Treatment of HIV Infection

The history of the treatment of HIV is probably one of the most re­markable success stories in the history of medicine. HIV treatment is cer­tainly the most important development in a widespread, serious disease in the last twenty years.

Right now we have no rigid rules about when to start treatment. The decision to begin treatment should be strongly influenced by your CD4 count and, to a much lesser extent, your viral load. Most important, treatment should start when you are ready to stick with it. Partial treat­ment is worse than no treatment. In no other disease in medicine is the penalty for noncompliance worse.

We now know that drugs can reverse the loss of CD4 cells, decrease the amount of HIV, and delay or prevent the onset of the complications that accompany HIV infection. Drugs directed against HIV are called antiviral or antiretroviral drugs.

HAART, Highly Active Antiretroviral Therapy

The history of antiviral drugs against HIV infection begins with AZT and a clinical trial whose initial results were spectacular.

A large group of people with HIV infection was given either AZT or an inert pill called a placebo. The results were analyzed by an over­sight board, an independent group of scientists who had access to in­formation that the researchers doing the trial did not have. The reason for oversight boards is to find out if the drug is too bad to continue giv­ing it, or too good to continue giving the placebo; either way, the trial will be stopped. In the case of AZT, when the oversight board first re­viewed the results, the group getting the placebo had 19 deaths and the group getting AZT had only 1—a person who happened to have stopped taking the medicine.

Unfortunately, the benefit from AZT was short-lived, a problem that became painfully clear in the next few years. We now know that HIV became resistant to AZT. But we didn’t know that then and we took a decade to prove it. During that decade and the next, drug companies in­troduced many daughter drugs of AZT. The daughters are sometimes called “D drugs,” because of some of their names—ddI, d4T, ddC, 3TC, and ABC—but are more properly called nucleoside analogs. All these drugs are chemically related and inhibit the same enzyme, reverse tran­scriptase, by which HIV turns its RNA into the cell’s DNA, thereby pro­ducing new viruses. The nucleoside analogs are informally called nukes, or NRTIs (nucleoside reverse transcriptase inhibitors).

The nucleoside analogs all inhibited the multiplication of HIV, but not enough that they could be used alone. And as HIV continued to mul­tiply, it mutated, changing its genetic structure. Billions of new viruses every day meant millions of mutants; inevitably, a few of the mutants were resistant to the nucleoside analogs. After enough multiplication, the strains of virus vulnerable to the nucleoside analogs were killed off and only the resistant strains were left. These resistant strains of the virus meant that the nucleoside analogs were either less effective or not effec­tive at all. HIV was now multiplying the way it did before AZT existed. Shortly after ddI, the second AIDS drug, was introduced, we tried giv­ing two nucleoside analogs instead of just one—that is, ddI and AZT combined. The combination seemed to work better than when either drug was given alone. But in retrospect, it’s difficult to know how much we achieved during the first ten years of the treatment of HIV infection, except that by 1996, we had five drugs that added an estimated six ex­tra months of life.

In 1996 a second class of drugs was introduced, called the nonnu­cleoside reverse transcriptase inhibitors, or NNRTIs. As the name sug­gests, NNRTIs inhibit the same enzyme as the nucleoside analogs but are chemically different. The first NNRTIs were nevirapine, introduced in 1996; delavirdine, in 1997; and efavirenz, in 1998.

Meanwhile in 1996, the FDA also approved a third class of drugs, called protease inhibitors or TIs. PIs inhibit an enzyme called protease that HIV uses to assemble new viruses. Saquinavir was approved by the FDA in December 1995 and after that, a lot more PIs were approved (see table 3).

So during those next ten years, what did we learn about these treat­ments? We learned a lot. We learned that the goal of treatment is to in­hibit the virus, and the more inhibition, the better. The fewer viruses, the fewer mutations. And the fewer mutations, the less likely the mutations are to confer resistance to drugs. So the rate of resistance mutations de­pends on the viral load; and a viral load of 50 is better than 500, which is better than 5,000. In fact, with a viral load that’s undetectable—that’s less than 50—resistance appears to be nil. An undetectable viral load and the absence of resistance mutations result from good adherence to treatment, which should keep people in good health for years, probably decades.

We learned that people need to take two or three active drugs: not one and not four, but sometimes two and usually three. And those drugs are not just any two or three but specific combinations of specific drugs, because we also learned that some combinations work and some do not.

We learned that adherence to the treatment regimen is the key, more important with drugs for HIV than with drugs for most other diseases. First, people need to take at least 95 percent of the prescribed doses to achieve the treatment’s goal, which is no detectable virus for a sustained

*NRTI = nucleoside reverse transcriptase inhibitor NNRTI = nonnucleoside reverse transcriptase inhibitor PI = protease inhibitor

EI = entry inhibitor

period.

We learned that the drugs have side effects. Some side effects were short-term, like allergic reactions and stomach problems. But some were long-term and came as a complete surprise, like the lipodystrophies that change people’s blood lipids and appearances.

We learned that many people fail treatment. They fail the first treat­ment, and when the regimen is changed, they fail that too. And with each failure thereafter, the treatment becomes progressively more compli­cated and the probability of success progressively less.

We learned—and were surprised—that failed treatment was still better than no treatment. We know this because when people fail treat­ment and have a high viral load and then stop taking drugs altogether, the viral load immediately jumps even higher and the CD4 count falls faster.

We learned that to improve adherence to treatment, the treatments need to be simplified. People have three problems with adhering to treat­ment: the pill burden (the total number of pills), the number of times per day the pills have to be taken, and the necessity of taking the pills with or without food. New drug development has focused on these problems. To reduce the pill burden, pills were combined; taking the drugs three times a day became completely antiquated; and the need to take drugs with food was either eliminated or more carefully defined.

We learned that treatment decisions are complicated and require the physician to have substantial skill and experience. The physician has to choose among twenty drugs—and more are coming—that vary sub­stantially in toxicity, potency, and patterns of resistance. The stakes are high and so are the patients’ expectations. HIV care has become a spe­cialty. People with HIV infection should be seeing specialists.

From all that we learned in those ten years, we put together certain guidelines for the drugs’ use. These federal guidelines are based on good, authoritative sources, experts in the field who laboriously review the data from reports and tests and then write recommendations. Nearly every country in the world has national guidelines for the use of these drugs, and fortunately, they all say the same, or nearly the same, things. (The exception is the World Health Organization guidelines for devel­oping countries, where drug prices drive many of the decisions.) The U.S. federal guidelines are constructed by a panel of about thirty-five ex­perts who communicate by teleconference every month to keep the guide­lines up-to-date, in what is clearly the fastest-moving field in medicine. These guidelines (and other federal HIV guidelines for topics like HIV prevention and nonoccupational exposure) are available at www.aidsinfo.nih.gov.

