TB PCR in BAL and EBUS-TBNA

This is a prospective multi-centre study across the UK including London, Birmingham, Leicester and Manchester which account for the highest proportion of TB cases in the UK. These are centres routinely investigate patients with suspected TB and routinely use Xpert Ultra in their practice.

Any patient who is suspected of having pulmonary or mediastinal TB who is routinely undergoing a bronchoscopy or EBUS-TBNA for clinical purposes will be recruited. These patients will be identified mainly by the TB services but also via Accident and Emergency, wards and clinics. Inclusion criteria include any consenting adult ≥18 years of age suspected of having pulmonary or mediastinal TB who is undergoing a routine clinical bronchoscopy for BAL or EBUS-TBNA.

Any patient identified who need a bronchoscopy or EBUS-TBNA for suspected TB will have this done in the local hospital. Patients will follow the routine follow up arrangements in line with the local trust guidelines but the researchers will review any routine follow up data up to 3 month post-procedure if available. There will be no additional procedures or visits as the researchers will merely use routinely available clinical data and follow up data if available.

The study aims to recruit a minimum of 323 participants undergoing a BAL, and 323 participants undergoing an EBUS-TBNA for a minimum of 6 months or until the recruitment targets are reached. The researchers will review any routine follow up data for a minimum of 3 months post-procedure. There will be no additional study follow up visits.

The study outcome is the diagnostic performance of Xpert Ultra in BAL and EBUS-TBNA samples in patients with suspected TB against conventional modalities available (smear microscopy, culture, cytology) using a subgroup analysis of the different clinical categories attributing to the likely TB diagnosis incorporating follow-up data and the clinician's decision to treat. A clinical expert panel will review the culture-negative cases blindly to categorise each case into a clinical diagnosis category.

Category Criteria:

1. Culture confirmed TB: Microbiological culture of MTB, and clinical and radiological findings suggestive of TB.
2. Highly probable TB: Clinical and radiological features highly suggestive of TB and unlikely to be caused by other diseases, a decision to treat made by a clinician, appropriate response to therapy, and histological evidence if available.
3. Clinically indeterminate diagnosis: Final diagnosis of TB neither highly probable nor reliably excluded.
4. Highly unlikely or TB excluded: Other diagnosis made other than TB (e.g. sarcoidosis, cancer or lymphoma).

Prospective data will be collected from Imperial College Healthcare NHS Trust, London North West University Health Care NHS Trust, Chelsea and Westminster Hospital NHS Foundation Trust, Royal Free London NHS Trust, The Hillingdon Hospitals NHS Trust, Barts Health NHS Trust, University Hospitals of Birmingham NHS Foundation Trust, University Hospitals of Leicester NHS Trust, Manchester University NHS Foundation Trust which together serve a diverse cultural population with a high prevalence of TB in the UK and offer EBUS-TBNA services.

Clinical (patient demographics, medical history, TB symptoms, previous TB, history of exposure to TB, medications, HIV status and immunosuppression), microbiological, cytological data, radiology and biomarkers of infection will be collected throughout the study and documented on the following case report forms (CRFs). A paper or electronic CRF will be used as a clinical data collection tool which will comply with GCP, FDA CFR-21 Part-11, and HIPAA.

This data will be collected by the local clinical team, a clinical research fellow or research nurses using participant hospital records or from the participant directly. Personal data will be kept in pseudo-anonymised form with a link code which can be used to refer to the participant's information. This link code will only be available to the clinical or clinical research team. There will be an audit trail of the staff entering the data.