Free Web Hosting Provider - Web Hosting - E-commerce - High Speed Internet - Free Web Page

Acquired immunodeficiency syndrome is a worldwide pandemic. Since the initial identification of five cases of AIDS in Los Angeles in 1981, the number of reported cases has increased progressively. Approximately half of all cases of HIV-1 infection are in sub-Saharan Africa. HIV-1 is spreading most rapidly in Asia, with over five million cases estimated. The epidemic continues to expand in South America but appears to have plateaued in North America and Europe.

HIV-1 is transmitted by intimate sexual contact, blood-borne contamination, and vertical transmission. The worldwide epidemiology of HIV-1 is a direct reflection of the transmission patterns of the virus. Initially concentrated in homosexual men, the epidemic in the United States has gradually shifted to involve a greater proportion of persons who have acquired HIV-1 through injection drug use or intimate heterosexual contact. The number of women, African Americans, and Hispanics infected with HIV-1 has grown steadily. In developing countries, the virus is spread primarily through heterosexual contact and affects males and females on a more equal basis than in Western countries.

The efficiency of HIV-1 transmission varies with the route, the quantity of virus, and certain phenotypic characteristics of the virus that are incompletely defined. Sexual contact is a relatively inefficient mode of transmission. The estimated risks of sexual transmission of HIV-1 are one in 300 for male-to-male transmission; one in 500 for male-to-female transmission; and one in 1,000 for female-to-male transmission. However, it should be emphasized that these are only average risks. Some infected persons appear to be much more efficient than others at transmitting the virus to sexual partners, and some persons remain uninfected despite extensive exposure to the virus through unprotected sexual contacts. Factors accounting for differences among persons with respect to the ability to acquire or transmit HIV-1 have not been fully delineated. The presence of other sexually transmitted diseases (e.g., chancroid) increases the risk of transmission of HIV. It is less clear whether the stage of disease in the transmitting partner plays a strong role in determining the transmission risk. Epidemiologic and biologic data indicate that persons homozygous for altered chemokine receptor genes have a lower risk of infection; the degree to which heterozygotes are protected is much less significant. At the outset of the epidemic, concerns were raised that HIV-1 might be transmitted by casual contact in the workplace, at schools, and in nosocomial settings. However, after a decade and a half, the continued concentration of the virus in persons with the same risk behaviors identified at the beginning of the epidemic is strong evidence that HIV-1 is not transmitted by casual contact.

Knowledge of the modes of HIV-1 transmission has had an important impact on curbing the spread of the virus. Transmission by intimate sexual contact has been greatly reduced by the proper use of condoms. Needle-exchange programs reduce the risk of transmission of HIV-1 associated with injection drug use. Transmission of HIV-1 through transfusion of blood products and organ transplantation has been almost eliminated in settings where high-risk donors are deferred and HIV-1 serologic screening is performed. In the United States, the estimated risk of acquiring HIV-1 after receiving one unit of blood is one in 500,000. The use of zidovudine (also known as AZT) in the perinatal period reduces the risk of HIV-1 transmission by approximately 67%. Postexposure prophylaxis with zidovudine after needle-stick injuries reduces the risk of acquiring HIV-1 by approximately 80%.

AIDS is the last stage of a disease process that has an average course of a decade. The initial stages of HIV-1 infection may be subclinical or associated with a mononucleosis-like syndrome characterized by a combination of symptoms that may include fever, a fleeting maculopapular eruption, aseptic meningitis, and lymphadenopathy. This constellation of symptoms is noted in 40% to 60% of patients with primary HIV-1 infection. The primary infection syndrome generally follows acquisition of the virus by 2 to 8 weeks and resolves spontaneously over 2 to 4 weeks. Most patients experience no overt clinical manifestations of illness for 6 to 10 years after primary infection. Despite this clinical quiescence, viral replication occurs at a rapid rate. The rate of disease progression varies from person to person and appears to depend on both viral and host factors. If heterozygosity for chemokine receptors plays a role in determining the natural history in HIV-1 infection--a controversial hypothesis--the effects are modest. Patients who experience symptomatic primary infection have a significantly worse prognosis than those who undergo asymptomatic seroconversion.

On any given day, about 10 billion viral particles are produced in an infected person. Ninety-nine percent of these viral particles are produced by activated CD4⁺ T cells, which are killed when the virus enters the lytic stage of infection. This massive amount of viral replication, coupled with the replicative infidelity of HIV-1, results in both a major increase in the turnover of CD4⁺ T cells and the rapid generation of diverse viral quasispecies. The accelerated rate of CD4⁺ T cell turnover gradually results in increasing immunodeficiency and the onset of opportunistic complications of HIV-1 disease.

A formal diagnosis of AIDS is made when the CD4⁺ T cell count falls below 200 cells/mm³ or when a patient experiences his or her first AIDS-defining opportunistic infection or neoplasm. However, this distinction has little bearing on clinical management and is useful primarily for historical and epidemiologic reasons. As HIV-1 infection progresses, patients are at increased risk for clinical manifestations directly associated with the disease, including HIV encephalopathy (also known as AIDS dementia complex) and HIV-1-associated wasting. Death is usually the result of wasting or an opportunistic infection or neoplasm.

HIV-1 infection is usually diagnosed after a positive result on serologic testing that was prompted by a suggestive clinical finding or by known behavioral risk factors of the patient.

Male homosexuals ,intravenous rug abusers ,sexual partner of infected persons hemophiliacs treated with blood products and recipient of blood transfusion , clinicians must be prepared to suggest HIV-1 testing to any patient suspected to be at risk. HIV-1 testing should be strongly recommended to virtually all pregnant patients. A lesser risk is associated with nosocomial occupational exposures and with a history of blood transfusion or organ transplantation, especially if the procedure took place. Clinical syndromes that should prompt consideration for HIV-1 testing are any of the HIV-1-associated opportunistic infections (e.g., Pneumocystis carinii pneumonia, cryptococcosis, and cytomegalovirus [CMV] infection), opportunistic neoplasms (e.g., Kaposi's sarcoma), and so-called softer signs, including oral or vaginal thrush, hairy leukoplakia, herpes zoster (especially in persons younger than 50 years), new or worsening eczema or seborrheic dermatitis in an adult, and unexplained lymphadenopathy or weight loss. Laboratory abnormalities that should prompt testing for HIV-1 are unexplained leukopenia (especially with concomitant lymphopenia) and thrombocytopenia. On occasion, asymptomatic persons are diagnosed with HIV-1 infection when they test positive on donating blood or when they undergo examination on applying for employment or life insurance.

