Nontuberculous mycobacterial (NTM) infection is an important cause of cervical lymphadenitis in children without Calmette–Guérin bacillus (BCG) vaccination.1 NTM cervical lymphadenitis usually occurs between 1 and 5 years of age.1–3 The typical clinical presentation is an otherwise healthy child with unilateral nontender cervical mass.2,3 The differential diagnoses range in urgency from normal reactive lymph nodes to Mycobacterium tuberculosis (MTB) infection and malignancies.3,4 Eventually, a classic scrofula develops with skin discoloration, thinning and a draining fistula.2,3 Surgery is considered the primary treatment of choice.2–4 Delayed diagnosis increases the risk of nerve damage and poor cosmetic outcome.2,4 Currently, there is no reliable rapid and noninvasive diagnostic method available. Childhood NTM cervical lymphadenitis remains a diagnostic challenge.
The commercial T-SPOT.TB (ELISPOT, Oxford Immunotec Ltd, Oxfordshire, United Kingdom) test is intended to detect patients sensitized to MTB.5 We have previously reported on modifications applied to this test, including an additional purified protein derivative (PPD; Statens Serum Institut, Copenhagen, Denmark) antigen mixture stimulation of peripheral blood mononuclear cells (PBMCs).5 The PPD mixture contains dozens of mycobacterial antigens that cross-react with the BCG strains, MTB and many NTM species.3,6
During the diagnostic work-up of non-BCG–vaccinated children with suspected NTM cervical lymphadenitis, we observed high reactivity to the PPD antigens without reactivity to MTB specific antigens. This observation encouraged us to evaluate the potential of the modified T-SPOT.TB test for diagnosing NTM cervical lymphadenitis in children.
MATERIALS AND METHODS
The study was conducted according to the guidelines set by the Institutional Review Board of Helsinki University Central Hospital. On review of the hospital records, we found 21 children who had been diagnosed with NTM cervical lymphadenitis in the Hospital District of Helsinki and Uusimaa (HUS) in Finland between March 2009 and January 2012. The patient records and laboratory results of these children were reviewed retrospectively. All children were non-BCG–vaccinated and presented with unilateral nontender cervical mass, either healing after excision or progressing to drainage lasting up to several months. Seventeen had been tested with the modified T-SPOT.TB method. In 10 of these cases, the diagnosis had been confirmed by a positive bacteriological isolation of NTM. These culture-confirmed NTM cervical lymphadenitis cases that had been tested with the modified T-SPOT.TB method formed the disease group (n = 10, 5 male and 5 female). None of the enrolled cases were immunocompromised.
Specimens for conventional cultures (n = 10) had been collected from affected lymph nodes through a biopsy, a fine-needle aspiration or a draining fistula. Mycobacterial species had been identified with GenoType Mycobacterium CM assay (Hain Lifescience, Nehren, Germany). The bacterial isolations were Mycobacterium avium (8/10) and Mycobacterium malmoense (2/10). The age range in the disease group was 15–38 months with a median age of 31 months. All were born in Finland to Finnish parents. The duration of the lymphadenitis from the first parental observation to the test date ranged from 1 to 8 months with a median of 2 months.
As a control group, we wanted healthy nonsymptomatic non-BCG–vaccinated children tested with the same method. We reviewed the hospital records of all children younger than 5 years tested in HUS in 2009 and 2010. During that period, a total of 99 children had been tested, including 49 BCG-vaccinated, 18 with unclear BCG status and 32 non-BCG–vaccinated. From the 32 non-BCG–vaccinated we excluded those diagnosed with medical condition (n = 11, eg, rheumatoid arthritis, leukemia or pyogenic, mycobacterial or viral infection). The remaining healthy non-BCG–vaccinated children formed the control group (n = 21, 8 boys and 13 girls).
The control group had been tested because of MTB contact tracing (n = 16) or pre-BCG–vaccination investigations (n = 5). The age range in the control group was 4–54 months with a median age of 21 months. Seventeen were born in Finland and 4 abroad.
The commercial T-SPOT.TB test contains peptide mixtures of early secretory antigenic target-6 (EST-6) and culture filtrate protein-10 (CFP-10) proteins that are highly specific for MTB. In the modified method, PBMCs were stimulated with nonspecific PPD antigen mixture in addition to the ESAT-6 and CFP-10. Briefly, 250,000 PBMCs in 100 μL were stimulated with PPD in 50 μL (final concentration, 22 μg/mL). The rest of the assay was performed as recommended by the manufacturer. We calculated the lymphocyte ratio for each PBMC specimen and expressed the results as a number of reactive spots per 106 lymphocytes. We set-up the cut-off of 25 spots/106 lymphocytes for all antigen stimulations.5,6
The test results of 31 children (10 belong to disease group and 21 control group) were analyzed. In the disease group, the reactivities of PBMCs to PPD antigen stimulation ranged from 30 to 455 spots/106 lymphocytes with a median of 231 spots/106 lymphocytes, whereas in the control group, the corresponding reactivities ranged from 0 to 195 spots/106 lymphocytes with a median of 8 spots/106 lymphocytes. The test result distributions of the 2 groups are presented in Figure 1.
In all the disease group patients, PBMCs were reactive to PPD, whereas in the healthy control group, reactivity to PPD was observed in 4 of 21 samples. Calculated from these results, the estimated sensitivity and specificity of the test would be 1.00 (95% confidence interval: 1.00–1.00) and 0.81 (0.64–0.98), respectively.
One child in the disease group had a borderline positive reactivity to CFP-10 antigens (26 spots/106 lymphocytes). This child’s PBMCs were nonreactive to ESAT-6 antigens and highly reactive to PPD (455 spots/106 lymphocytes). M. avium was isolated from this child’s lymph node specimen, and the result of the Xpert MTB/RIF test (Cepheid, Sunnyvale, CA) was negative.
