Report 5 -  4th International Workshop on HIV Drug Resistance and Treatment Strategies
Written by Jules Levin
Sitges, Spain, June 12-16 2000

Structured Therapy Interruptions

This article discusses several studies presented at the Resistance Workshop and others presented at the recent Salvage Therapy Workshop. The first article discusses Felipe Garcia's pilot open-label study of 10 individuals who have 3 STIs. Although CD4s can decline during the STI, some individuals appear to have control of HIV 8 months after interruption. He also discusses a study using hydroxyurea in STI.  Jan Van Lunzen reported at Sitges on the CD4 dysfunction that can occur after one therapy interruption in individuals with well suppressed HIV on HAART with or without IL-2. The third article below discusses a report at the Workshop on detecting resistance with very sensitive testing method when conventional genotyping assay was unable to detect resistance in individuals in STI following HAART failure. This highlights a limitation of conventional resistance testing.

Continuous and spontaneous control of plasma viral load after 8 months without therapy and 3 cycles of therapy interruption in chronic HIV-1; correlation with induced cytotoxic (CTL) and helper HIV-1 specific immune responses

Felipe Garcia reported in an oral presentation for a Spanish research group which included Doug Nixon from the Aaron Diamond Research Center in NYC, on a small open-label non-randomized study of 10 individuals and three cycles of a structured therapy interruption. The objective of the study was to analyze the dynamics of viral load rebound (VL) and HIV- specific immune responses after three consecutive cycles of structured therapy interruption (STI). And, to assess the safety of this approach.

Ten antiretroviral naÔve patients with chronic HIV-1were recruited from the Spanish EARTH-1 Study . Their baseline CD4 was >500, and VL was >10,000 copies/ml (range 14,700-504,000). They were treated with d4T, 3TC, and ritonavir or indinavir for 52 weeks.  After one year of HAART all had a plasma VL <20 copies/ml and agreed to be enrolled in this STI pilot study. All had been <20 c/ml for >32 weeks. One patient was lost for follow-up after the first stop and another patient became pregnant after the second stop.

At baseline, VL was required to be <20 c/ml and therapy interruption was for 4 weeks or until VL increased to >200 c/ml. Everyone was reintroduced to therapy for 6 months. At week 28 a second therapy interruption was started if VL was <20 c/ml. Therapy was reintroduced 1 month after VL >200 c/ml if VL did not drop spontaneously. Therapy was reintroduced and maintained for 6 months and a third interruption was introduced if VL was <20 c/ml for a third and last time. This time therapy was only reintroduced when VL was about 10,000 c/ml or greater. At weeks 0, 28, and 96 (first, second, and third stop, respectively), plasma VL was <20 c/ml in all cases and below 5 c/ml in 7 of 10 cases (first stop), and in 7 of 9 cases after the second stop and third stop (one patient was lost to follow-up). A rebound in VL was detected in all cases with a mean (SE) doubling time (DT) of 2.23 (0.32) in the first stop, 3.38 (1) days in the second stop and 3.25 (0.38) in the third stop (p=0.05, for the comparison between DT in first versus third stop). In two patients, DT increased from 2.48 to 8.66 and from 3.85 to 8.66 days, respectively. 

At the second stop, in 4 of the 9 patients, VL rebounded to similar levels as baseline (week -52) and dropped spontaneously thereafter (0.8, 0.8, 1.3, and 2.09 log copies/ml, respectively). These 4 patients developed strong and broad HIV-1 specific CTL responses and a strong CD4 lymphocyte proliferative response to HIV-1 antigens.

After the third stop, a rebound in plasma viral load was detected in all cases. Six of the 8 patients had a VL set-point significantly lower than baseline (from 0.5 to 1.7 log). And, 4 of 8 patients had VL < 10,000 c/ml (range 650 to 9,000) after 8 months off therapy. Recovering of specific CTL and CD4 lymphocyte response was detected in 5 of the 8 patients. After the first, second and third stops, genotypic or phenotypic resistance to reverse transcriptase or pritease inhibitors was not detected.

HIV- Specific Immune Responses at Stop 3

At baseline while on HAART, none of the 8 patients had a CD4+ proliferative response against HIV-1 p24 antigen, nor did they have a CTL HIV-1 specific response. During the third interruption, 6 of 8 had a CD4+ proliferative response against HIV-1 p24 antigen and 5 of 8 had a CTL HIV-1 specific response. Garcia reported that HIV-1 specific immune response (both CTL and CD4 responses) correlates with spontaneous control of viremia after stops 2 and 3.

