Improvements in outcome following treatment for acute leukemia have derived from various sources: the introduction of new chemotherapeutic agents; the development of effective drug combinations; the use of multistage approaches to induction consolidation, and maintenance therapy to optimize durable control; and advances in supportive care to reduce treatment-related mortality. Certainly, infectious morbidity and mortality have plagued effective antileukemic therapy for many years. As a result, control of infectious complications has been an important area of clinical study. Through the concerted efforts of many clinical investigators, progress has occurred, and, today, infectious mortality is an infrequent complication of antileukemic therapy.
The article by Drs. Sarkodee-Adoo, Merz, and Karp provides a concise overview of developments in the management of infections in patients with acute leukemia. The authors emphasize preventive and therapeutic strategies against bacterial and fungal infections. They also touch briefly on the topic of adjunctive measures in the management of neutropenic infections, such as the use of growth factors and granulocyte transfusions. In their last paragraph, they highlight promising new directions to fortify the mucosal defense barriers and control inflammatory cytokines.
With so much accomplished, does any work remain? As Sarkodee-Adoo et al point out, the fundamental principles of infection management were set down more than a decade ago. Although much is the same today (as the authors note), much has also changed. More options are currently available, due primarily to the introduction of programmable infusion pumps; the development of new broad-spectrum antibiotics with both oral and intravenous (IV) formulations, as well as antibiotics with long half-lives, permitting once-daily dosing; the widespread availability of home care companies; and the availability of hematopoietic growth factors to speed recovery from neutropenia. With societal imperatives to control health care costs and with patients wanting greater consideration of quality-of-life concerns (eg, shorter hospital stays), there has been a shift in supportive care efforts from the inpatient to the outpatient setting.
Risk-Stratified Antibacterial Therapy
Regarding antibacterial strategy, the authors ask the question: Does one size fit all? Clearly, most of our principles for managing neutropenic fever have approached the problem in a unitary manner. However, insights from the study of risk factors in neutropenic patients treated for solid tumors have caused us to rethink this approach.
Aside from the classic notions that duration and depth of neutropenia are the major risk factors, it is clear that other factors are associated with morbidity and mortality from neutropenic infections. These include the presence of comorbidity, whether or not the fever occurred in the inpatient or out-patient setting, and the degree of control of the underlying neoplastic disease.[1,2]
From such observations, new strategies to manage neutropenic fever have evolved. They typically include one of two basic approaches. In the first, the patient receives IV antibiotics in the hospital and is reassessed 24 to 72 hours later; the patients physiologic status, response to antibiotics, and the results of blood cultures and other diagnostic tests are used to determine whether continued inpatient management is needed or whether outpatient antibiotics can be substituted. The second approach involves consideration of outpatient antibiotic therapy at the outsetif the patient falls into a low-risk category.
Although such approaches have been successfully used to select treatment strategies in neutropenic solid tumor patients,[3-8] these studies, unfortunately, did not address acute leukemia. The degree of mucositis has been identified as a distinct risk factor for infection in leukemia patients,[9,10] one which, undoubtedly, would be an important discriminator between low risk and high risk.
Recently, two groups of investigators evaluated oral vs IV antimicrobial regimens in neutropenic cancer patients, including those with leukemia.[11,12] Unfortunately, the leukemia patients were not analyzed separately from the solid tumor patients; presumably, however, such patients fared as well. It should also be noted that the oral therapies were administered in the inpatient setting; in an editorial, Finberg and Talcott called it premature to switch to the outpatient setting without validation in prospective trials.
These studies raise the possibility that we may be able to identify a subgroup of patients with low-risk acute leukemia, in whom outpatient oral or IV therapies may be possible. Clearly, further studies are warranted to test new strategies and identify low- and high-risk subgroups.
Risk-Stratified Antifungal Therapy
The value of antifungal prophylaxis in patients with acute leukemia has been debated. Fluconazole(Drug information on fluconazole) (Diflucan) appears to be quite efficacious in reducing the rate of mucosal candidiasis. However, a large, randomized, placebo-controlled trial showed that this agent produced only a marginal reduction in the risk of hematogenous candidiasis. The frequency of systemic fungal infections in patients with acute leukemia ranges from 3% to 42% and appears to be lower than the frequency observed in bone marrow transplant recipients. Most studies conducted to date generally either have lacked sufficient statistical power to demonstrate a benefit or have evaluated a very heterogeneous patient population, in which patients with acute leukemia and bone marrow transplant recipients were analyzed together. Interestingly, a recent randomized trial of oral fluconazole in patients with febrile neutropenia showed a marked reduction in the incidence of superficial fungal infections, invasive fungal infections, and fungal mortality in a subset of patients with leukemia.
