Dani-Margot Zavasky MD
Julie L. Gerberding MD, MPH
Merle A. Sande MD
Essentials of Diagnosis
When the first clusters of cases of Pneumocystis carinii pneumonia and Kaposi's sarcoma were identified in 1981, it was not known that this was just a prologue to one of the most significant pandemics in the history of mankind and that, < 20 years later, 50 million people would be affected. Our knowledge of the disease started with the dramatic growth in the number of cases of unusual opportunistic infections in patients without prior inherited immunodeficiency disorders. Acquired immunodeficiency syndrome (AIDS) entered the medical and colloquial vocabulary. Not until 1983 was the offending agent identified as a virus. Known initially as lymphadenopathy-associated virus, then human T-cell lymphotropic virus type III, and AIDS-associated retrovirus, the microorganism acquired its current name, human immunodeficiency virus (HIV), in 1986. See Chapter 40 for a discussion of the virology of HIV and other retroviruses.
AIDS has been now reported in more than 150 countries. The United Nations estimates that as of December 2000, 58 million people have been infected with HIV. 22 million people have already died, of which 3 million died in 2000. As we are entering the third decade of AIDS history, the epidemic continues to spread rapidly through several continents. A total of 16,000 persons are infected per day worldwide, including 1600 children < 15 years of age. Women constitute over 40% of the adult cases. About 7000 people die from complications of HIV infection every day. Most infected persons live in either sub-Saharan Africa (> 70%) or South/Southeast Asia (> 20%).
In recent years, the epidemic has been particularly explosive in South/Southeast Asia and Central and Eastern Europe. The predominant mode of transmission appears to be heterosexual contact in Asia and injection drug use in Europe. Increasing syphilis incidence in Central/Eastern Europe signifies an increase in sexually transmitted diseases (STDs) of all kinds, including HIV, in this part of the world. There were 700,000 HIV infected patients by the end of 2000 compared with just 420,000 at the end of 1999.
In the United States, 920,000 AIDS cases and 460,000 deaths from AIDS were reported through December 2000. The total number of infected individuals in the United States probably exceeds 1 million; 45,000 of which became infected in 2000. The epidemiology of HIV infection is changing in the United States as the pandemic evolves. In 1992, 50% of all cases occurred in gay men and 25% in intravenous drug users, whereas 15% were derived from heterosexual contacts. Currently, the greatest increase in viral spread is occurring in the heterosexual population, similar to that observed in sub-Saharan Africa. The trends also confirm increase of infection in blacks and Hispanics, with blacks constituting ~41% of all adult cases. Also, women now account for 23% of all adult cases in the United States. The epidemic appears to affect predominantly women of color, particularly black women of African descent living in the southeastern part of the United States, in small rural communities (as opposed to the large cities, where HIV transmission in this same population occurred in the 1980s). Cases of AIDS in Hispanic women of Puerto Rican ancestry also predominate in the Northeast, especially in the states of Connecticut, New York, and New Jersey. However, in the Southwest and the West, Hispanic women of mostly Mexican and Central American descent have the same infection rate as white women. The facilitators of transmission in this heterosexual population again appear to be multiple sexual partners and the exchange of sex for money and drugs, especially crack cocaine.
In 1997, HIV/AIDS dropped from the first to the second leading cause of death in young adults in the United States. In 2000 it wasn't even one of the top 10 causes of death in young adults. The death rate in black men is fourfold that in white men and is also higher in black women than white men. Improved medical management of HIV disease resulted in an overall 23% decrease in AIDS deaths, especially among homosexual men, beginning in 1996. Also, between 1992 and 1995, perinatal AIDS cases decreased by 27% as a result of the widespread use of maternal and infant antiretroviral prophylaxis with zidovudine (AZT).
