Secondary Logo

Journal Logo

Clinical Science

Safety, Pharmacokinetics, and Antiviral Activity of AMD3100, a Selective CXCR4 Receptor Inhibitor, in HIV-1 Infection

Hendrix, Craig W MD*; Collier, Ann C MD; Lederman, Michael M MD; Schols, Dominique PhD§; Pollard, Richard B MD; Brown, Stephen MD; Jackson, J Brooks MD*; Coombs, Robert W MD, PhD; Glesby, Marshall J MD#; Flexner, Charles W MD*; Bridger, Gary J PhD**; Badel, Karin BScN**; MacFarland, Ronald T PhD**; Henson, Geoffrey W PhD**; Calandra, Gary MD**for the AMD3100 HIV Study Group

Author Information
JAIDS Journal of Acquired Immune Deficiency Syndromes: October 1st, 2004 - Volume 37 - Issue 2 - p 1253-1262
doi: 10.1097/01.qai.0000137371.80695.ef


Current antiretroviral (ARV) therapy for HIV-1 infection most commonly uses combinations of drugs that suppress HIV replication via inhibition of 1 of 2 viral-specific enzymes, reverse transcriptase (RT), or protease (PR).1 A recently approved ARV drug, enfuvirtide (T-20), an HIV fusion inhibitor, is the first approved drug to inhibit HIV infection through action at the cell surface rather than by inhibiting viral enzymes. Although the other classes of ARV drugs decrease HIV RNA and prolong life, many patients do not tolerate these medications well or develop antiviral resistance. The identification of new classes of well-tolerated ARV drugs with unique mechanisms of action thus remains an important therapeutic objective.

HIV entry is a multistep process, which makes it an attractive therapeutic target. HIV infects target cells via interactions between the viral envelope surface glycoprotein gp120 interaction and the target cell receptor (CD4) and a coreceptor (CXCR4 or CCR5), followed by gp41-mediated fusion.2 Of the compounds reported to interfere with these early steps in the HIV replication cycle, only the fusion inhibitor enfuvirtide has been approved for use in the United States. Enfuvirtide potently suppresses HIV RNA in HIV-infected subjects in combination with an optimized background regimen.3–5 The bicyclam AMD3100 (l,1′-[1,4-phenylenebis(methylene)]-bis-1,4,8,11-tetraazacyclotetradecane) exhibits potent and selective inhibition of X4-tropic HIV replication in vitro by binding to the chemokine receptor CXCR4.6–8 Historically, X4-tropic viruses have been defined as T-cell–tropic (T-tropic) and syncytium-inducing (SI) HIV viruses. The equating of X4 and SI viruses is outdated, because many R5 viruses are SI for cells that express CCR5. AMD3100 does not bind to CCR5, which mediates entry of R5-tropic viruses, historically defined as macrophage-tropic (M-tropic) and non–syncytium-inducing (NSI) HIV viruses.9–12 X4 strains are considerably more pathogenic, and their appearance late in HIV infection correlates with a decline in CD4+ T-cell counts and more rapid disease progression.13–16 In contrast to R5-tropic viruses, X4 strains rarely emerge as the predominant virus population after horizontal transmission, although it is not clear whether they are not transmitted or if they are held in check by negative selection.14,17–19 In vitro, AMD3100 inhibits replication of T-tropic strains of HIV (concentration of the drug required to inhibit HIV replication by 50% [IC50] between 1.0 and 10 ng/mL), provides complete protection (>99.9%) of lymphocytes and monocytes from HIV infection at concentrations of 10 to 30 ng/mL, and inhibits syncytium formation between cocultures of persistently HIV-infected HUT-78 cells and uninfected MOLT-4 cells (1–5 μg/mL).9,20,21 In a study using the simian/HIV (SHIV) macaque model, after infection with X4 and R5 SHIV, CD8+ T-cell–mediated control of infection is more complete for X4 than for R5, possibly explaining early R5 dominance; late-stage X4 resurgence is postulated to occur as a direct result of CD8+ depletion or disruption of immune homeostasis such that X4 predominates.22

The in vivo anti-HIV activity of AMD3100 has been demonstrated in the severe combined immunodeficiency (SCID)-Hu Thy/Liv mouse model of HIV infection using viral p24 capsid antigen (p24 Ag) production in transplanted human thymocytes as the end point. At subcutaneous doses greater than 1 mg/kg/d, AMD3100 decreased p24 Ag levels of a primary isolate in a dose-dependent manner. In continuous subcutaneous infusion studies using Alzet pumps, steady-state plasma levels of 40 ng/mL (1.1 mg/kg/d) and 117 ng/mL (3.8 mg/kg/d) were associated with significant (P = 0.004 and 0.003, respectively) reductions in plasma p24 Ag levels up to 80% or more without clinical signs of toxicity or weight loss.23

The first human study of AMD3100 evaluated the safety, bioavailability, and single-dose pharmacokinetics of progressively higher doses of AMD3100, from 10 to 80 μg/kg, administered intravenously over 15 minutes in successive cohorts of 3 to 5 healthy volunteers.24 Five volunteers also received a single subcutaneous injection (40 or 80 μg/kg), and 3 volunteers also received oral doses (80 or 160 μg/kg). AMD3100 was well tolerated in all volunteers by all routes of administration. Six of the 12 subjects (50%) had mild transient symptoms, primarily gastrointestinal in nature. Most subjects experienced a transient dose-dependent increase in white blood cell (WBC) count 1.5 to 3 times baseline (BL) values within 6 hours of the infusion, which returned nearly to BL by 24 hours after dosing. The pharmacokinetic profiles were dose proportional, with an estimated half-life of 3.6 hours. Absorption after subcutaneous administration was good (87% bioavailability), but no drug was detectable in the blood after oral dosing.

We undertook this next study to determine the safety, pharmacokinetics, and antiviral effect of AMD3100 administered as a continuous 10-day infusion in subjects with stable HIV infection.


