icon-folder.gif   Conference Reports for NATAP  
  18th HIV Drug Resistance Workshop
June 9-12 2009
Ft Myers Florida
Back grey_arrow_rt.gif
HIV Reservoirs and Potential Eradication
  XVIII International HIV Drug Resistance Workshop, Fort Myers, Florida, USA, June 9-13, 2009
Mark A. Wainberg, Ph.D.
Professor and Director,
McGill University AIDS Centre

HIV Reservoirs and Potential Eradication

The first session of the conference dealt with the very important subject of HIV reservoirs and potential eradication. The keynote speaker for the session was Dr. David Margolis of the University of North Carolina at Chapel Hill. Dr. Margolis explained that the major problem in regard to potential eradication of HIV is the rare, persistent proviral genome that lies sequestered within host cells. To address this issue in the context of na´ve T cells as well as activated CD4+ cells and memory CD4+ cells. Dr. Margolis explained that attempts have been made to perturb persistence of low level plasma viremia by intensification with such drugs as Raltagravir onto already effective antiretroviral regimens consisting of two NRTIs and a NNRTI. Unfortunately, such intensification has not led to changes in regard to the ability of persistent plasma viremia to continue. In addition, Dr. Margolis explained that low-level plasma viremia is not significantly affected by intensification with other drug classes such as PIs, NNRTIs, or Enfuvirtide.

In certain studies, a transient increase in numbers of 2 LTR circles or episomal viral DNA have been reported in some patients following use of raltagravir. However the results are inconsistent and it is possible that not enough drug may have been used to make a difference. Thus, there seems to be little hope for the moment that we will be easily able to deal with the problem of persistent viral expression in long-lived cells. It is interesting that a recent paper in Nature Medicine (2009) by the group of R-P Sekaly shows that HIV reservoirs at least seem to be limited to lymphoid rather than other types of tissues such as brain, kidney, etc. This may, at least, somewhat narrow the problem.

Dr. Margolis also asked whether or not a specific cell type might be targeted in regard to this problem. He raised a series of three questions as follows:

1. Does homeostatic proliferation of cells by mitosis play a role in persistent viremia?

2. Are some cells resistant to apotosis? This would allow virally infected cells to continue to proliferate?

3. What about the possibility that direct cell to cell viral infection takes place?

Indeed, most observers would agree that the third of the above possibilities probably occurs on a regular basis. Again however, there does not seem to be any clear indications of what the field should do to obtain eradication of HIV from infected hosts.

This subject is of utmost importance, as there is general agreement that we must continue to strive to eradicate HIV from the body to attain a cure. In this context, a recent Swiss study showed that residual viremia is due to latent cells and not merely to continuing viral replication. Dr. Margolis explained that efforts to activate latently infected pools of lymphocytes have so far not been successful, nor has it been possible to depress the pool size in regard to latently infected lymphocytes. He explained that more efforts need to be made in regard to better understanding integration sites and what the role might be of flanking sequences in regard to viral gene expression. He also made mention of the potential to use activation markers such as TNFα and NFκb. In this context, both valproic acid (VPA) and phytohemagglutinin (PHA) can affect gene expression of multiple cellular activators in many different ways. However, the results obtained are often difficult to understand due to a wide variety of effects that might be mediated by use of these agents. Thus, it is difficult to understand how any simple activation might work or what the role might be of T host transcription factors in regard to sequestration of viral DNA.

Dr. Margolis also suggested that modifications of viral RNA transport might provide some answers. This is especially important in view of the fact that the LTRs seem to be stably integrated into host cells. In this context, more attention could be paid to histones and chromatin. As an example, the enzyme histone deacetylase-1 might be important, since it might contribute to quiescence of the HIV genome.

Dr. Margolis has himself performed a number of resting CD4 outgrowth experiments, and, using VPA, has been able to recover virus after stimulation at levels that are equivalent to those attained through use of PHA. Possibly, the use of histone deacetylase inhibitors might also allow for stimulation of viral production to take place from latently infected cells.

In summary, we need new approaches to target latent reservoirs through efforts at activation. Is it possible that such compounds as IL7 might play a useful role in this regard, in view of the ability of this substance to promote cell proliferation? In contrast, is it also possible that the use of such mediators might promote cell division and hence give rise to a greater abundance of latently infected cells? Possibly, the use of humanized mice that can be treated with antiviral drugs will represent a useful model toward further exploration of this subject.

