Ever since blue-ribbon panels and big AIDS meetings began inviting Charles Flexner to speak, the wider HIV community has found out what his Johns Hopkins colleagues already knew: Flexner possesses not only an acute understanding of antiretroviral pharmacokinetics, but also a rare knack for explaining this complicated field to the uninitiated, and for warning us what those complications may portend.

This reporter first heard Flexner at a February 28, 1996, gathering of the FDA's Antiviral Drugs Advisory Committee, which had been mustered to decide which dual nucleoside combinations outmuscle nuke monotherapy. Specifically, the Committee was charged to determine whether zidovudine (ZDV) plus didanosine (ddI) does more than ZDV or ddI alone, then whether ZDV plus zalcitabine (ddC) beats out ZDV or ddC alone.

As quaint as that exercise may seem today, at the time the assembled experts struggled to assimilate several complex analyses of ACTG 175, Delta 1 and 2, and CPCRA 007 that showed--statistically--that ZDV and ddI offered little advantage over ddI alone. As the Committee edged nervously toward the notion that, well, maybe ZDV/ddI isn't so great, Flexner reminded colleagues that in a few minutes attention would turn to ZDV/ddC, which the statisticians had just showed was a whole lot better than ddC alone or ZDV alone, but not quite up to ZDV/ddI. Did the Committee really want to entomb itself with the ZDV/ddI-versus-ddI data, Flexner wondered. "A foolish consistency," he cautioned, "is the hobgoblin of little minds--and large committees." That little ray of wisdom turned out to be the turning point that saved the Committee from an embarrassing decision.

Flexner now practices the science, and art, of HIV medicine as Associate Professor of Medicine (Clinical Pharmacology and Infectious Diseases) and Associate Professor of Pharmacology and Molecular Sciences at The Johns Hopkins University School of Medicine in Baltimore. He gathered up a BS in biology at Stanford and an MD at Hopkins before working at Stanford University Hospital, the Laboratory of Viral Diseases at NIH, and finally Hopkins. An active member of the AIDS Clinical Trials Group (ACTG), Flexner chaired the ACTG Pharmacology Committee from 1993 to 1995. He now serves on the International Scientific Review Committee of the International AIDS Society.

Flexner's perceptive analyses of antiretroviral developments can be read regularly at the Hopkins HIV site (http:// www.hopkins-aids.edu/) and often in Medscape's HIV meeting coverage (http://hiv.medscape.com/). With Stephen Piscitelli, PharmD, he devised and updates two interactive Web sites, one of AIDS drug interactions and another on medication scheduling (see "What's on the Web about drug-drug interactions?"). His online monograph, "Pharmacokinetics for Physicians--A Primer," is a must-read (see "What's on the Web about drug-drug interactions?").

"Life is complicated," Flexner said at a 1998 Conference on Retroviruses lecture. "Pharmacology makes it more so." But HIV pharmacology seems a little less complicated, thanks to his work.

Therapeutic drug monitoring: not yet--and not ever?

JIAPAC: Proponents of therapeutic drug monitoring (TDM) list several reasons why measuring plasma levels of protease inhibitors (PIs) may prove to be a valuable clinical tool.

1. PI concentrations correlate with antiviral effect, and concentrations vary from person to person.
2. Elevated PI levels may warn of toxicity.
3. TDM may give hints of poor adherence.

But you've been skeptical that TDM will have a role in the HIV clinic. Why?

Flexner: First of all, a number of studies do correlate concentrations of protease inhibitors and viral load^1-4--either extent of viral load response, or prediction of who's more likely to fail therapy. No one should be surprised that such a correlation exists. In fact, if such a correlation did not exist, we would have to question whether protease inhibitors were really active drugs, because a basic tenet of pharmacology is that there's a quantitative relationship between exposure to the agent and either benefit or toxicity.

The more complex question is whether having that information can lead to any strategies for improving treatment outcome by individualizing the dose of a drug in a given patient. And that's something that has not been very well examined. If you look at the history of therapeutic drug monitoring in general, there are precious few settings in all of medicine in which therapeutic drug monitoring is used routinely. And in most settings it's used to avoid drug toxicity, not to ensure that the drug is achieving its desired effect.

