Fungal infections have emerged as a world-wide health care problem in recent years [1], owing to the extensive use of broad-spectrum antibiotics [2], long-term use of immunosuppressive agents, increasing use of hyperalimentation and indwelling devices [3] and the increasing population of terminally ill, debilitated and immunocompromised patients [4]. In tune with this general trend, there has been a phenomenal rise in the occurrence of fungal lung infections over the last two decades [5], a significant fraction of which is community-acquired [1]. Specific diagnosis of fungal pneumonia assumes importance in view of the different therapeutic strategies involved and the higher mortality associated with acute invasive fungal infection. Unfortunately, diagnosis of fungal respiratory infection is always difficult, owing to the lack of pathognomonic clinical features, contamination of the non-invasive samples like sputum with normal commensal flora and difficulty in obtaining invasive samples like translaryngeal aspirate and lung biopsy. A previous study has shown that Candida colonization could be found in the respiratory samples obtained by Bronchoalveolar lavage (BAL), endotracheal aspirate or protected specimen brushing in critically ill patients [6]. Since patients with chronic lung pathology provide a suitable nidus for fungal colonization, we hypothesized that screening such patients for fungal colonization of the respiratory tract would enable us to identify individuals requiring closer monitoring for the development of possible complications like acute invasive fungal infection or dissemination via hematogenous spread. The present study was designed with the purpose of characterizing the fungal pathogens associated with particular pulmonary diseases, and correlating between the individual fungal pathogens and the profile of risk factors, radiological presentations and clinical conditions in these patients.

The literature is replete with reports of fungal lung infections occurring in the background of overtly immunocompromising conditions like transplantation, haematological malignancy, neutropenia and HIV/AIDS. Though several authors have examined the incidence of infection with specific fungal agents in the context of chronic lung pathology, studies on pulmonary mycoses, as a group, have been relatively limited. In this study we report the profile of fungal colonization of the respiratory tract in patients presenting with different chronic respiratory conditions in a tertiary care teaching hospital located in the Himalayan region of India. We observe that 46.7% of the patients recruited in our study were culture-positive for fungal agents and 35% demonstrated presence of anti-fungal precipitin antibodies. Fungal colonization was irrespective of the underlying lung pathology and could not be predicted by the presence of any specific co-morbid condition, risk factor or radiological presentation. Moreover, a normal chest X-ray could not be used to exclude the possibility of accompanying pulmonary fungal infection. Interestingly, the identity of the fungi recovered was different between the major diagnostic categories with significant predilection being observed for colonisation with Aspergillus sp. in patients suffering from Bronchogenic carcinoma and for Candida sp. among patients with Tubercular sequelae.

Though it is difficult to make a confirmed distinction between fungal colonisation of the lung and active fungal infection [1], we considered our patients to be colonised since none of our patients satisfied the criteria for "proven", "probable" or "possible" Invasive Fungal Disease, as proposed in the revised definitions of invasive fungal disease from the European Organization for Research and Treatment of Cancer/Invasive Fungal Infections Cooperative Group and the National Institute of Allergy and Infectious Diseases Mycoses Study Group (EORTC/MSG) Consensus Group [8]. None of these patients showed any radiological abnormality which could not be explained by their underlying lung pathology. Further follow-up study on patient progress is required to determine whether such colonization is a prelude to the development of co-morbidity due to invasive or disseminated fungal disease and if so, to work out the time kinetics and nature of such progression. However, none of our patients developed invasive fungal infection till the completion of the study.

