The Body PRO Covers: The 10th Conference on Retroviruses and Opportunistic Infections

CROI 2003: When, and With What, to Initiate Antiretroviral Treatment

February 14, 2003

One of the ongoing themes of all AIDS conferences is how they inform decision making for two of the most pivotal treatment issues in the life of someone with HIV: when to initiate antiviral treatment, and with what regimen.

Like almost all prior meetings, the Retrovirus conference featured research regarding both of these issues. Although treatment will not necessarily dramatically change as a result of this meeting, there was the addition of new treatment options, particularly around the issue of regimens to consider.

When to Start

Added to the list of documented concerns regarding why the decision of when to start treatment is difficult, is the issue of toxicity of treatment. If antiretrovirals were completely free of side effects, were easy to take on a daily basis, and rarely if ever resulted in resistance, there would likely be a completely different tone to the conversations about initiating treatment. However, it is clear that over time, HIV does significant damage. Thus, when to initiate treatment becomes a balance of the need to stop the progression of HIV and the tremendous benefits seen with treatment, versus an understanding of the cumulative toxicity of treatment itself. When and if treatment ever becomes devoid of problems and tradeoffs, it would likely be desirable to provide treatment no matter what the CD4+ count.

At this meeting, we saw evidence of one more important side effect to consider in both when to treat, and ultimately, with what regimen. Over the past several years, the results of many studies on medication side effects have been discussed at AIDS conferences, including changes in body fat distribution (lipodystrophy), problems of elevated blood sugar, and increases in blood lipids such as cholesterol. Given that changes in a patient's cholesterol can be expected to take years to be relevant, a large study was conducted and presented at this meeting. The "D:A:D" 1 (Data on the Adverse Effects of Drugs) study was an analysis of many large cooperating cohorts of patients receiving care or research at many sites in several continents, including the U.S. and Europe. The focus of this analysis was to assess what impact treatment has on increasing the risk for heart attacks, and what risk factors alter that risk. In all, the researchers monitored the outcomes for about 23,000 people who have been on treatment for an average of two and a half years. The baseline age of the group averaged 39, and 24 percent were women.

In order to study what role, if any, antiviral treatment has, they used the data from the 19 percent of patients who were not on treatment; this group served as the comparison for the 81 percent who were on antivirals. These data were then analyzed to see if there was any evidence of heart attacks as a result of continuous antiretroviral treatment. It is important to note that about 60 percent of people monitored were past or current smokers -- a known risk factor for increasing rates of heart attacks. The deleterious impact of smoking was seen here as well, in fact it doubled the risk of a heart attack. In their analysis, the researchers did find a small but real impact of antivirals increasing the risk of heart disease -- there was a 26 percent increase in the risk of a heart attack for those on continuous antiviral treatment as compared to those not on treatment. The risk appeared to increase with each year someone was on treatment -- increasing in some way the validity of the link to antivirals as causative.

It is important to appreciate that the rate for heart attacks overall was only about three in 1,000 persons per year of monitoring -- so that the absolute risk of a heart attack is still quite low. In addition, at this time, the researchers are unable to identify if this problem was due to any specific medications or even specific classes -- people were on a large variety of antiviral medications and this initial presentation did not analyze the risk by class of drug. Of note is that the researchers did find that this effect of antivirals was present even when controlling for blood lipid levels, although having high cholesterol does independently increase the risk by about 16 percent. Triglycerides were found to have no impact on the risk of heart attacks in this study. And men were almost three times more likely to have a heart attack in this study as compared to women.

A completely unexpected finding was also reported. People with lipodystrophy were found to have a 40 percent reduced risk of heart attacks. This observation remains unexpected and unexplained since elevated lipids and body fat alterations often run together. Nevertheless, these results do add to the real balance that treating HIV entails, since if nothing else, it provides clear incentives to continue the search for new and hopefully less complicated medications that suppress HIV with fewer tradeoffs.