The current serologic tests, most of which are based on enzyme-linked immunosorbent assay (ELISA) technology, are highly sensitive and specific for HIV-1. Positive results by ELISA are confirmed by a second test, such as Western blot assay, that detects antibodies to specific HIV-1 antigens. If the patient was recently exposed to the virus, serologic testing should be repeated at 6 weeks, 12 weeks, and 6 months after the exposure to allow sufficient time for potential seroconversion. When a diagnosis of primary HIV-1 infection is under consideration, testing for HIV-1 p24 antigen or HIV-1 RNA in plasma or serum is recommended. This allows for detection of infection in the so-called window period before the appearance of HIV-1 antibodies. False positive test results on HIV-1 RNA assays can occur, however, particularly when the RNA level is near the lower limit of detection in the assay used; a confirmatory test is especially important.

Once a diagnosis of HIV-1 infection is established, it is always in the patient's best interest that he or she be referred to a clinician with specialized expertise in HIV-1 medicine. The initial evaluation should include an assessment of the patient's clinical status, counseling about transmission risk to other persons, and a discussion of whether HIV-1 testing should be offered to persons whom the patient may have exposed to the virus.

The patient should be carefully assessed for clinical symptoms and signs of HIV-1 infection, and a laboratory evaluation should be made to stage the severity of infection and assess the risk of reactivation of opportunistic pathogens . Certain clinical features, including thrush or prominent systemic symptoms known as B symptoms, suggest more advanced disease and a propensity for more rapid disease progression. The CD4⁺ T cell count is the most useful indicator of immunologic dysfunction and the immediate risk for opportunistic infections, whereas the level of HIV-1 RNA in plasma is the best predictor of the rate of clinical and immunologic progression . Such prognostic information is often of interest to the patient and provides useful insights to help formulate the initial antiretroviral chemotherapeutic strategies.

The risk of HIV-1-associated opportunistic infections with specific viral, fungal, bacterial, and mycobacterial pathogens increases progressively as the CD4⁺ T cell count falls below 200 cells/mm³ In most cases, the clinical syndromes produced by opportunistic pathogens represent reactivation of latent infections acquired earlier in life that become symptomatic as cellular immunity declines. Thus, distribution of clinical syndromes in a given population depends in part on the reservoir of latent pathogens in that population. In the United States, the most commonly observed opportunistic infections are P. carinii pneumonia , CMV infection , disseminated Mycobacterium avium complex infection , and Toxoplasma gondii encephalitis . The most common neoplastic complications are Kaposi's sarcoma and B cell lymphoma . Certain pathogens, such as P. carinii and M. tuberculosis, are more likely to present as initial manifestations of HIV-1 infection at higher CD4⁺ T cell counts (around 200 cells/mm³), whereas other pathogens, such as CMV and T. gondii, are generally seen at lower CD4⁺ T cell counts (50 to 100 cells/mm³). Of importance is that the occurrence of these infections varies greatly from patient to patient.

It is therefore important to establish the presence of pathogens that may reactivate if preventive measures are not taken as immunologic function declines. Patients should also be assessed for pathogens that share transmission routes with HIV-1, such as T. pallidum and hepatitis B virus (HBV). Morbidity from complications of opportunistic infections has been greatly reduced by more aggressive use of primary and secondary prophylaxis and the improved ability of clinicians to recognize early clinical manifestations, thus facilitating effective early therapy of individual bouts of infection.

The initial medical management of HIV-1 infection is determined primarily by the stage of the illness and entails antiretroviral chemotherapy and prophylaxis against opportunistic infections through vaccination and chemotherapeutic interventions. Vaccinations produce better immunologic responses in persons with higher CD4⁺ T cell counts. Although data are insufficient to indicate a specific CD4⁺ T cell count below which any particular vaccine loses efficacy, the consensus is that vaccine responsiveness declines progressively as the CD4⁺ T cell count falls from 500 cells/mm³ to 200 cells/mm³. Certain live attenuated vaccines, such as oral polio vaccine and vaccinia virus vaccine, should not be administered to HIV-1-infected persons. Antiretroviral chemotherapy is associated with an improvement of cell-mediated immune responsiveness, including responsiveness to vaccines. Thus, if vaccinations are contemplated in persons who are candidates for antiretroviral chemotherapy, the vaccinations should be delayed until 4 to 6 weeks after the antiretroviral chemotherapy is initiated. This approach has the advantage of blunting any transient bursts of viral replication that may be associated with vaccine-induced T cell activation. Although data are lacking regarding clinical efficacy, most experts recommend use of pneumococcal vaccine in all previously unvaccinated patients.

Hepatitis B vaccination is recommended by some experts but should be reserved for patients who are at increased risk for HBV. Influenza vaccination has become controversial on the basis of data indicating that in some patients, vaccination is associated with a transient rise in plasma HIV-1 RNA; however, more recent investigations have questioned these data. Because influenza infection would probably have a more profound and prolonged effect on HIV-1 plasma RNA levels than vaccination, it is advisable to offer influenza vaccination to persons who are likely to be immunologically responsive while undergoing antiretroviral chemotherapy.

Chemoprophylaxis for opportunistic pathogens is an important component of the initial patient encounters. Patients who test positive on purified protein derivative (PPD) tuberculin skin testing should receive prophylactic isoniazid therapy . Although PPD testing may be falsely negative in persons with advanced HIV-1 infection who reside in areas where the risk of tuberculosis is high, empirical isoniazid prophylaxis is not recommended in such cases. Prophylaxis for P. carinii pneumonia is recommended for patients with a history of P. carinii pneumonia and those with CD4⁺ T cell counts below 200 cells/mm³ . The presence of thrush, systemic B symptoms, or both is an independent predictor of P. carinii pneumonia and should also prompt the initiation of primary prophylaxis. Primary prophylaxis for T. gondii should be considered for HIV-1-infected persons who are seropositive for this organism . Because T. gondii seldom causes encephalitis in persons with CD4⁺ T cell counts below 50 cells/mm³, prophylaxis can be delayed until later stages of AIDS. The use of trimethoprim-sulfamethoxazole, dapsone, or atovaquone for P. carinii prophylaxis also provides prophylaxis for T. gondii.

