HIV Treatment Bulletin

Tuberculosis diagnostic pipeline

Javid Syed

After the doldrums of 40 years in which tuberculosis (TB) research languished without much funding or focused scientific attention, the last decade has brought a flurry of activity to the field for the development of new tools. The creation of the Stop TB Partnership in 2000 led to the first (2001-–2005) and second (2006-–2015) global plans to stop tuberculosis, the formation of the New Diagnostics Working Group within the partnership, and in parallel the creation of the Foundation for Innovative New Diagnostics (FIND), a product development partnership focused on new tuberculosis (and more recently, malaria and sleeping sickness) diagnostic development. Through these entities the Stop TB Partnership has aimed to harness new scientific tools and approaches to accelerate the discovery, development, approval, and distribution of new diagnostic technologies that could make diagnosing TB more accurate, faster, and more reliable. Until recently it was expected that there would be off-the-shelf technology approaches that could be easily adapted and scaled to diagnose TB in resource-limited settings—the so-called low-hanging fruit.

Sensitivity and specificity

Sensitivity is the ability of a diagnostic test to accurately identify the condition it is attempting to diagnose. The lower the sensitivity of the test the greater the chances that people with the condition will not be accurately identified by the test; this can lead to false negative results (in which TB is present but not detected).

Specificity is the ability of a test to accurately rule out the condition the test is seeking to diagnose. The lower the specificity the greater the chances that people without a condition will be falsely diagnosed as having it (the test will detect TB, but it is not there). Therefore, low specificity will lead to a greater number of false positive results.

An ideal test will have both—high sensitivity and high specificity.

Ten years later, the low-hanging fruit has yielded a small but flavorful harvest. Many advanced technologies are now ready to launch or are in place in central reference laboratories and in some referral laboratories in tertiary (usually major city hospital) settings. Both “rapid” (2–4 week) liquid culture techniques and nucleic acid amplification tests (which detect Mycobacterium tuberculosis, or MTB, DNA sequences) are available if the proper financial and technical support is provided. However, there has as yet been no breakthrough diagnostic test to revolutionize TB diagnostics in peripheral health post or community point-of-care settings. These peripheral health post settings are the most decentralized health care facilities that have inconsistent access to running water, electricity, trained laboratory workers, or any laboratory equipment and yet provide care for the largest number of people with TB. To date, TB is most commonly diagnosed with a 125-year-old sputum-smear microscopy test despite the fact that it has low sensitivity, is unable to diagnose TB disease if the bacterial load in the sputum is low (smear-negative TB), is unable to diagnose TB that is not in the lungs (extrapulmonary TB), can pick up acid-fast bacilli that are not TB, and cannot distinguish between drug-sensitive and drug-resistant TB. FIND estimates that only about 20% of TB cases worldwide are detected with sputum-smear microscopy (World Health Organization 2006). The failure of the most common diagnostic test to pick up more than half of TB cases, and the lack of facilities, meant that in 2008 only 61% of all forms of TB (smear-negative, smear-positive, and extrapulmonary) were reported by national TB programs to the World Health Organization (World Health Organization 2009b). Where programs do not exist no one will receive care, and even when programs do exist, if they rely on sputum-smear microscopy they will miss half the cases overall and much more among children and people with HIV who have higher levels of smear-negative and extrapulmonary TB. Cultivating the TB bacillus on solid or liquid growth media, or TB culture, is still considered the gold standard for diagnosis. Though the culture method is very sensitive, it too can be nonspecific, as nontuberculous mycobacteria will also be detected in culture, and subsequent speciation tests using a lateral flow (dipstick) TB antibody test such as the Tauns test are required to definitively identify the growing organisms as MTB. Culture does detect smear-negative and drug-resistant strains, and can even detect extrapulmonary TB if the right sample is drawn, but it is still far from ideal as it takes an average of up to two weeks to become detectable with rapid liquid culture and four weeks to grow to visible levels on solid media (Dorman 2010). Culture also requires skilled laboratory staff, electricity, and biosafety infrastructure in order to be performed safely and accurately. TB culture testing is not available or accessible for most people with TB who live in low-income countries and access care at health posts.

To address these challenges, from 2007 to 2009, the Strategic and Technical Advisory Group for TB (STAG-TB), the group that advises the World Health Organization (WHO) on new policies including those for the uptake of new TB diagnostics, has approved at least seven new diagnostic approaches for use in high-TB-burden settings. These tools and strategies address some of the most pressing diagnostic challenges in TB care—the need to improve the sensitivity over sputum-smear microscopy and accelerate identification of TB and drug resistance. Despite the improvements offered by these new tools, the impact on TB control has not yet changed the situation on the ground in many places because most of the newly recommended diagnostics are not simple, robust, or cheap enough to use in the field. Increasingly, there is a recognition that though the rapid and sensitive liquid culture and nucleic acid amplification tests (NAATs) are great and urgently need to be rolled out, the real revolution in TB diagnostics will only occur when tools designed for the higher levels of health systems are complemented by an easy-to-use point-of-care diagnostic that is sensitive, specific, fast, cheap, robust, and safe (Cruciani 2004; World Health Organization 2007, 2009a, 2009b)

halving global TB incidence from 1990 levels by 2015, nor the Stop TB Partnership’s target of having six new TB drugs approved by then. There are an estimated 2 billion people infected with TB around the world, compared with approximately 33 million infected with HIV (UNAIDS 2009; World Health Organization 2009). Over the past 23 years the U.S. Food and Drug Administration (FDA) has approved 24 new compounds to treat HIV infection. During that same time, the FDA approved just one new drug—rifapentine—to treat TB from an existing class of drugs, the rifamycins.