We must acknowledge that guidelines are not rules. Every person with HIV infection is different, and exceptions are normal. Knowing which idiosyncrasies require a detour is the “art” of medicine. This art comes with experience, along with keeping current through a vast array of medical journals, Internet sites, and medical conferences.

When to Start Treatment

The U.S. federal guidelines recommend HIV treatment for all people with an AIDS-defining complication or a CD4 cell count below 200. In fact, all guidelines for Europe, Africa, South and Central America, and Asia make this same recommendation. A CD4 cell count below 200 de­fines the risk of most complications of HIV infection; it defines AIDS. Taking treatment before the CD4 count goes this low, however, provides a margin of protection and prevents the infrequent complications that occur even when the CD4 cell count is above 200. We often recommend beginning treatment at a CD4 cell count of 350. In short, antiretroviral treatment is a must for a person with a CD4 cell count below 200, and such treatment is seriously considered when the CD4 count is between 200 and 350.

Several issues influence the decision to take treatment when CD4 counts are between 200 and 350. (1) Do you want to be treated? Some people are determined to wipe out that virus as soon as possible; others want to delay the drugs as long as possible. (2) What is your viral load? A high viral load generally means faster progression through the disease and more mutations in the virus. Some experts think that treatment should be considered for anyone with a viral load over 100,000 copies per milliliter. (3) What is your CD4 cell count? A count of 210 is obvi­ously quite different from a count of 340. (4) Are you ready for a com­plex regimen that requires tenacious adherence? Nonadherence to the regimen takes the risk that your virus will become resistant and tarnish the success of the treatment. (5) What is your CD4 slope? The CD4 slope is the trajectory of the CD4 count over time: a rapid decline is an obvi­ous cause for concern.

For the person with HIV infection, the decision to begin treatment requires a certain amount of introspection, commitment, and realism.

What “No Detectable Virus” Means and Why You Want It

The viral load, which measures the concentration of HIV in the blood, can now be detected at a lower concentration than it could by earlier tests. The lowest level that older tests could detect was 500 copies per milliliter of blood. Newer tests can go down to 20 or 50 copies per mil­liliter. “Undetectable” means that the concentration of HIV in the blood is lower than the test’s limit of detection; in other words, with the older tests, “undetectable” would mean below 500, and with the newer tests, below 50. Most HIV experts now define “undetectable” as below 50 copies per milliliter.

You want no detectable virus because those low concentrations in­dicate that the virus is reproducing so slowly that drug-resistant muta­tions are unlikely. The higher the viral load, the more new viruses are made, and with more new viruses the chance of mutations is greater. A person with a viral load of 10,000 has a viral reproduction rate 10 times higher than a person with a viral load of 1,000, a mutation rate 10 times higher, and a resistance rate 10 times higher. A person with a viral load below 50 does not appear to be making enough new viruses to cause clinically meaningful resistance. So a combination of drugs that brings about a viral load of less than 50 should work indefinitely.

You can usually tell soon after treatment starts how well it will work. In the first week, the viral load should drop ten-fold; in 8 to 16 weeks, it should be less than 500; and by 24 weeks, it should be unde­tectable. Sometimes these goals are achieved more slowly. These early weeks are critical because the first treatment regimen has the highest probability of success.

How to Get “No Detectable Virus”

You have to take all your pills regularly, and that’s not easy. The pills can be numerous and toxic, the pill schedule can be demanding, and the side effects are unpleasant. We know that most people find adhering to even simple drug schedules difficult: the average person with high blood pressure, for example, takes only about half the recommended doses. With HIV the difficulties of adherence are much greater. And unfortu­nately, the consequences of nonadherence are more severe. If HIV is sup­pressed only incompletely, the remaining virus develops resistance to the drugs. The person with high blood pressure who misses several doses will still, when he or she wants to resume taking pills, have a drug that works. The person with HIV infection who misses pills risks resistance, and once resistance develops, that drug is usually lost to that person for­ever. Nonadherence might even risk losing the effectiveness of a whole class of drugs for the person.

One study showed that if you take 95 percent of the pills (e.g., 19 out of 20), you’ll have an 81 percent chance of getting no detectable virus. If you take only 80 percent of the pills (e.g., 16 out of 20), your chances nose-dive to 30 percent. On the basis of this and other studies, we conclude that occasional missed doses at the 5 percent level proba­bly won’t create resistance; resistance might require weeks or months of missed or reduced doses. This risk might be quite different for NNRTIs. So the risk seems to vary with the drug, the viral load, and how many other drugs you’ve taken.

A few rules are important:

• Always take your medication in full dose, unless your doctor tells you otherwise.

• Understand your regimen. Know what you are taking and when. If you are unsure, ask. Your resources are your physician, nurse, physician assistant, or pharmacist. They are best if they are affiliated with your care, and best if they are “HIV-savvy.” The regimens used generally don’t vary. For example, atazanavir (Reyataz) is two 200 mg capsules a day, always with food. “Boosted” atazanavir is two 150 mg capsules a day, with 100 mg of ritonavir (Norvir), again with food. The reason for the ritonavir is to increase the amount of atazanavir in your body by slowing your body’s attempt to elimi­nate it. Ritonavir effectively triples the concentration of atazanavir. The reason for the food is that it increases the amount of drug your gut absorbs by 70 percent; without food, you get about one-third as much drug. We don’t make these recommendations simply because they seem to make sense; they are based on facts established by so­phisticated pharmacology studies that test people in laboratories, carefully measuring the amount of drug their bodies absorb with and without food, and with and without ritonavir. The standard methods and results of these studies are all part of the one-thousand- page package submitted to the FDA for the drug’s approval.

• Know the language. Each HIV drug has a chemical name, a trade name given by the company that manufactures it, and a three-letter abbreviation. For example, efavirenz (chemical name), Sustiva (trade name), and EFV (three-letter abbreviation) are all different names for the same drug. Some clinics and pharmacy books, such as the Physicians’ Desk Reference, have photographs of the drugs; HIV clinics almost always have wall charts with photographs of the HIV drugs; but because both the books and the clinics have a prob­lem keeping up, you’ll often find that one of your drugs isn’t pic­tured because it’s too new. The abbreviations are standardized and often used to indicate specific regimens. For example, TDF/FTC/ LPV/r means tenofovir, emtricitabine, lopinavir, and low-dose ri­tonavir; this sounds like a lot of different drugs, but because several drugs are now packaged in one pill, it’s actually just Truvada (TDF plus FTC) and Kaletra (LPV plus r).

