Mike Youle and Julio Montaner reported their use of multi-drug regimens, commonly called Mega-HAART to treat individuals with extensive prior treatment experience and advanced HIV. Both Montaner and Youle initiated treatment with induction therapy. That is, they would start individuals off with a large number of drugs. As many as they could tolerate, recycling drugs they had previously used. A rationale behind Mega-HAART is that perhaps by bombarding HIV with so many drugs, even if the person had resistance to them, the virus would have a hard time replicating. The feeling is that a person may have a minimal level of sensitivity to each drug, or to many of them, and thereby by using enough drugs there might be adequate suppression on the virus to stop it from replicating. After a period of time, possibly 12 weeks, drugs could be selected to be eliminated from the regimen. If viral load started to increase after eliminating some drugs, these drugs could be placed back into the regimen. The researchers said side effects do emerge but that they are generally manageable. Youle says that by giving adequate attention to the patients in dealing with these regimens the patients could succeed.

Interesting presentations were made at this Workshop on the use of lactate testing for lactic acidosis, the utility of resistance testing when viral load is below 1000 copies/ml, and the use of indinavir+ritonavir at a dose of 800 IDV+ 200 RTV twice daily for use in salvage situations. Reports on these discussions will follow.

Prolonged Viral Suppression After Introduction of a Post HART Salvage Regimen. Mike Youle, from the Royal free Centre for HIV Medicine in London, reported an update on his salvage therapy study using Mega-HAART in highly treatment experienced individuals. He has been reporting on his use of Mage-HAART for a few years. He studied 92 efavirenz-naÔve patients who previously failed a PI containing regimen. About 40% had exposure to two protease inhibitors.

At baseline median CD4s were 112 (range 48-200): 24 patients had <50 CD4s (25%), 17 patients had 50-100 CD4s (19%), 27 had 100-199 (30%), and 24 had 200 CD4s or greater (25%). The median viral load was 302,000 (5.5 log) copies/ml (range 76,000-661,000: 9 patients had <10,000 (10%), 15 had 10,000-99,000 (16%), 35 had 100,000-499,000 (38%), and 33 had >500,000 (36%). And 58% had a previous AIDS diagnosis. Patients had previously used a median of 5 drugs: 3 nucleosides, 1 protease inhibitot (mean 1.7), and 17% (16) previously used nevirapine. So these were a group with advanced HIV.

Patients received a median of 5 drugs: 92% efavirenz, 80% ddI, 75% started off with hydroxyurea, 90% a double PI regimen, 75% ritonavir, 73% indinavir, 16% saquinavir, 12% nelfinavir, 36% d4T, 26% 3TC, 5% AZT, 5% abacavir.

The follow-up of these patients is a median of 41 weeks (range 31-71) with a maximum of 95 weeks. Viral load was tested every 6-7 weeks. Youle reported, at a median follow-up of 39 weeks Kaplan-Meier estimates and an intent-to-treat had 85% (n=68) reached <400 by 30 weeks, and 78% reached <50 copies/ml by 60 weeks. During the study some of the patients had switched to another regimen. "We started re-salvaging the salvage" Youle said. Some patients who failed ritonavir+indinavir are receiving foscarnet and ABT-378. The proportion of patients whose viral load rebounded after initially reaching below 400 copies/ml is about 20-23%. The number of evaluable patients is small, only 11 at 64 weeks. At week 24 the number of evaluable patients was 32 and 18% had rebounded above 400 (2 consecutive viral loads above 400). Youle also reported that using an intent-to-treat analysis about 40% virologically failed by week 50. One reason for the seemingly high 40% rate is that this includes patients who never reached <400 but who achieved a low but detectable viral load. Youle said he considers reducing viral load to several thousand a success, in these individuals with extensive prior experience who've few treatment options remaining, even if they could not attain undetectable. And I agree. Over time there are increasing numbers of people who failed at some point. That's when regimens are changed. Previous use of nevirapine (17%) was a factor in virologic failure. About 60% of patients experienced a CD4 rise of about 100 by 54 weeks follow-up.

34 (37%) of patients experienced one or more adverse events leading to stopping one or more of the drugs in their regimen. 10 patients experienced one or more new AIDS defining diseases (3 non-hodgkins lymphoma, CMV 2, 6 others). There was 1 death from PML (at VL<50 and CD4 >100 rise).

Youle discussed why this approach worked. He said a number of patients received induction therapy followed by maintenance therapy. That is, patients initially received as many drugs as possible that they were not intolerant to, and at about 12 weeks some drugs were removed from the regimen. Some patients were admitted into the hospital. Dual PI was used in the vast majority of patients: 800/200 and 800/400 IDV/RTV. There have been a lot of treatment interruptions but this analysis was unable to show a benefit of treatment interruption. There is constant contact with patients. They were explained that this could be their last opportunity to succeed with treatment. They have every access to physicians and nurses. They've intensfied therapy for a lot of patients who have had viral load blips.

