Occurrence of tuberculosis among people exposed to cattle in Bangladesh

Abstract Background Tuberculosis (TB) has been an important public health concern in Bangladesh. The most common cause of human TB is Mycobacterium tuberculosis, while bovine TB is caused by Mycobacterium bovis. Objective The objective of this study was to determine the frequency of TB in individuals with occupational exposure to cattle and to detect Mycobacterium bovis among cattle in slaughterhouses in Bangladesh. Methods Between August 2014 and September 2015, an observational study was conducted in two government chest disease hospitals, one cattle market, and two slaughterhouses. [Correction added on 27 June 2023, after first online publication: In the preceding sentence, the year “2014” has been added after the word “August”.] Sputum samples were collected from individuals who met the criteria for suspected TB and had been exposed to cattle. Tissue samples were collected from cattle that had low body condition score(s). Both humans and cattle samples were screened for acid‐fast bacilli (AFB) by Ziehl–Neelsen (Z‐N) staining and cultured for Mycobacterium tuberculosis complex (MTC). Region of difference (RD) 9‐based polymerase chain reaction (PCR) was also performed to identify Mycobacterium spp. We also conducted Spoligotyping to identify the specific strain of Mycobacterium spp. Results Sputum was collected from a total of 412 humans. The median age of human participants was 35 (IQR: 25–50) years. Twenty‐five (6%) human sputum specimens were positive for AFB, and 44 (11%) were positive for MTC by subsequent culture. All (N = 44) culture‐positive isolates were confirmed as Mycobacterium tuberculosis by RD9 PCR. Besides, 10% of cattle workers were infected with Mycobacterium tuberculosis in the cattle market. Of all TB (caused by Mycobacterium tuberculosis) infected individuals, 6.8% of individuals were resistant to one or two anti‐TB drugs. The majority of the sampled cattle (67%) were indigenous breeds. No Mycobacterium bovis was detected in cattle. Conclusions We did not detect any TB cases caused by Mycobacterium bovis in humans during the study. However, we detected TB cases caused by Mycobacterium tuberculosis in all humans, including cattle market workers.


INTRODUCTION
Tuberculosis ( in Bangladesh (Zaman, 2010). According to the WHO, Bangladesh is one of 30 countries with a high TB burden worldwide, with an estimated annual incidence rate of 221/100,000/year in 2021 (Global Tuberculosis Report, 2021).
Zoonotic TB is a form of TB in people caused primarily by Mycobacterium bovis, which belongs to the Mycobacterium tuberculosis complex (MTC). Other potential MTC strains that cause zoonotic TB in Asia include Mycobacterium orygis identified from hospitalised TB patients in India (Duffy et al., 2020), dairy cattle, and rhesus macaque in Bangladesh (Rahim et al., 2017). Mycobacterium bovis causes chronic TB in cattle and other mammalian hosts (LoBue et al., 2010), affecting the production of milk and meat for consumption in these animals (Gutiérrez et al., 1995;Higino et al., 2011). Humans can become infected with Mycobacterium bovis through direct contact with infected animals, either by airborne transmission or by consuming infected unpasteurised milk or meat products (Grange & Yates, 1994). People in specific occupations such as farmers, veterinarians, slaughterhouse workers, and butchers have an occupational risk for zoonotic TB (Adesokan et al., 2012;Robinson et al., 1988).
Globally there were an estimated 140,000 new human TB cases and 11,400 deaths caused by Mycobacterium bovis in 2020 (Global Tuberculosis Report, 2020). In South-East Asia, the estimated number of zoonotic TB cases was 43,400 in 2020 (Global Tuberculosis Report, 2020, Ramos et al., 2020). The actual burden of zoonotic TB is unknown due to the lack of surveillance data in most low-income countries (Müller et al., 2013;Olea-Popelka et al., 2016;Wedlock et al., 2002). People occupationally exposed to cattle, such as livestock farmers and abattoir workers, are at a higher risk of contracting TB (Khattak et al., 2016). Evidence suggests that 2% of livestock farmers and 25% of abattoir workers were infected with TB in Pakistan (Khattak et al., 2016). Another study reported high prevalence (76%) of latent TB infection among dairy farm workers (tested by tuberculin skin test) exposed to cattle in Mexico (Torres-Gonzalez et al., 2013).
There is evidence that 10% of livestock traders were infected with TB in Nigeria (Hambolu et al., 2013). Likewise, about 2% of slaughterhouse workers were reported to be infected with TB in Iraq (Al-Thwani & Al-Mashhadani, 2016).
Bovine tuberculosis is an important cattle health problem in lowand middle-income countries. Globally, an estimated 7.4% of cattle had positive reactions to tuberculin skin tests (Grace et al., 2012). The tuberculin skin test (TST) is the primary screening test used to identify cattle infected with bovine tuberculosis. The screening test is not likely perfect, considering its sensitivity (the ability of the test correctly identify animals with the bovine tuberculosis) and specificity (the ability of the test to correctly identify animals without bovine tuberculosis) (Praud et al., 2015). The sensitivity and specificity of the TST can be influenced by a variety of factors, including the strain of the bacteria causing bovine tuberculosis, purified protein derivative products, the subjective injection, measurement ability of the test performer, the age and immune status of the animal being tested (De La Rua-Domenech et al., 2006;Kleeberg, 1960;Snider, 1982;Schiller et al., 2010). Bovine tuberculosis surveillance is rare in in Bangladesh. However, several cross-sectional studies reported the prevalence of bovine tuberculosis in cattle, ranging from 3% to 28% according to TST in different geographical areas in Bangladesh (Biswas et al., 2017;Islam et al., 2021;Islam et al., 2020;Islam et al., 2007;Mahmud et al., 2014;Pharo et al., 1981;Samad & Rahman, 1986 (Ullah et al., 2015). This high animal-human density creates opportunities for close contact between TB-infected livestock and humans during handling of animals and production of animal products, specifically milking, herding cattle and goats, slaughtering, handling skins and hides, moving cow dung, and plastering walls with dung or mud (Grace et al., 2012). However, there is no published data on the frequency of TB in humans who are occupationally exposed to cattle at the human-animal interface in Bangladesh. We hypothesised that individuals having occupational exposure to cattle at the human-animal interface are at high risk for contracting TB in Bangladesh.

