How well are doctors managing the complexities of treatment? Not very well, suggest the experts, who don't blame doctors as much as the difficulties posed by HAART. Even experts admit the current federal guidelines for HIV treatment serve as a good, but rough, guide for managing patients. In reality, there are so many novel drugs and factors to consider that many doctors are struggling, and failing, to make sound treatment decisions on issues including dosing, toxicity, drug tolerance, and potential drug interactions.

A recent study conducted by pharmacists at Albany Medical Center Hospital in New York highlights the problem. Over 31 months, doctors there prescribing HIV antiretroviral drugs made a hundred errors in 81 patients. Almost three quarter (73 percent) of the errors were ranked as serious or severe, and others were considered clinically significant (21 percent). That's scary.

The most common prescribing errors were underdosing (46 percent) and overdosing (34 percent). Half the errors (50 percent) involved patients taking protease inhibitors, and many of the rest involved the better-known nucleoside analog drugs like AZT. A closer look shows that these errors increased from 2 percent in 1996 to 14 percent in 1998-paralleling the arrival of protease drugs.

The study was one of several reports on managing HIV, a hot topic at the recent ICAAC infectious disease meeting in San Diego. There, doctors and advocates packed themselves like sardines into an interactive workshop that reviewed the guidelines on when to start and switch therapy, when to stop using HIV prophylaxis, and what to choose for salvage therapy when a given regimen has failed. The session was chaired by two experts in the field, Michael Saag of the University of Alabama at Birmingham, and David Cooper, an Australian researcher. Both stressed the importance of individualizing therapy, or tailoring existing guidelines to fit the patient-not the reverse. As Saag pointed out, the time to begin therapy may simply be when the patient is ready. Trying to initiate a complex, potentially toxic regimen to someone who's not prepared is a recipe for disaster, he said, since missing doses or poor adherence can lead quickly to drug resistance and failure.

To get an idea of how well physicians could make on-the-spot treatment decisions, session participants were asked to interactively give answers to test cases. A majority of participants were clinicians with large HIV caseloads who, not surprising, disagreed quite consistently about what to do when different factors were added to sample test scenarios. These factors included changing viral-load and T-cell counts, the onset of infections, drug resistance, and side effects. As several speakers pointed out, right now there's no single correct answer when it comes to treatment, only the best guess. Or as one doctor joked, "It's the blind leading the blind-down a dark and crowded street."

Among the pearls of wisdom was this advice: Expect failure. Given the realities of HIV treatment, said Saag: "If we clinicians today aren't prepared for failure, we're not doing our job." That advice applies to individuals with HIV as well. Being prepared for problems means anticipating ways of overcoming them, which increases one's chances of success with a given course of therapy.

Based on federal guidelines, the goal of antiviral combination therapy remains to reduce HIV activity to as low a level as possible for as long as possible. There's increasing evidence that driving viral load to below 50 copies increases the chance of keeping the virus down over time. Other studies show even a threefold decrease in viral load has health benefits. Beyond that, the best shot is one that looks carefully at a patient's medical and treatment histories, tolerance of a given regimen, and factors that might be influencing the drugs' ability to work, including nonmedical issues like cost and convenience.

In a recent survey of problems affecting patient adherence, Calvin Cohen, research director of the Community Research Initiative of New England, cited a dozen reasons why people failed to religiously follow their drug regimens, including emotional and psychosocial factors such as depression or a distrust of Western medicine in general. He also noted that "these circumstances may change over time and therefore should be reassessed frequently." Several surveys show the most common reason for poor adherence is forgetting to take one's pills. That's a problem, says Cohen, that can be addressed by simple strategies like getting patients to use a portable alarm.

The point is, just as there's no blueprint for how therapies will work, there's no blueprint for how a given patient will manage HIV therapy. And as HIV activists point out, the same can be said of physicians. Some are more on top of things than others. That's why the federal guidelines recommend seeking care from a doctor who's experienced with HIV treatment.

