HIV RNA in plasma rebounds within days during structured treatment interruptions

Abstract

Objective: 

To evaluate time to viral rebound in patients undergoing repeated structured treatment interruptions (STI).

Method: 

Fourteen chronically HIV-infected patients enrolled in the Swiss–Spanish Intermittent Treatment Trial (SSITT) underwent frequent blood sampling. Patients underwent four cycles of 2-week STI, followed by 8-week retreatment with the identical antiretroviral treatment (HAART) used before STI. At the fifth cycle, treatment was stopped for a longer period. Before each new STI, plasma viral load (VL) had to reach < 50 copies/ml. VL was measured during day 0 (last day on HAART) and on days 4, 8 and 14 during all five STI.

Results: 

During the first cycle, plasma HIV RNA increased to > 50 copies/ml (range, 67–88) in five patients at day 4, in eight patients (> 100 copies/ml) at day 8 and in 12 patients (> 100 copies/ml) at day 14. Cumulative analysis of the frequency of detectable HIV RNA at days 4, 8 and 14 compared with day 0 for all five cycles revealed nine patients with VL > 50 copies/ml [13 of 54 samples tested (24.1%); P = 0.14] at day 4, 11 patients [33 of 58 samples tested (56.9%); P < 0.0001] at day 8 and 12 patients [53 of 65 samples tested (81.5%); P < 0.0001] at day 14.

Conclusions: 

Significant viral replication can be induced during 1 week STI, and this may increase the risk of the emergence of drug resistance during long-term cycling. Therefore, short-term cycling strategies such as 1-week-on, 1-week-off treatment, although conceptually intriguing, should still be regarded as investigational and should be restricted to rigorously controlled clinical trials ideally involving patients who have never failed treatment before.

Introduction

Structured treatment interruptions (STI) are investigated in clinical trials for several reasons: to test whether protective immune responses can be induced, to limit drug exposure in order to minimize toxicity and reduce costs and to test whether response to salvage regimens in patients with multiple treatment failure can be increased.

A number of clinical trials using highly diverse protocols of STI have been conducted in patients who have initiated their first highly active antiretroviral treatment (HAART) either during primary [¹] or during chronic HIV infection [^2–12]. In these studies, the protocols used have varied widely with regard to the number and the duration of STI. However, a general finding is that virus rebounded in basically all patients. A recent study reported that, in patients with undetectable plasma viral loads (VL) after HAART, repeated cycles of 1-week-on followed by 1-week-off HAART did not lead to rebound of VL during the drug-free period in any of the eight subjects investigated [¹³]. Since HIV replication did not resume within 1 week after treatment cessation, emergence of drug-resistant virus during this type of STI was deemed as unlikely [¹³].

The present study examines the viral rebound in 2-week STI periods.

Methods

Fourteen patients enrolled in the Swiss–Spanish Intermittent Treatment Trial (SSITT) [¹²] underwent frequent sampling during STI [¹⁴]. Patients had been taking HAART for ≥ 6 months with varying regimens: seven patients (patients 102, 104, 105, 112, 116, 118, 120) were taking zidovudine plus lamivudine plus indinavir; one patient (patient 130) was taking zidovudine plus lamivudine plus nelfinavir; one patient (patient 109) was taking zidovudine plus lamivudine plus ritonavir; two patients (patients 121, 127) were taking stavudine plus lamivudine plus nelfinavir; two patients (patient 111, 125) were taking stavudine plus didanosine plus nelfinavir; and one patient (patient 128) was taking zidovudine plus didanosine plus nelfinavir. All the patients had had no breakthrough of viraemia, had undetectable VL (HIV RNA < 50 copies/ml) for 11 to 32 months (mean 25 months), and a CD4 cell count of ≥ 300 × 10⁶ cells/l (mean, 739; range, 347–1269). Extensive baseline characteristics have been reported elswere [¹⁴]. Patients underwent four cycles of 2-week STI, followed by 8-weeks of retreatment with the identical drug combination used before STI. At the fifth cycle, treatment was stopped for a longer period [¹²]. VL in plasma was measured by the ultrasensitive Amplicor Monitor test version 1.5 (Roche Diagnostics, Rotkreuz, Switzerland) at baseline (day 0; last day on HAART) and on days 4, 8 and 14 during all five STI cycles. To minimize assay variability for each patient, batchwise HIV RNA testing was performed on frozen (−80°C) plasma samples for each complete STI cycle. None of the patients enrolled in the present study had received immunomodulators at any time before HAART, while taking HAART or during STI.