TB is most common in poor countries and communities; despite over nine million cases each year, the vast majority of patients are unable to pay for the drugs. Because TB is an airborne infectious bacterium, curing TB disease is a public health responsibility yet governments have not allocated the necessary resources to cover the cost of TB diagnosis and treatment. A major rationale for the 50-year drought

Ten years later, the low-hanging fruit has yielded a small but flavorful harvest. Many advanced technologies are now ready to launch or are in place in central reference laboratories and in some referral laboratories in tertiary (usually major city hospital) settings. Both “rapid” (2–4 week) liquid culture techniques and nucleic acid amplification tests (which detect Mycobacterium tuberculosis, or MTB, DNA sequences) are available if the proper financial and technical support is provided. However, there has as yet been no breakthrough diagnostic test to revolutionize TB diagnostics in peripheral health post or community point-of-care settings. These peripheral health post settings are the most decentralized health care facilities that have inconsistent access to running water, electricity, trained laboratory workers, or any laboratory equipment and yet provide care for the largest number of people with TB. To date, TB is most commonly diagnosed with a 125-year-old sputum-smear microscopy test despite the fact that it has low sensitivity, is unable to diagnose TB disease if the bacterial load in the sputum is low (smear-negative TB), is unable to diagnose TB that is not in the lungs (extrapulmonary TB), can pick up acid-fast bacilli that are not TB, and cannot distinguish between drug-sensitive and drug-resistant TB. FIND estimates that only about 20% of TB cases worldwide are detected with sputum-smear microscopy (World Health Organization 2006). The failure of the most common diagnostic test to pick up more than half of TB cases, and the lack of facilities, meant that in 2008 only 61% of all forms of TB (smear-negative, smear-positive, and extrapulmonary) were reported by national TB programs to the World Health Organization (World Health Organization 2009b). Where programs do not exist no one will receive care, and even when programs do exist, if they rely on sputum-smear microscopy they will miss half the cases overall and much more among children and people with HIV who have higher levels of smear-negative and extrapulmonary TB. Cultivating the TB bacillus on solid or liquid growth media, or TB culture, is still considered the gold standard for diagnosis. Though the culture method is very sensitive, it too can be nonspecific, as nontuberculous mycobacteria will also be detected in culture, and subsequent speciation tests using a lateral flow (dipstick) TB antibody test such as the Tauns test are required to definitively identify the growing organisms as MTB. Culture does detect smear-negative and drug-resistant strains, and can even detect extrapulmonary TB if the right sample is drawn, but it is still far from ideal as it takes an average of up to two weeks to become detectable with rapid liquid culture and four weeks to grow to visible levels on solid media (Dorman 2010). Culture also requires skilled laboratory staff, electricity, and biosafety infrastructure in order to be performed safely and accurately. TB culture testing is not available or accessible for most people with TB who live in low-income countries and access care at health posts.

To address these challenges, from 2007 to 2009, the Strategic and Technical Advisory Group for TB (STAG-TB), the group that advises the World Health Organization (WHO) on new policies including those for the uptake of new TB diagnostics, has approved at least seven new diagnostic approaches for use in high-TB-burden settings. These tools and strategies address some of the most pressing diagnostic challenges in TB care—the need to improve the sensitivity over sputum-smear microscopy and accelerate identification of TB and drug resistance. Despite the improvements offered by these new tools, the impact on TB control has not yet changed the situation on the ground in many places because most of the newly recommended diagnostics are not simple, robust, or cheap enough to use in the field. Increasingly, there is a recognition that though the rapid and sensitive liquid culture and nucleic acid amplification tests (NAATs) are great and urgently need to be rolled out, the real revolution in TB diagnostics will only occur when tools designed for the higher levels of health systems are complemented by an easy to use point-of-care diagnostic that is sensitive, specific, fast, cheap, robust, and safe (Cruciani 2004; World Health Organization 2007, 2009a, 2009b)

Table 1. Diagnostic Tests Approved by the WHO, 2007–2009*

Recommended Approach Name of Test Sponsor/Developer Technique Measures Health Systems in Which Test Is Most likely to Be Used Year of WHO Approval
Liquid culture MGIT BD Diagnostic Systems Automated liquid culture TB growth and drug-resistant TB Reference laboratories 2007
Rapid speciation test Capilia test Tauns, Standard

Diagnostics, and FIND

Lateral flow techno-

logy that uses anti-bodies to detect MTB

MTB DNA Reference laboratories 2007
Revised case definition of a sputum-positive pulmonary TB case to at least one TB bacilli in one sputum sample Special Programme for Research and Training in Tropical Diseases (TDR) Strategy to increase sensitivity of sputum-smear microscopy TB bacilli Peripheral laboratories 2007
Line-probe assays for

multi-drug-resistant (MDR) TB

INNO-Lipa

GenoType MTBDRplus

Innogenetics

HAIN

Lifescience

Line-probe assay that requires culture

Line-probe assay that can be done

on sputum

Rifampicin resistant mutation in MTB DNA

Isoniazid- and rifampicin-resistant mutation in MTB DNA

Peripheral laboratories 2008
Front-loaded sputum-

smear microscopy

TDR Strategy to prevent dropouts in the diagnostic process by reducing number of clinic visits needed for sputum-smear microscopy TB bacilli Peripheral laboratories 2009
Light-emitting diode

(LED) microscopy

LED adaptor for existing microscope

Primo Star iLED microscopy

LW Scientific

and Fraen

Carl Zeiss Inc.

and FIND

Fluorescent

microscopy

Fluorescent

microscopy

TB bacilli

TB bacilli

Peripheral laboratories 2009
Noncommercial culture and drug susceptibility testing (DST) Microscopic

observation drug suscepti-bility (MODS)

Nitrate

reductase

assay

Colorimetric

DST

Academic laboratories

Academic laboratories

Academic laboratories

Inverted light microscopy that detects TB growth

Solid culture; TB growth causes color change

Solid culture; TB growth causes color change

TB growth and drug-resistant TB

TB growth and drug-resistant TB

TB growth and drug-resistant TB

Reference laboratories 2009

*Sources: World Health Organization 2007, 2008b, 2009b

What Is in the TB Diagnostic Pipeline?

