In early 2007, the DHHS Guidelines panel met and agreed to a complete review and update of the guidelines. Members of the panel, which include treating physicians, researchers, government officials and community members, were assigned to various subcommittees to review the document and recommend changes to the full panel. Two sets of updates have been issued: one in December 2007 and one in January 2008. This article reviews the changes thus far.
The panel made subtle, but important changes in the recommendation for when to start HIV treatment. The previous guidelines recommended that treatment be discussed when a person's CD4 count fell below 350 cells/ml, and be started in anyone whose count fell below 200. The guidelines now recommend starting treatment before a person's CD4 count falls below 350. Treatment is also recommended -- regardless of CD4 count -- for pregnant women, anyone with an AIDS-defining illness, people co-infected with hepatitis B virus (HBV) that requires treatment, and anyone diagnosed with HIV-associated nephropathy (HIVAN).
Many scientists, doctors and activists now think that HIV treatment should be started earlier than the current recommendations. Citing a lack of conclusive data, the panel was not willing to make this recommendation. Rather they included a discussion about the possible benefits and risks of starting treatment at higher CD4 counts.
The major points of the discussion were:
Factors weighing against early treatment would be:
The best time to start HIV treatment remains one of the major unanswered questions in HIV treatment. There are two significant obstacles to answering this question: one philosophical and one practical. Philosophically, there's no consensus that there even could be a single recommendation that will be appropriate for all, or even most, people with HIV. Some say that important factors needed to guide treatment decisions vary significantly from person to person. No recommendation, or set of recommendations, could possibly account for this. Others would counter that, while a person's unique situation must factor into all treatment decisions, broad recommendations can nonetheless can help guide and inform anyone making the decision to start HIV treatment for the first time.
Philosophical considerations aside, there's a practical issue with the data. The guidelines panel's recommendations are data driven. The gold standard of bio-medical research is the prospective, randomized, controlled clinical study. No such study exists on the question of when to start. Designing and implementing one is challenging. Such studies would need to be large and long-term to gain useful information. This begs the question of the applicability of the information once the study is done. For example, if a when to start' study had been done in the late 1990s, final data would just now be reportable. We would have a lot of information on Crixivan (indinavir), Viracept (nelfinavir) and Zerit (stavudine) regimens. How helpful would that be given the range of classes and individual drugs that are now available?
The panel's recommendation to start treatment when CD4 counts fall to 350 was based on data that show a clear benefit to people starting treatment at this level. The document cites two studies in particular. The first is the ART cohort -- a large, multi-centered study with over 60,000 person years of follow-up. (A person year is a statistical measure that simply multiplies the number of people in the study by how many years each person has been followed.) The ART study found that people who started treatment with CD4 counts between 200 and 350 had a much higher risk of opportunistic disease and death than those who started with CD4 counts below 200.
They also cited the SMART study -- a large, prospective, randomized, controlled trial of continuous HIV treatment vs. CD4 guided intermittent treatment. SMART was halted early when it was found that people in the intermittent treatment group had higher rates of opportunistic and non-opportunistic diseases and death.
The data were similarly strong for pregnant women, people with HIVAN, and HBV co-infection that required treatment. The use of HIV treatment by pregnant women has reduced the risk of mother-to-child HIV transmission from around 25% to less than 1% in some settings. The risk of kidney damage in people with HIVAN has been shown to be more closely related to being off treatment than either HIV levels or CD4 count.
The situation for people with HBV co-infection requiring treatment is a bit different. Many of the most common drugs used to treat HBV have anti-HIV activity as well. If a person were on HBV drugs, like Viread (tenofovir), Epivir (lamivudine) or entecavir, they run the risk of developing resistance to HIV drugs.
Project Inform believes that HIV treatment decisions should be driven by a combination of the best available data a person's unique life circumstances and medical history, and personal preference. We also recognize that in most diseases, particularly infectious diseases like HIV, earlier treatment usually leads to better treatment outcomes. There is no evidence to suggest this would not be true for HIV disease.
The main factors leading to delaying treatment in HIV disease have been the demands of lifelong treatment, concerns over side effects, and the fear of drug resistance. With over two dozen HIV drugs on the market today -- many of which are simpler to take and seem to have fewer side effects than the older generation of drugs -- this situation may be changing.
