Culture positivity of smear negative pulmonary and extrapulmonary tuberculosis- A study from North Kerala, India.

INTRODUCTION: The identification of smear negative pulmonary and extrapulmonary tuberculosis continues to remain a diagnostic challenge. This study was conducted in a tertiary care setup in north Kerala to isolate and identify mycobacteria by culture from radiologically and clinically suspected cases of smear negative pulmonary and extrapulmonary tuberculosis.
METHODS: A total of 200 samples (100 pulmonary and 100 extrapulmonary) were processed and cultured by automated (MB/BacT) and conventional methods. Heat stable catalase test, nitrate reduction test and detection of MPT 64 antigen were done to aid species identification.
RESULTS: Overall culture positivity was 7% (14 isolates - 8 pulmonary and 6 extrapulmonary) of which 92.9% (13) of the isolates were Mycobacterium tuberculosis and 7.1% (1) was Mycobacterium fortuitum (identified by molecular typing). Detection rate by automated method was 7% (14) and by conventional method was only 1.5% (3).
CONCLUSION: Despite its shortcomings and low positivity, culture still remains the gold standard for the diagnosis of EPTB and SNPT. However, automated liquid cultures have better isolation rates than the conventional LJ culture. Subjecting these isolates to rapid diagnostic tests like antigen detection and LPA can aid in the early institution of appropriate treatment regimen. Copyright:

Introduction

Tuberculosis (TB) is a major contributor of ill-health worldwide and has continued to remain a global health problem since its discovery. In 2014, the actual number of clinically diagnosed pulmonary TB and extrapulmonary tuberculosis (EPTB) cases notified in India alone comes to an astounding figure of 730600 among 1609547 total notified new and relapse TB cases.[1] Therefore, a significant 45% of the TB epidemic in India is contributed by smear negative pulmonary tuberculosis (SNPT) and EPTB. Most of the EPTB cases are smear negative and constitutes about 15-20% of all cases of TB in immunocompetent patients and in Indian studies, EPTB constitutes 45-56% of all the cases of tuberculosis in persons with AIDS.[234]

Though smear-negative TB is less infectious due to its paucibacillary nature, a delay in the diagnosis and treatment of smear-negative, culture-positive TB cases could contribute to increase in disease transmission and morbidity as established by various studies.[567] Furthermore, the dual HIV/TB epidemic has complicated the clinical case scenarios of SNPT and EPTB.[8]

Direct smear microscopy which forms the cornerstone of TB control programs in middle and low income countries has a poor track record in EPTB, paediatric TB and in patients co-infected with HIV and TB, due to reduced bacillary load in these patients.[910]

Though smear negative TB is primarily diagnosed by thorough history and physical examination along with radiological and laboratory evidences suggestive of TB, the misdiagnosis rates have been estimated as high as 35-52% without a standardized diagnostic protocol.[11] Thus, the definitive diagnosis of smear negative TB still depends upon culture isolation and identification of mycobacteria.

Smear negative TB is a relatively neglected area of study compared to smear positive TB. This study, to the best of our knowledge, is the first among published articles to deal with isolation and identification of mycobacteria from SNPT and EPTB in the state of Kerala.

Methods

A total of 200 samples (100 pulmonary and 100 extrapulmonary), which were smear negative for acid fast bacilli (AFB) from patients with clinical and radiological features highly suggestive of pulmonary TB and EPTB, were included in this study. Samples from patients who were already on anti-TB drugs were excluded. Other relevant co-morbid conditions, blood investigations (total count, ESR, RBS and HIV testing) and radiological investigations were recorded where necessary. Patients with positive culture were followed up for clinical response to anti-TB drugs.

Specimen collection

Pulmonary samples and extrapulmonary samples were collected in leak proof sterile containers and under aseptic precautions, wherever necessary. The samples were processed as soon as possible. In case of delay, they were refrigerated at 4°C for not more than 24 hours and then processed accordingly.

Specimen processing

All specimen processing and inoculation was done in Biosafety cabinet class II A2. Direct smears prepared from the specimens were stained by Ziehl-Neelsen (ZN) technique. Specimens from sterile sites were inoculated directly without decontamination. Urine samples were centrifuged at 3000 g for 15 minutes and the deposits were used for further processing. Tissue biopsy specimens were ground using sterile mortar and pestle with minimum amount of sterile distilled water before decontamination. Specimens were decontaminated and concentrated using Modified Petroff's method.[12] In this study, double the volume of 4% sodium hydroxide was used for pulmonary samples, whereas, equal volume of 2% sodium hydroxide was used for decontaminating extrapulmonary samples.