The reason we don't need therapeutic drug monitoring most of the time is because the antiretrovirals developed have a therapeutic index large enough to make individualizing the dose largely unnecessary. If you favor therapeutic drug monitoring, you must believe that antiretrovirals as a class are fundamentally different from all other drugs used to treat chronic disease. Even if you worry that the therapeutic indices of some current agents are too narrow, you have to admit that we're in the midst of an evolution away from drugs with relatively narrow therapeutic indices, and toward second- and third-generation drugs that have broader therapeutic indices and better pharmacokinetic profiles. Therapeutic drug monitoring is less likely to be useful for these newer agents.

JIAPAC: But what about TDM's value for the antiretrovirals we already have?

Flexner: Most of the data on correlation between drug levels and outcome involve indinavir. Indinavir probably has the worst therapeutic index of all the approved protease inhibitors in the sense that it has a short half-life and variable pharmacokinetics. Indinavir is actually quite a potent antiretroviral drug, but it has a pharmacokinetic profile that could stand some improvement.

In our clinic we're not using indinavir as a single protease inhibitor much anymore. The current trend here, and in many other clinics, is to combine indinavir with ritonavir to slow its clearance and enhance its pharmacokinetics. And when you do that, the therapeutic index of indinavir looks quite good. Most people appear to respond to that combination, and most appear to tolerate it reasonably well. Then one has to question whether all of these data collected from patients in indinavir monotherapy trials are relevant to the way we're using the drug.

It seems increasingly likely that the most common cause of treatment failure is noncompliance or partial adherence, not inadequate pharmacokinetics. Therapeutic drug monitoring does little to address compliance issues. So it seems to me that patients would be better off and society would be better off if we put more energy into developing drug regimens that are easier to take and easier to tolerate and easier to adhere to, and less energy into worrying about whether concentrations of an antiretroviral in an individual can be used to modify the regimen for that individual.

I'm sympathetic with the focus on antiretroviral drug concentrations because I think they're important to understand. But I think the place they're most important is in drug development, not after the drug is out there in the clinic. And even if therapeutic drug monitoring might seem to make sense in selected situations, with the speed at which antiretroviral therapy is changing, today's studies may be irrelevant to tomorrow's therapies.

JIAPAC: Are there specific instances when TDM might make sense, for example, for someone in whom compromised liver function might push PI levels into the toxic range?

Flexner: No one has been able to define the correlation between protease inhibitor concentration and toxicity. We've mainly focused on associating low concentrations with poor virologic outcome. In the long run the toxicity issue may loom larger, and I think it already has, but right now there's not a whole lot of data about that.

Most of the data suggest that patients with mild to moderate liver disease don't have clinically significant changes in drug clearance and don't require dose modification for most antiretrovirals. But I think patients with severe or end-stage liver disease are in a different situation and may require dose adjustment.

The biggest problem is that if you're adjusting the dose to achieve a target concentration, what's the concentration range you want to achieve? No one has been able to define a concentration range adequately--either the low end or the high end--for any antiretroviral.

Keeping clinical tabs on adherence and resistance

JIAPAC: What's your opinion on other kinds of patient monitoring that are talked about a lot these days? One is coming up with a way to monitor adherence better, and the second is monitoring for resistance-associated genotypic changes or for loss of susceptibility to an antiretroviral.

Flexner: I'll address each of those in turn because they're really very different.

I think there may be a role for passive adherence monitoring, where you simply keep track of how many doses of drug a patient is taking over time and then provide that information to the patient to help improve adherence. Perhaps patients can time their pill taking to some daily activity rather than trying to take doses by the clock, or some set of external cues could make it easier for them to take their drugs as prescribed.

There may also be a role for active adherence monitoring, that is, for using devices that warn patients when they're late taking their dose or warn them if they try to take the dose too early. There are devices out there already that have that capacity. Studies need to be done to show that they actually provide benefit, and we don't have such studies yet.

I don't think you need this kind of intervention for everybody because there are many patients who take their medications quite nicely without any special reminders. And it is a minor inconvenience to have to carry your pills around in some kind of adherence-monitoring device.

With respect to resistance testing, I think it will have a role in developing more effective regimens for patients, especially as resistance become more prevalent in the population. The problem we have right now, in my opinion, is that the phenotypic and genotypic tests available have insufficient sensitivity and specificity to answer all our clinical questions.