Testing of donor blood and the use of recombinant or specially treated coagulation factors have now virtually eliminated blood products transfusion as a source of infection in developed countries. Transmission of infection by blood is now largely associated with the sharing of needles and syringes by injection drug users. In some areas of the world, the seroprevalence of HIV infection among injection drug users has been as high as 70%. Transmission of infection to health care workers is rare but occurs after 0.3% of exposures to blood during needle punctures.
Transmission of HIV does not occur through a casual (nonsexual) contact with infected individuals or through insect vectors. The virus has been detected in saliva, tears, urine, and breast milk, but with the exception of breast milk, these fluids are not sources of infection.
PREDICTORS OF DISEASE PROGRESSION
The course of HIV disease is dictated by the severity of the individual's immune deficiency and the resulting complications. The recent impact of highly active antiretroviral therapy (HAART) on the slowing progression of HIV infection has been dramatic; however, therapy is complicated and expensive, thus not available to the vast majority of patients worldwide that would benefit.
The determinants of progression of the HIV disease have been under intense scrutiny by the research community. Results of prior studies emphasized the role of high viral load, presence of particularly virulent strains (eg, those that produce syncytium formation in culture) and low T-lymphocyte function as predictors of rapid disease progression. Recently, several coreceptors for HIV have been identified. The two most important of these are CCR5 and CXCR4. Individuals homozygous for a deletion in the CCR5 gene are resistant to infection regardless of multiple exposures to the virus. Some infected persons who are heterozygous for this deletion have been noted to have a prolonged AIDS-free survival. In current clinical practice, however, two main markers are used to monitor disease progression and to evaluate the need for and response to treatment: CD4 lymphocyte count and viral load.
CD4 Lymphocyte Count (Immunologic Monitor)
CD4 lymphocytes are the major cellular target for HIV. Both absolute CD4 counts and the rate of decline have prognostic predictive value and are used to determine the need for antiretroviral therapy and for opportunistic infection prophylaxis. CD4 counts do not accurately indicate viral load per se, but they predict the short-term risks for progression to AIDS-related illness.
Quantitative HIV RNA (See Table 23-2) (Virologic Monitor)
The quantitative titer of circulating HIV RNA (viral load) is the single best predictor of long-term progression and response to treatment. Viral load is used to assess the steady-state concentration of HIV, the response to treatment, and, most likely, tissue HIV RNA production.
Two main methods are generally used to measure viral load: reverse transcriptase-polymerase chain reaction (RT-PCR) and branched-chain DNA (bDNA) amplification. (See Table 23-2) By 2001, three commercial assays based on these methods or modifications of these methods were used to measure viral load: Amplicor HIV Monitor (RT-PCR, lower limit of detection = 400 viral copies/mL), Quantiplex HIV assay (bDNA, lower limit of detection = 500 copies/mL), and Nuclisens (lower limit of detection = 40 copies/mL). Assays with even more sensitivity such as Amplicor Ultrasensitive (lower limit of detection = 20 copies/mL) and Quantiplex HIV 3.0 (lower limit of detection = 50 copies/mL) are currently undergoing clinical evaluation. To optimize interpretation, two measurements of viral load should be obtained on two separate occasions in the same laboratory using the same assay. When possible, viral load testing should be delayed until 4 weeks after intercurrent infections are treated, acute illnesses have resolved, and following an immunization.
Current guidelines for viral load testing suggest that the pretreatment measurement be obtained to evaluate the patient's prognosis and that tests be repeated every 3–4 months until treatment is started. Once treatment is initiated, a measurement should be obtained 4–8 weeks later to determine whether the expected initial therapeutic response (0.5–0.75 log reduction in titer) has been achieved. Another measurement should be made 3–4 months later to determine whether HIV replication is fully suppressed (below the level of detection). If not, then adherence should be assessed and treatment modifications considered. Once the patient is on a stable antiviral regimen, viral load tests should be obtained every 3–4 months to monitor ongoing virologic response.
HIV infection can produce a huge array of clinical syndromes. In general, these syndromes correlate with the duration of illness and severity of immunosuppression. The major clinical syndromes and their treatment are described in Box 23-1. The following sections describe the most common symptoms associated with HIV infection at various stages of infection.