Primary prophylaxis for other HIV-1-associated pathogens is more complex but should be an integral component of the initial patient management strategy. Primary prophylaxi against M. avium complex infection in AIDS patients has become increasingly important and should be initiated when feasible and tolerable for the patient. Rifabutin, clarithromycin, and azithromycin have partly prevented M. avium complex infection. Most clinicians reserve these agents for persons with CD4⁺ T cell counts below 100 cells/mm³ because the agents add cost, complexity, and toxicity to the management of patients with AIDS, particularly in the era of HIV-1 protease inhibitor therapy and its attendant drug-drug interactions [see Antiretroviral Chemotherapeutic Agents, HIV-1 Protease Inhibitors, below]. In addition, the use of macrolides for primary prophylaxis may induce resistance, making them ineffective treatment for subsequent established infections. Nonetheless, it has been shown that one of these agents, clarithromycin, confers a survival benefit.

Oral ganciclovir therapy reduced the risk of CMV retinitis in one randomized, placebo-controlled trial in which patients were followed prospectively by ophthalmologists. Because of ganciclovir's cost and toxicity, however, CMV prophylaxis with ganciclovir is not routinely recommended.

It was once possible to offer a rather simple algorithm for the use of antiretroviral chemotherapeutic agents in the management of HIV-1 infection. However, as the number of antiretroviral agents with nonoverlapping toxicity and resistance profiles increases, the number of possible combinations and sequences of antiretroviral drugs is increasing even more rapidly. The proper use of antiretroviral chemotherapeutic agents requires an understanding of current concepts of HIV-1 pathogenesis; a knowledge of virologic, immunologic, and clinical tools for staging the illness; and an in-depth understanding of the activity, pharmacology, and toxicity and resistance profiles of approved and experimental antiretroviral agents. Although no single approach is appropriate for all patients, several principles exist that can guide the clinician in this aspect of patient management .

Clinical latency is not synonymous with viral or immunologic latency. The exact time at which therapy should be initiated depends on several factors: (1) the degree of immunodeficiency as judged by the CD4⁺ T cell count, (2) the risk of disease progression as predicted by the plasma HIV RNA level, (3) the presence or absence of clinical symptoms, and (4) the willingness of the patient to commit to a complex medical regimen for a prolonged period. On the basis of these considerations, most experienced clinicians recommend antiretroviral therapy for all infected persons with symptoms and all asymptomatic HIV-1-infected persons with a CD4⁺ T cell count below 500 cells/mm³. The greater risk of disease progression for persons with CD4⁺ T cell counts above 500 cells/mm³ and with 5,000 to 10,000 copies/mm³ of HIV-1 RNA in plasma suggests that antiretroviral chemotherapy should be offered to asymptomatic persons in this category as well, though prospectively generated clinical data are not yet available to support this recommendation.

Response to antiretroviral therapy is currently considered achieved if plasma HIV-1 RNA levels are driven below 50 copies/mm³. The durability of the antiviral response depends critically on the ability of the regimen selected to drive viral replication rates to a level below which resistant viral quasispecies are detected. Because this requires the ability and willingness of the patient to adhere to a complex regimen of antiretroviral chemotherapeutic agents, individualization with respect to when therapy should be initiated is crucial.

In most patients, at least two nucleoside analogue reverse transcriptase inhibitors and a potent protease inhibitor are required to achieve sufficient suppression of viral replication. Although less aggressive regimens may initially suppress viral replication, it is not possible to accurately determine which patients will respond to less aggressive regimens. Because failure to achieve near-complete suppression of viral replication results in the emergence of resistance to drugs in the chosen regimen, it is not generally in the patient's interest to start antiretroviral chemotherapy with less aggressive regimens. Antiretroviral drugs must be taken according to a relatively rigid schedule to achieve maximal benefit; some clinicians have therefore advocated withholding protease inhibitors in patients who are likely to have difficulties with compliance. This approach is unsound: the use of two nucleoside analogues alone, though associated with substantial clinical benefit, is universally associated with the emergence of resistant virus. It is usually better to defer therapy until a patient is able to take a complex regimen (or until less complex regimens are devised) than it is to initiate therapy with an inferior regimen.

The specific choice of drugs should be determined by the ability of the patient to comply with the regimen prescribed. Combinations of nucleoside analogues should be chosen on the basis of nonoverlapping toxicities and, when possible, to maximize the number of mutations the virus must incorporate to replicate in the presence of all the drugs in the regimen. The three most frequently chosen regimens are zidovudine and lamivudine, stavudine and lamivudine, and didanosine and stavudine. Recently completed clinical trials suggest that these regimens have the same general level of potency. The combination of stavudine and zidovudine should be avoided because of antagonism between these two agents. The choice of a protease inhibitor to be combined with the nucleoside analogues should also be based on considerations of tolerability and potency. Currently, ritonavir, indinavir, and nelfinavir are the only protease inhibitors with sufficient potency to be included in an initial regimen. Although direct randomized comparisons have not been done, ritonavir is probably slightly more potent than the other two drugs. However, complaints by patients taking the required dosage of ritonavir, 600 mg twice a day, suggest that better compliance is achieved by treating patients with either indinavir or nelfinavir. Indinavir is possibly slightly more potent than nelfinavir, but significant effects of food on the bioavailability of indinavir make it an unwise choice in patients who are unable to follow a relatively rigid schedule with respect to drug administration and food intake. The choice of the optimal initial protease inhibitor in the individual patient requires a detailed discussion with the patient regarding toxicities and inconveniences. It is extremely important to emphasize to patients that the initial choice is provisional and that if they experience side effects that limit adherence, they should promptly contact the physician to determine whether an alternative drug would be preferable.

A large number of clinical trials are under way to determine the efficacy of other types of initial regimens. These include regimens with two nucleoside analogues and a nonnucleoside reverse transcriptase inhibitor (NNRTI), regimens with two protease inhibitors, and regimens with protease inhibitors and NNRTIs. Although several studies have demonstrated that such regimens have excellent acute antiviral effects, the longer-term durability of these regimens has not been established with the same degree of certainty as the combination regimens with two nucleoside analogues and a protease inhibitor.

