According to the latest data published by the Spanish Epidemiological Vigilance network (Red de Vigilancia Epidemiológica), in mid-July 2009, 3340 new cases of TB had been notified.2

According to provisional data published by this national center, in 2009, 6070 cases of TB were registered. However, these numbers should be taken with a grain of salt as, despite the fact that TB is a disease requiring mandatory notification, it is estimated that at least one-third of cases are not reported.

The data that are available about the natural history of TB suggest that in the first two years after the infection by Mycobacterium tuberculosis, between 5% and 10% of infected individuals develop tuberculosis disease.3 With a proper immune response by the infected individual, the bacillus can remain inactive for decades or even for a life-time. Consequently, the diagnosis and treatment of tuberculosis infection will be more effective if they are aimed at individuals at greater risk for the progression from infection to tuberculosis disease, including recently infected individuals and immunosuppressed patients.4

Individuals infected by M. tuberculosis react to PPD with a delayed hypersensitivity response mediated by cells (especially T lymphocytes), and after 48–72h an induration appears in the area of the injection. This hypersensitive response remains for life, although it may be reduced in the elderly, as well as in certain clinical alterations. Repeated PPD in a non-sensitized individual does not alone trigger the immune response.

The Mantoux technique5 entails the intradermal injection of 2 units of PPD RT-23 tuberculin (0.1ml) with a 27-gauge needle on the inside of the forearm in an area where there are no skin lesions. In order for the technique to be correct, a papule measuring 6–10mm in diameter should be produced at the time of the injection. This is the most common PPD method.

The results are read 72h after the injection by measuring the cross-sectional diameter of the induration following the longitudinal axis of the forearm. The measurement is taken in millimeters. In cases where there is no induration and instead only erythema, it is interpreted as 0mm.

According to the Spanish Society of Pulmonology and Thoracic Surgery (SEPAR), the following indurations are considered positive6:

• -

  In non-vaccinated persons, ≥5mm.

• -

  In persons vaccinated with BCG, this poses the problem of discerning whether a tuberculin induration is due to a tuberculosis infection or instead to a response to the antigens that are shared with the BCG vaccine (M. bovis BCG). In this situation, certain clinical conditions are taken into account, considering the PPD positive with a diameter ≥5mm if, in addition to being vaccinated, the subjects either live with or maintain frequent contact with bacilliferous patients, present chest radiography with lesions suggesting former TB that had never been treated, are infected by HIV or are patients with pneumoconiosis.

• -

  In the rest of subjects vaccinated with BCG, the results are determined to be an infection and not a secondary reaction to the vaccine if the size of the induration is >15mm. In general, the vaccine reaction does not usually cause indurations of more than 14mm, and it is considered that the greater their size and the longer the time passed since the vaccination, the more likely it is due to a tuberculous infection and not a vaccine reaction (the studies in this direction indicate that the interference of BCG on tuberculin reaction can be limited some 10–15 years after vaccination).4

The greater the probability of being infected or of developing the disease (for instance, in cases of recent contact or those infected with HIV), less should the vaccination history influence the interpretation of the test. In patients infected by HIV, a PPD <5mm does not exclude the diagnosis of infection as it could be due to the situation of anergy caused by the compromised immune response.

Tuberculin reactions with vesicles or necrosis in the inoculation area are also considered indicative of tuberculosis infection, regardless of the size of the induration or the vaccination history.

In cases of contact studies, the interpretation is rather simplified, as the vaccination history should not be contemplated and an induration ≥5mm is considered indicative of tuberculosis infection.

Table 1 summarizes the situations that can cause false positives of the test.

There are certain circumstances depending on the individual that may trigger false negatives with PPD, such as a concurrent viral infection, vaccinations with live viruses, immunosuppressive situations or treatment with drugs that diminish the immune response.

Table 2 summarizes the situations that could lead to false negatives of the test.

In some individuals, basically in older people infected years beforehand or in persons vaccinated in childhood, an initial PPD may be negative and a repeat test 7–10 days later may show positive. This “booster” phenomenon implicates that the first PPD can have a memory effect in the immune system, which, after a second test, produces the immune response. The definitive result of the test is considered that of the second reading.

