CONCLUSIONS 1. What are the goals of initial ARV therapy? To achieve sustained maximal viral suppression with minimal side effects. Strong promotion of patient adherence to therapy is crucial to attain this goal. 2. Should all patients be treated with maximal aggressive therapy? 2.1. Is monotherapy still acceptable in HIV management? The group rules out any kind of monotherapy, except in those cases of HIV-infected pregnant women, where the goal is to decrease perinatal transmission and treatment of the mother is not indicated if rejected or is unavailable. However, triple therapy is advocated for most pregnant women. Some advocate suspension of all antiretroviral therapy in the first trimester since this is the time of organ formation and no agent has established safety for the infant in the first trimester. 2.2. Is dual therapy with two reverse transcriptase inhibitor nucleosides (NRTIs) acceptable? If local conditions (limited budget, unavailability of drugs, defective health care services, high probability of non-adherence, toxicity, etc.) make triple therapy unfeasible, dual combinations with two NRTIs may be acceptable as a suboptimal option for:   a) stable (non-progressive) patients already on dual therapy, with good clinical response and undetectable viral load      (VL);   b) symptomatic patients or those with advanced immunodeficiency (<200 CD4) or with very high VL are eligible for      dual therapy as a less than optimal, yet only possible, therapeutic option;   c) HIV-infected patients with tuberculosis (TB), based on concerns about interaction between rifampin and protease      inhibitors (PI). In the latter a combination of two NRTIs (e.g., ddI plus AZT) and nevirapine may be an alternative.      The use of 3TC in dual combinations should be avoided in order to preserve this NRTI for triple combinations. Close      clinical and laboratory monitoring is necessary for patients on dual therapy, because of the high likelihood of non      compliance. 3. Is there any argument that favors deferral of ARV therapy? 3.1. Is it appropriate to use triple combination therapy when VL measurements are not available? Both prognosis and evaluation of therapeutic response require VL monitoring, and this technology should be available in HIV care. It is difficult to monitor triple combination in asymptomatic patients with no baseline and follow-up VL measurements. In symptomatic patients and those with low CD4 counts, triple therapy may be offered even without VL measurements. 3.2. How does compliance with combination therapies affect final outcome and viral resistance development? Patient compliance to the antiretroviral regimen is critical for a favorable outcome. Non-adherence is the major risk for viral resistance development. 3.3. How can compliance be enhanced? Patient and caregivers education is important. The regimen should be tailored to the patients' everyday activities. A long term goal is regimens with once daily administration to permit directly observed therapy (DOT), similar to that used for tuberculosis. 4. How can current recommendations based on HIV pathogenesis be applied in various socio-economic scenarios? 4.1. Can NIH/IAS-USA guidelines be applied in Latin America? The guidelines can be applied wherever the necessary laboratory and therapeutic resources are available. Modifications in terms of local conditions are acceptable, but some may require suboptimal options according to current scientific data. The group's opinion is that application of those guidelines will result in economic benefits in the long run. However, cost-effective analysis should be done to confirm this opinion. Countries that choose not to provide today the highest standard of care are doomed to pay a high price in the future in terms of both human lives and costs for treating opportunistic infections. This is an issue that countries should address when ranking priorities for health care financing. 4.2. What are the minimum preconditions -at a national, regional, institutional and professional level- for the distribution of, and compliance with, ARV treatments? HIV management should involve adequate attention to issues other than ARV recommendations, such as those listed below: - Supportive infrastructure that enjoins providers, patients pharmaceutical companies, goverment and NGOs - Prevention campaigns - HIV testing and counseling - Prevention and treatment of opportunistic infections and other complications - Patient education - Training for health care providers - Massive diffusion of information on critical issues - Improvement of health and social support services - Psychosocial support for patients - Patient confidentiality - Physician and patient joint decisions - Reliable drug supply - Requirement for appropriate standards of quality control, including toxicity tests and proof of efficacy for new drugs   and new drug formulations. 4.3. If budgetary constraints and national health priorities are considered, is it feasible to treat all patients? Should priorities be established for receiving treatment? If so, who should be treated first? Priorities should be established only in the context of budgetary constraints. A rational order of priority proposed is: first is the prevention of perinatal transmission; the second, the treatment of symptomatic HIV-infected patients; third, prophylaxis for healthcare workers with workplace exposure. The highest priorities should result from an evidence-based medical approach; this applies to all three mentioned categories. Next is treatment of asymptomatic patients based on CD4 count and viral load assays. Victims of sexual assault should be considered, although there is no evidence of the efficacy of treatment in this setting. 5. Ethical Issues 5.1. Are there patients non-eligible for therapy? No decision should be made on the exclusive basis of demographic profile or risk groups. No population can be said to be non-eligible for ARV therapy. Anticipated non-compliance represents a relative contraindication. Patients with a CD4 > 500/mm3 and viral load of < 10,000 copies/mL may be observed. For individuals who are difficult to reach, such as the homeless, people living in rural areas or IV-drug users, programs should be established to facilitate their management. In areas with no health care providers with extensive expertise in HIV-infection management, programs for ready consultation should be developed. 5.2. Should healthcare employers provide prophylaxis to individuals with accidental occupational exposure? If accidental occupational exposure to HIV occurs, the employer should offer health care providers the best available prophylaxis, at the very minimum AZT for one month (based on evidence from the Centers for Disease Control and Prevention, USA). A further suggestion is to establish hot-lines in order to deal with accidental exposures. 