-infected INRs, including mesenchymal cell therapy, interleukin-2 (IL-2) therapy, maraviroc, atorvastatin, niacin, artesunate, hyperimmune bovine colostrum (HIBC), probiotics, and traditional Chinese medicines. However, none of these had been proven with efficacy in improving CD4⁺ T cell recovery and reducing immune activation. Therapeutic approaches to HIV-suppressed INRs remain unsettled.

Added value of this study

We previously reported the efficacy of Chinese herbal Tripterygium wilfordii Hook F in INRs. (5R)-5-hydroxytriptolide (LLDT-8) is a novel immunosuppressant modificated from triptolide, the major bioactive compound to reduce HIV-related immune activation. Our study is the first randomized, double-blind, placebo-controlled clinical trial assessing the effect of LLDT-8 in HIV-infected immunologic non-responders. The patients were 1:1:1 assigned to receive oral LLDT-8 0.5 mg or 1 mg daily, or placebo combined with antiretroviral therapy for 48 weeks. In the full analysis set, the primary endpoint of CD4 increase in the 1 mg daily group was significantly higher than that of the placebo group, particularly in those over 45 years old. The change in interferon-γ-induced protein 10 also significantly differed between 1 mg daily and placebo groups.

Implications of all the available evidence

LLDT-8 enhanced CD4 recovery and alleviated inflammation in long-term suppressed INRs, providing them a potential therapeutic option for HIV INR patients. The benefit was superior with LLDT-8 dosage 1 mg daily, and in older participants. Future larger size of clinical studis and pharmacological research into underlying mechanisms are needed to better understand its effectiveness.

Introduction

The advance in antiretroviral therapy (ART) has greatly improved the life of people living with HIV (PLWH). Most PLWH receiving effective ART could achieve varied levels of immune reconstitution.^1,2 However, an estimated 20–30% patients fail to achieve adequate CD4 T-cell recovery despite long-term virological suppression.³ These so-called immunological non-responders (INRs) are associated with increased risks of opportunistic infections, malignancies and non-AIDS comorbidity compared with immunological responders.^4,5

Incomplete immune reconstitution in HIV infection is a complex phenomenon, reflecting the crossover of immunodeficiency and underlying immune activation. Late presenters with low nadir CD4 count before ART initiation have a higher risk of becoming INRs.^3,4,6 Contributing factors also include those associated with reduced lymphocyte production like older age, reduced thymopoiesis and lymphopoiesis, and cytokine dysregulation.^7,8 On the other hand, persistent immune activation resulted from HIV reservoirs, co-infections, and microbial translocation, contributes to immune-senescence and apoptosis of CD4 T cells.^8,9 Several therapeutic strategies have been developed with limited benefit. Some aim to enhance thymic activity or lymphocyte proliferation through recombinant human interleukin (IL)-2,¹⁰ IL-7,^11,12 growth factors^13,14 or keratinocyte growth factors.¹⁵ Others targeted at the abnormal inflammatory status using immunomodulator agents such as chloroquine, hydroxychloroquine,^16,17 and metformin.^18,19 However, none of these proved benefit in prospective clinical studies.^9,20

Tripterygium wilfordii Hook F (TwHF) is a plant-originated Chinese herbal medicine used as an immune modulating agent approved by China National Medical Products Administration (NMPA), and has been used in autoimmune diseases.21, 22, 23 We conducted a pilot study in 18 INRs where TwHF was co-administrated with ART for 12 months, reported on the safety profiles and observed a mean elevation of 88 cells/mm³ in CD4 counts with a reduction of T-cell activation markers.²⁴ Subsequent transcriptional and proteomic studies demonstrated that triptolide was the major bioactive component in TwHF, which could reduce the activity of interferon (IFN)-signaling pathway. Such modulation may benefit INRs, who are featured with upregulation of these pathways as reported in our previous work.^25,26

TwHF is a botanical drug with diverse chemical components. Over 400 metabolites have been isolated and characterized from TwHF.²⁷ The extracted TwHF coformulation consists of unfixed compositions, limiting its application in clinical studies. Therefore, its main bioactive component triptolide has been extracted, and modified into a novel compound (5R)-5-hydroxytriptolide (LLDT-8) with immunosuppressive activity and reduced toxicity.²⁸ LLDT-8 has been studied in patients having rheumatoid arthritis, with anti-inflammation effect and favourable safety profile. We conducted a multicentre and randomized phase II study (CTR20191397; Clinicaltrials.gov ID: NCT04084444), to evaluate the safety and efficacy of LLDT-8 (Shanghai Pharmaceuticals Holding Co., Ltd., SPH) in improving CD4 reconstitution and reducing inflammation in adult INRs in China.

