Bacteriology of Bronchoalveolar Lavage Samples in Lower Respiratory Tract Infections

Introduction

Bronchoalveolar lavage (BAL) is a low-risk invasive procedure in which sterile normal saline is instilled into small segments of the lung and subsequently collected using a suction apparatus for analysis.¹ Lower respiratory tract infections (LRTI’s) account for approximately 20 to 24 percent of all deaths attributed to acute respiratory illnesses.² The symptoms range from mild to severe and may include weakness, fever, dyspnea, chest pain, and cough.³ While pneumonia manifests in the elderly, LRTIs are more prevalent in early life.⁴

A variety of lower respiratory tract infections, like acute or chronic bronchitis, community-acquired pneumonia (CAP), and pneumonia in immunocompromised individuals, are among the most commonly encountered infectious diseases. Therefore, the role of the microbiologist in laboratory-based diagnosis of these conditions is critical.⁵

Bronchoalveolar lavage is a valuable diagnostic sample, despite being collected through a low-risk invasive method. It is used to diagnose respiratory infections caused by bacteria, including Mycobacterium species, viral and fungal lower respiratory tract infections, and various non-infectious diseases/conditions. The sensitivity and specificity of BAL samples surpasses those of various non-intrusive methods in the diagnosis of pulmonary infections. As with any invasive procedure, BAL requires strict adherence to standard precautions.⁶

Materials & Methods

This prospective observational study was conducted over a one-year period at a 1250-bed tertiary care teaching hospital. The study was carried out in the Department of Microbiology following ethical clearance (FMIEC/254/2023). BAL samples were collected from immunocompetent adult patients aged 18 years and above, with suspected LRTIs and no underlying comorbidities. Patients were recruited from both the wards and ICUs of the hospital. Individuals below 18 years of age and those diagnosed with viral, fungal, or mycobacterial infections were excluded. A total of 75 BAL samples were analyzed. Sample size requirement was calculated using the formula, n = z2p(1-p) / e2, Z = 1.96, at 95% confidence interval, P = 98%7, e = 2.25% (allowable error) and n = 75.⁷

The BAL samples collected were macroscopically examined in the laboratory for appearance, colour, consistency and presence of any debris. A good quality smear was prepared from the purulent portion of the sample using the Gram staining technique on a clean, grease-free glass slide. The semi-quantitative loop method was used to culture the BAL samples.7 A 3 mm external diameter loop, carrying 2 μL of the BAL sample, was used to inoculate MacConkey agar, blood agar and chocolate agar plates. These inoculated plates were incubated at 37°C for 24 hours. Samples exhibiting ≥ 20 colonies were interpreted to have a colony count of ≥ 10,000 CFU/mL and were considered to show significant growth.⁸ The growth characteristics of the isolates were observed and the colonies were identified using standard biochemical tests. Antibiotic susceptibility testing was performed using appropriate antibiotics based on the isolated organism. In cases where no growth was observed on any of the culture plates after 24 hours of incubation, the sample was reported as ‘No growth’.

Statistical analysis was conducted using frequency and percentage methods.

Results

Of the 75 BAL specimens tested for bacterial infection, 33 (44%) showed significant microbial growth. No respiratory bacterial pathogens were isolated in 40 (53.3%) samples. Two samples (2.6%) showed no growth after 60 hours of incubation. Among 33 culture positive samples, 17 (51.51%) were from men and 16 (48.48%) were from women.

The bacterial isolates were obtained from patients across various age groups. Of the 33 isolates, 3 (9.09%) were from patients aged 18-30 years, 4 (12.12%) from those aged 31-40 years, 7 (21.21%) from the 41-50 year age group, 8 (24.24%) from patients aged 51-60 years, and 11 (33.33%) from those aged ≥61 years.

Among the 33 BAL samples that showed significant bacterial growth, all isolates were Gram-negative bacteria. The distribution of isolates was as follows: Klebsiella pneumoniae (11 isolates, 33.33%), Pseudomonas aeruginosa (7 isolates, 21.21%), Escherichia coli (3 isolates, 9.09%), Haemophilus influenzae (2 isolates, 6.06%), Stenotrophomonas maltophilia (2 isolates, 6.06%), and one isolate each (3.03%) of Burkholderia pseudomallei, Pseudomonas stutzeri and Enterobacter cloacae. Five isolates (15.15%) were identified as Acinetobacter baumannii. All cases were clinically correlated with radiological findings and inflammatory marker data, confirming significant isolates rather than colonizers.