Thus, given the benefits and risks of treatment mentioned previously, there continues to be a general consensus that treatment should be initiated before CD4+ counts fall below 200 cells/mm3. At this meeting we saw yet another analysis confirming that those who start treatment with a CD4+ count below 200 do not do as well in terms of AIDS-related illness as compared to those who start above this threshold2. However, another cohort, this one from Spain, even found some differences in those who started treatment with a CD4+ count above 350, versus waiting until 2003. Despite these unresolved issues, there is nevertheless a clear benefit to treatment, even when treatment is initiated at lower counts. An analysis of the ART Cohort Collaboration from Europe documents quite clearly that while baseline CD4+ counts are important, the response to treatment is seen even just six months later. They noted that even someone who initiates antivirals with a baseline CD4+ of 25 can appreciate a 50 percent reduction in the risk of AIDS-related complications with as little as a 25 cell increase from treatment at six months. Similarly, someone who has a viral load before treatment of over 100,000, and achieves suppression to <500 in six months will experience a 70 percent reduction in the risk of illness as a result of this effective treatment4.

Therefore, at this point, the question concerning how far above a count of 200 cells/mm3 should HIV treatment be initiated remains the subject of ongoing debate. All the more so because, as we know, most people remain clinically well despite these low CD4+ counts, and, as noted above, there are clear risks of treatment. Randomized studies are underway which attempt to address this question. It is expected to be many years before there is clarity on this particular issue. While we can control HIV even at CD4+ counts lower than 200 on several antiviral combinations, it remains unclear to what degree immune reconstitution is fully protective when treatment is postponed to these lower levels. Thus, there still remains a general recommendation to initiate treatment before CD4+ counts fall below 200, assuming someone is diagnosed with a CD4+ count above this level.

Finally, researchers also re-examined the role of viral load in choosing treatment. What seems clear now is that some, although not all, regimens are less successful at higher viral loads. One study presented data showing that patients with a viral load >100,000 remained at higher risk5 despite treatment, but adherence to antivirals, as well as utilizing more potent regimens, could minimize the importance of this viral load. Most if not all of these studies were conducted prior to the availability and use of more recent treatment options, including the use of efavirenz (EFV, Sustiva) and "boosted" protease inhibitors (PIs) such as lopinavir/ritonavir (LPV/RTV, Kaletra), which have shown impressive potency and success in controlling HIV at even these higher viral loads.

Therefore, it remains fair to state that the progress made in identifying potent combinations provides flexibility concerning when treatment should be initiated, although there are some data to suggest that there may still be an advantage to treatment started sooner rather than later in the process.

What Combinations to Start On

As a result of even more head to head studies comparing some of the better treatment options, we have evidence to support a number of initial regimens as superior choices. Virtually all of the initial combinations that have been studied are based on a combination of two nucleosides, in combination with either a third NRTI (abacavir [ABC, Ziagen]), a non-nucleoside RT inhibitor (NNRTI), or a protease inhibitor (PI). Most of the differences between these combinations have been demonstrated when looking at people with higher viral loads (for example over 100,000) and/or when looking at those with lower CD4+ counts before treatment. And while there are differences seen in success rates, it is useful to note that a treatment is only successful when someone takes it daily as prescribed, and feels well while taking it. Clearly it is difficult to convince anyone to remain on a combination if they don't feel well while on it. Thus, having several options to choose from will remain critically important given the need to match each person to a regimen that is most successful for them. While it would be ideal to somehow be able to predict who will do best on any selected medication, it is impossible to know in advance who will tolerate each medication. With rare exceptions, such as genetic testing to assess the risk of a hypersensitivity reaction to ABC, the occurrence of side effects to treatment can only be known once a patient is on the treatment.

As a result of several comparison trials, combination regimens based on two NRTIs plus either EFV or the boosted PI LPV/RTV have emerged as the "front runners," i.e., the most well-established potent initial regimens to choose from. Studies that have compared EFV to other choices have consistently shown that EFV is either similar or superior to other choices. In prior meetings we have seen data showing that the combination of two NRTIs plus EFV is in general superior in the numbers with viral suppression to:

  • nelfinavir (NFV, Viracept) from ACTG 3846
  • boosted saquinavir (SQVsgc, Fortovase) (FOCUS trial) 7
  • boosted amprenavir (APV, Agenerase) (CLASS study) 8
  • indinavir (IDV, Crixivan) (Dupont 006) 9
  • ABC (ACTG 5095)

At this meeting we saw the first results10 of a head to head study comparing another NNRTI, nevirapine (NVP, Viramune), to EFV always in combination with d4T (stavudine, Zerit) and 3TC (lamivudine, Epivir). This study contained four arms, since EFV was compared to NVP as either two pills taken once daily (QD) or one pill taken twice a day (BID). (The fourth arm, taking both NVP and EFV with the two NRTIs did not do as well on several measures and thus is not a contender for first line use.)