Regardless of the regimen that is employed, all patients should be seen 2 to 3 weeks after the initiation of therapy to discuss compliance and to assess the antiviral response. Although the maximal effects of potent regimens will not be evident for as long as 20 weeks, a substantial reduction in plasma HIV-1 RNA levels should be evident by the second week of therapy. If a reduction in HIV-1 RNA levels is not observed, the clinician should ask the patient whether he or she has been able to adhere to the regimen. If side effects are limiting adherence, this is an excellent opportunity to make changes in the drugs. If plasma HIV-1 RNA levels have declined by 1.0 log₁₀ or more by this time, it is reasonable to reassess the patient 12 to 16 weeks after the initiation of therapy. If the regimen has lowered plasma HIV-1 RNA levels to less than 50 copies/mm³, and if the patient is tolerating the regimen, patients can be followed at intervals of 8 to 12 weeks to assess the degree to which suppression of viral replication is maintained. If plasma HIV-1 RNA levels are several thousand copies or more, it is generally advisable to seek an alternative regimen if one is available. In patients with plasma HIV-1 RNA levels in an intermediate range of 50 to 1,000 copies/mm³, it is reasonable to recheck plasma HIV-1 RNA levels in 4 weeks. If the plasma HIV-1 RNA level has continued to fall, this sequence can be repeated. Studies are under way to address the wisdom of intensifying regimens for patients with very low but measurable levels of plasma HIV-1 RNA.

Once the initial regimen adequately suppresses viral replication, patients should be followed at intervals of 2 to 3 months for toxicity and for loss of antiretroviral effect as evidenced by plasma HIV-1 RNA levels. Note that vaccinations and intercurrent infections can transiently raise plasma HIV-1 RNA levels, and therapeutic decisions should not be made on the basis of a single plasma HIV-1 RNA level. When suppression of viral replication is lost, the clinician should alter the regimen, generally by substituting at least two new drugs for components of the previous regimen. The precise level of plasma HIV-1 RNA at which suppression of viral replication should be considered lost has not been established. It is clear that patients continue to derive substantial clinical and immunologic benefits when low levels of plasma HIV-1 RNA are detected. However, smoldering viral replication in the presence of selective pressure of antiretroviral drugs inevitably leads to increased drug resistance. This is particularly troublesome in the case of protease inhibitors in that resistance to other drugs in the class is progressively acquired because mutations in the viral protease gene result when low levels of viral replication occur in the presence of the initial protease inhibitor. Agents in secondary and tertiary regimens should be chosen on the basis of tolerability and the probability of cross-resistance with drugs previously given to the patient. Although clinical progression and a substantial loss of CD4⁺ T cells have been used as indicators of antiretroviral drug failure, these parameters are less sensitive than virologic parameters and should no longer be used as primary indicators of drug failure.

Of importance is that although the goal of antiretroviral therapy is to achieve almost complete suppression of viral replication, this is not achievable in all patients, especially those in late stages of infection. Antiretroviral drug therapy in later stages of AIDS is complicated by increased drug intolerance, resistance of the virus to available drugs on the basis of prior therapy, and an increased probability of drug-drug interactions. In general, however, even in the late stages of AIDS, most patients can be treated with one or more antiretroviral drugs. The patient will still derive measurable clinical benefit if the plasma HIV-1 RNA level is at least 0.3 to 0.5 log₁₀ lower than it would be without therapy. Therapy should be stopped if no drugs can be tolerated within the context of other agents that are required in a given patient or if no evidence of antiretroviral activity can be demonstrated by HIV-1 RNA quantitation techniques.

Antiretroviral agents have been targeted primarily at two viral enzymes: reverse transcriptase and the HIV-1 protease. Agents have been developed under the assumption that specific inhibition of essential steps of the viral life cycle that involve these enzymes can be achieved without a significant impact on the metabolism of host cells.

The nucleoside analogue reverse transcriptase inhibitors have proved to be a useful although relatively weak class of HIV-1 replication inhibitors. In general, a reduction in plasma HIV-1 RNA levels of less than 1 log₁₀ is observed when these drugs are initiated as monotherapy in previously untreated persons. Nonetheless, even this degree of reduction of viral replication is associated with a transient increase in the level of CD4⁺ T cells and, depending on the patient population studied, a modicum of clinical benefit.

Zidovudine was developed as a potential antineoplastic agent before the AIDS era. In patients with HIV-1 infection, it is administered at a dosage of 600 mg orally daily in two divided doses.

Pharmacokinetics Zidovudine is rapidly absorbed by the gastrointestinal tract and accumulated by lymphocytes, monocytes, and other cells. It must be phosphorylated intracellularly by host cell enzymes to its triphosphate analogue to achieve antiviral activity. Although the plasma half-life of the parent molecule is less than 1 hour, the 6-hour half-life of zidovudine triphosphate in the intracellular compartment allows for 12-hour dosing intervals. Zidovudine is glucuronidated by hepatic enzymes and is excreted in the urine.

Toxicity The major dose-limiting toxicity of zidovudine is suppression of myelopoiesis and erythropoiesis. Macrocytosis of the red blood cell series develops with such regularity over the first 6 to 12 weeks of therapy that increases in the mean corpuscular volume can be taken as a direct indicator of compliance. The development of anemia and granulocytopenia is related to dosage and disease stage, though some persons in earlier phases of HIV-1 disease may develop severe anemia or granulocytopenia. Bone marrow suppression is exacerbated by the concurrent use of other myelosuppressive drugs, such as ganciclovir and antimetabolic agents used in the treatment of HIV-1-associated oncologic complications.

In addition to the effects of zidovudine on the bone marrow, the drug is associated with anorexia, headache, or both in approximately 15% of patients. In about half of affected patients, these adverse effects resolve within 2 weeks after initiation of therapy. However, in as many as 10% of patients, these symptoms may be so troublesome that the drug must be discontinued. In general, these side effects are not amenable to additional pharmacologic interventions directed at relief of symptoms.

A wide array of less common side effects of zidovudine have also been reported. Although extremely rare, the most serious of these toxicities is hepatic failure, which may be mediated by a mechanism similar to that which occurs in persons treated with fialuridine. Because this complication appears to occur more frequently in patients with preexisting liver disease, obesity, or both, questions have been raised about whether there is a true causal relation between zidovudine use and hepatic failure. Nonetheless, clinicians who use zidovudine and other nucleoside analogue reverse transcriptase inhibitors should be aware of this toxicity and its association with these underlying conditions.