• Be aware of the side effects. All drugs have them, and they’re gen­erally well-known because the FDA requires a drug’s toxicity pro­file before licensing it. With HIV drugs, side effects are important: most studies show that 30 to 40 percent of people need to change regimens because of the drugs’ side effects. Some side effects, peo­ple know about—nausea, diarrhea, headache. Some are detected primarily by laboratory tests. Some occur immediately; some occur only after months or years of taking the drugs. It’s good to know what side effects to expect. Ask.

• Pay close attention to drug interactions, the effect of one drug on another. Drug interactions are common with all PIs and NNRTIs. For example, a person taking ritonavir died from the street drug Ec­stasy, because ritonavir increases the blood levels of Ecstasy by up to ten-fold. St. John’s Wort can cause severe interactions with all protease inhibitors because it decreases their concentrations to the point where they are useless. Sometimes a cardiologist or an internist is prescribing one set of drugs while the HIV physician prescribes another set. Almost any drug for ulcers is a disaster for atazanavir, which needs stomach acid for absorption. Simvastatin (Zocor), commonly used for blood cholesterol problems, can’t be given with any protease inhibitors because the protease inhibitors make the blood levels of Zocor so high that the drug could become toxic. Most of these drug interactions are well worked out and easily found in various charts or on the Internet. The point is, make sure all your physicians know all your medications. You can also look up the drugs you’re taking and their interactions on the Internet: some sites allow you to enter two drugs and rapidly learn what drug interaction takes place.

• Pay close attention to instructions to take the drugs with meals or on an empty stomach.

• As a general rule, when you have a problem with part of the regimen, it’s best to stop all drugs. Thus, if you stopped one drug of a three- or four-drug regimen because that drug ran out or caused toxicity, you should stop all drugs until a full regimen can be restarted. This recommendation is complicated because two commonly used drugs, efavirenz and nevirapine, last a long time in the body. So when you stop AZT/3TC/EFV immediately, the AZT and 3TC are gone from your body in one day, but the EFV is still there two weeks later. This long period of efavirenz at low levels could possibly trigger resistance. Note that this problem with varying times for clearing a drug out of your body applies only to efavirenz and nevirapine. In any case, HIV physicians deal with these complexities with tricks that are themselves complex, variable, and evolving. In general, then, not stopping the full reg­imen will inhibit HIV only partially and HIV will become resis­tant.

• The most important drug regimen is probably the first one taken. The reason is, if you become resistant to one drug, you may also be resistant to other drugs in the same class. This is called cross­resistance, and it happens especially with NNRTIs.

We have worked hard to get treatments that can be taken twice or even once a day, that will be well tolerated, and that will work. Research into adherence shows—as you’d expect—that the best treatments are sim­ple, and that complicated treatments cause confusion and nonadherence.

Research has also been done to learn which patients are most ad­herent. Being good at taking pills does not appear to be related to gen­der, race, socioeconomic status, or level of education. Men or women; black, white, or Hispanic; rich or poor; educated or not—we can’t pre­dict who will be able to adhere to the HAART regimen. In fact, studies done with AIDS doctors show that most of them cannot adhere to the treatments they regularly prescribe for their patients. The only people who seem predictably bad at adhering to treatment are active substance abusers like alcoholics or injection drug users, people who are mentally ill, and adolescents.

And to make predicting adherence even harder, studies find that most people tend to overstate their adherence. They know their doctors think adherence is important, and they want their doctors to see them as good patients. But research that actually monitors people taking pills reveals that people aren’t sticking to the regimen as strictly as they say they are. It’s best to be honest.

Sticking to the Regimen

Because the first round of treatment is the best bet for getting no de­tectable virus, you shouldn’t risk failing at it. Under no circumstances should you “give treatment a whirl.” Don’t start the regimen until you’re ready to stick to it. Tell your doctor if you think you can’t. This is seri­ous stuff and you may get only a limited number of shots at effective drugs. The following are guidelines that might help adherence:

• People should learn their regimens and those regimens’ nuances: which pills and when, which need to be taken with food and which need to be taken while fasting, and what the relevant drug interac­tions are—can you take the HAART drugs with Prilosec, for exam­ple, or with Lipitor?

• Drugstores and clinics have weekly pill boxes with sectioned con­tainers: most pill boxes have seven daily sections, each divided into three to five time compartments. Weekly pill boxes not only remind you of the pills you need to take, but help you keep count of whether you’ve taken them all.

• In any case, it is a good idea to count out pills weekly and to write down when you missed a dose.

• The dose that’s hardest to remember is the midday dose. Pills taken once daily seem easy; twice daily are only a little harder; and three times daily, much harder. Most HIV treatment is now pills taken once or twice daily.

• Try to work pill taking into your daily routine. Pills taken twice a day could be taken when you brush your teeth, and you could re­member them by keeping them next to your toothbrush.

• A number of reminder systems are available: watches with alarms, timers, pagers, automatic dialing phone systems. The latter is a sys­tem where you program the telephone to ring regularly and de­liver a recorded message like “take your pills,” or “take two yellow football-shaped tablets and an orange pill.” These reminder systems can help, but in general they don’t work: because people depend on them, the system fails, and the people miss doses.

• Ask a spouse or colleague or friend to help you remember.

• Some people draw clocks with the names of the pills or colors for each pill marked at the proper times and in relation to meals.

• One common problem with adherence is the drugs’ side effects, es­pecially those effects felt immediately, like nausea and diarrhea. (Many people with HIV infection say they didn’t feel sick until they got treated.) Try not to stop or miss doses because of trivial side ef­fects. Most side effects are dealt with easily. A few, however, are im­portant indicators of more serious complications, so make sure your health care provider knows every possible side effect you might be having.

• Two major impediments to adherence are depression and sub­stance abuse. If you are depressed or are abusing drugs or alcohol, your health care provider absolutely must know. Both conditions can be treated.

Missing Doses

When HAART fails, the cause is usually poor adherence—people miss too many doses of the drug. Poor adherence not only allows HIV to get out of control, but it also permits resistant viruses to flourish; and resis­tant HIV limits which drugs will work. Balancing adherence and resis­tance is a tricky equation and depends on the drugs.