Study Design

This was a phase 1/2, multicenter, open-label, dose escalation study of AMD3100 administered as a 10-day continuous infusion of AMD3100 at a dose of 2.5, 5, 10, 20, 40, 80, and 160 μg/kg/h. Cohorts of 3 or more patients were enrolled into each dose level in succession, with escalation based on predetermined safety and antiviral end points. The first 8 subjects were enrolled without regard to SI/NSI phenotype, and these evaluations also were not performed on 2 subsequent patients. Beginning in the 10-μg/kg/h cohort, 2 enrollment strata were defined to speed enrollment and spare volunteers with X4 virus (those believed to be the only patients with an expectation of an antiviral response) for antiviral activity analysis at higher doses. Stratum 1 enrolled subjects regardless of SI/NSI status, with NSI patients to be used only for the safety-evaluable cohort, whereas SI patients were to be used for the efficacy- and safety-evaluable cohorts. Stratum 2 patients were required to have SI phenotype and were included in the efficacy-evaluable and safety-evaluable cohorts. Stratum 1 required a minimum of 3 subjects for completion, whereas stratum 2 required a minimum of 2 subjects before escalation. The local institutional review board approved the study protocol at all centers, and all subjects provided written informed consent to participate. A total of 40 subjects were enrolled at 6 participating centers: Johns Hopkins University Drug Development Unit, University of Washington AIDS Clinical Trials Unit (ACTU), Case Western Reserve University/University Hospitals of Cleveland, University of Texas at Galveston, Cornell Clinical Trials Unit, and AIDS Research Alliance (Los Angeles).


Subjects enrolled had HIV infection without acute complications and were on stable or no ARV therapy for 4 weeks preceding the study. Subjects in the efficacy-evaluable cohort had a stable plasma HIV RNA level >5000 copies/mL (with less than 0.5 log10 difference over 2 weeks) and SI phenotype based on MT-2 assay.25 All subjects had CD4+ T-cell counts >50 cells/mm3 (later lowered to >10 cells/mm3) and a Karnofsky performance score of ≥80%. The use of concomitant medication for chronic medical conditions, including ARV drugs, did not preclude entry into the study. The dose and regimen were to be maintained without change from screening through study day 18; at that time, any or all medications could be changed as required.

Study Procedures

Consenting subjects were screened to assess their health by way of history and physical examination and laboratory evaluations. Eligible subjects were enrolled in the study, which included assignment to stratum 1 (SI/NSI status independent) or stratum 2 (SI only) based on screening by the MT-2 assay for SI/NSI status after the initial 12 subjects were enrolled. Enrolled subjects were admitted to an inpatient unit for 11 nights. On the morning after admission, subjects began a continuous 10-day infusion of AMD3100 at the prescribed dose administered via infusion pump at 40 mL/h (daily solutions were prepared by dilution of a 10-mg/mL solution of AMD3100 in 0.9% saline). Blood for safety, pharmacokinetics (including terminal blood sampling after cessation of the infusion), and antiviral activity was collected at specified times throughout the study. Continuous cardiac monitoring began on admission and continued throughout the inpatient phase of the study.

HIV RNA was measured in plasma samples from all study participants before the initiation of study drug, at day 6, at the termination of study drug (day 11), at day 18, and at day 39. A second assessment of MT-2 status (SI/NSI) and viral tropism phenotype (X4, dual tropic, or R5) was measured on most study participants from blood collected before the initiation of study drug and at the termination of study drug (day 11). Patients returned for follow-up visits 1 week, 1 month, and 3 months after discharge from the inpatient unit.

In stratum 1, at least 3 subjects were dosed in a given dose cohort. Once a cohort was filled, entry to the next cohort awaited completion of the previous cohort; review of safety data; and a determination made regarding attribution of any grade 3, grade 4, or serious adverse events to study drug. Dose escalation continued until safety and efficacy end points were reached. If more than 1 patient within a cohort experienced an adverse event of grade 3 or 4 that was judged to be related, or possibly related, to study medication, the cohort size was doubled. Escalation continued unless 2 or more subjects in the expanded cohort experienced grade 3 or 4 adverse events that were related or possibly related to study drug. The highest dose at which 1 or fewer subjects experienced drug-related toxicity was considered the maximum tolerated dose (MTD). The antiviral activity end point was defined as a ≥1.0-log10 decrease in plasma HIV RNA in at least 2 subjects within a dose cohort.

AMD3100 Assay

Plasma concentrations of AMD3100 were determined by high-performance liquid chromatography (HPLC) with electrochemical detection using a validated assay (MDS Laboratories, Saint Laurent, Quebec, Canada). After addition of the internal standard (AMD2763), plasma samples were extracted with methyl t-butyl, followed by acidification with trifluoroacetic acid and back extraction. Chromatography was performed with an Isochrom pump and PLRP-S column (5 μm, 250 mm × 4 mm) using an acetonitrile/water (33.7%: 65.5%) mobile phase containing 0.5% t-butyl ammonium hydroxide, 0.3% sodium lauryl sulfate, and EDTA. Electrochemical detection was carried out using a Coulochem II electrochemical detector. AMD3100 standards were prepared in the range of 5 to 250 ng/mL using pooled human plasma (lower limit of quantitation (LLOQ) = 5 ng/mL). Standards and samples were extracted and assayed in an identical manner using an internal standard (AMD2763). Quality control samples (15, 120, and 200 ng/mL) were used to determine interday precision (%CV) and accuracy (% nominal concentration). The quality control sample %CV ranged from 4.4% to 8.6%. The % nominal concentration ranged from 97.9% to 100.4%. Intraday precision and accuracy (from assay validation) ranged from 1.1% to 5.7% and from 96.5% and 98.4%, respectively. Correlation coefficients for standard curves from sample analysis runs were equal to or greater than 0.9914 for AMD3100.

