[Skip to Navigation]
Sign In
Aug 2011

Usefulness of Interferon- γ Release Assays in the Diagnosis of Erythema Induratum

Author Affiliations

Author Affiliations: Department of Dermatology and Skin Science, University of British Columbia (Drs Vera-Kellet and Dutz), and Division of Tuberculosis Control, British Columbia Centre for Disease Control (Drs Peters and Elwood), Vancouver, Canada; and Department of Dermatology, Pontificia Universidad Cat ólica de Chile, Santiago (Dr Vera-Kellet).

Arch Dermatol. 2011;147(8):949-952. doi:10.1001/archdermatol.2011.183

Background Erythema induratum (EI) manifests as tender indurated nodules on the lower legs. It may be associated with concomitant active tuberculosis (TB) and is considered a hypersensitivity reaction to mycobacterial antigens. However, the results of Mycobacterium cultures are rarely positive, and the tuberculin skin test is of limited usefulness in populations exposed to bacille Calmette-Gu érin (BCG). Interferon- γ release assays (IGRAs) are alternatives to the tuberculin skin test and have high specificity. We explored the use of IGRAs as an adjunct in the diagnosis of EI. We describe 5 patients with positive tuberculin skin test results and a history of vaccination against BCG or TB in whom IGRAs supported the diagnosis of EI.

Observations All patients were initially seen with tender nodules on the lower legs and a history of BCG vaccination or TB. Tuberculin skin test results were positive, and chest radiographic results were normal. The results of Mycobacterium cultures were negative in all patients, and biopsy specimens were compatible with EI. Interferon- γ release assays were performed in all patients and supported initiation of anti-TB treatment in 4 of 5 patients.

Conclusion Interferon- γ release assays may have value as an adjunct in the diagnosis of EI, particularly in the setting of prior BCG exposure.

Erythema induratum (EI) is an uncommon condition that manifests as tender indurated nodules on the lower legs. It was originally described by Ernest Bazin in 1861 as nodular lesions on the legs in association with tuberculosis (TB).1 The term tuberculid refers to a cutaneous eruption that arises in response to Mycobacterium tuberculosis or M tuberculosis antigens and is suspected to be immunologic in origin. Unlike cutaneous TB (ie, chancre, verrucosa, and scrofuloderma), M tuberculosis is seldom found by staining or culture of biopsy specimens in tuberculids; hence, a presumptive diagnosis of EI is made on the basis of characteristic clinical morphologic appearance, positive tuberculin skin test (TST) result, compatible histological findings, and historical evidence of TB exposure.

Erythema induratum is uncommon in North America but is prevalent in India, Hong Kong, and some areas of South Africa. For example, the most common cutaneous manifestation of TB in Hong Kong between 1993 and 2002 was EI (86.4% [127 of 147 TB cases]).2 These areas have a higher incidence of TB, and vaccination against bacille Calmette-Gu érin (BCG) is common. The TST can be used in nonvaccinated populations to investigate the presence of active or latent TB. However, it is of limited usefulness in populations exposed to BCG. Interferon- γ release assays (IGRAs) are alternatives to the TST and have high specificity for identification of M tuberculosis infection. The US Centers for Disease Control and Prevention3 has recently determined that IGRAs are the preferred method of testing for TB in individuals who have received BCG vaccination. Therefore, we examined the use of IGRAs as an adjunct in the diagnosis of EI among individuals previously exposed to BCG vaccination or TB.


We describe 5 patients with suspected EI, a positive TST result, and previous BCG vaccination or TB. Interferon- γ release assays (T-SPOT Assay TB test [T-SPOT]; Oxford Immunotec, Oxford, England; and QuantiFERON-TB Gold test [QTF]; Cellestis, Inc, Valencia, California) were performed in all patients. Erythema induratum was diagnosed when a patient was seen with a history of a painful nodular eruption on the lower limbs, pathologic findings were consistent with nodular vasculitis, the TST result was positive, and the eruption responded to anti-TB medication. Polymerase chain reaction (PCR) –based diagnosis of TB was unavailable in our jurisdiction. Interferon- γ release assays were interpreted according to guidelines published by the US Centers for Disease Control and Prevention.3


Five patients with suspected EI were followed up by 2 of us (C.V.-K. and J.P.D.) between January 1, 2005, and December 31, 2009. Clinical characteristics of the patients are given in the Table.