Resistance testing is most useful when you find that a patient has a mutation or set of mutations that confers high-level resistance to one or more drugs. But if the genotypic or phenotypic test is negative, that's not very helpful because those tests are only sensitive enough to pick up virus variants that constitute 10 percent or more of the circulating pool. So if virus is present at a low level, say 5 percent of the circulating pool, and the patient is placed on a drug that would select for that resistant virus, the therapy is still likely to fail.

Another problem with phenotypic testing is how we define resistance. The two available resistance assays define resistance as a 2.5-fold or 4-fold increase in the IC₅₀, that is, the concentration of drug required to inhibit viral replication by 50 percent. If resistance to a drug increases by a factor of 4 but blood levels in the patient are still 20 times higher than the IC₅₀ of that resistant mutant virus, the patient should still respond to treatment with that drug, if the patient is taking the drug. And so phenotypic testing may produce false positives and suggest that a patient's virus is resistant to a drug when in fact it's not, because of failure to take into account the achievable drug concentration in the patient.⁵

If you look at antibiotic sensitivity testing, for the past several decades we have made a conscious effort to tailor our reports of antibiotic sensitivity to achievable drug concentrations in patient plasma. So a "resistant bacterium" means that bacterium requires inhibitory concentrations of an antibiotic that are not easily achieved throughout a dosing interval in a human being. We need to do the same kind of thing for HIV phenotypic resistance testing.

Keys to cutting the risk of ominous drug interactions

JIAPAC: Because so many antiretrovirals are licensed now, they're getting combined in ways that haven't been studied. When a trial combines two drugs that haven't been looked at together in a large number of patients, it's not uncommon for negative interactions to get uncovered. What can clinicians do to be on guard for such unforeseen interactions when they're putting together a new regimen for people who have gone through two or three courses of therapy?

Flexner: The first general principle for the clinician is to try to prescribe a combination that you're familiar with, preferably one for which there's substantial clinical data or enough clinical data for you to understand the benefits and the risks of the regimen.

However, we all encounter situations where we have to use a combination for which there is not a lot of clinical data, but that regimen seems to make the most sense for the patient sitting in the office at the time. When we do this, we need to warn the patient that there could be some unexpected interactions between these drugs or unexpected toxicities, because this is a combination that has not been widely used. We also have to monitor those patients more frequently by bringing them back to the clinic more often than we would if they were taking a well-studied regimen.

Of course clinicians also have to keep up to date on drug-drug interactions. [See "What's on the Web about drug-drug interactions?"]

JIAPAC: Stephen Piscitelli's finding that St. John's wort greatly lowers levels of indinavir⁶ got a lot of attention because it underscores a real problem. There are lots of people taking these so-called natural remedies, although very little is known about their interactions with antiretrovirals. What should clinicians be telling their patients about these remedies? Should they be telling them, "Don't take any of this stuff"?

Flexner: I think telling your patients "Don't take any of this stuff" is probably the wrong approach, because we know that many of our patients feel they derive a benefit from "natural remedies," health foods, or "neutriceuticals." If you tell them "You can't have that," they'll probably go out and take it but not tell you that they're taking it.

What I would promote instead is a therapeutic alliance between the physician and the patient where we maintain good communications and honesty. Otherwise we'll both be in the dark and not be able to get the most out of our relationship. So I would encourage patients to tell me whatever they're taking, warn them of the possibilities of unexpected toxicities or even drug interactions involving complementary medicines, and keep an open mind about the potential role that health foods, and all foods for that matter, may play in unanticipated toxicities or unanticipated poor response to a therapeutic regimen.

A year ago, if I had a patient on indinavir who wasn't responding the way I expected, and the patient told me, "Oh, doc, by the way, I'm taking St. John's wort," I would have thought nothing of it. I probably would have thought this is a noncompliant patient. Today I know that, because this person was taking St. John's wort, indinavir levels would have been substantially decreased, and that's why this patient wasn't responding to therapy.

So keep an open mind, and be prepared to suggest to patients who are aren't responding adequately to treatment that the problem could be caused by one or more of their health foods. Tell patients up front that we might want to consider "nutriceuticals" just as we would a prescription drug. If a possible interaction problem arises, we would withdraw those agents individually and see if that doesn't fix the problem. Then we could add back the ones that didn't appear to be causing the problem.