Pharmacokinetic Analysis

Blood for pharmacokinetic analysis was collected immediately before dosing and at 1, 2, 4, 8, 24, and 48 hours after initiation of infusion; before bag changes on day 4, 6, 8, and immediately before discontinuation of infusion; and at 30 minutes and 1, 2, 4, 8, 12, and 24 hours after discontinuation of the infusion. Noncompartmental analyses were performed using PhAST (version 2.3-001) software. Individual pharmacokinetic parameters, clearance (Cl), maximum concentration (Cmax), steady-state concentration (Css), terminal elimination half-life (t1/2), and area under the concentration-time curve extrapolated to infinity (AUC0–∞) using the log-linear trapezoidal rule were calculated using all available time points and then summarized by dose cohort. Pharmacokinetic parameters were correlated with dose to test dose proportionality (Pearson correlation coefficient).


HIV RNA in blood plasma was measured using the commercial Roche HIV-1 RNA Amplicor Monitor Assay, with a limit of quantitation 400 copies/mL.

Peripheral Blood Mononuclear Cell Coculture Assay

HIV was cultivated by mixing unstimulated patient lymphocytes with mitogen-stimulated donor lymphocytes from healthy HIV-negative individuals. Donor cells were stimulated for 3 days with the mitogen phytohemagglutinin (PHA; 2 μg/mL), and the lymphocytes were then maintained in the presence of interleukin-2 (IL-2; 25 IU/mL). The patient lymphocytes were mixed with PHA-stimulated donor lymphocytes without prior PHA stimulation.

Culture medium containing patient cells (2–5 × 106) mixed with donor lymphocytes (5 × 106) and IL-2 (1 ng/mL) (R&D Systems) were incubated in a culture flask. The supernatant virus stock was taken at 1 and 2 weeks after the start of the coculture. Virus production was measured using an enzyme-linked immunosorbent assay (ELISA) to detect the HIV p24 capsid protein (DuPont-Merck Pharmaceutical Co., Wilmington, DE), and virus stocks were diluted into small aliquots and stored at −80°C.

AMD3100 Susceptibility in a Peripheral Blood Mononuclear Cell Assay

Virus stocks were evaluated for their sensitivity to AMD3100 in freshly isolated peripheral blood mononuclear cells (PBMCs) from a healthy donor as previously described.11 Virus production in the supernatant was measured by HIV-1 core antigen by p24 Ag ELISA. The data were expressed as IC50.

Coreceptor Use

If the virus production was sufficiently high (>1 × 104 pg of p24 Ag/mL), the virus stock (at different dilutions) was added to U87.CD4 cells transfected with either CXCR4 or CCR5. The human astroglioma U87 cells were kindly provided by Dan R. Littman (Skirball Institute of Biomolecular Medicine, NY) and were cultured in Dulbecco modified Eagle medium (Life Technologies) containing 10% fetal bovine serum (FBS; BioWhitaker), 0.01 M of HEPES buffer (Life Technologies), and 0.2 mg/mL of gentamicin (G-418 sulfate; Life Technologies). Coreceptor use on U87.CD4.CCR5 or U87.CD4.CXCR4 cells was determined by the method of Bjorndal et al.26 Briefly, the U87.CD4-transfected cells were seeded in 24-well plates (2 × 104 cells/well in 1 mL of culture medium), and 1000, 5000, or 10,000 pg of p24 Ag of the HIV-1 isolates was added. The cytopathic effect of virus replication in the cell cultures was evaluated microscopically at days 5 through 7 after infection. Culture supernatants were collected at days 5 through 7 and analyzed for HIV-1 core antigen by p24 Ag ELISA.

MT-2 Assay

MT-2 cells were infected with subjects’ HIV isolates derived from PBMC coculture. Cell cultures were monitored for syncytium formation for up to 14 days according to established AIDS Clinical Trials Group laboratory protocol.25 During the study, for purposes of strata-specific enrollment, these assays were performed at 2 clinical centers: Johns Hopkins University (J.B.J.) and University of Washington (R.W.C). Poststudy assessment of tropism shifts within individuals was performed at Rega Institute for Medical Research (D.S.).

Statistical Analysis

Safety, pharmacokinetic, and virologic data from each group of patients were summarized using descriptive statistics, including means, standard deviations, medians, and ranges. For calculating HIV RNA changes, the mean values of log10-transformed HIV RNA at BL and at completion of therapy have been used.


Subject Characteristics

A total of 40 patients were enrolled, all of whom were included in the safety evaluation cohort. Based on results of the enrollment MT-2 assay, 18 patients were infected with NSI viruses (45%); 12 were infected with SI viruses (30%, 5 in stratum 2 and 7 in stratum 1); and SI status was unknown for the remaining 10 patients (25%), who were primarily in the lower dose cohorts (9 of 10 patients). The 12 SI-positive patients were used to evaluate antiviral activity (efficacy-evaluable cohort).

The demographic characteristics of the safety population were as follows: age of 40.2 ± 6.4 years (mean ± SD), weight of 79.5 ± 18.6 kg (mean ± SD), 98% male, and 43% white. These characteristics did not vary significantly across the dosing cohorts. Subjects had a Karnofsky status of 91.3 ± 6.5 (mean ± SD). The BL HIV RNA level was 120,104 ± 164,272 copies/mL (mean ± SD; range: 6803–714,364 copies/mL), and the BL CD4+ cell count was 246 ± 205 (× 106) cells/L (mean ± SD; range: 0–691 × 106 cells/L). The efficacy-evaluable cohort (all 12 patients with initial SI phenotype) did not differ from the overall cohort in terms of demographics or HIV-related disease characteristics, including CD4+ T-cell counts (P > 0.05).

At entry, 17 patients had been receiving approved ARV medications at stable doses for at least 1 month with stable HIV RNA. One of these patients was later excluded from the antiviral analysis because of missing MT-2 stratification data, documented poor ARV compliance before study entry, falling HIV RNA values between the screening visit and study entry (although within protocol limits), and a large reduction in plasma HIV RNA (1.53 log10) during the study thought to confound the antiviral assessment.