Looking toward the future, these issues mask a more serious debate that's erupted over long-term management of HIV, one that's not likely to be resolved quickly. While the current "hit early, hit hard" model works well to drive down viral load in the short term, is it creating more problems over time? That's a concern put forth by HIV immunologist Jay Levy of the University of California at San Francisco, a well-respected researcher who argues that our current approach to HIV is based on the aggressive chemotherapy model used in cancer-a method that may not work as well when drugs have to be taken for a long time.

In a September letter to the British medical journal Lancet, Levy challenged his colleagues to reconsider the risk and benefits of early use of HIV drugs in all patients. He's worried that the longer some people with chronic infection take toxic HIV drugs, the greater their risk of developing side effects, as well as problems with adherence and drug resistance. In some cases where people are healthy and their viral loads are stable, delaying treatment might be a better choice.

Levy's critique of early treatment doesn't apply to people who are treated immediately after exposure to the virus; for this group the benefits are clear, he says. The same holds for people with advanced HIV disease. But for people who are healthy and have been infected for months or years, early use of drugs may limit the ability of their immune system to generate a potent response to the virus. And by giving drugs early, we may be pushing HIV to mutate faster than it might without the pressure of drug selection, says Levy; that's a Darwinian concept that means the virus mutates its genes to escape being killed by drugs.

Levy's ideas aren't new; we've always assumed there would be risks associated with these effective but toxic therapies. Topping the list of side effects is "protease paunch"-a disfiguring redistribution of body fat, called lipodystrophy, that's linked to metabolic and lipid problems and which turns some people's blood to near butter-resulting in skyrocketing cholesterol, triglyceride, and glucose levels that increase the risk of heart attack, diabetes, and damage to the pancreas. Initially linked to Crixivan use, lipodysrophy has been reported in many people on protease inhibitors, including Fortovase and Norvir, and in some people taking other HIV drugs.

In a study presented earlier this year by Cooper's Australian team, up to 64 percent of their patients on triple-drug therapy developed lipid problems. New data from ICAAC shows that the triglyceride levels in some people who develop lipodystrophy may be many times beyond the limit of what's considered normal. That's a clear health risk.

A new study by A. Veny and colleagues in Spain also presented at ICAAC backs Levy's concerns about a potential cumulative risk associated with long-term use of HIV therapies. Veny's group followed 422 patients on HIV therapies, 158 of them taking protease inhibitors. After six months, those taking protease drugs had a 3.2 percent risk of developing lipid problems, but at 18 months the estimated risk had jumped to 29 percent. By 30 months, patients had a 75 percent risk of having lipodystrophy-a figure rivaling the very high numbers seen in Cooper's Australian group. More research is needed, and quickly, to establish whether this is a common trend.

While some researchers worry privately that Levy's public airing of an opinion will discourage people from initiating or staying on HAART, others see it as a welcome litmus test of whether and how we might need to change or modify our approach to long-term treatment as we move forward. In October, the treatment advocacy group Project Inform published an excellent review of Levy's concerns, which the organization felt didn't differ that dramatically from those of most scientists, or even from federal guidelines. The current HIV guidelines do say that delaying treatment is an option for those with a high, stable T-cell count and low viral load (see Guidelines). As Project Inform concluded, "Nothing in Dr. Levy's letter would suggest that people who have already started treatment should stop." Instead, for those who are thinking of starting a regimen, "Levy's points serve as an important reminder that treatment is a serious decision that should be entered carefully and only when the patient is fully informed."

To date, Levy says he's gotten "almost all positive responses" to his letter from colleagues, though he frankly admits, "People who disagree probably won't write to me." Taking it to the next step, he thinks the only way to get answers fast to these critical questions is to do some experiments. One might be a clinical trial to study the risks versus benefits of delaying early treatment for chronically infected people. Another approach would be a retrospective study to see what's happened to people who haven't gone on therapy right away. A third avenue is one already in the works: taking people off therapy once they have no detectable virus, and possibly using alternative approaches like immune boosters or a vaccine to help the immune system control a low-level viral infection (see "Slouching Toward A Cure?").

Looking to the future, Levy's ideas point to a very different scenario for long-term treatment. So go on, jump into the debate.