Reappearance of Minority K103N HIV-1 Variants after Interruption of ART Initiated during Primary HIV-1 Infection

Background In the Zurich Primary HIV infection study (ZPHI), minority drug-resistant HIV-1 variants were detected in some acutely HIV-1-infected patients prior to initiation of early antiretroviral therapy (ART). Here, we investigated the reappearance of minority K103N and M184V HIV-1 variants in these patients who interrupted efficient early ART after 8–27 months according to the study protocol. These mutations are key mutations conferring drug resistance to reverse transcriptase inhibitors and they belong to the most commonly transmitted drug resistance mutations. Methodology/Principal Findings Early ART was offered to acutely HIV-1-infected patients enrolled in the longitudinal prospective ZPHI study. Six patients harboring and eleven patients not harboring drug-resistant viruses at low frequencies prior to ART were included in this substudy. Minority K103N and M184V HIV-1 variants were quantified in longitudinal plasma samples after treatment interruption by allele-specific real-time PCR. All 17 patients were infected with HIV-1 subtype B between 04/2003 and 09/2005 and received LPV/r+AZT+3TC during primary HIV-1 infection (PHI). Minority K103N HIV-1 variants reappeared after cessation of ART in two of four patients harboring this variant during PHI and even persisted in one of those patients at frequencies similar to the frequency observed prior to ART (<1%). The K103N mutation did not appear during treatment interruption in any other patient. Minority M184V HIV-1 variants were detected in two patients after ART interruption, one harboring and one not harboring these variants prior to ART. Conclusion Minority K103N HIV-1 variants, present in acutely HIV-1 infected patients prior to early ART, can reappear and persist after interruption of suppressive ART containing two nucleoside/nucleotide analogue reverse transcriptase inhibitors and a ritonavir-boosted protease inhibitor. Trial Registration Clinicaltrials.gov NCT00537966


Background
Twenty years after emergence of the AIDS epidemic and a decade after introduction of highly successful combination antiretroviral therapy (cART) HIV-1 remains one of the three leading infectious diseases in the world causing a high burden of morbidity, mortality and costs [1,2].
Despite an enormous body of knowledge accumulated, a number of key questions regarding HIV-pathogenesis remain unresolved. To date it is still not fully understood what factors determine the viral setpoint in a given individual. Even less is known about factors determining viral transmission. It has been recognized that longitudinal studies of HIV-infected patients early on after primary HIV infection (PHI) may be key to better understanding of HIV-pathogenesis [3][4][5][6][7][8]. Insights of early events between virus and host may potentially be crucial for development of vaccines, new drugs, microbiocides and new strategies optimizing long-term antiretroviral therapy. Despite 10 years of cART to date it remains unknown whether early-cART initiated during PHI can alter the course of disease in delaying disease progression [5,[9][10][11][12][13][14].
Thus, taken together there is a strong need for systematic longitudinal pathogenesis studies in humans infected with HIV-1 addressing transmission and early virus-host events.

Early antiretroviral treatment after primary HIV infection.
The optimal time to start antiretroviral therapy for HIV-infected patients remains unknown [15][16][17][18]. Even less clear is whether treatment of PHI early on is beneficial in terms of delaying disease progression. Recent work provided some rationale that early intervention during PHI might indeed be beneficial. Yerly et al. [19], demonstrated that viral rebound and setpoint upon interruption of cART in chronically infected patients is predicted by proviral DNA levels at time of treatment stop. Furthermore, Strain et al. [20] showed that reduction of the size of the latent reservoir is significantly more pronounced in early-cART treated PHI patients when compared to initiation of cART in chronic infection. Furthermore, this work also demonstrated that the decay of proviral HIV-DNA is similar to the decay of cell infectivity assessed by terminal dilution co-culture. In conjunction, one could hypothesize that a reduction of the latent reservoir indeed could translate into lowered viral setpoints and subsequent delay of disease progression. For this reason, proviral DNA as a valuable surrogate marker of the long lived cellular reservoir and potentially of the viral setpoint needs to be revisited in the context of primary HIV-infection. Furthermore, potentially other viral markers such as different HIV-RNA species [21][22][23][24][25][26][27][28] may be of additional value. A recent observational study suggested that early-cART treated PHI patients exhibited lower viral setpoints after stop of early-cART compared to controls when adjusted for baseline CD4 counts. Interestingly the CD4 count benefit observed, persisted for up to 72 weeks [9]. Another very recent study reported that a median duration of HAART of 1.1 years was associated with a persistent effect on viral load and CD4-T-cell count for up to 5 years after seroconversion [14]. In contrast to these results a short term treatment of 24 weeks did not show any benefit on viral setpoint and CD4 count [13]. In the near future, we mainly depend on observational studies because randomized clinical trials in this setting are not feasible [5].

