Diagnostic Tools Used Routinely for Diagnosis of Active TB

Acid fast staining of clinical material, followed by smear microscopy, remains the most frequently used microbiological test for detection of TB. The major drawback of sputum smear microscopy is its poor sensitivity, estimated to be ∼70% in a recent systematic review [2]. However, the sensitivity of sputum smear microscopy is clearly less in many field settings and may be as low as ∼35% in some settings with high rates of TB and HIV coinfection [3]. Compounding the limitation of poor test sensitivity is inadequate or absent test quality assurance in some resource-constrained settings, further driving down the yield of microscopy, driving up the laboratory workload as more sputum tests per patient are performed in an effort to reach a diagnosis, and increasing diagnostic delay and patient loss to follow-up. Drug susceptibility status cannot be ascertained from smear microscopy.

The HIV pandemic has brought into focus the inadequacy, from individual and public health perspectives, of sputum smear microscopy as the cornerstone of TB diagnosis in low- and middle-income settings. HIV infection dramatically increases the incidence, severity, and mortality risk of active TB [4, 5, 6]. At the public health level, failure to diagnose disease in a large proportion of HIV-infected patients with smear-negative TB may contribute to transmission [7] and further stresses health care and/or personal resources, because individuals remain in the health care system without correct diagnosis and treatment or exit the system and probably die.

Culture of M. tuberculosis in clinical specimens is substantially more sensitive than smear microscopy. Culture can be performed using solid media, such as Lowenstein-Jensen, or liquid media, such as that used in commercially available automated systems. Until the recent advent of molecular tests for drug resistance (described in the next section), isolation of M. tuberculosis with use of culture was a prerequisite for subsequent phenotypic drug-susceptibility testing. The Achilles heel of culture is the long time to results (10–14 days for liquid culture and 3–4 weeks for solid culture), which is a consequence of the long doubling time of M. tuberculosis. Currently available culture methods are technically demanding, require implementation of biosafety practices and equipment to prevent inadvertent infection of laboratory personnel, and have relatively high per-test prices. The Global Tuberculosis Report 2008 documents the stunning lack of culture facilities in the government health sector in most developing countries, some of which have a single laboratory that may or may not be well resourced or quality assured [8].

Tuberculin skin testing using purified protein derivative and chest radiographs are used as adjuncts to smear microscopy (and culture, if available) in some settings; however, the former have poor sensitivity and specificity for active TB, and the latter are often not available at the point of primary patient care. Trials of antibiotics directed against common bacterial pneumonia pathogens are often recommended in TB program diagnostic algorithms but are also fraught with problems and may lead to lengthy diagnostic delay.

New Diagnostic Technologies and Tools

There is a clear need for development, introduction, and effective implementation of cost-effective new tools that contribute to improvement in patient-centered outcomes and public health and that perform well for HIV-infected and HIV-uninfected individuals. The Stop TB Partnership Working Group on New TB Diagnostics has placed priority on accurate, simple new tools for TB case detection, rapid identification of drug-resistant TB, and reliable detection of latent TB infection [9].

Against a backdrop of increased need fueled by HIV infection and drug-resistant TB, advances in biology (including the solution of the M. tuberculosis genome in 1998 [10]), and emergence of public-private partnerships involved in global health, engagement in TB diagnostics development and evaluation has increased among public health and academic groups, government funding agencies, and importantly, the industry sector. The result has been an expansion in the number of promising diagnostic tests under development, including 2 new tests (in addition to liquid culture) that have been endorsed for use by the World Health Organization. Table 1 lists some of the more promising new technologies and tests currently in demonstration or late-stage evaluation phase [11, 12].

Beyond New Technology

Successful implementation of new tools will depend on more than technological innovation (Figure 1). At the research level, rigorous implementation of well-designed, bias-minimized studies and complete and accurate reporting are essential for appropriate decision making by the health care community charged with implementing tests for individual patient evaluation or recommending tests for TB program use. The Standards for Reporting of Diagnostic Accuracy Initiative has developed standards and tools for improving the quality of reporting of diagnostic accuracy studies, to best allow the reader to detect the potential for bias in a study and to assess the generalizability and applicability of study results [13]. Assessment of test impact on relevant clinical outcomes should become a routine component of late-stage evaluation research. Diagnostic accuracy is arguably just a surrogate for patientand public health-oriented outcomes. Economic assessments are important during the continuum of diagnostics research. During device development, cost estimates can guide the need for device modifications to facilitate use in settings with the greatest need. Later, during the evaluation and demonstration phase, cost-effectiveness and cost-benefit analyses can provide information critical to policy development and implementation. Operations and health systems research is also needed to understand how to effectively implement new tools in relevant settings where existing access to and delivery of health care are weak.