Written informed consent was obtained from all patients according to the guidelines of the Ethics Committee of the University Hospital Zurich.

Results

Immediately before the first STI, all 14 patients had VL of < 50 copies/ml. During this first cycle, plasma HIV RNA rose to > 50 copies/ml in five patients (range, 67–88) at day 4, in eight patients (> 100 copies/ml) at day 8 and in 12 patients (> 100 copies/ml) at day 14 (Fig. 1a). The cumulative frequencies of detectable HIV RNA from day 0 (baseline of each STI cycle) to days 4, 8 and 14 were analysed for all patients during the five cycles (Fig. 1a). VL was ≥ 50 copies/ml at day 4 in nine patients [13 of 54 samples tested (24.1%); P = 0.14, chi-square test probing disparity to day 0], at day 8 in 11 patients [33 of 58 samples tested (56.9%); P < 0.0001] and at day 14 in 12 patients [53 of 65 samples tested (81.5%); P < 0.0001]. Seven patients had VL ≥ 50 copies/ml at the baseline of one or more cycles [8 of 66 time points analysed (12.1%)]. A continuous increase of viraemia over cycle baseline values was observed (Fig. 2). The increase in mean VL at days 8 and 14 compared with that at cycle baseline was highly significant (Fig. 1b). Similarly, the mean relative increases of VL at these time points were highly significant (Fig. 2).

Fig. 1.:   Synopsis of plasma HIV RNA levels detected at day 0, 4, 8 and 14 of structured treatment interruptions (STI). (a) HIV RNA levels stratified by patient and with regards to amounts of HIV RNA detected for each sample at a given time point. (b) HIV RNA levels shown separately. Horizontal lines represent mean values. P values show results of unpaired t-tests. Patient 104 was excluded (e) from the study because he did not reach HIV RNA < 50 copies/ml after the first reinitiation of treatment. Genotypic resistance testing 23 weeks later, at a viral load of 2730 copies/ml revealed mutations associated with reverse transcriptase (184V) and protease inhibitor resistance (46L, 54V, 63P, 82A). Successful salvage treatment with abacavir, stavudine, efavirenz, and lopinavir/ritonavir was started. The patient was treated for depression during the first year while taking HAART and admitted to a period of non-adherence during that time. As determined retrospectively, the 184V mutation was present at high frequency at the rebound of week 2 [¹⁶]. Consequently, the emergence of drug resistance most likely was not induced by STI in this patient. Some data are not available (na). HIV RNA was transiently detected at low levels in nine patients at baseline (on treatment) of STI cycles 2–5. However, the study protocol was not violated because the decision to stop treatment for each STI cycle was made based on the VL measured at 1 week before treatment was interrupted. All these VL were retrospectively tested batchwise to obtain best possible quality of data.

Fig. 2.:   Relative increases in HIV RNA levels at days 4, 8 and 14 of structured treatment interruptions (STI). Log ratios were calculated by dividing the viral load measured at days 4, 8 and 14 by the on-treatment values (day 0) for the same patient in the corresponding STI cycle. If day 0 VL were nondetectable, the median detection limit of all undetectable VL (=7) was chosen. Horizontal lines represent mean values of the log ratios of relative HIV RNA increases. P values show results of one sample t-tests against log₁₀ copies/ml values (1 = no change).

Discussion

These findings clearly show that viral rebound can occur within 8 days of STI in a substantial proportion of patients (11 of 14 in our study). This rapid rebound of viraemia despite prolonged suppression of plasma viraemia before STI suggests that activation of the latent viral reservoir and/or upregulation of residual viral replication under therapy occurs frequently. Our observations raise concerns that the risk of drug resistance selection during extended periods of cycling may be significantly higher than previously reported [¹³]. Martinez-Picado et al. [¹⁵] have reported the emergence of the 184V mutation in two patients undergoing STI in a trial with a different STI schedule. From the 14 patients examined in the present study, we could detect minor variants of the M184V mutation in eight patients at least once during the five STI cycles (K. Metzner, personal communication; [¹⁶]). However, whether these minor populations harbouring the 184V mutation will be of clinical relevance remains to be determined.

