... from the 3rd International Workshop on Salvage Therapy for HIV Infection.

Rescue with ritonavir? One approach to rescue therapy has been to combine other protease inhibitors with ritonavir (Norvir) in an attempt to improve virologic response to antiretroviral therapy by improving drug trough levels. One study sponsored by Merck Research Laboratories and ViroLogic measured the [IC.sub.95] values of drugs from in vivo samples against a panel of 20 protease inhibitor (PI) resistant viruses. The researchers found that ritonavir co-administered with saquinavir (Fortovase), amprenavir (Agenerase), or indinavir (Crixivan) increased drug trough levels of these drugs by 2, 7, and 24-68 fold, respectively. The researchers speculate that the ritonavir/indinavir and possibly ritonavir/amprenavir combinations may be rescue options for patients with PI resistant virus. Abstract 2.

Amprenavir/ritonavir/efavirenz. The combination of amprenavir/ritonavir was investigated as a means of countering the reduction in plasma levels of amprenavir when co-administered with efavirenz (Sustiva). Researchers in France added low-dose ritonavir to a rescue regimen of amprenavir/efavirenz in 7 highly treatment experienced patients. All 3 drugs were started simultaneously in the following doses: amprenavir 450 mg twice daily, ritonavir 100 mg twice daily, and efavirenz 600 mg once daily. Two nucleoside reverse transcriptase inhibitors (NRTIs) were added afterward (specific drugs not listed). After 1 month the mean viral load decreased 2.08 log, and 6 out of the 7 patients had viral loads less than 1000 copies/mL. CD4 T cells increased by an average of 69 cells/[mm.sup.3]. The boosting effect of ritonavir seemed to occur earlier than the lowering effect of efavirenz. Overall amprenavir levels were higher and more stable than with the currently recommended dosing of 1200 mg twice daily. Clearly, this regimen must be evaluated for long-term duration of virologic suppression. Abstract 15.

Indinavir/ritonavir. Research sponsored by Merck & Co. evaluated a combination of indinavir (800 mg) and ritonavir (200 mg) with reverse transcriptase inhibitors twice daily as a rescue regimen in treatment experienced patients (Abstract 27). They conducted a retrospective chart review from 41 PI experienced patients, which revealed that 73% were also experienced with nonnucleoside reverse transcriptase inhibitors (NNRTIs). These patients were treated with indinavir/ritonavir as part of their rescue therapy. Median CD4 T cell counts were 258 cells/[mm.sup.3] and median viral load was 30,015 copies/mL. At 3, 6, and 9 months on therapy, 51.2%, 58.6%, and 57.1% of patients at each timepoint had viral loads less than 400 copies/mL. (The total number of patients reviewed at 6 and 9 months decreased because some of the original 41 patients were still in follow-up.) Another study, in cooperation with researchers at the University of Miami, retrospectively found that half of the 24 reviewed patients who started rescue therapy with indinavir/ritonavir responded to the therapy, i.e. had viral loads less than 400 copies/mL (Abstract 7). Responders and nonresponders did not differ in prior length of time on antiretroviral therapy, prior time on PI-based therapy, baseline viral load (199,333 copies/mL) or baseline CD4 T cell count (239 cells/[mm.sup.3]). Also, after a mean follow-up of 22 weeks, there were no differences with respect to use of an NNRTI, numbers of total and new drugs used, or numbers of drugs used to which viral resistance was detected. No difference was seen between the two groups based on genotypic or phenotypic resistance to indinavir. Adherence was determined to be the best predictor of a favorable response.

Saquinavir/ritonavir follow-up. After 3 years of follow-up in a cohort of 39 patients receiving saquinavir/ritonavir as part of their rescue regimen, investigators from the University of Miami reported on the duration of viral suppression with this regimen. Patients (30 males, 9 females) had been on prior antiretroviral therapy for an average of 39.6 months and had taken an average of 4.4 antiretroviral drugs. Also, all patients had been treated with most of the NRTIs available and most (35) had used at least one PI. Responders were defined as having a greater than 0.5 log copies/mL decrease in viral load. Most of the patients (85%) responded to the therapy. The researchers identified two patterns of response: sustained (46%, with median follow-up of 22 months) and partial (38%, with viral load eventually returning to baseline). They found that adherence and magnitude of viral load decrease were predictive of sustained responses, while baseline CD4 T cell and viral load values and addition of NNRTIs to the regimen were not predictive. Abstract 11.

Rationale for amprenavir/saquinavir/ritonavir. Based on Glaxo Wellcome-sponsored research, investigators have proposed a potential rescue regimen of amprenavir (600 mg), saquinavir (800 mg) and ritonavir (100 mg) twice daily. This regimen would have a total pill burden of 9 capsules twice a day. The ritonavir is included to increase drug trough levels. The proposed regimen is an attempt to take advantage of differences in resistance mutations to amprenavir and saquinavir. Research in vitro indicates that amprenavir-resistant virus (exhibiting a mutation at position 50) is sensitive to saquinavir. Also, saquinavir-resistant virus (with mutations at positions 48 and 90) remains susceptible to amprenavir. The researchers also point out that the amprenavir/saquinavir combination has shown synergistic activity against wild-type and reverse transcriptase resistant virus in vitro. Unfortunately, this combination has not yet been evaluated in vivo for pharmacokinetic data, efficacy or safety. Abstract 16.