Multi-Drug Rescue Therapy (MDRT) in Three Cohorts of HIV-Positive Individuals. Julio Montaner reported on his use of MDRT (multi-drug rescue therapy), also known as Mega-HAART, in his affiliation with the Canadian HIV trials Network and the British Columbia Centre for Excellence HIV/AIDS in Vancouver, Canada. He reported on an observational study conducted in a single, university-based referral clinic. Patients who failed several regimens were offered MDRT with at least 5 antiretrovirals including up to 2 protease inhibitors, 2 NNRTIs, 4 NRTIs.The MDRT regimens were adopted according to the patient's history of previous antiretrovirals (ART) exposure, tolerability, and baseline laboratory profile. Three cohorts or groups of patients were started: cohort 1 (n=106) started between 8/97-6/98; cohort 2 (n=68) started between 7/98-12/98; and cohort 3 (n=71) started between 1/99-5/99. Their baseline viral loads were, 62,000, 56,500, and 63,700, respectively. The probability of sustaining viral load <400 copies/ml among responders was calculated using Kaplan-Meier methods.

Patients in cohort 1 received 5-6 drugs and had previous ART exposure of 27-55 months. Their average baseline CD4 count was 180. Patients in cohort 2 had 5-8 drugs in their regimen and previous ART exposure of 23-59 months. Their average baseline CD4 count was 200. Patients in cohort 3 received 6-7 drugs in their regimen and had 31-62 months of ART exposure. Their average baseline CD4 count was 190.

The median viral load following initiation of MDRT was 2.95 log (891 copies/ml) and 2.80 log (630 copies/ml) at week 36-46 for cohort 1 and 2, respectively, while cohort 3 had median viral load of 3.08 log (1200 copies/ml) at week 25-35. Intent-to-treat analysis showed that 35%, 44%, and 36% of patients had viral load values <400copies/ml between weeks 25 and 35 of follow-up for cohort 1, 2, and 3, respectively. Median change in CD4 was -10 for cohort 1 and 0 for cohort 2 and 3. The median change in percentage CD4 was 0% for cohort 1, 0% for cohort 2, and 1% for cohort 3.

A favorable viral response is defined as having at least 2 consecutive viral load measures <400 after starting MDRT. Among 106 cohort 1 participants, 52 (49%) responded, while 41 (55%) of 75 in cohort 2, and 34 (49%) of 69 in cohort 3 responded. Multivariate logistic regression analysis for cohort 1 showed that every log copies/ml increase in baseline viral load decreased the patient's likelihood of responding to MDRT by 55%.

Among responders, the probability of remaining suppressed at 24 weeks from initial suppression of viral load <400 was 76.8% for cohort 1, 81.8% for cohort 2, and 78.8% for cohort 3.

Among cohort 1, phenotypic testings showed that 59% of the patients had some degree of decreased susceptibility to 7 or more ART drugs at baseline. Subanalysis of 95 cohort 1 patients with a phenotypic resistance profile available showed that baseline susceptibility to 3TC, d4T, ddI, and saquinavir were most strongly correlated with a favorable VL response.

When tolerability in cohort 1 was examined, the most common lab abnormalities were alterations in liver function tests, neutropenia, anemia and changes in lipid profile. Montaner reported that most adverse events were mild to moderate and did not interfere with continuation of treatment. Subjective adverse events, although less frequent, were more often serious and they involved most frequently the gastrointestinal tract.

Montaner reported that median viral load at week 47-57 for cohort 1 was about 1000 copies/ml; about 1000 copies/ml for cohort 2 at week 36-46; and for cohort 3 also about 1000 copies/ml at week 25-35.

As stated above, Montaner used mega-HAART induction for about 12 weeks before stripping away some drugs. John Mellors raised the issue that selecting a regimen by using resistance testing up front may be preferable because you can select the drugs a person is sensitive to and thus use less drugs. Montaner responded by saying that for individuals with such extensive prior treatment experience, genotypic mutations may not predict which drugs to use in regimen. Montaner said adverse events were frequent but manageable.

Montaner concluded that MDRT induced a 30-40% antiviral response in heavily pre-treated patients. Lower viral load at the time of initiating MDRT remains as the strongest predictor of virologic response. He suggested that using baseline resistance testing in addition to historical, clinical, and lab evidence may help in regimen selection when recycling ART drugs. Full genotypic and phenotypic analyses are underway to help further characterize predictors of response.