Clinical utility The first clinical trials of zidovudine for the treatment of HIV-1 infection were conducted with persons in the late stages of infection, before the era of primary and secondary prophylaxis for AIDS-associated opportunistic infections. In these patients, the drug had a major short-term impact on disease progression and survival. Zidovudine therapy was associated with increased CD4⁺ T cell counts, improvements in cell-mediated immunity (as assessed by delayed-type hypersensitivity reactions), and a major decrease in disease progression and mortality. In subsequent studies performed in healthier patient populations in the P. carinii prophylaxis era, zidovudine monotherapy was associated with more modest clinical benefits. Although some investigators have interpreted these studies as evidence that zidovudine is not associated with clinical benefit, the results are consistent with the emerging concepts of disease pathogenesis. Zidovudine has modest antiviral activity when used singly, and zidovudine-resistant viral quasispecies emerge as a result of ongoing viral replication in the presence of error-prone reverse transcription. These clinical trials demonstrated that zidovudine monotherapy delays clinical evidence of disease progression in previously untreated persons with CD4⁺ T cell counts below 500 cells/mm³. Zidovudine also prevents or reverses HIV encephalopathy. Reductions in mortality have been demonstrated only in persons in the late stages of AIDS. Therapeutic regimens with didanosine alone or combinations of zidovudine and didanosine, lamivudine, or zalcitabine are associated with significantly better outcomes than zidovudine monotherapy; thus, few indications remain for the use of zidovudine as a single agent.

Didanosine was the second nucleoside analogue reverse transcriptase inhibitor to be evaluated as treatment for HIV infection. Didanosine is dosed by weight. Patients weighing 60 kg or more should receive 200 mg twice daily; those weighing less than 60 kg should receive 125 mg twice daily. Didanosine can also be administered in a single daily dose.

Pharmacokinetics Didanosine is administered orally as an inosine prodrug and is compounded with a buffer directed at gastric acid because of the acid lability of dideoxyadenosine. Dideoxyadenosine is further metabolized into the triphosphate derivative in the intracellular compartment. Cerebrospinal fluid levels are about 20% of serum levels. Didanosine is formulated as a chewable tablet that is not acceptable to all patients because of its large size and firm consistency; a reduced-mass tablet is now available. Absorption of didanosine is decreased if taken with food.

Toxicity The major toxicities associated with didanosine are pancreatitis and peripheral neuropathy. These side effects are dose related and are more often seen in persons with advanced disease. Didanosine-associated peripheral neuropathy is generally reversible if the patient stops taking the drug. Many patients tolerate resumption of the drug at a reduced dose. Unfortunately, persons who have experienced peripheral neuropathy in association with didanosine are at a greater risk for the development of this complication in association with zalcitabine or stavudine.

Didanosine-associated pancreatitis is potentially life threatening. Pancreatitis occurs more often in patients with late-stage disease, patients with diabetes, and patients who have had previous bouts of pancreatitis (to whom the drug should generally not be prescribed). Pancreatitis may occur at any time after initiation of the drug and may occur with such rapid onset that regular monitoring of the serum amylase or lipase level is not useful in preventing this complication. Patients who begin didanosine therapy should be warned about peripheral neuropathy and pancreatitis and instructed to discontinue the drug immediately if neuropathy or abdominal pain develops. Patients who experience abdominal pain should seek medical attention and be evaluated for pancreatitis before didanosine therapy is resumed.

Clinical utility Didanosine therapy has been evaluated in persons with moderate to advanced manifestations of HIV-1 infection. Many of these investigations studied persons who had previously used zidovudine. It was demonstrated that didanosine is superior to zidovudine in antiviral and immunomodulatory effects and provides additional clinical benefits to patients who have used zidovudine. Didanosine was also demonstrated to be superior to zidovudine in previously untreated patients with CD4⁺ T cell counts of 200 to 500 cells/mm³. This study examined patients who were in earlier stages of HIV-1 infection than patients in a previous study in which zidovudine appeared to be superior to didanosine as initial therapy. The differences in these two studies may be attributable to disease stage (in view of the preferential phosphorylation of zidovudine in activated lymphocytes) or to the longer study duration of AIDS Clinical Trial Group 175 (which found a more durable effect for didanosine). Although these trials suggest that didanosine monotherapy provides measurable clinical benefits, monotherapy with nucleoside analogues is no longer an acceptable initial regimen.

Zalcitabine (also known as dideoxycytosine, or ddC) is a nucleoside analogue reverse transcriptase inhibitor that exhibits potent antiretroviral activity in vitro. Dose escalation of zalcitabine is limited by peripheral neuropathy, however, and zalcitabine is therefore used only in combination regimens or for the treatment of patients who are intolerant of or unresponsive to zidovudine with either didanosine or lamivudine. Zalcitabine is administered at a dosage of 0.75 mg three times daily.

Pharmacokinetics Zalcitabine is readily bioavailable after oral administration and exhibits an intracellular half-life of 3 to 4 hours as the triphosphate derivative, which is sufficient to allow dosing three times daily. Bioavailability of the drug is not significantly affected by food.

Toxicity Peripheral neuropathy is the major dose-limiting toxicity of zalcitabine. This complication, which is related to dose and disease stage, occurs at doses below those that exhibit maximal antiretroviral activity in vivo. When the drug is stopped promptly, peripheral neuropathy is usually reversible. After resolution of symptoms, many patients tolerate reintroduction of the drug in reduced doses. Few patients experience subjective or objective evidence of zalcitabine toxicity other than peripheral neuropathy.

Clinical utility Because of the dosing limitations, zalcitabine has limited utility as a single agent and, with few exceptions, should not be used alone. It has been shown to be as good as or better than didanosine in persons with advanced AIDS who are intolerant of zidovudine therapy or for whom zidovudine therapy has failed. Because of its convenient formulation and dosing schedule, zalcitabine has been used extensively in combination regimens for persons with advanced AIDS who are intolerant of other antiretroviral chemotherapeutic agents.

Like zidovudine, stavudine is a thymidine analogue with significant antiretroviral activity in vitro. The drug has been investigated in patients with moderate to advanced HIV-1 infection, especially those with previous zidovudine experience. Stavudine therapy is based on the weight of the patient. Persons weighing 60 kg or more should be started on 40 mg of stavudine twice daily; persons weighing less than 60 kg should receive 30 mg twice daily.

Pharmacokinetics Stavudine is well absorbed after oral administration and exhibits an intracellular half-life of 3.5 hours as the triphosphorylated derivative, which allows dosing twice daily. The drug achieves CSF levels that are in the range of 25% to 50% of serum levels.