So-called tuberculin conversion is the situation in which an individual with PPD considered negative is then later positive. This represents the acquisition of a recent tuberculous infection and is defined as “converter”. Functionally, it is defined as an individual who goes from having a tuberculin test of <5mm to having one of ≥5mm, with a difference of at least 5mm in less than 2 years.

PPD, as all diagnostic tests, should only be used in persons in whom a possible therapeutic intervention could be based on the results.3 In TB, there are only two possibilities for therapeutic intervention: treatment of the patients and chemoprophylaxis, or preventive treatment of infected patients with a high risk for developing TB.

Table 3 shows the indications for PPD according to SEPAR.6

Interferon gamma is an important molecule for the control of tuberculosis infection, and its participation is essential in the immune response that protects against said microorganism. This cytokine, produced by the CD4+, CD8+, and NK T lymphocytes, activates the infected macrophages, with the consequent release of IL-1 and TNF-α, limiting the growth and the multiplication of the mycobacteria. The individuals with deficiencies in the receptors or in the genes that codify the synthesis of this molecule are more susceptible to having mycobacterial infections with greater frequency and with greater severity.

There are two commercial techniques for the in vitro diagnosis of tuberculosis infection: QuantiFERON-TB-Gold In Tube (Cellestis®, Victoria, Australia)8 and T-SPOT.TB (Oxford Immunotec®, Oxford, United Kingdom).9 The first generation of QuantiFERON-TB, approved by the Food and Drug Administration (FDA) in the year 2001, detected the release of interferon gamma in response to PPD. In 2004, the FDA approved the second generation of this diagnostic test, known as QuantiFERON-TB Gold, which, unlike the first generation did not use PPD as mycobacterial antigens, but instead used synthetic peptides that simulate more specific antigens, such as Early Secreted Antigenic Target-6 (ESAT-6) and Culture Filtrate Protein-10 (CFP-10). These two molecules are codified by the RD-1 region of the genome of M. tuberculosis and significantly increase the specificity compared with PPD. These antigens are absent in M. bovis BCG and in the majority of the non-tuberculous mycobacteria (except Mycobacterium kansasii, Mycobacterium marinum, Mycobacterium szulgai) (Table 4).

Currently, the third generation of this test is on the market, known as QuantiFERON-TB Gold In Tube (QFT-GIT), which incorporates a third mycobacterial antigen: TB 7.7, and tubes specifically designed for taking blood samples for this test (Table 5).

• 1.

  QFT-GIT: In this technique, 3 specific tubes are used that come with the reagent kit (one of the tubes includes the ESAT-6, CFP-10, and TB 7.7 specific tuberculous antigens, called the problem tube; another contains phytohemagglutinin, which is the positive control tube; and the third contains no reagents, which is the negative control tube). A total of 3ml of blood is required (1ml per tube) and the blood is extracted directly into the tubes themselves. Later, after having shaken the tubes, they are incubated for 18–22h at 37°C, after which time the tubes are centrifuged and the plasma obtained is used for the enzyme immunoassay that detects and quantifies the interferon gamma released by the lymphocytes of the patient. This step can be done either manually or totally automatically. After incubation, the plasma may be stored for several weeks without affecting the results, which, if necessary, would facilitate the organization of the work load of the laboratory. The technique uses specific software for the emission of the results.

• 2.

  T-SPOT.TB: In order to carry out this technique, 8–10ml of heparinized blood is used. In the laboratory, and following the indications of the manufacturer, the mononuclear layer is separated, in which, after washing and later cell count, the number of cells will be adjusted to a quantity of 250000cells/ml. This quantity will be used as inoculum in the 4 wells that are included in this test (2 wells with the ESAT-6 and CFP-10 antigens, and the other 2 for the positive and negative controls). The plate is incubated for 18–22h at 37°C in a CO2 incubator, after which time the immunospot is done, allowing for the quantification of the number of interferon producing cells (observed as spots, each spot representing the footprint of an individual T-lymphocyte interferon secretor). An interpretation algorithm developed by the manufacturer facilitates the emission of the results. Technically, T-SPOT.TB requires more blood, more preparation time and is more painstaking than QFT-GIT. Furthermore, it does not allow for working with the samples in a differed manner. The guidelines recommended by the manufacturers of these tests for their interpretation are shown in Tables 5 and 6.8,9

  Table 6.

  Interpretation Criteria for QuantiFERON-TB Gold In Tube (QFT-GIT) Test.

