BACKGROUND INFORMATION The use of the rapid diagnostic test Xpert MTB/RIF (Xpert) for
simultaneous detection of Mycobacterium tuberculosis and Rifampicin resistance has been
scaled up in recent years globally, but real-life effectiveness studies could not
demonstrate the expected beneficial impact on patient important clinical outcomes such
as: time to diagnosis, time to treatment initiation and mortality among others.

The ideal rapid diagnostic tool should demonstrate better sensitivity and most
importantly contribute to the detection of more TB cases and decreased time to diagnosis
at decentralized Health Care facilities closest to the homes of the patients.

The new highly sensitive Xpert® MTB/RIF Ultra (Ultra) assay on GeneXpert® platform have
the potential to overcome the limitations of Xpert by providing molecular TB testing with
an improved sensitivity operating at peripheral facilities.

The presented study will assess the clinical impact and operational characteristics of
Ultra at decentralized Primary Health Care (PHC) level, compared to existing routine
diagnostic algorithm based on Xpert at specialized central TB service points (TB
Cabinets).

OBJECTIVES AND PURPOSE In 2010, the World Health Organization (WHO) endorsed Xpert
MTB/RIF, a rapid molecular diagnostic test with good performance characteristics[2,3]. In
smear-positive patients, the test has a pooled sensitivity of 89%, whereas the pooled
sensitivity in smear-negative, culture-positive patients is only 67%. Specificity is
excellent in both groups at 99%[4].Xpert has been successfully rolled out in most high TB
burden countries. However, efficacy trials showing excellent results were not further
supported by real world effectiveness studies done after implementation of test by
National TB Programs. In addition, tests in most cases have not been brought to the point
of care level and thus could not pose a significant change.

In 2014, the WHO convened a meeting of stakeholders to establish consensus on identifying
high-priority target product profiles (TPPs) for new tuberculosis diagnostics[5]. The
identified high priority TPPs included; 1) a point-of-care non-sputum-based test capable
of detecting all forms of TB by identifying characteristic biomarkers or bio signatures
(the biomarker test), 2) a point-of-care triage test, which should be a simple, low-cost
test that can be used by first-contact health-care providers to identify those who need
further testing (the triage test), 3) a point-of-care sputum-based test to replace smear
microscopy for detecting pulmonary TB (the smear-replacement test) and 4) a rapid
drug-susceptibility test that can be used at the microscopy-centre level of the
health-care system to select first-line regimen-based therapy (the rapid DST test).

The newly developed Ultra cartridge along with GeneXpert system, both products of
Cepheid, Inc. - have the potential to become the smear-replacement test as outlined in
the TPPs by overcoming operational challenges and the suboptimal sensitivity of the
present Xpert. The new Ultra cartridge can detect TB with sensitivity similar to standard
liquid culture [6]. As shown by Dorman et al for tuberculosis case detection, sensitivity
of Xpert Ultra is superior to that of Xpert in patients with paucibacillary disease and
in patients with HIV. The Ultra has also enhanced ability for detection of rifampicin
resistance[7]. As a downside, the higher sensitivity of Ultra leads to the detection of
non-viable mycobacteria, mainly in individuals with recent history of TB. The false
positive results are seen predominately in samples with small amount of bacilli ("trace
call" positive) which need to be interpreted according to recommendations by the WHO and
in the light of the local epidemiology[8]. Furthermore, introducing a new diagnostic
technology might lead to better patient care by shorter time to diagnosis, rather than
through specific improved performance characteristics[9].

The proposed project will investigate the next generation molecular test Ultra with
improved sensitivity onGeneXpert platform adjusted to environmental and clinical needs at
point-of-care level.

HYPOTHESES

1. Symptom-based prioritization of patients and testing with Ultra on GeneXpert
platform placed at PHC facility will substantially decrease time to diagnosis and
time to treatment initiation of tuberculosis patients.

2. The proportion of bacteriologically confirmed TB cases initiated on TB treatment
within 7 days from study enrolment will be at least 20% higher in the intervention
arm compared to control arm.

3. Ultra on GeneXpert platform performed at PHC level will decrease initial
lost-to-follow up by 30%.

4. Operations characteristics of Ultra will show feasibility to be implemented at
Primary Health care level.

MAIN OBJECTIVES OF THE TB DIAGNOSIS STUDY AT PHC LEVEL:

1. To assess the impact of symptom-based prioritization of patients and testing with
Ultra (intervention) in comparison to the standard algorithm (control) on time to
diagnosis and time to treatment initiation.

2. To determine the proportion of participants who have bacteriologically confirmed TB
and have TB therapy initiated within 7 and 30 days of enrolment in intervention arm
compared to the control arm.

3. To evaluate the impact of Ultra (intervention) in comparison to the standard
algorithm (control) on pre-treatment loss to follow-up.

4. To investigate operational characteristics of Ultra used on GX1 and GX4 platforms in
PHC facilities in Georgia.

Trial subject population

Presumptive pulmonary TB patients attending selected PHC facilities will be invited to
participate in the study.