5.3. Are there ways of increasing access to funding for HIV therapy? If so, should these efforts be channeled through governments, private agencies, academia/research, or NGOs? A program is just being undertaken in Mexico between the federal government and NGOs, aimed at raising money through non-profit foundations that collect funds from different sources and then invest them. The money is then used to buy drugs for patients attended at public hospitals and non-profit health centers. Pharmaceutical industry-supported therapeutic trials are a very advantageous way to obtain drugs, but there should be an agreement ensuring drug supply after the protocol is completed. Such agreement should be seen as a major issue in terms of the acceptability of any treatment protocol. It would be unethical to enroll a patient in a therapeutic trial, if continuity of the evaluated regimen cannot be ensured after completion of the protocol. Efforts should be made to promote Expanded Access Programs providing there is a continuing source of drugs. 6. What are the latest options for initial ARV therapy? 6.1. What are the current and future roles of hydroxyurea in ARV therapy? Recent studies showed the achievement of substantial of plasma VL and lymph node VL supression with combination therapy including ddI, hydroxyurea and a PI. Further research is needed to establish hydroxyurea's role in combination therapies. New members of the three existing classes (NRTI, NNRTI, and PI) have shown promising results in clinical trials. These data hopefully give a broader picture of a greater number of therapeutic options in the near future. RECOMMENDATIONS A. Naïve, asymptomatic patient, CD4 >500, VL not available Most importantly, VL should be determined. If it is not possible, the patient must be closely monitored with CD4 counts. If CD4 levels begin to drop significantly (>30%) or the patient becomes symptomatic, treatment should be started without VL determinations. B.1. Naïve, asymptomatic patient, CD4 >500, VL >10,000 copies This is a low-priority candidate for therapy. The strength of the recommendation will depend on VL: the higher the VL, the stronger the recommendation. B.2. Naïve, asymptomatic patient, CD4 >500, VL 5,000-10,000 copies Priority of this candidate for therapy is lower than that of patients with VL >10,000 copies. VL should be monitored. C. Naïve patient, CD4 and VL not available, with zoster and oral candidiasis This patient should be treated. D. Naïve patient, CD4 and VL not available, with any AIDS-defining illness This patient should be treated. The recommendation is stronger than in C. E. Patient currently treated with AZT, clinically stable, CD4 = 350 This patient should be switched to triple combination or, if this is not feasible, at least to two new NRTIs. F. Patient currently treated with 2 NRTIs, clinically stable, CD4 = 350, undetectable VL Probably no medication needs to be added if the combination is acceptable. However, it would not be a mistake to add a PI if the physician in charge thinks that the patient will benefit with triple combination. G. Patient currently treated with 2 NRTIs, with decreasing CD4 count and/or increasing VL This patient should receive a triple combination with at least one new NRTI and a drug of a different class (PI or NNRTI). H. Patient currently on triple combination therapy, with decreasing CD4 count and/or increasing VL This patient should receive a new regimen with 3 or 4 drugs, including at least 2 new drugs. If CD4 and VL return to pretreatment levels and no other therapeutic options are available, or if the regime is highly detrimental to the patient's quality of life, the attending physician should consider discontinuation of ARV treatment or the patient's inclusion in a clinical trial protocol. Discontinuation should be also considered for patients with advanced AIDS dementia. I. Acute Retroviral Syndrome Adequate diffusion of information on acute retroviral syndrome (ARS) to promote detection by primary care physicians is important to make possible timely diagnosis and appropriate referral of affected patients. There are no long-term clinical trials demonstrating that ARV therapy is beneficial for patients with ARS. Based on scientific data, most experts would treat these patients with a maximal suppressing combination of 3 drugs. J. Miscellaneous recommendations J.1. The combination of hard gel capsules of Saquinavir plus 2 NRTIs is inferior to other triple combinations of 2 NRTIs with other PI. Saquinavir plus Ritonavir with one or two NRTIs is an acceptable first line combination. J.2. Generics. We strongly endorse establishments of standards of quality control for all licensed drugs, including ARV drugs. The data presented should demonstrate adequate bioavailability, show in vitro and in vivo activity and limited toxicity. Data intended to endorse certification or inclusion in protocol regimes should be original and related to the actual product to be licensed using assays that are made according to industrial standards. Pharmaceutical companies should make any data on drug quality available to health care providers. J.3. Three fold changes in VL and a 30% difference in CD4 counts are considered significant and should be taken into account. J.4. Treatment of HIV-infected patients should be managed by specialists in infectious diseases and experts defined by other acceptable criteria. J.5. It is advisable to involve local community organizations in the process of implementing recommendations. J.6. It is recommended that mechanisms be established to update guidelines to accomodate rapid changes based on availability of new drugs and new data. J.7. It is recommended to hold periodical meetings of the panel adding experts from countries not present this time, if possible on occasion of major meetings, like those of the Pan-American ID Society. J.8. It is recommended that this document be widely distributed, and made available to AIDS specialists and other health providers in English, Spanish, and Portuguese. The resulting publication should include the background papers presented at this meeting. Diffusion should be made through all available means. J.9. Consistent with conclusions 3.3. and 4.2, and recommendation I., the panel suggests to work out educational programs on HIV for appropriate care providers in order to improve early recognition, and timely and appropriate referral to the specialists. Also, the panel recommends to implement patient-education activities, with active participation of groups of patients. J.10. It is also strongly advisable to develop coordinate efforts to monitor resistance.