Antibiotic susceptibility and resistance patterns of the isolated organisms 

A panel of 10 antibiotic discs were tested for all the isolates using Kirby-Bauer disk diffusion method (Table 1), according to the Clinical Laboratories Standard Institute (CLSI) guidelines.⁸ As majority of isolates were Gram negative bacilli, Amoxycillin / Clavulanate (10 µg), Cefazolin (30 µg), Gentamicin (10 µg), Trimethoprim/ sulfamethoxazole (1.25/23.75 µg), Amikacin (30 µg), Ciprofloxacin (5 µg), Piperacillin/ Tazobactam (100/10 µg), Imipenem (10 µg), Meropenem (10 µg), Tigecycline (15 µg) were tested. Few isolates were resistant to one or more antibiotics. The results of the sensitivity test indicate that isolates demonstrated highest sensitivity to Gentamicin, Amikacin, Ciprofloxacin, Piperacillin / Tazobactam, Imipenem, Meropenem and Tigecycline, whereas higher resistance patterns were observed for antibiotics like Amoxycillin / Clavulanate and Cefazolin. Tigecycline was 100% effective against all the isolates.

K. pneumoniae showed the highest susceptibility to Piperacillin/Tazobactam (90.90%), Imipenem (90.90%), Meropenem (90.90%) and Tigecycline (90.90%). The highest resistance was observed to Cefazolin (36.36%) and Trimethoprim/Sulfamethoxazole (36.36%). P. aeruginosa and P. stutzeri were found to be 100% susceptible to all antibiotics tested.

E. coli demonstrated 100% susceptibility to Amoxicillin/ Clavulanate, Trimethoprim/Sulfamethoxazole, Gentamicin, Amikacin, Piperacillin/Tazobactam, Imipenem, Meropenem, and Tigecycline. Highest resistance was observed against Cefazolin (66.66%). The A. baumannii complex showed maximum susceptibility to Tigecycline (100%), while exhibiting maximum resistance to Amoxycillin / Clavulanate (100%) and Cefazolin (100%).

H. influenzae was highly susceptible to Trimethoprim / Sulfamethoxazole (100%), Imipenem (100%) and Meropenem (100%). It showed highest resistance to Amoxicillin/ Clavulanate (50%). E. cloacae complex isolate was susceptible to all the second and third-line antibiotics and was resistant to first line antibiotics. S. maltophilia showed highest susceptibility to Trimethoprim/ Sulfamethoxazole (100%) and Ciprofloxacin (100%).

One isolate of B. pseudomallei was resistant to Ciprofloxacin and Levofloxacin but sensitive to Carbapenems. Among the 33 bacteria isolated, 4 (12.12%) were identified as multidrug-resistant (MDR), while the other 29 (87.87%) were non-MDR. The A. baumannii complex was the most common MDR isolate, accounting for 3 cases (9.09%) of the total isolates, followed by K. pneumoniae, which contributed to one MDR isolate (3.03%).

Discussion

The examination of BAL cytology, cell counts, and culture provides valuable information regarding lung infections, inflammatory conditions, immune systems, and malignant diseases. Other bronchoscopic techniques such as endobronchial or transbronchial biopsies, bronchial washings, bronchial brushings, and transbronchial needle aspiration are frequently performed in conjunction with BAL.⁹ Quantitative culture techniques have been introduced to address challenges in interpreting culture results due to potential contamination. This study evaluated the sensitivity and specificity of different thresholds of colony forming units (CFU) per milliliter of BAL fluid for the diagnosis of pneumonia.¹⁰ The bacteria commonly associated with LRTIs are: P.aeruginosa, K.pneumoniae, S.aureus, Acinetobacter sp., S.pneumonia, E.coli, Enterobacter sp., Haemophilus influenzae, Moraxella sp., Citrobacter sp., and Nocardia sp.⁶

The quality check of a BAL sample is conducted using Gram’s stain. All samples with greater than 1% salivary epithelial cells when observed under low power objective, with absence or presence of bacteria in less than 1-2 / oil immersion field suggest oropharyngeal contamination and such a samples were not processed further.⁸ Several other low-risk invasive sampling methods are available for diagnosing LRTIs in both ventilated and non-ventilated patients. These include the guarded mini-BAL procedure performed blindly, or non-bronchoscopically (non-B-mini-P-BAL), protected alveolar lavage (PAL) (a variation of BAL), and protected specimen brushing (PSB). Each of these sampling methods has a different threshold for bacterial growth quantification, as reported by various authors, typically ranging from 10³ to 10⁵ CFU/mL, based on the procedure used.⁹ CFU counts can be diluted by factors of 10 to 100.⁹

