Differentiation between lymphadenitis caused by mycobacteria other than tuberculosis (MOTT) and extrapulmonary tuberculosis is a common diagnostic problem in immunocompetent children presenting with local lymphadenitis and a positive tuberculin skin test. In young children 9 of 10 cases are caused by MOTT, and recent reports suggest an increasing incidence of nontuberculous lymphadenitis.1, 2 Today, the most frequently isolated MOTT species is M. avium complex1, 3, 4 Diagnosis of nontuberculous lymphadenitis is based on detection of mycobacteria in the surgically removed lymph nodes. The sensitivity of culture detection of mycobacteria in tissue ranges between 30 and 50%,4, 5 and mycobacterial growth might take up to 8 weeks. Here we report the noninvasive diagnosis of cervical lymphadenitis caused by M. avium in 2 children by PCR analysis of gastric aspirates before surgery.
MATERIALS AND METHODS
Clinical samples. Surgically removed lymphatic tissue from 13 children with suspected mycobacterial lymphadenitis admitted to the Children's Hospital of Heidelberg University between 1995 and 1997 was prospectively examined. In 4 children 1 to 4 gastric aspirates were available for analysis to exclude extrapulmonary tuberculosis.
Gastric aspirates were obtained 1 to 3 days before surgical excision of the lymph nodes. The macroscopically involved tissue was intraoperatively separated into four equal sections for histopathology, general bacteriology, mycobacterial microscopy and culture and molecular analysis.
Microscopy and culture isolation. Concentrated smears of gastric aspirates and homogenized lymphatic tissue were examined by fluorescence microscopy, and positive results were confirmed by the Kinyoun method.6 Of each specimen two solid cultures on Stonebrink and Loewenstein-Jensen medium as well as one liquid culture in Middlebrok 7H9 with OADC enrichment were inoculated.7 Cultures were incubated at 35°C for up to 8 weeks, and MOTT culture isolates were submitted to the German Reference Center for Mycobacteria for species identification.
DNA preparation. Under sterile conditions lymphatic tissue was cut into small pieces and homogenized by using a conical tissue homogenizer (Wheaton®; Neolab/Migge Laborbedarf, Heidelberg, Germany). Macroscopic particles were pelleted by centrifugation, and nucleic acids were extracted with 100 μl of the supernatant directly and at 1/100 dilution. After centrifugation at maximal speed for 15 min, lysis of mycobacteria and purification of nucleic acids was performed according to a modification (omission of the lysozyme treatment) of the protocol published by Smith et al.8 Gastric aspirates were submitted to the standard NALC-NaOH decontamination procedure, and sediments were processed as described above.
PCR. One-half of the purified DNA was submitted to a novel PCR assay using primers p694 5′-TCG AGC ACC TCC AGC AAG GC-3′ and p794 5′-TCA CTA CCG AGA GGA ACA TC-3′, developed in our laboratory. The primer sequences were derived from the published sequence of the insertion sequence IS12459 (Accession No. L33879). The specificity of the selected primers for M. avium was demonstrated with the use of purified DNA of 60 mycobacterial isolates and type strains including species most frequently encountered in mycobacterial lymphadenitis Mycobacterium intracellulare, Mycobacterium malmoense, Mycobacterium scrofulaceum, Mycobacterium kansasii and Mycobacterium fortuitum (data not shown). The amplification was performed under standard conditions for 35 cycles of 1.5 min at 94°C, 1.75 min at 60°C and 2.0 min at 72°C. The resulting PCR products were analyzed by gel electrophoresis on an 8% polyacrylamide gel (Sigma-Aldrich, Deisenhofen, Germany). After transfer to a nylon membrane specificity of the amplified fragment was demonstrated by hybridization with a fluorescein-labeled diagnostic oligonucleotide p894 5′-GCG ATG GCC CTG GAC CAG TC-3′ at 68°C by using the ECL oligolabeling and detection kit (Amersham Buchler, Braunschweig, Germany) as recommended by the manufacturer. Mock specimens and negative controls were included in all steps of DNA extraction and amplification to recognize false positive results caused by laboratory cross-contamination.
DNA fingerprint analysis. Extraction of genomic DNA from mycobacterial cultures and PvuII restriction fragment length polymorphism was performed as described.9 The IS1245-specific probe was created by amplification with primers p694 and p794.