Study settings, design and period
Between August 2014 and September 2015, an observational study was conducted in two government chest disease hospitals, one cattle market, and two slaughterhouses in Dhaka (Figure 1 with or without production of sputum despite the administration of a nonspecific antibiotic (National Tuberculosis Control Programme, 2004) and exposed to cattle while feeding, handling or cleaning at cattle farm in any part of their life >2 weeks before the interview.
After obtaining written informed consent, a medical technologist collected three consecutive sputum specimens from each TB-suspected participant after oral gurgling with sterile water, and specimens were kept in labelled screw cap disposable plastic bottles. Additionally, a trained medical technologist screened cattle market workers with suspected TB (persons with a persistent cough for 3 weeks or more, with or without production of sputum) following the guidelines of the Bangladesh National Tuberculosis Control Program (NTP) for suspected TB (National Tuberculosis Control Programme, 2004) and who were exposed to cattle in the past year. Cattle market workers were considered to be exposed to cattle at the market if they had contact with them during feeding, selling meat, and cleaning market places in which cattle were sold at least 2 weeks prior to the interview. After screening cattle workers with suspected TB, medical technologists collected three consecutive sputum specimens from each participant after oral gurgling with sterile water, and specimens were kept in labelled screw cap disposable plastic bottles. Sputum samples were collected in sterile containers and stored at an appropriate temperature of 2-8 • C before being transported to the Mycobacteriology Laboratory at the International Centre for Diarrhoeal Disease Research, Bangladesh (icddr,b), where they were stored at −20 • C until testing.
Regarding the number of people to be selected for sputum sample collection, we assumed that cattle exposed group had a 3% prevalence of tuberculosis caused by Mycobacterium bovis (Adesokan et al., 2012) with a 1.5% precision, and 95% confidence level, a sample size of 497 was calculated.

Sampling strategy for animals
We collaborated with Dhaka City Corporation South (under the Ministries of Local Government and Rural Development) to conduct the study. The field team, comprising trained veterinarians, field research assistants, visited each slaughterhouse twice a week in the early morning. A trained veterinarian assessed the overall physical status of animals using a five-point body condition score (BCS) scale, where animals were classified as emaciated (BCS 1), thin (BCS 2), normal (BCS 3), muscular (BCS 4), and fat (BCS 5) (Katale et al., 2013;Msangi et al., 1999). The BCS score of 2 was used to screen animals for sample collection. The animals had a BCS score of 2, indicating emaciation to thinness, and they had a sunken appearance with visible ribs, hips and backbone (Edmonson et al., 1989). Veterinarians collected tissue samples that revealed visible granulomatous lesions and/or caseous masses from the lungs, liver, intestines, and lymph nodes of slaughtered cattle that had emaciated to thin body condition. The tissue samples were collected in sterile containers and stored at an appropriate temperature of 2-8 • C before being transported to the Mycobacteriology Laboratory at the icddr,b where they were stored at −20 • C until testing.