Table. Clinical Characteristics of the Patientsa
Table. Clinical Characteristics of the Patientsa
Table. Clinical Characteristics of the Patientsa

An example of a case with prior BCG exposure is patient 5 (Table), a 51-year-old man with a 6-month history of recurrent tender nodules on the posterior aspects of his legs (Figure, A) and the lateral aspects of his thighs. His TST result was positive (Figure, B). Results of a skin biopsy specimen were compatible with necrotizing panniculitis (Figure, C). Mycobacterium culture results were negative. His chest radiographic results were normal. He had been vaccinated against BCG. Interferon- γ release assays were performed, and the results of the T-SPOT and the QTF were positive. The patient received 9 months of isoniazid and vitamin B6 therapy and experienced complete resolution of the lesions.

Figure. Patient 5 (Table). A, Posterior left calf showing a proximal subcutaneous nodule with overlying erythema. B, Volar left forearm at 48 hours after tuberculin skin test. C, Histopathologic findings of granulomatous lobular panniculitis (hematoxylin-eosin, original magnification  ×200).

Figure. Patient 5 (Table). A, Posterior left calf showing a proximal subcutaneous nodule with overlying erythema. B, Volar left forearm at 48 hours after tuberculin skin test. C, Histopathologic findings of granulomatous lobular panniculitis (hematoxylin-eosin, original magnification ×200).

An example of a case without prior BCG exposure is patient 3 (Table), a 49-year-old woman with an 8-month history of tender nodules on her lower legs. In 1987, the results of a sputum culture had been positive for M tuberculosis, and she had completed standard treatment that year. Erythema induratum was suspected 22 years later, and a skin biopsy specimen showed mixed lobular and septal panniculitis compatible with EI. Results of her Mycobacterium culture were negative, her TST result was strongly positive, and her T-SPOT result was positive. Her condition improved after a further course of anti-TB therapy.


Erythema induratum usually manifests as recurrent flares of violaceous nodules on the legs in young to middle-aged women. The lesions are cold, are sometimes painless, have a tendency to manifest central ulceration, and may heal with scars. The most frequent locations of these lesions are the posterior aspect (ie, the calves) and anterolateral areas of the legs. The feet, thighs, arms, and face can be involved.4 The histologic features of EI include lobular panniculitis associated with granulomatous inflammation with focal necrosis, vasculitis, and septal fibrosis. These findings are not specific; they vary depending on the age of the lesion and overlap with other forms of panniculitis.4 Nevertheless, the presence of septal granulomatous inflammation and lobular granulomatous inflammation is characteristic of EI and contrasts with the primary septal inflammation in erythema nodosum.

The association of EI with TB used to be controversial because mycobacteria are rarely cultured from skin lesions.1 The term nodular vasculitis is used to describe chronic inflammatory nodules of the legs that show histopathologic changes similar to those of EI but are unassociated with TB.5 The absence of acid-fast bacilli, detectable by microscopy, in tuberculid skin lesions may be the result of few bacilli. Another theory is that the lesions are secondary to a systemic reaction induced by circulating M tuberculosis fragments. Evidence that M tuberculosis is involved in the pathogenesis of EI is supported by the detection of M tuberculosis DNA in skin lesions, a strongly positive TST result, and resolution of the skin lesions with anti-TB therapy.6 If skin biopsy specimens are evaluated for M tuberculosis using pathogen-specific DNA testing by PCR with primers for the repeat sequence IS6110, M tuberculosis is found in 25% to 100% of cases.7-9 False-negative PCR test results have been ascribed to degraded target DNA, diminished PCR amplification secondary to the formalin used, or insufficient target DNA.10 In the series of cutaneous TB cases from Hong Kong, 5 of 7 patients with EI who underwent PCR testing had positive results.2