Methods

Study design

The study was a multicentre, randomized, double-blinded, placebo-controlled, dose-finding phase II study in long-term HIV-suppressed INRs. The study was carried out at Peking Union Medical College Hospital, Chinese Academy of Medical Sciences (PUMCH, CAMS, Beijing), and other eight hospitals in China (Table S1). The study was approved by NMPA, and by Institutional Review Board of PUMCH and other participating hospitals.

The study was initiated on August 30, 2019, and completed on July 5, 2022. The trial consisted of a screening-baseline period and a 12-month follow-up. A Data and Safety Monitoring Board meeting was held on March 18, 2022, when the interim data and events were reviewed and the study was recommended to continue without adjustment.

Participants

Eligible participants were PLWH aged 18–65 years old, who had received effective ART for at least four years and had plasma viral load below level of detection for at least 3.5 years, with their CD4 counts consistently below 350 cells/mm³. Patients were excluded if they had other types of immunodeficiency or severe co-morbidities. Those who received immunosuppressants or immunomodulators within six months before screening were also excluded. Detailed inclusion and exclusion criteria are presented in Supplementary Table S2. Contraception for participants during the study was required.

Due to the outbreak of coronavirus disease 2019 (COVID-19) during study period, some participants received SARS-CoV-2 vaccinations. Amendment was made for exemptions to recruitment under such circumstances. Subgroup analysis was performed for possible impact of SARS-CoV-2 vaccination on study objectives.

All participants provided written informed consent before participation in the study.

Randomisation and masking

At the time of randomization, eligible participants were randomly assigned in a 1:1:1 ratio to receive LLDT-8 0.5 mg daily, LLDT-8 1 mg daily or placebo combined with ongoing ART. A block randomization with stratification by center was utilized for this study. The randomization list was computer-generated using the PROC PLAN procedure of SAS software (version 9.4, SAS Institute Inc., United States) and applied to allocation concealment using an electronic Interactive Web Response System (DAS for IWRS, version 5.0, Beijing BioVoice Technology Co., Ltd.) by independent third-party statistical company.

The randomization system assigned every participant a unique number that remained unchanged throughout the study and labeled with the investigational drugs. The drugs were outwardly indistinguishable in packaging, label, and content. All participants, investigators, and outcome assessors were masked to group assignment.

Procedures

On study day 1, eligible participants were randomly assigned (1:1:1) to receive LLDT-8 0.5 mg daily, LLDT-8 1.0 mg daily or placebo for a consecutive 48 weeks. The study drug was recommended to be taken in the morning on an empty stomach, and be separated from participants' daily ART. Participants were followed and assessed at Week 4, 12, 24, 36 and 48 of treatment. On each scheduled visit, the investigator would assess the participant's general status, newly-onset symptoms and signs, and adverse events. Laboratory assessment was performed as depicted in Supplementary Table S3.

On each visit, number of research drug pills taken since previous visit was assessed. In the final analysis, the exposure level of study drug was defined as the numbers of drug pills taken relative to the numbers prescribed. A level between 80 and 120% was considered acceptable for data interpretation, and the drug adherence is one of the PPS criteria.

Participants’ ART mostly composed of two nucleoside reverse transcriptase inhibitors (NRTIs) combined with a non-nucleoside reverse transcriptase inhibitor (NNRTI), a boosted protease inhibitor (PI/r) or an integrase strand transfer inhibitor (INSTI), in multi-tablet or single-tablet formulation. Participants were encouraged to stay with original regimens, and modifications were allowed if needed for their HIV care.

Outcomes

The primary endpoints of the study included 1) Absolute change of CD4 counts from baseline to week 48; 2) Proportions of participants achieving a CD4 increase ≥50 cells/mm³ from baseline to week 48; and 3) changes of selected inflammatory markers from baseline to week 48, including interferon-γ-induced protein 10 (IP-10), high-sensitivity C-reactive protein (hsCRP), and IL-6.

The secondary endpoints included 1) change in CD4/CD8 ratio from baseline to week 24 and week 48; 2) proportions of participants achieving a CD4 increase ≥20% from baseline to week 24 and week 48; 3) proportions of participants (baseline CD4 count <200 cells/mm³) achieving a CD4 count over 200 cells/mm³ at week 24 and 48, respectively.

Exploratory endpoints were assessed in three Beijing-located centers, including changes in activation markers CD38 and HLA-DR expression levels of CD8 T cells from baseline to week 24 and 48.

Safety endpoints included occurrence of adverse events (AEs) and serious adverse events (SAEs) over the study period. Adverse events included those related to the study drug; all serious adverse events; and abnormal laboratory test values at each study visit.