In addition to protecting the health and well-being of the person living with HIV, ARV treatment also reduces the risk of HIV transmission. While treatment decisions should be primarily guided by the needs of a person living with HIV, the health and well-being of the larger community can also benefit. This could be particularly important for people in relationships with an HIV-uninfected partner, or people with multiple sexual partners.
The recommended first line treatment paradigm remains the same: a backbone of two NRTIs plus either a boosted PI or an NNRTI. The list of drugs appropriate for first line use was changed significantly. The fixed-dose combination pill Combivir was downgraded from preferred to alternative based largely on lipoatrophy data. Epzicom was upgraded from alternative to preferred based largely on the availability of HLA testing that reduces the fear of abacavir hypersensitivity reactions. Several protease inhibitors changed position as well.
Below is a summary table. Note that this is only for first line use.
ddI + 3TC or FTC
The decision to move Combivir from preferred to alternative was based largely on data that showed a higher risk of lipoatrophy, or loss of fat (usually in the face, arms, legs and buttocks) in people taking regimens that contain zidovudine (AZT), which is one of the drugs in Combivir. Of all the NRTIs in wide use, zidovudine carries the highest risk of lipoatrophy. The combination of Videx EC (didanosine, ddI) + Epivir is also listed as an alternative.
The decision to move Epzicom from alternative to preferred was due to the development of a genetic test, called HLA testing. This can fairly accurately predict a person's risk of developing a serious allergic reaction to abacavir, which is one of the drugs in Epzicom. This reaction, called abacavir hypersensitivity reaction or HSR, has been a major deterrent for many people to use this drug. Recently several studies, including PREDICT and SHAPE, have shown that a genetic variation, called HLA-B5701, is powerfully predictive of a person's likelihood of developing abacavir HSR. The updated guidelines include guidance on using HLA testing; emphasizing it should not be used as a substitute for clinical vigilance.
A recent finding reported elsewhere in this issue has reopened the question of Epzicom's position in the guidelines. As we reported here, the large cohort D:A:D study found an increased risk of heart attack (myocardial infarction) for people taking abacavir and another NRTI, Videx. The overall risk was still small, but the difference was considered statistically significant. This finding was surprising. Some have questioned the significance, pointing out that other studies looking at abacavir haven't found an increased risk of heart attack. Whether this new finding leads the guidelines panel to change their recommendations on using abacavir remains to be seen.
The Norvir (ritonavir) boosted protease inhibitors Lexiva (fosamprenavir) and Reyataz (atazanavir) now join Kaletra as preferred options for first line protease inhibitors (PIs). Boosted Invirase (saquinavir) was moved from the not-recommended to alternative category. These changes resulted from a series of head-to-head studies of these PIs vs. Kaletra, which had been the only preferred first line PI. These studies show that each of these PIs are fairly closely equal to Kaletra (the technical term for these studies is non-inferiority) for people taking HIV drugs for the first time. Boosted Invirase was not put in the preferred category because these data, from the GEMINI study, were still preliminary and not yet published in a peer-reviewed journal.
Unboosted Lexiva and Reyataz, along with once daily Kaletra and once daily Lexiva were also added to the alternative category. The data on these regimens are not quite as strong as for those in the preferred category. However, people may choose those to avoid Norvir, or for more convenient dosing.
Unboosted Invirase, Viracept (nelfinavir), Aptivus (tipranavir) and Prezista (darunavir) are not recommended for first line treatment. It is possible that this will change for Prezista as preliminary data from a head-to-head study vs. Kaletra for first line treatment has shown promising results. The guidelines point out that these data are preliminary, and the dose being studied is not commercially available.
Sustiva (efavirenz) remains the sole preferred NNRTI for first line use. Viramune (nevirapine) is listed as an alternative option, largely due to concerns over liver toxicity. Rescriptor (delavirdine) and Intelence (etravirine) are not recommended for first line treatment.
Several other drugs are not recommended for first line treatment. In most cases, not including the integrase inhibitor Isentress (raltegravir) and fusion inhibitor Fuzeon (enfuvirtide) is due simply to a lack of data for the drugs used in this way. In the case of the entry inhibitor Selzentry (maraviroc), it was due to results from a head-to-head study vs. Sustiva, where Selzentry failed to prove as good as Sustiva.