Resistance to reverse transcriptase inhibitors: Polymerase domain

This area of research represents basic science studies, some of which may or may not be shown to have clinical relevance in the future. Abstract 17 by K. Das et al was on the topic of high resolution structures RNA: DNA complexes together with reverse transcriptase and either Tenofovir or dATP. The authors determined the crystal structures of complexes that included either wild-type or mutated RT. The mutation that was chosen to investigate these structures was at position D498N, because this mutation blocks potential RNaseH activity. The latter is essential for HIV replication and is thus a legitimate target for future drug development. The results show that these ternary complexes can easily be generated and provide new insights into binding of the active form of tenofovir, i.e TFV-DP at the polymerase active site in reverse transcriptase. These findings provide further evidence for the excellent antiviral activity of Tenofovir and add to our understanding of rationally-based evidence to support the continuing use of this drug.

In Abstract 18, Dr. KL White of Gilead Sciences described studies of a prodrug termed GS-9131, that is converted into GS-9148, and that is currently being developed by Gilead for potential therapeutic use. The authors of this study have selected out the Q151L mutation in tissue culture after exposure to GS-9148 for over 100 days. This mutation apparently confers approximately 20-fold resistance against each of GS-9148 and GS-9131. The accumulation of additional mutations at positions L74I, K70N/E and L187F/M led to higher levels of resistance. Interestingly, the Q151L mutation confers hypersusceptibility to both ZDV and TFV and viruses containing the Q151L substitution are efficient in regard to replication capacity. Thus, the Q151L resistance mutation may represent a novel pathway toward development of resistance to GS-9148. The fact that Q151L may render TFV and ZDV hyperactive against HIV replication suggests that future combination therapy involving GS-9131 with either of these agents may be possible. (from Jules: development of this drug has been put on hold by Gilead)

In Abstract 19, M Gotte and colleagues discuss mechanisms of resistance against acyclovir that might be conferred by the V75I mutation in reverse transcriptase. This group has previously documented that acyclovir, which is licensed for the treatment of herpes virus infections, may also display anti-HIV activity as a consequence of the phosphorylation of acyclovir by herpes virus type 6 kinase, thus, converting acyclovir into a reverse transcriptase inhibitor. Selection pressure in tissue culture has now led to the identification of the V75I, M184V, and T69N mutations as substitutions that might impact on the efficacy of acyclovir as an anti-HIV drug. The results of studies by this group also document that the V75I mutation results in diminished anti-HIV acyclovir activity on the basis of discrimination, i.e the reactive form of acyclovir is no longer well recognized by RT molecules that contain V75I. It is also possible that several different resistance pathways may affect acyclovir sensitivity and vulnerability.

Abstract 20 by B. Marchand et al dealt with mechanisms of inhibition and resistance aimed at enhancing the activity of location-deficient reverse transcriptase inhibitors. This group has studied a novel compound termed EFdA and has demonstrated that it can select for resistance on the basis of mutations at positions M184V, T165R, and I142V. The degree of resistance seems to become larger as these mutations accumulate. It appears as though both discrimination and excision mechanisms may be involved in the development of resistance against this compound. It is not certain whether this compound will have potential for clinical development.