New agents seem scarce. Data presented on novel compounds suitable for rescue therapy seemed lacking; many agents are still too early in development or have not been tested long enough to establish efficacy. Some new drugs risk being duplicates of what is already available, thus offering no benefit to patients in need of rescue therapy. Several abstracts at the workshop presented data on a few potentially novel therapies. Researchers at the VA Medical Center (Pittsburgh), the University of Pittsburgh and the University of California, San Diego presented in vitro data on alkylglycerol foscarnet analogs (Abstract 13). Although these compounds have limited utility because of kidney toxicity and lack of oral bioavailability, batyl alcohol analogs of these compounds show favorable activity and resistance profiles and will presumably be tested further. WF10, a macrophage regulator, is being studied by a multicenter group in Canada and the US for use in patients with CD4 T cell counts less than 50 cells/[mm.sup.3] (Abstract 17). Early clinical data suggest that fewer opportunistic infections occur in patients receiving this compound. Investigators have randomized 202 patients so far and are monitoring primary clinical endpoints (i.e. incidence of opportunistic infections and death). Also, researchers in Greece in cooperation with Schering-Plough Pharmaceuticals are studying intermittent dosing of interferon-alpha in combination with highly active antiretroviral therapy (HAART). Data so tar (in asymptomatic, treatment naive individuals) show only that there is no evidence of early drug failure or cross-resistance with reverse transcriptase and protease inhibitors (Abstract 18).

Adherence is vital. Authors of several abstracts came to similar conclusions that nonadherence to treatment regimens is predictive of viral rebound, AIDS progression or even death. Canadian investigators reported results from a study of 815 male and 135 female HIV-infected individuals who were antiretroviral naive (Abstract 28). These individuals started therapy with 2 NRTIs and either an NNRTI or PI. The researchers measured adherence by dividing the number of months of documented prescriptions dispensed by the number of months of follow-up in the first year of therapy. Nonadherence was measured per 10% decline in drug adherence. Their analysis indicated that treatment adherence--independent of physician experience or baseline clinical markers, which were not specified in the abstract--was a predictor of AIDS progression and death. This same group found that nonadherence and less-experienced physician care were associated with viral rebound (Abstract 10). Another study by Italian investigators found that the "benefit of monitoring HAART patients by genotypic resistance testing is limited by patient adherence [as assessed by questionnaire] and available treatment options" (Abstract 30). A group of French researchers reported a lower risk of viral rebound independently associated with increased age, low baseline viral loads, baseline CD4 T cell levels greater than 500 cells/[mm.sup.3], or total or almost total adherence to therapy (Abstract 5).

L-acetyl-carnitine for neuropathy. Research sponsored by Blond-McIndoe Laboratories indicates that oral L-acetyl-carnitine may improve symptoms of peripheral neuropathy commonly associated with some antiretroviral agents. Evidence of increased cutaneous nerve regeneration was seen in skin biopsies from the legs of 4 subjects compared to baseline. Improved neuropathic symptoms, such as pain, were also reported. Further research should be done to reproduce these results in a greater number of subjects. Abstract 36.

HAART: more (rebounds) equals less (response). Researchers from Kaiser Permanente in San Francisco reported on the success rates of rescue treatment after 2 or more viral rebounds on HAART. Thirty-two patients met study criteria (at least 1 year or more of follow-up "after initiating a third HAART regimen or again failing and needing further HAARTs"). Other requirements included having a genotypic test at the beginning of the third HAART regimen and the use of a new PI or NNRTI with at least one unused NRTI during the first and second HAART regimens. Success (viral load being less than 50 copies/mL) after a year or more of follow-up was 60% for patients on their third HAART regimen, 36% for those on their fourth HAART regimen, and 0% for those on fifth or sixth HAART regimens. Also, success was associated with fewer reverse transcriptase mutations but not with the number of protease mutations or the number of antiretroviral drugs used. Abstract 8.

Mega-success? London investigators reported results from an observational study of 82 HIV-infected patients on mega-HAART therapy (Abstract 23). To be included patients had to have viral loads greater than 1000 copies/mL while on a PI containing HAART regimen. Median CD4 T cell count was 110 cells/[mm.sup.3] and median viral load was approximately 316,000 copies/mL. Subjects had taken an average of 5 prior drugs; 10 had taken nevirapine. (The authors did not specify whether the nevirapine experienced patients had actually failed the NNRTI or their length of experience on the drug. In addition, they did not specify how many of these nevirapine experienced patients rebounded.) Patients were switched to efavirenz containing regimens mainly in combination with didanosine (Videx), other NRTIs, indinavir/ritonavir, and hydroxyurea. The researchers reported that 85% of subjects reached viral loads less than 400 copies/mL by 30 weeks; 78% reached viral loads less than 50 copies/mL by 60 weeks. Also, 60% of subjects saw CD4 T cell increases greater than 100 cells/[mm.sup.3] by 54 weeks of follow-up. Viral rebound (above 50 copies/mL) was experienced by 18% of subjects at 30 weeks and 21% of subjects at 60 weeks. Incidence of toxicity was not stated. The authors acknowledge that the rates of sustained viral suppression in this cohort exceeds that in most reported studies. In contrast, Canadian researchers reported results from a study of patients on "multi-drug rescue therapy" that included up to 9 drugs (Abstract 24). Preliminary results showed sustained antiviral response in 30-40% of heavily pretreated patients.