Silicosis and Silicotuberculosis Among Small Scale Gemstone Miners in Northern Tanzania

Background Artisanal and small-scale mining (ASM) produces a quarter of the global supply of key metals for the electronics industry and provides direct employment for an estimated 45 million people. Approximately 10 million are in sub-Saharan Africa (SSA), of whom over 1 million work in Tanzania. ASM is characterized by occupational insecurity and low levels of safety and environmental protection, but there is little research into its risks, particularly respiratory and communicable diseases.

Silicosis is an occupational respiratory disease caused by inhalation of respirable crystalline silica (RCS). It usually presents irreversibly after decades of exposure, although accelerated disease associated with high initial exposures can lead to respiratory failure within 1-2 years. To date, silicosis has been examined in only two ASM populations: in Brazil, where radiographic evidence was reported in 37% of 348 current and ex-miners, and in Zimbabwe, where a prevalence of 11% was reported among 514 current miners.

Silicosis and RCS exposure increase the risk of pulmonary tuberculosis (PTB). In miners with silicosis, human immunodeficiency virus (HIV) is also known to multiplicatively increase the risk of TB. Correspondingly, a high PTB prevalence of between 3% to 14% has been reported in several studies of sub-Saharan African ASM workers. The role of miners in community PTB transmission is of considerable public health importance but remains unclear; whole genome sequencing (WGS) is the most accurate available method for assessing this. HIV prevalence across small-scale mining populations ranges from 1.8-18%. The role of HIV in the development and progression of silicosis is unclear and requires further study, with conflicting evidence on whether it reduces or increases the risk of silicosis.

The World Health Organization (WHO) provides a strong recommendation for screening among silica-exposed populations. This recommendation is based on an estimated number needed to screen of 36 in high incidence settings and an increased mortality from TB amongst miners with silicosis, with a corresponding 82% reduction in mortality in screened compared to unscreened miners with TB. Despite this recommendation, there is limited evidence, particularly from Sub-Saharan Africa and in the last two decades, of optimized methods for doing this. Beyond screening, differentiating silicosis, tuberculosis, and silicotuberculosis is clinically challenging, leading to misclassification across all diagnoses. Symptoms are common among miners; for example, in ASM TB screening studies, between 23-39% meet symptomatic criteria for presumptive TB. There is also a growing appreciation that subclinical TB is common and plays an important role in transmission. A screening tool that, either alone or in tandem with current approaches, improves the sensitivity and specificity for TB would be useful diagnostically and programmatically, potentially allowing allocation of resources towards higher risk individuals. One recently investigated method which has shown promise in meeting the WHO target product profile of 90% sensitivity and 70% specificity for a triage tool is C-reactive protein (CRP). CRP testing can be performed at the point of care for less than $2-3. Most recently, WHO guidance has recommended CRP (with a threshold of >5mg/L) may be used in screening for TB disease among adults and adolescents living with HIV.

The tanzanite gemstone is mined exclusively in a 5x2 km strip of land in Mererani, Northern Tanzania. Miners use deep shafts and basic techniques. Three sources of evidence raise concern that the burden of silicosis and PTB among the estimated 9,000-12,000 ASM tanzanite miners is high. First, a recent cross-sectional survey of 330 miners estimated a silicosis prevalence of 30% on chest radiograph and a microbiologically confirmed PTB prevalence of 6% among miners and 8% in the surrounding community. Miners had a median age of 35 years and just 5 years of employment. Second, a case series from the local tertiary referral hospital found a primary diagnosis of silicosis or silicotuberculosis made up 22% of respiratory admissions; over half were under 45 years old and 63% required oxygen therapy. Third, in an occupational clinic for miners and their families, 22% of 315 miners had silicosis on chest X-ray (CXR) and 18% of 243 tested had confirmed PTB.

Study Rationale

There is clear evidence of significant occupational risks resulting in a high burden of aggressive silicosis in young tanzanite miners and a high burden of tuberculosis among miners and the community. By using a prospective cohort study design, the investigators will be able to describe the rate of change of silicosis, which is important for understanding the disease burden. Importantly, for the case of occupational research, prospective study reduces the risk of healthy worker bias.

