Location: Research was performed at the Clinic of Small Animal Medicine, LMU University of Munich, Munich, Germany.
Efficacy and Adverse Effects of the Antiviral Compound Plerixafor in Feline Immunodeficiency Virus-Infected Cats
Version of Record online: 30 MAR 2012
Copyright © 2012 by the American College of Veterinary Internal Medicine
Journal of Veterinary Internal Medicine
Volume 26, Issue 3, pages 483–490, May-June 2012
How to Cite
Hartmann, K., Stengel, C., Klein, D., Egberink, H. and Balzarini, J. (2012), Efficacy and Adverse Effects of the Antiviral Compound Plerixafor in Feline Immunodeficiency Virus-Infected Cats. Journal of Veterinary Internal Medicine, 26: 483–490. doi: 10.1111/j.1939-1676.2012.00904.x
- Issue online: 2 MAY 2012
- Version of Record online: 30 MAR 2012
- Manuscript Accepted: 27 JAN 2012
- Manuscript Revised: 4 JAN 2012
- Manuscript Received: 28 JUL 2011
Bicyclam derivatives inhibit feline immunodeficiency virus (FIV) replication through selective blockage of chemokine receptor CXCR4.
CXCR4 antagonist plerixafor (AMD3100, 1,1′-bis-1,4,8,11-tetraazacyclotetradekan) alone or combination with adefovir (PMEA, 9-(2-phosphonylmethoxyethyl)adenine) safe and effective for treating FIV-infected cats.
Forty naturally FIV-infected, privately owned cats.
Materials and Methods
Prospective, placebo-controlled, double-blind clinical trial. Cats randomly classified into 4 treatment groups. Received AMD3100, PMEA, AMD3100 in combination with PMEA, or placebo for 6 weeks. Clinical and laboratory parameters, including CD4+ and CD8+ cell counts, FIV proviral and viral load measured by quantitative polymerase chain reaction (qPCR) evaluated. Additionally, FIV isolates from cats treated with AMD3100 tested for drug resistance.
FIV-infected cats treated with AMD3100 caused significant decrease in proviral load compared to placebo group (2.3 ± 3.8% to 1.9 ± 3.1%, of blood lymphocytes P < .05), but did not lead to improvement of clinical or immunological variables; it caused a decrease in serum magnesium concentration without clinical signs. No development of resistance of FIV isolates to AMD3100 found during treatment period. PMEA administration improved stomatitis (stomatitis score [degree 1 – 100] PMEA group: 23 ± 19 to 11 ± 10, P < .001; AMD3100 + PMEA group: 12 ± 17 to 3 ± 5, P < .05), but did not decrease proviral or viral load and caused anemia (RBC [×106/μL] PMEA group: 9.07 ± 1.60 to 6.22 ± 2.16, P < .05; AMD3100 ± PMEA group: 8.80 ± 1.23 to 5.84 ± 1.58, P < .001).
Conclusions and Clinical Importance
Administration of CXCR4 antagonists, as AMD3100, can induce reduction of proviral load and may represent viable treatment of FIV-infected cats. Combination treatment with PMEA not recommended.