Laboratory assessment

A central laboratory was set in PUMCH, Beijing, China. For all participants, the PUMCH Lab was involved in virological assessments (except for the screening ones), inflammatory marker testing (hsCRP, IL-6 and IP-10). In addition, the T-cell subsets including CD8 activation levels in the three Beijing centers, were analyzed in PUMCH Lab. CD4 counts of participants outside Beijing were measured at each clinical center. To ensure the consistency and quality of T-cell subset analysis, inter-laboratory quality control was carried out and checked at designated points during the study.

Viro-immunological analysis

Plasma HIV-1 viral load was measured at PUMCH Lab by COBAS AmpliPrep/COBAS TaqMan V2.0 RT-PCR (Roche). The lower detection limit is 20 copies per milliliter. T-cell subsets were measured using six-color FACSCanto flow cytometer (BD Biosciences) or CytoFlex flow cytometer (Beckman Coulter Life Science) at participating centers, including T cells (CD45+CD3+), CD4 T cells (CD45+CD3+CD4+), and CD8 T cells (CD45+CD3+CD8+). In PUMCH Lab, exploratory CD8 markers (CD45+CD3+CD8+CD38+ and CD45+CD3+CD8+HLA-DR+) were further measured for participants in three Beijing centers.

Plasma levels of hsCRP, IL-6, and IP-10 were measured in PUMCH Lab, through Simple Plex immunoassay per protocols (R&D Systems, Inc).

Statistical analysis

The efficacy analysis was based on the full analysis set (FAS), which included all randomized patients with at least one posttreatment efficacy assessment, and the per-protocol set (PPS), which included patients who completed the study with proper study drug exposure and without any major protocol violations. The safety analysis was based on the safety analysis set (SAS), which included all randomized patients who received at least one dose of study drug. All safety results were descriptive, listed and summarized.

The primary efficacy differences between treatments on CD4 counts and inflammatory marker level were assessed through the t-test and analysis of covariance (ANCOVA) model. The ANCOVA model, including the baseline CD4 counts as the covariate, study group and center as fixed factors, generated the Least square means (LSMEANs) and relative 95% confidence interval (95% CI) in both FAS and PPS. Treatment centers with less than 10 enrolled patients were merged in this model. Cochran–Mantel–Haenszel (CMH) chi-square test was used to compare the effects on CD4 improvement and other categorical data across different treatments in both FAS and PPS.

Missing values in primary efficacy analysis (i.e., CD4 counts and levels of inflammatory markers) were imputed with the last recorded data before the missing assessment. No imputation was performed in the secondary and exploratory analysis for missing values, differences between study groups were analyzed with unimputed data.

Statistical analysis was performed using SAS (version 9.4) and SPSS (version 25.0). All statistical tests were two-sided, and P values ≤ 0.05 were considered statistically significant for treatment differences.

Role of the funding source

This work was supported by the CAMS Innovation Fund for Medical Sciences (CIFMS) (2021-I2M-1–037), the National key technologies R&D program for the 13th five-year plan (2017ZX10202101), and Shanghai Pharmaceuticals Holding Co., Ltd. The funder of the study reviewed and commented on the initial concept sheet but had no role in the development of the final protocol, data collection, data analysis, data interpretation, writing of the report, or in the decision to submit the paper for publication.

Results

From Dec 25, 2019 to Jul 9, 2021, a total of 277 PLWH were screened at nine centers, and 151 eligible patients were recruited and randomized (Fig. 1). Among them, two patients were found to have baseline CD4 counts over 350/mm³ before, so they did not receive assigned treatment and were excluded from the final analysis. A total of 46, 51, and 52 participants were allocated to LLDT-8 0.5 mg (LT8), LLDT-8 1.0 mg (HT8), and the placebo (PL) group, respectively. The process of enrollment and numbers of patients included in different analyses were depicted in Fig. 1.

Fig. 1.

Flow diagram for study design and enrollment. ∗ Of the 126 participants who were not randomly assigned in initial screening for study eligibility, 59 did not meet inclusion criteria, 52 met the exclusion criteria, and 15 had other reasons. † The two participants received no drug for being confirmed with CD4+ T cell >350 cells/mm³ at screening (359 and 436 cells/mm³ separately), which was against the inclusion criteria.

Demographic and clinical features of all participants were summarized in Table 1. The vast majority of participants were men (97.4%). The median age was 41 years old (range 22–64), with 61.1% under the age of 45. Participants received a median of 6.1 years of ART, ranging from 4.1 to 14.9 years. Most patients were receiving a NNRTI-based regimen (70.5%), most frequently the national free regimen tenofovir, lamivudine plus efavirenz (45.6%). Participants receiving boosted PI/r-based and INSTI-based regimens accounted for 18.1% and 11.4%, respectively. On enrollment, all participants were virologically suppressed, and the median CD4 count was 248 cells/mm³ (range 18–347), comparable among three study groups. Around 1/4 participants had a CD4 count below 200 cells/mm³. Moreover, CD4 counts at two separate points before recruitment were also collected, showing a stable level during the previous year. The median baseline CD4/CD8 ratio was 0.45.