With protease inhibitors, taking most of the doses but not control­ling HIV will lead to resistance. Taking even fewer doses—less than half the doses prescribed—will also result in failure to control HIV but will not lead to resistance: in that case HIV can thrive without resorting to resistance. With nevirapine and efavirenz, missing occasional doses is not as dangerous because they last a long time in the body. Missing a lot of doses, however, will cause HIV both to go out of control and to be­come resistant. With these drugs, you should take every dose, and if stopping is necessary, stop them all at once (unless instructed otherwise).

The general rule, however, is to stick carefully to the regimen but not to panic. Because we pay such enormous attention to adherence, some people worry that a few missed doses will mean harsh penalties. But penalties come not from a few missed doses, but from repeated in­consistencies: taking some drugs but not the whole regimen; or losing the full dosage by taking the drug with food when it should have been taken while fasting. In general, most of the drugs can be taken a few hours late without losing effectiveness. If you forget a dose, you can take it a few hours later. If you’ve missed by more than a few hours, just wait until the time for the next dose. The exception to this is a pill taken once a day: take it as soon as you notice missing the dose, or you may have a long period with no drug on board. If you have delayed until the time of the next dose, don’t “double up”—it is better simply to lose one dose. But do avoid missing a single daily dose consistently. If you have been vomiting, you should stop taking the pills until you can keep them down. If you have diarrhea, take the drugs as usual; diarrhea has little effect on drug absorption.

If you do not take your pills with or without food, whichever is re­quired, you risk the drug not getting into your system. If you have trou­ble taking the pills with or without food as required, tell your doctor, who can prescribe a more realistic regimen. Treatment may need to be stopped temporarily for many reasons. You may run out of pills, or the pharmacy may run out of them. You may have gastroenteritis or surgery. In general, it’s always best to stop all drugs at once. The only drugs that are possi­ble exceptions are efavirenz and nevirapine because they stay in the body a long time. Make no decisions without talking to your doctor.

What Drugs to Treat With

The more we learn about the effects and long-term side effects of the drugs and about the durability of the viral response, the more compli­cated the treatment of HIV infection has become. But these are the goals of treatment:

1. Suppress the virus as long as possible: we keep track of suppres­sion by measuring the HIV viral load—the concentration of HIV in the blood—usually at three- to four-month intervals.

2. Maintain or get back a good immune system: we keep track by measuring the CD4 cell count, which is the barometer of immune function. Success in keeping the immune system healthy and the CD4 cell count high depends on our ability to suppress the virus, be­cause it is the virus that kills the CD4 cells.

3. Prevent the HIV-related complications: these are the complica­tions, like pneumocystis pneumonia or cryptococcal (fungal) menin­gitis, that first led to recognition of HIV. These complications virtu­ally never happen when the CD4 cell count is above 200.

4. Avoid the side effects of the drugs used to accomplish goals 1-3. These side effects are summarized below.

5. Maintain as many treatment options as possible: some treatment decisions must balance risk against benefit. A low CD4 cell count risks a major complication, so we need to attack HIV aggressively, even if we have to change treatment many times. With a high CD4 cell count, however, we may have the luxury of going slower, thereby having the benefit of more options down the line. Generalizing about these decisions is difficult because individual risk-benefit cal­culations are so highly variable.

6. Stop HIV transmission: the likelihood of transmission depends on the risk behavior and the viral load. If we could decrease both, we would be able to slow or stop this epidemic.

HIV treatment has large gray zones and is different for different peo­ple. But we do have a consensus, based on vast experience and substan­tial review, on a few recommended regimens for initial treatment. After making sure the person with HIV infection has realistic expectations of what can be done, the best success for achieving goals 1 to 3 comes from the first regimen.

The usual first regimen is two nucleosides combined with a protease inhibitor, or a boosted protease inhibitor, or a nonnucleoside reverse transcriptase inhibitor. The two nucleosides are sometimes called the “nucleoside backbone,” and the drug from the other two classes, a PI or a NNRTI, is just called “the third drug.” When a PI is used, it is often “boosted” with ritonavir. Ritonavir is also a PI, but it is used because it increases the blood levels of the other PI, that is, it “boosts” the PI. In this case, the doses of ritonavir are so low that it does not really count as part of the treatment. In short, the first regimen usually consists of three drugs for HIV and is sometimes called “triple therapy.”

When to Stop Treatment

Treatment is stopped in three situations. One is when stopping is planned or elective. Sometimes stopping treatment is an option because the person has responded to it so well. For example, when treatment started, the CD4 count was 200 and the viral load was 50,000; four years later, the virus is undetectable and the CD4 count is 800. In a case like this, stopping has generally been safe and effective. The general re­sult is that once the drugs are all stopped, the virus quickly rebounds to pretreatment levels and the CD4 count drops rapidly for two to three months, then levels off. Treatment always needs to be restarted, usually when the CD4 count is down to 300 to 350. Studies of more than 700 patients show that such “drug holidays” can last an average of eight to twelve months and that when treatment restarts, the patients respond well. Just remember that when the viral load is high, transmission be­comes a greater risk. And don’t be alarmed at the precipitous drop in the CD4 count in the first months.

The other situation in which treatment is stopped is when people feel that it’s not working and they want to stop. Many regimens are hard to tolerate or don’t control the virus. The viral load is high, and resis­tance mutations exclude the likelihood that any new combination of drugs will work any better. But even though the drugs cost $10,000 to $15,000 per year, cause side effects, and don’t appear to work, they’re still working. When the drugs are stopped, the viral load shoots up, usu­ally ten-fold, and the CD4 count drops fast. So even though a treatment appears to fail—assuming no other drug options are better—it should still be continued.

Unplanned drug discontinuations are different and common. These may happen because of gastroenteritis, bad side effects, an operation or surgical procedure, or a pharmacy running out of a drug. In these cases, the best rule is to stop all drugs simultaneously and restart treatment when the full regimen is again feasible. Be sure to let your health care provider know.

When to Change Treatment

The best way to measure the response to treatment is to measure the vi­ral load after the treatment begins. Although we listed six goals for treat­ment, the most important is number 1, suppressing the virus. The aver­age person with HIV infection, without treatment, produces about 10 billion new HIVs every day. These viruses aggressively invade CD4 cells, multiply inside them, then kill them. We can kill the HIV virus with “triple therapy,” but we do not have drugs that reconstitute the CD4 cell count. So the drugs against HIV are the key not only to suppressing the virus, but also to maintaining a good immune system and preventing complications.