human immunodeficiency virus
feline immunodeficiency virus
plerixafor, 1,1′-[1,4-phenylenbismethylene]-bis(1,4,8,11-tetraazacyclotetradecane)-octachloride dihydrate
quantitative polymerase chain reaction
Crandell feline kidney cells
ribosomal deoxyribonucleic acid
highly active antiretroviral therapy
feline leukemia virus
50% effective concentration
50% tissue culture infective dose
red blood cell count
natural killer cells
Feline immunodeficiency virus (FIV) was first detected in 1986, and is recognized as a common infectious agent of domestic cats worldwide. Mode of infection, cell types infected, dispersal of virus in the body, and time course of infection in cats closely resemble those features in human immunodeficiency virus (HIV) infection in people. Both HIV and FIV use a chemokine receptor for infecting cells.[3, 4] Chemokine receptors belong to the group of transmembrane proteins, in which signal transmission is afforded through rapid influx of calcium into the cell. They are also essential co-receptors for HIV and FIV during infection of primary susceptible CD4+ lymphocytes. In late-stage HIV infection, viral isolates mainly use the CXCR4 receptor for cell entry. In FIV infection, CXCR4 also is one of the major co-receptors.[4, 7] By blocking the chemokine receptors, infection of cells by HIV or FIV can be prevented.[4, 8, 9]
Bicyclams are low molecular weight nonpeptidic compounds that bind selectively to the chemokine receptor CXCR4, thereby preventing interaction of this receptor with other ligands, such as HIV or FIV, and inhibiting entry of these viruses into the cell.[10-13] Plerixafor (1,1′-[1,4-phenylenbismethylene]-bis(1,4,8,11-tetraazacyclotetradecane)-octachloride dehydrate, AMD3100, JM3100), is the prototype compound among the bicyclams. Plerixafor is commercially available1 and used for stem cell mobilization in humans. In vitro studies showed that AMD3100 efficiently inhibits FIV replication.[12, 15]
Highly active antiretroviral treatment (HAART) protocols in which compounds of several classes are combined are currently the mainstay of treatment of HIV-infected people. This has led to a dramatic improvement of prognosis and survival in infected humans. Although several anti-HIV compounds have been shown to inhibit FIV replication in cell culture, antiviral chemotherapy is not widely used in FIV-infected cats. Zidovudin (AZT, azidothymidine, 3′-azido-2′, 3′-dideoxythymidine) is the only drug currently applied to some naturally FIV-infected cats, but is not very effective and can be associated with severe adverse effects.[17-19] Multidrug protocols have only been used in case reports, and in vivo efficacy studies in naturally FIV-infected cats are missing for most HIV compounds, including chemokine receptor blockers. All drugs so far investigated in controlled field studies in cats are nucleoside analogs, and combination with other drug classes has not been investigated.
Therefore, the aim of the present study was to evaluate efficacy and adverse effects of the bicyclam plerixafor and the nucleoside analog adefovir (9-(2-phosphonylmethoxyethyl)adenine (PMEA),2 alone or in combination in FIV-infected cats in a placebo-controlled, double-blind study specifically looking at (1) quality of life measures and clinical signs, (2) changes in CD4+ and CD8+ counts, (3) decrease in proviral and viral load, (4) adverse effects, and (5) development of drug resistance.
Materials and Methods
The study was designed as a 6-week placebo-controlled, double-blinded, clinical trial. It fulfilled the general German guidelines for prospective studies with owners' consent. Forty cats were included and randomly assigned to 1 of 4 groups of 10 cats each. Ten cats received placebo only, 10 cats received AMD3100 and placebo, 10 cats received PMEA and placebo, and 10 cats received AMD3100 and PMEA.
Cats were only included in the study if they were FIV-infected and had clinically evident stomatitis. All cats had serum antibodies against FIV p24 antigen detected by ELISA3 and detectable provirus in blood measured by quantitative polymerase chain reaction (qPCR). Only cats for which the owners gave their consent for participation were included. Cats were excluded if they were coinfected with feline leukemia virus (FeLV), if they were in a moribund condition (Karnofsy's score < 30%), or if they were aggressive.
Thirty-six of the 40 cats (90%) were domestic short-hair cats and 4 (10%) were long-hair mixed-breed cats. Age of cats ranged between 1 and 11 years (median 6.4 years). All animals were neutered; 28/40 (70%) were male, 12/40 (30%) were female.
AMD31001 and its respective placebo were injected SC at 0.5 mg/kg q12h; PMEA2 and its respective placebo were injected SC at 10 mg/kg twice weekly. Sterile phosphate-buffered saline (PBS) was used as placebo and as solvent for PMEA and AMD3100. Preparation and encoding of all compounds was performed at the Rega Institute for Medical Research (JB). All other investigators were masked until completion of statistical analysis of all data.
To determine efficacy and detect adverse effects, proviral and viral load, CD4+ and CD8+ counts, clinical signs, and laboratory parameters were monitored throughout the study. A physical examination was performed in all cats before treatment and then weekly during the trial. A numerical scoring system (0 = no clinical sign; 100 = most severe signs) was designed to semiquantify severity of stomatitis and conjunctivitis. To enable investigators to assess the overall health of the animals objectively, the Karnofsky's score adapted to cats was used. To detect adverse effects, a complete blood count (CBC)4 and serum biochemistry analysis5 were performed every second week. Proviral and viral load and CD4+ and CD8+ cell counts were measured every 2nd week as follows. Lymphocytes were stained for cell surface expression of CD3, CD4, and CD8 as described, and counted using a fluorescence-activated cell sorter.6 In general, 30,000 events were acquired and analyzed using 2 different software programs.7 CD4+ and CD8+ cell counts and the CD4 : CD8 ratio were calculated.