Data collection
Field research assistants collected demographic information (age, sex) from human participants using a separate, structured questionnaire.
Veterinarians collected demographic information (age, sex, type of breed) of sampled cattle from animal owners/abattoir worker using a structured questionnaire.

Laboratory testing
2.2.1 Human sputum and cattle tissue processing, staining and culture Human sputum and cattle tissue samples were obtained according to the standard operating procedures and guidelines (Petroff, 1915;Zaman et al., 2006). Human sputum specimens were digested and decontaminated (Petroff, 1915) and then inoculated on Lowenstein-Jensen (L-J) slant cultures. The L-J slants were incubated at 37 • C for up to 8 weeks and visibly examined once per week for contamination and growth of Mycobacterium spp colonies. The cattle tissue samples were processed following the technique described earlier (Rahim et al., 2007). In brief, small pieces of tissue were sliced by a sterile surgical blade and homogenised in sterile 5 mL phosphate-buffered saline (PBS) using a tissue homogeniser in a bio-safety cabinet. The homogenate was mixed with an additional 10 mL PBS and allowed to settle for 15 min at room temperature. Approximately 5 mL of supernatant was collected and decontaminated following Petroff's methods (Petroff, 1915). Finally, the pellet was re-suspended in 1 mL PBS and used for culture on two Lowenstein-Jensen (L-J) slants with and without sodium pyruvate. The processed samples were inoculated simultaneously in two L-J slants; one with glycerol but no sodium pyruvate for the growth of M. tuberculosis and another L-J contained sodium pyruvate but not glycerol for the growth of Mycobacterium bovis (Corner & Nicolacopoulos, 1988). Human sputum and cattle tissue samples were considered culture negative if no visible mycobacterium colonies were seen on L-J slants following 8 weeks of incubation at 37 • C (Banu et al., 2012). Human sputum and cattle tissue specimens were also examined for acid-fast bacilli (AFB) using Ziehl-Neelsen (Z-N) staining and light microscopy ( Figure 2) following methods described earlier (Organization, 1998).

Anti-mycobacterial drug susceptibility test
Drug susceptibility testing (DST) was done on culture-positive isolates following conventional proportion susceptibility methods (Canetti et al., 1969). MTC isolates were tested for susceptibility to isoniazid, rifampicin, ethambutol, and streptomycin according to the method described by Canetti et al. (1969). An isolate was considered to be resistant to a given drug when any growth of 1% or more above the control was observed in a quadrant plate containing the drug (Banu et al., 2012). Pyrazinamide (PZA) susceptibility testing was also done on all MTC isolates using the methods described by Rahman et al. (2017).

2.3.1
Mycobacterium tuberculosis-specific region of difference 9 (RD9) analysis Polymerase chain reaction (PCR)-based on RD9 deletion analysis was performed using forward: 5t' CGATGGTCAACACCACTACG-3t' and F I G U R E 2 Study flowchart.
Each reaction mixture was prepared in a total volume of 20.0 μL, con-

2.3.3
Quality assurance of the testing All the procedures, including sample preparation, decontamination, and culture, were done in the Mycobacteriology Laboratory of icddr,b (https://www.icddrb.org/). All the procedures were done in aseptic conditions using a class II A2 Biosafety cabinet and air-locked centrifuge to avoid any cross-contamination or aerosol generation. Also, laboratory personnel wore appropriate personal protective equipment while processing the specimens. Moreover, we used appropriate controls included in the L-J slant culture method and the Z-N staining.
During inoculation into L-J slant culture, we always used PBS as a negative control and H37Rv as a positive control. For Z-N staining, we referenced known graded positive and negative slides.

Data analysis
We summarised the categorical variables using frequencies and percentages. We reported the means with standard deviations of continuous variables for symmetric distributions and medians with interquartile ranges (IQRs) for asymmetric distributions. We calculated the overall proportion of TB (based on positive AFB staining and/or culture results) in humans and cattle with 95% confidence intervals (CI).

RESULTS
A total of 412 humans sputum samples were analysed in this study (Table 1). Of the 412, 382 and 30 samples were collected from the hospitals and cattle market, respectively. The median age of human participants was 35 (IQR: 25-50) years, 75% were male (Table 1), and the median body mass index (BMI) was 17.9 (IQR: 16.7-19.6). Approximately two-thirds of the participants were aged 16-45 years (Table 1).
Most participants (85%) reported receiving the Bacillus Calmette-Guerin (BCG) vaccine in the past. All human participants reported that they had been exposed to cattle the previous year.  (Table 3). The median BCS of cattle was 2 (IQR: 2-2.5) ( Table 3).
Most of the cattle (76%) were female (Table 3), and the majority of the cattle (67%) included in the study were pure indigenous breeds. During this study, no cattle tissue samples had granulomas lesions. All the cattle tissue samples were neither AFB nor culture positive for MTC.