Cross-sectional evidence is highly suggestive of a high burden of aggressive silicosis in young tanzanite miners. To understand and ultimately to prevent the risks, prospective study is required. PTB prevalence is also high among miners and the local community.

This study aims to characterize the progression of silicosis among a cohort of tanzanite miners and measure PTB prevalence and transmission patterns in a cross-sectional, screened population of miners and community members. Given a high PTB prevalence amongst miners, improved triage methods are needed, and utilizing CRP represents a feasible, cost-effective approach to this.

Importantly, cost-effective prevention strategies for silicosis and TB are available. A better knowledge of the extent of the silicosis burden in the Tanzanite communities and improved understanding of potential barriers and opportunities for implementation will inform targeted and appropriate preventive strategies.

This study is part of a long-running partnership between Kibong'oto Infectious Diseases Hospital (KIDH) and the mining community that provides clinical care and performs clinical research. Through the identification of occupational respiratory diseases, silicosis and tuberculosis, miners will have access to care and treatment. In the longer term, this study, dependent upon its findings, and collaboration with Imperial College, provides a platform for interventional studies potentially aimed at reducing RCS exposure and silicosis and tuberculosis disease. By identifying groups of miners or characteristics of miners who face higher risks of disease - for example, those miners who are drillers - the investigators may be able to intervene effectively and prevent the risks associated with these groups. Importantly, in the case of tuberculosis, preventing disease among a group of miners may also reduce the risk of transmission within the community - understanding this dynamic is also an aspect of the study.

More broadly, the ASM community is a vital part of the global economy. By promoting a safe and secure working environment, the study directly informs the United Nations Sustainable Development Goals (SDG 8.8). Given the large global artisanal mining population - 1 million of whom are based in Tanzania - the study would be of global importance and applicability.

Objectives

Primary Objectives

• Broad: To test the hypothesis that small-scale tanzanite miners have a high rate of rapidly progressive silicosis, characterized by changes in chest radiology and markers of respiratory impairment over an 18-month period. Furthermore, the investigators hypothesize that the rate is directly related to high RCS exposure and is modified by the presence of PTB and HIV.
• Specific: The investigators shall measure change in the International Labour Organization (ILO) classification of pneumoconioses over an 18 month in a cohort of 410 small-scale tanzanite miners and test whether this is related to cumulative RCS exposure.

Secondary Objectives

1. Broad: To test the hypothesis that small-scale tanzanite miners have a high prevalence of PTB which is related directly to high exposures to RCS during mining activity and is modified by the presence of silicosis and HIV.

   • Specific: The investigators shall measure the prevalence of microbiologically confirmed PTB and test whether this is related to cumulative RCS exposure in a cross-sectional study by the end of quarter 1 of the study period.

2. Broad: To test the hypothesis that the proportion of clustered PTB cases among tanzanite miners and the community is high and that the proportion of community cases transmitted from miners is higher than the proportion transmitted within the community.

   • Specific: The investigators shall use whole genome sequencing (WGS) methods to observe whether cases of PTB from a cross-sectional and 18-month cohort study are clustered and if a community member is more likely to have contracted TB from a miner or a community member.

3. Broad: To test the hypothesis that occupational disease-related morbidity amongst tanzanite miners results in 'catastrophic' economic loss.

   • Specific: Using standardized tools, the investigators shall measure the costs of TB and silicosis treatment in an 18-month cohort of miners and compare these costs to a standard definition of catastrophic costs for TB treatment.

4. Broad: To test if adding C-reactive protein (CRP) measurement to standard PTB screening algorithms in tanzanite miners improves its sensitivity and specificity.

   • Specific: The investigators shall test the sensitivity, specificity, and area under the receiver operating curve of CRP compared to a culture reference standard for the diagnosis of TB in a cross-sectional study by the end of quarter 1 of the study period.

5. Broad: To use whole genome sequencing (WGS) to determine rates of Mycobacterium tuberculosis (Mtb) drug resistance in the mining community and examine what proportion of drug-resistant cases are associated with silicosis and clustering over an 18-month period.

