Swiss-Spanish Strategic Interruption Study in Chronic Infection in Group Well Controlled Virologically:

only 17% were able to control HIV <5000 copies/ml after several interruptions\; 80 patients have had 4 interruptions and median viral load during interruptions has not declined. But additional studies will be looking at interruptions.

Bernard Hirschel from the Division of Infectious Diseases at University Hospital in Geneva Switzerland reported again on his study of Strategic Treatment Interruptions. He first reported on this study in Durban this past Summer. This is the largest study to date on STIs and was received by many researchers in Durban as strongly suggesting that the idea that STIs in people with chronic infection would not stimulate the hoped for immune response that might control HIV off therapy. This study is in a unique population of people: study participants had viral load <50 copies/ml, so they had good viral control, and he feels they were compliant when on therapy in study after interruptions. In addition, participants were treatment-naÔve prior to the therapy they were on at the beginning of this study, so they should not have pre-existing resistance. This is in contrast to interruptions in people treated during acute infection or shortly after infection, as studied by Eric Rosenberg and mentioned below.

A main hope for the goal of STIs is that repeated interruptions would stimulate an immune response by exposing the person's immune system to HIV antigen. After Hirschel reported his findings in Durban, hopes were dashed about this goal. I had a talk with Dr Hirschell here in Glasgow. He feels that his study shows some people were able to control HIV off therapy after repeated interruptions but many were unable to. Whether this control by a few was random or if there are discernible reasons remains unknown. Except for 1-2 persons in whom resistance developed all others were able to regain virologic control after going back on therapy following interruption. At first he said to me that he thinks the interruptions were safe, meaning that except for 1 or 2 people everyone was able to regain viral control. However, I suggested to him what about the possibility of resistance developing over the longer term, or the longer term consequences of reseeding reservoirs in which viral load had been severely reduced. He admitted that these could be risks that we as of yet we have not been able to assess adequately. He also said that if a person had pre-existing resistance from treatment prior to being undetectable on their current regimen, then interruptions can be risky in that resistance could emerge and it may be more difficult to regain virologic control.

There is some degree of controversy about the safety of taking "drug holidays" or treatment interruptions. Some doctors feel that if a person has reached the bitter end and cannot take medication anymore- often called adherence fatigue- then interrupting medication is acceptable. They also feel it's not too risky, depending on your CD4 and pre-therapy viral load. After stopping therapy, a person's CD4 count may decline fairly substantially. A decline in CD4 count of 30% is expected, so if CD4 count is 600 it could decline to 400. Once CD4s decline to about 200, PCP prophylaxis should be taken.  I think most doctors will strongly discourage you from taking such a break, and feel interruption is risky. The risks are unknown: resistance, unable to regain viral control, implications of reseeding reservoirs. In Durban a number of doctors and researchers warned against the unknown outcome from taking interruption. Having said all this here is the data Hirschel reported at Glasgow. Personally, I think an interruption is risky whether you are undetectable and virally well under control or not. Hirschel also expressed to me the hope that immune therapy such as a vaccine may help in controlling HIV off drugs. Several studies will be exploring this but so far this is just a hypothesis.

Hirschel described the types of interruptions. Patients who started HAART during primary HIV infection (acute infection, very early after infection). The goal in this circumstance is low viremia or undetectable without HIV drugs. A recent published letter to the journal Nature  from Eric Rosenberg (Rosenberg et al. Nature 2000; Vol. 407, Sept. 28 issue) reported on a study in which 5 of 8 individuals had low viremia (<500 copies/ml) after repeated interruptions following initiation of therapy in acute infection. The partial control in HIV seen in this study was seen along with increases in HIV-specific CTL immune responses. It is believed that such an immune response is what is sought to control HIV. Opinions are mixed about this data. The study was very small and some feel the results don't prove much. The cause-effect relationship is far from proven. There is no randomization, and no before & after comparison. Others feel the results may be reliable. But the same individuals think that after these individuals are off therapy for a while they will lose control of HIV and have to restart therapy.