Toxicity Peripheral neuropathy, which has been observed in about 15% of patients receiving stavudine, is the major side effect. This toxicity is dose related and is more frequently observed in patients with advanced HIV-1 infection and in patients who have experienced previous nucleoside analogue-associated peripheral neuropathy. Stavudine-associated peripheral neuropathy is reversible if the drug is discontinued promptly; about half of patients tolerate resumption of the drug at half the original dose. Hepatitis is also observed in association with stavudine it is seen in about the same fraction of patients in whom peripheral neuropathy develops and is almost always reversible with discontinuance of the drug. Although hepatitis occurs in a significant percentage of patients, it is seldom life threatening if the drug is stopped on recognition of significant elevations in serum hepatocellular enzymes. Stavudine causes macrocytosis of red blood cells but is seldom associated with anemia.

Clinical utility Stavudine was initially developed for use as monotherapy in zidovudine-intolerant patients or those with prior zidovudine experience. In the largest clinical trial comparing stavudine with zidovudine, when persons who had taken zidovudine for an average of 18 months were given stavudine, the CD4⁺ T cell count increased by about 50 cells/mm³, which was greater than the increase seen in persons who remained on zidovudine. There was a strong trend toward clinical benefit, though the difference in disease progression and death between the two study arms was not statistically significant. Recent studies of the combination of stavudine with didanosine or lamivudine, however, have firmly established the utility of stavudine in combination regimens. ^41,42

Lamivudine is well tolerated and results in acute reductions in plasma HIV-1 RNA levels of approximately 1.5 log₁₀. However, a single mutation in reverse transcriptase at position 184 results in a 100-fold to 1,000-fold decrease in susceptibility to lamivudine. Allowing any measurable degree of viral replication in the presence of the drug results in the rapid emergence of resistant mutations. Thus, lamivudine should be used only in patients in whom almost complete suppression of viral replication can be expected. Although the currently recommended dosage of lamivudine is 150 mg every 12 hours, daily dosing is a feasible option based on the pharmacology of the drug.

Pharmacokinetics Lamivudine exhibits potent antiretroviral activity in vitro. The drug is rapidly and completely absorbed after oral administration and is phosphorylated intracellularly to a triphosphorylated derivative. The intracellular half-life of the triphosphorylated active metabolite is longer than 12 hours.

Toxicity Lamivudine is occasionally associated with suppression of the erythroid and myeloid elements of the bone marrow. However, this complication is dose and disease-stage related and is less frequently observed with lamivudine than with zidovudine.

Clinical utility Although lamivudine rapidly selects for resistant viral isolates with a mutation at position 184 in the reverse transcriptase, this mutation compromises the ability of the virus to simultaneously make mutations that are associated with resistance to zidovudine. Mutations that antagonize the simultaneous development of resistance to two or more antiretroviral agents have also been observed with didanosine and zidovudine and are termed suppressor mutations. Because of these suppressor mutations, combination therapy with lamivudine and zidovudine results in an increase in CD4⁺ T cell counts and a much more pronounced and prolonged reduction in viral load than does therapy with zidovudine alone. More recent data indicate that lamivudine can be combined with stavudine with excellent antiviral effects in vivo. The addition of lamivudine to regimens with nucleoside analogues was recently shown to have a profound effect on disease progression and death, resulting in 50% decreases in morbidity and mortality.

Abacavir is the first guanosine analogue nucleoside reverse transcriptase inhibitor to be used clinically. Currently in phase II/III trials, abacavir is extremely potent in vivo, resulting in reductions in plasma HIV-1 RNA levels of up to 2 log₁₀ in previously untreated patients. Abacavir is generally well tolerated. Toxicities that have been encountered in a minority of patients are mild headache and nausea. The only major concern is a shocklike syndrome that occurs in some patients who are rechallenged with the drug after experiencing a febrile rash during the early period of dosing. The mechanism of this toxicity is unknown, but the clinical manifestations are dramatic. Thus, patients experiencing a rash and systemic symptoms associated with abacavir should never be rechallenged with the drug. Appropriate clinical settings for the drug are being investigated. Because the in vivo activity of the drug is compromised by mutations at positions 65, 74, and 184, prior treatment with didanosine, lamivudine, or both may decrease the utility of abacavir. Preliminary data suggest that antiviral activity is still seen, however, if phenotypic susceptibility is reduced no more than fourfold, compared with wild-type virus, as a result of prior treatment with other drugs.

These drugs are a class of structurally diverse HIV-1 replication inhibitors that were developed from drug screening for activity against the HIV-1 reverse transcriptase. NNRTIs, including nevirapine, delavirdine, and efavirenz (DMP-266), exhibit significant antiretroviral activity in vitro and in vivo and rapidly generate viral variants with significantly reduced susceptibility. As in the case of lamivudine, the propensity for the virus to develop high-level resistance to NNRTIs, if it is allowed to replicate in the presence of selective pressure, necessitates that use of these agents be restricted to situations in which complete suppression of viral replication is anticipated.

Nevirapine acutely suppresses plasma HIV-1 RNA levels by as much as 2.0 log₁₀. When nevirapine is given as monotherapy, however, the virus rapidly becomes resistant to the drug by the incorporation of a single mutation at position 188 or 190 of the reverse transcriptase enzyme, with complete loss of activity of the drug resulting within 2 to 4 weeks. The drug is generally well tolerated, but the major adverse effect is a maculopapular eruption that occurs in 10% to 15% of patients within the first 2 weeks of drug initiation; the lesion usually resolves without the need to discontinue therapy. In some patients, however, the drug induces a Stevens-Johnson-like syndrome. In such cases, nevirapine should be stopped immediately, and the patient should not be rechallenged with the drug.

In cases of incomplete suppression of viral replication, nevirapine has modest and transient effects when part of a combination regimen. Regimens that succeed in driving plasma HIV-1 RNA levels to 50 or fewer copies/mm can be enhanced by the addition of nevirapine. Nevirapine induces the hepatic cytochrome P-450 system, thereby enhancing the metabolism of most HIV-1 protease inhibitors; thus, the dosage of drugs such as indinavir should be adjusted when used in combination with nevirapine.

Delavirdine is a potent inhibitor of HIV-1 replication in vivo, but as in the case of nevirapine, high-level resistance to the drug is encountered within the first several weeks of dosing if it is administered in regimens that do not completely suppress viral replication. Delavirdine is also associated with a rash in 10% to 15% of patients. It is not clear whether the prior occurrence of a rash with nevirapine increases the likelihood of a rash with delavirdine. Management of patients who develop a rash with delavirdine is identical to that with nevirapine.