   • Specific: The investigators shall describe the proportion and characteristics of drug-resistant tuberculosis isolates from an 18-month cohort study as identified by whole genome sequencing.

6. Broad: To elicit stakeholder understanding of the causes of occupational respiratory disease and the acceptability of interventions designed to reduce their risks.

   • Specific: The investigators shall use qualitative methods, including interviews and focus groups, to explore stakeholder understanding of occupational respiratory disease and possible interventions by the end of quarter 2 of the study.

   Methodology

   To achieve the primary objective, the investigators will establish a representative prospective cohort study of 410 small-scale tanzanite miners and measure their baseline prevalence of silicosis, PTB, and HIV infection using validated methods. The primary outcome measure will be change in chest radiograph using the International Labour Organization (ILO) scale over the 18-month study period, starting from Q1 2023 and ending in Q3 2024. The investigators will also measure respiratory symptoms and lung function through spirometry. The questionnaire is designed to take 20 to 25 minutes.

   To measure PTB prevalence amongst miners (secondary objective 1), the investigators will screen miners in the cohort for PTB at enrolment and at 18 months using a comprehensive screening approach of symptoms and chest radiography, followed by a molecular test called GeneXpert (Xpert) and sputum culture. PTB will be defined as a positive result on culture or Xpert.

   During the study follow-up, miners will be invited to attend the occupational health clinic at Kibong'oto Infectious Diseases Hospital (KIDH) if they are concerned they have PTB; TB and HIV testing and treatment is free at the clinic.

   To investigate clustering of PTB within the mines (secondary objective 2), the investigators will undertake WGS on culture-positive Mycobacterium tuberculosis (Mtb) samples from the mining cohort and their contacts (>18 years of age). The investigators will determine the proportion of cases arising from transmission within mines by identifying clustered cases compared to non-clustered cases. The WGS pipeline will also ascertain the proportion of miners with anti-TB drug resistance (secondary objective 5) and speciate the TB isolates.

   In a linked cross-sectional survey, the investigators will enroll approximately 2400 miners and community members from an ongoing community-based screening program led by KIDH, and from incident TB cases from the local Mererani health center and KIDH clinic. Amongst symptomatic individuals, the investigators will perform Xpert testing and Mtb culture, with positive cultures proceeding to WGS. The investigators will collect demographic and PTB risk factor data on enrolled participants and combine this dataset with the miner's cohort to create a combined dataset of Mtb WGS samples. By performing WGS on all positive cultures, the investigators can determine the proportion of community cases arising from miners based on their phylogenetic association. The investigators will then investigate risk factors amongst clustered vs non-clustered cases and whether this is modified by being a miner/non-miner. The cross-sectional questionnaire is designed to take less than 5 minutes.

   To investigate change in economic status (secondary objective 3), the investigators will use questionnaire methods to measure changes in occupational status, monthly estimated income, and healthcare expenditure of miners in the cohort over an 18-month period. The investigators will then determine if catastrophic economic loss has occurred using a standard definition. The addition of CRP and phone screen testing (PoST) to triage testing amongst miners will allow testing of whether the diagnostic accuracy of screening for the presence of PTB amongst miners can be improved (secondary objective 4).

   A subgroup of stakeholders will be invited to take part in semi-structured qualitative interviews and focus groups (secondary objective 6). The number of participants will be determined by saturation of data; however, theoretically is expected to be approximately 20 individuals and, similarly, 4-5 focus group sessions of 4-8 individuals. Interviews and focus groups will be expected to last approximately 20 and 60 minutes respectively. These stakeholders will be asked for their understanding of the causes and etiology of respiratory diseases amongst themselves and their work colleagues and regarding the potential acceptability of interventions designed to reduce their risk of occupationally acquired disease. The investigators will use the capability, opportunity, motivation, behavior (COM-B) model to understand from a theoretically informed perspective, opportunities and barriers to interventions. Interviews (conducted in Swahili) will be audio recorded, transcribed, anonymized, and analyzed using thematic analysis.

   Finally, in participants who provide specific consent, sputa and Mtb culture isolates from both studies and serum from the cohort study will be retained in -20 C freezers at KIDH for future studies. Separate ethical approval will be required for future studies on these retained samples.