Overall the study showed that NVP and EFV were similar in outcomes, although, on some measures of treatment success, NVP and EFV were equal while on other measures EFV did somewhat better than what was seen with NVP. There are important differences in side effects as well as trends in viral potency, and these differences also help focus a discussion of options. For example, there is an extremely small but real risk of serious toxicity seen with NVP, including two deaths noted from the approximately 800 people taking the drug: one from hepatitis, and the other as a complication of a serious rash called Stevens-Johnson. There was also a trend for more consistent suppression at high viral loads for EFV over NVP, although this difference did not achieve statistical significance. It was also shown that there were more instances of liver toxicity when NVP is dosed once daily over twice daily (and over EFV), suggesting that it is reasonable to consider using NVP twice daily to minimize hepatic complications from this agent. One other interesting outcome is related to the number of patients who needed to discontinue treatment due to side effects usually associated with EFV only -- specifically, those involving sleep or mood. There was only a 2 percent greater rate of discontinuing EFV over NVP due to such side effects, which is perhaps lower than would be predicted based on the predictable toxicities associated with these agents. Overall, while this study served to improve the overall impression of NVP as more potent than was previously understood, it is unclear how the results of this trial will alter preferences between these drugs given their differences.

Additional trials showed impressive and favorable potency when using newer unboosted PIs. The PI APV has been reformulated to a much-improved tablet currently called "908" or fos-APV. In a head to head study11 comparing 908 to NFV, the 908 arm performed consistently better in terms of viral suppression when compared to NFV in a study in which both where given with an NRTI backbone of 3TC and ABC. This outcome included consistent results at very high viral loads for 908 over NFV, showing that 908 did as well at high viral loads as compared to what was seen at lower viral loads. This was distinct from the response on NFV, which did not do as well at highest viral loads.

This impressive and somewhat unique demonstration of the potency of 908 was shown for people initiating therapy, and was partially supported by a second study12 in PI-experienced patients. In that trial, a head-to-head comparison of boosted 908 versus LPV/RTV in PI-experienced patients demonstrated that LPV/RTV was superior in percentage suppressed: 48 percent of patients' viral loads were <50 copies on LPV/RTV while 40-42 percent of patients' viral loads were suppressed on 908/ritonavir (RTV, Norvir), with the 908/RTV dosed either once daily or twice daily respectively. U.S. Food and Drug Administration (FDA) review of 908 should hopefully be underway and approval can be anticipated sometime this year, at least in the U.S.

Just after the Retrovirus meeting ended, another trial revealed some important results informing practice. This was the trial called ACTG 5095 and it is a currently ongoing trial with treatment naive patients. It compared AZT [zidovudine, Retrovir]/3TC/ABC (Trizivir) to AZT/3TC (Combivir)/EFV to AZT/3TC/ABC plus EFV. Although the two arms containing EFV are continuing at this point to see if there are any differences between them, the study team recently announced that due to inferior viral suppression in the AZT/3TC/ABC alone arm, they have stopped that arm and are in the process of altering the regimen for patients on that arm. They did not report a lot of detail, but what they did say was that they found about 10 percent less viral suppression in the AZT/3TC/ABC arm versus the AZT/3TC plus EFV arm. Also of note is the discovery of some evidence of ongoing loss of suppression found on the AZT/3TC/ABC arm which was not clearly explained or predicted by the baseline viral load. In prior trials, AZT/3TC/ABC did not do as well at higher viral loads, but was as potent as other standard combinations at lower viral loads. It is not yet clear from the analyses presented if that was seen here or if there was less suppression at any baseline viral load. These results now suggest caution when relying on this triple regimen as stand-alone therapy -- as there appears to be a slightly higher virologic failure rate than with other available choices. It is noted by this study that AZT/3TC/ABC can certainly continue to be useful as a three-drug backbone for a four-drug regimen, if that is desirable.