Abstract 21 by C. Dobard and colleagues was on the topic of recombinant viruses derived from subtype B that contain a subtype C reverse transcriptase, in regard to whether or not enhanced development of the K65R resistance mutation may take place in subtype C viruses. These authors constructed recombinant viruses-containing subtype C reverse transcriptase in a subtype B background. They then performed tissue culture selection experiments using the drug tenofovir and showed that both subtype B and subtype C viruses were equally adapt at developing the K65R substitution. Moreover, this occurred quickly and only required approximately 40-45 days in each case. However, ultrasensitive analysis seemed to indicate that the subtype C viruses may have had a predilection to develop the K65R mutation more rapidly than the wild-type subtype B strains. This topic is, in fact, of clinical relevance in view of the fact that other groups have published data indicating that subtype C viruses may select the K65R mutation under drug pressure with a variety of compounds faster than occurs in subtype B (Brenner BG, Oliveira M, Doualla-Bell F, Moisi DD, Ntemgwa M, Frankel F, Essex M, and Wainberg MA. HIV-1 Subtype C Viruses Rapidly Develop K65R Resistance to Tenofovir in Cell Culture. AIDS. 20:F9-F13, 2006). More importantly, however, there have now been several clinical reports that attest to the fact that the K65R mutation seems to develop to greater extent in patients with subtype C infections than has been the case with subtype B, (e.g. Hosseinipour MC, van Oosterhout JJ, Weigel R, Phiri S, Kamwendo D, Parkin N, Fiscus SA, Nelson JA, Eron JJ, Kumwenda J. The public health approach to identify antiretroviral therapy failure: high-level nucleoside reverse transcriptase inhibitor resistance among Malawians failing first-line antiretroviral therapy. AIDS. 2009 1;23(9):1127-34). This paper showed that high proportions of individuals in Malawi, infected by subtype C viruses, and failing a regimen of co-formulated Triamune (d4T/3TC/Nevirapine) have developed K65R and follows an earlier study by F. Doualla-Bell and colleagues from Botswana who also reported high levels of K65R presence among smaller numbers of individuals who had failed regimens of either d4T/3TC/ddI or d4T/3TC/NVP. The upshot is that K65R may be rapidly selected in subtype C viruses, an issue that assumes vast importance since co-formulated Triamune is the most widely prescribed HIV medication in the world. Should it be true that K65R selection by stavudine occurs more frequently with subtype C viruses, this might prove to be disastrous for individuals living in regions of sub-Saharan Africa in which subtype C viruses are predominant. Further research is necessary to clarify this subject. It is also relevant that at least one manuscript (Coutsinos D, Invernizzi CF, Xu H, Moisi D, Oliveira M, Brenner BG, Wainberg MA: Template Usage Is Responsible for the Preferential Acquisition of the K65R Reverse Transcriptase Mutation in Subtype C Variants of Human Immunodeficiency Virus Type-1. J Virol 83(4):2029-33, 2009) has shown that a template-based mechanism may be responsible for the faster selection of K65R by subtype C viruses.

Abstract 22 by D. Nissley et al dealt with the topic of HIV reverse transcriptase fidelity variants. This group has studied the existence of RT variants that affect fidelity and has shown that these map to certain expected regions of HIV RT but to others that are less obvious. Fidelity variants were shown to cluster in the fingers region of RT and may exert their influence through a long range effect on nucleotide discrimination at the enzymatic active site. Other mechanisms may also be responsible for the fact that different RT molecules may display different levels of fidelity. This work indicates that approaches that may augment RT fidelity and prevent resistance mutations are still considered meritorious.

In Abstract 23, V. Svicher et al investigated the appearance of either the M184V or M184I mutations in patients who had failed regimens consisting of either FTC/TDF or 3TC/TDF. Some of the individuals in question had previously experienced therapy with 3TC prior to switching to FTC. The findings document that patients were less likely to develop the M184V mutation if they failed FTC + TFV than if they failed regimens containing 3TC + TDF (39% vs 56%). This abstract is one of several findings that suggest that FTC may be a superior drug to 3TC with regard to forestalling the development of drug resistance, a subject that may be of potential economic significance in view of the fact that 3TC will probably become a generic product during the 2011 calendar year. It may be anticipated that third party payers may therefore try to reimburse for 3TC rather than FTC in the context of TDF-containing regimens. The current study may provide strong arguments why the use of FTC should be encouraged over 3TC and why pressures from third party payers to make a substitution from FTC to 3TC should be resisted.

In Abstract 24, CF Perno and colleagues studied the development of resistance mutations against lamivudine in the reverse transcriptase of Hepatitis B virus (HBV). They showed that such mutations can be selected even at very low levels of virus replication and observed this in a series of 25 individuals who had been on lamivudine monotherapy as well as in 39 individuals who first received lamivudine and then switched to lamivudine/adefovir. Individuals in the latter category were also prone to develop more complicated patterns of resistance-associated mutations in regard to lamivudine.