Viral and host factors influence the HIV-1 viral setpoint
HIV infection is characterized by continuous viral replication at a high rate, which combined with the error rate of the reverse transcriptase [29,30], frequent recombination [31, 32], and host selection pressure leads to a high genetic diversity in infected individuals [33][34][35][36][37]. However, the level of diversity between individual patients can vary considerably [36,[38][39][40]. After primary HIV infection each individual reaches a steady state level of plasma viral load which is referred to as viral setpoint. These setpoints between individuals can vary more than 1000 fold [41]. The switch in coreceptor usage from CCR5 to CXCR4, which occurs in approximately 50% of patients, is associated with more vigorous viral replication in vivo exemplified by an increase in plasma viremia levels in HIV-infected individuals and more rapid disease progression [41][42][43][44][45][46]. Various viral and host properties may contribute to the observed diversity: these include differences in virulence, subtype, immunogenicity and replication capacity of the transmitted viruses, the quasispecies composition of the infecting inoculum (transmission of single versus multiple quasispecies), host genetic factors such as chemokine receptor polymorphisms, HLA types and gender differences [47][48][49][50][51]. HIV-specific CD4 and CD8 responses also contribute to control of viremia, however, optimal measurements for functional HIV-1 specific CD8 responses in humans have still to be determined. In particular, various recent studies in humans [10,[52][53][54] in contrast to earlier ones [55] were not able to link levels of HIV-specific CD8 responses with the viral setpoint. Neutralizing antibodies arising within weeks to months after PHI are also thought to contribute to the viral setpoint. Evidence, however, is mostly indirect by escape of neutralization responses and viral evolution of the env gene most likely driven by neutralizing antibodies directed against env [35,37,56]. Recently, we have shown that higher neutralizing antibody titers against autologous viruses indeed did correlate with viral setpoint reached in patients undergoing structured treatment interruptions [38]. More convincingly, first direct evidence for neutralizing activity in vivo was demonstrated by our passive immunization study, where a cocktail of three broadly neutralizing antibodies was administered to patients with suppressed viremia. Upon stop of cART, viral rebound could be delayed significantly, demonstrating that neutralizing antibodies indeed can have a direct effect on viral replication in vivo [57,58] [59].
In addition to viral factors and immune responses it has been well described, that a variety of host genetic factors are influencing HIV-1 disease progression in a negative or positive way. Polymorphisms at the following genotypic locations have been found to influence disease progression: CCR5 delta 32, CCR5 G-2455A, CCR2 V64I, RANTES G-403A, RANTES C-28G, MIP-1a T113C, SDF-1 3'A. Moreover, a variety of HLA alleles have been found to accelerate AIDS (HLA-A, B, C homozygosity [60,61], HLA-B*35 [62,63], HLA-A*01-B*08-DRB1*03 [64,65] and others have been demonstrated to delay disease progression such as HLA-B*57 and HLA*B27 [66]. Of interest one HLA association was even shown to correlate with maintenance of viral suppression in patients treated early after seroconversion; HLA-DRB1*13-DQB*6 [67]. Recently, a whole genome association study of major determinants for host control revealed three new relevant polymorphisms which can explain approx. 15% of the variation among viral setpoints in untreated asymptomatic patients [68]. For these reasons, it is evident that predictors of the viral setpoint cannot be studied anymore without also assessing these important factors -at least in partin modulating the course of HIV-1 disease. Nevertheless, it has been shown many times that effects of these host genetic traits are subtle within single individuals and large patient cohorts are normally needed to demonstrate such effects. Thus, for our project we do not focus exclusively on host genetics, but rather will perform these analysis to exclude a potential bias as we have done so before [24,69].
Taken together, prerequisites to comprehensively study viral and host factors are well defined patient groups that are followed longitudinally.