Compartment Penetration: Is It Really So Important?

JIAPAC: Pharmaceutical companies apparently see an advantage to proving that their antiretrovirals can penetrate compartments like semen and cerebrospinal fluid (CSF). But you recently reviewed some compartment research and proposed that differential penetration might not mean much clinically.⁷ Why not?

Flexner: Up until now we've largely depended on empiric observations about compartment penetration, and there are conflicting data out there on whether differential penetration makes any difference clinically. Why are the data conflicting? First of all, many of the studies have involved inadequate sampling to quantitate the extent of penetration into CSF or semen compared with plasma. We're beginning to fix that problem and more data are emerging, so we now have a better quantitative picture of the relative penetration of drugs into cerebrospinal fluid and semen.

Some drugs actually appear to be concentrated in semen, that is, their semen concentration is higher than their blood plasma concentration. Almost all drugs, no matter what class, have lower concentrations in the CSF than in plasma. That's due in part to the blood-brain barrier, in part to the different proteins in those two compartments, and in part to the physicochemical properties of the drug itself.

But if penetration into the CSF is an important indicator of the ability of the drug to suppress virus in the CSF, one might predict that there would be a lot more cases of HIV brain disease in patients who had adequate plasma concentrations of drug but low CSF concentrations. Even indinavir, which has the highest CSF penetration of any of the protease inhibitors, achieves CSF levels only 20 to 40 percent of those in plasma over most of a dosing interval. And if you think about it, a 5-fold overall reduction in indinavir dose systemically would likely lead to treatment failure.

Most studies show that when viral load falls in the plasma, it falls in parallel in the CSF. Now the argument has been made that maybe the viral RNA measured in the CSF isn't coming from the brain, it's coming predominantly from the plasma, and that's why it looks like it falls at the same rate as the plasma virus falls. However, if CSF penetration is inadequate, we ought to be seeing more neurologic treatment failure with drugs that penetrate the central nervous system (CNS) poorly. And the bottom line is that we haven't seen that: ddI penetrates the CNS poorly, but we're not seeing a lot of ddI resistance compared with other nucleosides. All of the protease inhibitors, except indinavir, penetrate the CSF poorly, and yet the only study I've seen that suggested a differential failure rate for one protease inhibitor compared with the others centered on saquinavir, and saquinavir is no worse at getting into the CSF than ritonavir, nelfinavir, or amprenavir.⁸

What's happening to these drugs is complicated, and there's a lot more going on than simply whether or not the drug gets into the cerebrospinal fluid. Empiric observation suggests that brain disease in patients on HAART is a not an epidemic problem. Until someone develops a drug that penetrates the CNS substantially better than others, and shows that that drug is associated with a better long-term response rate and less resistance simply because of its CNS penetration, the clinical importance of CNS penetration will remain an unanswered question.

Switching from PIs to NNRTIs: What do the data say?

JIAPAC: I heard a talk you gave when protease inhibitors were first coming into widespread use. One of the points you made was that, because protease inhibitors were studied so quickly and approved pretty quickly, there might be unforeseen side effects that would become evident only when the drugs got used by a larger population. And that turned out to be true. Now a lot of clinicians are worried about the protease inhibitors and they're switching their patients to nonnucleosides (NNRTIs), so now nonnucleosides are getting used by a larger population. Might the same thing happen with that class of drugs?

Flexner: I hate to be right about this sort of thing. If you put a drug into practice without a lot of data, you're going to make some discoveries down the road that you didn't anticipate early on. I think that's inevitable. Whenever possible, I would base a treatment decision on available data. It may turn out that switching patients from protease inhibitors to NNRTIs will result in amelioration of some PI-associated toxicities, like lipodystrophy syndrome. But it could also be worse to switch patients from PIs to NNRTIs.

We need more studies, we need more data. In the interim, I would be very cautious about switching someone off a protease inhibitor to avoid lipodystrophy if that person was tolerating the regimen and if the viral load was suppressed. The apparent risk of cardiovascular disease induced by the kinds of lipid changes associated with protease inhibitors is relatively small, particularly compared with the risks of having a high viral load.