Fever (with or without night sweats) is common in patients infected with HIV, especially in the setting of advanced disease. The absolute CD4 cell count can help to guide the evaluation and differential diagnosis.
Altered Mental Status
Changes in mentation involve cognitive dysfunction, decreased level of consciousness, and delirium and psychosis. Altered mental status can be caused by both infectious and noninfectious disorders. Differential diagnosis, again, is influenced by the known degree of immunosuppression, that is, the stage of HIV disease. Most infections occur in patients with CD4 counts < 200.
The mainstay of evaluation includes history (especially of substance abuse), neurological examination, neuroimaging (computer tomography or MRI), cerebrospinal fluid examination, and thorough review of medications. Altered mental status at any CD4 count can be caused by bacterial meningitis, neurosyphilis, as well as by noninfectious etiologies such as metabolic disorders, seizures, or trauma. Recreational drug use and prescription medication toxicity are other possibilities.
BOX 23-1 Clinical Syndromes, Opportunistic Infections, and Neoplasms Caused by HIV
Headache may pose a challenging diagnostic dilemma. Besides a vast array of opportunistic CNS infections and malignancies, it can result from muscle tension, or a systemic illness without a direct affliction of intracranial structures. Headache can be induced by a medication frequently used for treatment of HIV disease, for example, AZT, or by migraine not related to HIV infection. At any stage of the HIV disease, patients are more susceptible to bacterial meningitis or sinusitis as well as to neurosyphilis. Other opportunistic infections tend to occur at specific levels of immunosuppression.
Swallowing difficulty (dysphagia) is frequently accompanied by pain (odynophagia). Patients complain about the sensation of food sticking. Appropriate diagnosis and treatment are important not only to alleviate the discomfort but also to prevent weight loss and malnutrition. At any stage of HIV disease, patients can develop reflux esophagitis. However, esophageal complaints and disorders are more typical for late to advanced disease.
Diarrhea is a commonly encountered clinical problem in HIV-infected patients and may be caused by HIV itself. Symptoms range from frequent loose stools to fulminant diarrhea, producing profound weight loss, malabsorption, and intravascular fluid depletion. A thorough evaluation is warranted because virtually any gastrointestinal pathogen may be found.
Both small bowel and colon can be affected, causing enteritis or colitis, respectively. Symptoms of enteritis include profuse, watery diarrhea often with symptoms of bloating, nausea, and periumbilical cramping. Volume depletion and malabsorption may be marked. Colitis presents with fever, lower abdominal pain, tenesmus, and passage of frequent, small volume stools with mucus and sometimes blood. Characteristics of symptoms can guide the choice of initial diagnostic procedure but are unreliable to predict the most likely pathogen. CD4 cell count can, again, be useful in outlining differential diagnosis.
At any CD4 cell count, the following agents should be considered: Clostridium difficile; Salmonella, Campylobacter, Shigella, and Cyclospora spp.; Entamoeba histolytica; Giardia lamblia; Isospora belli; enteroviruses; and Strongyloides stercoralis. Most of the antiretroviral drugs also cause diarrhea.
The differential diagnosis of lower respiratory symptoms is extensive and largely depends on the stage of HIV disease with consideration to current PCP prophylaxis, history of travel, or area of residence. Tuberculosis can occur with various clinical signs, depending on the degree of immunosuppression and should be considered in every HIV patient regardless of CD4 count. Appropriate isolation precautions should be undertaken according to clinical suspicion. Irrespective of CD4 counts, typical bacterial causes of pneumonia should also be considered, especially Streptococcus pneumoniae, Haemophilus influenzae, Legionella pneumophila, Mycoplasma pneumoniae, and Chlamydia pneumoniae.
