Report from IDSA, Infectious Disease Society of America, Philadelphia , Nov 21

Structured Therapy Interruption
Jules Levin, NATAP

Eric Rosenberg reported preliminary data at IDSA on Sunday morning on 2 individuals who started HAART during acute infection as part of his original 25-person cohort (see backgrounder below), and who interrupted therapy. These two are the first to interrupt therapy 2 or more times. An additional 7 individuals are in earlier process of interrupting therapy. They may be during their first interruption now or shortly after that point.

Rosenberg reported that after multiple interruptions there was partial and improved but not complete control of viral replication during therapy interruption. It appears viral control may be associated with an improved immune response as measured by increases in HIV specific CD4s and improved CTL responses in both individuals. New data was reported on CTL responses and viral load data during therapy interruption. Additional studies are ongoing and planned by a number of researchers looking at therapy interruption in chronic infection, and with various combinations of immune-based therapies.

Subject One. Subject one interrupted therapy 3 times and was an interesting case. The first interruption was planned and it took about 3 weeks for virus to begin to replicate, whereupon it went up to 17,000. After restarting therapy (protocol mandated restarting therapy at 10,000) viral load promptly went back down to undetectable and stayed there. Due to acute hepatitis A the person had an unplanned second interruption. Again, it took 2 weeks for virus to return whereupon virus shot up to 37,000. He was still too ill to restart therapy and within 72 hours fell to 6,000. 48 hours later his viral load was 400. Over the next 2-3 weeks slowly viral load started creeping up (it was still below 10,000) and therapy was restarted because it wasn't a planned interruption. Viral load rapidly declined to below detection after restarting therapy.

Rosenberg used a stimulation index (SI) to measure HIV specific CD4 response. A stimulation index above 10 is considered significant. Subject one had a strong SI of 80 prior to first interruption. After second interruption he peaked with a 100 SI. But after second unplanned interruption SI went to over 800, which was one of the highest indexes ever seen. They also looked at CTL responses to measure immune response. During first year on therapy the person did generate significant CTL responses—measured in breadth (number of epitopes responded to) and magnitude. After the first interruption while he was off therapy they got stronger in magnitude. After the second interruption they got even stronger.

After interrupting therapy for the third time, this person had a brief blip up to 12,000 copies/ml, and then promptly fell to <5,000 copies/ml for 2 months. He had a second blip in his viral load up to about 12,000, but they think it's because he was experiencing an infection. His therapy fell almost uniformly to less than 5,000. He went for a period of 4 months and at a viral load of 3,600 reelected to restart therapy.

Subject 2. The second person had two interruptions. The first interruption lasted 2-3 weeks during which he had low viral load but then it rebounded to 116,000 copies/ml. Therapy was re-started and viral load was kept below limits of detection. When therapy was discontinued a second time the person went 4 months consistently with viral load below 5,000 copies/ml. Despite being below 5,000 the person elected to restart therapy at the 4-month period “to try for a better result”.

At baseline subject 2's SI was about 20 but after first interruption there was a boost in SI to over 120. During his first year while on therapy prior to interruption he had small CTL responses (recognized 3 epitopes). During the first interruption the CTL response increased in magnitude a little. After restarting therapy and during the following second interruption, there was a “clear-cut” boost in the level of CTL responses as measured by both the breadth (where he now recognizes 6 epitopes) and “clearly” in the magnitude of response.

Both individuals experienced partial control of viral replication but not complete control, as viral load was detectable although low.

Cautionary Editor's Note:
Rosenberg cautioned that it is premature to try therapy interruptions outside of controlled human studies like the ones he and others are conducting. This concept of therapy interruption is not well understood. Rosenberg's study is in individuals treated with HAART during acute infection. People treated during chronic infection may not be as responsive. After acute infection a person's ability to mount an HIV-specific CD4 response appears to wane. For the few individuals reported at conferences who appear to have short term benefit reflected by partial control of viral replication without therapy, many individuals do not get this benefit. So far there is no measurable way to identify who may respond well to interruptions. We don't yet know long-term effects of therapy interruption. Although it appears in these studies that viral load declines back to undetectable after restarting therapy following an interruption, are there long-term harmful consequences? We don't know yet. Although drug resistance has not been observed during the interruptions, will it emerge after repeated interruptions or over time? What are the long-term consequences of exposing reservoirs, such as lymph tissue, to virus after reducing virus levels with HAART)? A number of researchers are conducting studies to understand therapy interruption. Although this study and others are in individuals treated with HAART during acute infection, studies in individuals with chronic infection are planned. In addition, interruption studies for individuals taking HAART and Remune are also ongoing.

Background given by Rosenberg at IDSA:
CD4+ T-helper cell responses are critical for the maintenance of cellular immunity. Virus specific T-helper responses are typically weak or absent in progressive HIV infection. Robust HIV specific T-helper cell responses are observed in individuals controlling virus without therapy (a small select group called long term non-progressors). Rosenberg said recently we've shown that HIV-specific T-helper cell responses are inversely correlated with viral load in individuals with chronic infection. He hypothesized that HIV-specific CD4 responses are generated and lost during primary or acute HIV infection, and that the initiation of HAART during acute infection may restore these immune responses. To test this hypothesis they have began to generate a cohort of individuals with acute HIV-infection. Over the past 2.5 years they have diagnosed 25 persons prior to or at the time of sero-conversion. They've had either a negative HIV-1 ELISA or a weakly positive HIV-1 ELISA, and a negative or evolving Western Blot. And the diagnosis was primarily made by very high viral loads. 23 of 25 individuals were male and the average age was 35 (range 22-60). 19 are caucasian, 3 African-Americans, and 3 Hispanics. 24 acquired infection via sexual contact (3 by oral sex).

The average viral load was 13.3 million copies/ml (range 250,000 to 95 million) at time of diagnosis during early stage of acute infection. Average CD4 count was 418 (range 42-1023). All individuals were treated with HAART at the time of diagnosis. At baseline, prior to initiating therapy only 3/25 individuals had significant SI. One person's was just above 10 with 2 had indexes of about 100. HAART was started and at 2.5 months after therapy started 22/25 had significant SI (above 10), with the average being 42. At 6 months the average SI was 44 and 22/25 had SI above 10. At one year they had 15 evaluable subjects and all but 2 generated highly significant HIV specific CD4 responses. These two individuals had highly drug resistance virus and their viral load decline was slower than others in group. The SI was similar to the SI observed in a group of 9 LTNPs who did not receive therapy. Rosenberg and colleagues wanted to test therapy discontinuation. Will HIV-specific immune responses generated and maintained be enough to control viral replication? And if virus returns once therapy is discontinued, can this further boost the immune system?