The most common subjective complaints, regardless of attribution to study drug, included diarrhea (48%), flatulence (43%), headache (40%), nausea (35%), abdominal pain (33%), abdominal distention (25%), tachycardia (25%), dizziness (25%), and paresthesias (23%). There were no apparent dose-related trends for most adverse events. The occurrence of gastrointestinal disorders, however, showed some evidence of a dose-related trend, with a frequency of 17 (71%) of 24 patients in the low-dose group (up to 20 μg/kg/h) and 16 (100%) of 16 patients in the higher dose cohorts (40–160 μg/kg/h), mostly attributable to flatulence (25% low dose, 69% higher doses) and loose stools (12% low dose, 31% higher doses). Extremity and perioral paresthesias also seemed to be dose related, with none occurring among 24 patients in the low-dose group and frequencies of 38%, 80%, and 67% in the 3 highest dose cohorts, respectively.

Of 40 subjects enrolled, 8 patients had study drug discontinued or interrupted for serious adverse events (SAEs) or other adverse events (Table 1). Four subjects permanently discontinued study drug because of adverse events: (1) 1 patient with thrombocytopenia with a platelet nadir of 7000 cells/μL (BL of 174,000 cells/μL) received platelet transfusion (5-μg/kg/h cohort), after which platelets immediately rose to 41,000 cells/μL, followed by a slow resolution to 91,000 cells/μL 12 days later; no concomitant medications were used, antiplatelet antibody tests were negative, and bone marrow biopsy was normal; (2) 1 case of PICC line–associated staphylococcal infection associated with mild thrombocytopenia (91,000 cells/μL); the patient received antibiotics (20-μg/kg/h cohort); (3) 1 patient experienced premature ventricular contractions (PVCs) up to 25 per minute that were unifocal and uncoupled, and hypocalcemia (40-μg/kg/h cohort); and (4) 1 patient had a panic attack and paresthesias (80-μg/kg/h cohort). Apart from the PICC line infection, these events resolved after cessation of dosing the study drug and may have been drug related. The patient with PVCs had higher drug levels than the rest of the cohort (the levels were above the average for the 80-μg/kg/h cohort and below that of the 160-μg/kg/h cohort).

Table 1
Table 1:
Study Drug Discontinuation or Interruption for Reason of Adverse Event

Among those with transient drug interruptions, 1 patient had drug interrupted after PVCs (a brief run of bigeminy and then unifocal and primarily uncoupled in character) were noted. After 100 minutes, drug infusion was restarted with only occasional PVCs observed, and none were seen by the time of discharge. In this patient, low serum magnesium was treated with oral supplements. When the patient was seen 7 days after the completion of treatment, PVCs were noted, suggesting a preexisting condition. The other 3 patients with drug interruption had postural hypotension, postural tachycardia with nausea and vomiting, elevated liver enzymes, and orthostatic hypotension (see Table 1).

In addition to the thrombocytopenia and PICC line infection noted previously, other SAEs occurred in 5 subjects; all of these followed cessation of study drug by at least several weeks, and none were attributed to AMD3100. All but 1 of these SAEs were infections and included: disseminated varicella-zoster virus (event on day 74, BL CD4+ T cells = 197), gastroenteritis (day 34, BL CD4 = 66), pneumonia (day 32, BL CD4 = 28), cryptococcal meningitis (day 69, BL CD4 = 10), and elevated gamma glutamyl transferase (day 39). These infections were not unexpected in patients with such low CD4 counts. There was no association between markedly elevated AMD3100 drug levels and these AEs. Three additional subjects left the study for personal non–study-related reasons.

Vital sign abnormalities, including hypertension (67%), hypotension (25%), and tachycardia (47%), were recorded transiently in many patients, although there were no dose-related trends observed. There were only a few additional patients with minor findings of potential clinical significance, all of which returned to BL during or after AMD3100 infusion as follows: anemia (2 patients) and hypocalcemia (1 patient).

We observed a dose-related increase in WBCs, neutrophils, lymphocytes, and monocytes in all subjects (Fig. 1A). The maximum increase (mean ± SD) was 2.7-fold (±0.9), and the end of infusion increase was 2.2-fold (±0.8) (P < 0.001). The elevation was also prolonged, demonstrating a persistent 1.3-fold elevation (cohort mean) at the 18-day follow-up (P < 0.001). Using pharmacodynamic modeling, this leukocytosis (mean WBCs on study drug) reached a plateau (maximum effect [Emax]) at 3.4 (95% confidence interval [CI]: 2.9–3.9) times BL with a half-maximal effect (EC50) estimated to occur at a dose level of 11.6 μg/kg/h (95% CI: 3.5–19.6). Effects in the leukocyte subsets were similar. CD4+ and CD8+ T cells were also elevated on study drug, a mean (±SD) of 1.65 (±0.75) and 1.72 (±0.70) times BL, respectively; no clear dose relation was identified. These observations were not associated with clinical events.

A, Dose-dependent leukocytosis is displayed as the increase in white blood cell counts (response) during AMD3100 infusion relative to baseline, expressed as the mean of the ratios for each individual (solid circles) versus dose. The line represents the predicted dose-response relation according to the maximum effect (Emax) model fit to the data. B, AMD3100 concentration (log scale) is shown over time for each dose cohort. The AMD3100 infusion began at 0 hours and ended at 240 hours.


Of the 40 study subjects, 37 contributed to estimates of Css and 29 had sufficient time on study to contribute to estimates of Cl, t1/2, and AUC0–∞ (Table 2; see Fig. 1B). Css ranged from a mean of 33 ng/mL at the lowest dose (2.5 μg/kg/h) up to a mean of 3282 ng/mL at the highest dose (μg/kg/h); dose proportionality was demonstrated (r2 = 0.95, P < 0.05). The AUC0–∞ also demonstrated dose proportionality over the full range of doses studied (r2 = 0.95, P < 0.05). Clearance was similar across all dose cohorts, with the means ranging from 0.046 to 0.088 L/h/kg, demonstrating dose independence (r2 = 0.1481). The estimates for half-life were also similar among all dose cohorts, with a mean of 8.6 hours (range: 8.1–11.1 hours) among dose cohorts. Half-life was also dose independent (r2 = 0.0384).