Culture

Each processed sample was inoculated onto two Lowenstein – Jensen medium (LJ) slopes as well as automated MB/BacT liquid culture bottles and incubated at 37 °C. The LJ medium was checked daily for 7 days and weekly thereafter for 12 weeks. LJ media with contamination i.e. growth other than mycobacteria, were discarded after confirmation by ZN and gram staining.

Samples from MB/BacT bottles flagged positive were confirmed for growth of AFB by ZN staining and gram staining (to exclude contamination). Subculture was done on LJ medium from positive bottles. In this study, MB/BacT bottles were declared negative for AFB only after incubation for an extended period of 60 days, as the samples collected were smear negative.

Identification of the isolates

Preliminary identification of the isolates was done from LJ medium. Heat stable catalase test was done to differentiation between Mycobacterium tuberculosis complex (MTC) and nontuberculous mycobacteria.[12] Nitrate reduction test was done to differentiation between Mycobacterium tuberculosis and Mycobacterium bovis.

The Detection of MPT64 antigen for MTC was done by immunochromatographic method (SD BIOLINE TB Ag MPT64 Rapid) as per the manufacturer's instructions. The nontuberculous mycobacterium was subjected to molecular typing for identification at Rajiv Gandhi Centre for Biotechnology, Thiruvananthapuram.

Drug susceptibility testing

It was done by proportion method on drug incorporated LJ medium.[12] The concentration of the drugs used were 0.2 mg/L, 40 mg/L, 4 mg/L, and 2 mg/L for INH, RMP, EMB, and SM, respectively.

The isolates were taken to Intermediate Reference Laboratory, Thiruvananthapuram for performing LPA (GenoType MTBDRplus, Hain Lifescience, Nehren, Germany) to confirm sensitivity pattern to INH and RMP.

Results

A total of 200 smear negative pulmonary and extrapulmonary samples (100 each) received during the study period of 1 year. Pulmonary specimens included 88 sputum samples and 12 bronchial washings. Pleural fluid, urine and tissue biopsy samples constituted the major share of extrapulmonary specimens [Figure 1].

Figure 1

Description and number of the 100 extrapulmonary samples processed

Culture and identification of isolates

Overall culture positivity was 7% in smear negative tuberculosis (14 isolates). Culture positivity in SNPT was 8% and in EPTB was 6%. Of the 8 culture positive pulmonary samples, 7 were sputum and 1 was bronchial washings. The 6 culture positive extrapulmonary samples are shown in Table 1.

Table 1

Culture positive specimens *M. fortuitum was isolated from urine sample

Specimen	Number of isolates
Sputum	7
Bronchial washings	1
Skin biopsy	1
Rectal biopsy	1
Pleural fluid	1
Pus from pelvic abscess	1
Pus from lumbar abscess	1
Urine	1
Total isolates	14

Of the 14 isolates, one was positive for heat stable catalase and negative for MPT 64 antigen which suggested atypical mycobacterium. It was identified to be Mycobacterium fortuitum after subjecting to RT-PCR and DNA sequencing in a reference laboratory. The rest of the 13 isolates (92.9%) were positive for MPT 64 antigen indicating infection by MTC species. Nitrate reduction test was positive for all 13, which excluded infection by Mycobacterium bovis.

Solid vs automated liquid culture

Isolation rate by automated MB/BacT liquid culture was 7% and by conventional method on LJ media was only 1.5%. The isolation of AFB by automated culture and LJ media from various specimens have been given in Table 2. The mean time for detection of AFB was 34.4 days by automated method as compared to 56.3 days by LJ culture.

Table 2

Isolation by conventional LJ culture and automated MB/BacT liquid culture from various samples

Specimen type	Total number proceeded	Isolation by MB/BacT	Isolation by LJ media
Sputum	88	7	3
Bronchial washings	12	1	-
Bone marrow	6	-	-
Tissue biopsy	22	2	-
Ascitic fluid	2	-	-
Pericardial fluid	4	-	-
Lymph node aspirate	4	-	-
Pus	6	2	-
Pleural fluid	30	1	-
Urine	25	1	-
CSF	1	-	-
Total	200	14 (7%)	3 (1.5%)

Of the 200 specimens, contamination in both automated and LJ medium was seen in 5 samples. Additional 7 samples grew contaminants on LJ medium. The contamination rate by automated method was 7% (14 samples) and by solid LJ culture was 10.5% (21 samples).

Drug sensitivity testing

All the 13 M. tuberculosis isolates were sensitive to INH, RMP, EMB and SM. The LPA also showed concordant results with all 13 samples showing sensitive pattern to INH and RMP.