Case Report 1. An 11-month-old girl presented with 1 day of fever of 39°C and enlarged cervical lymph nodes. The prior medical history was unremarkable with frequent upper airway infections but no serious illnesses. On admission the examination revealed a bilateral enlargement of the cervical lymph nodes and a slight pharyngitis. Laboratory tests showed a high sedimentation rate of 80 mm, leukocytosis of 24 800/mm3 and an elevated C-reactive protein of 1.3 mg/dl. The tuberculin skin test was positive with 0.8 cm induration. The child had not received Bacillus Calmette-Guérin vaccination. Serology for lymphotropic viruses including cytomegalovirus, Epstein-Barr virus and HIV were negative. A chest radiograph showed no pathology, and ultrasound examination of the enlarged lymph nodes demonstrated multiple lymph nodes without signs of necrolysis or calcifications. Treatment with cefuroxime (100 mg/kg) did not result in clinical or sonographic changes. Thus gastric aspirates were obtained, and the macroscopically affected lymphatic tissue was surgically excised and submitted to histopathologic, microbiologic and molecular genetic analysis. Histopathologic changes with signs of granulomatous infection suggested a tuberculous or mycobacterial lymphadenitis. The excised tissue and one of three gastric aspirates obtained before surgery were positive by mycobacterial culture and PCR for M. avium. At discharge the child was given clarithromycin because of bilateral involvement. One year after the surgical treatment the child remains disease-free.
Case Report 2. A 17-month-old boy was admitted with a 3-week history of left cervical lymphadenitis that was unresponsive to oral amoxicillin. The tuberculin skin test showed an induration of 1 cm. The clinical examination was unremarkable besides grossly enlarged, nontender lymph nodes on the left side high up in the neck. A chest radiograph was normal, and ultrasound examination demonstrated bilateral enlarged lymph nodes with some calcification. With the exception of a moderately elevated sedimentation rate laboratory values were within normal limits. To rule out extrapulmonary tuberculosis three gastric aspirates were obtained. The enlarged lymph nodes were surgically removed. PCR analysis of one gastric aspirate was positive for M. avium; however, cultures remained negative for mycobacteria. The clinical diagnosis of mycobacterial lymphadenitis was confirmed by granulomatous changes on histopathology, PCR and culture of the excised tissue which grew M. avium. In spite of postoperative treatment with azithromycin, recurrence of the lymphadenitis required a second surgical intervention to remove remaining infected tissue. The further clinical course was unremarkable.
During a 24-month period 13 children were treated at the Children's Hospital of Heidelberg University for suspected mycobacterial lymphadenitis. The diagnosis could be confirmed by positive histopathology and culture isolation of the pathogen for 7 children between 11 months and 4.8 years of age. Acid-fast microscopy of excised lymph nodes was positive in 5 of these children. In 1 child who showed abscess formation Staphylococcus aureus was isolated from lymphatic tissue. Catscratch disease was proved by positive serology in a 14-year-old boy. In spite of typical granulomatous disease, culture and acid-fast microscopy remained negative in the other 4 children. A summary of the results is presented in Table 1.
Analysis of the excised lymphatic tissue by a novel PCR assay for M. avium was positive in four cases, three of which were also culture-positive (Table 1). PCR results were available while the children were still in the hospital, whereas cultures results were positive after 2 to 4 weeks.
All PCR results were confirmed by hybridization with a diagnostic oligonucleotide. Mock samples were included into every step of the sample preparation to control for false positive results. In addition strict adherence to separation of sample preparation, PCR setup and amplification areas further reduced the risk of carryover contamination. The performance of the PCR assay was monitored by inclusion of positive amplification controls. However, false negative results caused by inhibition of the PCR reaction could not be excluded given that no internal control was introduced into individual samples.
In two children infection with M. avium was diagnosed before surgical intervention by PCR of gastric aspirates. The diagnosis was confirmed by histopathology, acid-fast microscopy, PCR and culture of the excised lymph nodes. In one child cultures of lymphatic tissue and gastric aspirate grew M. avium. Results of IS1245 DNA fingerprint analysis of these isolates showed identical patterns with multiple bands (Fig. 1), thus proving clonality of the isolates from both sources. The gastric aspirate of a third child with positive findings on histopathologic examination was positive by the M. avium PCR assay. However, culture of both lymphatic tissue as well as gastric aspirate remained negative.
Cervical lymphadenitis caused by MOTT is an increasing clinical problem. Usually the diagnosis of mycobacterial lymphadenitis is confirmed by surgical excision and analysis of the affected lymph nodes. The analysis of the excised tissue by a novel PCR assay allowed early diagnosis of M. avium while the children were recovering from surgery. One additional case of M. avium lymphadenitis could be diagnosed based on PCR and histopathology results that was missed by mycobacterial culture. This is in good agreement with results for detection of pulmonary tuberculosis in children reported by Smith et al.8 that show an increased diagnostic sensitivity when results of culture and PCR are combined.