The standard test for active or latent infection with M tuberculosis is the TST. Previous BCG vaccination can confound the results of a TST interpretation because it increases the likelihood of a positive TST result.11 A meta-analysis12 of the evidence from 1966 to 1999 for the effect of BCG vaccination on TSTs in individuals without active TB showed that positive TST results with indurations of at least 15 mm are more likely to be the result of TB infection than of BCG vaccination. Four of our 5 patients had a history of BCG vaccination, and 2 of 4 had TSTs with indurations less than 15 mm, rendering the results uninterpretable.

Interferon- γ release assays are a reasonable option for patients who want more specific evidence of latent infection before undergoing a complete course of anti-TB treatment, which has potentially toxic effects. Two different IGRAs are available to measure responses to the Mycobacterium -specific antigens: early secretory antigenic target 6 (ESAT-6) and cultured filtrate protein 10 (CFP-10). The QTF is based on a whole-blood enzyme-linked immunosorbent assay, and the T-SPOT is based on an ex vivo overnight interferon- γ enzyme-linked immunospot technique to measure peripheral blood mononuclear cell –derived interferon- γ.13 The tests are similar in terms of antigens (ie, ESAT-6 and CFP-10) and incubation time (ie, overnight or 16-24 hours). Because the genes encoding these antigens are found in the region of difference 1 of the M tuberculosis genome, which is not present in the genomes of BCG, Mycobacterium bovis, or Mycobacterium avium, IGRAs do not yield false-positive results in populations exposed to BCG or M avium.14,15 In contrast, infections with Mycobacterium marinum, Mycobacterium kansasii, and Mycobacterium szulgai cannot be distinguished from M tuberculosis using IGRAs because these organisms share the same antigen set.16 The TST uses a mix of antigens from M tuberculosis, causing false-positive reactions in patients with BCG vaccination or M avium sensitization. Further advantages of IGRAs compared with the TST are that IGRAs require only 1 patient visit and that these assays are ex vivo tests, reducing the risk for adverse effects and eliminating potential boosting when testing is repeated.17 The disadvantages of IGRAs compared with the TST include higher cost, the need for careful handling of drawn blood to maintain the viability of lymphocytes, and the requirement of a fully equipped laboratory with appropriately trained staff.

The number of interferon- γ –producing antigen-specific effector T cells decreases significantly in most patients with TB between diagnosis and treatment completion, but IGRA reversion is exceptional.18-20 Patients without IGRA reversion have a significantly higher baseline TST result.19 It has been suggested that IGRAs may be used to monitor treatment efficacy or risk of relapse. In this regard, it is notable that patient 3 (Table), who had previously been treated for TB and thereafter had developed EI, had a strongly positive T-SPOT result. Patients with QTF scores in our series had a high level of response (range, 1.97-7.86 IU/L; positive cutoff, >0.35 IU/L), suggesting that high immune reactivity to M tuberculosis may correlate with the development of EI.

Chee et al21 demonstrated a disparity in the performance of IGRAs among different racial/ethnic groups, with an increased likelihood of indeterminate QTF results among Malays and Indians compared with Chinese patients, as well as diminished T-cell responses to the M tuberculosis –specific antigens in Malays (with T-SPOT assay) and Indians (with T-SPOT and QTF assays). In their study, factors independently associated with indeterminate QTF results and with positive T-SPOT results were female sex, age older than 60 years, and non-Chinese (ie, Indian or Malay) race/ethnicity. Of note, our patient with discordant T-SPOT and QTF results (patient 2 in the Table) was a woman from India. Furthermore, she had an indeterminate response to anti-TB therapy and could not unequivocally be diagnosed as having EI.

