Based on the detailed data extracted from 9 studies including 248 immunocompromised and 276 immunocompetent PC patients, our systematic review and meta-analysis provide a comprehensive description of clinical manifestations and CT findings in immunocompromised and immunocompetent PC patients.
The clinical manifestations in patients with PC are generally nonspecific, and patients can even be asymptomatic. Among the immunocompromised patients, asymptomatic infection occurred in up to 30.2%. Immunocompetent patients appeared to have a higher rate of asymptomatic infection (up to 40.8%). However, there was no significant difference in the proportion of asymptomatic cases between immunocompetent and immunocompromised groups. Some investigations have reported asymptomatic PC in more than 50% of patients[18, 19]. The reason for this difference might be differences in patient screening processes based on comorbidities or immune status. PC usually elicits several nonspecific respiratory symptoms, such as cough, expectoration, chest pain, and dyspnea. The incidence of these nonspecific respiratory symptoms was not significantly different between the immunocompetent and immunosuppressed PC patients. Nonspecific respiratory symptoms are likely to cause delays in PC diagnosis and subsequent proper treatments. Systemic syndromes with an overall low incidence, such as fever and headache, seemed to occur more frequently in immunosuppressed patients. The reason for more frequent systemic symptoms in immunocompromised patients might be due to the deficiency in immune surveillance in immunocompromised patients, resulting in failure to elicit a cryptococcus immune response. This immune surveillance deficiency results in more pulmonary exudative and necrotizing pathological changes, as well as the intrapulmonary or even systemic spread of cryptococcus[20, 21]. In immunocompetent patients, cryptococcal infection tends to be localized due to phagocytosis by macrophages and granulomatosis formation. Therefore, immunocompetent patients tend to have mild pulmonary dissemination and fewer systemic symptoms.
According to thin-section chest CT images, nearly half of the immunocompromised PC patients had bilateral lung involvement. Even though less than one-third of the immunocompetent PC patients expressed bilateral lung involvement, the difference between the immunocompetent and immunocompromised patients was nonsignificant (p = 0.052). However, several retrospective studies revealed that immunocompromised PC patients are more likely to show bilateral lung lesions on chest CT than immunocompetent patients[11, 13]. Our meta-analysis failed to reveal a positive association, possibly because only 6 studies directly extracted data on unilateral or bilateral lung involvement, and the relatively small amount of data may cause bias. More than half of the total patient population showed a peripheral distribution of lesions. Our analysis indicated that PC tended to involve the lower lobes in both immunocompetent or immunocompromised patients, which is consistent with several retrospective studies. Although upper lung involvement was not predominant in PC, it was relatively more common in immunocompromised patients.
The common radiological features of PC, such as air bronchogram signs, halo signs, cavity, pleural effusion, ground-glass attenuation, consolidations, enlarged mediastinal lymph nodes, nodules and masses mimic other pulmonary infectious diseases and even malignant tumors. This imaging finding similarity is one of the reasons for the delay in PC diagnosis. Single or multiple nodules were the most commonly observed chest CT findings in PC patients according to our analysis and several other reports[13, 23, 24]. Among these imaging findings detected by chest CT, cavitation, enlarged mediastinal lymph nodes, and ground-glass attenuation were more common in immunocompromised patients. Pleural effusion was rare in PC patients, especially in immunocompetent patients, with an incidence of only approximately 5%. Single or multiple nodules were the most common imaging finding in both immunocompromised and immunocompetent PC patients. Similar to other pulmonary infectious diseases that induce the formation of granulomatous nodules during the disease process, the presence and architecture of granulomas, which present as “nodules” in chest CT imaging, are likely related to cryptococcus infection and intact host immune status. Consistent with other studies, our analysis concluded cavities sign within nodules, masses or other lesions occurred significantly more frequently in immunocompromised than in immunocompetent PC patients. The difference may be caused by the inability to mount an effective immune response to localize cryptococcal infection in immunocompromised patients. The proliferating microorganisms destroy the adjacent lung tissue and promote the formation of cavities[20, 25]. Ground-glass attenuation was observed to be more common in immunocompromised patients. Several investigations of pulmonary fungal infections found that pulmonary exudation might be consistent with pulmonary hemorrhage caused by fungal infection. These pulmonary exudative lesions contain pathogenic microorganisms[25–27]. This imaging difference is also evidence of the inability to localize pulmonary cryptococcal infection in immunocompromised individuals. Mediastinal lymph node enlargement, which was more frequently encountered in immunocompromised individuals, is probably due to mediastinal lymphadenitis caused by the lymphatic spread of microorganisms in immunocompromised patients.
In conclusion, based on the limited available data, the immunocompromised PC group had a higher proportion of older adults (≥ 60 years) than the immunocompetent PC group. Among the nonspecific respiratory syndromes, we were unable to identify any symptoms that were significantly different between immunocompromised and immunocompetent PC patients. Nevertheless, several systemic symptoms, such as fever and headache, were more common in immunocompromised individuals. According to thin-section CT findings, lesions tended to be peripherally distributed in the lower lobes. However, upper lobe involvement was observed more frequently in immunocompromised patients. Similar as immunocompetent PC individuals, solitary or multiple nodules were the most common appearances in immunocompromised patients. Air bronchogram signs, halo signs, consolidations, and masses are possible radiologic features in immunocompetent and immunocompromised PC patients. Cavitation, ground-glass attenuation, and enlarged mediastinal lymph nodes are common radiologic features in immunocompromised PC patients but not immunocompetent individuals.
We acknowledge several limitations of our study. (1) A limited number of studies had available data related to the comparison of clinical and imaging features in immunocompromised and immunocompetent PC patients. Thus, for our analysis, we extracted only 9 suitable studies. (2) In terms of the “immunocompromised” definition, the majority of studies defined “immunocompromised” status based on the presence of comorbidities or concomitant medications that caused immunosuppression. A few studies considered a combined evaluation of immune cell counts and immunoglobulin levels. Nevertheless, our study results will help clinicians identify the potential cryptococcal pneumonia and recognize the differences in clinical manifestations and radiographic findings in immunocompetent and immunocompromised PC patients.