Management of Cryptococcosis: How Are We Doing?

Perfect discusses the implications of a national prospective study (the "CryptoA/D" study) of the factors influencing clinical presentation and outcome of patients with cryptococcosis.


Perspectives
February 2007 | Volume 4 | Issue 2 | e47 I n a new research article published in PLoS Medicine, Francoise Dromer and colleagues [1] report on a French national prospective study (the "CryptoA/D" study) of the factors infl uencing clinical presentation and outcome of patients with cryptococcosis (an infection caused by the pathogenic fungus Cryptococcus neoformans). This is just the type of prospective study that is needed to help us understand how we are doing in the management of a life-threatening infection. Often our randomized, blinded, and controlled studies do not refl ect actual outcomes in the general medical community, because study participants are biased by a series of entry criteria and other issues. Thus, prospective studies to collect real-time data during clinical management become important gauges in how successful we are and may help us defi ne what we need to improve.

The New Study
Dromer and colleagues present a multicenter surveillance of an entire country's experience of cryptococcosis management over a four-year period. Their study of 230 patients came to three major conclusions.
First, from the initial assessment, higher severity of illness and poorer outcome occurred in: (1) men; (2) HIV-positive patients; and (3) patients infected with serotype A strains. These three factors capture three potential issues of: genetic susceptibility and/or hormonal infl uences, host immunity, and strain virulence or microbial fi tness in cryptococcosis. Many years ago, the importance of male sex and outcome of cryptococcosis was observed in controlled animal experiments [2]. In cryptococcosis, host immunity infl uences the production of disease and disease outcome, and cellmediated immunity is always the focal point in this infection. The host immune system, which can be altered by HIV infection or highdose corticosteroids, controls the appearance and severity of cryptococcal infection ( Figure 1). Finally, the particular invading cryptococcal strain may have some impact on disease. In animals cryptococcal strains vary in their ability to produce disease with the same inoculum, and in the Dromer et al. human study serotype A strains were more likely than serotype D strains to produce severe disease. Further validation of the importance of specifi c strains on disease is illustrated by the large outbreak of Cryptococcus gattii on Vancouver Island, Canada, in immunocompetent hosts with a probable recombinant hypervirulent strain [3].
The second fi nding of this study is the direct mycological impact of treatment. With the use of an endpoint at two weeks of mycological failure (defi ned as at least one cultured sample from the patient containing viable C. neoformans), the researchers found that failure was associated with initial dissemination of infection, high serum antigen titers, and the lack  of fl ucytosine use during induction therapy. It is clear that patients with a high burden of organisms are at risk for failure, and initial cultures of several body sites and antigen titers can help identify these high-risk patients. This study also continues to validate the strategy of using the combination therapy of amphotericin B plus fl ucytosine [4]. The benefi ts of adding fl ucytosine are related to its faster "sterilization" of tissue and fl uids [5], and the combination approach is associated with fewer relapses compared with monotherapy [6].
Third, this study identifi ed the three factors that contributed to reduced three-month survival: (1) abnormal neurology, (2) abnormal brain imaging at baseline, and (3) underlying hematological malignancy. When there is more disease produced by infection, the outcome is worse and our vigilance in care must be even greater. On the other hand, for many of these invasive fungal infections the fi nal arbiter of success will be the underlying disease and unfortunately, this issue may be diffi cult to control. For example, several decades ago Kaplan et al. noted that there was a poor prognosis for the combination of cryptococcosis and malignancy [7]; this fact has not changed.

Implications of the Study
This study allows us to observe where we are in the management of cryptococcosis in the era of HAART (highly active antiretroviral therapy) and with access to current antifungal drugs and supportive care. Unfortunately, the failure rates remain substantial, with a mortality of about 12% at three months. Furthermore, there are undefi ned costs of care and morbidity from this infection. The question is: How do we do better?

Next Steps
The answer is we must perform careful evidence-based studies to address a series of important clinical questions on patients at high risk for failure (see Box 1). We can then adjust our treatment guidelines to better fi t the individual patient. This evidencebased directed research strategy will be good for the future of medicine but does not necessarily help the patient today. Therefore, the basic message of Dromer et al. rings as true as the message from the prognostic studies of cryptococcal meningitis with amphotericin B treatment in 1974 [8]. In the initial management of cryptococcosis, the assessment of the burden of yeasts in the host from its site of infection(s) to its quantity of yeasts determined through cultures, antigen loads, and radiographic appearances will give clinicians a prediction of the diffi culties that they might face. Furthermore, the initial philosophy in the high-risk patient is to provide therapies that effi ciently eliminate yeasts from the host. The sugar-coated killer whose sweetness sickens must be stripped away leaving only the underlying disease to deal with [9].
In their study, Dromer and colleagues have provided us with a careful gauge of how we are doing with this life-threatening infection.
In some respects, clinicians should be congratulated for their skills in caring for this deadly infection. On the other hand, there still remains a need for better strategies and treatments for cryptococcosis in all patients who suffer from the disease. Specifi c guidelines for management of cryptococcosis are helpful [10], and studies regarding risk factors provide insight, but until we have further evidence-based understanding, cryptococcosis treatment in some patients can have bedside nuances requiring judgments and adjustments one patient at a time.