The modified T-SPOT.TB method with additional PPD stimulation is a new promising noninvasive diagnostic technique for childhood NTM cervical lymphadenitis.
Mycobacterial culture is regarded as the diagnostic standard for NTM lymphadenitis, but it is impaired by poor recovery (~50%) and slow growth rate of the bacteria.3,7,8 Other diagnostic tools include staining, histopathological examination and nucleic acid amplification techniques.3,7 Fine-needle aspiration or tissue biopsy is essential in obtaining diagnostic material, so, currently, some form of invasive intervention is required for reliable diagnosis.3,7,9 A positive tuberculin skin test (TST) may indicate an NTM infection, but alone cannot differentiate between MTB and NTM infection.10 Previous diagnostic algorithms have suggested a positive TST coupled with a negative interferon-gamma release assay as a noninvasive diagnostic method for low-endemic MTB regions, but in 30% of NTM infections, the TST presents a false negative result.8,10 Intradermal skin tests with proteins derived from NTM have been suggested as an alternative.8 However, the required antigens are currently not commercially available, and some studies have demonstrated that 6–32% of asymptomatic children have a high reactivity in skin tests to NTM proteins.7,8
The T-SPOT.TB has been used in Finland since 2004. From the very beginning, we applied modifications to the test, among which stimulation with PPD was initially designed to serve as an additional positive control.5 At that time, 98% of the Finnish population was BCG vaccinated. A healthy BCG-vaccinated subject shows reactivity to PPD but not to MTB specific antigens. In 2006, the universal BCG vaccination of newborns was changed to a risk-group based approach. In a non-BCG–vaccinated child with lymphadenitis, reactivity to PPD without reactivity to MTB specific antigens is highly suggestive of NTM infection.10
Our study presents the clinical and laboratory results obtained from investigation of 31 children (10 belong disease group and 21 control group) tested with the modified T-SPOT.TB method. The data analysis shows promising sensitivity (1.00) and good specificity (0.81). However, because of the small sample size, the calculated values of the test should be viewed with caution. We suspect that the positive reactivity of the 4 healthy children in the control group resulted from an asymptomatic exposure to NTM. The modified T-SPOT.TB seems to compare well with other diagnostic tests in non-BCG–vaccinated children in low MTB incidence settings and when MTB infection has been excluded. Only 1 patient with cervical MTB lymphadenitis was diagnosed during the study period in HUS. The advantage of our approach is that a presumptive diagnosis is achieved rapidly without multiple appointments and before invasive diagnostic techniques. The possibility of MTB infection, which was our initial reason for performing the modified T-SPOT.TB test, is investigated concomitantly.
Our study holds limitations. It was retrospective, and the available clinical information was variable. Our sample size was small because NTM cervical lymphadenitis is relatively rare. The median parent-reported duration of lymphadenitis before performing the test was 2 months. Further studies are required for assessing the time frame when the test is applicable at the earliest. Further prospective studies will reveal the suitability of the method in regions with varying NTM exposure and higher MTB prevalence, in other clinical presentations of NTM infection and in children with other than non-mycobacterial causes of cervical lymphadenitis.
This study is the first report of the T-SPOT.TB with additional PPD stimulation for diagnosing NTM cervical lymphadenitis in children. It presents a promising diagnostic technique, and further prospective studies are needed to confirm our findings.
The authors wish to thank Dr. Aino Ruohola for the helpful comments on the manuscript.
1. Romanus V, Hallander HO, Wåhlén P, et al. Atypical mycobacteria in extrapulmonary disease among children. Incidence in Sweden from 1969 to 1990, related to changing BCG-vaccination coverage. Tuber Lung Dis. 1995;76:300–310
2. Loeffler AM. Treatment options for nontuberculous mycobacterial adenitis in children. Pediatr Infect Dis J. 2004;23:957–958
3. Hatzenbuehler LA, Starke JR. Common presentations of nontuberculous mycobacterial infections. Pediatr Infect Dis J. 2014;33:89–91
4. Scott CA, Atkinson SH, Sodha A, et al. Management of lymphadenitis due to non-tuberculous mycobacterial infection in children. Pediatr Surg Int. 2012;28:461–466
5. Tavast E, Salo E, Seppälä I, et al. IGRA tests perform similarly to TST but cause no adverse reactions: pediatric experience in Finland. BMC Res Notes. 2009;2:9
6. Yang H, Kruh-Garcia NA, Dobos KM. Purified protein derivatives of tuberculin–past, present, and future. FEMS Immunol Med Microbiol. 2012;66:273–280
7. Bruijnesteijn Van Coppenraet ES, Lindeboom JA, Prins JM, et al. Real-time PCR assay using fine-needle aspirates and tissue biopsy specimens for rapid diagnosis of mycobacterial lymphadenitis in children. J Clin Microbiol. 2004;42:2644–2650
8. Lindeboom JA, Kuijper EJ, Prins JM, et al. Tuberculin skin testing is useful in the screening for nontuberculous mycobacterial cervicofacial lymphadenitis in children. Clin Infect Dis. 2006;43:1547–1551
9. Lindeboom JA. Conservative wait-and-see therapy versus antibiotic treatment for nontuberculous mycobacterial cervicofacial lymphadenitis in children. Clin Infect Dis. 2011;52:180–184
10. Staufner C, Sommerburg O, Holland-Cunz S. Algorithm for early diagnosis in nontuberculous mycobacterial lymphadenitis. Acta Paediatr. 2012;101:e382–e385