Current status of treatment interruptions in HIV-infected patients.
Originally, structured treatment interruptions were studied systematically for three reasons: Firstly, to test the autovaccination hypothesis, secondly, to reduce toxicity, and thirdly, to reduce exposure to antiretroviral drugs and therefore reduce costs. The autovaccination hypothesis has been rejected in chronically infected as well as in patients infected during primary HIV infection, meaning that short term-treatment interruptions, despite the fact that they could induce HIV-specific T-helper,cytotoxic CD8 and humoral immune responses were not able to sustainably lower viral setpoints [10,53,54,[70][71][72]. CD4-guided treatment interruptions in lower CD4 strata were associated with increased frequencies of HIV-, non-HIV-related morbidity and non-HIV related mortality in two large trials [73,74]. However, increased morbidity and mortality were not seen in higher CD4-strata [75][76][77]. No data of structured treatment interruptions in large clinical trials performed in patients with primary HIV-infections are available to date. However, the fact that morbidity and mortality seen in the large trials were dependent at least in part on the CD4 stratum in which the cycling took place suggests that the risk for interruption of early ART introduced during PHI is minimal because patients with early-cART in most cases do reach normal or almost normal CD4 counts within the first year of treatment. Thus, taken together, the risk of harming patients who stop early PHI treatment is not existent or very low, given that high CD4 counts are reached after early PHI treatment, and that negative treatment interruption effects seen in chronic HIV-infection were very rare in general and CD4-dependen at least to some extent.

Current status of the Zurich PHV infection cohort
The Zurich PHI study has already been started at the beginning of 2002. First, patients were enrolled in the Merck study 520, Protocol No 013-00 (EK 851). At that timepoint we have already applied for a study extension in addition to the Merck protocol and this was accepted under the EK 851 and later on 1086). From our own research initiative, several manuscripts have already been published over the last 3 years: This study so far has been highly successful and for this reason it was decided to continue with our own protocol after Merck has stopped to enrol patients. To date we have enrolled more than 140 patients with documented primary HIV infection within the first 5 years. An interim analysis of 62 patients [78] focusing on clinical aspects like delay to diagnose HIV on first physician contact due to difficulty to recognize the "acute retroviral syndrome" (ARS), symptoms and signs, laboratory abnormalities, primary resistance and mode of transmission revealed that clinical characteristics of the Zurich PHI cohort are comparable with other published PHI studies [79][80][81]. Approximately 90% of patients so far have chosen to undergo early-cART. To this end 50 patients have stopped antiretroviral treatment. Primary resistance has only been detected in 4% so far and approx. 70% were infected with a subtype B strain. Criteria for diagnosis of PHI are: 1) acute retroviral syndrome [78] and negative or indeterminate westernblot in the presence of a positive p24Ag and/or detectable plasma HIV-1 RNA. 2) documented seroconversion with or without symptoms within 90 (acute infection) or 180 days (recent infection). 3) possible ARS, positive westernblot and detectable HIV-RNA, and a negative HIV-gp120 avidity assay [82,83] or a detuned assay [84] (recent infection). Overall, 73 patients (63%) had a negative or indeterminate westernblot, respectively 97 patients could unambiguously be identified to be acutely infected. Thus, the Zurich PHI cohort is well characterized and suitable for the work proposed here.

Patient population
As per 30 July 2007 148 patients were enrolled under the former protocols approved by the Ethics committee (No. EK 852 and EK 1086). All patients enrolled previously will be asked to sign the informed consent of the current project (INFZ-ZPHI 01.01, Version 1) retrospectively. We expect to enroll another 150 patients by the end of 2011. This is a realistic scenario assuming the same rate of new HIV-1 infections as we have observed over the last 5 years.