A second type of interruption is during chronic infection for people with detectable viral load ("failing therapy"). The goal in this case is avoiding toxicity, freedom from restrictive adherence ("drug holiday"), and saving costs. There is also the hope of recovery of wild type virus thinking that after wild type virus returns a person may respond better to the next regimen. However, Veronica Miller and her German colleagues reported this year at the Salvage Therapy Workshop that her retrospective analysis of patients who took such a drug holiday was not successful. Median CD4s declined from 207 to 93, 17 new OIs developed, and viral load increased. It took some time for CD4s to return. And its been reported by several researchers that although resistance may not be seen by conventional resistance testing, by using senstive testing methods resistance can be found.

A third type of interruption, which is the subject of Hirschel's study, is where patients with effective HAART during chronic infection interrupt therapy to achieve low viremia without drugs. They want to avoid toxicity and improve quality of life.

Hirschel's study is called SSITT (Swiss-Spanish Intermittent Therapy Trial). In this study patients were on combination ART, with undetectable VL for >6 months and VL <50 copies/ml at screening. No changes in treatment because of virologic failure while on HAART. CD4 count >300 at screening. Not on NNRTIs, and no treatment prior to HAART. The study consists of 4 cycles of 2 weeks off therapy followed by 8 weeks back on therapy. After 40 weeks treatment is stopped. The endpoints of this study are--how many patients lose their indication for therapy (defined as VL <5000) from week 52 to 92, after 12 to 52 weeks off therapy.

Patients in study had duration of HAART at study entry for median 26 months (8 to 44.5 months), and undetectable viremia for median of 21 months (6-43 months). At the start of SSITT they had a median CD$ of 727 (316-1600), median CD8 830  (198-2700), and median CD4/CD8 ratio of 0.87 (0.21-3.89). Their viral load was well controlled as 99/128 had viral load <20 copies/ml, and 29 had 20-49 copies/ml.

Hirschel reported results so far in study. As you can see in table, median viral load did not go down during interruptions.

Viral Load After 2 weeks Treatment Interruptions

Clearly, sequential interruptions did not result in VL decline. As well, 2.5% to 12.5% had VLs >5 log (100,000 copies/ml) during interruption.

No rebound:

High rebound:

No significant differences.

Hirschel said failure did not look like it was due to resistance. Patients were able to regain viral control (<50 copies/ml) after treating for more than nine weeks following interruption. I asked Hirschel if anyone was unable to regain viral control and he said no except for 2 patients. One patient was compliant with 3TC, NFV and AZT, and failed at week 9 with a VL of 141 copies/ml. He continued initial treatment and after 23 additional weeks of therapy VL was 2730 copies/ml. Multiple resistance mutations were present (184 [RT], 46, 54, 63, 82A [protease]). He switched to RTV, SQV, EFV, Abacavir, and d4T and VL was <10 copies/ml after 6 weeks. A second patient had resistance but was found to have been previously treated with AZT prior to HAART.

Failure Upon Retreatment After Interruption

The protocol defined failure before week 40 as unable to regain viral control (<50 copies/ml) within 9 weeks of restarting therapy, but Hirschel said in essence all were able to regain control with longer treatment. The most important risk factor was a high VL at rebound (RH of failure is 3.5 per log increase in VL at rebound). A total of 24 patients were failures.

54 patients have now reached week 52. 9 patients (17%, CI 8-30%) have viral load <5000 copies/ml, and 19 patients have VL >5000 copies/ml. Obviously, there is no clear trend, and few patients have HIV under control. 17 stopped before week 40 mostly because their viral load did not drop quickly enough to <50 copies/ml after 9 weeks restarting treatment following interruption (n=12); 4 due to protocol violations, and 1 with ARS (acute retroviral syndrome) during first interruption. 9 restarted treatment between week 40 & 52 mostly because of high viral load (n=8), and 1 because of low CD4 count.

Hirschel told me that he will be conducting immunology testing to try to discover if there are differences between those with less than or greater than 5000 copies/ml. This is to see if its possible to predict who may respond to interruptions by controlling HIV without drugs. There were several patients several months after stopping treatment (after week 40) who had low viral load („1000 copies/ml) but pre-ART had higher viral load.