  Interpretation	Null Controla	Response to TBb	Mitogenc
  Positived	≤8.0	≥0.35IU/ml and ≥25% of the null control	Any
  Negativee	≤8.0	<0.35IU/ml or <25% of the null control	≥0.5
  Indeterminatef	≤8.0	<0.35IU/ml or <25% of the null control	<0.5
  	>8.0	Any	Any

  Based on Cellestis Limited. QuantiFERON-TB Gold©.

  a

  Concentration of interferon gamma (IFN-γ) in plasma incubated without antigens.

  b

  Concentration of IFN-γ in plasma stimulated with a group of peptides represented by ESAT-6, CFP-10, and TB 7.7, except the null control.

  c

  Concentration of IFN-γ in blood plasma stimulated with mitogen, except the null control.

  d

  The interpretation indicates that the infection by Mycobacterium tuberculosis is probable.

  e

  The interpretation indicates that the infection by M. tuberculosis is not probable.

  f

  The interpretation indicates an uncertain probability of the infection by M. tuberculosis.

The new in vitro diagnostic techniques for tuberculosis infection offer important advantages over PPD: they do not present interferences with the BCG vaccine, the subjectivity of the interpretation is avoided, an appointment for reading the results is avoided and they incorporate a positive control that provides valuable information for interpreting an apparently negative test as either a true negative or indeterminate as a result of technical errors or due to immunosuppression.

With the lack of a true reference test for the diagnosis of tuberculous infection, it is difficult to establish the sensitivity and the specificity of these new diagnostic techniques.

In order to resolve the problem of sensitivity, three strategies have been used: (1) evaluate the patients who have active TB and therefore should be infected; (2) evaluate the individuals who have been in contact with tuberculous patients and stratify them according to degree of exposure; (3) analyze the agreement between the IGRA and PPD determination tests.10,11

An important facet of the transmission of TB is that the risk for infection is mainly determined by the frequency, duration and the proximity of the contact with the person diagnosed with tuberculosis. Therefore, for a new technique to be considered more sensitive and more specific than PPD, it should be more closely related with the level of exposure and should not take into regard the state of vaccination (BCG). This theory has been used to compare the degree of precision of the IGRA determination techniques with PPD in contact studies,11,12 which have concluded that these new techniques correlate equally well or even better than PPD, regardless of BCG.

In one of the largest studies published to date that included 535 subjects,12 the results of QTF-GIT and T-SPOT.TB were not interfered with by the history of vaccination with BCG, although this did occur with the tuberculin test, which demonstrates the greater specificity of the former. In addition, IGRA also correlated better with the exposure to TB disease.13–18 In one of the studies carried out in exposed children,18 a dose-response relationship was found between the bacilli load in the sputum and the positivity of the IGRA. Those who had been in contact with the patients with a greater bacilli load in sputum more frequently had positive IGRA and tuberculin tests.19–22

Inflammatory bowel disease (IBD) includes a heterogeneous group of disorders that run their course with chronic inflammation of the gastrointestinal mucosa. Up to one-third of patients with IBD have a severe form of the disease that requires the use of immunomodulatory drugs or biological agents for the control of the inappropriate response of the immune system in these patients. Tumor necrosis factor alpha (TNF-α) is a proinflammatory cytokine that plays an important role in the pathogenesis of IBD. Infliximab (IFX), an IgG1 monoclonal anti-body that specifically binds with TNF-α, has been demonstrated to be effective for inducing and maintaining remission, both in Crohn's disease as well as in ulcerous colitis.

In 2001, the FDA reported on 70 cases of TB out of the 147000 patients treated with IFX throughout the world, both for IBD as well as for rheumatoid arthritis. Two-thirds of the cases were found in Europe, and Spain was the European country with the greatest number of affected patients.32 The majority had an extrapulmonary location, with disseminated disease in 24% of cases. This type of tuberculosis disease is associated with immunosuppression, which suggests that TNF-α plays an important role in the response of the host against TB, which includes the formation of granulomas.33 Given that the majority of the cases appeared after starting treatment with IFX, that the prevalence of TB in their countries was low and that they had not recently been exposed to patients diagnosed with TB, they concluded that TB was caused by the reactivation of an LTI.34 These findings uphold the recommendation for adopting measures to rule out the existence of LTI, previous to the establishment of treatment with anti-TNF-α antibodies. In our country, the Spanish Group for Crohn's Disease and Ulcerous Colitis (Grupo Español para la Enfermedad de Crohn y la Colitis Ulcerosa—GETECCU) published their own recommendations35 in 2003, which were revised in 2006,36 according to which patients diagnosed with LTI should receive treatment with isoniazid before initiating treatment with biological drugs. The results obtained to date show an important reduction in TB cases within this population.37–39