Patient enrollment:
Patients with documented primary HIV-1 infection who agree to participate in the current study will receive the following routine screening at baseline: total blood count, routine chemistry laboratory (e.g. creatinine, liver enzymes, lipid profile, c-reactive protein, serum-glucose), HIV-1 RNA, p24Ag, HIV-genotypic drug resistance test, Western-blot, (HIV-screening if still necessary), CD4 count, and further laboratory testing belonging to the standard procedure for patients presenting with new HIV-infection (Hepatitis, A, B, C serologies, CMV-, syphilis-, toxoplasmosis-serologies etc.). All these tests are covered by regular health insurance because they reflect standard of care. acute retroviral syndrome [78] (ARS) and negative or indeterminate Westernblot in the presence of a positive p24 Ag and/or detectable plasma HIV-1 RNA documented seroconversion with or without symptoms within 90 days. or B) recent HIV-1 infection, defined as: possible ARS, positive Westernblot and detectable HIV-RNA, and a negative HIV-gp120 avidity [82,83], respectively detuned assay [84]. documented acute HIV-1 infection, however, referral to our center more than 90 days after presumed date of infection.

Study site
This is an open label, non-randomized monocenter, observational study, conducted at the University Hospital Zurich Division of Infectious Diseases and Hospital Epidemiology Rämistrasse 100 8091 Zürich

Patients with intervention:
Intervention a): Early combined antiretroviral treatment (early-cART): All patients with documented acute HIV-1 infection, presenting within 90 days after presentation of ARS or documented seroconversion will be offered early combination antiretroviral treatment with a standard firstline cART regimen as recommended in treatment guidelines [85] containing drugs that are approved by Swissmedic. Treatment will be continued until at least 12 months complete HIV-RNA suppression in the plasma has been achieved (HIV-plasma RNA < 50 copies/ml for at least 12 months (see Figure 1).

Intervention b):
Interruption of early-cART: After one year of fully suppressive early-cART patients can stop cART and then will be switched to the observational study arm (see Figure 1).

Patients without intervention:
All patients with documented acute HIV-1 infection who do not want to have early-cART, respectively patients with documented recent infection will be enrolled in the observational study group only (see Figure 1).

Drug regimes selected for early-cART
Initial treatment consists of a ritonavir-boosted protease inhibitor (PI) combined with two nucleoside reverse transcriptase inhibitors (nRTI) (e.g. lopinavir/ritonavir: 2 x per day 200/50 mg in combination with zidovudin/lamivudin: 2 x per day, 300/150 mg). This is a well established, potent drug combination with excellent efficacy as a first line treatment. This treatment is chosen because the boosted protease inhibitor has a high genetic barrier, which is particularly important because NNRTI resistance mutations transmitted, may hamper the whole NNRTI based first line treatment from the beginning (von Wyl et al, Archives of Internal Medicine, 2007, in press). If side effects or toxicity do arise, respectively results from the resistance test recommends a different drug regimen, different early-cART regimens are chosen. In this case the following alternative boosted-PI regimens will be chosen (atazanavir/ritonavir: once daily 300/100 mg, darunavir/ritonavir: 2 x per day 600mg/100mg). If a boosted-PI regimen is not possible, a non-nucleoside-reverse-transcriptase inhibitor (NNRTI) based regimen (e.g. efavirenz: once daily 600mg or nevirapin: 200mg 2 x per day, or once daily 400mg) is also possible. As alternative NRTI regimens, abacavir and lamivudin: once daily 600/300mg, or tenofovir and emtricitabin: once daily 300mg/200mg will be used.

Safety of cART
Safety of cART will be monitored the same way as this is done in clinical routine: week 2, 4, 8, 12 and thereafter every three months. In women who undergo early-cART and are pregnant a drug regimen will be chosen, which excludes efavirenz.

Study flow chart
a) The diagram below shows a schematic outline of the study: Appendix I (page [15][16]: This table depicts patient visits and planed amounts of blood that will be drawn from patients who will undergo early-cART. Furthermore, in the right column a synopsis of the total amount of blood per timepoint (blood for clinical routine purposes and study blood) is shown.
Appendix II (page [17][18]: This table shows patient visits and planed amounts of blood that will be drawn from patients who will interrupt their early-cART treatment after at least one year of successful suppression (plasma HIV-RNA <50 RNA copies/ml for at least 12 months). Furthermore, in the right column a synopsis of the total amount of blood per timepoint (blood for clinical routine purposes and study blood) is shown.
Appendix III (page [19][20]: This table shows patient visits and planed amounts of blood that will be drawn from patients who have not chosen to start early-cART treatment or who have presented with "recent infection". Furthermore, in the right column a synopsis of the total amount of blood per timepoint (blood for clinical routine purposes and study blood) is shown.

c) Specimen Collection and Processing
Blood specimens will be obtained according to the tables of appendix I -III. EDTA-Blood from all time points will be separated into cells and plasma according to standard procedures. Cellular material is divided into aliquots that are stored either in DMSO medium in liquid nitrogen to yield viable cells for immunological analysis, or stored at -70°C as dry cell pellets to provide material for genetic analysis and viral load measurements. All plasma samples are stored at -70°C.