Finally, there was some new data13 presented at the Retrovirus conference concerning atazanavir (ATZ), another PI in development. Data currently supports ATZ's use at least for initial therapy. The positive features of ATZ include a lack of lipid disturbances, as well as an interesting resistance pathway, which, if it develops as a result of using ATZ as the first PI, will result in a lack of cross-resistance to other PIs. There is a demonstration, shown only by resistance testing, that the mutation conferring resistance to ATZ results in hypersusceptibility, or potentially greater activity, from other PIs14. The other important feature of this agent is the demonstration of, not only a lack of disturbances to the lipid profile of those taking it, but a reversal in the lipid disturbances associated with other agents after a patient switches to ATZ. There are ongoing questions concerning the data documenting ATZ's true potency profile. Future data will allow us to understand ATZ's role in treatment when it is boosted by RTV, especially after a patient has used other PIs. Research results answering these questions are expected at future meetings and will broaden our understanding of ATZ's role in HIV treatment.

There were also a few important presentations addressing the choice of which two NRTIs to pair with the "third agent." Comparative trials showed differences between tenofovir (TDF, Viread) versus d4T when paired with 3TC and EFV. These differences are summarized on this Web site -- in sum, there were nearly identical response rates seen in naive patients starting on either TDF or d4T with 3TC/EFV, but there were some toxicity differences noted, including higher blood lipid measurements seen on d4T. In addition, there was a reported difference in the number of patients who developed lipodystrophy in this "blinded" trial. There was a 12 percent rate reported on patients on d4T versus an only 1 percent rate at 96 weeks for those patients on TDF. While resistance was rare in both arms, the controversy about the earlier use of TDF raised by some concerns the development of the K65R mutation, which if it occurs, can lead to some degree of NRTI cross resistance slightly more often than what is noted in the d4T arm15.

Summary: Regimens to Choose From

In prior meetings, a few studies showed concern for the lesser performance, as well as higher toxicity potential from the combination of d4T and ddI (didanosine, Videx). Therefore, the five dual-NRTI pairs generally favored at this point, based on available data, are 3TC with either AZT (co-formulated as Combivir), d4T, TDF, ddI or ABC. Combinations that avoid the use of 3TC up front are infrequently selected at this point, given the lack of data to favorably recommend any of these.

Choices regarding the "third agent" can be divided into two categories: those with adequate potency, and those with ample potency. The former can be understood as those medications with a reduced suppression rate for people with high initial pre-treatment viral loads. The drugs in this category would include the NRTI ABC, and all "unboosted" PIs such as NFV, with the possible difference of 908, based on the NEAT study data. Medications with potency at a higher viral load would include boosted PIs such as LPV/RTV, as well as the non-nucleoside EFV. Based on the 2NN study, the data on NVP were similar to EFV, although just a bit less successful than what was seen with EFV at higher viral loads, allowing clinicians to decide the importance of the small differences observed.


  1. Friis-Møller N., Weber R., D'Arminio Monforte A. et al. Exposure to HAART Is Associated With an Increased Risk of Myocardial Infarction: The D:A:D Study. Presented at: 10th Conference on Retroviruses and Opportunistic Infections; February 10-14, 2003; Boston, Mass. Abstract 130.

  2. Sterling T., Chaisson R., Keruly J. et al. Improved Outcomes With Earlier Initiation of HAART: Longer Follow-Up of an Observational Cohort Study. Presented at: 10th Conference on Retroviruses and Opportunistic Infections; February 10-14, 2003; Boston, Mass. Abstract 570.

  3. Ferrer E., Santamarina E., Santin M. et al. Progression to AIDS/Death Is Higher in Patients Initiating Antiretroviral Therapy With CD4 Counts Below 350 Cells/µL. Presented at: 10th Conference on Retroviruses and Opportunistic Infections; February 10-14, 2003; Boston, Mass. Abstract 910.

  4. Sterne J.A.C., May M., Costagliola D. et al. Does It Matter Where You Came From? Prognosis of Patients Starting Potent Therapy, According to Initial Response. Presented at: 10th Conference on Retroviruses and Opportunistic Infections; February 10-14, 2003; Boston, Mass. Abstract 181.

  5. Wood E., Hogg R.S., Yip B. et al. The Impact of Baseline Plasma HIV RNA and Adherence on Response to Therapy and Mortality After the Initiation of HAART. Presented at: 10th Conference on Retroviruses and Opportunistic Infections; February 10-14, 2003; Boston, Mass. Abstract 182.