Table 2
Table 2:
Mean Pharmacokinetic Parameter Estimates, Excluding Subjects Who Did Not Receive Complete 10-Day Intravenous Infusion

Antiviral Effect

No patient on the study demonstrated a 1.0-log10 or greater reduction in HIV RNA by day 11 (Table 3). The total RNA change for all patients for whom RNA measurements were available was a median of +0.03 log10 copies/mL (range: 0.38 to −0.87 log10 copies/mL, n = 36). One patient exhibited an HIV RNA reduction of 0.9 log10 HIV RNA by day 11, a significant (>0.5 log10) change from BL, although below our predefined level of antiviral response (1.0 log10). This patient also received AMD3100 at the highest dose (160 μg/kg/h). The HIV RNA levels at screening, day 6, and day 11 were 3.8, 3.2, and 3.0 log10 copies/mL, respectively. This patient’s HIV RNA level continued to decline after cessation of drug treatment on day 11 down to a nadir of 2.5 log10 copies/mL on day 18 and returned to the pretreatment level by day 39 (3.9 log10 copies/mL). The steady-state plasma concentration of AMD3100 in this patient was 2.9 μg/mL. He was not receiving additional ARV medications.

Table 3
Table 3:
Baseline Viral Phenotype and Change in HIV-1 RNA from Day 1 to Day 11 for the Efficacy-Evaluable Cohort


This study describes the clinical safety, pharmacokinetics, and antiviral effect of AMD3100, a CXCR4 receptor antagonist that blocks the entry of X4-tropic HIV-1. HIV-infected patients were enrolled into this dose escalation study and received open-label AMD3100 by intravenous infusion for 10 days. The higher dose cohorts were enriched with patients who were infected with SI phenotype viruses that we expected to be responsive to the X4-specific antiviral effects of AMD3100 previously seen in vitro and in the SCID-Hu Thy/Liv mouse model.23 The study was terminated after enrollment of the third patient in the 160-μg/kg/h cohort, primarily because of the perceived lack of antiviral effect (no subject’s total HIV RNA dropped by 1.0 log10, our predefined antiviral efficacy measure) and the occurrence of unexplained PVCs in 2 patients.

The only subject with a measurable reduction in HIV RNA (0.9 log10) also received the highest AMD3100 dose in the study (160 μg/kg/h). It is also important to note that this subject had a prolonged reduction in virus 1 week beyond cessation of AMD3100 in the absence of other ARV drugs. We believe this prolonged antiviral effect to be real, supported by the absence of other ARV drugs and the parallel observation within the whole study cohort of a prolonged elevation of WBCs and subsets, also believed to be an effect of CXCR4 inhibition.27,28

The absence of an antiviral effect in the other SI subjects at high doses could be explained by the presence of a mixed- or dual-tropic (X4/R5) viral population not detectable by the evaluable cohort–defining MT-2 assay in contrast to a pure X4 population in the 1 subject showing the 0.9-log10 reduction. In fact, half of our evaluable SI patients had X4 and R5 virus detected in the combined PBMC and coreceptor assay. The absence of change in HIV RNA level as a result of the presence of an unchanging R5 virus population unresponsive to CXCR4 inhibition, even with a substantial change in the X4 virus population (as a result of CXCR4 inhibition), is mathematically consistent given the large proportion and magnitude of X4 virus change required to observe a total HIV RNA reduction using the Roche Amplicor Assay. In a hypothetic dual-tropic patient receiving a CXCR4 antagonist, to observe what would be considered a significant 0.5-log10 copies/mL reduction in HIV RNA, the X4 virus would have to fall far more than 0.5 log10 and account for the major proportion of the virus present.29 For example, a 0.5-log10 drop in total RNA based solely on X4 reduction requires a 1.0-log10 reduction in X4, which must initially account for 76% of the total virus population. If the reduction in X4 virus is 2.0 log10, the X4 population must account for 70% of the total virus population to see the same 0.5-log10 reduction. Further increasing the magnitude of a pure X4 reduction still requires X4 virus in excess of 68% to see a total RNA reduction of 0.5 log10.

After our study was completed, our patient samples were evaluated by an assay that is a modification of a method previously reported to detect the presence of X4 and R5 virus independently more sensitively and quantitatively (PhenoSense, Virologics).30 According to this investigational assay and in contrast to the PBMC assay, our single patient with the antiviral response was also the only patient with pure X4 virus in the study. All other patients of our efficacy-evaluable cohort were determined to have dual- or mixed-tropic virus; 8 of 15 patients with dual- or mixed-tropic virus (including 3 not in the efficacy-evaluable cohort, because they were not SI at BL) had purely R5 phenotypes by day 11, suggesting significant reductions in the X4 virus population present.31 Despite this, none of these patients had a 1.0-log10 or greater reduction in the total HIV RNA. Although the application of this assay is investigational, its preliminary results are consistent with our suggestion that dual- or mixed-tropic virus populations, not identified by phenotypic assays we used, made it highly unlikely that a modest anti-X4 effect could be detected by the HIV RNA assay. Quantitative X4 and R5 specific assays will be essential in future antiviral studies of chemokine inhibitors.

