Early treatment and interruption makes no difference to disease progression, study finds

The largest study yet done on the whether immediate antiretroviral therapy (ART) may benefit people with recent HIV infection has found that immediate treatment followed by treatment interruption produced no reduction in the subsequent viral load ‘set point’ of people starting ART in early infection, and no reduction in the expected set point of patients starting treatment in acute infection, before seroconversion.

The ACTG 371 study did find that, in people who interrupted their treatment after a year of treatment, a large minority (40%) of patients, and nearly half (48%) of those with a baseline viral load of under 100,000 copies/ml, still had a viral load less than 5000 six months after ceasing to take their initial treatment regimen.

However the researchers were unable to say if their strategy produced improved immune control and clinical benefit compared to similar patients who do not start ART. They conclude: “Early ART followed by treatment interruption does not produce the profound alterations in disease course that were originally suggested by uncontrolled studies.”

The study recruited 121 patients, 50 with acute HIV infection (diagnosed within the first month, before antibody seroconversion) and 71 with non-acute but recent infection (diagnosed within the subsequent two months).

The study found that there was no significant difference in rates of viral suppression, viral rebound and CD4 decline during treatment interruption between patients who started treatment in acute infection and patients in early infection. This dashes hopes raised by smaller studies that pre-seroconversion treatment might produce permanent reductions in viral load and disease progression.

The researchers comment: "It seems highly unlikely that that any trial will enrol patients closer to the actual moment of infection than ACTG 371.” Trials based on trying to find even earlier cases and treat them would not “address the reality of primary infection as seen in the clinical setting," they add.

The study gave all patients a standardised regimen of antiretrovirals for a year. If patients’ viral load was below 50 at this point, they could then choose to interrupt their treatment. The primary endpoint was defined as a viral load below 5000 24 weeks after starting treatment interruption.

Patients were restarted on treatment if they had one viral load over 50,000 or three consecutive viral loads over 5000. One further cycle of treatment interruption was offered if patients achieved a viral load under 50 again and maintained a viral load under 400 for at least two months.

ACTG 371 originally started in 1999 as a study of an induction/maintenance strategy of treatment and subsequent interruption for patients in early infection. It was modified into a smaller study and patients were stratified into those with acute and recent infection as time went on.

The trial's age is reflected in the fact that its initial regimen was one that would now not be recommended, namely d4T, 3TC, abacavir and boosted amprenavir. Patients were allowed to substitute ddI for d4T and nelfinavir for boosted amprenavir in cases of individual drug toxicity and subsequently to substitute any other NRTI (including later ones such as tenofovir) for any NRTI toxicity but other protease inhibitor substitutions and NNRTI substitutions were not allowed.

Toxicity due to this regimen appears to have produced an initial high drop-out rate, such that only 73 out of the initial 121 patients recruited (60%) actually started a treatment interruption. Of the other 40%, half discontinued study treatment within the first year and half did not interrupt the study treatment.

Participants were 95% male and 70% white. There were no differences between those with acute and those with recent infection, except that those with acute infection had, unsurprisingly, a higher median viral load (210,300 versus 42,700). There were also no differences in the patients who actually started a treatment interruption except for viral load. Patients with acute infection who started a treatment interruption had a lower median viral load (109,000) than patients who did not start an interruption.

The proportion of the 73 patients interrupting treatment who achieved the primary endpoint of a viral load under 5000 after 24 weeks was 40%, 43% in those with acute and 38% in those with recent infection, a non-significant difference. Forty-eight per cent of patients with baseline viral loads under 100,000 achieved the endpoint, including 57% of those with acute infection (eight out of 14 patients).

In terms of secondary endpoints:

• Median time to treatment resumption was ten weeks in acute infection and 14 weeks in recent infection.
• Peak viral load during interruption was 28,000 in acute infection and 36,000 in recent infection. In the case of patients with recent infection, this was virtually identical to baseline viral load and there was therefore no indication of a reduction in viral load ‘set point’.
• Time to peak viral load during interruption was eight weeks in both groups of patients.
• During the initial phase of treatment interruption, viral loads doubled every week in patient with acute infection and tripled in those with recent infection.
• CD4 counts decreased from 893 to 728 and 829 to 739 in acutely- and recently-infected patients respectively.

None of the differences between acutely- and recently-infected patients was statistically significant.

The researchers conclude that their study failed to establish “that early treatment enhances virologic suppression”.

However, they also point out that it does not establish whether immediate and non-interrupted versus deferred treatment for patients diagnosed in early infection will improve clinical outcomes, nor does it establish “whether ART, if initiated in primary infection, should ever be discontinued”.

Several randomised studies are currently underway to answer these questions.