PCR analysis was almost twice as sensitive as culture in detecting M. avium from lymphatic tissue and gastric aspirates (seven PCR positives compared with four culture positives). The high sensitivity of the assay is based on the repetitive nature of the amplified target sequence IS1245.9 The high specificity for M. avium seems to be an obvious disadvantage of this method; however, M. avium complex is reported to be the most frequently isolated MOTT causing cervical lymphadenitis in children1, 4 and was responsible for one-half of confirmed cases in our study. The number of cases caused by the second species within the M. avium complex, Mycobacterium intracellulare, remains to be determined.
In two cases PCR allowed the detection of M. avium in gastric aspirates before surgery with culture-proved mycobacterial lymphadenitis. In immunocompetent children with primary tuberculosis, isolation of M. tuberculosis from gastric aspirates has been observed even in the absence of pulmonary involvement.10, 11 Therefore the detection of M. avium in gastric aspirates in both children suggested a pathogenetic role of the isolated organism. The positive PCR of gastric aspirate in a third child with granulomatous lymphadenitis suggests M. avium to be the causative agent. However, the lack of culture confirmation from the lymphatic tissue does not allow us to rule out colonization or contamination with the organism.
DNA fingerprint analysis based on the number and genomic location of repetitive elements has proved to be a reliable tool for the study of tuberculosis by identification of specific strains.12-16 Recently a similar technique has been described for the analysis of M. avium based on the insertion sequence IS1245.9 In our study, for one child culture isolates from gastric aspirate and lymphatic tissue were available (Patient 5 in Table 1). The isolates could be proved to be clonal by IS1245 DNA fingerprint analysis. To our knowledge this is the first report providing direct evidence for a pathogenetic link between detection of MOTT in gastric aspirates and mycobacterial lymphadenitis in immunocompetent children. No changes in the chest radiographic results or indication for internal fistula formation was detected in either child.
Thus far the port of entry of cervical lymphadenitis in children caused by MOTT has proved elusive. Only rarely did histopathologic examination of the tonsils reveal granulomatous inflammation as indication of the initial focus.4 Our results might indicate that colonization of the mucous membranes with MOTT represents the first step of mycobacterial lymphadenitis in young children. Because a correlation between upper respiratory infections and MOTT disease has been observed,4 inflammation of the respiratory tract may foster the infection in colonized children. However, MOTT are found ubiquitously in the environment, and the rate of colonization in immunocompetent healthy children is unknown. Recently a prospective study of 106 children with cystic fibrosis demonstrated colonization with MOTT in 6.6%,17 but only rapid growing mycobacteria were isolated. Even though a retrospective analysis of adult patients with pulmonary tuberculosis showed isolation of M. avium complex in 11%,18 cultures of gastric aspirates from children with suspected pulmonary tuberculosis usually are negative for MOTT. Because frequently enlarged lymph nodes are detected by palpation and ultrasound examination bilaterally, an alternative explanation for the detection of the causative mycobacterium outside of the main disease focus might reflect systemic migration of infected macrophages.
As demonstrated by the second case, a considerable number of relapses occur because of incomplete excision.4, 5 Residual infected tissue is not uncommon because of frequent involvement of deep cervical vascular or neural structures. In addition follow-up studies of surgical intervention show permanent damage of the facial nerve in some cases,5 and in small children the long term cosmetic results are not always satisfying. Promising results have been obtained with clarithromycin and rifabutin for treatment of mycobacterial lymphadenitis caused by MOTT.19 With this in mind our results suggest that analysis of gastric aspirates by culture and molecular genetic techniques should be performed before surgery in suspected mycobacterial lymphadenitis. Confirmation of our results in a large number of children might provide a rational basis for an alternative approach to diagnosis and treatment of nontuberculous disease in immunocompetent children.
This work was supported by Deutsche Forschungsgemeinschaft Grant 1921/3-1,2.
1. Grange JM, Yates MD, Pozniak A. Bacteriologically confirmed non-tuberculous mycobacterial lymphadenitis in south east England: a recent increase in the number of cases. Arch Dis Child 1995;72:516-7.
2. Kuth G, Lamprecht J, Haase G. Cervical lymphadenitis due to mycobacteria other than tuberculosis: an emerging problem in children? ORL J Otorhinolaryngol Relat Spec 1995;57:36-8.