Table 1.

Characteristics of the participants at baseline, according to full analysis set^a.

Characteristics	Total (N = 149)	LLDT-8 1 mg (N = 46)	LLDT-8 0.5 mg (N = 51)	Placebo (N = 52)
Male sex — no. (%)	145 (97.3)	46 (100)	48 (94.1)	51 (98.1)
Age, median (range) — yr	41 (22–64)	41 (28–64)	41 (26–61)	44 (22–64)
Age group — no. (%)
18–45 yr	91 (61.1)	29 (63.0)	35 (68.6)	27 (51.9)
45–65 yr	58 (38.9)	17 (37.0)	16 (31.4)	25 (48.1)
Ethnicity Han — no. (%)	142 (95.3)	44 (95.7)	48 (94.1)	50 (96.2)
BMI — kg/m2	22.3 ± 2.5	22.1 ± 2.2	22.5 ± 2.9	22.2 ± 2.4
ART duration, median (range) — yr	6.1 (4.1–14.9)	6.1 (4.1–11.7)	6.1 (4.1–14.9)	6.0 (4.2–10.7)
ART regimens — no. (%)
NNRTI-based	105 (70.5)	31 (67.4)	39 (76.5)	35 (67.3)
TDF+3TC + EFV	68 (45.6)	19 (41.3)	27 (52.9)	22 (42.3)
PI/r-based	27 (18.1)	12(26.1)	6 (11.8)	9 (17.3)
INSTI-based	17 (11.4)	3 (6.5)	6 (11.8)	8 (15.4)
CD4 lymphocyte count, median (range) — cells/mm3	248 (18–347)	249 (18–343)	243 (87–347)	249 (119–341)
CD4 lymphocyte stratification— no. (%)
<200 cells/mm3	38 (25.5)	10 (21.7)	15 (29.,4)	13 (25.0)
≥200 cells/mm3	111 (74.5)	36 (78.3)	36 (70.6)	39 (75.0)
CD8 lymphocyte count, median (range) — cells/mm3	518 (134–1611)	527 (134–1196)	495 (177–1383)	538 (255–1161)
CD4/CD8 ratio, median (range)	0.45 (0.10–1.61)	0.45 (0.10–1.61)	0.47 (0.21–1.27)	0.45 (0.15–1.26)

^aThe analysis set for participants who underwent randomization in this study. Percentages may not total 100 because of rounding. Plus–minus values are means ± SD. BMI denotes body mass index; SD, standard deviation; ART, antiretroviral therapy; NNRTI, non-nucleoside reverse transcriptase inhibitor; TDF, tenofovir; 3TC, lamivudine; EFV, efavirenz; PI/r, ritonavir-boosted protease inhibitor; INSTI, integrase strand transfer inhibitor.

During the study, all participants remained virologically suppressed. Seven participants had changed their ART per clinical needs as shown in Table S10, mostly due to adverse side effect. Data of these participants were included in the final analysis.

In the FAS population, the CD4 count at Week 48 was 303 cells/mm³ (95% CI 274, 333) in HT8 group, 299 cells/mm³ (95% CI 277, 320) in LT8 group, and 281 cells/mm³ (95% CI 261, 302) in PL group, respectively. The mean absolute increase of CD4 count from baseline was 63 cells/mm³ in HT8 group (95% CI: 41, 84), 49 cells/mm³ in LT8 group (95% CI: 29, 68), compared with 32 cells/mm³ in PL group (95% CI: 13, 51). All groups gained statistically significant CD4 increase compared to the baseline (Supplementary Table S4). However, CD4 increase in HT8 group was significantly higher than that of PL group (p = 0.036). The LT8 group also gained more CD4 cells than PL group, but the difference was not significant at 48 weeks (Fig. 2B, Table 2). The HT8 group exhibited a robust CD4 increase during the first month of treatment and kept relatively stable afterwards, while CD4 T cells in LT8 group tended to rise gradually over 48 weeks (Fig. 2A). In addition, changes of CD4 counts over the study were more consistent in the higher dose 1 mg group (Fig. S3 and Supplementary Table S4). PPS-based analyses showed similar results, as shown in Fig. S1 and Supplementary Table S5.

Fig. 2.

The CD4⁺T-cell recovery and cytokine changes among FAS population. Pre-screening timepoints 1 and 2 were within one year before the screening. Data were shown and exhibited as mean and 95% CI. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001; ns, not significant when compared to the placebo group.

Table 2.

Primary efficacy analyses^a.