Proviral and viral load was determined by Taqman qPCR. Deoxyribonucleic acid (DNA) or ribonucleic acid (RNA) were extracted from 200 μL of whole blood or 140 μL of plasma, respectively, using 2 extraction kits.89 The FIV proviral load in peripheral blood mononuclear cells (PBMC) was quantified using qPCR measuring PCR product accumulation through a dual-labeled fluorogenic TaqMan probe. Because a broad range of FIV subtypes likely occur in the area where the study was performed, 3 different assays targeting a broad range of FIV subtypes (1010p, 1372p, and 1416p) that have been described previously,[24-26] were used in the present study. The 25-μL PCR mixture contained 10 mM Tris (pH 8.3), 50 mM KCl, 3 mM MgCl2, 200 nM deoxyadenosine triphosphate (dATP), deoxycytidine triphosphate (dCTP), deoxyguanosine triphosphate (dGTP), 400 nM deoxuridine triphosphate (dUTP), 300 nM of each primer, 200 nM of the fluorogenic probe, and 2.5 units of Taq DNA polymerase. After initial denaturation (2 minutes at 95°C), amplification was performed with 45 cycles of 15 seconds at 95°C and 60 seconds at 60°C. The PCR reaction and the on-line measurement of the emitted fluorescence were performed on a sequence detector system.10 The copy number per PCR reaction was calculated using sequence detection software11 utilizing a series of 4-fold dilutions of genomic DNA derived from a Crandell feline kidney (CrFK) cell line infected with FIV (Petaluma). The absolute DNA content per PCR reaction was estimated by a second real-time PCR assay targeting the 18S ribosomal deoxyribonucleic acid (rDNA) genes. The relative proviral load was calculated by setting the value at beginning of therapy as 100% and then relating all subsequent measurements to this.
The FIV viral load in plasma was quantified with reverse transcriptase (RT) qPCR using the same primers and probes as above. The 25-μL one-step RT qPCR mixture contained 12.5 μL 2× reaction buffer,12 300 nM of each primer, 200 nM fluorogenic probe, 0.5 μL SuperScript III RT/Platinium Taq Mix, and 5 μL of sample. After a reverse transcription step (30 minutes at 42°C) followed by a denaturation step (10 minutes at 95°C), amplification was performed with 45 cycles of 15 seconds at 95°C and 60 seconds at 60°C. Reverse transcription and amplification were performed in a sequence detection system.13 The copy number per RT qPCR reaction was calculated by sequence detection software 11 utilizing a series of 10-fold dilutions of in vitro transcribed RNA as described.
Development of Resistance
A blood sample for virus isolation was analyzed in all cats receiving AMD3100 and all cats receiving AMD3100 and PMEA on days 0 and 42. Virus isolates were titrated and used in an antiviral assay. Thus, 50% effective concentration (EC50) values were determined by adding different concentrations of AMD3100 to thymocytes cultured in 96-well plates with 100 50% tissue culture infective doses (TCID50) of FIV per well. Virus production was measured by p24 ELISA, and the highest concentrations of AMD3100 inducing 50% p24 levels were estimated. The EC50 values on days 0 and 42 were compared.
Clinical stomatitis and conjunctivitis scores, modified Karnofsky's index values, all laboratory parameters, lymphocyte subset numbers, and proviral and viral load were compared among the 4 treatment groups by use of one-way ANOVA conducted using SPSS.14 Whenever a statistically significant difference was detected among groups, the change over time was determined by calculating the difference between the mean values at the end and beginning of therapy. The Kruskall-Wallis Test was then performed to analyze differences among separate groups. To ensure that the 4 groups were comparable concerning the variables investigated before treatment initiation, groups were compared on day 0 using one-way ANOVA. A P value < .05 was considered significant for all analyses.