Though Candida sp. and Aspergillus sp. constitute the bulk of fungi reported in pulmonary mycoses, their relative proportion and species distribution have shown considerable geographical variation. In the present study, Candida albicans was the most frequent isolate, being recovered from 42.9% of patients, followed by Aspergillus flavus (21.4%), Aspergillus fumigatus (14.3%), Aspergillus niger (10.7%), Candida tropicalis (7.1%) and Cryptococcus neoformans (3.6%). In a similar study on Candida infection in chronic pulmonary conditions, Phukan et al. reported an isolation rate of 50% from sputum specimens, with C. albicans and C. tropicalis predominating [9]. In studies on pulmonary Aspergillosis occurring in chronic lung diseases, Kurhade et al[10] and Shahid et al[11] have reported isolation of the fungus from 16.3% and 14.7% of cases of chronic respiratory diseases, using sputum and BAL samples respectively. The species distribution of Aspergillus showed some difference between these studies and the present study. Aspergillus fumigatus accounted for 80% of the isolates in the former study and Shahid et al[11] reported Aspergillus fumigatus as the predominant species being isolated from 69.2% cases. Our findings are consistent with reports from other centres in India, in which isolation rate of A. flavus largely exceeded that of A. fumigatus, underscoring the importance of geographical differences in fungal recovery [12].

The performance of serological tests in the diagnosis of pulmonary mycoses is dependent on a number of factors including the choices of antigen and assay and the immune status of patients. Discordant result between culture and immunodiffusion assays was observed by us in 18 of the 60 (30%) patients. Among them, immunocompromising conditions were observed in 8 of the 12 patients who were culture-positive but immunodiffusion-negative. In view of the ubiquity of common fungi, serological evidence has been reported to assist in confirming fungal infection and add to the clinical significance of fungal recovery [13]. However, interpretation of serological assays becomes difficult in the presence of immunocompromised clinical status [10, 14] and invasive fungal infections [15]. In our study, 5 of the 6 patients who were immunodiffusion-positive but culture-negative had bronchogenic carcinoma. The sixth patient had pleural effusion of unresolved aetiology, in whom the possibility of bronchogenic carcinoma could not be excluded. Fungal recovery may be difficult in bronchogenic carcicoma because of non-communication of the lesion with main bronchus [11] and might have been improved in our study on including multiple specimens. Similar observations have been reported by Malik et al who found serology to be more sensitive than culture for the detection of Aspergillosis in patients with bronchogenic carcinoma [16]. We observed seropositivity for Aspergillus antigen in 15% of recruited patients, which is in broad agreement with figures reported in similar studies [10, 11, 13, 17, 18].

Owing to the presence of pre-existing respiratory pathology in the recruited patients, it is difficult to establish a causal relationship between the culture-positivity and radiological abnormality. However, 50 of our patients had radiological abnormalities in their chest X-ray, of whom 23 patients were culture-positive for fungal agents. On the other hand, 5 of our culture-positive patients had normal Chest X-ray. Bronchoscopic examnation of all of these 5 patients revealed the presence of central tumors, which could account for the normal appearance on chest X-ray.

In the present study, we report for the first time, serological evidence of Candidiasis in patients with chronic respiratory diseases. Immunodiffusion test with cytoplasmic antigen of Candida albicans was positive in 12 of the 60 patients (20%). Of them, 8 patients were culture-positive for Candida and the remaining 4 patients were culture-negative. Sero-reactivity to the 47 kD cytoplasmic antigen has been shown in previous studies to discriminate between patients with deep-seated invasive Candidiasis and superficial Candidiasis and has been proved to be a valuable adjunct in the diagnosis of invasive candidiasis [19, 20]. In view of these findings, we used the cytoplasmic antigen of Candida albicans in an immunodiffusion format and detected precipitin antibodies in 57.1% of our patients.

The point prevalence of fungal colonization was found to be almost 50% in our patients. Since no association was observed with clinical diagnosis, co-existing risk factors and radiological findings, it is imperative that the possibility of accompanying fungal colonization should be explored meticulously in such patients. An interesting predilection was observed for Aspergillus sp. and Candida sp. to preferentially infect patients with Bronchoenic carcinoma and Tubercular sequelae respectively. Till further evidence is generated to establish the prognostic implications of fungal colonization with certainty, these patients should be closely monitored, in order to enable earlier detection of invasive or disseminated fungal infection.