Efficacy outcome	LLDT-8 (1 mg)	LLDT-8 (0.5 mg)	Placebo
Primary endpoints
Change in the CD4 counts at Wk 48, cells/mm3
Mean	62 (40, 84)	50 (29, 70)	29 (13, 45)
LSMEAN	63 (41, 84)	49 (29, 68)	32 (13, 51)
LSMEAN Difference	14 (−14, 43)b; 30 (2, 59)c	16 (−10, 43)c	–
P-value	0.338b; 0.036b	0.224c	–
Proportion of patients with the CD4 count increased by ≥ 50 cells/mm3 at Wk 48
No. (%)	23 (50.0)	23 (45.1)	18 (34.6)
P-value	0.584b; 0.077c	0.302c	–
Change in inflammatory factors (IP-10, hsCRP, IL-6) at Wk 48d
IP-10, pg/ml
Mean	−72.1 (−99.9, −44.3)	−42.5 (−64.3, −20.7)	−22.8 (−48.7, 3.1)
LSMEAN	−72.1 (−97.7, −46.5)	−42.5 (−66.8, −18.2)	−22.8 (−47.1, 1.5)
LSMEAN Difference	−29.6 (−64.9, 5.7)b; −49.3 (−84.6, −14.0)c	−19.7 (−54.0, 14.7)c	–
P-value	0.100b; 0.007c	0.260c	–
hsCRP, mg/L
Mean	−2.7 (−8.0, 2.5)	−0.5 (−1.0, −0.1)	−0.3 (−1.2, −0.6)
LSMEAN	−2.7 (−5.7, 0.2)	−0.5 (−3.3, −2.3)	−0.3 (−3.1, −2.5)
LSMEAN Difference	−2.2 (−6.3, 1.8)b; −2.5 (−6.5, 1.6)c	−0.3 (−4.2, 3.7)c	–
P-value	0.288b; 0.234c	0.895c	–
IL-6, pg/ml
Mean	−2.1 (−7.2, 2.9)	0.7 (−0.6, 1.9)	2.6 (−2.4, 7.6)
LSMEAN	−2.1 (−6.3, 2.0)	0.7 (−3.3, 4.6)	2.6 (−1.3, 6.5)
LSMEAN Difference	−2.8 (−8.5, 3.0)b; −4.7 (−10.5, 1.0)c	−1.9 (−7.5, 3.6)c	–
P-value	0.340b; 0.106c	0.493c	–
Secondary endpoints
Change in the CD4/CD8 ratio at Wk 24
Mean	0.04 (0.01, 0.07)	0.04 (0, 0.08)	−0.01 (−0.04, 0.01)
LSMEAN	0.04 (0.01, 0.07)	0.04 (0.01, 0.07)	−0.01 (−0.04, 0.02)
LSMEAN Difference	0 (−0.04, 0.04)b; 0.05 (0.01, 0.10)c	0.05 (0.01, 0.10)c	–
P-value	0.987b; 0.017c	0.015 c	–
Change in the CD4/CD8 ratio at Wk 48
Mean	0.05 (−0.01, 0.11)	0.04 (−0.01, 0.09)	0.04 (0, 0.07)
LSMEAN	0.05 (0, 0.10)	0.04 (−0.01, 0.09)	0.04 (−0.01, 0.08)
LSMEAN Difference	1.1 (−0.06, 0.08)b; 2.2 (−0.05, 0.08)c	0.01 (−0.06, 0.07)c	–
P-value	0.809b; 0.646c	0.829c	–
Proportion of patients with the CD4 count increased by ≥ 20% at Wk 24
No. (%)	21 (45.7)	26 (51.0)	10 (19.2)
P-value	0.833b; 0.005c	<0.001c	–
Proportion of patients with the CD4 count increased by ≥ 20% at Wk 48
No. (%)	23 (50.0)	24 (47.1)	18 (34.6)
P-value	0.874b; 0.240c	0.423c	–
Proportion of patients with the CD4 count <200 cells/mm3 at baseline and increased to ≥ 200 cells/mm3at Wk 24e
No. (%)	6 (60.0)	12 (80.0)	5 (38.5)
P-value	0.252b; 0.657c	0.023c	–
Proportion of patients with the CD4 count <200 cells/mm3 at baseline and increased to ≥ 200 cells/mm3at Wk 48e
No. (%)	9 (90.0)	11 (73.3)	9 (69.2)
P-value	0.769b; 0.339c	0.827c	–

Abbreviations: LSMEANS: least-squares means; CI: confidence interval; IP-10: Interferon gamma-induced protein 10; hsCRP: high-sensitivity C-reactive protein; IL-6: interleukin 6.

^aThe analyses were set for the FAS population. Mean, LSMEAN, and LSMEAN difference were represented as data and 95% CI. LSMEANs Difference and P-values were ^bversus the LLDT-8 (0.5 mg) group; or ^cversus the placebo group.