Products in the diagnostics pipeline attempt to address the most pressing challenges faced by TB control by making diagnostic tools available to the lower rungs of health care systems and by developing new tools that are more sensitive, specific, and faster than currently available tools to confirm TB infection or disease and identify drug-resistant strains. It is disappointing that there are no new tools in development currently targeted at the health post level, where the greatest number of people with suspected or active TB are seen.

After the introduction of a number of new tools and technologies in the last few years, it appears that the low-hanging fruit has nearly all been harvested. None of the diagnostic tools in this year’s pipeline are appropriate for use at the health post or the point of care. A number of tools covered in last year’s Pipeline Report, such as light emitting diode (LED) microscopes and the LED adaptor for diagnosing TB using fluorescent microscopy; the strategy to collect sputum samples on the same day to prevent attrition of TB patients during the diagnostic pathway; and the microscopic observation drug susceptibility (MODS) test to detect TB and drug resistance using an inverted light microscope were recommended by STAG-TB in 2009 for use in TB control programs.

Diagnostic tools or strategies from last year’s Pipeline Report that were dropped this year include fluorescent vital dye staining, sputum concentration strategies to improve the sensitivity of sputum-smear microscopy, the MPT-64 skin patch test that Sequella was developing for detection of latent TB infection, and the MTB DNA test that FIND is still working on with Spaxen and University College London.

This year we focus on technical approaches to TB diagnosis that are the most likely to be ready for review by the STAG-TB in the coming three years, and for which there is at least some peer-reviewed literature.

The impact of a diagnostic tool in reducing disease and death is defined not only by its sensitivity and specificity but also by the health system level at which it can be used. Therefore, we examine these tools according to where within health systems they will be deployed:

  • Health posts with or without consistent sources of water and electricity, trained laboratory workers, or laboratory equipment—these serve 60% of people in search of TB care.
  • Peripheral laboratory settings that can conduct sputum-smear microscopy, have some trained staff but limited infrastructure and biosafety systems—these serve about 25% of people needing TB services.
  • Reference laboratories, with the most trained staff, highest biosafety levels, and most reliable clean water and electricity supplies, usually capable of carrying out at least TB culture if not NAAT as well—accessible to at most 15% of people (O’Brien 2009).

Table 2. TB Diagnostic Test or Processes in the Pipeline, 2010

Name of Test or Process Sponsor/Developer Technique Measures Estimated Date of WHO Review
Peripheral Laboratories
Manual loop-mediated isothermal amplification process (LAMP) Eiken Chemical and FIND Manual nucleic acid MTB DNA 2011
Clearview Lipoarabinomannin (LAM) antigen enzyme-linked immunoabsorbent assay (ELISA) Inverness Medical Innovations ELISA to detect LAM antigen in urine MTB LAM antigen 2012
Reference Laboratories
GeneXpert MTB/RIF Cepheid, FIND and UMDNJ* Automated nucleic acid amplification test MTB DNA, rifampicin resistance sequences 2011
QuantiFERON-TB Gold Test Cellestis Interferon-gamma release assay Immune cell response to latent TB infection 2011
T-SPOT.TB Oxford Immunotec Interferon-gamma release assay Immune cell response to latent TB infection 2011

*UMDNJ-University of Medicine and Dentistry of New Jersey

The health post setting

As Treatment Action Group (TAG) has documented in the past five years, despite significant progress at the higher-tech level suitable for reference and some peripheral laboratories, investment in the basic and applied science necessary to discover and develop a true point-of-care test for TB disease is shockingly inadequate. There is no peer-reviewed information available to shed any further light on a tool that is likely to be approved by STAG-TB in the next three years for use at the heath post (Harrington 2008, Treatment Action Group 2009).

Peripheral laboratories

The Eiken LAMP Nucleic Acid Test

Sputum or other fluid samples are subjected to a loop-mediated isothermal amplification process (LAMP) to amplify and detect TB DNA for active disease diagnosis. This test is easier than previous NAAT TB tests as it does not require heating or cooling—the LAMP device can amplify DNA at a constant temperature of 65 degrees Celsius. A prototype of this LAMP test developed by Eiken Chemical was studied by FIND in Peru, Bangladesh, and Tanzania. We previously reported in 2009 that LAMP had a sensitivity of 97.7% in culture-positive, smear-positive specimens. The sensitivity was poor (48.8%) in smear-negative, culture-positive specimens (Boehme 2007). Subsequently the test has been redesigned and is currently being studied. Data from the redesigned product may be reviewed by the WHO in 2011.

Advantages: The LAMP test is easier to conduct than other NAATs due to its isothermal nature. Sample preparation and test readout takes less than two hours. In feasibility studies, laboratory technicians with no prior experience with NAAT could learn to conduct this test in about one week. No DNA contamination was observed even when the test was conducted in one room without biosafety cabinets.

Limitations: This test is manual and requires trained laboratory staff, electricity, and laboratory infrastructure, which will prevent its use below peripheral laboratory settings. The redesigned version of this test needs to be validated to address the lower sensitivity that was previously observed among smear-negative, culture-positive patients.