We know that NNRTI-containing regimens and nucleoside-containing regimens are now associated with lipodystrophy syndromes, although the individual components of the syndromes may be different with different regimens. But we don't know what the relative risks of those syndromes are right now, and each of them may be risky in different ways. I suggest we focus on the facts: Our enemy is still the virus, and regimens that suppress viral replication completely for the longest time are going to do patients the most good. The health risks of recurrent HIV replication appear to be substantially higher than the health risks of elevated cholesterol and triglycerides for most of our patients.

JIAPAC: But there have been a few small studies showing different indicators of possible cardiovascular disease down the road--carotid intima media thickening, arterial plaque studies, and so on. Have those preliminary findings changed your opinion about how readily PI-induced hyperlipidemia might progress to heart disease?

Flexner: I think all of those studies are hard to put into clinical context because they've involved highly selected patients who may have been at the highest possible risk of developing cardiovascular disease.

Until recently, there has been little attempt to modify other cardiovascular risk factors in people taking antiretroviral combinations. Ten years ago I would never advise my HIV-infected patients to quit smoking because the virus was going to kill them, I thought, long before their cigarette smoking did. Today the tables are completely turned. In one study life expectancy for a 40-year-old man who's taking HAART and has a fully suppressed viral load is reported to be at least 30 years and may be nearly normal now.⁹ So now we have to start thinking about all the other things that might contribute to the risk of heart disease. In particular, cigarette smoking and lifestyle modification--diet and exercise--come to the front. Whereas in the past we kind of told our patients to enjoy themselves, now we're going to have to rethink that.

Granted, antiretroviral therapy and/or HIV infection might be causing atherosclerotic risks in other ways. Maybe simply being HIV infected is a new cardiac risk factor or an atherosclerotic risk factor that we have to identify. But I would be very cautious interpreting data that show that patients on HAART get coronary artery disease and get carotid plaques because, quite frankly, we now have a lot of middle-aged men and women on HAART and--guess what--middle-aged men and women get coronary artery disease and have carotid artery plaques.

So I think for me the important thing is to begin to put relative risk and benefit into context for the individual patient, rather than stating, "We're now seeing cardiovascular risk factors in patients on HIV therapy; therefore let's do everything we can to get them off this therapy and get them onto something that doesn't carry those risks."

If you show me an effective HAART regimen that carries no cardiovascular risks, then I'll start to listen to that kind of talk. But until then, I think, we need to focus on what's best for the patient in the here-and-now. And for most of my patients, what's best for them in the here-and-now is to keep the virus suppressed for as long as possible.

Sorting through the side effects of nukes, nonnukes, and hydroxyurea

JIAPAC: Now, of course, nucleosides have been implicated in lipodystrophy. But when I look at all of the studies that have been presented so far, it's hard for me to see if one particular nucleoside is any worse than another. How do you appraise those data?

Flexner: Again I think we need to be cautious about how those data are interpreted because nucleosides were approved at different times and have been used in different patient populations for different lengths of time. There's been a lot of bad press about stavudine (d4T), but we need to keep in mind that for a while we have tended to use stavudine for our sickest patients because we thought it was their best option as a nucleoside analog. So you need to sort out what's cause and what's effect when you begin to look at nucleoside toxicity.

We're all renewing our interest in nucleoside toxicity. We went through a time when we thought that protease inhibitors were relatively nontoxic, and we now know that that's not the case; we went through a time when we thought that NNRTIs were nontoxic, and we now know that that's not the case; and we went through a time when we thought that nucleosides were the least malignant form of therapy that we had available, and we now know that that's not the case--although we were beginning to understand more clearly the mechanism of zidovudine toxicity before protease inhibitors and NNRTIs came along. The first description of nucleoside-induced lactic acidosis and steatosis syndrome predated the approval of the first protease inhibitors.

Just as we need to be cautious about how we define lipodystrophy syndrome, and what it is and how you respond to it, we need to be cautious about how we define nucleoside toxicity and how you respond to it. There have been a lot of reports of elevated lactic acid levels in patients on nucleoside analogs. Most of those studies are hard to interpret because the lactic acid levels were not properly drawn--they're venous lactates that were not placed immediately on ice and that were drawn from a site with a tourniquet applied, which can artificially elevate lactate levels. There are also reports of elevated anion gaps with nucleoside therapy, trying to equate that with lactic acidosis. But that's a very nonspecific way to look at elevated lactates.

