Aboulafia provides an extensive review of the occurrence of and treatments for bone marrow disorders that complicate HIV infection and AIDS. Understanding of the pathogenesis of these disorders is increasing, and the availability of recombinant colony-stimulating factors (CSFs) has, in many ways, facilitated the treatment of HIV-1 infection and its complications. Nonetheless, numerous critical questions remain regarding the optimal use of these expensive and powerful reagents.
When Should CSF Therapy Be Inititated?
In patients with neutropenia, when should CSF administration be initiated? Recent data from a controlled trial of granulocyte CSF (G-CSF, filgrastim(Drug information on filgrastim) [Neupogen] in HIV-1-infected persons with absolute neutrophil counts (ANCs) from 750 to 1,000 cells/mL demonstrated a decreased incidence of bacterial infections among persons assigned to G-CSF. In this trial, all persons were treated with G-CSF if their ANC fell below 500 cells/mL. These results notwithstanding, many clincians would be loathe to treat all patients with an ANC less than 1,000 cells/mLwith G-CSF.
Better predictors of the risk for bacterial infection are needed. In this regard, the results of an AIDS Clinical Trial Group trial (ACTG 815) indicate, not surprisingly, that both macrolide and trimethoprim(Drug information on trimethoprim)-sulfamethoxazole prophylaxis decreases the risk of serious bacterial infections and that lower performance status is associated with a higher risk for such infections. Surprisingly, persons under 40 years of age and women tended to be at greater risk. Perhaps for now, routine administration of G-CSF should be offered to persons with an ANC less than 500 cells/mL (a point below which the risk for serious bacterial infections increases dramatically) and to those with an ANC less than 1,000 cells/mL and a history of recurrent bacterial infections.
Other Critical Questions
In patients who receive G-CSF therapy, if both the degree and duration of neutropenia predict the risk of infection, how high must the ANC climb to prevent morbidity and for how long must the ANC remain in the safe range? Could strategies that promote weekly and brief neutrophil increases provide clinical benefit with fewer injections and lower cost?
Tantalizing data suggest that granulocyte-macrophage CSF (GM-CSF, sargramostim(Drug information on sargramostim) [Leukine, Prokine]) administration may increase the antifungal and antimycobacterial activities of mononuclear phagocytes,[4,5] and a brief anecdotal report suggests that administration of GM-CSF may help clear mucosal infection with Candida species in persons with AIDS.
Can GM-CSF enhancement of monocyte function compensate, in part, for the profound T-cell deficiency associated with advanced AIDS? Can monocyte activation decrease the risk of opportunistic infections in this population, and what is the effect of this strategy on the pro-inflammatory cytokine network and on HIV-1 RNA?
Although most patients experience a brisk neutrophil response to administration of GM-CSF or G-CSF, will patients who fail to respond to G-CSF respond to GM-CSF, and vice versa?
Anemia, Transfusion, and Prognosis in AIDS Patients
As Dr. Aboulafia reminds us, the anemia of AIDS may have many etiologies, although drug- and infection-related causes are most common. Interestingly, in several older studies, anemia appeared to be an independent predictor of mortality in persons with AIDS with or without Mycobacterium avium intracellulare infection.[7-9] Whether anemia was an indirect marker of severity of disease or was related to other predictors of poor outcome, such as plasma HIV-1 RNA levels, is not known. Yet, transfused patients with HIV-related anemia also tended to have a particularly poor prognosis.
A greater transfusion requirement also may be reflective of more advanced disease. However, in vitro data suggest that allogeneic leukocytes contained within packed red blood cell units used for transfusion may activate HIV expression through a mixed lymphocyte reaction, and a preliminary report indicates that transfusion of packed red blood cells to persons with HIV-1 infection is associated with small but significant rises in plasma HIV-1 RNA levels.
Whether leukocyte reduction strategies will affect HIV-1 RNA levels or long-term outcome among transfused HIV-1 infected patients may be answered by the results of the National Heart, Lung and Blood Institutes Viral Activation by Transfusion Study (VATS) now ongoing in 11 US medical centers. Also needed and targeted by this study is an examination of the potential mechanisms whereby anemia and transfusion as a treatment for anemia may affect long-term outcome and risk for opportunistic infections in patients with advanced HIV-1 infection.
For now, limiting transfusions to HIV-infected persons with symptomatic anemia is prudent. Administration of recombinant human erythropoietin(Drug information on erythropoietin) (rHuEPO [Epogen, Procrit]) may decrease transfusion requirements without affecting plasma HIV levels. However, it is not clear that a target hematocrit level in the mid-30s is either necessary or justifiable. Rather, the hematocrit should be maintained at a level that is compatible with patient comfort.