Clearview LAM antigen ELISA

Lipoarabinomannin (LAM) is a TB cell wall protein excreted in urine. This test, developed by Inverness Medical Innovations, is currently marketed in an enzyme-linked immunoabsorbent assay (ELISA) format as the Clearview LAM Antigen ELISA. Antibodies to LAM present in the test well bind any protein found in urine and when a reagent is added a color change indicates a positive readout (Inverness Medical Innovations 2010).

Several recent studies have examined this test in settings with high burdens of TB and HIV. One assessment of LAM’s utility was nested in a prospective study designed to determine the predictors and causes of death among hospitalized TB suspects. All enrolled patients were given HIV antibody and LAM antigen tests. Of 499 persons enrolled, 422 were HIV-positive. The LAM test was 59% sensitive and 96% specific among all patients with culture-confirmed TB. Among people with HIV the LAM test was 67% sensitive and 95% specific and the LAM test’s sensitivity was highest in people with a low number of CD4 cells. When stratified by CD4 levels, sensitivity was 55% for those with CD4 greater than 200; 14% for those between 150 and 200; 56% for those between 100 and 150; 71% for those between 50 and 100; and 85% for those lower than 50. Among the 193 confirmed TB cases in this study LAM alone was more sensitive than sputum-smear microscopy alone (32% vs. 16%), though neither sensitivity was optimal. Both LAM and sputum-smear microscopy together identified 75% of the confirmed TB cases (Shah 2009).

An earlier version of the LAM test was also studied in Harare, Zimbabwe, in those suspected of having TB and in registered TB patients who were recruited from hospital settings and tested for HIV and TB. The HIV prevalence in this study population was 77% and the TB prevalence in enrolled patients regardless of HIV status was 49%. LAM test sensitivity was 52% in HIV-positive and 21% in HIV-negative persons. Compared to the 52% sensitivity of LAM, the 60% sensitivity of the sputum smear was higher for people with TB and HIV. In this study the sensitivity of the test was not stratified by CD4 cell count (Mutetwa 2009).

A third study of LAM was conducted among South African HIV-positive adults who had not been on antiretroviral therapy (ART) and were not diagnosed with TB. Sputum and urine were tested for TB, using fluorescence microscopy and liquid culture for sputum and LAM ELISA for urine. In this study the sensitivity of the sputum smear alone was 14% and did not differ by CD4 cell count, while the LAM test was 38% sensitive. LAM sensitivity was strongly associated with lower CD4 cell counts. Among those with fewer than 50 CD4 cells, LAM was 67% sensitive, while among those with CD4 between 50 and 100 it was 35% sensitive, with just 4% sensitivity in those with CD4 counts over 100. The combined sensitivity of the sputum smear and LAM was highest at 67%, among those with fewer than 50 CD4 cells (Lawn 2009).

Inverness is developing a rapid LAM antigen test that uses the same lateral flow platform that is in its Determine test kit for HIV-1 and HIV-2. The goal is to detect LAM in unprocessed urine within 20 minutes. No additional information is yet public regarding this approach.

The LAM studies indicate that this test will only be useful in TB/HIV patients with very low CD4 counts—which may be helpful, as this population is especially likely to be sputum-smear negative or have extrapulmonary TB. If LAM testing is implemented, it will be important to define an algorithm by which to optimize its use in the diagnostic pathway. The Tuberculosis Clinical Diagnostics Research Consortium, which was established in 2009 through a seven-year grant to the Johns Hopkins University School of Medicine from the National Institute of Allergy and Infectious Diseases (NIAID), will study the LAM test to determine its feasibility.

Advantages: The LAM test provides a result in less than 3 hours’ time (compared with 14 days for rapid liquid culture or MODS). The urine test has the advantage of being noninvasive and the sample is easier to collect than sputum. It is most sensitive among HIV-positive TB patients with very low CD4 cell levels, a population at increased risk of being missed by the sputum-smear test because of greater chances of having extrapulmonary and smear-negative TB. When used with sputum-smear microscopy, the combined sensitivity of the two tests appears to provide important data to help rapidly identify patients in urgent need of TB treatment.

Limitations: The sensitivity of the test in people who are not HIV–positive or with CD4 counts above 100–200 is quite poor. The current test is in an ELISA format and is not appropriate for use in lower rungs of the health care system.

The reference laboratory setting

GeneXpert MTB/RIF

Cepheid is developing the GeneXpert MTB/RIF test in partnership with FIND and the University of Medicine and Dentistry of New Jersey. This closed-system NAAT was initially developed to detect anthrax for the U.S. Postal Service and is now being modified to diagnose TB bacilli and rifampicin-resistant strains. The test cartridge, into which sputum or another fluid sample is administered, contains all the reagents, does not require much sample processing, and is able to detect rifampicin-resistant TB in less than two hours.

Initial results of the test reported in 2009 were very promising, showing its sensitivity for TB detection to be 99.1% in smear-positive and culture-positive specimens and 80% in smear-negative and culture-positive specimens. The high degree of sensitivity of the test, especially in smear-negative and culture-positive specimens, requires that the test is run three times on each sample. Unpublished data presented by FIND at the South African TB Conference in June 2010 showed that when each sample is only tested once the test’s sensitivity is 67.2% in smear-negative and culture-positive specimens while the sensitivity in smear-positive cases remains very high at 99%. The specificity of the test was 95.7% in initial studies conducted in Latvia and Peru; its ability to detect rifampicin resistance had 100% sensitivity and 96.7% specificity (O’Brien 2009, Roscigno 2010).

