Skip to content

U.S. and French Investigators Alike Are ‘Shocked, shocked’ to Learn of Maintenance Protocol’s Failure

Might the Brits get it right?

First it was kidney stones. Blood in the urine. Hyperbilirubinemia. Yellowed whites of the eyes. An unquantifiable array of quality-of-life issues. The occasional gastric reflux. Annoying but only infrequently life-threatening. A Letter to Physicians warned of the appearance of diabetes. Reports of hypogonadism emerged. Cholesterol and triglyceride levels sky-rocketed. Then came “Crix gut.” “Protease paunch.” Facial wasting: hollowed out cheek and eye cavities. Something called “buffalo hump”: strange disfiguring accumulations of fat around the back of the neck and shoulders. In fact, the up-dated list of possible side-effects from Crixivan now includes: nephrolithiasis (kidney stones), hyperbilirubinemia, hyperglycemia (high blood sugar), acute hemolytic anemia (rapid breakdown of red blood cells), interstitial nephritis (inflammation of the kidney tissues), anaphylaxis (allergic reaction), increased bleeding in persons with hemophilia, alopecia (hair loss), liver failure, ketoacidosis, abdominal pain, nausea, diarrhea, vomiting, acid regurgitation, dry mouth, loss of appetite, dry skin, insomnia, taste perversion, rash, sore throat, skin discoloration, fatigue, weakness, flank pain and headache. And that’s just from one drug. Forget the kidneys. What about the liver? And what are ritonavir, saquinavir and nelfinavir doing to the gastrointestinal tract? Most recently we’ve received reports of life-threatening anaphylactic shock (and several deaths) related to 1592 (abacavir). Then came word of DMP-266 (efavirenz) treated monkeys giving birth to one-eyed, cleft-palated young — some without a brain! Add to this the peripheral nerve damage from ddI, d4T and ddC. The pancreatic toxicity from ddI and 3TC. The anemia/leukopenia from AZT and 3TC. And god knows what chromosomal damage from all of these impostors of the body’s natural DNA building blocks, and one has to ask, At what point will the damage from these drugs outweigh the benefit?

Several clinical studies in the U.S. and Europe have been investigating the concept of a subtractive approach to anti-HIV therapy, in which people receive aggressive treatment (“induction” therapy) with three or four drugs for a period of time in order to try to suppress their viral load to below the limit of detection (anywhere from 50 copies/mL to 500 copies/mL). Then the study participants are switched to a less intensive one- or two-drug regimen (“maintenance” therapy) to try to maintain that suppression.

While the idea of so-called maintenance therapy has been controversial since it was first proposed over a year ago, there are a number of reasons worth investigating the possibility. In other fields of medicine — and, indeed, in other areas of HIV medicine — this induction/ maintenance method is well established. Many cancers and opportunistic infections are successfully treated with an initial period of intensive therapy, followed by less intensive maintenance therapy thereafter. If it is possible to successfully step down anti-HIV therapy in this way, the potential benefits might include:

  • Reduced risk of long-term toxicities
  • Reduced risk of the development of resistance to multiple agents
  • Less onerous daily regimens
  • Lower cost

At the 5th Conference on Retroviruses and Opportunistic Infections in Chicago last month, data were presented from the first two induction/maintenance-type trials to produce results, one from France and the other from the U.S. Both were quick to fail.

Both trials started with the three drugs AZT, 3TC and indinavir (Crixivan). The French trial (ANRS 072 a/k/a “Trilege”) enrolled 397 treatment naïve individuals and switched study participants after 12 weeks of induction to the maintenance regimens of AZT+3TC or AZT+indinavir. The American trial (ACTG 343) included 309 participants, both treatment naïve and AZT-experienced (more or less a 50-50 mix) and switched them to maintenance regimens (AZT+3TC or indinavir monotherapy) after 24 weeks’ induction. Both trials were stopped early because the maintenance regimens were less effective than continuing the triple therapy, as measured by the proportion of participants with HIV RNA >500 copies/mL in the French study and >200 copies/mL in the U.S. trial (see table at top of first page).

