According to current estimates, a third of the world population is a carrier of latent Mycobacterium tuberculosis infection,1 which can become reactivated, particularly in immunocompromised patients. The estimated incidence of latent tuberculosis infection (LTI) in Spain in patients with psoriasis treated with anti-tumor necrosis factor (TNF) agents is 29%.2 In the best case scenario, only 10% of those infected will develop active tuberculosis (TB) during their lifetimes.3,4 However, there are factors that may significantly increase this percentage and so latent infection is considered a major obstacle for worldwide elimination of TB. Public health measures and, ultimately, clinical management of TB infection could be improved if we had a better knowledge of latent and reactivated states of M tuberculosis.

On the other hand, in the past 10 years, development of biological therapies as led to a major change in the treatment of some chronic inflammatory skin diseases, and psoriasis in particular, but also hidradenitis suppurativa after the recent approval of adalimumab in that indication. In 1998, the Food and Drug Administration approved the use of infliximab in patients with psoriasis who are refractory to conventional immunomodulatory treatment.5 Since then, more than 20 new drugs have become available for the treatment of immune-mediated inflammatory diseases (IMIDs) in which TNF and its receptors play a key role in acute and chronic inflammatory responses.6

TNF is important for immune response, and so drugs that target this molecule may increase the risk of infections and TB reactivation.

Postmarketing surveillance of the first biological agents to be approved (infliximab and etanercept) soon revealed the appearance of associated cases of TB.7 In Spain, since February 2000, a surveillance program with active data acquisition in the Biological Products Database of the Spanish Rheumatology Society (BIOBADASER)8 has been implemented and in October 2008, the Spanish registry of systemic psoriasis treatments (BIOBADADERM) was set up as part of the pharmacovigilance strategy.9

Evidence from reported cases showed a strong association between infliximab therapy and increased risk of active TB. This soon forced guidelines and recommendations to be established such as the one published in 2003 by the Spanish Work Group for Crohn Disease and Ulcerative Colitis (GETECCU),10 for the prevention of TB in candidates for infliximab treatment.

Diagnosis of LTI and preventive treatment with isoniazid (INH) for 9 months has been established as a means to reduce the probability of progression to active TB,11,12 and it has been shown that implementation of LTI screening protocols in candidates for anti-TNF therapy has reduced the incidence of TB by more than 78%.13 However, new cases have continued to arise, even after preventive treatment with INH,14 and this has forced the protocols to be updated and attempts to improve the sensitivity and specificity of the diagnostic tests and treatments in this population, with a high risk of progression to active TB.15

The availability of abundant and recent information on new biological therapies and the lack of guidelines agreed among different Spanish scientific societies was the rationale behind the need for a consensus document, published in 2016. This document aimed to collate existing information and recommendations on the basis of evidence available and consensus of a group of experts.16 The objective of the present document is to update and extend the information in the earlier document and adapt it to dermatology patients.

The document has been drafted by designated experts from the participating societies (Spanish Academy of Dermatology and Venereology, Spanish Society of Pneumology and Thoracic Surgery, Spanish Society of Digestive Pathology, Spanish Society of Rheumatology, and Spanish Society of Infectious Diseases and Clinical Microbiology) who spend a large part of their time on the study and control of candidates for biological therapies and who have particular experience in this field. The most recent editions of Spanish and international guidelines for biological diseases, diagnosis and treatment of LTI, and databases of Medline and Cochrane up until July 2017 were consulted. With the information obtained, recommendations were established based on the classification of the Infectious Diseases Society of America,17 as well as the strength of the recommendation18 (Table 1).

Currently, the biological agents with an indication approved in Spain for psoriasis are infliximab, etanercept, adalimumab, ustekinumab, secukinumab, and ixekizumab.19–31 Adalimumab has also recently been approved for the treatment of hidrosadenitis suppurativa.32 There are no other biological agents approved for diseases specifically of the skin. Table 2 summarizes the main indications for each biological agent and the date of approval in Spain.19–24

Currently, new therapeutic options are emerging for the treatment of psoriasis such as brodalumab33 (IL17 receptor antagonist); guselkumab,34 tildrakizumab35 and risankizumab36 (anti-IL23); and tofacitinib37 (Janus kinase 1 and 3 inhibitor). The biological agents currently available have already been shown to be effective in diseases other than psoriasis (sarcoidosis, lipoid necrobiosis, granuloma annulare, pityriasis rubra pilaris), and some of these may be approved in these indications in the near future.

Corticosteroid therapy in systemic diseases and immunosuppressant protocols to prevent graft rejection in transplantation have demonstrated in the past that TB is a real problem for patients treated with immunosuppressants38 (Table 3). The development of active TB in this population is a clear cause for concern: diagnosis may be delayed due to an atypical presentation,7,39 patients may present with adverse effects due to the underlying diseases or drug interactions, and the potential for TB transmission to the community should not be forgotten, including infection of other immunosuppressed patients who receive similar medical care in shared treatment areas.