The many clinical studies of various treatments have shown that the initial triple therapy will kill about 90 percent of all HIV viruses in the body within one to two weeks, and 99.9 percent within one month. We expect that the viral load will go from an average of about 50,000 to less than 50—that is, to “no detectable virus”—within 16 to 24 weeks. Many studies now indicate that the best predictor of long-term success with treatment is the viral load after 24 weeks of treatment.

When the viral load decreases, the body makes more CD4 cells. The CD4 cell count increases rapidly, often by 40 to 50 in the first two to three months and by an average of 100 to 150 per year thereafter. How­ever, this increase is different in different people. Some people control the virus well, but their CD4 cell response is blunted. Other people have the opposite: a relatively poor control of the virus but a robust jump in the CD4 cell count. These variations are poorly understood, and so far we have few guidelines. In general, we know that killing the virus is critical: killing the virus is the only way we know to raise the CD4 cell count. Controlling the virus also helps prevent HIV-associated compli­cations. But if the virus is under control and the CD4 cell count remains disturbingly low, we sometimes change treatment to see if the CD4 count will rise.

Treatment is changed for two major reasons. The first is “virologic failure,” which means the failure to achieve a low viral load, measured at periodic intervals after treatment has started. In the case of one of the recommended treatment combinations, the triple therapy, we have two possible explanations. First, the virus may be resistant to one or more of the drugs. We can conduct a resistance test and get a very good read on why a treatment may not be working. Second, the drugs might not be getting to the places in the body where HIV is found. The most common reason is inadequate adherence. Adherence has always been the most dif­ficult challenge of triple therapy and the most common cause of treat­ment failure. People with HIV infection hear so much about taking pills as prescribed, more than anyone with any other disease, because the con­sequences of nonadherence are so punishing.

However, reasons other than poor adherence may cause inadequate drug levels. One reason might be pharmacological—that is, the drug simply is not getting into the system—and this might occur because the drug is taken with food when it should be taken without, or taken on an empty stomach when it needs to be taken with food. We have worked hard to develop treatments that do not have this “food effect,” and only a small number of drugs still do. Another reason might be an interaction with another drug: an antibiotic, or a cholesterol pill, or blood pressure medicine may have decreased the concentration of the HIV drug. Drug interactions are most common and most pronounced with the protease inhibitors. And since protease inhibitors also have a marked effect on the concentrations of many non-HIV medications, you should be sure that all the physicians participating in your care are aware of all the drugs you are taking. Sorting this out is extremely complicated—thou­sands of drugs are currently in the market place—but many computer programs or Internet sites allow you to enter the two drugs and rapidly learn what drug interaction takes place.

If failure to achieve a low viral load—virologic failure—is the most frequent cause for changing treatment, the second most frequent is in­tolerance of the drugs’ side effects. Any of the three drugs you take can have a multitude of side effects. The method of managing the side effects depends on the severity of the reaction—which may have potential long­term consequences. Some side effects are so severe that future use of the drug carries a substantial risk, and the drug can no longer be safely taken. Other drugs are stopped because the side effect is annoying, or gets worse over a period of months or years, or seems intolerable be­cause it makes the quality of life so poor. Such side effects might be se­vere headaches, uncontrollable diarrhea, or severe nausea. Decisions are simplified because we now have so many drugs: if a regimen, especially the first one, is so awful that it interferes with your usual life activities, it is usually relatively easy to find alternatives. It may get progressively more difficult, though, when you lose the option of taking certain drugs because of either resistance or intolerance.

Long-term studies of people taking triple therapy indicate that about 40 to 50 percent of them will need to change the regimen. About half of these changes are the result of resistance, and the other half arise from drug intolerance. This is not to say that the drugs do not work at all; the drugs do work, but they do not meet our stringent expectations for both keeping the virus controlled and maintaining a good quality of life.

What to Change To

When people encounter virologic failure, the usual alternative is to mea­sure the resistance of the virus to various drugs and then to select drugs based on that “resistance profile.” In general, finding alternative regi­mens after one round of treatment fails is pretty easy; it is a bit harder after two rounds, and it gets more difficult as the number of rounds in­creases. The reason is that resistance accumulates: you develop resis­tance to one drug, then another, and so on, and those resistant strains of the virus persist in the body. Sometimes viruses develop “class resis­tances,” meaning they decrease the activity of an entire class of drugs.

How to Test for Resistance

Tests for resistance predict which drugs will work against your strain of HIV. Actually, what the tests do is predict which drugs won’t work. The two types of resistance tests are phenotypic tests and genotypic tests.

The phenotypic test, done only rarely, checks whether your own strain of HIV has become resistant to a given drug. The test treats your strain of HIV with that drug, then compares the results to another, unmutated HIV, called a “wild type virus,” treated with the same drug. If 4 times more of a given drug are needed to inhibit your strain than the wild type, your strain is considered resistant to that drug. If 4 to 10 times more drug are needed to inhibit your strain, your strain is considered very resistant.

The genotypic test is more complicated. HIV is error-prone; when it reproduces, it makes lots of mistakes. Specifically, some of its genes mu­tate. Some mutations, called resistance mutations, confer resistance to certain drugs.

The convention for reporting resistance mutations is a confusing se­ries of numbers and letters: one is K103N. The number in the middle in­dicates the particular sequence of amino acids (called a codon) on the gene at which the mutation occurs—in this case, codon 103. The first letter refers to the amino acid that is supposed to be at codon 103 (K, here, means lysine). The second letter refers to the amino acid that is ac­tually there as a result of mutation (N means asparagine). So the K103N mutation on the gene for reverse transcriptase means that on codon 103, lysine has been replaced by asparagine. Such changes in the sequence of amino acids can make the virus behave differently.

A virus with a K103N mutation, for instance, won’t be affected by efavirenz (Sustiva). Although efavirenz will continue to kill most strains of the virus, it won’t kill the strains with a K103N mutation. Those mu­tated strains will thrive, then gradually predominate. Eventually, the per­son will have enough HIVs with the K103N mutation that genotypic tests will detect the mutation, showing that efavirenz will no longer lower the person’s viral load.

The mutations that affect NNRTIs are often knockouts; that is, the mutations render a drug completely useless: K103N on the reverse tran­scriptase gene makes the NNRTIs efavirenz and nevirapine (Vira- mune) useless. The mutations that affect the nucleoside analogs often simply reduce their effectiveness: K65R on the reverse transcriptase gene reduces the effectiveness of didanosine (ddI), tenofovir (TDF), and abacavir (ABC) but has no effect on the effectiveness of zidovu­dine (AZT).