Interferon- γ release assays are being evaluated as tools to aid in the detection of occult M tuberculosis infection, such as in the setting of tumor necrosis factor inhibitor therapy. Findings in our patient series suggest that IGRAs may have value as an adjunct in the diagnosis of EI, particularly in the setting of prior BCG exposure. All patients with suspected EI and a positive QTF test result responded to anti-TB therapy. The relative value in the cost-efficacy and specificity-sensitivity of DNA testing compared with IGRA testing for EI remains to be studied. Furthermore, the occurrence of prior BCG vaccination in 4 of 5 patients studied herein suggests that this may be a risk factor for the development of EI. Alternately, provenance from an area where TB is endemic (and where BCG vaccination is administered) may be the sole risk factor for the development of EI. A comparison of EI in TB-exposed populations with vs without prior BCG exposure may elucidate this possibility. Interferon- γ release assays indicate only sensitization to TB and are not used to diagnose active infection as a rule. Although none of our patients had clinically active TB infection, positive TST and IGRA results should prompt clinical examination to rule out active infection. Finally, although we highlight the usefulness of IGRAs to support a diagnosis of EI in the setting of known previous BCG or TB exposure, TSTs remain a valuable tool for the diagnosis of EI in the absence of BCG exposure.

Back to top
Article Information

Correspondence: Jan P. Dutz, MD, FRCPC, Department of Dermatology and Skin Science, University of British Columbia, 835 W 10th Ave, Vancouver, BC V5Z 4E8, Canada (dutz@interchange.ubc.ca).

Accepted for Publication: December 13, 2010.

Author Contributions: Dr Dutz had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Vera-Kellet and Dutz. Acquisition of data: Vera-Kellet, Elwood, and Dutz. Analysis and interpretation of data: Vera-Kellet, Peters, Elwood, and Dutz. Drafting of the manuscript: Vera-Kellet and Dutz. Critical revision of the manuscript for important intellectual content: Vera-Kellet, Peters, Elwood, and Dutz. Study supervision: Dutz.

Financial Disclosure: None reported.

Funding/Support: Dr Vera-Kellet was supported by a scholarship from the Chile Bicentennial Scholarship (Becas Chile) and Pontificia Universidad Cat ólica de Chile. Dr Dutz was supported by a senior scientist award from the Child and Family Research Institute and is a senior scholar with the Michael Smith Foundation.