Our findings of phenotype-specific HIV RNA reduction and prolonged AMD3100 effect also mirror data from 2 clinical studies of CCR5 antagonists, SCH-C (Schering Plough) and UK-427 (Pfizer).32,33 Similar to our study, each of these studies observed (1) HIV RNA reductions in patients with pure R5 variants as expected with a CCR5 inhibitor (pure X4 responsive to a CXCR4 inhibitor in our study), (2) failure to reduce HIV RNA in patients with dual-tropic virus despite an NSI phenotype, and (3) a prolonged antiviral suppression. This prolonged antiviral response has also been seen with CCR5 inhibitors against HIV in vitro (SCH-C) and against simian immunodeficiency virus (SIV) in a macaque model (CMPD 167).33,34 Perhaps these chemokine-binding inhibitors have a long residence time (slow off-rate) from the receptor, allowing the antagonist to inhibit the infection of new target cells by HIV or SIV even after the drug has apparently been cleared from the plasma. Alternatively, prolonged blockade of CXCR4 or CCR5 could lead to upregulation and increased production (to maintain homeostasis) of the corresponding natural chemokines, for example, stromal derived factor (SDF) or regulated-on-activation normal T cell expressed and secreted (RANTES), respectively. In addition to inhibiting viral replication, AMD3100 inhibits SDF-1α binding and signaling via CXCR4, whereas SCH-C inhibits binding and signaling of RANTES and macrophage inflammatory protein (MIP)-1α to CCR5.7–9,20,35–37 These chemokines also inhibit viral replication in a chemokine coreceptor-specific way via a direct competitive antagonism of virally encoded gp120 binding to the receptor or by a secondary mechanism in which the chemokine internalizes the receptor.7,20,35–41 In either case, increased production of chemokines could have an additive or synergistic antiviral effect during treatment and a longer term antiviral effect after drug treatment. Recent studies in HIV-1–infected subjects suggest that high plasma SDF-1α levels and low CXCR4 expression on T lymphocytes are associated with long-term nonprogression of disease.42,43

The clinical impact of eliminating X4 virus is not known. Epidemiologic studies associate SI phenotype and X4 virus with rapid rates of CD4+ T-cell decline and progressive late-stage disease, whereas NSI phenotype and R5 virus predominate in early-stage disease.44–48 Recently, X4 virus has been identified as an independent risk factor for CD4 decline and clinical disease progression.49,50 Despite having the seemingly more “benign” phenotype, patients whose virus seems to use CCR5 (NSI phenotype) exclusively throughout their disease course still progress to AIDS. Whether there is clinical value in having the ability to turn back the evolutionary clock from the more virulent X4 population to an R5 population, more commonly associated with slowly progressive or nonprogressive disease, remains to be tested. Use of quantitative phenotype-specific evaluations in longitudinal clinical studies may be helpful to address these questions.

AMD3100 was not the first CXCR4 antagonist to be tested in clinical trials. The small-peptide CXCR4 inhibitor ALX40-4c, administered by short intravenous infusion, was evaluated in a phase 1/2 trial in 40 HIV-positive patients.51 ALX40-4c was well tolerated, and in a similar manner to our results, the compound did not cause significant or consistent reductions in HIV RNA by the end of the 1-month treatment period (including 12 patients confirmed to harbor SI virus by the MT-2 assay).

Safety and pharmacokinetic issues impeded further development of AMD3100 for HIV infection. The most common and only dose-related adverse effects were gastrointestinal symptoms (mostly a sense of abdominal fullness and some loose stools) and paresthesias (perioral and in the extremities). These might be troublesome in an otherwise well outpatient population. Further, the episode of thrombocytopenia (which was possibly study drug related), the orthostatic changes, and at least 1 of the 2 instances of ventricular ectopy remain unexplained. In terms of pharmacokinetics, AMD3100 demonstrated well-behaved first-order kinetics across the range of doses tested and has a favorable 9-hour half-life, likely amenable to once- or twice-daily dosing. The parenteral route limitations are a significant obstacle, however, and oral analogues are clearly needed.

In summary, the clinical trial of AMD3100 suggests that CXCR4 is a legitimate target for the development of anti-HIV agents. One patient, the only patient with a pure X4-tropic virus population (who also received the highest dose of AMD3100 administered [160 μg/kg/h]), responded with a 0.9-log10 reduction in HIV RNA by day 11, which further decreased to 1.3 log10 copies/mL by day 18. It may not have even been possible to detect X4-specific changes in the other patients if dual- or mixed-tropic phenotypes were common. Preliminary results from a new coreceptor tropism assay (PhenoSense Assay) confirm the results that the single antiviral response patient was the only patient with a pure X4 virus population and suggest that X4 virus was reduced below the level of detection by day 11 in more than half of the other patients identified to have X4 virus present in a mixed-tropic virus population. Gastrointestinal symptoms and paresthesias became common at higher doses of AMD3100. Despite a favorable 9-hour half-life, the drug cannot be given orally. Development of AMD3100 continues for stem cell mobilization,27,28 whereas an orally bioavailable CXCR4 antagonist (AMD070) is in trials for HIV infection.


The authors thank the study participants, the technical assistance of Sandra Claes and Eric Fonteyn, and Joan Dragavon and Pax Ortega (University of Washington Retrovirology Laboratory). They also recognize the significant contributions to the study by Ella Redpath, N. Jeanne Conley, Joanne Stickler, Valery Hughes, the General Clinical Research Center staff at the participating institutions, and the Harborview Medical Center telemetry and 3 East nursing staff.