Epidemiological profile

64% of the study group comprised of males with a male: female ratio of 1.78:1. Majority (72%) of the patients were in the age range of 21-60 years [Figure 2]. HIV co-infection was present in 17 patients with SNPT and 1 patient with EPTB. Of the 18 samples from HIV infected patients (17 sputum samples and 1 bone marrow), the isolation rate was 16.7% (3 samples) and all 3 were isolated from sputum.

Figure 2

Age and Gender distribution of the study population

At the same time, out of the 100 suspected SNPT, 97 had Chest X-ray (CXR) findings consistent with TB and 3 had non-specific findings.

Discussion

In the present study, the culture positivity of SNPT was 8%. As mentioned earlier, this subgroup of TB patients can act as a source of disease transmission. It is a well-known fact that among the MTC species, M. tuberculosis is the most common cause of human infection worldwide. In the Indian setup, M. bovis has been implicated in human infection too.[1314] Apart from these two, there have been no reports of cases from India due to other species of the M. tuberculosis complex, except for a single case report from New Zealand of transmission of M. orygis from a woman of Indian origin to a cow.[15] Hence, with the exclusion of M. bovis by nitrate reduction test in all the 13 isolates positive for MPT 64 TB antigen, we can conclude with high possibility that all the 13 isolates are M. tuberculosis species. Immunochromatographic assay kit for MPT 64 TB antigen is a simple, reliable, rapid identification kit with 97-100% sensitivity and specificity, as demonstrated by various studies, which can markedly reduce the turn-around time in MTC identification and aid in the proper management of atypical mycobacterial infections.[161718]

Out of the 6 EPTB isolates in this study, one urine sample isolate, negative for MTC by immunochromatographic test, was identified as Mycobacterium fortuitum after molecular typing from Rajiv Gandhi Centre for Biotechnology, Thiruvananthapuram. Urinary tract infection can be rarely caused by M. fortuitum and a case of nephritis mimicking renal tuberculosis have also been reported.[1920] M. fortuitum is a common saprophyte found in environmental and nosocomial sources.

Solid vs automated liquid culture comparison

Various studies have already demonstrated earlier time of detection, better isolation rates and lower contamination rates with automated liquid culture as compared to conventional culture for SNPT and EPTB similar to the present study.[21222324] A multicentric study from India demonstrated that automated liquid culture system had significantly higher sensitivity than solid LJ culture in pulmonary TB patients, independent of their HIV status. It also demonstrated significant reduction in the time to detect mycobacteria compared to LJ culture, thus speeding up the final diagnosis of both pulmonary and EPTB.[25]

All the 13 M. tuberculosis isolates were sensitive to the 4 first line anti-TB drugs by proportion method on LJ medium. In addition, LPA was also performed on the 13 culture isolates which showed sensitive pattern to INH and RMP, thus demonstrating good concordance with conventional proportion method. All 13 patients clinically improved on first line anti-TB drugs (INH, RMP, EMB, SM and pyrazinamide) under the 6 month regime of DOTS (directly observed treatment short course). Studies have shown primary drug resistant tuberculosis to be low in Kerala and in this study, no drug resistance was found in the 14 smear negative TB isolates to INH, RMP, EMB and SM.[2627] A major limitation of this study was that drug sensitivity testing to pyrazinamide could not be performed.

As seen from Figure 2, this study population comprised of 128 males and 72 females. Other Indian studies have also shown similar male predominant pattern in tuberculosis which can be attributed to the gender dependent sociocultural factors, affecting the risk of exposure to TB and the ease of access to healthcare facility.[2628] In this study, 16.7% of HIV co-infected patients gave positive cultures from smear negative sputum samples. In India, the magnitude of smear negative TB in HIV infected patients is underestimated and there is a substantial diagnostic delay which increases the morbidity and mortality.[25]

CXR consistent with pulmonary TB were observed in 97 patients and 3 patients had non-specific findings but the latter group had clinical features highly suggestive of TB (evening rise of temperature, night sweats, weight loss, fatigue) with raised ESR. Culture was positive in one of these 3 patients with non-specific findings who was also HIV positive. Radiological findings may be altered in the presence of HIV co-infection proportional to the degree of immunosupression.[29]

Conclusion

Smear negative TB which are paucibacillary forms pose a challenging issue in the diagnosis and management of TB. Culture, despite its shortcomings and low positivity, still remains the gold standard for the diagnosis of EPTB and SNPT. However, automated liquid cultures have better isolation rates than the conventional LJ culture. Subjecting these isolates to rapid diagnostic tests like antigen detection and LPA can aid in the early institution of appropriate treatment regimen. For a wholesome effort to stop TB transmission, primary care physicians should be sensitized to the role of culture in the diagnosis of EPTB and SNPT and more research initiatives in this direction is a necessary step in the combat against TB.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.