TRIAL DESIGN

1.1 Primary and secondary endpoint

Primary endpoints:

1. (i) Time from enrolment to TB diagnosis: time interval between first approach to PHC
and TB diagnosis Time interval between first approach to PHC positive TB diagnosis
based on laboratory results (i.e. smear, culture, Xpert, Ultra) or clinical
judgement (ii) Time from enrolment to TB treatment initiation Time interval between
first approach to PHC and initiation of a TB treatment

2. Proportion of participants who have bacteriologically confirmed TB and have TB
therapy initiated within 7 (30) days of enrolment Numerator: Participants with
positive smear, culture, Xpert or Ultra result and who are on TB treatment 7 (30)*
days after enrolment Denominator: All participants of the study arm.

Secondary endpoints:

1. Number of presumptive TB cases at the PHC facility Number of people approaching PHC
with cough more than two weeks.

2. Proportion of patients with early stage TB disease referred to the NCTLD TB cabinet
Early stage disease will be defined as no cavitation on chest x-ray and negative
smear microscopy results

3. Time from onset of symptoms to diagnosis Time interval between first TB symptom and
TB diagnosis based on laboratory results (i.e. smear, culture, Xpert, Ultra) or
clinical judgement

3 Proportion of patients with unfavourable outcomes after enrolment and loss to follow up
measured at 6 months Unfavourable outcome is defined as lost to follow-up, death,
treatment failure, and patient transfer 4 Proportion of patients with bacteriologically
confirmed TB after 6 months Numerator: Participants with positive smear, culture, Xpert
or Ultra result 30 days after enrolment Denominator: All participants of the study arm. 5
Proportion of patients diagnosed with TB from PHC sites Numerator: Number of all
confirmed TB cases from all PHC site Denominator: Number of all confirmed TB cases
diagnosed during trial period in RCT catchment area (self-referred, referred from
specialized secondary, tertiary level clinics) 6 Operational characteristics of Ultra:

1. Error rates

2. Platform failure

3. User appraisal

1.2 Trial design

This is a pragmatic, prospective, parallel cluster randomized trial in PHC facilities,
routine diagnostic settings of the Georgian health system.

1.3 Measure to minimize bias

The objective of the study is to assess Ultra in routine use. The study-related
procedures will be embedded into the routine practice at the PHC facilities to avoid any
influence by the study. Enrolment criteria will be defined identically for both
intervention and non-intervention arms to avoid selection bias.

Selection of study sites will not consider factors such as urban representation. PHC
facilities with rationale amount of registered beneficiaries (population attributed to
receive health care at selected HCF) and number of TB suspects are located only at city
levels. The capital city Tbilisi will not be considered for the study due to a different
set-up of patient/sample flow in comparison to the rest of the country.

1.4 Study duration and duration of subject participation

Study duration will be 24 months:

- 1 month preparation (IRB approvals, agreement set-up)

- 3 months initiation: staff recruitment, qualitative data collection tool
development, study staff training, eCRF and online database development;

- 12 months: Prospective enrolment of the patients at randomized PHC facilities; End
of 12 month, collection of qualitative data on performance characteristics using
specially developed data collection tool

- 6 months follow up of individual study subjects;

- 2 months data collection, analysis and distribution of study results. Manuscript
publications: within three months after study completion

1.5 Recruitment

The most common symptom of Pulmonary TB "cough more than two weeks" will be a sign for
"presumptive TB case". This symptom will be added to standard registration log at the
entrance of every PHC facility in both intervention and control arms. Cough information
along with standard registration details will be filled by registration desk.

In the intervention PHC facilities, patients with cough for more than two weeks on
initial presentation will immediately be referred to the study nurse. The study nurse
will obtain informed consent. All consenting patients will obtain one on spot sputum
sample for further investigation with Xpert/Ultra on the site. After testing, cases with
positive Ultra results will be referred to TB cabinets. Notification of positive cases
will also be sent to TB cabinets and regional coordinators. In non-intervention PHC
facilities, patients with "cough more than two weeks" will be added to the standard
registration log at the entrance and will continue with the standard approach for
examinations and referral without having Xpert Ultra testing.

All study participants from intervention and control arm will receive the same standard
of care for tuberculosis diagnostics and consecutive treatment at TB cabinet level. TB
diagnosis will further be confirmed with standard examinations according to National
Guidelines approved by the MoH, which includes additional clinical examination, chest
x-ray and laboratory examinations (culture and DST), TB treatment based on molecular or
phenotypic DST and further follow up. The only difference will be: treatment will be
initiated based on Ultra results in the interventional arm and on MTB/RIF results or
clinical decision in control arm.

Enrolment will continue for a total of 12 months. Clinical and contact data (address,
phone numbers) will be collected from participants on day 0.

1.6 Inclusion criteria

- Presumptive pulmonary TB patients, as defined by the NTP treatment guidelines.

- Adults 18 years old and above

1.7 Exclusion criteria

- Patients with symptoms which are only attributable to extra-pulmonary TB

- Patients below the age of 18 years

1.8 Criteria for discontinuation of trial site

Reasons for the early closure of a study site by investigator may include but are not
limited to:

- Failure of the site staff to comply with the protocol, the requirements of the
IRB/IEC or local health authorities, the sponsor's procedures, or GCP guidelines,
including the failure to meet capacity requirements outlined at the start of the
study.

- Inadequate recruitment of participants by the site

- Discontinuation of further study intervention development