   Study Procedure

   Persons attending the TB screening program who meet the eligibility criteria for either the cross-sectional survey alone or both cross-sectional and cohort studies will be identified by study staff involved in the screening program and offered the opportunity to take part in the studies. Should they express interest, an informed consent process and subsequent study enrollment will be undertaken by different staff members. The consent process will include the provision of a participant information sheet to the potential participant. The participant will be provided time to read this document or, should they be illiterate, it will be read to them by the study staff member.

   Both the cohort study and cross-sectional survey will utilize routine data collected as part of standard TB screening. This will include demographic data, symptom data, chest X-ray (CXR), HIV screening, and (where applicable) a GeneXpert (Xpert) result. Informed consent will be captured on paper forms. All results will be made available to the participant as they are completed, for example, CXR reports (by the onsite radiographer) and Xpert results will become available on the day and will be provided to the participant.

   In addition to participants identified during the TB screening program, contacts within the mines and the homes of persons identified to have TB will be contact traced for TB and invited for testing. For the miners, the investigators will do this by contacting the mine managers and asking them to invite fellow teammates or other close contacts of the identified case. For home contacts, if they are able to attend the screening program, the investigators will ask the case to encourage their family contacts to attend. If they are more distant, the investigators will offer two options: either labeled sputum samples can be taken home and brought to the screening program, or the investigators will contact local TB control officers in the case household's area, who will arrange for testing. Identified cases will be linked to local treatment programs. Furthermore, Xpert-positive incident TB attending Mererani health center/KIDH during a three-month period around the time of the study will be offered the opportunity to enroll in the cross-sectional survey, should they meet its eligibility criteria.

   For cohort participants, face-to-face study visits will take place at month 0 and 18 months, to coincide with TB screening program visits. These dates are currently estimated at Q1 2023 and Q3 2024. Additional 6 and 12-month telephone visits will be arranged. These and the 18-month visit are detailed in the study protocol.

   Routine data collection for screening participants in the cross-sectional survey includes:

   • A questionnaire from the National TB and Leprosy program asking about demographic data and TB symptoms, administered by study staff and transcribed electronically using REDCap software
   • A digital CXR (all miners and community members who are WHO symptom screen positive) with initial reports provided by the onsite radiographer
   • Voluntary counseling and testing for HIV with linkage to care
   • For those with a positive symptom screen for TB, sputum testing for TB with GeneXpert (Xpert)

   Additional testing and data collection as part of the cross-sectional survey will include:

   • For miners, finger prick testing for CRP using the point of care iChroma™ M3 device
   • For participants screening positive for TB (estimated approximately 20-30% based on previous studies), a series of two sputum samples will be requested which will be pooled and then split with 1ml taken forward for Xpert testing (as noted in routine testing) and the remaining half retained.
   • For community participants in the cross-sectional study, if the Xpert sample is positive, the second retained sample will be taken for Mtb culture.
   • For miners in the cross-sectional or cohort study participants, all retained sputum samples will undergo Mtb culture.
   • All positive Mtb culture results from either study will proceed to WGS.
   • Simple techniques to enhance sample production will be taught to all staff. Procedures for obtaining, handling, and processing sputum samples will follow standardized KIDH laboratory guidelines.
   • All positive Mtb culture samples will be stored at -20 C according to standard KIDH lab protocols.
   • Swab samples from participants' mobile phone screens will be collected. Real-time PCR will be used to identify Mtb in the PoST sample similar to methods described in recent studies. This will be performed at a later date and results will not be used clinically as this is experimental.

   Additional testing and data collection as part of the cohort study will include:

   • A questionnaire asking about respiratory symptoms and past respiratory medical history, occupational related dust exposures, and economic status. This will also be administered by study staff.
   • Venepuncture with samples sent for full blood count (FBC), renal function and electrolytes (U+E), and CD4 count and viral load (if HIV positive and no result available in the last 12 months, or clinically indicated) and a single (5 ml) EDTA and serum sample retained for storage. If CD4 count is \<350 cells/mm3 or HIV viral load is >1000 copies/mL, then participants will be linked to their local HIV clinic for further management as per national guidelines.
   • Pre- and post-bronchodilator spirometry will be undertaken by a trained member of the study team with data recorded electronically

   For all participants who identify as smokers, a brief smoking cessation intervention will be provided.

