Robust, reliable and resilient: designing molecular tuberculosis tests for microscopy centers in developing countries

Keywords::

• microscopy centers
• molecular diagnostics
• point-of-care testing
• test development
• tuberculosis

Nucleic acid amplification tests (NAATs) for the diagnosis of tuberculosis (TB) have been in clinical use for almost two decades. Yet, for the most part, because of high cost and stringent infrastructure requirements, their use has been limited to reference laboratories in high-income countries. Widely held as the biggest recent advance in TB diagnosis, the Xpert® MTB/RIF assay (Cepheid Inc., Sunnyvale, CA, USA) entered the market in 2009, and was endorsed by the WHO in 2010 [1]. Since then, over 3 million test cartridges have been used worldwide, with South Africa alone accounting for over half of this volume [101].

Unlike conventional PCR-based assays, the Xpert MTB/RIF assay is an automated, cartridge-based nested PCR designed for use in district and subdistrict level laboratories [2]. With no requirement for manual DNA extraction, clean rooms or stringent quality assurance for prevention of amplicon contamination, this technology is the first TB NAAT that can be deployed outside of the high-level or reference laboratory [3]. These features, combined with high accuracy [4], strong WHO policy endorsement [1], and subsidized instrument and cartridge prices for high-burden countries [102], has made Xpert MTB/RIF the first NAAT that is actively being scaled up in high-burden countries for TB diagnosis as well as rapid detection of drug resistance [1,101].

Although Xpert MTB/RIF is a potentially game-changing technology, it has its limitations. First, the implementation of this test at the district level still means limited access to many patients who may not reach that tier of the healthcare system [5]. Even if they do, access may occur after several weeks or months of diagnostic delay, during which TB transmission may occur. Furthermore, since the most important goal of point-of-care (POC) testing is to make a treatment decision in the same clinical encounter or visit [6], Xpert MTB/RIF may be less helpful for decentralized POC testing programs at lower levels of the healthcare system (e.g., primary care) where patients with TB symptoms initially seek care. Another important concern with the implementation of the Xpert MTB/RIF technology is its high overall cost for national TB programs in low-income countries [7], especially for decentralized deployment of the test [8].

A more affordable molecular POC test at the primary care level will greatly help to reduce diagnostic delays and curb TB transmission in the community [9]. Inspired by the success of Xpert MTB/RIF, a variety of newer, ‘fast-follower’ NAATs are now entering the market and claim to be ‘POC NAATs’, that is, intended for more decentralized settings than the Xpert MTB/RIF assay [10,11]. If they are designed for POC use, then, at the very least, they should be deployable in peripheral microscopy centers where currently the majority of initial TB testing takes place by use of sputum smear microscopy. These are usually primary health centers with attached small laboratories with technicians trained to do microscopy, and often staffed by physicians or nurses who can initiate TB treatment. Thus, these health centers are a level higher than health posts or outpatient clinics that have no attached laboratories and are unlikely to be staffed by physicians.

National TB programs in high-burden countries are heavily reliant on microscopy centers and thousands of such centers have been established for smear microscopy TB testing (Figure 1). For example, India alone has over 13,000 designated microscopy centers in the public sector, where over 15 million sputum smears are stained and read every year by trained microscopy technicians [103]. These microscopy centers are linked to decentralized treatment units where first-line TB drugs are available and TB treatment can be initiated and monitored by community-based directly observed therapy providers.

Figure 1. Examples of peripheral TB microscopy centers in India (B & C), Uganda (A) and Kenya (D).

Because microscopy centers are usually embedded in or attached to primary health centers, they are closer to patients than district or subdistrict level hospitals and laboratories. This, in turn, suggests that TB can be diagnosed earlier at the microscopy center level. Therefore, these centers should be the ideal place for the implementation of a novel molecular assay that is more sensitive than smear microscopy and has a faster turnaround time to support more rapid initiation of anti-TB therapy in an already established infrastructure.

The big issue then is: are the so-called POC NAATs for TB deployable in peripheral microscopy centers? Are they designed for such settings? For example, can they survive the high temperature and frequent power outages that are likely in such settings? Will manual sample processing and DNA extraction prove to be too big a hurdle for basic laboratories?

To summarize the current state of peripheral microscopy centers, we recently conducted a survey of microscopy centers in 22 highest TB burden countries [12]. We surveyed multiple respondents from each country and asked them to complete a simple questionnaire, keeping in mind a typical, peripheral microscopy center. The results of the survey are summarized in Figure 2, which highlights scarcity of infrastructure (e.g., temperature control, uninterrupted power), lack of basic equipment (e.g., biosafety hood, centrifuge) and limited skills at the level of peripheral microscopy centers in all high-burden countries, although BRICS (i.e., Brazil, Russia, India, China and South Africa) countries fare better than the others.