^dLLDT-8 (1 mg) group included 46 patients; LLDT-8 (0·5 mg) group included 51 patients; placebo group included 51 (1 missing baseline data) patients.

^eLLDT-8 (1 mg) group included 10 patients; LLDT-8 (0·5 mg) group included 15 patients; placebo group included 13 patients.

At Week 48, 23 of 46 (50%) participants in HT8, 23 of 51 (45.1%) in LT8, and 18 of 52 (34.6%) in PL group had a CD4 count increase over 50 cells/mm³ (Table 2, Fig. 2C). Both LLDT-8 groups had more participants achieving a discriminating CD4 increase, and the proportion increased with higher LLDT-8 dosing, but was statistically insignificant. At week 24, more participants in HT8 group achieved a CD4 increase ≥20%, however, such superiority became unremarkable at Week 48 (Table 2). Changes of CD4/CD8 ratio were comparable between the three groups at Week 48.

CD4 increase in the high-dose LLDT-8 group was more significant in participants over 45 years old. There were 17, 16, and 25 participants aged over 45 in the HT8, LT8 and PL groups, respectively. After 48 weeks of 1 mg LLDT-8 treatment, a mean increase of 96 cells/mm³ (95% CI: 54, 137) were achieved in participants over 45 years old, compared with 20 cells/mm³ in PL group (p < 0.001, Fig. 2G–H and Supplementary Table S6).

The efficacy of LLDT-8 was also more prominent in participants with a baseline CD4 count 200–350 cells/mm³. In these participants, a consistent superiority of CD4 increase was observed in HT8 over PL. In contrast, in those with a baseline CD4 count below 200 cells/mm³, effect of LLDT-8 was insignificant at Week 48 (Fig. 2I–J, Table S7).

18 of 46 participants in HT8 group, 21 of 51 in LT8 group and 18 of 52 in PL group received inactive SARS-CoV-2 vaccination (Sinopharm, or Sinovac) during the study. Among them, 10, 11 and 10 participants in HT8, LT8 and PL group received at least one vaccine within three months before the last visit. Stratified analysis showed that benefit from high-dose LLDT-8 in CD4 counts became insignificant in participants receiving SARS-CoV-2 vaccination regardless of the time point (Table S8).

At Week 48, the mean change of serum IP-10 level was −72.1 pg/ml (95% CI −97.7, −46.5) in HT8 group, −42.5 pg/ml (95% CI −66.8, −18.2) in LT8 group, and −22.8 pg/ml (95% CI −47.1, 1.5) in PL group, respectively. LLDT-8 treatment led to reduction of IP-10 levels in both treated groups, and 1 mg daily LLDT-8 induced more substantial IP-10 decrease compared with the placebo (p = 0.007, Fig. 2D, Table 2). HsCRP and IL-6 also showed slight decrease at Week 48 in the HT8 group, but with limited clinical indications since they were both within normal ranges (Fig. 2E–F, Table 2).

Association between changes in IP-10 levels and CD4 counts at each time point were analyzed (Fig. S2). Interestingly, dynamics of plasma IP-10 level changes showed similar shape to that of the CD4 count changes. Of note, in the HT8 and PL groups, levels of IP-10 decrease were significantly associated with CD4 increase (Pearson's r = 0.18, p = 0.005).

The cellular activation levels of CD8 T cells were measured in 62 participants from the three Beijing centers, including 19 in HT8 group, 21 in LT8 group and 22 in PL group. The surface expressions of CD38 or HLA-DR on CD8 T cells were comparable among the three groups (Fig. S4 and Tables S9A-9B).

All participants stayed virologically-suppressed during the study. Treatment-emergent adverse events (TEAEs) were reported in 41 of 46 (89.1%) participants in HT8 group, 43 of 51 (84.3%) in LT8, and 42 of 52 (80.7%) in PL group. A total of 610 events were reported, and 164 of them were presumably considered possibly, probably, or related to the study drug. The most frequently reported (>5%) drug-related TEAEs included hyperlipidemia, increased alanine transaminase (ALT), decreased white blood cells, decreased neutrophils, increased gamma-glutamyl transferase (GGT) and hepatic steatosis (Table 3). Numbers of TEAEs were comparable between the three groups. TEAEs ≥ Grade 3 were slightly more frequent in the HT8 group, mostly neutrophil decrease and dyslipidemia, yet no severe infection occurred, nor did any drug-related TEAEs lead to study discontinuation. No drug-related SAEs were reported.

Table 3.

Safety analyses^a.