3. Wallace RJ Jr, Cook JL, Glassroth J, Griffith DE, Olivier KN, Gordin F. Diagnosis and treatment of disease caused by nontuberculous mycobacteria: American Thoracic Society statement. Am J Respir Crit Care Med 1997;156:S1-25.
4. Wolinsky E. Mycobacterial lymphadenitis in children: a prospective study of 105 nontuberculous cases with long-term follow-up. Clin Infect Dis 1995;20:954-63.
5. Wright JE. Non-tuberculous mycobacterial lymphadenitis. Aust NZ J Surg 1996;66:225-8.
6. Smithwick RW. Laboratory manual for acid-fast microscopy. 2nd ed. Atlanta, GA: Centers for Disease Control, 1976:9-10.
7. Kent PT, Kubica GP. Public health mycobacteriology: a guide for the level III laboratory. Atlanta, GA: US Department of Health and Human Services, Public Health Service, 1985.
8. Smith KC, Starke JR, Eisenach KD, Ong LT, Denby M. Detection of Mycobacterium tuberculosis
in clinical specimen from children using a polymerase chain reaction. Pediatrics 1996;97:155-60.
9. Guerrero C, Bernasconi C, Burki D, Bodmer T, Telenti A. A novel insertion element from Mycobacterium avium, IS1245
, is a specific target for analysis of strain relatedness. J Clin Microbiol 1995;33:304-7.
10. Pierre C, Olivier C, Lecossier D, Boussougant Y, Yeni P, Hance AJ. Diagnosis of primary tuberculosis in children by amplification and detection of mycobacterial DNA. Am Rev Respir Dis 1993;147:420-4.
11. Khan EA, Starke JR. Diagnosis of tuberculosis in children: increased need for better methods. Emerging Infect Dis 1995;1:115-23.
12. van Embden JDA, Cave MD, Crawford JT, et al. Strain identification of Mycobacterium tuberculosis
by DNA finger-printing: recommendations for a standardized methodology. J Clin Microbiol 1993;31:406-9.
13. Small PM, Shafer RW, Hopewell PC, et al. Exogenous reinfection with multidrug-resistant Mycobacterium tuberculosis
in patients with advanced HIV infection. N Engl J Med 1993;328:1137-44.
14. Theisen A, Reichel C, Rüsch-Gerdes S, et al. Mixed-strain infection with a drug-sensitive and multidrug-resistant strain of Mycobacterium tuberculosis.
15. Horn DL, Hewlett D Jr., Haas WH, et al. Superinfection with rifampicin-isoniazid-streptomycin-ethambutol (RISE)-resistant tuberculosis in three patients with AIDS: confirmation by polymerase chain reaction fingerprinting. Ann Intern Med 1994;121:115-6.
16. Edlin BR, Tokars JI, Grieco MH, et al. An outbreak of multidrug-resistant tuberculosis among hospitalized patients with the acquired immunodeficiency syndrome. N Engl J Med 1992;326:1514-21.
17. Fauroux B, Delaisi B, Clément A, et al. Mycobacterial lung disease in cystic fibrosis: a prospective study. Pediatr Infect Dis J 1997;16:354-8.
18. Epstein MD, Aranda CP, Bonk S, Hanna B, Rom WN. The significance of Mycobacterium avium
complex cultivation in the sputum of patients with pulmonary tuberculosis. Chest 1997;111:142-7.
19. Berger C, Pfyffer GE, Nadal D. Treatment of nontuberculous mycobacterial lymphadenitis with clarithromycin plus rifabutin. J Pediatr 1996;128:383-6.
THE BURTIS BURR BREESE AWARD
This award is in the memory of Burtis Burr Breese, M.D., who died in 1998, to celebrate his contributions to pediatric teaching and clinical practice. The epidemiology, diseases and care of patients within his daily practice were sources of continual fascination, curiosity and a belief that private practice offered a unique opportunity for controlled research to answer questions that could contribute to the understanding and care of pediatric diseases. He also believed that such endeavors brought constant stimulation, quality and enjoyment to the routine of pediatric practice. Furthermore, with a facility of language and a pleasure in writing, he shared the results and his thoughts in his many published studies spanning a half-century. Through these studies of infectious diseases, especially streptoccal infections, Dr. Breese pioneered office-based research.
This award is presented yearly by the Pediatric Infectious Diseases Society to the author of a paper published in The Pediatric Infectious Disease Journal® that best illustrates the principles and practices of Dr. Burtis Burr Breese.