Garcia described two examples of patients in the study--one who responded to the STI and the other who did not. Subject #2 had a baseline VL of 27, 300 c/ml. After the first interruption VL increased to 3000 c/ml. After the second interruption VL rebounded to similar level as baseline VL but VL dropped spontaneously to 200 c/ml. Therapy was reintroduced, and after the third stop VL increased to 6000 c/ml, but spontaneously dropped to 500 c/ml and has maintained this for 8 months. In subject #12, STI was not successful. At baseline on HAART VL was about 27,000 c/ml. After the first interruption VL rebounded to 300,000 c/ml, but went to undetectable after therapy was reintroduced. After the second stop VL increased to 70,000 c/ml and again was reduced to undetectable after therapy was reintroduced. After the third interruption VL rebounded to 40,000 c/ml and remained between 10,000 and 40,000 during the 8 months of follow-up without antiretroviral therapy.

Safety

After the first, second and third stops, genotypic and phenotypic resistance to reverse transcriptase or protease inhibitors were not detected. A quick virological response was observed after reintroduction of the same antiretroviral treatment after the first and second stops, and in the two patients who restarted ART after the third stop. CD4 T lymphocytes dropped significantly after the first, second and third stops to a level similar to the level before starting ART.

Garcia finished his talk with an additional discussion of a different cohort of 5 patients who were recruited in response to individual requests and had similar baseline characteristics, STI schedule, and data consistent with the experiences of the first cohort reported in this presentation. Five chronically infected individuals had CD4s at baseline of >500 c/ml and >10,000 c/ml (range 34,000 to 130,000). They had HAART (d4T+3TC+indinavir) for 52 weeks and then 3 cycles of STI. In 1 of the 5 patients VL has not rebounded after 6 months off therapy following 3 cycles of STI. A rebound in VL was observed in 4/5 after the last interruption of therapy. In 2 of the 5 patients, after 6 months off therapy VL set-point was <0.5 log below baseline VL (-1.7 to 3.84 log). And, 2 of 5 had VL <10,000 c/ml (<5 and 3470).

Garcia concluded by saying that these data could be proof of concept about the likelihood of inducing an effective and sustained specific immune response against HIV-1 antigens in chronic HIV-1 infected patients.

Commentary

This study was conducted in a unique population of individuals, who were in relatively early disease with relatively high CD4s and who had VL <20 copies/ml for 12 months. Several points, which I think are important, were raised following Garcia's talk in the session. Randomized well-controlled studies are indicated to explore STIs in this setting. It was mentioned that there is data from a cohort in which individuals had spontaneous viral load reductions while not on therapy. Are the VL reductions seen in these studies after the third interruption due to immune responses caused by therapy interruption or spontaneous reductions as seen in the cohort mentioned?  

Hydroxyurea and STI

In a poster abstract Garcia reported on a study investigating whether the association of hydroxyurea (HU) with STI could prevent VL rebound after cyclic interruption of ART and/or help to recover the specific immune response. 18 chronic HIV-1 infected patients treated with d4T/3TC+indinavir for 52 weeks and VL< 20 c/ml for at least 32 weeks were randomized to receive d4T/ddI/IDV/HU (n=9) versus d4T/ddIIDV (n=9) during 6 months. Thereafter, 3 consecutive STI cycles were undertaken, separated by periods of two months with the same triple ART (which was reintroduced when VL increased >200 c/ml in the first stop and after two weeks in the second and third stops). Plasma VL, genotypic resistance, lymphocyte immunophenotyping, CD4 lymphocyte proliferative responses to mitogens and to HIV-1 antigens and HIV-1 specific cytotoxic T lymphocyte (CTL) responses were assessed.

At day 0 of the first, second and third stops, plasma VL was <20 c/ml in all cases, and <5 c/ml in 6 of 9 in the HU group and in 5 of 9 in the ART group. A rebound in plasma VL was detected in all cases after the first stop with a mean (SE) doubling time of 2.08 (0.38) days in the HU group, and 2.67 (0.8) days in the ART group (p=0.52). The second and third rebound (evaluated at week 2 after stopping ART) was similar in magnitude to the first rebound (Wilcoxon test, p=0.54), both in HU and ART groups. 5 of 9 in the ART group and 5 of 9 in the HU group developed strong and broad HIV-1 specific CTL responses and CD4 lymphocyte proliferative response to HIV-1 antigens after the second and third stops (which were not present at baseline or after the first stop). After the first, second and third stops, mutations associated with resistance to RT or protease were not detected. VL dropped <20 c/ml in al patients at week 8 after reintroduction of the same ART after all 3 stops. 

Garcia concluded that use of HU with STI neither prevents VL rebound nor changes VL rebound dynamics after interruption of ART. HU also was not effective in increasing the control of viral replication after 3 interruptions of therapy, nor in inducing a higher proportion of specific immune responses against HIV-1 antigens.

Rapid rebound of viral replication and T cell dysfunction after stopping HAART when viral load is well suppressed with and without IL-2

Jan van Lunzen and a German research group reported on a study of the kinetics of viral rebound in connection with parameters of T cell proliferation, activation, and functionality after stopping therapy. Therapy was stopped in all patients from a randomized, controlled trial who discontinued HAART with (6 patients) or without (5 patients) IL-2 after having achieved plasma viremia <25 copies/ml for 13-23 months. The patients were enrolled in the COSMIC trial (n=56) where they had baseline CD4s of 480 and viral load of 4.4 log (25,000 copies/ml). All patients showed T cell subsets in the normal range compared to age-matched normal controls and rare or no productive HIV replication in lymphoid tissue as well as viral load below the limit of detection in the CSF at the time of treatment interruption.