Ollert MW, Thomas P, Korting HC, Schraut W, Braun-Falco O. Erythema induratum of Bazin: evidence of T-lymphocyte hyperresponsiveness to purified protein derivative of tuberculin: report of two cases and treatment.  Arch Dermatol. 1993;129(4):469-4738466218PubMedGoogle Scholar
Ho CK, Ho MH, Chong LY. Cutaneous tuberculosis in Hong Kong: an update.  Hong Kong Med J. 2006;12(4):272-277Google Scholar
Mazurek GH, Jereb J, Vernon A, LoBue P, Goldberg S, Castro K.IGRA Expert Committee; Centers for Disease Control and Prevention (CDC).  Updated guidelines for using interferon gamma release assays to detect Mycobacterium tuberculosis infection: United States, 2010.  MMWR Recomm Rep. 2010;59(RR-5):1-25Google Scholar
Mascar ó JM Jr, Baselga E. Erythema induratum of Bazin.  Dermatol Clin. 2008;26(4):439-445, v18793975PubMedGoogle Scholar
Montgomery H, Leary PA, Barker NW. Nodular vascular diseases of the legs: erythema induratum and allied conditions.  JAMA. 1945;128(5):335-341Google Scholar
Lighter J, Tse DB, Li Y, Borkowsky W. Erythema induratum of Bazin in a child: evidence for a cell-mediated hyper-response to Mycobacterium tuberculosis Pediatr Infect Dis J. 2009;28(4):326-32819238114PubMedGoogle Scholar
Hsiao P-F, Tzen C-Y, Chen H-C, Su H-Y. Polymerase chain reaction based detection of Mycobacterium tuberculosis in tissues showing granulomatous inflammation without demonstrable acid-fast bacilli.  Int J Dermatol. 2003;42(4):281-28612694493PubMedGoogle Scholar
Tan SH, Tan BH, Goh CL,  et al.  Detection of Mycobacterium tuberculosis DNA using polymerase chain reaction in cutaneous tuberculosis and tuberculids.  Int J Dermatol. 1999;38(2):122-12710192161PubMedGoogle Scholar
Vieites B, Su árez-Pe ñaranda JM, P érez Del Molino ML,  et al.  Recovery of Mycobacterium tuberculosis DNA in biopsies of erythema induratum: results in a series of patients using an improved polymerase chain reaction technique.  Br J Dermatol. 2005;152(6):1394-139615949033PubMedGoogle Scholar
Quir ós E, Bettinardi A, Quir ós A, Pi édrola G, Maroto MC. Detection of mycobacterial DNA in papulonecrotic tuberculid lesions by polymerase chain reaction.  J Clin Lab Anal. 2000;14(4):133-13510906765PubMedGoogle Scholar
Rowland K, Guthmann R, Jamieson B, Malloy D. Clinical inquiries: how should we manage a patient with a positive PPD and prior BCG vaccination?  J Fam Pract. 2006;55(8):718-72016882448PubMedGoogle Scholar
Wang L, Turner MO, Elwood RK, Schulzer M, FitzGerald JM. A meta-analysis of the effect of bacille Calmette Gu érin vaccination on tuberculin skin test measurements [published correction appears in Thorax. 2003;58(2):188].  Thorax. 2002;57(9):804-80912200526PubMedGoogle Scholar
Kobashi Y, Mouri K, Yagi S,  et al.  Clinical evaluation for diagnosing active TB disease and transitional change of two commercial blood tests.  Scand J Infect Dis. 2008;40(8):629-63418642159PubMedGoogle Scholar
Adams LV, Waddell RD, Von Reyn CF. T-SPOT. TB Test ® results in adults with Mycobacterium avium complex pulmonary disease.  Scand J Infect Dis. 2008;40(3):196-20318274952PubMedGoogle Scholar
Menzies D, Pai M, Comstock G. Meta-analysis: new tests for the diagnosis of latent tuberculosis infection: areas of uncertainty and recommendations for research [published correction appears in Ann Intern Med. 2007;146(9):688].  Ann Intern Med. 2007;146(5):340-35417339619PubMedGoogle Scholar
Kobashi Y, Mouri K, Yagi S,  et al.  Clinical evaluation of the QuantiFERON ®-TB Gold test in patients with non-tuberculous mycobacterial disease.  Int J Tuberc Lung Dis. 2009;13(11):1422-142619861017PubMedGoogle Scholar
Pai M, Zwerling A, Menzies D. Systematic review: T-cell –based assays for the diagnosis of latent tuberculosis infection: an update.  Ann Intern Med. 2008;149(3):177-18418593687PubMedGoogle Scholar
Bosshard V, Roux-Lombard P, Perneger T,  et al.  Do results of the T-SPOT. TB interferon- γ release assay change after treatment of tuberculosis?  Respir Med. 2009;103(1):30-3418977647PubMedGoogle Scholar
Lee SH, Lew WJ, Kim HJ,  et al.  Serial interferon-gamma release assays after rifampicin prophylaxis in a tuberculosis outbreak.  Respir Med. 2010;104(3):448-45319879123PubMedGoogle Scholar
Ribeiro S, Dooley K, Hackman J,  et al.  T-SPOT. TB responses during treatment of pulmonary tuberculosis.  BMC Infect Dis. 2009;9(1):e2319250549PubMedGoogle Scholar
Chee CBE, Gan SH, KhinMar KW,  et al.  Comparison of sensitivities of two commercial gamma interferon release assays for pulmonary tuberculosis.  J Clin Microbiol. 2008;46(6):1935-194018400912PubMedGoogle Scholar