Efficacy of the Compounds
There was no significant difference in any evaluated variable among the 4 groups before treatment initiation on day 0 (Table 1). At the end of the study, all FIV-infected cats were still alive. During the treatment period, the clinical status of all cats improved in all groups. Among clinical variables, no significant difference was seen in the body weight, the Karnofsky's score, or the conjunctivitis score among treatment groups. A statistically significant difference was detected only for the stomatitis score. Improvement in stomatitis score for cats receiving PMEA only (P < .001) or PMEA with AMD3100 (P < .05) was significantly greater than for cats receiving placebo (Fig 1). By contrast, no significant difference in improvement of stomatitis was detected between cats receiving AMD3100 and cats receiving placebo.
|Variables||Placebo||AMD3100||PMEA||AMD3100 + PMEA|
|Body weight (kg)||Mean||5.3||5.0||5.0||4.7|
|Karnofsky's score (%)||Mean||83.1||93.4||81.4||85.9|
|Stomatitis score (degree 1–100)||Mean||9.0||4.2||23.2||12.0|
|Conjunctivitis score (degree 1–100)||Mean||3.4||4.0||5.8||5.3|
|CD4+ cell count (/μL)||Mean||417||532||545||1,002|
|CD8+ cell count (/μL)||Mean||379||503||530||357|
|(% infected PBL)||SD||18.17||3.81||1.76||6.86|
|Virus load (copy RNA/mL plasma)||Mean||348,773||5,007||1,269||88,518|
No significant changes in absolute CD4+ or CD8+ counts or the CD4 : CD8 ratio were observed among groups. There were also no significant differences in viral load among groups. However, significant differences in proviral load were detected among groups; compared to the placebo group, the relative proviral load decreased significantly in the AMD3100 group (P < .05), but a relative increase was observed in cats receiving PMEA and AMD3100 (P < .05) compared to cats receiving only AMD3100 (Fig 2).
A significant decrease in red blood cell (RBC) counts (Fig 3), hemoglobin concentration, and hematocrit was found in cats receiving PMEA in comparison to cats receiving placebo (P < .05) and compared to cats receiving AMD3100 (P < .05), as well as in cats receiving both compounds (PMEA + AMD3100) in comparison to the placebo group (P < .001) and in comparison to cats receiving AMD3100 (P < .001). No significant differences were found in the white blood cell counts. Among all biochemical parameters evaluated, only serum magnesium decreased significantly in cats receiving AMD3100 (P < .05) and in cats receiving PMEA and AMD3100 (P < .05) compared to placebo-treated cats as well as in cats receiving PMEA and AMD3100 compared to cats receiving PMEA only (P < .05) (Fig 4). Changes in concentrations of calcium and other electrolytes were not significantly different among groups.
Development of Resistance
In 15 cats receiving AMD3100 (7 cats) or AMD3100 and PMEA (8 cats), virus isolation was possible on day 0 and day 42. The 50% effective concentration (EC50) values against AMD3100 were determined for 15 FIV isolates. For the FIV isolates that could be recovered from the AMD3100-treated cats, EC50 values before (day 0) and after (day 42) treatment were 52 ± 46 nM and 68 ± 89 nM, respectively. The EC50 values for the FIV isolates before and after AMD3100 + PMEA treatment were 135 ± 64 nM and 200 ± 150 nM, respectively. In both groups of virus isolates there was no statistically significant difference in the susceptibility to AMD3100 before or after drug treatment revealing that no drug resistance development occurred within the treatment period.
Both drugs used in this study showed evidence of efficacy but also mild adverse effects. No significant difference in overall health status was detected between cats receiving AMD3100 and cats receiving placebo or PMEA. However, the stomatitis score improved in cats from the PMEA-treated and AMD3100/PMEA-treated group. By contrast, stomatitis in cats treated solely with AMD3100 did not improve. This might be because AMD3100 exerts an antiretroviral effect against FIV but is not active against other viruses, such as feline herpesvirus or calicivirus considered important cofactors in the development of FIV-associated stomatitis. Instead, as demonstrated in earlier studies, an improvement or cure of stomatitis in FIV-infected cats was seen in treatment with PMEA or related drug.[17, 29] Nucleoside phosphonates such as PMEA are effective in vivo against a broad spectrum of retroviruses, including HIV and FIV, and several DNA viruses including certain herpesviruses.[29, 31] Antiproliferative effects and immunomodulatory properties (eg, through intensification of natural killer cell [NK] activity and interferon production)[33, 34] also occur during PMEA treatment. Because herpesvirus and calicivirus are common causes of stomatitis in cats, PMEA might have exerted a direct antiviral effect on these viruses. Alternatively, it is possible that PMEA inhibited excessive growth of the oral mucosa because of its antiproliferative properties. This could explain the improvement of stomatitis in cats of both groups receiving PMEA in contrast to the other groups. It should be noticed that cats treated with PMEA started with more severe average stomatitis scores (although this difference was not statistically significantly different between the groups; see Table 1) which markedly resolved by day 42 and that it cannot be fully excluded that improvement of stomatitis was not a direct effect of the drug but rather a coincidental cycling in the degree of stomatitis as clinical signs may fluctuate with or without treatment.