Obviously, more sensitive assays could have been employed and study participants not switched until the lowest possible viral load threshold was crossed. But in a Trilege sub-analysis, even among those study participants whose viral loads were suppressed below 50 copies/mL, 19% on the AZT+3TC arm and 13.8% on the AZT+indinavir arm broke through after a mean of 8.5 months on the maintenance therapy regimen.

While not the hoped-for results, the Trilege and 343 results are far from the last word on the induction/maintenance model. Investigators for a similar U.K. induction/maintenance study, called ProCom, note that the duration of the induction phase may be of critical importance in determining the success or failure of such an approach. The U.K. study will also be using different drugs: nelfinavir, soft-gel saquinavir (Fortovase), ddI and d4T. A fourth study, “ADAM,” of Amsterdam’s Joep Lange will throw DMP266 into the maintenance therapy mix. Finally, a Glaxo-Wellcome study (curiously underwritten by the tax-payer funded ACTG) will explore a 36-week induction regimen of all four of G-W’s antiretrovirals (AZT, 3TC, 1592 and 141W) in acute seroconverters. Maintenance regimens in the study (ACTG 371) will consist of double or triple combinations of these agents.

In reality, the study design of both Trilege and 343 were flawed from the outset. And more than a few enlightened lookers-on (certain AIDS activists and M.I.T.’s mathematical modeler Lawrence Wein among them) had vociferously predicted failure from the outset. Not only was 343’s decision to randomly assign AZT pre-treated individuals to a maintenance regimen consisting solely of AZT+3TC a tad suspect, but Trilege’s choice of a mere 3 months of induction therapy blatantly defied the current understanding of viral clearance in lymphoid tissue. As François Raffi, lead investigator for the Trilege study, confessed in a heavy French accent, “We now know from the data of Ashley Haase’s group that even as late as 6 months into therapy there is still actively replicating HIV-RNA in the lymphoid tissue.” In a presentation last October in Hamburg, former NIAID researcher Giuseppe Pantaleo suggested it would be necessary to wait until participants have “undetectable” viral load in a series of lymph node biopsies before successfully reducing the number of drugs. His elegant algorithm calls for 12-18 months of induction therapy at which time serial lymph node biopsies would be performed utilizing the most sensitive techniques. If no RNA and no DNA are detected, Pantaleo says that stopping therapy altogether may be a legitimate option. If only DNA is detected, a less aggressive regimen should keep the virus suppressed, Pantaleo explained. On the other hand, if any evidence of on-going replication (i.e., detectable HIV-RNA) is found, Pantaleo believes that the original aggressive regimen would need to be continued.

Others argue that switching to entirely new drugs for one’s maintenance regimen makes more sense than simply subtracting one or two of the drugs used as part of the induction regimen. ACTG 343’s Diane Havlir went so far as to predict, “We may need to employ immune modulators in combination with antiretroviral agents.”

In a March 13 letter to U.K. AIDS advocates, the ProCom organizers announced that after reviewing the results from Trilege and ACTG 343 they had decided to extend the induction phase of ProCom to 28-40 weeks, and that study participants will switch to the maintenance phase only after achieving two viral load results <50 copies/mL. The committee also decided to restrict recruitment to treatment naïve individuals. With perhaps a lingering hint of crow on her tongue, Havlir conceded that working out a successful induction/maintenance treatment model for HIV infection “will be much harder than we initially thought.” A New England investigator, one of many in the Chicago Sheraton audience who had quietly looked on — months ago — while the 343 deck had been stacked to virtually assure its failure, observed, “Pity it took 300 patients to prove that.” Pity, indeed.

Maintenance Therapy: Percentage of Patients Breaking Through

Lower or Longer, Little Difference

Maintenance Therapy Group Duration of Induction
3 months* 6 months**
AZT+3TC+IDV 1.6% 3.0%
AZT+3TC 19.0 23.0
IDV ND 23.0
*Includes the 191 (of 277) Trilege study participants whose HIV-RNA was suppressed to <50 cps/mL
**Includes all 309 ACTG 343 participants (whose HIV-RNA was suppressed to <200 copies/mL)
ND=not done


Back To Top