Various dermatologic diseases are extremely common in HIV infection, affecting ~90% of patients. Some of the conditions occur in immunocompetent individuals but are more severe and resistant to therapy in HIV-infected persons. Other skin conditions are unique to HIV disease and tend to follow the degree of immunosuppression. Patients with HIV disease tend also to have a preponderance for drug sensitivity reactions compared with an HIV-negative population.
In the past 2 years, tremendous progress has been made in treatment of HIV infection. New combination treatment regimens can suppress virus replication below the level of detection in many individuals. However, cure is still not possible.
Antiretroviral treatment guidelines are undergoing rapid evolution, but some consensus about general principles has been achieved. First, the major goal of treatment is to suppress viremia and prevent immunosuppression. To reduce the risk that drug-resistant isolates will emerge, therapy should be implemented only when the patient is ready to adhere to a treatment regimen. Treatment is recommended for all patients with symptomatic HIV infection (eg, AIDS, thrush, and fever of unknown origin). For asymptomatic patients, the decision is more complex. Potential benefits of starting therapy in these patients include the following:
However, these potential benefits must be weighted against treatment risks:
Most experts believe that treatment should be offered to asymptomatic patients when immunosuppression has developed (ie, CD4 < 350) or when viral load is high (ie, > 30,000 copies/mL with bDNA assay or 55,000 copies/mL with RT-PCR), especially if the patient desires treatment and is likely to adhere to it. (See Table 23-3)
Monotherapy is no longer recommended for HIV treatment. A combination of three antiviral drugs, to fully suppress plasma HIV below the level of detection, is now recommended. Antiviral drugs are grouped into three categories: (See Table 23-4)
Table 23-4 lists the drugs that were approved for HIV treatment by 1998. Adherence, side effects, and drug-drug interaction should be considered when choosing a regimen. All medications are started at the same time at the full dose (although dose escalations are needed for ritonavir, nevirapine, and ritonavir in combination with saquinavir).
Table 23-1. Laboratory tests commonly used in the diagnosis of infection with HIV-1.
Table 23-2. Detection and measurement of HIV-1.
Table 23-3. Timing of antiretroviral treatment.
Table 23-4. Quick reference guide to antiretrovirals
Table 23-5. Summary of suggested initial treatment of HIV infection.1
Numerous clinical trials confirmed that antiretroviral combination therapy results in greater reductions of viral RNA copies, less frequent emergence of resistance, and a more sustained therapeutic response. Available data and clinical experience support using a combination of two NRTIs along with a PI as the most potent combination available (HAART). However, the choice of drugs has to be individualized based not only on the patient's virologic and immunologic response but also on the patient's tolerance of side effects, ease of use, potential drug-drug interactions, and cost. With growing numbers of new antiretrovirals, the recommendations about the best combination therapy continue to be modified. Alternatives to the two NRTI-one PI combination are indicated in Table 23-5).
For suggestions when initial treatment fails, See Table 23-6. Responses to failure.
Medical care should focus on providing comfort, preventing the opportunistic infections, and managing the new complexities of antiretroviral therapy. In the terminal stage of HIV disease, treatment of the virus itself along with opportunistic infections and especially the malignancies is often palliative and should be instituted after careful consideration of the potential complications of therapy. A thorough discussion of these facts with the patients will facilitate realistic treatment expectations. The use of life-support measures, including cardiopulmonary resuscitation and mechanical ventilation, should also be addressed as early as possible. The survival rate for AIDS patients with respiratory failure has actually improved in the past few years, and intensive care is certainly not contraindicated. However, the expected benefit from aggressive measures must be balanced against the overall quality of life in the face of progressive disease.
Throughout the course of HIV disease, there is an active viral replication and concomitant stimulation and destruction of the immune system, even during the clinically asymptomatic period. The rate of progression of the infection that eventually culminates in complete destruction of the CD4 lymphocytes is determined by a balance between the pathogen virulence factors and the host genetic make-up and immune defenses against HIV.