  6. Robbins G., Shader R., Smeaton L. et al. Antiretroviral Strategies in Naive HIV-Positive Subjects: Comparison of Sequential Three-Drug Regimens (ACTG 384). Presented at: XIV International AIDS Conference; July 7-12, 2002; Barcelona, Spain. Abstract LbOr20A.

  7. Montaner, J.S.G., Saag M.S., Barlyski C. et al. FOCUS Study: Saquinavir QD Regimen Versus EFV QD Regimen 48-Week Analysis in HIV-Infected Patients. Presented at: 42nd Interscience Conference on Antimicrobial Agents and Chemotherapy; September 27-30, 2002; San Diego, Calif. Abstract 167.

  8. Bartlett J.A., Johnson J., Herrera G. et al. ABC/Lamivudine (ABC/3TC) in Combination With EFV (NNRTI), APV/Ritonavir (PI) or Stavudine (NRTI): ESS40001 (CLASS) Preliminary 48-Week Results. Presented at: XIV International AIDS Conference; July 7-12, 2002; Barcelona, Spain. Abstract TuOrB1189.

  9. Levy R., Labriola D., Ruiz N. Low Two-Year Risk of Virologic Failure With First Regimen HAART. Presented at: 8th Conference on Retroviruses and Opportunistic Infections; February 4-8, 2001; Chicago, Ill. Abstract 325.

  10. van Leth F., Hassink E., Phanuphak P. et al. Results of the 2NN Study: A Randomized Comparative Trial of First-Line Antiretroviral Therapy With Regimens Containing Either NVP Alone, EFV Alone or Both Drugs Combined, Together With Stavudine and Lamivudine. Presented at: 10th Conference on Retroviruses and Opportunistic Infections; February 10-14, 2003; Boston, Mass. Abstract 176

  11. Nadler J., Rodriguez-French A., Millard J. et al. The NEAT Study: GW433908 Efficacy and Safety in ART-Naive Subjects, Final 48-week Analysis. Presented at: 10th Conference on Retroviruses and Opportunistic Infections; February 10-14, 2003; Boston, Mass. Abstract 177.

  12. DeJesus E., LaMarca A., Sension M. et al. The Context Study: Efficacy and Safety of GW433908/RTV in PI-Experienced Subjects With Virological Failure (24-Week Results). Presented at: 10th Conference on Retroviruses and Opportunistic Infections; February 10-14, 2003; Boston, Mass. Abstract 178.

  13. Murphy R., Pokrovsky V., Rozenbaum W. et al. Long-Term Efficacy and Safety of ATZ With Stavudine and Lamivudine in Patients Previously Treated With NFV or ATV: 108-Week Results of BMS Study 008/044. Presented at: 10th Conference on Retroviruses and Opportunistic Infections; February 10-14, 2003; Boston, Mass. Abstract 555.

  14. Colonno R., Rose R., Cianci C. et al. Emergence of ATZ Resistance and Maintenance of Susceptibility to Other PIs Is Associated With an I50L Substitution in HIV Protease. Presented at: 10th Conference on Retroviruses and Opportunistic Infections; February 10-14, 2003; Boston, Mass. Abstract 597.

  15. Staszewski S., Gallant J.E., Pozniak A.L. et al. Efficacy and Safety of TDF DF (TDF) Versus Stavudine (d4T) When Used in Combination With Lamivudine and EFV in Antiretroviral Naive Patients: 96-Week Preliminary Interim Results. Presented at: 10th Conference on Retroviruses and Opportunistic Infections; February 10-14, 2003; Boston, Mass. Abstract 564b.


Tell us what you think of The Body's conference coverage!

This article was provided by TheBodyPRO. It is a part of the publication The 10th Conference on Retroviruses and Opportunistic Infections.

Please note: Knowledge about HIV changes rapidly. Note the date of this summary's publication, and before treating patients or employing any therapies described in these materials, verify all information independently. If you are a patient, please consult a doctor or other medical professional before acting on any of the information presented in this summary. For a complete listing of our most recent conference coverage, click here.

The content on this page is free of advertiser influence and was produced by our editorial team. See our content and advertising policies.