Most of us would say that lactic acidosis, steatosis, and pancreatitis--the really severe complications of nucleoside therapy--are still quite rare, though we need to have a renewed vigilance for them. But don't go overboard in thinking that everybody who comes into the office with a cough, a cold, or a bellyache has a nucleoside lactic acidosis syndrome.

I think we're going to see data in the next several years that will help us sort out the relative toxicities of different nucleoside combinations and which ones represent the highest risk to the patient. Right now I think it's a little hard to know how to interpret the cross-sectional studies that have been reported. The principle that should still be applied is to give the patient the drug combination that they're most likely to tolerate, most likely to take, and that will most likely suppress their virus below the detectable limit for as long as they take their drugs.

JIAPAC: Some people I talk to feel that nonnucleosides are still pretty clean as far as side effects profiles go, at least compared with protease inhibitors. But you've mentioned NNRTI side effects a couple of times now. What do you see as the biggest toxicity problems with NNRTIs?

Flexner: I think there are several areas where nonnucleosides could stand improvement. The most popular nonnuke right now is efavirenz. It's got substantial cognitive toxicity. We can get around that by giving the drug at night, but some patients still have problems during the day. There has been liver toxicity associated with nonnucleoside reverse transcriptase inhibitors, and we still have the rash problem. In most patients we can get around the rash by treating through, so that's not a terrible thing. But there have been cases of Stevens-Johnson syndrome reported with nevirapine.

Beyond side effects, a critical unanswered question about nonnucleoside RT inhibitors is long-term treatment success. Even though efavirenz and nevirapine have wonderful pharmacokinetic properties, and even though they're very potent in suppressing HIV if taken in combination with nucleosides, each of those drugs is only one amino acid change away from high-level resistance. So we need to get out of the phase III study results, which are done in highly selected, highly motivated patients, and look at performance of these drugs in a more unselected patient population. We need to know if nonnucleosides perform as well at a year, and two years, and three years in unselected patients as they have in their phase III studies. I am optimistic that NNRTIs will turn out to be very useful drugs in the long run, but again we need more data.

JIAPAC: Then we come to hydroxyurea. ACTG 5025 surprised a lot of people because patients went into it with a relatively high CD4 count, and yet there were cases of pancreatitis and elevated liver enzymes, a few people died, and the study had to be stopped.¹⁰ First of all, what do you think happened in that study? And, second, are you getting the sense that the negatives with hydroxyurea may be starting to outweigh the potential positives?

Flexner: I can't answer your second question with a whole of confidence, because I don't think we have enough data yet. Let me just make some comments about that.

We did a large ddI/hydroxyurea combination study in the AIDS Clinical Trials Group, ACTG 307, which Joe Eron and Ian Frank cochaired.¹¹ In that study, in 140 patients followed for over six months, we saw no pancreatic toxicity. Those patients all had CD4⁺ counts above 200, and they were only taking ddI and hydroxyurea. They weren't taking other antiretrovirals in the initial phase of the study, so maybe they represent a special case. But Franco Lori has also given hydroxyurea in a lot of settings, mostly to people with early disease, and has seen very little toxicity with ddI/hydroxyurea combinations.

The patients who developed life-threatening or fatal pancreatitis on hydroxyurea that I'm aware of were all taking ddI, d4T, and hydroxyurea. And that combination may be more dangerous than ddI and hydroxyurea alone, because we know that d4T can cause pancreatitis, ddI can cause pancreatitis; put those two drugs together with hydroxyurea and it's perhaps not surprising that you see more frequent and severe pancreatitis than you would otherwise anticipate. We may discover that hydroxyurea's toxicity is more pronounced in some regimens than in others.

The way we think hydroxyurea exerts its benefit on the antiretroviral activity of ddI is by changing the ratio of the intracellular ddATP triphosphate to the dATP triphosphate. And since it's probably the ddATP triphosphate that mediates toxicity as well as activity, then that strategy should not only increase antiretroviral response, but it should increase the risk of toxicity. The hope was that you could combine hydroxyurea with low doses of ddI and other drugs with nonoverlapping toxicity and not get into trouble. Whether or not that will prove true in large, unselected patient populations remains to be seen.