The overall affect of these changes is that people have more options for constructing their first HIV drug regimens. This is good news for people living with HIV. More options mean a better chance to construct a well studied regimen that fits a person's unique needs. More changes might be forthcoming, as companies continue to study their drugs against those in wide use for first line treatment.
The section on treatment for people with extensive experience taking HIV drugs dose not include a simple chart of recommended, alternative and not recommended drugs. Treatment decisions for this group are more complex and must be guided by treatment history, resistance testing and other factors.
Several important changes were included in this update. The first is information on the newly approved drugs Intelence, Isentress, Prezista and Selzentry. All of these have been extensively studied in treatment experienced people. The guidelines include a summary of what is known about each of these new drugs, including some data from studies.
A second change was subtle but significant -- strengthening the treatment goals' section to include language on undetectability and avoidance of serial monotherapy. The goal of reducing anyone's HIV levels to below the limit of detection has always been there. Now the availability of multiple new, potent drugs that when used in combinations are likely to get almost anyone's HIV levels to undetectable. For years, only one new drug became available at a time. This meant that people with few or no treatment options were forced to take each new drug as it became available, often as the only fully active drug in their regimen -- a situation called serial monotherapy. The approval of four potent new drugs, all either from new classes or able to overcome resistance to older drugs has given almost everyone a window of opportunity to combine multiple new, fully active drugs and hopefully reduce their HIV levels to undetectable.
New language was added on the potential for simplification', or reducing the number of drugs a treatment experienced person is taking, when they are on a suppressive regimen. For many years, the approach to treating people with multi drug resistant HIV was to treat with as many drugs as could be tolerated, sometimes called the kitchen sink approach. Some research has shown that when treatment experienced people are on stable suppressive regimens, they can sometimes safely reduce the number of drugs in their regimen -- to reduce the risk of side effects and improve quality of life.
Third, more information was added about the risks of treatment interruptions, flowing mostly from SMART, but including other studies like DART, PART and TRIVICAN. The bulk of the data on treatment interruptions show them to be risky. This is especially true for people who have ever had an AIDS-defining illness or CD4 count below 200. Recent data presented at CROI reinforces the growing unease with treatment interruptions.
The guidelines changed some of the language around diagnosing and treating acute HIV infection. There is a growing body of evidence on the importance of this earliest phase of HIV disease. It is thought that people in the acute phase of HIV infection -- when their immune systems have not yet mounted a full response -- are at a particularly high risk of passing HIV onto others. For one thing, people are very likely to be unaware of their HIV status at this point, and therefore may be less likely to practice safer sex and drug use practices. Some evidence also shows that some of the things that happen during the acute phase, like HIV's invasion into the gut, may predict a person's long-term outcomes. The guidelines added language to help physicians identify people who might be in acute HIV infection, and stressed the importance of testing for NNRTI drug resistance which is more commonly transmitted than other kinds of resistance.
Several changes were also made to the section on treatment for people co-infected with Mycobacterium tuberculosis or TB. TB remains a common co-infection for people living with HIV, especially those in jail or who are homeless or living in shelters or other crowded situations. There is a growing problem with multi-drug and extensively drug resistant TB as well. The guidelines added recommendations on treating active and latent TB, as well as drug sensitive and drug resistant TB.
The Guidelines are meant to help HIV treating physicians better manage care and treatment decisions. The document is under constant review by a dedicated volunteer group. The panel strives to incorporate the best evidence in an ongoing fashion. The stronger the evidence, the stronger the recommendations.
Sometimes changes in the guidelines can appear to lag behind the cutting edge. This is sometimes due to simple time constraints. Other times, it has to do with strength of evidence. The process of reviewing and updating the guidelines is both fascinating and daunting. The large number of drugs and the many studies of their use require constant vigilance by panel members.
No set of guidelines, however well researched or debated, can ever replace the clinical insight of an HIV experienced clinician. One of the main audiences for the guidelines is treating physicians who do not specialize in HIV, or who only treat a small number of people living with HIV. The guidelines are just a tool to help physicians and people living with HIV make better decisions around their care and treatment. It is a living document, undergoing constant scrutiny and revision.
Currently, four community members are included in the panel to ensure that the user's perspective is always part of the discussions. Paul Dalton, Project Inform's Director of Treatment Information and Advocacy, is a community member of guidelines panel and is the author of this article. This article was not written on behalf of the Guidelines Panel. The entire document can found here.