1. Yeni PG, Hammer SM, Carpenter CC, et al. Antiretroviral treatment for adult HIV infection in 2002: updated recommendations of the International AIDS Society-USA Panel. JAMA. 2002;288:222–235.
2. Berger EA, Murphy PM, Farber JM. Chemokine receptors as HIV-1 coreceptors: roles in viral entry, tropism, and disease. Annu Rev Immunol. 1999;17:657–700.
3. Kilby JM, Hopkins S, Venetta TM, et al. Potent suppression of HIV-1 replication in humans by T-20, a peptide inhibitor of gp41-mediated virus entry. Nat Med. 1998;4:1302–1307.
4. Lalezari JP, DeJesus E, Northfelt DW, et al. A controlled Phase II trial assessing three doses of enfuvirtide (T-20) in combination with abacavir, amprenavir, ritonavir and efavirenz in non-nucleoside reverse transcriptase inhibitor-naive HIV-infected adults. Antivir Ther. 2003;8:279–287.
5. Lalezari JP, Henry K, O’Hearn M, et al. Enfuvirtide, an HIV-1 fusion inhibitor, for drug-resistant HIV infection in North and South America. N Engl J Med. 2003;348:2175–2185.
6. Bridger GS, Skerlj RT. Bicyclam derivatives as HIV inhibitors. In: De Clercq E, ed. Advances in Antiviral Drug Design, vol. 3. New York: Elsevier; 1999:161–229.
7. Hatse S, Princen K, Bridger G, et al. Chemokine receptor inhibition by AMD3100 is strictly confined to CXCR4. FEBS Lett. 2002;527:255–262.
8. Gerlach LO, Skerlj RT, Bridger GJ, et al. Molecular interactions of cyclam and bicyclam non-peptide antagonists with the CXCR4 chemokine receptor. J Biol Chem. 2001;276:14153–14160.
9. Schols D, Struyf S, Van Damme J, et al. Inhibition of T-tropic HIV strains by selective antagonization of the chemokine receptor CXCR4. J Exp Med. 1997;186:1383–1388.
10. Dragic T, Litwin V, Allaway GP, et al. HIV-1 entry into CD4+ cells is mediated by the chemokine receptor CC-CKR-5. Nature. 1996;381:667–673.
11. Deng H, Liu R, Ellmeier W, et al. Identification of a major co-receptor for primary isolates of HIV-1. Nature. 1996;381:661–666.
12. Alkhatib G, Combadiere C, Broder CC, et al. CC CKR5: a RANTES, MIP-1alpha, MIP-1beta receptor as a fusion cofactor for macrophage-tropic HIV-1. Science. 1996;272:1955–1958.
13. Connor RI, Ho DD. Human immunodeficiency virus type 1 variants with increased replicative capacity develop during the asymptomatic stage before disease progression. J Virol. 1994;68:4400–4408.
14. Tersmette M, de Goede RE, Al BJ, et al. Differential syncytium-inducing capacity of human immunodeficiency virus isolates: frequent detection of syncytium-inducing isolates in patients with acquired immunodeficiency syndrome (AIDS) and AIDS-related complex. J Virol. 1988;62:2026–2032.
15. Schuitemaker H, Koot M, Kootstra NA, et al. Biological phenotype of human immunodeficiency virus type 1 clones at different stages of infection: progression of disease is associated with a shift from monocytotropic to T-cell-tropic virus population. J Virol. 1992;66:1354–1360.
16. Simmons G, Wilkinson D, Reeves JD, et al. Primary, syncytium-inducing human immunodeficiency virus type 1 isolates are dual-tropic and most can use either Lestr or CCR5 as coreceptors for virus entry. J Virol. 1996;70:8355–8360.
17. Zaitseva M, Blauvelt A, Lee S, et al. Expression and function of CCR5 and CXCR4 on human Langerhans cells and macrophages: implications for HIV primary infection. Nat Med. 1997;3:1369–1375.
18. Michael NL, Nelson JA, Kewal Ramani VN, et al. Exclusive and persistent use of the entry coreceptor CXCR4 by human immunodeficiency virus type 1 from a subject homozygous for CCR5 delta32. J Virol. 1998;72:6040–6047.
19. Fenyo EM, Morfeldt-Manson L, Chiodi F, et al. Distinct replicative and cytopathic characteristics of human immunodeficiency virus isolates. J Virol. 1988;62:4414–4419.
20. Donzella GA, Schols D, Lin SW, et al. AMD3100, a small molecule inhibitor of HIV-1 entry via the CXCR4 co-receptor. Nat Med. 1998;4:72–77.
21. De Clercq E, Yamamoto N, Pauwels R, et al. Highly potent and selective inhibition of human immunodeficiency virus by the bicyclam derivative JM3100. Antimicrob Agents Chemother. 1994;38:668–674.
22. Harouse JM, Buckner C, Gettie A, et al. CD8+ T cell-mediated CXC chemokine receptor 4-simian/human immunodeficiency virus suppression in dually infected rhesus macaques. Proc Natl Acad Sci USA. 2003;100:10977–10982.
23. Datema R, Rabin L, Hincenbergs M, et al. Antiviral efficacy in vivo of the anti-human immunodeficiency virus bicyclam SDZ SID 791 (JM 3100), an inhibitor of infectious cell entry. Antimicrob Agents Chemother. 1996;40:750–754.
24. Hendrix CW, Flexner C, MacFarland RT, et al. Pharmacokinetics and safety of AMD-3100, a novel antagonist of the CXCR-4 chemokine receptor, in human volunteers. Antimicrob Agents Chemother. 2000;44:1667–1673.
25. Division of AIDS, National Institute of Allergy and Infectious Diseases, National Institutes of Health and Collaborating Investigators. DAIDS Virology Manual for HIV Laboratories [Internet]. Available at:
26. Bjorndal A, Deng H, Jansson M, et al. Coreceptor usage of primary human immunodeficiency virus type 1 isolates varies according to biological phenotype. J Virol. 1997;71:7478–7487.
27. Hübel K, Liles WC, Broxmeyer HE, et al. Leukocytosis and mobilization of CD34+ hematopoietic progenitor cells by AMD3100, a CXCR4 antagonist. Support Cancer Ther.(In press)