New programmatic approaches, including revised clinical algorithms for TB diagnosis, may be needed to maximize the impact of new tools. For example, should rapid molecular tests for drug resistance be performed for all persons with suspected TB during initial evaluation, be reserved for use in the initial evaluation only of persons with suspected TB with risk factors for drug resistance, or be used in some other place in a diagnostic algorithm? In populations with a high prevalence of HIV infection, should urine-based antigen detection tests be used solely for evaluation of symptomatic persons with suspected TB, or should they also play a role in routine screening of HIV-infected persons [14]? To date, most TB diagnostic test development has focused on maximizing sensitivity and specificity to rule in or confirm a TB diagnosis. On the other hand, a test with an exceedingly high negative predictive value might have use in ruling out TB and, thereby, allowing efficient triage of patients and resources; such a test would require careful assessment to determine its optimal use in clinical algorithms.

Laboratory capacity needs to be strengthened, especially in resource-limited settings. Although some aspects of laboratory strengthening will vary according to the characteristics of the new tests, there are, nevertheless, general unmet needs, including those for training at technologist and management levels, retention of trained personnel, enhancement of quality-assurance systems, enhancement of results-reporting mechanisms, and reliable mechanisms for instrument maintenance and supply procurement. Strengthening of HIV (and in some instances TB) laboratory capacity under the US President's Emergency Plan for AIDS Relief and related programs serves as a useful model, as does the successful collaborative programundertaken by the Foundation for Innovative New Diagnostics, Partners in Health, the World Health Organization, and the Lesotho Ministry of Health and Social Welfare to strengthen the National Reference Laboratory of Lesotho [12].

A number of important barriers exist with respect to the full engagement of industry and investors in diagnostics development. Barriers include uncertainty about the size and/or accessibility of the TB diagnostics market, especially in developing countries; complex regulatory processes; unfavorable intellectual property rights protections; and in some instances, lack of knowledge about the types of tests that are most needed and likely to be relevant. A recent analysis indicates that, worldwide, >US $1 billion is spent annually on TB diagnostics [15]. Of interest, approximately three-quarters of all diagnostic tests are performed outside established market economies; however, this testing burden accounts for only one-third (∼US $326 million) of the current market, because the most common tests performed are sputum smear microscopy and chest radiograph, which have relatively low per-test costs [15]. In countries with established market economies, tuberculin skin testing is the most frequently used TB test, in accordance with the relatively low rates of TB and relatively high prioritization of detection and treatment of latent TB in those settings. The blood-based interferon-γ release assays, including the QuantiFERON-TB Gold tests (Cellestis) and T-SPOT.TB (Oxford Immunotec), can detect TB with a higher degree of specificity than can the tuberculin skin test and are now approved for use in the United States and a number of other countries. To date, in the United States, use of these tests as replacements for or adjuncts to the tuberculin skin test has not been widespread, but momentum appears to be growing.

Funding for TB diagnostics research by the top 40 TB research funding institutions in 2007 was estimated at US $41.9 million, less than the amount for TB drugs (US $170 million) and TB vaccines (US $71.2 million) and more than the amount for TB operational research (US $36.8 million) [16]. This level of funding falls woefully short of the Global Plan to Stop TB's recommendations of at least US $900 million per year for research and development of new tools for TB diagnosis, treatment, and prevention [17]. Support for technical assistance to national TB programs as they implement and monitor new tools cannot be underestimated. Funding estimates aside, it is clear that important advances in TB diagnosis have recently been made, and potentially useful new tools are emerging; continued and augmented investment will be required to successfully implement the most promising of these tools in the settings where they are most needed and to maintain a robust pipeline that will ultimately yield the tools that revolutionize TB diagnosis.

Tools in the Pipeline: Transformative or Incremental Gains?

Most of the tools in demonstration or late-stage evaluation are sputum based and, thus, are likely to result—at best—in incremental gains in TB case detection. Their yield is expected to be suboptimal for patients with TB who have only extrapulmonary TB, who have respiratory disease in which a relatively large burden of organisms is not in communication with the airways, and who cannot provide a respiratory specimen for testing. Nevertheless, effective implementation might, over time, have a substantial impact on TB control through detection of a very high proportion of individuals with capacity to transmit infection to others (provided diagnosis is sufficiently prompt and treatment is available). Highly accurate, simple to perform, point-of-care tests amenable to testing of readily available clinical specimens, such as urine or blood, and with ability to detect and predict active TB anywhere in the body, in combination with effective preventive and treatment strategies (as described in other articles in this Supplement), are needed. Truly transformative change will require more than a perfect sputum test, but a really good sputum test would be a step in the right direction.

Conclusions

What's new in TB diagnostics? A lot, but not enough. The future is brighter as several promising new tools enter the demonstration and late evaluation stages. But the need is great, and important barriers remain in translating technical advances into meaningful and sustainable improvements in individual and public health in settings hardest hit by TB.