Patients with chronic inflammatory rheumatic disease have a high risk for TB, which has increased even more after the incorporation of biological therapy. In most cases, the disease is caused as a consequence of the reactivation of a latent infection; therefore, its incidence varies considerably depending on the prevalence of infection in the area studied.

The practice of systematic screening, both for infection as well as for tuberculosis disease prior to the start of biological therapy, has led to a lower incidence of TB in this population. The experience of the Spanish cohort of patients treated with biological agents (BIOBADASER) of the Spanish Rheumatology Society shows a reduction in TB cases since the implementation of the official recommendations in 2002 when compared with previous years (117 vs 522 cases/100000 persons-year, respectively). However, we must also highlight an important increase in the use of etanercept (human protein made up of receptor p75 of the tumor necrosis factor and the Fc portion of the human IgG1) during the second period of the study, in detriment to IFX, which could explain part of the reduction in incidence. Even when assuming the effectiveness of these measures, the incidence of TB in that patient collective continues to be almost five times higher than that of the general Spanish population. Several causes could explain this fact: inadequate application of the protocols, lack of compliance with the treatment for the tuberculosis infection, exogenous reinfection, and the limitations of PPD for diagnosing latent infection in the clinical context of these patients.

The experience with IGRA in the diagnosis of LTI in the disease mediated by inflammatory mechanisms is still rather limited and comes from small-scale cross-sectional studies centered on evaluating the agreement between PPD and IGRAs, without correlating the results with the risk factors for tuberculosis infection. They conclude that the agreement between PPD and IGRA is low and that the disagreeing results (PPD+, IGRA−) are due to previous vaccination with BCG.37–41

As for the correlation between IGRA and TB risk factors, in a recent study of 142 patients with disease mediated by inflammatory mechanism,37 the IGRAs were more closely related with risk factors for LTI than tuberculin. Moreover, the positivity of PPD correlated with the previous vaccination with BCG, not true for IGRA. Martin et al.39 compared the two IGRA tests in patients with rheumatoid arthritis and they observed that both QTF-GIT as well as T-SPOT.TB correlated with TB risk factors.

Based on current evidence, it can be concluded that, in patients with inflammatory disease mediated by immune mechanisms who follow an immunosuppressant treatment, the specificity of IGRA is greater than PPD.

Patients with chronic renal failure (CRF) who require hemodialysis or peritoneal dialysis are an example of a population that characteristically manifests cutaneous anergy to PPD antigens and that has a high risk for developing active TB42–45. This risk is approximately 10–25 times greater for the reactivation of an LTI in comparison with the general population.46–48 What is more, hemodialysis units are places where TB could be disseminated more easily.49

It is known that CRF is associated with numerous alterations of the immune system, the majority related with the alteration in cell immunity.50,51 Among them are the reduced proliferative response of the lymphocytes, the deficiency of interleukin-2, peripheral B lymphocyte deficiency and increased cell apoptosis.52–54 A recent study of 203 patients with CRF undergoing hemodialysis55 compared three diagnostic methods for TB infection (PPD, T-SPOT.TB, and expert evaluation). PPD showed a very low sensitivity and only presented positivity in one out of every five patients. T-SPOT.TB was positive in approximately three out of every four patients with risk factors for LTI. The sensitivity, backed by the panel of experts who confirmed the LTI cases, was nearly 75%.

Regarding peritoneal dialysis, the study by Palomar et al.56 states that these new techniques measure a type of immune response different from that involved in the delayed hypersensitivity response to PPD, and, unlike what happens in the contact study, in immunodepressed patients recent tuberculous infection is as important as remote infection. The study concludes that IGRAs complement PPD, as performing both simultaneously increases the diagnostic probability of TB.

For a person with positive PPD, the risk for developing active TB is estimated at 5%–10%.57 However, there are very few longitudinal studies that allow us to make conclusions about the capability of IGRAs for predicting the risk of developing active TB.

