A vaccine would certainly be ideal for preventing infection by HIV and thus for avoiding AIDS — the late stage of HIV infection, when immunity is severely impaired. Yet the near-term prospects for a vaccine are poor, and people who contract the virus need care. For the immediate future, then, many scientists are concentrating on improving therapy.

Until only a few years ago, HIV infection was everyone's worst nightmare — it was almost invariably a progressive, lethal disease that completely robbed its victims of dignity. Most medical interventions focused on treating pneumonias and other serious "opportunistic" infections that stemmed from immune failure, not on controlling HIV itself.

Since late 1995, however, several related advances have led to a profound shift in the prospects for most patients who receive treatment today. Notably, scientists have gained a much fuller understanding of how HIV behaves in the body and a better sense of how to shackle it. Two classes of potent drugs have joined the anti-HIV arsenal, and tests that can directly monitor viral levels have been introduced, enabling physicians to assess a therapy's effectiveness rapidly. Together these advances have made it possible to treat HIV infection aggressively and to improve health and survival — at least in the industrial nations, where the intensive therapy is now widely available.

Initially, the signs of a sea change were anecdotal: extraordinary but increasingly frequent tales of people who were plucked from the brink of death to resume vigorous, productive lives. More recently, statistics have borne out the anecdotes. Between the first half of 1996 and the first half of 1997, deaths from AIDS in the U.S. declined by 44 percent. In roughly the same period the frequency of HIV-related hospitalizations and major HIV complications dropped markedly as well.

Nevertheless, both scientific understanding and treatment remain far from perfect. For instance, researchers do not know whether the impressive responses to therapy can be sustained. Treatment is burdensome and costly, which puts it out of reach of certain patients. In addition, some fraction of people who receive the best available care respond poorly. For such reasons, the search continues for ways to make therapy more universally effective and accessible.

The ultimate goal, of course, is a cure. Investigators are unsure whether that aim is feasible. But many of us are cautiously optimistic that we are, at last, beginning to accumulate the weaponry needed to manage HIV as a bearable, chronic disorder, somewhat akin to diabetes or hypertension.

Recommendations for optimal therapy seek to halt viral replication indefinitely — something inconceivable just three years ago. To meet this target, patients must usually take three or even four carefully selected drugs twice or more a day, exactly as prescribed. These general guidelines, and more specific recommendations, derive from current knowledge of HIV's activity in untreated patients.

How HIV Harms

The virus spreads from one person to another usually through sexual intercourse, direct exposure to contaminated blood, or transmission from a mother to her fetus or suckling infant. In the body, HIV invades certain cells of the immune system — including CD4 T lymphocytes — replicates inside them and spreads to other cells. (These lymphocytes, named for the display of a molecule called CD4 on their surface, are central players in immunity.)

At the start of an infection, hefty viral replication and the killing of CD4 T cells are made manifest both by high levels of HIV in the blood and by a dramatic drop in CD4 T cell concentrations from the normal level of at least 800 cells per cubic millimeter of blood. About three weeks into this acute phase, many people display symptoms reminiscent of mononucleosis, such as fever, enlarged lymph nodes, rash, muscle aches and headaches. These maladies resolve within another one to three weeks, as the immune system starts to gain some control over the virus. That is, the CD4 T cell population responds in ways that spur other immune cells — CD8, or cytotoxic, T lymphocytes — to increase their killing of infected, virus-producing cells. The body also produces antibody molecules in an effort to contain the virus; they bind to free HIV particles (outside cells) and assist in their removal.

Despite all this activity, the immune system rarely, if ever, fully eliminates the virus. By about six months, the rate of viral replication reaches a lower, but relatively steady, state that is reflected in the maintenance of viral levels at a kind of "set point." This set point varies greatly from patient to patient and dictates the subsequent rate of disease progression; on average, 8 to 10 years pass before a major HIV-related complication develops. In this prolonged, chronic stage, patients feel good and show few, if any, symptoms.

Their apparent good health continues because CD4 T cell levels remain high enough to preserve defensive responses to other pathogens. But over time, CD4 T cell concentrations gradually fall. When the level drops below 200 cells per cubic millimeter of blood, people are said to have AIDS.

As levels dip under 100, the balance of power shifts away from the immune system. HIV levels skyrocket, and microbes that the immune system would normally control begin to proliferate extensively, giving rise to the potentially deadly opportunistic infections that are the hallmarks of AIDS. Once such diseases appear, AIDS frequently becomes lethal within a year or two. (Opportunistic infections sometimes occur before the CD4 T cell level falls under 200; in that case, the appearance of the infections leads to a diagnosis of AIDS, regardless of the CD4 T cell level.)

Although patients typically survive HIV infection for 10 or 11 years, the course can vary enormously. Some die within a year after contracting the virus, whereas an estimated 4 to 7 percent maintain fully normal CD4 T cell counts for eight years or more and survive beyond 20 years.

At the cellular level, scientists also know how HIV invades and destroys CD4 T lymphocytes. The virus gains access to the interior of these cells (and certain other cell types) by binding to CD4 itself and to another molecule, a "co-receptor," on the cell surface. Such binding enables HIV to fuse with the cell membrane and to release its contents into the cytoplasm. Those contents include various enzymes and two strands of RNA that each carry the entire HIV genome: the genetic blueprint for making new HIV particles.

One of the enzymes, reverse transcriptase, copies the HIV RNA into double-strand DNA (a property that qualifies HIV as a "retrovirus"). Then a second enzyme, integrase, helps to splice the HIV DNA permanently into a chromosome in the host cell. When a T cell that harbors this integrated DNA (or provirus) becomes activated against HIV or other microbes, the cell replicates and also unwittingly begins to produce new copies of both the viral genome and viral proteins. Now another HIV enzyme — a protease — cuts the new protein molecules into forms that are packaged with the virus's RNA genome in new viral particles. These particles bud from the cell and infect other cells. If enough particles form, they can overwhelm and kill the cell that produced them.