Structure of the entire HIV-1 RNA genome

In Abstract 28, KM Weeks at the University of North Carolina presented data on novel mechanisms of understanding the HIV RNA genome. This research group has now developed novel chemical techniques that permit the analysis of primer extension reactions through use of 2' hydroxyl acylation reactions. They have generated a cDNA library and have developed plots that are termed SHAPE to determine RNA structure. The results have shown that there probably linkers between the proteins of retroviruses that are marked by regions of RNA enhanced structure. These authors have shown that 14 or 15 polyprotein linkers are encoded by very highly structured RNA. Furthermore, mutations that decrease RNA structure in these regions are associated with diminished replicative fitness. These authors have also identified packaging signals that seem to be due to clustering and believe that certain of these areas might serve as novel targets for future drug development. These enhanced methods for understanding RNA structure will help us to understand aspects of the HIV replication cycle that have hitherto been difficult to analyze. In addition, these methods may point the way toward further identification of potential future targets for drug discovery.

Resistance to reverse transcriptase inhibitors: connection/RNase H domains

In Abstract 29, KA Delviks-Frankenberry and colleagues reported on a series of polymorphisms within the HIV RT connection domain that are associated with high level resistance against ZDV. This group has been able to show that treatment experienced patients who have received zidovudine may have mutations that confer resistance to this drug that are independent of thymidine analogue mutation (TAMs) and that such mutations are located within the connection domain. Interestingly, certain of these mutations within the connection domain were able to confer higher levels of drug resistance against ZDV than are conferred by certain TAMs. It is possible that this may explain levels of natural resistance against ZDV in certain viral subtypes in which a connection domain may perhaps be different than is seen in subtype B viruses. In particular, the connection domain of A/E recombinant viruses stands out in this regard. It is also possible that this effect may be independent of the RT subtype in regard to reverse transcriptase, since it is not known to what extent RNase H activity may be impacted by connection domain mutations. The evidence now shows that certain of the connection domain mutations can diminish levels of RNase H mediated cleavage and underlines the fact that mutations within the connection domain may ultimately be of far greater importance than is now thought to be the case in regard to our understanding of drug resistance. Furthermore, it is conceivable that resistance genotyping may have to extend on a routine basis into both the connection domain and RNase H in order to provide a comprehensive understanding of resistance against certain NRTIs.

In Abstract 30, J Radzio and colleagues studied the N348A mutation in reverse transcriptase and documented that this substitution can be associated with resistance against two NNRTIs, NVP and EFV. Substitutions at N348I also gave rise to higher rates of excision in regard to ZDV triphosphate. In contrast, mutations that did not affect excision did not lead to lower levels of RNase H activity. This suggests that the mechanisms that involve the N348I mutation that are responsible for resistance to NVP and ZDV are probably different. Certainly, mutations that occur within the connection domain that limit sensitivity to ZDV may not always affect sensitivity to Nevirapine. The reverse of this is true as well. These authors also demonstrated that the viruses in question retain sensitivity to zidovudine in the absence of any TAMs.

In Abstract 31, VK Pathak and colleagues also studied connection domain mutations and have documented that these do not lead to enhanced levels of drug resistance against NNRTIs in treatment na´ve patients but do seem to increase overall levels of resistance in treatment experienced patients. They also suggest that there is likely a correlation with enhanced excision of ZDV. Furthermore, mutations that reduce RNase H activity seem to also reduce sensitivity to certain NNRTIs such as NVP and DLV but much less so in regard to EFV and ETV. They have also shown that a mutation at position D545N together with the 103N mutation may increase levels of resistance against NNRTIs to significant extent. They also obtained similar findings in regard to combinations of D545N and well established NNRTI mutations within RT at positions 181C and I79F. They postulate that a stalled inhibitor/RT template/primer complex may be involved on a mechanistic basis in the occurrence of such high levels of resistance.

In Abstract 32, J Brehm and colleagues discussed the virologic failure of regimens containing several NNRTIs + Efaverenz and found that this does not seem to be associated with the selection of mutations in either the connection or RNase H domains of RT. In particular, they identified changes at positions 467I and 530R that are sometimes associated with treatment failure. In addition, substitutions at positions 334H and 334Y may be seen in failure, but 334L, which may also be found at baseline, might possess a protective effect against subsequent development of resistance. In brief, 49% of all samples studied at failure had mutations in either the connection domain or in RNase H. In general, though not many significant differences were reported in regard to either the connection or RNase H domains. Both M184V and K103N were the most common mutations seen among individuals failing therapy.