There was a significantly reduced relative proviral load in AMD3100-treated cats in the present study. Although statistically significant, the effect was not very pronounced; however, it demonstrates for the first time antiviral efficacy of this compound in naturally FIV-infected cats. A more pronounced effect on the provirus load could have potentially been achieved during a longer treatment period. It is also possible that the virus load even further decreased more (and other parameters continued to improve) after treatment was stopped. This, however, was unfortunately not investigated in the study because cats were not available for further rechecks. The demonstrated antiviral efficacy is supported by cell culture data revealing that AMD3100 causes a dose-dependent inhibition of FIV replication comparable to that seen against HIV.[10, 12, 36, 37] In HIV-infected humans, efficacy was shown in a Phase II clinical trial but not further pursued mainly attributable to lack of oral bioavailability and an unexpected stem cell mobilizing effect. However, in contrast to results from the present study, data from an experimental study revealed that AMD3100 failed to decrease proviral load when administered to chronically FIV-infected cats. However, in this previous study, all cats were infected with one specific FIV isolate in an experimental setting, whereas in the present study cats were naturally infected and presumably with different FIV strains, which might explain the difference in proviral load development. Indeed, it has been reported that FIV can use alternative co-receptors. This could have influenced the outcome of the experimental study if one assumes that the efficacy of using an alternative coreceptor can differ depending the nature of the FIV strain investigated. However, it is difficult to explain why the combination of AMD3100 with PMEA was not effective in reducing viral load. Hypothetically, the immunomodulatory properties of PMEA[33, 34] might have led to increased production of FIV-infected lymphocytes and thereby, annihilated the decrease in proviral load caused by AMD3100.
In the present study, the number of provirus-containing cells declined with AMD3100 treatment. Therefore, the amount of viral RNA would also be expected to decline. However, no significant change in RNA was detected, even though the absolute viral load in the AMD3100-treated cats of both groups decreased. It is possible that treatment with AMD3100 for only 6 weeks is insufficient to produce a significant effect on new viral particle formation, and longer treatment duration would be necessary. However, this study design was chosen because relatively short treatment periods with highly effective antiviral combinations have been demonstrated to lead to significant improvement in HIV-infected people.
Unfortunately the antiproliferative properties of PMEA described above likely also contributed to the progressive decline in RBC counts seen in cats treated with PMEA alone or in combination with AMD3100. This is a common adverse effect of some acyclic nucleoside phosphonates, in particular PMEA.[17, 29, 41, 42] The active intracellular metabolite of PMEA (PMEApp) not only suppresses the activity of retroviral reverse transcriptases but also of some cellular DNA polymerases such as DNA polymerase-α. Because of the life-threatening anemia observed in the PMEA-treated cats, extended treatment with PMEA at the dose used here was not possible. Prolongation of the QT interval on ECG, with potential for development of arrhythmias has been reported as an adverse effect after repeated intravenous administration of AMD3100 to humans. However, in the present study, all cats were regularly examined by auscultation of the heart and no signs of cardiac arrhythmias were found. Likewise, in an experimental FIV trial, in which a higher dosage was used, no adverse cardiac effects were reported. This suggests that cats do not seem to be susceptible to potential adverse cardiac effects of AMD3100. Other adverse effects of AMD3100 seen in humans are mild transient gastrointestinal signs and an increased WBC count and serum magnesium concentration. By contrast, in the present study, cats receiving AMD3100 experienced a significant decrease in serum magnesium concentration although they did not show any clinical signs referable to this. Intestinal absorption of magnesium utilizes similar mechanisms as calcium. It is therefore possible that magnesium partially enters the cell with calcium. Physiologic signal transduction through CXCR4 causes a calcium influx into the cell. Binding of AMD3100 to CXCR4 has partial agonistic effects, and intracellular calcium concentrations increase proportionally with the AMD3100 concentration. This effect may have caused magnesium to shift intracellularly with a subsequent decline in serum magnesium concentrations in the present study. It is unclear, however, why serum magnesium concentrations are inversely affected in cats and humans. Serum calcium concentrations did not change significantly in the present study. To counteract a potential decrease of serum calcium through influx into cells after binding of AMD3100 to CXCR4, a fast mobilization of calcium from storage sites (ie, bone) is likely to occur immediately caused by the closely regulated calcium homeostasis. Initial clinical phase I trials in humans with AMD3100 showed an unexpected increase in the white blood cell counts. The drug specifically increased the CD34+ hematopoietic stem cell counts in peripheral blood by antagonizing the interaction of stromal-derived factor 1 (SDF-1) with its CXCR4 receptor. However, in cats, no such marked increase of white blood cell counts was observed.