There are several possible explanations for the differences observed in our study and the trial by Dybul et al. [¹³]. During the latter study, all patients were treated with stavudine plus lamivudine plus indinavir combined with ritonavir boosting, whereas in our study boosted protease inhibitor regimens were not included. Consequently, differences in drug potency may have had an impact on the outcome. Possibly the lower VL levels detected in the study by Dybul et al. [¹³] might be a result of the less-sensitive branched-chain DNA assay used. In general, assessing small numbers of patients can give rise to unrecognized confounding factors, which might have impacted on both studies. It is unlikely that the divergent findings are a consequence of the different sampling time points analysed. The doubling time of viral rebound in our patients was calculated to have a mean of 2.3 days [¹⁴]. Therefore, the delay of sampling by 1 day in our patients (day 8 compared with day 7 after STI by Dybul et al. [¹³]) cannot account for the observed differences. Moreover, the noted trend towards higher VL was already present in our study at day 4 (Figs 1 and 2). Potentially, in our study, the 2-week interruptions might have resulted in a greater replenishment of the latent reservoir than achieved by 1-week STI and thus caused more rapid rebound kinetics during subsequent STI. However, this seems unlikely since we observed an identical rapid rebound even during the first cycle of STI.

Other potential reasons for the different rebound kinetics observed in the patients studied by Dybul et al. [¹³] and here could be differing magnitudes of cellular and humoral immune responses in the respective cohorts. An extensive analysis of the interaction of HIV-specific immune responses and control of viraemia in the patients in the current study indicated that, although there was strong induction of HIV-specific CD8 immune responses during STI, these cytotoxic T lymphocyte responses could not predict viral growth and clearance rates during STI [^14,17]. This would suggest that factors other than HIV-specific cytotoxic T lymphocyte cause differences in viral rebound rates in chronically HIV-infected patients undergoing STI.

Finally, none of our patients had received immunomodulatory therapy before undergoing STI, whereas 7 of the 10 patients studied by Dybul et al. [¹³] had received interleukin-2 before undergoing STI. However, a recently published, carefully conducted prospective clinical trial of viral rebound in the latent reservoir or residual viral HIV RNA in lymph nodes after cessation of therapy with interleukin-2 in conjunction with HAART failed to find a lower frequency of viral rebound in patients treated with interleukin-2 and HAART compared with patients treated with HAART alone [¹⁰]. This would suggest that previous interleukin-2 treatment was not the factor responsible for lack of viral rebound in the study by Dybul et al. [¹³].

Larger clinical trials will be required to assess whether parameters such as drug regimen, immunmodulatory substances, levels of proviral HIV DNA, pretreatment plasma VL, cell-associated HIV RNA or immunological parameters, such as cellular or humoral immune responses, may predict the time point and magnitude of viral rebound and whether a group of patients with slow viral rebound kinetics can be defined.

In conclusion, our findings clearly demonstrate that significant viral replication can be induced during 1 week STI, which may increase the risk of the development of drug resistance during long-term cycling. Consequently, this therapy regimen still bears considerable risks and should not be considered as a general treatment option at the present time. Although the STI strategy with 1-week-on 1-week-off cycling is conceptually intriguing, a safe application of this strategy can only be performed in rigorously controlled clinical trials and in patients who have never failed treatment.

Acknowledgments

We are grateful to all patients for participating in this study and for their contribution to the better understanding of HIV pathogenesis. Moreover, we thank Friederike Burgener, Erika Schlaepfer-Nadal, Esther Beerli and Herbert Kuster for high-quality processing of clinical samples, Catherine Fagard and Michelle Lebraz for study coordination and Joseph Wong and Bruno Ledergerber for stimulating discussions and reviewing of the manuscript.

Sponsorship: This study has been financed in the framework of the Swiss HIV Cohort Study, supported by the Swiss National Science Foundation. (Grant 3345-062041), and was also supported by SHCS grant 290 (HFG), by SNF grant 3345-65168.01 (HFG, AT) and by a research grant from the Kanton of Zürich.