One concern is that the widespread use of fluconazole prophylaxis in all patients with acute myelogenous leukemia (AML) treated with intensive chemotherapy may lead to the emergence of resistant organisms, such as Candida krusei and Candida glabrata. Perhaps, targeted use of antifungal therapy in patients at high risk is most rational.
Two such strategies have been proposed. Fungal colonization is a known risk factor for invasive infection and usually precedes such infection. Thus, one proposal calls for fungal surveillance cultures of stool specimens to be obtained in order to identify patients at risk for fungemia; this subgroup of patients would then receive targeted prophylaxis.
The severity of mucosal damage produced by different antileukemic regimens has been shown to vary widely, and these differences account for var-iations in invasive fungal infection rates.[9,10] The second proposal is to stratify patients according to the mucositis potential of their treatment regimen: Groups receiving regimens that result in minimal mucositis would not receive prophylaxis and would be spared exposure to antifungals, while prophylaxis would be reserved for those given regimens that produce considerable mucositis.
Itraconazole(Drug information on itraconazole) (Sporanox), with its wider spectrum of action against the fluconazole-resistant Candida species, as well as filamentous fungi, has not fulfilled expectations, and a randomized trial of the cyclodextran oral solution of itraconazole (Sporanox) in patients with acute leukemia failed to demonstrate an impact on the incidence of mold fungal infections. An important limitation of itraconazole is its erratic bioavailability. In a recent, large, randomized trial, empiric therapy with IV itraconazole followed by an oral solution in neutropenic patients with hematologic malignancies was at least as efficacious and less toxic than ampho-tericin B (Fungizone). In this setting, IV itraconazole followed by an oral solution offers an alternative to amphotericin B(Drug information on amphotericin b).
Several scenarios bear mention: Even today, treatment-related mortality in the elderly with acute leukemia is substantial, posing a special challenge for the clinician. Certainly, this is a group in whom new strategies are needed and an ideal population in whom to test novel approaches.
Current therapies for AML typically consist of an intensive cytotoxic therapy, followed by bone marrow failure and mucosal barrier damage, which persists for 2 to 3 weeks and gradually recovers. That is quite a different scenario from the management of acute lymphoblastic leukemia (ALL), in which prolonged courses of cytotoxic therapies are administered. Although mucosal barrier damage is considerably less in treated ALL patients, neutropenia can persist for a substantially longer period. In these patients, the risks for infection are clearly different. Traditional criteria would call for months-long hospitalizations, and while some clinicians may prefer hospitalizing patients for the whole duration of induction chemotherapy, others feel comfortable following patients a on an outpatient basis after consolidationtherapy.
Patients with ALL are frequently given prophylactic fluoroquinolones and/or fluconazole, although this practice has not been evaluated in controlled trials. Prophylaxis with trimethoprim(Drug information on trimethoprim)-sulfamethoxazole for the duration of ALL treatment has also become a widely accepted standard of care. More studies are needed to optimize antimicrobial strategies in patients with ALL.
High priorities for future study include the identification of risk factors and the stratification of patients into low- and high-risk groups. An improved understanding of infectious risks will likely lead to refinements in antimicrobial strategies.
Prophylaxis would be inappropriate for low-risk patients, exposing them to needless toxicity and increased antibiotic cost. This practice might also lead to the emergence of antibiotic-resistant bacteriaa potential hazard to both the low-risk patients and others.
In contrast, prophylaxis or early therapy in subgroups of patients at high risk would reduce infectious morbidity and mortality. Moreover, this might actually reduce the risk of antimicrobial drug resistance, for two reasons: (1) the overall burden of infectious organisms would be decreased; and (2) the statistical likelihood that mutations conferring drug resistance would develop in a smaller population of organisms would be lower. Such efforts will hopefully allow us to tailor therapy to each patient according to need, thereby further reducing treatment-related mortality.