Background There is ample information on when to initiate highly active antiretroviral therapy (HAART) but there is little guidance on counseling patients about when to discontinue HAART as the end of life nears either from HIV or more commonly now a separate comorbid illness (see Fast Fact #213 and Fast Fact #214 on prognostication in HIV/AIDS). Since more patients now are dying with rather than from HIV/AIDS – decisions regarding the role of HAART at the end of life become more complex. A simple curative vs. palliative care paradigm is inadequate; the decision involves a nuanced discussion of HAART’s potential impact on quality of life within the context of the patient’s prognosis and goals of therapy. Simply put, as patients with HIV/AIDS get very close to death (e.g., last 2 weeks of life), the potential benefits of HAART diminish relative to treatment burden. The following is a review of HAART benefits/burdens near the end-of- life.

Benefits

1. Prevention of a retroviral syndrome. In one cohort study, over a third of patients who discontinued HAART experienced symptoms related to discontinuation (1). However, the onset of symptoms is often 1-2 months after discontinuation of HAART, and most patients do not experience symptoms, so this potential benefit should be balanced by considerations of treatment burden and life expectancy.
2. Treatment of syndromes or symptoms affected by viral load or viral ‘fitness.’ The severity of HIV-related neuropathy, and constitutional symptoms such as fatigue or weight loss, correlate directly to viral load (2). HAART may relieve or prevent symptoms by reducing the viral load or by changing the virus’s fitness.
3. Prevention of opportunistic infections. HAART reduces the incidence of opportunistic infections that impact the quality of life of patients with AIDS. In one prospective cohort study of patients with advanced HIV disease, there was a 66% reduction in the number of new non-CMV opportunistic events with HAART (3). HAART’s primary mechanism is decreasing viral load; however, HAART can also directly inhibit opportunistic organisms – e.g., inhibition of candida virulence factors by protease inhibitors (4). Again, the potential for benefit will depend on the extent of the patient’s disease and existing co-morbidities.
4. Delay the progression of AIDS Dementia Complex (ADC). Since ADC seems to be triggered by HIV replication, HAART is a key element in slowing progression. While the optimal antiretroviral therapy is not known, zidovudine has been shown in placebo-controlled trials to be of benefit (5,6).
5. Hope for a positive treatment effect. While it is difficult to objectively demonstrate improved outcomes based on psychological effects of continuing HAART, some clinicians and many patients find it difficult to not equate discontinuation of HAART with “giving up.” Or alternatively, patients may express that they do not want the dying process from a comorbid illness shortened by a seemingly preventable death from HIV. Finding the most appropriate means to achieve treatment goals often requires exploration of fundamental notions such as how the patient defines “hope” and how treatment goals will be served by a change in therapy.

Burdens

1. Cost. The high cost of these medications is a significant strain to a hospice system’s budget (typical regimen costs $250-500/ week). Since the role of HAART near the end of life is uncertain, hospice agencies may decide to decline payment. In some instances, patients may be admitted into hospice care for an alternate terminal diagnosis such as liver failure and remain eligible for medication coverage by their primary insurer, separate from their hospice benefit. A few patients can afford HAART meds out-of-pocket. Frequently, patients are forced to choose between hospice care and continuing HAART.
2. Side effects. Though HAART has improved over time and is associated with less side effects it still is not fully benign. GI intolerance, headache, and fatigue are common side effects. Gout, pancreatitis, myopathy; peripheral neuropathy, stomatitis, arthralgias; depression; and/or anxiety are less common side effects. When in doubt, consult with a clinical pharmacist or HIV specialist.
3. Pill burden. In a study of male veterans with a variety of chronic medical problems, perception of over-medication was associated with decreased quality of life (7). While many HAART regimens utilize considerably less number of pills per day, some patients still may find adherence with HAART challenging and burdensome.

Summary The decision to discontinue antiretroviral therapy for hospice patients with AIDS should be a part of comprehensive end-of-life care; how HAART serves the patient’s overall goals for care should be the primary consideration. Facilitating an informed discussion of benefits and burdens of HAART is an important means to help patients achieve their goals. In general, as the benefit to burden ratio for HAART has improved over time, HAART is often continued unless the patient is in the final weeks of life or the patient is experiencing individualized burdens.

References

1. Ahdieh GL, Silverberg MJ, Palacio H, et al.  Discontinuation of potent antiretroviral therapy:  predictive value of and impact on CD4 cell counts and HIV RNA levels.  AIDS. 2001;15(16):2101-2108.
2. Simpson D, Haidich AB, Schifitto G, et al.  Severity of HIV-associated neuropathy is associated with plasma HIV-1 RNA levels.  AIDS. 2002; 16:407-412.
3. Murphy EL, Collier AC, Kalish LA, et al.  Highly active antiretroviral therapy decreases mortality and morbidity in patients with advanced HIV disease.  Ann Inter Med.  2001; 135(1):17-26.
4. Munro CA, Hube B.  Anti-fungal therapy at the HAART of viral therapy.  Trends Microbiol.  2002; 10(4):173-177.
5. Clifford, DB.  AIDS dementia.  Med Clin N Am.  2002; 86:537-550.
6. Sidtis JJ, Gatsonis C, Price RW, et al.  Zidovudine treatment of the AIDS dementia complex:  results of a placebo controlled trial.  Ann Neurol.  1993; 33:343-349.
7. Fincke BG, Miller DR, Spiro A.  The interaction of patient perception of overmedication with drug compliance and side effects.  J Gen Inter Med.  1998; 13(3):182-185.

Version History:  This Fast Fact was originally edited by David E Weissman MD and published in November 2003. Re-copy-edited in April 2009; then again in June 2015; 4th Edition published in September 2024.
Conflicts of Interest: None to report.