Adverse events	LLDT-8 (1 mg)		LLDT-8 (0.5 mg)		Placebo
	No. of participants (%)	No. of eventsb	No. of participants (%)	No. of eventsb	No. of participants (%)	No. of eventsb
TEAEs	41 (89.1)	195	43 (84.3)	217	42 (80.8)	198
0 ∼ Wk 24	40 (87.0)	144	42 (82.4)	150	40 (76.9)	130
Wk 24 ∼ Wk 48	29 (63.0)	51	23 (45.1)	67	27 (51.9)	68
TEAEs with grade ≥3	11 (23.9)	14	8 (15.7)	11	5 (9.6)	5
Study drug-related TEAEs	26 (56.5)	67	19 (37.3)	46	24 (46.2)	50
Most frequent (≥ 5%) study drug-related TEAEsc
Hyperlipidemiad	15 (32.6)	16	6 (11.8)	8	5 (9.6)	7
Increased ALT	4 (8.7)	4	7 (13.7)	10	2 (3.8)	2
Decreased WBC	5 (10.9)	5	5 (9.8)	5	1 (1.9)	1
Decreased neutrophil	7 (15.2)	8	2 (3.9)	2	0	0
Increased GGT	2 (4.3)	2	3 (5.9)	3	2 (3.8)	2
Hepatic steatosis	4 (8.7)	4	0	0	1 (1.9)	1
Related TEAEs with grade ≥3	4 (8.7)	4	1 (2.0)	1	1 (1.9)	1
TEAEs leading to discontinuation	0	0	1 (2.0)	1	0	0
Related TEAEs leading to discontinuation	0	0	0	0	0	0
SAEs	0	0	2 (3.9)	2	1 (1.9)	1
0 ∼ Wk 24	0	0	2 (3.9)	2	0	0
Wk 24 ∼ Wk 48	0	0	0	0	1 (1.9)	1
Study drug-related SAEs	0	0	0	0	0	0
Death	0	0	0	0	0	0

^aThe analyses were set for the SAS population. Related TEAE category includes Possible, Probable, or Definite TEAEs which was presumed to be associated with study drug. TEAE denotes treatment-emergent adverse event; ALT, alanine transaminase; WBC, white blood cell; GGT, gamma-glutamyl transferase; SAE, serious adverse event.

^bNo. of events was defined as the number of occurrences of adverse events. Same adverse events in a participant at a consecutive time would be counted only once.

^cStudy drug-related TEAEs with an incidence rate ≥5% among the three groups were listed and generalized as preferred terms (PTs) according to MedDRA 24.0. PTs were ranked in descending order of the total number among the SAS population. The normal range of ALT was 9–50 U/L for male and 7–40 U/L for female; the normal range of WBC was 4.00–10.00×10⁹/L; the normal range of neutrophil was 2.00–7.50×10⁹/L; and the normal range of GGT was 10–60 U/L for male and 7–45 U/L for female.

^dDefinition of hyperlipidemia included hypercholesterolemia, hypertriglyceridemia, and the elevated concentrations of both cholesterol and triglyceride. The normal range of cholesterol was 2.85–5.70 mmol/L and normal range of triglyceride was 0.45–1.70 mmol/L. The incidence rates of hypercholesterolemia were 15.2% (7/46) at HT8 group, 5.9% (3/51) at LT8 group, and 5.8% (3/52) at placebo group. The incidence rates of hypertriglyceridemia were 4.3% (2/46) at HT8 group, 2.0% (1/51) at LT8 group, and 7.7% (4/52) at placebo group.

Discussion

Despite progressive advances in ART, INRs have always been a concern, since they account for a unneglected group associated with increased morbidity and mortality resulted from both immunodeficiency and immune activation. In this randomized controlled study, we evaluated the efficacy of LLDT-8, an analogue of extract from the Chinese herb TwHF, in improving immune recovery and reducing inflammation in INRs. We found that LLDT-8 administrated at 1 mg daily could enhance CD4 recovery in long-term suppressed INRs, especially in patients older than 45 years old. This effect was associated with a marked reduction of inflammatory markers. Current results confirmed LLDT-8 as a promising option for treating INRs, and further studies should be considered.

A range of treatment has been studied in INRs. However, most of them were observational and uncontrolled studies with limited evidence power, especially when evaluating CD4 counts that may fluctuate with multiple factors. In addition, the CD4 recovery following effective ART usually takes longer than the virological response, and may differ substantially between individuals.^1,2,29 To ensure the value of the study, we recruited patients suppressed for at least 3.5 years, and reviewed their CD4 records before screening, to make sure the insufficient CD4 level had been stable for a long period. We used a double blind, double dummy, and randomized design to maximally avoid the confounding impact and placebo effect. Efficacy of LLDT-8 was confirmed in both FAS and PPS analyses in our study, and it has been the only reported agent that showed efficacy in a strictly designed clinical trial. As an oral formulation with efficacy and favorable safety profiles, it would provide an option for CD4 reconstitution in INRs.