The clinical utility of delavirdine is similar to that of nevirapine. When delavirdine is used in potent combinations that achieve a high degree of viral suppression, sustained antiviral responses can result. Although these responses are impressive, combination regimens with nucleoside analogue reverse transcriptase inhibitors and delavirdine have not been compared with similar regimens with protease inhibitors. Many investigators believe that combinations of two nucleosides and an NNRTI are slightly less effective than similar regimens with potent protease inhibitors. Thus, most clinicians would not consider regimens with NNRTIs to be equivalent or analogous to similar regimens with protease inhibitors. Unlike nevirapine, delavirdine is a modest inhibitor of protease metabolism, and it has been suggested that it be used in combinations to convert thrice-daily protease inhibitors to twice-daily drugs.

Efavirenz, an NNRTI in late stages of clinical development, is also a potent antiretroviral agent. In contrast to nevirapine and delavirdine, efavirenz is seldom associated with a rash. The drug may be associated with confusion or light-headedness in some patients. These adverse effects are related to drug levels and are generally rapidly reversible. Although it appears that high-level resistance to efavirenz requires more than a single mutation, the optimal strategy for use of efavirenz entails complete suppression of plasma HIV-1 RNA. Efavirenz is similar to nevirapine with respect to its induction of protease inhibitor metabolism.

HIV-1 encodes an aspartyl protease--a homodimer composed of two 99-amino-acid segments--which cleaves the initially synthesized polyproteins of HIV-1 into functional subcomponents. In the absence of this proteolysis, noninfectious viral particles are synthesized and the replicative cycle is broken. More than a dozen investigative groups have developed potent inhibitors of the HIV-1 protease, which has proved to be an excellent target for structure-based approaches to the designing of effective therapeutic agents. The first protease inhibitors to be developed were often poorly soluble in aqueous solution and, thus, only minimally bioavailable. However, successive iterations of molecular design yielded a series of compounds that are highly bioavailable and inhibit the HIV-1 protease at nanomolar concentrations. These protease inhibitors have demonstrated potent antiretroviral activity in vivo and have been generally well tolerated. Acute declines in plasma HIV-1 RNA levels by 1.5 to 2.0 log₁₀ have been demonstrated after initiation of therapy with the protease inhibitors indinavir, ritonavir, and nelfinavir.

Saquinavir was the first protease inhibitor to undergo intensive clinical development. It is one of the most active inhibitors of HIV-1 replication in vitro in this class and was well tolerated in clinical trials. The major challenge to the development of saquinavir is a severe limitation on bioavailability imposed by extensive first-pass hepatic metabolism. The currently recommended dosage is 1,800 mg/day; when the drug is given at a higher dosage, however, antiviral effects equaling those of indinavir and ritonavir have been demonstrated. Combination therapy has been the focus of the initial clinical application of saquinavir because of its limited bioavailability and an extremely difficult synthesis that complicates commercial development of the drug. Saquinavir has been demonstrated to increase and prolong the antiviral and immune-enhancing effects of zidovudine and zalcitabine. In two recently completed clinical trials, the addition of saquinavir to regimens with nucleoside analogues was shown to have a significant effect on morbidity and mortality in both pretreated and treatment-naive patient populations. A formulation of saquinavir that increases the oral bioavailability approximately fourfold is now available. Saquinavir bioavailability is dramatically enhanced by the coadministration of ritonavir. This strategy results in substantial durable antiretroviral activity in vivo, but drawbacks are the unpalatability of ritonavir, hepatotoxicity (especially in patients who have preexisting liver abnormalities), and the cost of two protease inhibitors.

Indinavir is highly active in vitro and is much more bioavailable than saquinavir. The bioavailability of indinavir is greatly compromised if taken with food, however, and successful use of the drug requires assiduous attention to scheduling. In phase I clinical trials, the drug was shown to acutely decrease plasma HIV-1 RNA levels by 2 log₁₀, resulting in significant increases in CD4⁺ T cells, even in extensively pretreated patients. In many patients, indinavir is associated with transient hyperbilirubinemia that superficially resembles Gilbert syndrome. This complication is reversible and is not dose limiting. The major dose-limiting toxicity is the formation of renal stones containing indinavir. Although this complication was encountered in as many as 15% of patients in phase I trials of indinavir, more assiduous attention to hydration has greatly decreased the incidence of nephrolithiasis. In most situations, the occurrence of a stone does not necessitate discontinuance of indinavir. In vitro, indinavir is synergistic with nucleoside analogue reverse transcriptase inhibitors. The combination of indinavir, zidovudine, and lamivudine is particularly potent and reduced plasma HIV-1 RNA levels to less than detectable in 90% of the participants in a trial of persons who had previously received only zidovudine. In combination with lamivudine and zidovudine, indinavir produced substantial clinical benefit in patients with advanced disease in a recently completed clinical trial.

Ritonavir is a potent HIV-1 protease inhibitor that is highly bioavailable and has demonstrated significant antiretroviral activity in phase I clinical trials. Although extremely potent, ritonavir is difficult to administer because of subjective toxicities, including circumoral paresthesias and GI side effects. Ritonavir was the first protease inhibitor to demonstrate clinical benefit, with the report that it decreased mortality and disease progression in a group of persons with advanced HIV-1 infection. The use of ritonavir is also complicated by the drug's effects on the hepatic metabolism of a variety of other chemotherapeutic agents. This, combined with its unpalatability, has limited its clinical utility.

Nelfinavir is the most recently approved HIV-1 protease inhibitor. Although it may be slightly less potent in vivo than ritonavir and indinavir, its superior tolerability has made it a popular agent. Nelfinavir is well absorbed with meals and has few adverse effects other than diarrhea, which is usually controlled with over-the-counter antidiarrheal medications. Although studies are under way to determine whether nelfinavir can be administered twice daily, currently, it is most prudent to administer the drug at a dosage of 750 mg three times daily. With respect to its in vivo antiviral activity, nelfinavir is additive with reverse transcriptase inhibitors.

Amprenavir is a potent HIV-1 protease inhibitor yet to be approved by the Food and Drug Administration. This drug is well tolerated, has no significant food interactions, and results in reductions of roughly 2.0 log₁₀ at a dosage of 1,200 mg twice daily. Clinical trials are under way to determine its utility in combination with other antiretroviral chemotherapeutic agents.

Given the rapid rates of viral replication, the highly error-prone nature of the HIV-1 reverse transcriptase, and the inability of currently available antiretroviral agents to completely inhibit HIV-1 replication, the development of resistance to antiretroviral drugs has been an inevitable consequence of drug exposure. The first demonstration of viral variants with reduced susceptibility to zidovudine emerged in 1989. Since then, viral variants resistant to all other antiretroviral agents in active use have been demonstrated.