TNF is a potent modulator of early inflammatory response to a large variety of physical, environmental, infectious, and immunological stimuli. Specifically, TNF is a key part of the body's defense against M tuberculosis and, along with interferon (IFN) γ and other related cytokines, it plays an important role in the development and maintenance of granulomas as structures that compartmentalize and contain the tuberculosis bacillus during infection40,41 (Fig. 1). Inhibition of TNF and its regulatory network is the biological basis that appears to explain the 4- to 5-fold increase, on average, in TB incidence in patients with TB infection who receive infliximab.8 Cases of TB have also been reported with other lower affinity TNF inhibitors, such as etanercept, a fusion protein that binds to the soluble portion of the TNF receptor, and adalimumab, a monoclonal antibody like infliximab, but fully human.42 Only exceptionally, and without a clear causal relationship, has ustekinumab, a p40 protein inhibitor of IL12 and IL23, been related with possible reactivation of latent TB.43,44 IL17 inhibition, in contrast, does not appear to reactivate LTI according to in vitro models.45–47

Figure 1.

Graphical representation of the extensive involvement of TNF in immune response to M tuberculosis infection.

(0.17MB).

It is believed that granulomas are the structures that contribute to control of the intramacrophagic bacilli, as they provide a local environment in which antigen presenting cells (APCs) and lymphocytes interact to suppress growth and proliferation.7 A large part of the work for determining the TB reactivation mechanisms has focused on the role of TNF in the formation and maintenance of these structures. Bruns et al.,48 however, described an additional mechanism to explain the susceptibility to TB progression in individuals treated with infliximab. In cells from patients with a positive tuberculin test, they identified a subset of effector T cells, denominated CD8+ effector memory T cells (TEMRA), which are memory cells rich in granulysin, a protein with antimycobacterial properties, and they showed that TEMRA are selectively depleted by the effect of infliximab in patients with rheumatoid arthritis (Fig. 2).

Figure 2.

In absence of anti-TNF, cytotoxic effector memory T cells (TEMRA) release granulomas with cytolytic enzymes (perforin, granulysin) that destroy macrophages infected by M tuberculosis as well as intra- and extracellular bacilli. In presence of anti-TNF, transmembrane TNF of TEMRA is bound to this antibody, favoring lymphocyte apoptosis. Depletion of cytotoxic T cells therefore prevents growth control and allows dissemination of M tuberculosis.

Source: Adapted from Miller and Ernst.41

(0.28MB).

Cell-based immunity is essential for control of TB infection and, for a long time now, it has been recognized that CD4+ T cells are important mediators of M tuberculosis immunity. More recently, the focus has been on specific CD8+ cells, but the meaning of this finding has yet to be determined, particularly in humans. It is believed that CD8+ cells directly limit growth of mycobacteria by causing death of infected cells,49,50 and also indirectly, through release of cytokines that promote activation of macrophages and chemokines that coordinate recruitment of other cells. Other studies indicate that CD8+ cells are probably less important during the acute phase of infection by M tuberculosis but essential during the chronic phase, and so they may help prevent TB reactivation.51,52 On assessment of the contribution of CD8+ cells to M tuberculosis immunity, it is important to remember that mice do not express granulysin, the cytolytic protein that, in humans, contributes to the death of M tuberculosis.49 Thus, studies in murine models may underestimate the importance of the antimycobacterial properties of CD8+ cells if their role is to be extrapolated to human immunity. Fewer studies have been performed in humans, but in vitro experiments show that TEMRA exhibit a high avidity for stains specific to granulysin and show higher levels both of cytotoxicity and antimycobacterial activity compared with other T cell subsets.53

The study by Bruns et al.48 goes beyond the definition of TEMRA as mere effectors with potential antimycobacterial importance, and attempts to explain the consequences of TNF blockade on these cells based on observations of M tuberculosis growth in cell culture. With these observations, an apparently new mechanism is described in which TNF is neutralized, thereby facilitating progression of latent infection to active infection. Previous studies in animal models had shown variable increases in bacillary load in absence of TNF.54 The results of these studies indicated that TNF contributes to direct control of intramacrophagic mycobacteria and indirect suppression of mycobacterial growth by modulating the formation and persistence of granulomas.

Investigation of possible TB infection in candidates for biological treatment should start with evaluation of the potential risk of exposure to M tuberculosis. The highest-risk groups are summarized in Table 5.

