WASHINGTON, DC—The major problem with empiric antifungal therapy is that many cancer patients not in need of it receive it anyway, while many truly needing it may not be identified and never receive an adequate treatment course.
So said mycoses expert John R. Wingard, MD, Director of the Bone Marrow Transplant Program at Shands Cancer Center of the University of Florida, speaking here at the Interscience Conference on Antimicrobials and Chemotherapy (ICAAC) at a symposium on managing severe infections in cancer patients.
The symposium was developed in cooperation with the European Organization for Research and Treatment of Cancer (EORTC).
Dr. Wingard noted that despite all the promising new anti-fungal drugs on the horizon or already on the market, numerous studies show that invasive fungal infections remain a serious and growing threat in cancer patients.
“Although therapies that suppress cell-mediated immunity increasingly contribute to susceptibility, chemotherapy-induced myelosuppression remains the most common risk factor for invasive fungal infection. In that respect the new drugs haven't made much of a dent,” he noted.
For more than a decade invasive aspergillosis has been the leading cause of death from fungal infections in immunocompromised patients, and Candida remains the fourth most common cause of nosocomial bloodstream infections in North America, with a mortality rate that remains high.
Dr. Wingard reviewed the results of studies testing Aspergillus and Candida species against different formulations of Amphotericin B—the standard drug treatment for more than two decades—or new agents such as itraconazole, caspofungin, and voriconazole.
Results have been quite similar: The success rate—elimination of the fungal pathogen—ranged from about 35% to 50% regardless of which drug was used.
On the other hand, he emphasized, with the new compounds, nephrotoxicity was significantly less of a problem and there were fewer withdrawals from studies.
Generally, Dr. Wingard attributes the difficulties in managing immunocompromised cancer patients to the poor sensitivity and specificity of current diagnostics to reliably detect infections early in their course.
“Poor diagnostics,” he said, “force clinicians to depend on clinical suspicion to initiate antifungal therapy”—a form of Russian roulette, in his opinion.
At the same time, as clinicians should know, the practice of oncology has changed, Dr. Wingard said. “For example, today we routinely use hematopoietic growth factors to speed the recovery of neutrophils. And with bone marrow transplants we use stem cells to optimize neutrophil recovery.”
“As a result, neutrophil recovery is now outpacing the risk for these fungal infections, and certainly outpacing our ability to discern a benefit in antifungal therapy.”
In many cases, he stressed, antifungal therapy is not necessary, especially when the patient receives prophylaxis with fluconazole, an approach that eliminates Candida.
Role of Fever
According to Dr. Wingard, fever is a poor surrogate for judging which patient has an invasive fungal infection and which should rightly receive empirical antifungal therapy.
“Nor should we necessarily regard fever as a criterion of response, since recovery of neutrophils due to growth factors is the real reason people defervesce, not the particular antimicrobial agent in use,” he said.
“Once the patient has defervesced, we should continue antibiotics until the end of neutropenia and at the same time continue fungal surveillance to make sure defervescence is not followed by recurrence of the fever.”
Dr. Wingard reiterated that world studies show that with the empiric use of antifungal therapy, “we're overtreating many patients not in need of antifungals. These drugs can be toxic and expensive, and at the end of the day we're still left wondering who needed them but didn't receive them, or who received too much.”
The next step should be a move toward a targeted approach, to seek more clues to help in the decision about when to initiate antifungal therapy. Examples of important clinical clues, he said, are the signs and symptoms seen in some cancer patients with Candida infections such as skin lesions and liver lesions, or, with Aspergillus, respiratory ailments including sinusitis, and pulmonary symptoms—“and, of course, a positive culture provides strong presumptive evidence of aspergillosis.”
We can also rely on radiology, Dr. Wingard added, noting the example of the “halo sign”—a shadow surrounding a nodule indicating cavitation within a necrotic lesion—that may turn up in some patients with invasive aspergillosis.
“The idea is twofold: to be sure we use these aids carefully to make an accurate diagnosis so our Aspergillus patient will have an adequate course of therapy, and at the same time, to move as fast as possible because the lethal disease progresses rapidly,” Dr. Wingard said. “Despite the apparent contradiction here, we can try not to start treating before we know what's going on.”
“Many patients do not need antifungal drugs. I believe we can get smarter about this and target antifungal therapies to those who need them. For the future, there's clearly a need for new diagnostics that are noninvasive, accurate in distinguishing infected from noninfected patients, and capable of detecting infection early so as to optimize treatment outcomes.”
The European Approach
Claudio Viscoli, MD, from the University of Genoa, Italy, observed that when fever persists in a neutropenic patient and no pathogen is found—i.e., when nothing seems to work—younger or more nervous doctors tend to add or replace antibiotics without adequate reason, while experienced physicians search for a diagnosis and make no changes in therapy.
He described a recent study showing a flexible empiric strategy used by the EORTC: A total of 132 persistently febrile and granulocytopenic cancer patients with fever of unknown origin or a clinically documented infection unresponsive to empirical antibacterial therapy were randomized to either receive empiric Amphotericin B or to continue on antibacterial drugs.
There was no statistically significant difference between the two groups in terms of resolution of fever. Additionally, in a subset analysis, no death due to fungal infection occurred among patients receiving antifungal coverage compared with four deaths in the other group. The number of documented fungal infections was higher in patients not receiving amphotericin B, Dr. Wingard added.
The researchers concluded that no standard “predisposed” approach can be recommended in the persistently febrile and neutropenic patient. In such patients, they said, “every treatment modification should be tailored to the specific clinical situation and to local epidemiologic patterns.”