Prevention & Control
Vaccine against HIV is not yet available, and prevention of infection relies on controlling transmission of the virus. Control encompasses both prevention counseling and postexposure prophylaxis.
Assessment of risk factors as well as specific risk reduction information should be a part of routine primary care and a component of evaluation at the specific HIV testing centers and STD clinics. A supportive, nonjudgmental atmosphere, along with repeated, specific information targeting specific identified risk factors, is most effective.
Table 23-6. Responses to failure of treatment.
Injection drug users should be encouraged to join a drug rehabilitation program and offered evaluation for entry into methadone maintenance programs if available in the community. Short of discontinuation of drug injection, some behavioral modifications can reduce the risk of a potential HIV transmission; for example, using sterile equipment or cleaning drug paraphernalia, especially needles and syringes, with water and then bleach (for at least 30 seconds). Some communities also offer needle/syringe exchange programs. Eliminating needle sharing or reducing the number of sharing partners could further reduce the risk of transmission.
Promotion of safer sexual practices should be an essential part of prevention counseling. Such practices focus on implementation of consistent condom use and the reduction of the number of partners. Screening and treatment of STDs further contribute to reducing the risk of sexual transmission of HIV.
Counseling should also incorporate issues of contraception and a possibility of virus transmission from an infected mother to her baby. All pregnant HIV-infected individuals should receive antiviral treatment to prevent viral transmission to the baby.
Even though only AZT monotherapy was studied and proved to reduce the risk of HIV transmission to the neonate, a single drug antiviral regimen is no longer acceptable for treatment of the mother. Thus combination therapy should be used. Mothers should also be counseled about the potential risk of HIV transmission via breast milk, and alternative infant feeding should be initiated.
Prevention of occupational exposures should involve education and reinforcement of universal precautions and implementation of devices and safe disposal of contaminated materials that would minimize a possibility of exposure to infected body fluids. Postexposure prophylaxis with antiviral medications after an occupational high-risk exposure to contaminated body fluids has become widely accepted, and specific treatment protocols have been implemented by various health-care institutions. In general, postexposure prophylaxis should be initiated as soon as possible after a known or a potential exposure and should consist of a combination either of zidovudine and lamivudine or stavudine and didanosine. With either regimen, consideration should be given to the addition of a protease inhibitor (nelfinavir or indinavir) if exposure is especially risky (large volume/high titer) or if the source patient has a high viral load (> 50,000 copies/mL), has advanced AIDS, or has been previously treated with one or both nucleoside analogues in the recommended two-drug regimens. Triple therapy is the regimen preferred by most health-care providers.
Experience with occupational postexposure prophylaxis has led recently to considerations for expansion of prophylactic treatment of individuals with recent sexual exposure to HIV. The current proposed regimen is the same as for occupational exposure. Postexposure prophylaxis should only constitute a backup for failure of primary prevention measures. Counseling and implementation of risk reduction programs should always be the mainstay of prevention of HIV transmission.
Carpenter CJ et al: Antiretroviral therapy for HIV infection in 1997. JAMA 1997;277:1962.
Centers for Disease Control: Update: trends in AIDS incidence–United States, 1996. MMWR 1997;46:861.
De Roda Husman AM et al: Association between CCR5 genotype and the clinical course of HIV-1 infection. Ann Intern Med 1997;127:882.
Karon JM et al: Prevalence of HIV infection in the United States, 1984–1992. JAMA 1996;276:126.
Saag MS et al: HIV viral load markers in clinical practice. Nature Med 1996;2:625.
Sande MA, Volberding PA: The Medical Management of AIDS, 5th ed. WB Saunders, 1999.
Sande MA et al: The Stanford Guide to HIV/AIDS Therapy. 2001.
Vlahov D et al: Prognostic indicators for AIDS and infectious disease death in HIV-infected injection drug users. JAMA 1998;279:35.
World Health Organization: HIV/AIDS. The global epidemic. Wkly Epidemiol Rec 1997;72:17.