Hydroxyurea is not a drug that I would throw away yet, particularly in investigational circumstances, for two reasons. One is because I think it does make ddI a better antiretroviral drug. Whether it does the same for other nucleoside analogs remains to be defined. The second issue with hydroxyurea is that it's inexpensive, it can be given once a day, and although it does have toxicity, its major toxicity--neutropenia--is reversible, and one can easily monitor for that.

I know there's been talk about using hydroxyurea and ddI as an inexpensive, once-a-day antiretroviral combination in developing countries. I think we're going to have to be a little bit more cautious about that now because we are not able to monitor for toxicity in developing countries as well as we can monitor for it in the developed world. So we need to understand the long-term risks of that combination more precisely before we can advocate using it in clinical practice. I think studying it is still a good idea.

Who should treat HIV infection?

JIAPAC: When I look at all of these topics that you've talked about today, as a layman I'd say it's really obvious that treating HIV infection is just getting more and more complex. But I don't know as much about other infectious diseases. Is treating HIV infection a lot more complicated than treating other infectious diseases? And, if it is, should HIV be a subspecialty?

Flexner: This is a no-brainer. HIV management is definitely more complex than management of any other chronic infectious disease. In part this is because the virus is so dynamic and mutates so rapidly that we need a large number of drugs to control viral replication over long periods of time. This often turns out to be very complicated, because drug tolerability and drug interactions rear their ugly head.

The management of HIV right now is easily as complicated as giving combination chemotherapy to patients with solid tumors. If I ran a managed care company and came to you as a family practitioner or an internist and said, "You're now going to be responsible for giving combination chemotherapy to all your patients with breast cancer, colon cancer, and Hodgkin's disease," you would quit. And yet that's exactly the dilemma we're putting primary care providers in with the management of HIV disease in many managed care settings right now. There's a limited ability to consult specialists or refer to specialists, and I think that's a big mistake. Management of this disease has become so complex that it really needs to be in the hands of people who know what they're doing and are keeping up with developments in the field, which is difficult for most primary care providers to do.

You might ask, why treat HIV differently than we would any other chronic infection? Ten years ago we probably didn't need to treat it all that differently because we couldn't do that much about it. However, right now, with the drugs we have available in the drug store, in many patients we can suppress replication of their virus perhaps indefinitely, as long as we can come up with a regimen that they can take. And so to deny patients the best possible standard of care for this disease may be denying many people the chance for a substantial prolongation of life and a substantial improvement in their quality of life. I don't think that it's right any more to allow patients with chronic HIV infection to be managed by people who are not at the top of the art.

JIAPAC: Let me follow up on one thing you said, that we can suppress HIV almost indefinitely . . .

Flexner: I think in some patients we can suppress it indefinitely. I'll go out on a limb there.

JIAPAC: But lots of people would say, "You're crazy. Look at all these trials and cohort studies in which only 50 percent have a viral load under 400 after 12 months."

Flexner: Let me turn that around and say that means 50 percent of those patients have not had their viral load rise above 400. They're suppressed below 400, often below 50, and they're not apparently failing therapy. That tells us these drugs can work.

All of us can point to patients we have followed for three years or more on highly active antiretroviral therapy who have had their viral load suppressed to below the detectable limit, have no signs of failing therapy, and have essentially a normal quality of life right now. If I could present those types of data for some malignancies, people would be jumping up and down and screaming and yelling hooray. What this indicates is that we are really able to provide long-term control for this disease in some patients for whom that was not previously possible. And, in theory, as long as you can completely suppress replication of the virus, the virus won't mutate and become resistant to drugs. So if you can develop a tolerable regimen that patients can take on schedule and that keeps their viral load completely suppressed, their chances of developing drug resistance are vanishingly small, and their viral load should remain below the detectable limit indefinitely.

References and Notes

1. Hoetelmans RMW, Reijers MHE, Weverling GJ, et al. The effect of plasma drug concentrations on HIV-1 clearance rate during quadruple therapy. AIDS 1998 Jul 30;12(11):F111-5.