The CD4 homeostasis is maintained by balanced production and destruction. Early approaches to INRs aimed to promote production or release of CD4 cells, among which IL-2 was most thoroughly studied, but failed to demonstrate survival benefit.¹⁰ LLDT-8 works in the other way. It is a novel analogue of triptolide, the active gradient of TwHF, with lower cytotoxicity and major immunosuppressive activity.^28,30 Unlike IL-2 evoking proliferation and expansion of CD4 T cells, LLDT-8 mainly modulates chronic inflammation, which leads to reduced CD4 apoptosis or exhaustion.²⁶ Our previous studies showed that IFN signaling pathway and the signal transducer and activator of transcription 1 (STAT1) are particularly enhanced in INRs, which are also the major effector targets of TwHF and its bioactive triptolide.^25,26 In this study, LLDT-8 reduced plasma IP-10 levels, one of the most sensitive markers in HIV-associated inflammation.³¹ Moreover, in all study groups the curves of changes in IP-10 mimic those of CD4 alterations, indicating that inflammation status is probably one of the main driving forces for CD4 depletion in these patients. These findings were consistent with the previous reports in mice models and human PBMCs that LLDT-8 could effectively inhibit the IFN pathways.^28,32,33 Levels of CD8 activation by CD38/HLA-DR expression seemed not differentiated among the study groups, probably because they were already near the normal range at study baseline. The CD4/CD8 ratio was also comparable among three groups at 48 weeks, and a longer treatment duration might induce more significant changes in the ratio. Nevertheless, since chronic inflammation is also responsible for occurrence of non-AIDS-related complications, LLDT-8 may provide additional benefit in this population other than CD4 reconstitution.^20,34,35

Among all participants, those over 45 years showed more prominent response to LLDT-8, likely due to the older age where inflammation reduction may be more decisive on CD4 levels. This was supported by the CD4 trends in the control groups of different ages, which reflected the spontaneous CD4 changes without therapeutic interventions. However, it remains unclear why those with a baseline CD4 count 200–350/mm³ gained more increase with LLDT-8 administration. Since number of participants with baseline CD4 count <200/mm³ was relatively small in our study, a larger sample size of such patients may provide more information. In this study, the two LLDT-8 dose groups showed varied dynamics of CD4 increase, which may indicate different action mode in the two groups. Furthermore, subgroup analysis showed that SARS-CoV-2 vaccination during the study could compromise the efficacy of LLDT-8 in PLWH, probably due to the immunological variations caused by vaccination,^36,37 which may also bring unpredictable impact on CD4 counts. Therefore, further follow-up of these participants were extremely important in identifying long-term effects of LLDT-8 on inflammation and CD4 dynamics. It is very possible that a longer duration of LLDT-8 treatment may be necessary to maintain the CD4 levels since it mainly works against the inflammation persistent with HIV infection.

Our study has several limitations. First, we mainly assessed the benefit of LLDT-8 in CD4 counts at 48 weeks instead of the long-term clinical events. However, the surrogate value of CD4 T cells for HIV prognosis has been well-established, and we will follow all participants to evaluate the long-term post-treatment effect. Secondly, it was a pity that we were only able to measure CD4 counts without detailed subset composition in this study. Our previous studies of TwHF suggested that LLDT-8 mainly increases memory CD4 T cells, and could also bring down the activation levels of CD4 T cells.²⁶ More thorough analysis of the CD4 responses in future studies would provide more information of its mechanism. Lastly, the present study was mainly focused on the clinical aspect. The underlying mechanisms thereof should be further studied to better characterize actions of LLDT-8.

In conclusion, our study demonstrated that LLDT-8, a novel analogue of triptolide, could enhance CD4 recovery and reduce inflammation in long-term suppressed INRs. The benefit was superior with LLDT-8 dosage 1 mg daily, and in older participants. LLDT-8 provides an option for INRs, and more studies of its underlying mechanisms are warranted.

Contributors

WC, XSL, MZ, and TSL had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. WC, MZ, and TSL decided to publish the paper. WC and TSL provided input on the trial design. WC, XJS, LJS, AL, HXZ, NH, HXW, JC, BZ, MW, YL, PM, LYG, XCW, and JHY were responsible for patients selection and enrollment. XSL, YH, LFL, XDL, LL, and TZ were responsible for samples deliberation and analysis. XSL contributed to statistical analysis. WC drafted the manuscript. XSL, LFL, XDL, JPR, and TSL critically revised the manuscript. JPR and TSL gave valuable suggestions for data analysis. All authors contributed to conducting the trial.

Data sharing statement

After achieving approval from the Human Genetic Resources Administration of China and signing with contract, this trial data can be shared with qualifying researchers who submit a proposal with a valuable research question.

Declaration of interests

Min Zuo is the leader of Shanghai Pharmaceuticals Holding Co., Ltd., SPH and supported this research, he received no personal financial payment for this work. All other authors declare no competing interests.