   Participants for the qualitative element of the study will be recruited from the miners' cohort and relevant stakeholders based on purposive sampling and participant willingness to be involved. A separate patient information leaflet and informed consent process will be undertaken for this group. To attempt to reduce sampling bias, stakeholders will be invited through a consensus decision by the research team. Individuals and groups of miners will be invited from attendees to screening on multiple different days evenly distributed through the recruitment period.

   Follow Up During follow-up, cohort participants will receive two telephone follow-up appointments at approximately 6 and 12 months. These appointments are designed to ascertain their current location, brief health status, and occupational status.

   A second and final face-to-face study visit at month 18 will involve:

   • A questionnaire asking about change in occupational role and status, respiratory symptoms and new respiratory conditions, and change in economic status
   • Repeat sputum and testing using the same procedures as outlined in the month zero visit
   • A digital CXR
   • Spirometry with reversibility testing
   • Voluntary counseling and testing for HIV with linkage to care
   • Venepuncture with samples sent for full blood count (FBC), renal function and electrolytes (U+E), C-reactive protein (CRP), and a single (5 ml) EDTA and serum sample retained for storage.

   During the follow-up period, the mining cohort participants will be encouraged to attend the occupational clinic at Kibong'oto Infectious Diseases Hospital (KIDH) if they have any new respiratory symptoms that require investigation. They will be reimbursed for their travel expenses. TB and HIV testing and treatment at the clinic are free. As per standard TB diagnosis protocols at KIDH, sputum samples will be tested with GeneXpert (Xpert) and sent for TB culture. Positive culture results will be sent for whole genome sequencing (WGS). Outcomes for all participants who test positive for TB will be followed up using the Tanzanian National TB program electronic records system and contacted to encourage attendance if a loss to follow-up is noted. Follow-up will be limited to the duration of TB treatment (6 months for drug-sensitive TB, 9 months for multidrug-resistant TB (MDR-TB)).

   Sample Size

   The best estimate of the prevalence of silicosis in tanzanite miners is 30% following a median of 5 years employment in the mines. Local measurements of respirable silica indicate very high average exposure levels (up to 10.6 mg/m3). There are no local data on which to base sample size calculation for a study of disease progression; hence, the investigators have used information from longitudinal studies of workforces with comparably high silica exposures. In a study of slate pencil workers in India, with a baseline prevalence of 55%, 32% of 279 workers had progressive silicosis (defined as, using the International Labour Organization (ILO) classification, an increase in ≥2 steps in profusion of small opacities, or a ≥1 step increase in the size of small or large opacities, or the presence of new large opacities) over 16 months follow-up. This figure is close to that among the Hawk's Nest tunnellers in the US, many of whom died over a period of 18 months; and to those in populations of sandblasters in Turkey (82% progression over 4 years) and the Gulf of Mexico (52% over 2 years).

   On this basis, the investigators have conservatively estimated that 20% of a sample of tanzanite miners will show progression over 18 months; the total includes 1 in 3 workers with silicosis at baseline and an additional 10% with incident disease.

   Using a normal approximation of a binomial model, a confidence interval of +/-5%, and a significance level α=0.05 with power 95%, provides a sample size of 246 (estimated number of outcomes=49). To account for mine clustering, and using an intra-cluster correlation coefficient (ICC) of 0.01 (higher than the usual workplace ICC of \<0.003), the investigators applied a design factor of 1.49 (1 + ICC(n'-1) where n' = average cluster size), increasing the required sample size to 367 or 410 assuming a 10% loss to follow-up. Based on an estimated tuberculosis prevalence of 6%, and using the same approach, this study will be able to estimate prevalence to a 95% confidence interval of +/-2.3%. Likewise, assuming 1200 community members and a PTB prevalence of 8%, the investigators will be able to detect if 60% of community cases originate from miners (α=0.05, power 95%).