Factors Affecting the False Negative of Cryptococcal antigen in Patients with Cryptococcosis: Immunology and Radiology

Study objectives The purpose of this article is to investigate the factors associated with false negatives in cryptococcal antigen detection in patients with pulmonary cryptococcosis by studying the effects of immune status and radiology on cryptococcal antigen results. Methods The clinical records, serum cryptococcal antigen result, laboratory findings and computed tomography findings of 61 immunocompetent and 17 immunocompromised patients with a diagnosis based on biopsy confirmation of the presence of pulmonary cryptococcosis were reviewed during the course of the study. Chi-square test was used to analyze categorical variable. The independent t-test was obtained to analyze continuous variable. Logistic regression models was used to analyze the effects of immunity and radiology imaging on cryptococcal antigen results. We studied the differences between these two groups. No difference were found in baseline characteristics, clinical symptoms, and laboratory findings. Logistic regression analysis showed that the significant independent variables related to serum cryptococcal antigen result were the immune status and radiology findings. Our study indicates that immune status and lung imaging findings are important factors influencing cryptococcal antigen detection.


Introduction
Pulmonary cryptococcosis caused by infection with Cryptococcus neoformans C. Cryptococcus complexes is common in patients with immune dysfunction and it has become a emerging disease in patients with good immune function. [1][2][3]Some reports [4,5] revealed that pulmonary cryptococcosis occurs more frequently in immunocompetent patients than in immunocompromised patients.
Compared to its current high morbidity, the diagnosis of pulmonary cryptococcosis is still not fully diagnosed due to the limitations of diagnostic tools. Pulmonary cryptococcosis usually presents with nonspecific symptoms and computed tomography signs. [6][7][8] The methods used to confirm the infection are culture, direct microscopic, histopathology, serology, and molecular detection. [9, 10]Some research shows that serum CrAg is a useful diagnostic tool [11][12][13],and CrAg can be an effective non-invasive means of detecting pulmonary cryptococcosis. Research on serum CrAg was not a rare in the study of cryptococcal diseases. However, false negative serum CrAg is common among patients with Pulmonary cryptococcosis. The purpose of our study was to explore the factors associated with serum CrAg test.

Study subjects
We retrospectively reviewed the hospitalized patients with pulmonary cryptococcosis between

Statistical analysis
Statistical analyses were performed with SPSS software, version 20.0 .Continuous data are presented as the mean ± standard deviation or range. Chi-square test was used to analyze categorical variable.
The independent t-test was obtained to analyze continuous variable. Logistic regression models was used to analyze the effects of immunity and imaging on CrAg results. P < 0.01 was considered statistically significant.

Results
Patient demographic and clinical information is summarized in Sixty-one patients had no comorbidity, and 17 patients were considered to be immunocompromised with at least one predisposing condition, including kidney transplantationuse with immunosuppressive drugs using in 15 patients, Langerhans cell histiocytosis and Acute myeloid leukemia in 1 patient.
None of the patients included in this study had acquired immune deficiency syndrome (AIDS) at the time of diagnosis of pulmonary cryptococcosis. There was no significant difference in WBC, neutrophils, Hb, PLT, and plasma PCT level between two groups. Cough was the most common presenting symptom, occurring in 44 immunocompetent patients (72.13%) and 15 in immunocompromised patients (88.24%),followed by expectoration (55.74% vs 58.82%), chest pain (22.95% vs 23.53%),and fever (13.11% vs 17.65%).Nineteen people were asymptomatic,17 of whom had no immunodeficiency disease.

Serum cryptococcal antigen and human status
The relationship of serum cryptococcal antigen expression and human status were analyzed. For patients with normal immune function and immunodeficiency, the expression of cryptococcal antigen was compared and a significant correlation was found between the two groups. ( P < 0.01).And higher positive expression of cryptococcal antigen was observed in immunocompetent patients when compared with immunocompromised ones (Table 2). Table 3 showed that in the normal immune group, there were 47 diffuse extent lesion in the CrAg+ group and 7 in the CrAg-group (P < 0.01). Which demonstrated diffuse extent lesion was more common in CrAg+ group and limited extent lesion was more frequently observed in CrAg− group among normal immune group. On the other hand in Table 4, there were 4 diffuse extent lesion in the CrAg+ group and 10 in the CrAg-group among immunodeficiency group (P >0.01).Chest radiological findings and biopsy results in patients with pulmonary cryptococcosis were shown in Figure 2.

Radiological findings
Binary logistic regression analysis 5 The results of binary logistic regression analysis are summarized in

Discussion
Many studies [14,15] have shown that serum CrAg diagnosis of pulmonary cryptococcosis has a higher sensitivity and specificity for HIV-positive patients, but lack of studies on HIV-negative patients. Similar to previous reports, our study found that a relatively low proportion of cryptococcosis infected patients were immunocompromised patients. The most frequent symptoms in our patients with pulmonary cryptococcosis included cough and expectoration were also found in our series and in those in previously published reports. [4,[16][17][18][19] Our data on laboratory findings, including WBC, neutrophils, Hb, PLT, and plasma PCT level suggested that there were no significant difference between two groups (Table 1).It may confirm that cryptococcosis could dampen pulmonary neutrophil recruitment and inflammatory cytokine production in immunocompetent hosts, which affects the immunity of a healthy subject. [18,20] Our findings in normal immune patients have expanded the value of serum CrAg in the diagnosis of extensive disease in patients with normal immune function, supporting the results of previous studies that the most common CT findings of pulmonary cryptococcosis was multiple pulmonary nodules/masses of different diameters. [21][22][23][24] We consider that the results of our study suggest two important factors with clinical usefulness.
Firstly, serum CrAg was valuable for the diagnosis of extensive extent lesion in immunocompetent patients with pulmonary cryptococcosis. Secondly, cryptococcus infection is also common in normal immune populations; thus, Patents with pulmonary CT lesions should undergo CrAg examination.
On the other hand, our study had several limitations. First, it is a retrospective study with a small number of patients collected over a long period and our results should not be generalized to the worldwide population. Secondly, our study lacks the classification of cryptococcal varieties. Moreover, criteria for immunocompetency were based on the patient's medical records and thus may be prone to selection bias. Therefore, further study is wanted to confirm the findings reported here.
In conclusion, we have demonstrated that the immune status and lung CT findings of patients can      Figure 1 Flow diagram of the study patient selection.

Figure 2
The chest radiological findings and biopsy results in patients with pulmonary cryptococcosis.