Recent data corroborate the initial data and show the test’s high degree of sensitivity and specificity. In this newer study the test was able to detect all 23 commonly occurring rifampicin-resistant strain sequences. The study was conducted in Vietnam in 107 sputum samples of persons suspected of having tuberculosis and demonstrated that the GeneXpert was able to detect all 29 of the smear-positive and culture-positive cases—84.6% of the smear-negative, culture-positive samples that were identified through growth on solid media and 71.7% of smear-negative culture-positive cases grown on both solid and liquid media. The test identified all 25 of the culture-negative samples. The test was able to detect 98.4% of the 55 culture-positive cases among retreatment cases in Uganda and 100% of the 9 rifampicin-resistant cases. In addition to the test performance, the buffer in the test was shown to reduce the viability of the tuberculosis significantly, thereby reducing biohazard concerns (Helb 2010).

Advantages: The GeneXpert test addresses some of the biggest challenges in TB diagnostics. It is highly sensitive and specific in smear-positive TB cases as well as in detecting rifampicin-resistant strains. Its sensitivity in smear-negative cases is moderate, at 67.2%, when the test is done only once, and increases to 80% when the test is repeated three times on these specimens. The test requires minimal training of laboratory workers and provides results within two hours. As it is a closed-system test it does not require laboratories with high levels of biosafety and has low contamination concerns.

Limitations: The test requires a consistent source of electricity that will limit its use outside of settings where a regular power supply can be guaranteed. Currently the cost per test cartridge is over $20 and the instrument cost is a whopping $25,000. These costs will pose significant barriers for its uptake, though it may become cheaper if widely used due to economies of scale.

Immune-based tests for latent TB infection

Interferon-Gamma Release Assays

Interferon-gamma release assays (IGRAs) expose a blood sample to TB antigens that are specific for MTB and not found in the Bacille Calmette-Guérin (BCG) vaccine or in most environmental mycobacteria. The production of interferon-gamma (IFN-gamma), a protein produced by a primed immune system when it recognizes an antigen it has been previously exposed to, indicates TB infection. The three IGRAs that are available on the market are QUANTIFeron Gold test (QFT-G) and QUANTIFeron Gold in a Tube test (QFT-GIT), both produced by Cellestis, and T-SPOT.TB from Oxford Immunotec.

BCG vaccination does not cause false positives in the IGRA, as is often the case with the tuberculin skin test (TST). IGRAs are hoped to be more specific than TSTs in detecting TB infection.

IGRAs have been studied for a variety of diagnostic needs—from predicting risk of progression from latent to active disease to monitoring TB treatment response and TB diagnosis among the immunocompromised. A review article has concluded that the IGRAs were less sensitive in diagnosing latent TB infection (LTBI) where TB burdens were high (69%) compared with settings with low burdens of TB (83%). The pooled specificity of IGRAs was between 93 and 99%. Studies comparing TST and IGRA sensitivity in diagnosing TB infection at lower CD4 cell counts found that IGRAs—and especially T-SPOT.TB—were less prone to false negatives due to immune suppression, but that the IGRAs were also affected by low CD4 cell counts. One recent study showed that at CD4 cell counts lower than 100, IGRAs had high rates of indeterminate results. Other studies have shown that IGRAs are not useful in diagnosing TB disease or monitoring TB treatment success. Their utility in predicting risk of progression to active TB disease is unclear (Aabye 2009, Dheda 2009, Hoffman 2010, Lienhardt 2010, Pai 2010).

Some limited cost-effectiveness data suggest that IGRAs are best used after an initial TST to rule out false positive results. On the other hand, it might be cheaper to do a symptom screen for active TB, test for HIV, and administer isoniazid preventive therapy to those who are HIV-positive and without TB symptoms. Putting in two cumbersome and expensive tests would further complicate the diagnostic labyrinth that people in high-TB-burden settings must routinely navigate to get proper treatment. The cost-effectiveness of IGRAs is affected by LTBI prevalence and thus needs to be adjusted to the context in which the tests are being conducted (Pooran 2010).

The WHO convened an expert committee in July 2010 to assess available data and develop recommendations on whether to use IGRAs in detecting TB disease or LTBI. The IGRAs will also be discussed at the WHO STAG-TB meeting in 2010.

The QuantiFERON Gold Test and QuantiFERON Gold in a Tube

Cellestis has developed two versions of the QuantiFERON test (QFT), which has been approved by the U.S. Centers for Disease Control and Prevention for use in place of the TST.

These blood tests require the sample to be processed within 16 hours. The sample is incubated with the TB antigens for 16 to 24 hours and an ELISA measures the presence of IFN-gamma. The QFT-GIT offers an additional benefit over the QFT-G by already containing the TB antigen in the tube and is therefore easier to use (Cellestis 2010).

A meta-analysis has shown that the sensitivity of these tests in detecting TB infection was not very high, at 78% for QFT-G and 70% for QFT-GIT, though the specificity of both of the QFTs was 99% in non-BCG-vaccinated persons and 96% in persons who have been vaccinated with BCG (Pai 2008).

Advantages: QFTs are more specific than TSTs in identifying TB infection and provide results in 24 hours, rather than the 72 hours needed for TST and neither QFT-G nor QFT-GIT require a person to come back to the health center to have the test read. However, the 24 hours needed to get a test result do require some system through which the result can be communicated to those diagnosed with TB infection and lead to appropriate follow-up.

Limitations: The tests requires samples to be processed within 16 hours on an ELISA format, which prevents them from being used in health facilities at levels below reference laboratories. The tests have moderate sensitivity (70% for QFT-GIT and 78% for QFT-G) when compared to the T-SPOT.TB test (90%). Test accuracy is reduced at lower CD4 cell levels and the time to result is still too long at 24 hours (Pai 2008).