In absence of a gold standard for diagnosis of TB infection, current management of patients with autoimmune diseases who are candidates for anti-TNF treatment includes a directed medical history to investigate history of TB or latent infection, whether treated or not; assessment of the risk of exposure to possible active cases; the search for evidence of prior TB in the chest x-ray; and tests to detect LTI (tuberculin test [TT] or interferon-γ release assay [IGRA]).64

IGRAs are used as routine tests for diagnosis of latent infection in the United States of America, Canada, Australia, Japan, and some western European countries (United Kingdom, Italy, Germany, and Switzerland). Some countries use them instead of the TT, whereas others use them in combination. In Spain, it is recommended to complement the TT with IGRA to rule out false positives in candidates for chemoprophylaxis with prior BCG vaccination (to rule out false positives to tuberculin) and in those with a negative skin patch test and suspected immunosuppression (to rule out a false negative TT)75 (Fig. 3).

Figure 3.

BCG, bacillus Calmette-Guerin; IGRA, interferon γ release assay.

a Patients with CD4+ counts ≥ 500 cells/mL are considered immunocompetent.

b In cases of severe immunosuppression (CD4+ count < 200 cells/mL), the probability of true infection should be assessed individually.

c IGRA measurement can be repeated periodically (depending on the individual risk in each patient) without any booster effect.

Source: Adapted from the Diagnostic and Treatment Guidelines for Tuberculosis of the Spanish Society of Pulmonology and Thoracic Surgery (SEPAR)..75

(0.09MB).

Although numerous articles have been published in recent years, data on the performance of these tests are limited and contradictory.74–80 The data from those studies enable the conclusion that: a) agreement between TT and IGRA is weaker in patients with systemic inflammatory diseases than in healthy subjects due to the lower prevalence of positive TT in the first group66,81–86; b) the extent of tuberculin response is significantly lower in patients than in healthy controls66; c) discordance between TT+ and IGRA- is observed in vaccinated patients; d) discordance between TT- and IGRA+ is often observed in patients receiving corticosteroid therapy86 (Table 6).

Due to the vulnerability of these patients to TB when they are receiving anti-TNF treatment, it seems prudent to perform the 2 tests (TT and IGRA) in parallel to maximize diagnostic sensitivity of screening, at least, until evidence of the utility of IGRA in this population has been consolidated. It is important to mention that when the 2 tests are performed sequentially, IGRAs should be performed first due to the booster effect of the TT.87,88 It is not known how long the booster effect persists and whether it depends on the quantity of purified protein derivative administered or its characteristics.

Given that some patients with latent infection may have false negatives for IGRAs and true positives for TT, and vice versa, the dual screening strategy would enable clinicians to offer preventive treatment to all patients with latent infection (Table 7).

The results of the IGRAs in patients who are in treatment with anti-TNF agents remain difficult to interpret. One proposed strategy was to repeat the tests after starting biological therapy to detect false negatives. The available data from recent studies89–91 show conversions and reversions in both tests related to reproducibility problems and changes in the patients¿ immune status. Given this variability, it is not recommended to repeat the test for follow-up of patients during treatment.

The identification and treatment of LTI greatly reduces the risk of reactivation and, thus, has the potential to protect the health of infected individuals as well as preventing further infection by reducing the number of possible sources of contagion.92,93

After treatment of LTI, the durability of protection against reactivation is variable and depends on the prevalence of TB in the area and the risk of re-exposure. It is assumed that treatment for LTI can confer lifelong protection against disease. Among Eskimos in Alaska, for example, the protective effect of isoniazid (INH) has been shown to last more than 20 years.94,95

The treatment of latent infection is only started once active TB has been excluded.

Regimens for LTI treatment include INH for 9 months, rifampicin (RIF) monotherapy for 4 months, or INH with RIF for 3 months.96

For exposed patients who supposedly are infected with mycobacteria resistant to multiple agents, treatment of LTI should be individualized, based on the susceptibility profile of the index case.110,111 Two drugs should always be chosen to which the organism is susceptible, for example, a combination of pyrazinamide and ethambutol or a fluoroquinolone with antituberculous activity (such as levofloxacin).112 The duration of treatment has not been defined, although from 6 to 12 months seems to be the most reasonable. Patients with LTI due to multidrug resistant strains should receive carefully monitored treatment and be followed closely due to the risk of liver toxicity. In the case of extreme multidrug-resistant strains, with resistance to several drugs, prophylaxis in contacts may be very difficult and no regimens have been tested beyond the first-line drugs.

Patients in treatment for LTI should be supervised monthly with clinical examination and liver testing. Patients should also be educated about the symptoms of hepatitis and instructed to suspend treatment and seek medical attention quickly to reduce the risk of progression to serious disease.113

Treatment of LTI should be discontinued in symptomatic patients with serum transaminase levels in excess of 3 times the upper limit of normal or in asymptomatic patients with serum transaminase levels more than 5 times the upper limit of normal. Treatment should not be restarted under any circumstances.