In a study carried out in Germany with the participation of 601 contacts close to people with positive smears and positive cultures for M. tuberculosis, QFT-GIT performed better in the prediction of active TB58 than PPD, using a cut-point of 5mm. Five (2.3%) of the 219 contacts with indurations ≥5mm on PPD developed TB, while six (14.6%) of the 41 contacts with positive results for QFT-GIT developed the disease (P=.003). However, an unusually large proportion (59%) of the contacts had an induration (PPD) between 5 and 9mm. The proportion of those considered positive due to PPD with a cut-point of 10mm who developed active TB (5 out of 90 [5.6%]) was similar to the positive proportion of QFT-GIT (6 out of 41 [14.6%], P=.1). In addition, only 2 of the 6 contacts with positive QFT-GIT results that developed active TB had the diagnosis confirmed by culture. In another study, the sensitivity for predicting later active TB was not significantly different for the two tests.59

Another study of 339 immigrants in the Netherlands gave the results that PPD and QFT-GIT have a similar value for predicting active TB.60 There was a two-year follow-up of the contacts whose PPD was ≥5mm between 0 and 3 months after the diagnosis of the index patients. Nine (3.1%) out of 288 contacts with PPD ≥10mm developed active TB, compared with 7 (3.8%) out of 184 with PPD ≥15mm, 5 (2.8%) out of 178 with a positive QFT-GIT result and 6 (3.3%) out of 181 with a positive T-SPOT.TB developed the disease. The sensitivity for detecting the development of active TB over the course of the follow-up period was 100% for PPD with a cut-point of 10mm, 88% for a PPD with a cut-point of 15mm, 63% for QFT-GIT and 75% for T-SPOT.TB. Despite the fact that PPD with a cut-point of 10mm identified the greatest number of contacts who developed active TB (9 out of 9 [100%]), and QFT-GIT identified the least number of contacts that developed active TB (5 out of 9 [63%]), the sensitivities of the two tests were not statistically different (P=.08).

In conclusion, IGRAs, for the prediction of the development of tuberculous disease, do not seem to provide important differences compared to PPD.

Until now, many studies have been based on the study of contacts in TB. Initially, these were based on PPD, but since the incorporation of IGRA, these new techniques have been the topic of research on numerous occasions related with the contact studies.18,61–66 In two of these studies,18,66 the greater the recent exposure (greater duration of the exposure or greater number of acid-fast bacilli in the source sputum samples) was more associated with the positive results for IGRA than with PPD, which suggests that IGRAs could be better than PPD in detecting recent infection. In these studies, people with less exposure were more likely to be positive to PPD than to IGRA, which suggests that PPD could have been better than detecting the infection by means of IGRA for the detection of old/remote infection that was produced before (and therefore was not caused as a result of) recent exposure.63

In another research study,61 the proximity of the recent exposure (meaning the same room, another room, another dwelling) is more associated with the results of PPD than with the QFT-GIT results.

In the case of IGRA with regard to the study of contacts, we can conclude that although there are suggestions that their positivity could indicate recent infection, this is not proven and more studies will be needed to confirm this.

The studies carried out until now about IGRA did not do serial controls of said tests. In the study by Hill et al., which was performed by using the serial IGRA test in a series of contacts exposed to persons diagnosed with TB in a country with high prevalence (Gambia), it was observed that IGRA—more specifically ELISPOT—occasionally present conversions and reversions that occurred after the exposure to M. tuberculosis.67–71

In order to interpret the result of serial, we should first answer a series of questions:

• 1.

  What is the reproducibility of the response of the T cells in a specific timeframe? (variations in the inter-individual response).

• 2.

  What do we consider reversion and what cut-point should we use to define it?

• 3.

  What is the clinical significance and the prognosis of a reversion?

• 4.

  What is a conversion and what cut-point should be used to define it? How can it be distinguished from the non-specific variations of the T cell response over time?

• 5.

  What is the prognosis of an IGRA conversion? Do individuals with clear conversions (e.g. a large increase in the interferon gamma response over time) have a greater risk for the progression to active disease than the individuals with weaker conversions or negative results?