The mutations that affect protease inhibitors are either signature mu­tations that affect only specific drugs, or class mutations that affect all protease inhibitors. Class mutations are not usually knockout mutations like those affecting NNRTIs; that is, no single mutation renders the whole class of protease inhibitors completely useless. But several signature mu­tations, each conferring resistance to specific drugs, can add up so that the whole class of protease inhibitors becomes less effective over time.

In general, as the drugs become less effective, the concentration of resistant viruses increases, and so therefore does the probability of re­sistance mutations. In a person with a viral load of 20,000, the virus is multiplying ten times faster than in a person with a viral load of 2,000; and the probability of more resistance mutations is ten times greater.

The situation isn’t all bad. Although the M184V mutation on the reverse transcriptase gene is a knockout for lamivudine (3TC), the same mutation also seems to injure the virus—the virus simply doesn’t repro­duce as well. So even though the 3TC isn’t killing the virus, it is en­couraging the spread of the M184V mutation and thus reducing the virus’s ability to reproduce. That is, while people with K103N mutations should no longer take efavirenz or nevirapine, people with M184V mu­tations often take 3TC.

Resistance mutations are forever. Even though resistant strains seem to disappear when the drug they resist is stopped, they’re not really gone. They remain in the body in some sequestered haven, such as the lymph nodes, invisible to resistance tests and ready to rebound when you take that drug again. Say you take 3TC for three years and get the M184V mutation. If you stop 3TC, a resistance test six months later will show that the M184V mutation has disappeared. But M184V is actually still around in low concentrations; it has become what is called a minority species, and resistance tests don’t recognize it. If you take 3TC again, the M184V minority species will rebound and soon become a majority again.

The lesson is that decisions about treatment need to consider not only the results of current resistance tests, but also previous resistance tests and previous treatments that have failed.

Side Effects, or Toxicities, of HAART

HAART has been unexpectedly toxic. In fact, much of the research in the field of HIV infection is no longer devoted to the complications of immune suppression but to the complications caused by HAART. Peo­ple with HIV infection are now faced not with pneumocystis pneumo­nia or CMV retinitis, but with such side effects as high blood fat, pre­disposing them to heart disease, and a rearrangement of fat in their bodies.

Most side effects can be described in these three ways:

• Subjective versus objective: Subjective side effects—like gastroin­testinal complications, headaches, and weakness, that depend on people’s perceptions or feelings—are common and can be debilitat­ing, but their seriousness is often difficult for the physician to assess. Objective side effects are those that cause abnormal laboratory tests; they can be monitored for frequency and tested for confirmation.

• “Deal breakers” versus modifiers: Some side effects are “deal break­ers” that rule out the continued use of a drug; others simply mean the regimen needs to be modified or the symptoms need to be treated with another drug.

• Long-term versus short-term: Many of the drugs used for HIV in­fection have side effects that take months or years to develop and are still poorly understood. This topic is discussed in more detail be­low. Other side effects become apparent immediately. For example, gastrointestinal side effects are among the most common and usu­ally occur with the first dose; bad dreams with efavirenz, which are quite temporary, are noted with the first night’s sleep.

The major side effects associated with individual drugs are noted in table 4. This chapter deals with side effects by category.

Gastrointestinal Intolerance

Symptoms: The usual symptom is nausea, occasionally along with vomiting and abdominal pain. Diarrhea is also common. These side ef­fects are usually noted with the first dose of a drug or drug regimen.

Frequency: The frequency varies greatly, depending on the drug. This type of intolerance is most common with all of the PIs, AZT, and ddI. Diarrhea tends to be particularly frequent with LPV/r (Kaletra) and NFV (nelfinavir).

Treatment: Often tolerance of the drug improves with time. Nausea may improve when the drug is taken with food—though, since some drugs like unboosted indinavir or ddI (didanosine) must be taken on an empty stomach, food is not always an option. Many common medicines for gastrointestinal intolerance, such as Donnagel, Kaopectate, and Pepto-Bismol, are over-the-counter drugs that can be purchased without a prescription. For diarrhea, the standard is Imodium or calcium. These treatments usually clear up or at least improve the gastrointestinal symp­toms to the point where they can be tolerated.

Table 4. The Drugs of Highly Active Antiretroviral Therapy (HAART)

86

creases

87

The Protease Inhibitors

Table 4. (Continued)

88

other PIs in doses of

200-400 mg/day

89

The Entry Inhibitor

Warnings: Abdominal pain along with nausea and vomiting can also be a sign of such very important complications as pancreatitis or lactic acidosis. Pancreatitis is most frequently associated with ddI treatment and can cause severe pain and require hospitalization. If you are taking the drug known to cause this complication and the symptoms began oc­curring long after the drug was started, you may have pancreatitis. Lac­tic acidosis can cause weight loss along with the other symptoms; is al­most always associated with d4T, ddI, or, less frequently, AZT; and also occurs long after the drug was started (see below for a fuller descrip­tion).

Lactic Acidosis

Frequency: This is an unusual side effect associated with NRTI drugs, primarily d4T, ddI, and AZT. About 1 percent of people taking these drugs are diagnosed with lactic acidosis each year; the percentage for those on d4T is probably greater. The cause of this syndrome is not clearly defined, but it appears to be the toxic effect on mitochondria, the energy source of cells, when these drugs are used over a long time.

Symptoms: The most common complaints are the slow onset—af­ter months of treatment with the drugs mentioned—of rather nonspe­cific gastrointestinal and other symptoms, including nausea, appetite loss, stomach pain, vomiting, weight loss, and general fatigue.

Diagnosis: The first step is to suspect the syndrome, because its on­set can be subtle and its symptoms nonspecific. The diagnosis can be con­firmed with a blood test for lactic acid; it’s a standard test and available in most laboratories. The normal level is 2 (2 μmoles); a level of 2-5 is a bit high, 5-10 is serious, and a level over 10 could be regarded as a medical emergency.

Treatment: The usual treatment is to discontinue the drugs that ap­pear to cause the complication and to provide supportive care: intra­venous fluid for hydration, and support of various organs that may fail. Lactic acidosis takes a long time to come on, and it also takes a long time to disappear—weeks or months may elapse before the lactic acid levels are back to normal. The future use of drugs after this complication usu­ally is limited to drugs that were not implicated in the first place.

Skin Rash

Frequency: Skin rash is a common side effect of drugs in general and is especially common with nevirapine, efavirenz, fosamprenavir, tiprana- vir, and atazanavir. About 15 to 25 percent of the people taking these drugs report rashes. Most of the rashes are mild to moderate, and they go away if the drug is continued. In about 2 to 4 percent of the cases, the rash is so severe as to require discontinuing these drugs.