The T-SPOT.TB Test

This IGRA also exposes the sample to TB antigens and measures the presence of IFN-gamma to detect TB infection. Unlike QFT sample preparation, the T-SPOT.TB test requires centrifugation to extract mononuclear cells and these cells must be counted before the test is run. The sample preparation and test must be run within eight hours of collecting the sample (Oxford Immunotec 2010).

Advantages: There are some data that suggest that the T-SPOT.TB test is more sensitive than the QFTs and is also less likely than the QFTs or TST to be affected by CD4 cell levels. The T-SPOT.TB test does not react to prior BCG vaccination or most other environmental mycobacteria (Dheda 2009, Pai 2008).

Limitations: In the best case, the test will not be decentralized beyond the peripheral laboratory. Compared to the QFTs, the T-SPOT.TB test requires a greater degree of sample processing that will need more laboratory tools and greater level of skill among laboratory staff. The blood sample has to be processed within eight hours of being collected and this presents additional logistical challenges.

Policy and research recommendations

Develop clear policy recommendations and build laboratory capacity for rollout, uptake, and impact assessment of new diagnostic TB tools.

The ultimate goal of introducing new diagnostics is to reduce the burden of disease. Though at least seven new tools have been recommended since 2007, there has not been a significant uptake of these tools in national TB programs. This is likely due to a combination of factors that range from the general conservativeness of the TB community to embrace new tools, to the more legitimate need for laboratory capacity strengthening and clearer guidance that recommends specific tools within diagnostic algorithms appropriate for epidemiologic contexts. For instance, cost-effectiveness data that can show improvements in clinical outcomes will help national TB programs decide on the combination of tools most appropriate for their country’s TB epidemiology (Pai 2010).

Recent discussions at a Stop TB Partnership meeting to develop an operational research agenda for the Global Plan to Stop TB 2006–2015 discussed the need to address these research questions to ensure greater uptake of new tools. The NIAID-funded Tuberculosis Clinical Diagnostics Research Consortium will also be addressing this current gap by conducting accuracy and feasibility studies for diagnostic tools that are in late-stage development for which a proof of concept already exists but has not yet been tested extensively in clinical settings (Federal Business Opportunities 2010). The UNITAID-funded $87.5 million project Expanding Access to New Diagnostics for TB (EXPAND-TB) also aims to strengthen laboratory capacity in 27 countries by 2013. EXPAND-TB is a collaboration of the Global Laboratory Initiative, FIND, the WHO, and the Stop TB Partnership’s Global Drug Facility. These efforts are likely to ensure that the incremental improvements offered by new TB diagnostic tools are implemented to their greatest potential to reduce the burden of disease and death caused by TB.

Clarify not only which tools work but also which ones do not work.

A 2006 report by FIND and the WHO’s Special Programme for Research and Training in Tropical Diseases (TDR) estimated that the world annually spent $1 billion on TB diagnostics. In the absence of strong regulatory oversight of TB diagnostics, the WHO needs to provide clear guidance clarifying not only which tools are effective but also which ones are not useful to ensure funds are not wasted on unvalidated diagnostic tests. As part of this effort, the TDR conducted a systematic review of 19 serological tests in 2008 to examine their utility in diagnosing TB. The review showed that the serological tests were much less specific and sensitive than the sputum-smear test and should not be used for diagnosis. The WHO is convening an expert committee in July 2010 to examine all data available regarding the utility of serological tests to diagnose TB, and STAG-TB will likely provide clarification of what role these tests will or won’t play in TB control efforts (World Health Organization 2006, 2008a).

Accelerate targeted development of a point-of-care assay for all forms of TB disease.

New TB diagnostic development has opportunistically taken advantage of tools that were already in late-stage development. Many of the tools recommended by WHO put existing tools into public health facilities of countries with high-TB burden where they were not previously utilized or in lower rungs of the health system. However, the result has been that tools that fill certain critical gaps are only practical at district and referral laboratory settings. There is an increasing recognition that these tools at the higher levels of health systems will only contribute to a revolution in TB diagnostics if they are complemented by a point-of-care (POC) TB diagnostic, such as a dipstick, appropriate for use at the health post setting. However, a clear project-driven plan that is focused on a POC dipstick that has clearly defined minimum specifications has not been articulated. Such a product specification needs to drive a coordinated funding and scientific effort to meet this most urgent gap in TB diagnostics.

To develop such a product specification, TAG, M裩cines Sans Fronti籥s (MSF), and Partners in Health convened a group of experts in March 2009 that included basic researchers, product developers, laboratory technicians, activists focused on improving TB care, and TB program implementers. This group developed the minimum specifications required for a TB POC dipstick (Treatment Action Group 2009).

While discussing the TB POC specifications, a number of barriers to developing such a tool were identified. These included the need for specimen banks that have samples relevant for the development of a TB POC test from well-characterized patient populations as well as coordination and collaboration between research efforts to identify antibodies and antigens to assess a combination appropriate for a POC test. Though a number of specimen banks currently exist, such as those organized by FIND and the TDR, it was not clear whether they were sufficient and accessible. At the same time, though FIND has been conducting a systematic search to validate biomarkers for a TB POC test, a number of basic science researchers think that the antigens and antibodies appropriate for a good first-line POC tool are already known but that efforts between researchers has not been coordinated well enough to identify an appropriate combination.