Unfortunately, none of the studies published to date provide evidence on the cited questions, which seems to demonstrate that conversions and reversions are produced when the IGRA tests are serial and likewise when PPD are serial.68–71

These serial studies demonstrate at the same time that IGRAs are very dynamic tests and that the response of the T-cells, especially the weakly positive responses, tend to fluctuate over time, even in absence of specific treatment.

Although the data that are currently available are limited, they suggest that the positive results vary more than the negative. This is in part expectable because the positive results can vary in both directions, while the negative results can vary between 0 and the diagnostic cut-point. In short, these new diagnostic methods can be inherently prone towards conversions and reversions, and it still has not been determined whether this characteristic dynamism is an influence when the results are evaluated.

On the other hand, many of these serial tests have been carried out in countries with high incidences of TB. It is not clear whether similar findings will be found in countries where the exposure to TB is less frequent.

In general, the reversions are less frequent when the response to interferon gamma is strong.

In contrast, IGRA reversions are more frequent when both tests disagree (for example, positive IGRA, negative PPD). The discordant results are almost always weakly positive and are usually just above the cut-point.

Why do IGRA reversions occur? Some may reflect healing of the tuberculosis infection (spontaneously or secondary to treatment). Other reversions are caused by biological variations between individuals positive to IGRA, and also can be produced due to the variability in the procedures carried out in the laboratory.

In daily practice, the individuals who had had a positive result to IGRA will not undergo the test again, as neither will those individuals positive for PPD.

In that case, are reversions important outside the setting of research studies? Probably not, until we have a better understanding of the significance and the prognosis of said reversions.

Despite that fact that a high rate of IGRA conversions have been documented in high-risk populations with a high incidence of TB, there is no agreement in how to define the conversions.

With the data that are available, we cannot answer the following questions:

What degree of increase in interferon gamma indicates a new infection? How much of this variation is due to the test itself or to the biological variability? Should the same cut-point be used to define LTI and conversions? Some studies show that if a “negative-to-positive” cut point is used to define conversions, these can be greater with IGRA than with PPD, which can indicate greater sensitivity to the conversions (not necessarily for the diagnosis of LTI). However, a part of these may be due to non-specific variations around the cut-point. None of the completed studies take into consideration said variations.

The risk for developing active TB has been established in various cohort studies using PPD as a reference test. Likewise, thanks to controlled clinical assays, we know that initiating prophylactic treatment in persons with positive PPD reduced the risk for active disease.72

Unfortunately, there are no equivalent data for IGRA. The data that we have available are limited to a small study73 that was carried out between the family contacts of index cases and obtained as a result an association between persons with a strong response of interferon gamma to stimulation with ESAT-6 and later progression to active TB. Despite this, the prognosis of a positive result to IGRA has still yet to be determined.

What is the prognosis of an IGRA conversion? The conversion implies recent infection (incident), and the prognosis is different than the other infection that already existed previously (prevalent). On the other hand, the prognosis of a “strong conversion” can be very different than that of a “weak conversion”. There are no data that respond directly to the question of the prognosis of the conversions, but up-and-coming data suggest that responses to both ESAT-6 and CFP-10 antigens closely correlate with in vivo replication and with the progression from infection to disease. Recently, Andersen et al.74 have suggested a hypothesis by which high and/or increasing interferon gamma levels produced in response to ESAT-6 by the T cells in recently infected persons can be a sign of incipient disease, and therefore can serve as a prognostic marker of the later development of the clinical disease in the near future.

The study of contacts of patients with TB is recommended as a means for detecting infected persons who may later develop the disease. It has been demonstrated that the treatment of infected persons, fundamentally with isoniazid, diminishes the development of future TB cases.75 The effectiveness and cost-effectiveness of these programs are very influenced by the precision when identifying the individuals infected with risk for developing tuberculosis disease.76

The guidelines referring to the use of IGRA depend on each country. Thus, for example, the Centers for Diseases Control (CDC) in the United States recommend substituting PPD with IGRA in all cases. Contrarily, in the United Kingdom, the National Institute of Health and Clinical Excellence (NICE) recommends the use of IGRA in combination with PPD, but only in the cases in which the tuberculin was positive.77

A recently published study78 comparing the two tests currently available (QTF-GIT and T-SPOT.TB) concludes that, for the study of contacts, joint PPD/IGRA strategies are more economical than the use of either T-SPOT.TB–QTF-GIT or PPD alone.