Symptoms: Most mild to moderate rashes consist of blotchy, red spots on the trunk, arms, legs, and face. They may itch. Mild rashes can often be “treated through”; that is, the drug is continued and the rash usually resolves on its own. Treatment may include antihistamines and, occasionally, cortisone ointments. More serious rashes—those associ­ated with blisters, fever, and involvement of the mouth or eyes—man­date immediate medical attention. Any rash that is severe should be seen and managed in a medical-care setting.

Warnings: Two specific skin reactions that include a rash along with other findings indicate a severe and sometimes life-threatening compli­cation, either the abacavir hypersensitivity reaction or the nevirapine liver disease. They are both described below, but we mention them here because a rash may be an important part of the initial side effect. Also, some of the more extensive rash reactions seen with nevirapine or efavirenz, usually in the first sixteen weeks of treatment, may be severe enough to require permanent discontinuation of these drugs.

Liver Disease

Frequency and Description: All antiretroviral drugs may cause liver disease, but the mechanism is quite different for different classes of drugs. And although abnormal liver tests are noted in 5 to 15 percent of all people with HIV infection, most cases are quite trivial: the person has no symptoms and the problem is found by a routine test of liver func­tion. Many or most people do fine with continued treatment. The drug causing liver problems would be stopped only after a judgment call by the physician about how bad the liver test needs to be. This nonthreat­ening form of liver disease is a side effect of all protease inhibitors but occurs in only 10 to 15 percent of the people taking them. It is also a side effect of the NNRTIs nevirapine and efavirenz, occurring with a similar frequency. The nucleosides may also cause liver disease, but the process is different: the disease results from statosis, or fat deposits in the liver, caused by lactic acidosis.

A further form of liver disease, particularly important to recognize, occurs with nevirapine and is especially common in women who have a CD4 cell count prior to treatment that exceeds 250. This form of liver disease is accompanied by symptoms, usually gastrointestinal symp­toms, fever, and/or a rash. The great majority of cases occur during the first 16 weeks of treatment with nevirapine, and the drug must be stopped immediately. Because of this reaction, the standard of care is to test liver function frequently during the first 24 weeks of nevirap­ine.

Treatment: The treatment of these liver complications is generally restricted to deciding whether they are severe enough, or potentially se­vere enough, to discontinue the drug. Otherwise, except for supportive care, liver complications have no specific treatment.

Hepatitis and antiretroviral drugs: Some people taking antiretrovi­ral drugs have chronic hepatitis, either hepatitis C or hepatitis B. In these cases, liver disease is commonly associated with the changes in liver func­tion, and it is often impossible to tell whether the cause is the hepatitis virus or the antiretroviral drug. Most HIV care providers, who deal with this on a regular basis, try to determine the cause by using a variety of tests, sometimes a liver biopsy. One form of serious liver disease that is very important is chronic hepatitis caused by the hepatitis B virus. Three of the drugs commonly used for treating HIV are also effective against the hepatitis B virus: tenofovir (TDF), lamivudine (3TC), and emtric- itabine (FTC). So the good news is that you can treat both HIV and he­patitis B viral infections with the same drugs. The bad news is that dis­continuing any of these drugs may result in a flare-up of the hepatitis virus, a serious complication that requires taking some drug for hepati­tis B.

Warnings: The two liver diseases from antiretroviral drugs are the nevirapine (NVP) reaction and the flare-up of the hepatitis B virus. The nevirapine reaction generally occurs during the first 16 weeks of nevi­rapine treatment and is always associated with symptoms that can be verified by laboratory testing. The flare-up of the hepatitis B virus oc­curs when one of the three nucleosides that treat hepatitis B is discon­tinued.

Diabetes

Frequency: Protease inhibitors are associated with something called “insulin resistance,” which simply means that the body’s insulin oper­ates less effectively. About 3 to 5 percent of the people taking PI-based HAART get diabetes. Nearly all PIs can cause diabetes, but the fre­quency is variable: diabetes is particularly common with indinavir and infrequent or nil with atazanavir. The change in blood sugar takes place almost immediately after these drugs are given.

Symptoms: The usual symptoms of diabetes are increased appetite, large volumes of urine, and weight loss. In fact, we are usually aware of diabetes long before these symptoms are found. The reason is that a blood sugar is included in most of the standard tests that are done at 3- to 4-month intervals in most people taking HAART. Diabetes is defined by a blood sugar (glucose) level that is over 126.

Treatment: The standard treatment is a diabetic diet along with ex­ercise and weight loss, sometimes supplemented by pills referred to as “oral hypoglycemics.” Insulin treatment is rarely necessary unless the person is diabetes-prone as a result of genetics or other factors causing diabetes that are unrelated to antiretroviral treatment. Another option is to switch from the PI implicated to atazanavir or to a NNRTI-based antiretroviral regimen.

Warnings: Diabetes is usually asymptomatic and easily managed with diet and exercise, sometimes supplemented with oral drugs. For people with cardiovascular disease, diabetes is yet another risk factor for a stroke or a heart attack. Other people, particularly those who are diabetes-prone, may have unusual difficulty controlling their levels of blood sugar and may require insulin. These people might do best to change treatment, particularly when they have many options for equally effective antiretrovirals.

Fat Redistribution

Definition: Fat redistribution means a changed distribution of fat and comes in two forms. One, called fat atrophy, is the loss of fat in the arms, legs, and buttocks and in the face only above the upper lip. Fat at­rophy is usually a side effect of NRTIs, especially d4T but also ddI and AZT. The other, called visceral fat redistribution, is on the trunk: an ac­cumulation of fat on the back of the neck that looks like a buffalo hump; an increase in breast size in both men and women; and an increase in ab­dominal fat, particularly on the lower abdomen. Visceral fat redistribu­tion is usually a side effect of protease inhibitors.

Frequency: How often fat redistribution occurs is hard to judge be­cause it is totally subjective. A few laboratory measurements can con­firm what we see; but this is a cosmetic issue, and its importance is re­ally in the mind of the person with these changes. The methods to measure the changes are not very useful because they might miss some of the ones that the person thinks are most important and they might emphasize the ones that people will ignore. For example, the appearance of sunken cheeks is easy to see but hard to measure. Nevertheless, those who have tried to define the condition of fat redistribution have reported fat accumulation in 20 to 60 percent of people taking long-term PI-based HAART, and fat atrophy in 30 to 70 percent of people taking long-term d4T.