TAG has been working with partners to bring clarity to the above issues and define a way forward, calling for the NIAID to put out a request for information and ideas or convene a meeting to gather the best information and ideas available for antigens, antibodies, and diagnostic test platforms that could lead to a TB POC test. This open, public, and transparent process would allow for a clear assessment of what is currently known and what is needed to push the field forward—whether it is funding, the need for more basic science, a combination of both, or an entirely new approach.

To help create a well-informed advocacy effort for a TB POC test, TAG, MSF and the TB/HIV Working Group of the Stop TB Partnership are working with researchers based at the Imperial College in London to conduct an independent assessment of what is known about antigens, antibodies, and the technology platforms available and to clarify what infrastructure hurdles need to be addressed to develop a TB POC diagnostic tool. A report from this assessment to be completed in mid-2010 will inform TAG and our partner’s future advocacy for a TB POC diagnostic test.

Significantly expand research capacity.

Currently, research being conducted in TB drugs and vaccines does not sufficiently include diagnostic components. Such studies can provide important data on the utility of new TB diagnostic tools. Specimens from the study cohorts, if organized in a well-characterized sample bank, can also be a critical resource for biomarker discovery research. If diagnostic tools that already have proof of concept are incorporated into these studies, valuable data on specificity, sensitivity, and patient-relevant data could be collected that would help inform the WHO’s policy making. The sample bank created from the study population can also be used to identify and validate biomarkers that can predict treatment outcomes. Biomarkers that can reliably predict treatment outcome can significantly lower cost and time required for TB drug trials by reducing the time and effort required to perform follow up on study participants for at least 12 months to ensure that the regimen being studied is at least as effective as the existing standard of treatment.

To expand the research capacity for TB, TAG has urged NIAID to expand the mandate of its HIV clinical trial networks and sites to include TB-focused research. In 2009, NIAID announced its intent to expand the focus of the AIDS Adult Clinical Trials Group (ACTG) and the HIV Vaccine Trials Network to include TB and other diseases of importance for public health. TAG and other research activists have responded to the NIAID call for information highlighting the need for research to develop a POC tool for TB.

In June 2010 the National Institutes of Health, the Federal Drug Administration, and the Centers for Disease Control and Prevention, three of the leading U.S. government agencies involved in TB research and programs, organized a meeting to address the need for new TB diagnostics as well as the need to harvest samples from TB drug and vaccine trials to create a well-characterized sample bank that could be used to identify and validate biomarkers that can predict treatment outcomes. Though this initiative is in its inception, TAG commends the leadership of these three U.S. agencies coming together to address the gap in TB diagnostics—a gap that hampers not only TB programs but research. The creation of an accessible sample bank from a well-characterized cohort of study participants will greatly assist the efforts of diagnostic developers whose research priorities are in line with global TB control efforts.

Increase funding for TB diagnostics research.

In 2008 the world spent $49.7 million on TB diagnostics research and development, which is 10% of the total $491 million invested in TB research (Treatment Action Group 2010). TAG has tracked the resources invested in TB research and development since 2005. We have consistently advocated for an investment of $2 billion per year in TB research overall—including not only investment in new tools but also in basic science and operational research—in order to reach the targets set out in the Global Plan to Stop TB 2006–2015. Starting in 2006, TAG called attention to the initial inadequacy of the Global Plan’s research estimates, as they did not sufficiently account for basic science or operational research and were neither evidence-based nor sufficiently ambitious. The Stop TB Partnership is currently in the process of revising the research components of the Global Plan to incorporate basic science, operational research, and evidence-based budgeting for the discovery and development of new tools. This first evidence-based and comprehensive TB research plan will be completed by the end of 2010. TAG will continue to track the resources available for funding and advocate for research funders to better coordinate their portfolios to ensure that there is adequate, growing, and coordinated support for a comprehensive research agenda toward the elimination of tuberculosis.

Conclusion

The significant efforts of those working in TB diagnostics—most notably the Stop TB Partnership’s New Diagnostics Working Group, FIND, the TDR, and a number of academic laboratories and for-profit companies—have yielded a slew of new tools that have addressed important gaps in the current armamentarium. However, most of these tools have only addressed gaps at the reference and highly skilled peripheral lab levels and have not yet been translated into any significant measurable or reported improvement in TB case detection or treatment outcomes. This is in large part because the tools being developed are not appropriate for peripheral health post settings in low-income countries where most people with TB access care.

In the past decade, most of the efforts to develop diagnostic tools for TB have attempted to take advantage of recent scientific breakthroughs or to modify existing tools to better serve TB care and control efforts. Yet there is a need to move from this opportunistic strategy to a focused strategy driven by end-user-defined product specifications. Without such a focused and coordinated effort supported by researchers, funders, TB program leadership, and activist groups, there is a great danger that we will soon have consumed all the low-hanging fruit without successfully fertilizing the soil for the emergence of a revolutionary new crop of TB diagnostics to prevent, through rapid point of care diagnosis, the nearly 9 million new TB new cases and 2 million TB deaths that occur each year .

References

Aabye MG, Ravn P, PrayGod G, et al. The impact of HIV infection and CD4 cell count on the performance of an interferon gamma release assay in patients with pulmonary tuberculosis. PLoS One 2009;4(1):e4220.

Boehme CC, Nabeta P, Henostroza G, et al. Operational feasibility of using loop-mediated isothermal amplification for diagnosis of pulmonary tuberculosis in microscopy centers of developing countries. Journal of Clinical Microbiology 2007;45(6):1936–40.

Cellestis. QuantiFERON-TB Gold (In-Tube Method). Package insert. Retrieved 19 May 2010 from http://www.cellestis.com/IRM/Content/pdf/QuantiFeron%20US%20VerG_Jan2010%20NO%20TRIMS.pdf.

