Tuberculosis of the skin has become a rare disease in industrialized countries. Polymerase chain reaction (PCR) is a powerful diagnostic tool for mycobacterial infections of the skin, but it can fail, as demonstrated in this case.
A woman in her 80s was referred for surgical treatment of a cervical abscess. Similar abscesses erupted in the cervical region over the course of 2 years (Figure 1A). A needle aspiration biopsy was performed on a node at the left side of the neck, which measured 2 cm. The histopathologic report described a minor nonspecific inflammatory reaction, not suggestive of infection. Findings of the Mycobacterium tuberculosis PCR were negative. A culture was not performed. Two months later, the whole nodule was excised, including the adjacent inflamed skin. The resulting defect, with a diameter of 7 cm, was closed with a rotary-transposition flap. During this intervention, the thoracic nerve was injured resulting in an elevation palsy of the left arm. The histopathologic report of the excised tissue again showed a nonspecific inflammatory reaction; no microbiological analysis was conducted.
A, Multiple cervical abscesses recurred in the 2 years prior to presenting for care. B, A cold abscess at the neck was opened and drained to gain a specimen for histologic and microbiological analysis; multiple puckered scars are visible on the neck.
At presentation, the patient had puckered scars scattered over the neck in addition to an unusual “cold abscess” (Figure 1B). The clinical appearance was suggestive of scrofuloderma. Results of the Mendel-Mantoux test were positive (diameter, 20 mm), as were those from the interferon-γ release assay. However, PCR findings from the skin biopsy specimen and abscess material were negative for M tuberculosis. Histologically, no acid-fast bacilli could be detected by Ziehl-Neelsen staining.
Cervical sonography and magnetic resonance tomography revealed multiple abscesses in the lateral muscle loge. Chest radiography excluded pulmonary tuberculosis. Laboratory work showed an elevated level of C-reactive protein (115 mg/L; normal, <5 mg/L) but no other pathological findings. (To convert C-reactive protein to nanomoles per liter, multiply by 9.524.)
After 19 days, M tuberculosis was cultivated from the skin specimen (Figure 2). The strain was sensitive to isoniazid, rifampicin, pyrazinamide, ethambutol, and streptomycin.
Microscopic results from a liquid culture stained with Ziehl-Neelsen showing acid-fast M tuberculosis. The cord factor arrangement typical of M tuberculosis can be seen. The positive result was obtained 19 days after the specimen was collected. The arrow indicates a single bacterium.
Classic quadruple treatment with isoniazid, 300 mg/d; pyrazinamide, 1500 mg/d; ethambutol, 1200 mg/d; and rifampicin, 600 mg/d, was initiated. After 2 months, the regimen was reduced to isoniazid and rifampicin. After 4 months of the reduced regimen, all skin lesions had healed completely, leaving scars, and sonography revealed no remaining abscesses. Treatment was well tolerated, and at 24-month follow-up, no new nodules had evolved.
From 1% to 2% of tuberculosis cases are cutaneous tuberculosis (CTB).1 Tuberculosis cutis colliquativa, also known as scrofuloderma, is the most common CTB subtype in Europe.2 Scrofuloderma is a subcutaneous form of CTB manifesting with cold abscesses most commonly on the neck that spreads from underlying lymph nodes. Infection can also involve joints, bones, and epididymis.3 The same quadruple antibiotic therapy is used as in pulmonary tuberculosis. Before treatment is begun, possible multidrug resistance should be excluded.
Cutaneous tuberculosis can be caused by consuming cow milk contaminated with Mycobacterium bovis or by droplet infection with M tuberculosis. The correct diagnosis is often significantly delayed because CTB is not routinely considered in the differential diagnosis or because investigations fail to reveal the presence of M tuberculosis.4
Our case illustrates that scrofuloderma, though a rare disease in industrialized countries, should still be considered in the differential diagnosis of unusual abscesses and nodules of the neck. Skin testing and interferon-γ release assay can support the clinical diagnosis. Since PCR has been shown to have a limited sensitivity and specificity (eg, 88% sensitivity and 83% specificity5), there is a risk of failure to detect mycobacteria in skin samples by relying solely on PCR. Therefore, PCR should always be accompanied by culture.6
Corresponding Author: Ozan Haase, MD, Department of Dermatology, University of Luebeck, Ratzeburger Allee 160, 23538 Luebeck, Germany (firstname.lastname@example.org).
Published Online: April 23, 2014. doi:10.1001/jamadermatol.2013.10175.
Conflict of Interest Disclosures: None reported.
Haase O, von Thomsen AJ, Zillikens D, Solbach W, Kahle B. Recurrent Abscesses of the NeckScrofuloderma. JAMA Dermatol. 2014;150(8):909-910. doi:10.1001/jamadermatol.2013.10175