Endpoints and statistical analysis
Analysis of the parameters described under secondary study objectives will be performed upon availability of data. Exact description of these research projects are not feasible within this study protocol. Briefly, the following analyses are planned: • Viral setpoints of patients having stopped early-cART will be compared with viral setpoints from patients with primary HIV infection without early-cART and with viral setpoints from patients with documented time of seroconversion enrolled in the Swiss HIV cohort study (historic controls). A first interim analysis will be performed in 2009. • Moreover, the cART free interval between patients underoing early-cART (from timepoint of stopping their first cART to timepoint of initiation of second cART according to current guidelines) and patients without early-cART (timepoint of initiation of cART according to current guidelines) will be determined. • In addition, analysis of parameters described under secondary study objectives will be performed after the research data has been generated. In general, these parameters will be measured longitudinally, and kinetics of these parameters will be determined and compared between groups with and without early-cART.
Standard statistical methods will be used for these analyses. Details of the planned analyses have been described in Swiss National Science Foundation research proposals:

Time frame
The current study is the successor of two previous studies: Merck study 520 (EK 851) and the extension of that study (EK 1086). Patient enrollment for these two studies has been conducted between January 2002 -End of August 2007.
We will enroll patients under this new protocol as soon as it is accepted by the ethics committee. We expect to enroll patients from September 2007 -end of 2011. Laboratory investigations and the observational period will continue until 2016.

Informed consent.
Informed consent from each subject participating in the trial will be obtained before any trial specific procedures are performed on the subject. The aims, methods, benefits, and potential hazards of the trial will be explained to each participant before signing the informed consent. It will be made clear that each subject can withdraw their consent at any time and for any reason, without incurring any penalty or withholding of treatment on the part of the investigator. Signed informed consent forms will be kept on file by the investigator and a copy of it will be given to the patient upon request.
Patients who have already been enrolled in the EK protocol No 851 and 1086 will be consented again under this new protocol.
This trial will be registered in the trial registry of the University Hospital Zurich and in the international trial registry "clinical.trials.gov" (www.clinical.trials.gov).

Safety evaluations.
• Vital signs will be monitored at baseline (body temperature, pulse and blood pressure).
• Physical examination will be performed at baseline and later on if needed.
• Complete Blood Count (CBC) including white blood cell differential, and platelet count: at baseline • Chemistry including AST, ALT, bilirubin, creatinin, at baseline • Urine analysis (Uristix and sediment): at baseline and later on if needed. • CD4 count: at baseline • HIV-plasma RNA: at baseline • genotypic HIV drug resistance testing: at baseline For patients who will undergo early antiretroviral treatment safety laboratory (chemistry and CBC will be repeated at week 2, 4, 8, 12 and then every three months thereafter.
Viral load and CD4 monitoring occurs monthly until VL is undetectable and thereafter every three months.
For patients not undergoing early-cART, plasma HIV-RNA and CD4 count will be repeated every 3 months and CBC and chemistry will be repeated every 6 months as it is done in standard clinical care.
If the study participants encounter damage caused by the present study, such damage is covered by the insurance policy of the University Hospital Zurich.

Feasibility
The fact that 148 patients have already been recruited (January 1, 2002 -July 30, 2007) under the previous protocols (EK 851 and EK 1086) demonstrates that the study is feasible. The infrastructure and know-how in the clinic and research laboratories is in place. Thus, the capacity to conduct the study is available. Furthermore, all collaborations are working and fully active.

Finances
The current study is sponsored by the Division of Infectious Diseases and Hospital Epidemiology, University Hospital Zurich. Research expenses are covered by a variety of project specific separate research grants (Swiss National Science Foundation and private foundations).

Publications
If, at the end or during the trial, publications or presentations of the results of this trial are planned then all parties involved in the respective research work have the right to review the manuscript and abstracts before their submission. However, decision for submission of a manuscript or abstract is made by the principal investigator.