Cruciani M, Scarparo C, Malena M, et al. Meta-analysis of BACTEC MGIT 960 and BACTEC 460 TB, with or without solid media, for detection of mycobacteria. Journal of Clinical Microbiology 2004;42(5):2321–25.

Dorman, SE. New diagnostic tests for tuberculosis: Bench, bedside, and beyond. Clinical Infectious Diseases 2010;50(Suppl. 3):S173–77.

Dheda K, Van Zyl Smit R, Badri M, et al. T-cell interferon-gamma release assays for the rapid immunodiagnosis of tuberculosis: Clinical utility in high-burden vs. low-burden settings. Current Opinion in Pulmonary Medicine 2009;15(3):188–200.

Dinnes J, Deeks J, Kunst H, et al. A systematic review of rapid diagnostic tests for the detection of tuberculosis infection. Health Technology Assessment 2007;11(3):1–196.

Federal Business Opportunities. Tuberculosis Clinical Diagnostics Research Consortium. Retrieved 19 May 2010 from https://www.fbo.gov/index?s=opportunity&mode=form&id=65d23f889fa315169a530d3b1141ed09&tab=core&_cview=1.

Harrington, M. Creating Political Will and Scientific Momentum to Develop a Point-of-Care Test for TB. Paper presented at the Fourteenth TB/HIV Core Group Meeting of the Stop TB Partnership, 12 November 2008, Geneva.

Jones M, Story E, et al. Rapid detection of Mycobacterium tuberculosis and rifampin resistance by use of on-demand, near-patient technology. Journal of Clinical Microbiology 2010;48(1):229–37.

Hoffmann M, Ravn P. The use of interferon-gamma release assays in HIV-positive individuals. European Infectious Disease 2010;4(1):23–29.

Inverness Medical Innovations. Clearview TB ELISA LAM Specific Direct Urinary Antigen Test. Package insert. Retrieved 19 May 2010 from http://www.clearview.com/pdf/EN%20Clearview%20TB%20Elisa%20Booklet%20.pdf.

Lawn SD, Edwards DJ, Kranzer K, et al. Urine lipoarabinomannan assay for tuberculosis screening before antiretroviral therapy diagnostic yield and association with immune reconstitution disease. AIDS 2009;23(14):1875–80.

Lienhardt C, Fielding K, Hane AA, et al. Evaluation of the prognostic value of IFN-gamma release assay and tuberculin skin test in household contacts of infectious tuberculosis cases in Senegal. Plos One 2010;5(5):e10508.

Mutetwa R, Boehme C, Dimairo M, et al. Diagnostic accuracy of commercial urinary lipoarabinomannan detection in African tuberculosis suspects and patients. International Journal of Tuberculosis and Lung Disease 2009;13(10):1253–59.

O’Brien R. Progress in the development of new TB diagnostic tools—What is in the pipeline? Paper presented at the Fourth Scientific Symposium on the Occasion of World Tuberculosis Day, 22–23 March 2009, Berlin.

Oxford Immunotec. T-SPOT.TB. Package insert. Retrieved 19 May 2010 from http://www.oxfordimmunotec.com/8-UK.

Pai M, Zwerling A, Menzies D. Systematic review: T-cell-based assays for the diagnosis of latent tuberculosis infection: An update. Annals of Internal Medicine 2008;149(3):177–84.

Pai M, Minion J, Steingart K, et al. New and improved tuberculosis diagnostics: Evidence, policy, practice, and impact. Current Opinion in Pulmonary Medicine 2010;16(3):271–84.

Pooran A, Booth H, Miller RF, et al. Different screening strategies (single or dual) for the diagnosis of suspected latent tuberculosis: A cost effectiveness analysis. BMC Pulmonary Medicine 2010;10:7. Retrieved 14 May 2010 from http://www.biomedcentral.com/1471-2466/10/7.

Roscigno, G. Current and Future Development in TB Diagnostics. Paper presented at the Second South African TB Conference, 2 June 2010, Durban.

Shah M, Variava E, Holmes CB, et al. Diagnostic accuracy of a urine lipoarabinomannan test for tuberculosis in hospitalized patients in a high HIV prevalence setting. Journal of Acquired Immune Deficiency Syndrome 2009;52(2):145–51.

Stop TB Partnership, The Global Plan to Stop TB 2001–2005. Geneva, Switzerland: Stop TB Partnership, 2001.

Treatment Action Group. TAG 2009 Pipeline Report. New York: Treatment Action Group, 2009.

Treatment Action Group. 2009 Report on Tuberculosis Research Funding Trends, 2005–2008, 2nd ed. New York: Treatment Action Group, 2010.

World Health Organization. Diagnostics for tuberculosis: Global demand and market potential. Geneva, Switzerland: World Health Organization, 2006.

World Health Organization. Seventh meeting Strategic and Technical Advisory Group for Tuberculosis: Report on conclusions and recommendations. Geneva, Switzerland: World Health Organization, 2007.

World Health Organization (2008a). Laboratory-based evaluation of 19 commercially available rapid diagnostic tests for tuberculosis. Geneva, Switzerland: World Health Organization, 2008.

World Health Organization (2008b). Strategic and Technical Advisory Group for Tuberculosis: Report of the eighth meeting. Geneva, Switzerland: World Health Organization, 2008.

World Health Organization (2009a). Global tuberculosis control: A short update to the 2009 report. Geneva, Switzerland: World Health Organization, 2009.

World Health Organization (2009b). Strategic and Technical Advisory Group for Tuberculosis (STAG-TB): Report of the ninth meeting. Geneva, Switzerland: World Health Organization, 2009.