Abstract
Cryptococcosis, an invasive mycosis caused by Cryptococcus spp, kills between 20% and 70% of the patients who develop it. There are no vaccines for prevention, and treatment is based on a limited number of antifungals. Studying fungal virulence and how the host responds to infection could lead to new therapies, improving outcomes for patients. The biggest challenge, however, is that experimental cryptococcosis models do not completely recapitulate human disease, while human experiments are limited due to ethical reasons. To overcome this challenge, one of the approaches used by researchers and clinicians is to: 1) collect cryptococcal clinical isolates and associated patient data; 2) study the set of isolates in the laboratory (virulence and host-pathogen interaction variables, molecular markers); 3) correlate the laboratory and patient data to understand the roles fungal attributes play in the human disease. Here we review studies that have shed light on the cryptococcosis pathophysiology using these approaches, with a special focus on human disease. Isolates that more effectively evade macrophage responses, that secrete more laccase, melanize faster and have larger capsules in the cerebrospinal fluid are associated with poorer patient outcomes. Additionally, molecular studies have also shown that cryptococcal clades vary in virulence, with clinical impact. Limitations of those studies include the use of a small number of isolates or retrospectively collected clinical data. The fact that they resulted in very important information is a reflection of the impact this strategy has in understanding cryptococcosis and calls for international collaboration that could boost our knowledge.