2. Gieschke R, Fotteler B, Buss N, Steiner J-L. Relationship between exposure to saquinavir monotherapy and antiviral response in HIV-positive patients. Clin Pharmacokinet 1999 Jul;37(1):75-86.

3. Descamps D, Flandre P, Calvez V, et al. Mechanisms of virologic failure in previously untreated HIV-infected patients from a trial of induction-maintenance therapy. JAMA 2000 Jan 12;283(2):205-11. Available at: http://jama.ama-assn.org/issues/v283n2/full/joc90679.html. Accessed April 24, 2000.

4. Burger DM, Hoetelmans RMW, Hugen PWH, et al. Low plasma concentrations of indinavir are related to virological treatment failure in HIV-1 infected patients on indinavir-containing triple therapy. Antivir Ther 1998;3(4):215-20.

5. Investigators from Abbott and Glaxo reported examples of this phenomenon at the 7th Conference on Retroviruses for lopinavir (ABT-378) and abacavir, respectively. Kempf D, Xu Y, Brun S, et al. Baseline genotype and phenotypic do not predict response to ABT-378/ritonavir in PI-experienced patients at 24 and 48 weeks. Presented at: 7th Conference on Retroviruses and Opportunistic Infections; January 30-February 2, 2000; San Francisco. Abstract 731; Lanier ER, Melby T, St. Clair MH, et al. Potential clinical impact of small differences between Virco Antivirogram and ViroLogic PhenoSense assays for abacavir in 3TC-experienced patients. Presented at: 7th Conference on Retroviruses and Opportunistic Infections; January 30-February 2, 2000; San Francisco. Abstract 788.

6. Piscitelli SC, Burstein AH, Chaitt D, et al. Indinavir concentrations and St John's wort. Lancet 2000 Feb 12;355(9203):547-8.

7. Flexner C. Report from the 7th CROI: pharmacology and drug interactions: compartmentalizing antiretrovirals. Available at: http://www.hopkins-aids.edu/publications/report/mar00_6.html. Accessed April 15, 2000.

8. Grabar S, Pradier C, Le Corfec E, et al. Factors associated with clinical and virological failure in patients receiving a triple therapy including a protease inhibitor. AIDS 2000 Jan 28;14(2):141-9.

9. Justice AC, Chang CH, Fusco J, et al. Extrapolating long-term HIV/AIDS survival in the post-HAART era. Presented at: 39th Interscience Conference on Antimicrobial Agents and Chemotherapy; September 26-29, 1999; San Francisco. Abstract 1158.

10. Havlir D, Gilbert P, Bennett K, et al. Randomized trial of continued indinavir (IDV)/ZDV/3TC vs. switch to IDV/ddI/d4T or IDV/ddI/d4T + hydroxyurea in patients with viral suppression. Presented at: 7th Conference on Retroviruses and Opportunistic Infections; January 30-February 2, 2000; San Francisco. Abstract 456

11.Frank I, Boucher H, Fiscus S, et al. Phase I/II dosing study of once-daily hydroxyurea (HU) alone vs didanosine (ddI) alone vs ddI + HU. Presented at: 6th Conference on Retroviruses and Opportunistic Infections; January 31-February 4, 1999; Chicago. Abstract 402.

000510
IA000501

Copyright © 2000 - International Association of Physicians in AIDS Care. All rights reserved. http://www.iapac.org

ÆGiS is made possible through unrestricted grants from Roxane Laboratories, Inc., iMetrikus, Inc., the National Library of Medicine, and donations from users like you. Always watch for outdated information. This article first appeared in 2000. This material is designed to support, not replace, the relationship that exists between you and your doctor.

ÆGiS presents published material, reprinted with permission and neither endorses nor opposes any material. All information contained on this website, including information relating to health conditions, products, and treatments, is for informational purposes only. It is often presented in summary or aggregate form. It is not meant to be a substitute for the advice provided by your own physician or other medical professionals. Always discuss treatment options with a doctor who specializes in treating HIV.

Copyright ©1990, 2000. ÆGiS & the Sisters of Saint Elizabeth of Hungary. All materials appearing on ÆGiS are protected by copyright as a collective work or compilation under U.S. copyright and other laws and are the property of ÆGIS and the Sisters of Saint. Elizabeth of Hungary, or the party credited as the provider of the content.