The molecular mechanisms by which the virus develops resistance to antiretroviral chemotherapeutic agents are diverse and complex. In the case of zidovudine, resistance develops in a stepwise fashion, with multiple mutations required to confer a high level of resistance. As would be expected from the increasing size of the viral replicative pool with advancing disease, resistance tends to develop more rapidly in persons with more advanced infection; however, the rate at which resistance develops varies greatly from one patient to the next. The kinetics of the development of resistance are drug specific. In general, resistance develops much more rapidly to NNRTIs than to nucleoside analogues, though lamivudine is a clear exception. Resistance to didanosine and zalcitabine has been reported, but it appears to occur with less regularity than in the case of zidovudine.Resistance to HIV-1 protease inhibitors develops at a rate that is intermediate between that of resistance to nucleoside analogues and that of resistance to NNRTIs. Molecular studies suggest that the development of resistance in vivo may be a reflection of both the selection of preexisting variants and the evolution of resistant quasispecies. The flexibility of HIV-1 in the presence of selective pressure is best exemplified by the multiplicity of mutations the virus can incorporate in the presence of several drugs.

Certain molecular adaptations, resulting in suppressor mutations, appear to confer on the virus a significant selective disadvantage. This disadvantage provides the rationale for combination chemotherapy based on the specific selection of agents directed at invoking this mechanism. Such interactions have been demonstrated with the combinations of zidovudine and didanosine, zidovudine and lamivudine, and zidovudine and certain NNRTIs. Unfortunately, these interactions are only relative; the virus ultimately develops novel molecular variants that allow simultaneous resistance to agents in these combinations, often bypassing the suppressor mutation interactions.

Resistance to HIV-1 protease inhibitors is also complex, involving multiple mutations that confer stepwise increases in resistance. Phase I studies of HIV-1 protease inhibitors demonstrated that resistance develops more slowly if the virus is exposed at the outset to higher concentrations of the drug than if lower doses of the drug are used initially and escalated later. Although genotypic and phenotypic analyses of variants resistant to protease inhibitors indicate significant overlap among drugs of this class in terms of resistance, there are differences as well. The degree to which cross-resistance will complicate the clinical use of these agents has not yet been determined. In general, the degree of cross-resistance among protease inhibitors increases as the level of resistance to any member of the class increases. Careful monitoring is necessary to identify increased cross-resistance to HIV-1 protease inhibitors with continued drug administration in the presence of viral replication, so that future treatment options can be preserved to the greatest extent possible.

The clinical implications of resistance are also complex. In the case of NNRTIs, there is a close correlation between the emergence of resistant variants and the loss of antiviral activity in vivo. With nucleoside analogues - especially zidovudine - and protease inhibitors, in which multiple mutations are required to develop high-level resistance, the relation between resistance-associated genotypic and phenotypic changes in the virus and the loss of antiviral activity is much less straightforward. Resistance to an antiretroviral agent may produce other changes in the virus that have an independent impact on its behavior. For example, resistance to zidovudine is associated with more rapidly progressive disease, even in persons who are treated with other agents, such as didanosine; this suggests that the resistant phenotype gives the virus additional pathogenic qualities. On the other hand, viral variants that are 100-fold to 1,000-fold less sensitive to lamivudine appear to replicate with less vigor in vivo than the parent strain, suggesting that the position 184 mutation gives the virus a slight selective disadvantage.

The expense of susceptibility testing, along with the inability to extrapolate in vitro susceptibility results to in vivo application, has contributed to the conclusion drawn by most experts that in vitro susceptibility testing is not sufficiently developed for general application in the clinical management of patients. However, recent data suggesting an increasing prevalence of resistant viral variants (even in previously untreated patients) and improvements in technology make it likely that susceptibility testing will gradually become a part of clinical management over the next several years.

The demonstration that zidovudine therapy decreases the risk of perinatal transmission by 67% is one of the most important advances in antiretroviral chemotherapy. In this randomized, double-blind, placebo-controlled trial, zidovudine administered orally in the prenatal period, intravenously to the mother at time of delivery, and orally to the child after delivery reduced the probability of HIV-1 transmission from 25% to 8%, with little or no demonstrable toxicity to the mother or child. Although several studies have indicated that there is an increased risk of HIV-1 transmission by mothers with more advanced HIV-1 disease and with higher levels of plasma HIV-1 RNA, the effect of zidovudine in reducing transmission does not appear to be mediated through the drug's effect on maternal plasma HIV-1 RNA levels. In addition, there has been no determination of a threshold level of plasma HIV-1 RNA in the mother below which transmission of HIV-1 does not occur. Thus, antiretroviral chemotherapy should be administered to all HIV-1-infected pregnant women and their offspring. The choice of antiretroviral therapy in pregnant women is complex and should take into consideration both the goal of preventing perinatal transmission and the goal of optimally treating the mother. As in the case of nonpregnant women, the use of nucleoside analogue monotherapy is suboptimal, and in general, regimens that completely suppress viral replication should be prescribed. The effects of antiretroviral therapeutic agents on fetal anomalies and wastage have not been completely delineated, though evidence at hand suggests that these agents do not pose a major risk to the fetus. Ongoing studies are investigating the possibility that more aggressive antiretroviral therapy will further reduce perinatal transmission, and more data are being generated on recommendations for women who have been pretreated with zidovudine and who probably harbor zidovudine-resistant variants of the virus. While these studies are under way, it is prudent to involve an expert in antiretroviral chemotherapy, a pediatrician with particular emphasis in AIDS, and a high-risk neonatologist in the management of HIV-1-infected pregnant women.

1. The overall risk of acquiring HIV-1 after a percutaneous exposure with a sharp instrument contaminated with HIV-1- infected bodily fluids is approximately one in 300. The risk is increased when the injury involves a hollow-bore needle, when it involves a sharp instrument that has been in a body cavity (e.g., an artery or vein), and when it is more severe than a simple needle stick. In a retrospective case-control study, the use of zidovudine after a percutaneous exposure was associated with an 80% reduction in the risk of HIV-1 transmission. These data have prompted the Centers for Disease Control and Prevention to revise their recommendations to suggest that the risk of acquiring HIV-1 should be stratified on the basis of the nature of the injury and that, for all significant exposures, prompt chemoprophylaxis should be given, consisting of at least two agents to which the virus is unlikely to be resistant .