28. Liles WC, Broxmeyer HE, Rodger E, et al. Mobilization of hematopoietic progenitor cells in healthy volunteers by AMD3100, a CXCR4 antagonist. Blood. 2003;102:2728–2730.
29. Vandamme AM, Schmit JC, Van Dooren S, et al. Quantification of HIV-1 RNA in plasma: comparable results with the NASBA HIV-1 RNA QT and the AMPLICOR HIV monitor test. J Acquir Immune Defic Syndr Hum Retrovirol. 1996;13:127–139.
30. Petropoulos CJ, Parkin NT, Limoli KL, et al. A novel phenotypic drug susceptibility assay for human immunodeficiency virus type 1. Antimicrob Agents Chemother. 2000;44:920–928.
31. Schols D. AMD3100, a CXCR4 antagonist, reduced HIV viral load and X4 virus levels in humans [abstract 2]. Presented at the 9th National Conference on Retroviruses and Opportunistic Infections, Seattle, 2002.
32. Fatkentheuer G. Effect of short-term monotherapy with UK-427,857 on viral load in HIV-infected patients [abstract H-443]. Presented at the 43rd ICAAC Meeting, Chicago, 2003.
33. Reynes J. SCH-C. Safety and antiviral effects of a CCR5 receptor antagonist in HIV-infected subjects [abstract 1]. Presented at the 9th Conference on Retroviruses and Opportunistic Infections, Seattle, January 2002.
34. Veazey RS, Klasse PJ, Ketas TJ, et al. Use of a small molecule CCR5 inhibitor in macaques to treat simian immunodeficiency virus infection or prevent simian-human immunodeficiency virus infection. J Exp Med. 2003;198:1551–1562.
35. Hatse S, Princen K, Gerlach LO, et al. Mutation of Asp(171) and Asp(262) of the chemokine receptor CXCR4 impairs its coreceptor function for human immunodeficiency virus-1 entry and abrogates the antagonistic activity of AMD3100. Mol Pharmacol. 2001;60:164–173.
36. Palani A, Shapiro S, Clader JW, et al. Discovery of 4-[(Z)-(4-bromophenyl)-(ethoxyimino)methyl]-1′-[(2,4-dimethyl-3- pyridinyl)carbonyl]-4′-methyl-1,4′-bipiperidine N-oxide (SCH 351125): an orally bioavailable human CCR5 antagonist for the treatment of HIV infection. J Med Chem. 2001;44:3339–3342.
37. Strizki JM, Xu S, Wagner NE, et al. SCH-C (SCH 351125), an orally bioavailable, small molecule antagonist of the chemokine receptor CCR5, is a potent inhibitor of HIV-1 infection in vitro and in vivo. Proc Natl Acad Sci USA. 2001;98:12718–12723.
38. Tsamis F, Gavrilov S, Kajumo F, et al. Analysis of the mechanism by which the small-molecule CCR5 antagonists SCH-351125 and SCH-350581 inhibit human immunodeficiency virus type 1 entry. J Virol. 2003;77:5201–5208.
39. Labrosse B, Brelot A, Heveker N, et al. Determinants for sensitivity of human immunodeficiency virus coreceptor CXCR4 to the bicyclam AMD3100. J Virol. 1998;72:6381–6388.
40. Hatse S, Princen K, Vermeire K, et al. Mutations at the CXCR4 interaction sites for AMD3100 influence anti-CXCR4 antibody binding and HIV-1 entry. FEBS Lett. 2003;546:300–306.
41. Brandt SM, Mariani R, Holland AU, et al. Association of chemokine-mediated block to HIV entry with coreceptor internalization. J Biol Chem. 2002;277:17291–17299.
42. Soriano A, Martinez C, Garcia F, et al. Plasma stromal cell-derived factor (SDF)-1 levels, SDF1-3′A genotype, and expression of CXCR4 on T lymphocytes: their impact on resistance to human immunodeficiency virus type 1 infection and its progression. J Infect Dis. 2002;186:922–931.
43. Derdeyn CA, Costello C, Kilby JM, et al. Correlation between circulating stromal cell-derived factor 1 levels and CD4+ cell count in human immunodeficiency virus type 1-infected individuals. AIDS Res Hum Retroviruses. 1999;15:1063–1071.
44. Connor RI, Sheridan KE, Ceradini D, et al. Change in coreceptor use correlates with disease progression in HIV-1–infected individuals. J Exp Med. 1997;185:621–628.
45. Scarlatti G, Tresoldi E, Bjorndal A, et al. In vivo evolution of HIV-1 co-receptor usage and sensitivity to chemokine-mediated suppression. Nat Med. 1997;3:1259–1265.
46. Schramm B, Penn ML, Speck RF, et al. Viral entry through CXCR4 is a pathogenic factor and therapeutic target in human immunodeficiency virus type 1 disease. J Virol. 2000;74:184–192.
47. Malkevitch N, McDermott DH, Yi Y, et al. Coreceptor choice and T cell depletion by R5, X4, and R5X4 HIV-1 variants in CCR5-deficient (CCR5delta32) and normal human lymphoid tissue. Virology. 2001;281:239–247.
48. Tersmette M, Gruters RA, de Wolf F, et al. Evidence for a role of virulent human immunodeficiency virus (HIV) variants in the pathogenesis of acquired immunodeficiency syndrome: studies on sequential HIV isolates. J Virol. 1989;63:2118–2125.
49. Harrigan PR, Dong WY, Yip B, et al. Baseline HIV-1 V3 loop sequences as a predictor of immunologic decline and mortality in a large cohort starting initial HAART therapy. Presented at the International AIDS Society 2nd Conference on Pathogenesis and Treatment, Paris, July 2003.
50. Daar E, Kesler K, Lail A, et al. HIV co-receptor tropism and replication capacity predict HIV progression [abstract H-1722c]. Presented at the 43rd Interscience Conference on Antimicrobial Agents and Chemotherapy Meeting, Chicago, September 2003.
51. Doranz BJ, Filion LG, Diaz-Mitoma F, et al. Safe use of the CXCR4 inhibitor ALX40-4C in humans. AIDS Res Hum Retroviruses. 2001;17:475–486.

chemokine binding inhibitor; CXCR4 antagonist; antiretroviral HIV

© 2004 Lippincott Williams & Wilkins, Inc.