A common dilemma is whether to continue LTI treatment after having started complete treatment for suspicion of active TB which is finally not confirmed (negative cultures or histology). The duration of combination therapy in the initial phase administered may apply to the total duration of treatment for LTI. Subsequent options include interrupting treatment if RIF and pyrazinamide have been given for less than 2 months, continuing treatment with RIF for a total of 4 months (with or without INH), or continuing with INH treatment for a total of 9 months.114

In the case of regimens with INH (9 or 6 months), treatment can be restarted where it was left off if less than 3 months have been lost. Treatment for LTI should be reinitiated from the start if more than 3 months have elapsed. For treatment of LTI with regimens of 3 or 4 months, treatment can be restarted from where it left off if less than 2 months of the course remain.

The CDC recommends treatment of LTI of all patients who are planning to take a TNF inhibitor and who have a positive TT result (induration ≥5mm) or a positive IGRA result115,116 and in the case of patients with negative TT or IGRA but with evidence of previous TB disease in the chest X-ray (for example, regional fibrosis with or without enlarged hilar lymph nodes) or if there is epidemiological evidence of prior exposure to TB (for example, after having been in close contact with a TB case).

Patients who are candidates to receive anti-TNF agents with an indication for LTI treatment in general should receive standard therapy, that is, INH for 9 months. The recommendation about what the duration of therapy for LTI should be before starting anti-TNF has not been established, but most authors propose that patients should receive at least 1 month of treatment for LTI before starting anti-TNF therapy, whenever this is possible.14,117

Treatment with LTI will not protect against TB in the case of recently acquired infection except, perhaps, during the time that the treatment is administered. Therefore, patients who take anti-TNF agents should avoid exposure to TB (for example, occupational exposure such as in health centers with a high prevalence of TB, homeless shelters, prisons, travel to regions with a high prevalence of TB). Treatment with INH of latent infection will likewise not protect against reactivation of the infection by INH-resistant strains.118 There are no studies to enable recommendation of combined regimens of LTI treatment in these patients, although the RIF plus INH regimen has been used by expert groups in a large number of patients, without having detected cases of active TB after starting treatment.

Table 8 summarizes the recommendations of the panel of experts from the scientific societies participating in this multidisciplinary consensus document. The aim is to guide dermatology specialists in the assessment of risk of developing TB and to prevent disease through the diagnosis and treatment of LTI.

Figure 4 shows the algorithm proposed for assessment of TB infection in candidates for biological therapy.119,120

Figure 4.

Provisional algorithm for assessment of TB infection in candidates for anti-TNF treatment.

a Appropriate treatment for TB is defined as ≥6 months of treatment with first-line drugs, including ≥2 months of the combination of rifampicin, isoniazid, pyrazinamide and ethambutol. Appropriate treatment for latent infection can be administration of ≥6 months of isoniazid, 3 months of isoniazid and rifampicin, or 4 months of rifampicin alone.

b The risk of latent infection is determined by weighting factors such as exposure to known contagious case, age, country of origin, and work and social history of the individual, including travel to endemic countries and repeated exposure to collectives at risk (closed institutions, homeless, drug users).

c In those infected many years earlier, TT can be negative and become positive in a second TT (booster phenomenon). With the availability of IGRAs, it would seem more practical to use these as complementary tests rather than performing a second TT, both in patients aged more than 60 years and in immunosuppressed patients due to different causes, regardless of age.

d There are no conclusive data to establish a safe period between the start of treatment for latent infection and the start of anti-TNF therapy. Four weeks delay is considered usual practice and safe by most experts.

eActive TB treatment should be completed before starting biological therapy. See text for interpretation and extension of this algorithm.

Anti-TNF, anti-tumor necrosis factor; BCG, bacillus Calmette-Guerin; IGRA, interferon γ release assay; TB, tuberculosis; TT, tuberculin test.

Source: Adapted from Winthrop..120

(0.2MB).

Esteban Daudén declares he is a member of advisory boards, consultant, grant recipient, research support recipient, participant as an investigator in clinical trials, and recipient of fees for talks with the following pharmaceutical companies: Abbvie (Abbott), Amgen, Janssen-Cilag, Leo Pharma, MSD, Pfizer, Novartis, Celgene, and Lilly.

Carlos Taxonara Samso declares he is a consultant and speaker for MSD, AbbVie, Jannsen, Takeda, and Pfizer.

Francisco Javier López Longo declares having been an occasional speaker for Abbvie, Actelion, Bristol-Myers-Squibb, GSK, MSD, Pfizer, Roche Farma, and UCB and having received research funding from Abbvie and GSK.

The remaining authors declare that they have no conflicts of interest.