Abstract 33 by J. Vingerhoets and colleagues dealt with an analysis of mutations in the RT connection domain that were observed in individuals who participated in the DUET studies, in which both ETR (etravirine) and darunavir were used as important components of treatment regimens. The authors have demonstrated that each of the G33D, G335D, and A376S substitutions are associated with diminished responsiveness against ETR. However, connection domains mutations were not linked to any ETR resistance associated mutations but were more prevalent in samples that contained a high level of TAMs. These authors were unable to demonstrate any strong linkage between the existence of mutations in either the connection domain and/or RNase H with resistance against ETR. These findings are important because they provide assurance that ETR is likely to remain effective against viruses that contain classical mutations associated with resistance against EFV and NVP, and that the occurrence of novel mutations within either the connection or RNase H domains of RT are unlikely to affect responsiveness against ETR.

HCV protease inhibitor resistance

Abstract 34 was on the subject of the HCV NS3/4A a protease and its potential impact on anti-HCV compounds. The authors have shown that drug resistance is likely to occur in a manner that will weaken the binding of inhibitors but that substrate recognition and cleavage might nonetheless continue to take place under this circumstance. They argue for the development of protease inhibitor that will be able to fit better within the substrate binding region of the HCV protease because such inhibitors should then be less susceptible to the development of drug resistance mutations.

Global resistance and epidemiology

In Abstract 35, CM Parry and colleagues studied determinants that compensate for diminished fitness following the development of resistance to HIV protease inhibitors. This group employed a single cycle replication assay to compare the susceptibilities of two different regions of mutated gag and employed site-directed mutagenesis to identify changes in the viral matrix (MA) and capsid (CA) genes that might restore replication competence to viruses that contain mutated protease enzymes. The authors determined that changes in both MA and CA can confer high degrees of resistance against protease inhibitors and further determined that three amino acid changes within the MA protein can restore replication competence to appreciable extent. These findings are independent of the existence of cleavage site mutations and argue that further studies that are performed with full length gag will be necessary to better understand this issue. In particular, they have observed that changes within matrix protein at position 76, 79, and 81 may be particularly efficient with regard to restoration of replication capacity.

In Abstract 36, AD Revell and colleagues have presented an analysis of computational models that have been developed to help predict response to HIV therapy in resource-limited settings. The rationale for this study is that the use of model systems that do not depend on actual genotyping might be predictive of treatment responsiveness and/or failure. The authors have identified a number of variables that seem to have excellent predictive value in regard to responsiveness to therapy of which viral load levels are the most important. This work is interesting and may provide first steps toward alternative mechanisms of monitoring the impact of drug resistance in developing settings.

In Abstract 37, G. Hunt and colleagues monitored drug resistance patterns among HIV-infected children and adults who had failed boosted lopinavir-based regimens in South Africa. The authors were able to show that higher percentages of adults than children failed therapy without resistance mutations in regard to lopinavir. The most frequent mutation to appear among treatment failures was M184V followed by classic NNRTI mutations among individuals who had also been treated with NNRTIs. Among children developing resistance against lopinavir, the V82A and I54V mutations were most common. The existence of NNRTI mutations among individuals failing LPV/r may reflect earlier first line therapy with NNRTI-based regimens. One concern that was expressed is that individuals in South Africa may sometimes need to take anti-TB drugs such as Rifampin. The latter is of concern in view of the fact that rifampin can cut levels of LPV in the circulation by as much as 80%. Indeed, considerable numbers of the individuals followed in South Africa had received dual lopinavir/Rafampin therapy.

Abstract 39 by V von Wyl et al was on the topic of transmitted drug-resistance and provides insights from the Swiss national cohort study. These investigators completed a cross-sectional analysis that suggests that at least 11% of all sequences containing transmission-related mutations and one fifth of all transmission clusters probably originated from single patients. This work confirms earlier studies on transmission clusters and the dangers of onward transmission chains in regard to HIV disease. A very important point that is underlined by this study as well as others is that transmission of drug resistance can also occur from non-treated patients who are newly infected by HIV, who possess high levels of viremia at the time that likely transmission take place. (from Jules : a discussion at the meeting emerged during the sesssion related to the recent seemingly increased number of legal cases of criminalization of sexualy transmitted HIV. I twas said at the microphone by one of the meeting participants tht researchers in this area should be careful to protect the names of individuals of whose samples they use).