Bronchoscopy is one of the most effective and widely used methods for diagnosing disorders of the respiratory tract. Bronchoalveolar lavage (BAL) fluid can be collected during this procedure which can aid in diagnosis. BAL fluid provides greater diagnostic accuracy and sensitivity compared to other samples. Additionally, bronchoscopy is often necessary for older patients. By lowering the likelihood of contamination, the BAL technique enhances the effectiveness of pathogen detection.⁷ Lower respiratory tract infections are known to cause death and morbidity worldwide. The etiological agents and patterns of susceptibility to these infections differ by location. Furthermore, a conclusive diagnosis may not be achievable based solely on clinical observations.¹⁰

In this study, of the 33 bacteria isolated from BAL samples, 17 (51.51%) were from male patients and 16 (48.48%) were from female patients. These findings suggest that males are slightly more commonly affected by LRTIs than females. This increased susceptibility in males may be attributed to greater occupational and environmental exposure. Additionally, factors such as smoking, alcohol consumption, and tobacco use contribute to reduced local immunity in the respiratory tract by causing mucous plugging, airway collapse, respiratory muscle fatigue, and impaired mucociliary clearance.¹¹

The age group of 61 years and above showed the highest bacterial growth, accounting for 11 cases (33.33%), compared to the 18-30 years age group, which accounted for only 3 cases (9.09%). It indicates a significant association between age and bacterial growth. Similar observations were made in a study conducted by Adhikari et al.⁷ All the 33 bacterial isolates were Gram-negative organisms. The absence of Gram-positive bacteria may be attributed to the exclusion of out-patients in this study. Among the Gram-negative bacteria isolated, K. pneumoniae was the most frequently isolated organism (11 cases; 33.33%), followed by P. aeruginosa (7 cases, 21.21%). Comparable results have also been reported by Kumar et al., Padmaja et al., Patel et al., and Thananki et al. ^3,12,10,13

However, various studies reported conflicting results. Kim et al., reported Methicillin resistant S. aureus as the most common bacteria, followed by A. baumannii complex.¹⁴ In a study conducted by Khan et al., S. pneumoniae was the predominant Gram-positive bacterium, while P. aeruginosa was the most common Gram-negative isolate. In a study by Ahmadinejad et al., the commonly reported pathogen was A. baumannii, followed by P. aeruginosa. ^15,16

In the present study, the majority of subjects had normal bronchoscopy findings. This is not unexpected, as bronchoscopy is typically reserved for cases where other techniques have failed to yield a diagnosis or when there is clinical suspicion of conditions that can be directly visualized through bronchoscopy.³

Antibiotic sensitivity testing in our study demonstrated highest sensitivity of the isolates to Gentamicin, Amikacin, Ciprofloxacin, Piperacillin/Tazobactam, Imipenem, Meropenem and Tigecycline, while higher resistance was observed for antibiotics like Amoxycillin/ Clavulanate and Cefazolin. This observation aligns with the findings reported by Regha et al.¹⁷

Among the isolates, four were identified as multidrug resistant (MDR), of which three belonged to the A. baumannii complex and one was K. pneumoniae. These findings are consistent with those reported by Thomas et al., and Santella et al., who observed that A. baumannii isolates demonstrated the highest sensitivity to Colistin, with a resistance of less than 2%. One limitation of the present study was its exclusive focus on BAL samples.^18,19 The impact of the findings could be enhanced by comparing BAL samples with other respiratory samples.

Conclusion

BAL sampling is a highly sensitive method for isolating pathogens responsible for LRTIs. In our study, the most frequently isolated organism from BAL specimens was Klebsiella pneumoniae (11 isolates, 33.33%), followed by Pseudomonas aeruginosa (7 isolates, 21.21%), and the Acinetobacter baumannii complex (5 isolates, 15.15%). The findings of the present study highlight the importance of bronchoscopy as a key diagnostic tool for LRTIs. Non-fermenting Gram-negative bacteria remain major pathogens in respiratory infections. The emergence of multidrug-resistant organisms in BAL samples is a cause of concern, underscoring the crucial role of healthcare authorities and infection control measures in curtailing their spread. 

Conflict of interest

Nil

Source of support

Nil