DISCUSSION
Our findings provided a preliminary insight about TB among people exposed to cattle and those working in slaughterhouses in Dhaka city.
The study detected pulmonary Mycobacterium tuberculosis among 11% of humans who were presented to the chest disease hospitals with symptoms of TB, as well as 10% of workers in the cattle market. The proportion of Mycobacterium tuberculosis infection among cattle workers was higher (10%) in this study compared to another study (6.67%) conducted in Bangladesh (Rahman et al., 2015).  (67) Body condition score, median, IQR 2 (2-2.5) of Mycobacterium tuberculosis via droplets to buyers and fellow workers in the cattle market, where crowded and close contact is common.
We would recommend a periodic TB screening program for cattle market workers and conducting a health education campaign about transmission, diagnosis, treatment, and prevention of TB.
Our study found that two-thirds of Mycobacterium tuberculosis cases occurred among the 16-45 years age group in humans, which is a higher proportion compared to the report of the National Tuberculosis Control Program 2017 (63% among 15-54 years age group), but it was not statistically significant. This study also provided information on the spoligotype and drug susceptibility pattern of Mycobacterium tuberculosis isolates among humans.
Our study did not detect TB cases caused by Mycobacterium bovis among humans. It was possible that they had never been exposed to Mycobacterium bovis-infected animals or contaminated, unpasteurised animal products in the year prior to sample collection. It could also explain that all Mycobacterium tuberculosis isolates were identified from the sputum samples of suspected pulmonary TB (PTB) cases, which could have decreased the likelihood of detection of Mycobacterium bovis in this study. Because Mycobacterium bovis is associated with extrapulmonary TB, which limit the ability to detect Mycobacterium bovis in sputum sample. Studies suggest that Mycobacterium bovis infections cause a higher proportion of extra pulmonary TB (EPTB) compared to PTB cases in humans (9.4% vs. 2%) (Cosivi, 1998;Prasad et al., 2005;Rasolofo-Razanamparany et al., 1999). Also, other potential mycobacterium species, such as Mycobacterium orygis, cause zoonotic TB infections in humans in India (Duffy et al., 2020) and dairy cattle, and rhesus macaque in Bangladesh (Rahim et al., 2017) were not tested in our study samples. A future study could target EPTB cases to detect Mycobacterium bovis infection in individuals using whole genome sequencing (Duffy et al., 2020) or more comprehensive PCR (Duffy et al., 2020) in Bangladesh.
The study was not able to detect Mycobacterium bovis in cattle in the slaughterhouses investigated, given negative bacteriological and molecular assay test results. The absence of Mycobacterium bovis in cattle might be explained by the fact that 67% of sampled cattle were an indigenous breed that is naturally less susceptible to Mycobacterium bovis infection than nonindigenous breeds of cattle (Liston & Soparkar, 1917;Soparkar, 1926;Vordermeier et al., 2012). In fact, Bangladeshi indigenous cattle are still more prevalent (85%) than the cross-bred cattle (15%) (Hamid et al., 2017).

Previous studies have indicated that
Mycobacterium bovis infection is more common among nonindigenous than indigenous cattle breeds in Bangladesh (Islam et al., 2007;Hossain et al., 2012;Rahim et al., 2007;Samad & Rahman, 1986). Prior studies using a tuberculin skin test also reported that the prevalence of Mycobacterium bovis among cattle in Bangladesh ranges from 2.1% to 34% (Islam et al., 2010;Samad & Rahman, 1986;Uddin et al., 2014).

ACKNOWLEDGEMENTS
We would like to thank the veterinary doctors, physicians, abattoir workers, and veterinary field assistants who provided invaluable assistance with selection of subject participants and collecting samples and epidemiologic information for this study. We would also like to thank Gladys Leterme for her editorial assistance, as much of the English language checking, and given there are two coauthors, one of whom is the senior author, whose primary language is English, those were helped to review the English language in the manuscript.
The research protocol was funded by US Centers for Disease Control and Prevention (CDC) through their cooperative agreement no.
U01CI000628-05. icddr,b acknowledges with gratitude the commitment of the US CDC to its research efforts. icddr,b is also grateful to the Governments of Bangladesh, Canada, Sweden, and the United Kingdom for providing core/unrestricted support.

CONFLICT OF INTEREST STATEMENT
The authors declare no conflict of interest in doing these studies.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.

DISCLAIMERS
The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the Centers for Disease Control and Prevention or the Agency for Toxic Substances and Disease Registry.

ETHICS STATEMENT
The study protocol was reviewed and approved by the Ethical