Clinical observations suggest that intact cellular immune response is a major determinant in the outcome of infection with C. neoformans[20, 21]. Other immune mechanisms like antibodies to Cryptococcus antigens , antibody-dependent cell cytotoxicity , complement [24–27], and natural killer cells  also have a role in eliminating organisms. However, the frequency and severity of neurocryptococcosis and disseminated infection caused by C. neoformans in patients with AIDS clearly emphasizes the requirement for a T-cell–mediated immune response to this fungus [28, 29].
An adequate antigen capable of eliciting lymphocyte activation in vitro and available for investigation of C. neoformans–specific T-cell response has not yet been identified; nor has the best approach for interpreting results been defined. The classification of a patient as responder or non-responder to an antigen could be useful to the clinical immunologist for classifying patients as immunocompetent or immunodeficient. We have studied lymphoproliferation to two different antigens: soluble cell wall antigen (SAg) and heat-killed yeast antigen (KAg). Since the capsular polysaccharide itself, a known virulence factor present in the intact C. neoformans, can suppress the lymphocyte proliferation , we prepared antigens with non-encapsulated yeasts. Both tested antigens were immunogenic to PBMC from healthy volunteers, but heat-killed yeast (KAg) was more immunogenic than soluble cell wall antigen (SAg).
Since Mody et al. (1999)  observed a mitogenic effect of cell wall antigen from C. neoformans to PBMC, we investigated whether the observed lymphocyte stimulation could be nonspecific (mitogenic effect). For this purpose, PBMC from cord blood were stimulated with both antigens (data not shown). Neither KAg nor SAg elicited stimulation of the cord blood cells. Therefore, the lymphoproliferation from healthy volunteers could be specific to C. neoformans.
Our results were similar to those observed by Miller and Puck (1984) , demonstrating that normal individuals (volunteers) without continuous exposure to C. neoformans had significant specific lymphoproliferative response to fungal antigen. Since C. neoformans is common in the environment , all individuals, even the healthy ones, could be exposed to the fungus. As reported, the volunteers not occupationally exposed are also responsive to fungal antigens. The environment is the only contaminating source, which exposed them daily (subjects exposed to occupational contact were excluded).
Based on the Miller and Puck study  and according to our findings, Cryptococo exposure occurs in the environment regardless of occupational exposure. In our study we pointed out that newborns do not respond to fungus and that the antigen used in the assays does not provide mitogenic effect. Therefore, subjects with no occupational exposure that respond to the fungi are related to a consequence of environmental exposure, as suggested by Miller and Puck .
Since contaminated feces and pigeon droppings are major sources of natural exposure to C. neoformans[33, 34], we investigated the effect of this continuous exposure to specific lymphocyte response to C. neoformans antigen among poultry growers. They presented an increased lymphocyte response to C. neoformans antigen compared with non-exposed volunteers. It is interesting to note that this greater response could be only observed when data were analyzed after logarithmic transformation of stimulation index (log SI). When the data were calculated as delta cpm (data not shown) or SI, this statistical difference was not observed.
It is interesting to note that lymphoproliferative response to CMA antigen of peripheral blood mononuclear cell from poultry growers was higher than controls, regardless of the method used for data calculation. A possible explanation is that poultry growers are continuously exposed to C. neoformans (more than 10 years), increasing sensitized lymphocytes, which cross-react with C albicans antigen. Pietrella et al. (2002)  showed that immunized mice with mannoprotein from C. neoformans developed delayed-type hypersensitivity (DTH), a Th1 response, against lethal challenge with mannoprotein from C. albicans. As a matter of fact, there was a stronger and a moderate correlation (r = 0.81 and r = 0.7, respectively) of lymphoproliferative response by PBMC stimulated with CMA and KAg from poultry growers and patients with NCC, while PBMC from controls showed a poor correlation. Our data, as demonstrated by Pietrella et al. (2002) , also suggest a cross-reaction between epitopes of C albicans and C. neoformans in the T-cell response. On the other hand, the elevated lymphoproliferative response from poultry growers could be interpreted as a simple consequence of a better capability to respond or even a greater clonal expansion determined by continuously exposure to fungus.
In order to identify nonresponsive individuals or impairment of specific in vitro response to C. neoformans, the assay was also applied to HIV-seronegative patients with NCC. The patients’ specific lymphocyte response to KAg and pokeweed mitogen was not different from that observed with other groups (poultry growers or healthy volunteers), but it was lower for CMA, when compared with poultry growers. Our results were in accordance with those obtained by Levitz and North (1997) , who demonstrated that kinetics and magnitude of cellular immune response to C. neoformans antigen from patients and control group (without continuous exposure) are quite similar.
If we consider, as standardized in our laboratory, the 5th percentile of SI as the normal range cut-off value for classifying individuals as responders or nonresponders to the antigens (KAg and mitogen), 100% of patients could be considered responders, but only 50% were responsive to CMA. On the other hand, if we use the ROC analysis to obtain a cut-off value, the number of responders and nonresponders is quite different, mainly because the cut-off values were much higher than the 5th percentile of S.I. The ROC area, whose value represents the capability of the assay to discriminate subjects into two categories, showed us that the lymphotoxin alpha to C. neoformans antigen (like KAg) is a method that should not be used for diagnostic purposes, since this kind of method should present a ROC area of around 100%, high specificity, and high sensitivity.
Lymphocyte transformation assay stimulated by KAg from patients with NCC was not affected by TCD4+ cell count, and the intensity of response did not correlate with the clinical evolution of neurocryptococcosis. Cured patients without neurological sequelae presented a SI to KAg varying from 0.87 to 71, while patients who developed neurological sequelae or died presented high levels of specific lymphocyte response to KAg (SI, 1.27 - 368).
It might be expected that responses to short, relatively simple antigens such as synthetic peptides might indicate more precisely the difference between responders and nonresponders than the responses to complex antigens, since the potential number of epitopes and of responding cell clones would be much lower . In addition, our data have been taken from a study of an uncommon disease caused by a fungus to which a great proportion of the population has been exposed.
It is important to mention that the diagnosis applied to the NCC patients is only directed to the diagnosis of fungal disease . Although many Cryptococcosis cases are still diagnosed in immunocompetent individuals  the clinical management of patients with NCC is that cryptococcosis is an opportunistic disease. The evaluation of primary immunodeficiency after diagnosis is not indicated, it is not mentioned other immunological techniques (serological or cellular) to a diagnosis of the disease and its association with immunodeficiency.
Our results led us to conclude that specific lymphoproliferative response should be presented as SI, rather than raw cpm or delta cpm. Nevertheless, the comparison between two groups should be done by SI log transformation, which decreases variance. The use of mitogen stimulation reflects, with high sensitivity, nonspecific T-cell response capability, which was positive in all patients with NCC. The evaluated patients may not be classified as non-responders to C. neoformans by lymphoproliferative response assay. Considering all the results, lymphoproliferation assay applied to C. neoformans associated immunodeficiency diagnoses should be considered as a screening assay only. Furthermore, neurocryptococcosis is a disorder which has received large attention recently [39, 40].