Abstract 40 by J. Lidstrom et al was on the topic of zidovudine usage in resource-limited settings. In particular, these investigators conducted a study of the use of extended formulation ZDV alongside extended formulation NVP to prevent mother-to-child transmission in Malawi. The results show that the use of extended ZDV plus extended NVP significantly reduced the risk of development of resistance at 14 weeks in infants who became HIV infected. In order for this result to be obtained, HIV needed to be diagnosed and prophylaxis employed using antiretroviral drugs that were only terminated by the time the infant reached 6 weeks of age. The basis for this observation is that the continued use of antiviral drugs in HIV infected infants will likely select for drug resistance. These findings further indicate that the use of extended formulation ZDV is a good idea for infants carrying HIV infections.

In Abstract 41, ES Svarovskaia and colleagues examined the presence of the K103N mutation among treatment-na´ve patients who have received EFV as part of a triple regimen. The patients in this study were derived from the GS934 treatment protocol that had randomized individuals to receive either FTC/TDF/EFV or 3TC/ZDV/EFV. The authors used a realtime PCR assay and allele-specific PCR to examine the presence of the K103N substitution. The results revealed higher levels of K103N in treatment-na´ve patients than previously recognized and that the presence of this mutation was associated with a risk of virological failure within the group of individuals exposed to 3TC.

Abstract 42 by M. Lataillade et al provided novel insights in regard to the CASTLE study in which individuals had received boosted atazanavir together with TFV/FTC (Truvada) for treatment of HIV infection. The authors used pyrosequencing to examine virologic failure after 48 weeks of therapy and found that approximately 6.5% of individuals possessed the 184V mutation on the basis of pyrosequencing but that ordinary bulk sequencing was unable to detect this same mutation within the same individuals. Six of the nine individuals in whom M184V was detected went on to virologic failure. Certain of the individuals may have had transmitted drug resistance, which may also account for certain of the results obtained in this study. The pyrosequencing analysis indicates that as many as 30% of individuals possessed at least one mutation associated with transmitted drug resistance. Similar rates of transmitted resistance were identified in both B and non-B subtypes but transmitted drug resistance did not, in general, affect virologic responses for individuals on either boosted ATV or boosted LPV during the 48 week period of investigation.

In Abstract 43, JA Johnson and colleagues studied the use of bulk sequencing to identify resistance mutations in viral RNA and peripheral blood following use of single dose Nevirapine for prevention of mother-to-child transmission of HIV. Higher percentages of mutations were identified by allele-specific realtime PCR analysis then by bulk sequencing. The results also revealed that the presence of mutations that were identified by bulk sequencing was associated with treatment failure, which was highest among individuals who had started therapy within 6 months of receiving single dose Nevirapine.

In Abstract 44, T Wagner and colleagues studied the development of resistance mutations using an oligo-nucleotide ligation assay for HIV-1 DNA in regard to use of Nevirapine to prevent mother-to-child transmission of HIV-1. These investigators studied 181 women who had received single dose Nevirapine followed by HAART for a total of 9.5 months. Resistance against NVP was detected in 26% of these individuals. Virologic failure was mostly associated with mutations at positions K103N, Y181C, and G190A. This result makes clear that access to rapid, inexpensive assays to detect low levels of resistance against NVP prior to initiation of ART in women who had received single dose Nevirapine for PMTCC may enhance future responsiveness to antiretroviral therapy among such subjects.

Finally, Neil Parkin of the World Health Organization (WHO) delivered an impassioned lecture to close the conference. In his address, Dr. Parkin asked the audience to take cognizance of the fact that the benefits of antiretroviral therapy have now been extended to millions of people in developing country settings. While the occurrence of both acquired, i.e. treatment-related, and transmitted drug resistance will be inevitable consequences of the provision of ARVs throughout the world, WHO has established training and surveillance programs that link both developing and developed countries to ensure that proper monitoring of this problem will be in place. Dr. Parkin shares the concerns of many that the drugs that are currently available in many developing country settings may not be as good as those that are routinely used in western countries and expressed the hope that the situation will change over the next several years.


Mark A. Wainberg, Ph.D.
Professor and Director,
McGill University AIDS Centre