Viral load was frequently measured by the Ultrasensitive Amplicor assay. T cell proliferation and activation were determined by 4-color flow cytometry using a broad panel of Mabs (e.g. Ki-67, HLA-DR, CD38, CD25, FAS). Functional assays using intracellular cytokine expression (e.g. IL-2, IFN-y, TNF-a) were performed after stimulation with mitogens/SEB and recall antigens (e.g. CMV, TT). Tetramer staining directed against MHC class I and II molecules is being performed at present.

All patients showed a rapid rebound of VL after 4 to 8 weeks to baseline with a slight "overshoot" at week 4 despite prolonged control of viremia and normalization of peripheral T cell subsets. This was paralleled by a reincrease of proliferation and expression of activation markers on CD4 and CD8 cells (CD8>CD4). The kinetics of viral rebound and increase of T cell proliferation and activation as well as inversion of CD4/CD8 ratio were closely correlated with a slightly more rapid "response" in the CD8 T cell compartment. T cell responses to mitogen/SEB and recall antigens in vitro decreased over time indicating recurrence of functional defects. No influence of IL-2 therapy in the study was seen. There was no difference in responses between those receiving IL-2 or not receiving it. I think van Lunzen said T cell responses in vitro are reduced during peak VL rebound and increase again during steady state (5/6).

The authors concluded that rapid viral rebound of viral replication after stopping therapy is associated with inversion of CD4/CD8 ratio due to increased T cell turnover and activation. Functional defects reappear in parallel despite profound and sustained suppression of viral replication while on HAART.

Detection of Resistance During Therapy Interruption After HAART Failure; Difficulty in Detecting Low Level Minority Resistant Virus

Therapy interruptions are controversial even in situations where plasma viral load is well suppressed, but recently new information shed further light on the risks of therapy interruptions following HAART failure. Veronica Miller reported at the recent Salvage Therapy Workshop that during interruption for individuals who were on failing regimens, there were a number of new opportunistic infections, viral load rebounded in the majority of these patients, and CD4s decreased and it took a while to recover the losses. See the Salvage Therapy Workshop Reports on the NATAP web site for more details on her study. Although it was previously reported last year by Miller that resistance was not detectable following therapy interruption in this group, several researchers including Steve Deeks at the Retrovirus Conference this past February have reported that resistant virus can be found. See the Retrovirus Conference Reports on the NATAP web site to read Deeks' report.

At this meeting, A Hance and a French research group reported that small minorities of resistant virus were found in patients undergoing STI following HAART failure at times when conventional genotyping methods detect pure wild-type virus in plasma.  Conventional resistance testing assays can only detect virus present in 10-20% of total virus populations. This is a limitation of currently available resistance assays. Some may question the clinical utility of detecting virus at such low levels, but others will say that such low level resistance may soon grow out and emerge to become detectable by conventional assays. Using an Ultrasensitive PCR real-time detection method that relies on selective amplification and quantification of viruses carrying the V82A and L90M mutations in the protease enzyme, resistant viral quasispecies representing <0.1% (mutation L90M) and 1% (mutation V82A) of the total virus population could be detected. Hance reported that the accuracy of the results were fully validated by analysis of clones derived from parallel PCR reactions. In 4 patients who underwent STI, apparently complete reappearance of wild-type virus by 2-3 months was detected using conventional genotyping. When evaluated using the PCR procedure, resistant virus could still be detected for as long as 5 months after interruption of therapy, and was undetectable at 3 months in only 2/4 patients.

When both mutations (V82A and L90M) were initially present, the loss of viruses expressing V82A and L90M always occurred in parallel. To further evaluate whether the commonly used STI duration of 3 months was sufficient for complete washout of resistant virus, Hance and his research group quantified the persistance of the V82A and L90M mutations in a series of patients who had undergone STI for 3 months, and for whom conventional genotyping had revealed an apparent total reversion to wild-type virus. In the patients with L90M and/or V82A at baseline, resistant virus was still present at 3 months in 9/12 cases. No viruses with V82A or L90M were detected in patients who did not harbor the corresponding mutation at baseline.

  In a situation like that described above by Garcia and van Lunzen, where viral load is well under control, a doctor/researcher at this meeting told me that he will permit an interruption if the patient insists because they may be tired of taking the medications and therapy may have to continue for years.  Since a person may experience CD4 decline during therapy interruption, which could be as high as a 30% decrease from baseline as this doctor said he observes in his clinical practice, it is important to factor this in if a therapy interruption is undertaken. So, if a person has 300 CD4s at the time of therapy interruption a decline of 30% would leave about 200 CD4s. This may be a risky venture if CD4s in fact decline that low.