Clinical Experience With Pulmonary Nocardiosis In A Tertiary Care Hospital In Pakistan

Background Pulmonary nocardiosis is a rare disease. It usually affects immunocompromised patients. In this study we evaluated our clinical experience with pulmonary nocardiosis in our centre. Method It was a retrospective study done in Aga Khan University Hospital. Study was started after ethical approval. Cases from 2007 to 2017 were retrieved. Patient with diagnosis of Culture positive Pulmonary Nocardiosis with age more than 18 years included. Patient who were already on treatment of Nocardiosis at time of admission were excluded. Patients’ medical records were reviewed to identify epidemiologic, clinical, microbiologic, and radiographic features. Data was entered on SPSS V22. 56 patient met inclusion criteria but 12 had missing data so 44 patients’ data was included at end. 8 patients had disseminated disease. 52% were immunocompromised and 48% were immunocompetent. 31 were females. Diabetes mellitus was most common comorbidity. Fever and shortness of breath were most common presentations. Pleural effusion and consolidation were most common chest xray findings. Mean white cell count was 12.35±7 ×109/L. 70% patient had some degree of hyponatremia. Culture showed that all strains were sensitive to amikacin and 67% were sensitive to imipenum. Mean length of stay was 7.7days. 30 day mortality was 23%. 16% required ICU stay. There was significant difference between immunocompromized and immunocompetent group regarding 30 day mortality (p value 0.019) and ICU admission need (p value 0.021).

3 immunocompromised patients and mortality is often very high.

Background
Nocardia is one of the pathogenic bacteria that is gram positive, weakly acid fast, and aerobic. 1 This bacteria causes nocardiosis in humans and animals. Nocardia was first reported by Edmond Nocard in France in 1888 for Bovine farcy disease. 2 It contains total of 85 species and 25 of them are pathogenic. Nocardia Asteroides is most studied and pathogenic in humans. 50% infections are by Nocardia asteroides complex. In humans, mode of transmission can be Inhalational, via Skin cuts and Hospital acquired. 3 It usually cause disease in immunocompromised hosts but one third of the cases can be seen in immunocompetent patients also. 4 Pulmonary nocardiosis affects those people whose cellular immunity is depressed like patients with malignancies, human immunodeficiency virus infection, solid-organ, or hematopoietic stem cell transplantation, and those on longterm treatment with steroids. 5 The diagnosis of Pulmonary nocardiosis is often missed as there are no clear pathognomonic sign and symptoms. Presentation vary widely from subclinical disease to acute presentation with sepsis and in some cases chronic symptoms mimicking tuberculosis. 6 Mortality in immunocompromised hosts is quite high and patient end up needing intensive care support. 7 Incidence of pulmonary nocardiosis is 0.3 per 100000. 7 There is very limited studies from Pakistan regarding Pulmonary nocardiosis. We have collected data on presentation of disease, its risk factor and outcome in Pakistan.

Methods
This study was conducted in Aga Khan University Hospital in Pakistan. Retrospective data was collected from 2007 to 2017. ICD-9 code was used to identify all the cases with diagnosis of pulmonary nocardiosis during these 10 years. Patient with diagnosis of 4 culture positive Pulmonary Nocardiosis with age more than 18 years were included.
Patient who were already on treatment of Nocardiosis at time of admission were excluded.

Microbiological Data
Microbiological data has shown that 54%(24patient) had Nocardia asteroid species on culture while 46%(20 patients) has non asteroid species. Our lab only reported results as asteroides and non asteroides species. All strains were sensitive to amikacin while imipenum sensitive strains were 67%. Least sensitive antibiotic was ciprofloxacin (12%).
Details antibiotics sensitivities pattern is shown in Figure 1.

Outcome
Mean length of hospital stay was 7.7 ± 6 days. Sixteen percent (7 patients Pulmonary nocardiosis is regarded as to affect mostly with affected cellular immunity. In our study 48% of patient was immunocompetent and rest was immunocompromised. Most common cause of immune compromised state was long term steroid use 48%. Previous studies has shown immunocompromised hosts around 50-65%. 11 Diabetes mellitus was one of the common comorbidity that was found in our cohort. Poor control of diabetes can be one of the risk factor for having this disease as it was reported in previous cases reports. 12,13 None of our patient had bone marrow transplant (BMT). The risk of nocardiosis in BMT patient has decreased with the introduction of cyclosporine and trimethoprim/sulfamethoxazole (TMP-SMX) prophylaxis. In one centre, its rate was just 0.2%. 14 The clinical recognition of nocardiosis is difficult as it has relatively low incidence and a lack of pathognomonic symptoms. A high index of clinical suspicion needs to be exercised to diagnose pulmonary nocardiosis. There has been some differences in presentation among immunocompromised and immunocompetent hosts. Like pleural effusion was most common presentation in immunocompromised hosts while immunocompetent hosts have consolidation as most common xray finding. In previous case series, common finding in immunocompetent hosts were nodules and opacities. 15 Disease in immunocompromised host is slightly more severe than in immunocompetent hosts as they were found to be having more ICU admissions, more prolonged hospital stay and increased 30 days mortality. Microbiological data showed that there was mixed pattern of resistance. All strains were sensitive to amikacin while 1/3 rd was resistance to trimethoprim and imipenum. Therefore, it is always important to get final sensitivities when treating these infections. Most active agents in our study were linezolid and amikacin. A study in Spain also looked into resistance pattern. Almost same resistance pattern was seen in their study. 16 Limitation of this study are its being retrospective and we don't have PCR data to further classify Nocardia subspecies to see their differences. Regarding resistance pattern, we did not collect Minimum inhibitory concentration data (MIC) from our lab.

Ethics approval
Ethical approval was taken from hospital ethics committee of Aga Khan University hospital.

Consent for publication
Not applicable

Availability of data
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request    Figure 1 Nocardia antibiotics susceptibilities (Percentage)