Figure 2. Characteristics of peripheral microscopy centers in 22 high tuberculosis burden countries. Questions related to environmental conditions (is temperature or humidity not a concern?); infrastructure (is stable power supply, clean water supply present?); presence of equipment (are N95 respirator, micropipettes, refrigerator, incubator, centrifuge, hot water bath or biosafety hood present?) and skills (Are staff able to operate a micropipette or computer or perform a PCR test?); and the presence of means of communication (is landline, mobile network or internet present?). Additional questions asked about whether QA measures were established and which smear methods were currently used. Countries are sorted by increasing purchasing power parity. BRICS countries are Brazil, Russia, India, China and South Africa.

On the positive side, our survey showed that all high-burden countries have successfully established direct Ziehl–Neelsen microscopy with external quality assurance. Also, mobile phones seem to be widely available, opening the possibility of mobile health interventions to recall patients with positive results, notifying cases to TB control programs and being of use for supply chain management and quality assurance [13].

While the Xpert MTB/RIF technology was clearly not designed for the kind of laboratories shown in Figures 1 & 2, it is unclear whether fast-follower NAATs can actually be implemented in such settings either. For example, all the so-called POC NAATs, besides Xpert MTB/RIF, that are currently on the market still require manual sample processing and DNA extraction [11], and our survey suggests that this is likely to be challenging in most peripheral microscopy centers.

Based on our survey and the ground realities of microscopy centers in high-burden countries, we suggest a list of criteria (Box 1) that are important to ensure for sputum-based NAATs to be successfully implemented in POC testing programs at the level of microscopy centers, and to ensure same-day initiation of anti-TB therapy. Product developers should consider incorporating these criteria into their target product profiles (TPPs) for POC NAATs under development [14,104].

As newer TB assays are developed, product developers need guidance on the biggest unmet needs that should be addressed, the potential market for products that meet these needs and the TPP attributes that are critical for success in meeting the needs [14,15,104].

Box 1. Critical requirements for any sputum-based NAAT that is intended for point-of-care TB testing use at the level of peripheral microscopy centers in high-burden countries.

Accuracy

• • The assay should be more sensitive than sputum smear microscopy and ideally at least as sensitive as the Xpert MTB/RIF assay for the detection of pulmonary TB.

• • It should be at least as specific as smear microscopy and Xpert MTB/RIF for the detection of pulmonary TB.

• • The added ability to detect drug resistance is desirable and can be an add-on (reflex) test, if it is not integrated into the initial detection cartridge.

• • Turnaround time should allow for same-day treatment initiation (<1 h is preferable)

Operational aspects

• • It should not be more complex than smear microscopy.

• • Basic laboratory technicians with minimal training should be able to run the NAAT.

• • Manual, precision steps (especially for sample processing) should be kept to a minimum (i.e., similar to the Xpert MTB/RIF assay).

• • It should not require expensive or sensitive equipment.

• • It should not rely on cold chain or additional equipment (e.g., centrifuge, refrigerator, biosafety hood).

• • It should be possible to perform with reasonable throughput (minimally 15 tests per day; 30–40 tests per day may be ideal).

• • It should be able to handle multiple samples, preferably asynchronous and allow for walk-away operation.

Cost

• • It may be more expensive than smear microscopy, but should be cheaper than Xpert MTB/RIF.

Environment

• • It should be able to function at a high temperature (e.g., 40°C) and high humidity (e.g., 75%).

• • It should not require continuous power and be able to run on battery backup.

• • It should be implementable in a setting without separate clean rooms.

Biosafety

• • Since biosafety hoods are unlikely to be present, the NAAT should incorporate a mechanism to rapidly decontaminate sputum, which allows for subsequent work on the bench.

Quality assurance

• • It should have internal process controls and it should be easy to set up an external quality assurance system (e.g., testing of blinded panels).

• • Maintenance of the instrument should be inexpensive, easily doable at the field level and calibration should be feasible remotely or through swap-out of modules.

Training requirements

• • As with microscopy and Xpert MTB/RIF, periodic, short duration training should be sufficient to implement the NAAT.

Information and communication technology

• • It should allow for data export over mobile phone networks and thus allow for remote monitoring and direct notification of cases.

In conclusion, we advocate a robust, reliable, rapid, sputum-based NAAT/molecular test that is more sensitive than smear microscopy (and, preferably, as accurate as Xpert MTB/RIF), more affordable than Xpert MTB/RIF and can be successfully deployed as a same-day test and treat program at the level of peripheral microscopy centers in high TB burden countries. Based on our previous work, we have suggested a list of desirable attributes that will ensure the success of such a NAAT that can survive in resource-constrained, rough conditions, and potentially replace smear microscopy at the POC. In the longer term, we also believe that laboratory capacity should be strengthened in developing countries. This will require greater investments in healthcare infrastructure and human resources.

Financial & competing interests disclosure

M Pai serves as a consultant to the Bill & Melinda Gates Foundation (BMGF), but the views expressed in this article are his own. He also receives grant funding (grant OPP1061487) from BMGF for projects to assess the TB diagnostics market size and develop TTPs for POC testing. C Denkinger is supported by a Richard Tomlinson Fellowship at McGill University and a fellowship of the Burroughs – Wellcome Fund from the American Society of Tropical Medicine and Hygiene. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

No writing assistance was utilized in the production of this manuscript.

Notes

NAAT: Nucleic acid amplification test; TB: Tuberculosis.