Significance: Although this is a cosmetic effect, self-image can be a very important factor in the quality of life. The loss of fat in the face may particularly bother people.

Treatment: Sometimes plastic surgery is successful: injections into the area of fat loss in the face, or liposuction of buffalo humps, or breast reduction. These procedures tend to be expensive and are generally not covered by third-party insurance companies. The alternative is “switch therapy”; that is, changing to other drugs. Switch therapy is easiest with d4T, which has many equally effective alternatives from the same class, but the reversal of fat redistribution is very slow and, according to some studies, virtually nil one to two years later. For PI-associated fat accu­mulation, switching drugs may not be particularly effective. Although the fat changes are commonly ascribed to PIs, they have also been noted with NNRTIs.

Increase in Blood Fats

Definition: Antiretroviral drugs are associated with substantial changes in blood cholesterol and another high-risk fat, triglycerides. The typical changes are an increase in the total cholesterol, the LDL choles­terol (the bad cholesterol), and the triglycerides.

Frequency: These increases have been reported in 40 to 70 percent of people taking protease inhibitors, and they occur within days of be­ginning treatment. The exception is atazanavir, which is not associated with these changes. An elevated triglyceride finding is also noted with d4T.

Context: The trade-off sounds terrible: the progression of HIV in­fection is reduced, but the risk of cardiovascular disease rises. But the risks must be viewed from the proper perspective. For a person with a CD4 cell count of 200 and the average risk profile for heart disease, not taking antiretroviral drugs increases the risk of death in the next five years by 50 times. Nevertheless, the increase in blood fats needs atten­tion and can be serious, and it can be treated. Most important is the as­sessment of other risks for cardiovascular disease. Contributing factors include previous cardiovascular disease with a heart attack or stroke (the greatest risk for a second event), hypertension (high blood pressure), di­abetes, smoking, elevated blood cholesterol or triglycerides prior to tak­ing antiretroviral drugs, and genetics. Genetics here means having a first- degree male relative with cardiovascular disease before the age of 45 or a first-degree female relative with cardiovascular disease before the age of 55.

Treatment: The most important treatment is life-style changes, in­cluding weight reduction, exercise, diet modification, and discontinua­tion of smoking. All authorities emphasize the importance of these meth­ods to control the risks, which might exist quite apart from HIV and its treatment. Beyond that, LDL cholesterol can be controlled with drugs, usually statins. However, this advice is complicated by drug interactions between statins and protease inhibitors or NNRTIs. Protease inhibitors increase the levels of some statins to dangerous concentrations; and efavirenz reduces statin concentrations to the point where they may not be particularly effective. These complexities are not too difficult to sort out using the standard guidelines. Triglycerides may also be elevated, also contribute to the risk, and also can be managed with drugs. Another approach is switch therapy, since atazanavir, efavirenz, and nevirapine are relatively free of this side effect.

Neuropathy

Definition: Neuropathy is a side effect of d4T or ddI. It occurs in about 20 percent of people who take these drugs for more than a year and is more common in those who take both drugs together.

Symptoms: The symptoms are pain and tingling in the feet, but con­tinued use of the drugs may result in more pain and involvement of the arms as well. Eventually people might have relatively severe pain, caus­ing difficulty walking and eventually requiring narcotics if the drugs are not stopped. If they are stopped soon enough, the changes go away; if they are continued for a long time, the pain is permanent.

Treatment: There are a variety of treatments, but none are particu­larly effective. It should be noted that this neuropathy is identical to a neuropathy attributed to HIV infection itself.

Warnings: Peripheral neuropathy will continue to get worse as long as you take the drugs that cause it. Therefore it is important to report this complication as soon as it occurs.

Side Effects Due to Individual Drugs

The side effects reviewed above are “class reactions”: they are com­mon to many drugs within a class. Some important side effects, however, occur only with specific drugs.

Abacavir hypersensitivity reaction: About 5 to 8 percent of people taking abacavir have the abacavir hypersensitivity reaction. It is serious. The symptoms are high fever, a rash all over the body, nausea, chills, di­arrhea, vomiting, abdominal pain, and sometimes additional symptoms. It typically occurs early in the course of taking abacavir, usually around the ninth day of treatment; 90 percent of the reactions occur in the first six weeks. The drug should be stopped and not started again until the physician is certain that the symptoms were not from an abacavir hy­persensitivity reaction. A test dose can result in a much more serious re­action. The concern is that warning people about the seriousness of this reaction and the multitude of symptoms can cause great anxiety. People are tempted to discontinue the drug even when they have only a simple cold, gastroenteritis, or some other common complaint; or when they have side effects from another drug that was taken with abacavir, such as the rash that is seen with nevirapine or efavirenz.

An abacavir hypersensitivity reaction almost always involves two systems. A rash alone (skin only) or cold symptoms alone (respiratory tract only) almost never mean that you have this side effect. The stakes in diagnosing the abacavir hypersensitivity reaction are sizable because if that’s what you have, and if you take abacavir again, you may react so severely that you can never take the drug again. One possible ap­proach in difficult cases is to have abacavir administered in a clinic un­der observation; a characteristic feature of the reaction is that the next administration is always associated with more profound reaction. Some interesting studies suggest a genetic predisposition to the abacavir hy­persensitivity reaction, but this information has not yet evolved into a clinically useful tool to predict it or diagnose it.

Efavirenz-associated central nervous system complications: People who take efavirenz often experience an odd complication involving the brain: a “disassociated feeling,” along with dizziness, sleepiness, abnor­mal dreams, difficulty with concentration, or depression. Up to 50 per­cent of the people who take this drug have this reaction. The good news is that the reaction is generally limited to the first 2 to 3 weeks of treat­ment. The cause is completely unknown and there is no treatment. Per­sons taking efavirenz need to be warned about this side effect so that they will know that it is normal and can be expected to simply disap­pear in 2 to 3 weeks as they continue to use the drug.

Indinavir-associated kidney stones and a dry skin and mouth: Indi­navir may form crystals of indinavir in the kidneys and cause a kidney stone. This is actually noted in 15 to 20 percent of the people who take indinavir for a prolonged period. The symptoms are exactly like the symptoms of kidney stone due to other causes. It can be terrifying pain that requires narcotics for relief. This side effect is the reason that large volumes of fluid are required when you are taking this drug. Another odd reaction associated with indinavir is something called “sicca,” which involves dry skin, a dry mouth, dry eyes, and possible hair loss. The cause of the sicca syndrome is not known, but most people who take the drug will complain of being “dry.”