In the present study, no evidence of development of resistance of FIV against AMD3100 was found. Resistance of HIV against AMD3100 could be induced after 20–60 passages in the presence of the bicyclam drug in cell culture. The mechanism by which resistance developed is believed to be a mutation in the V3 region of gp120. Perhaps, the treatment period of 6 weeks in the present study was not long enough to select for mutations.
A limitation of the study is the relatively low number of cats per treatment group. However, as for several variables, a significantly different development between groups could be demonstrated despite the relatively small sample size, the study still provides valuable information. Based on the results of this study, additional studies including larger numbers of cats and potentially longer treatment periods are now advisable.
In conclusion, this study evaluated efficacy and adverse effects of the compound AMD3100 against FIV infection. It was shown that AMD3100 displayed a limited but significant antiviral activity in naturally FIV-infected cats and did not lead to clinically relevant adverse effects, without observed development of resistance during a 6-week treatment period. Because there are currently very few drugs with proven efficacy against FIV, and because most have significant adverse effects, AMD3100 could be considered as a sole therapy or as part of a multidrug treatment. However, use of AMD3100 in combination with PMEA cannot be recommended.
The authors thank K.U. Leuven (GOA 05/19) and University of Utrecht for partial financial support. We thank Prof Dr David Maggs, Davis, CA, for careful reading of the manuscript.
Mozobil, Genzyme Corporation, Cambridge, MA
Hepsera, Gilead Sciences, Foster City, CA
PetCheck Anti-FIV; IDEXX, Portland, MA
Cell-Dyn 3500; Abbott Laboratories, Abbott Park, IL
Hitachi 911; Roche Deutschland Holding GmbH, Grenzach-Wyhlen, Germany
FACS; Becton Dickinson, Heidelberg, Germany
Cell Quest 1.1.1. and Macintosh Quadra 650 hardware, Apple Inc, Cupertino, CA
QIAamp Blood Kit; Qiagen GmbH, Hilden, Germany
QIAamp Viral Kit; Qiagen GmbH
ABI 7700, Applied Biosystems, Foster City, CA
version 1.6.3., Applied Biosystems
SuperScript III Platinium One-Step Quantitative RT-PCR System, Invitrogen, Austria
ABI Prism 7700, Applied Biosystems
Statistical Package for the Social Sciences, Version 11.5, International Business Machines Corporation, Armonk , NY
- 15Preferential feline immunodeficiency virus (FIV) infection of CD4+ CD25+ T-regulatory cells correlates both with surface expression of CXCR4 and activation of FIV long terminal repeat binding cellular transcriptional factors. J Virol 2005;79:4965–4976., , , et al.
- 18AZT in the treatment of feline immunodeficiency virus infection: Part 2. Feline Pract 1995;23:13–20..
- 20Kombinierte anti-retrovirale Therapie (GART) bei einer mit dem felinen ImmunschwxE4chevirus (FIV)-infizierten Katze. Kleintierpraxis 2004;49:517–524., , , et al.
- 26Possibilities and limitations of molecular diagnostic methods in clinical microbiology: Feline immunodeficiency virus as an example. Wiener Tieraerztl Wochenschrift 2000;87:269–277., , .
- 27AZT in the treatment of feline immunodeficiency virus infections, Part 1. Fel Pract 1995;23:16–21., , .
- 38AMD3100, a CXCR4 antagonist, reduced HIV virus load and X4 virus levels in humans. In: 9th Conference on Retroviruses and Opportunistic Infections, Seattle 2002; 53., , , et al.
- 39Effects of bicyclam on FIV infection in vivo. In: 7th International Feline Retrovirus Research Symposium (IFRRS). Pisa, Italy; 2004., .
- 44AnorMED. Product information 2003. In: http://www.genzyme.com.