Bacterial Lymphadenitis at a Major Referral Hospital in France from 2008 to 2012
Updated information and services can be found at: http://jcm.asm.org/content/52/4/1161 These include: REFERENCES
CONTENT ALERTS
This article cites 30 articles, 11 of which can be accessed free at: http://jcm.asm.org/content/52/4/1161#ref-list-1 Receive: RSS Feeds, eTOCs, free email alerts (when new articles cite this article), more»
Information about commercial reprint orders: http://journals.asm.org/site/misc/reprints.xhtml To subscribe to to another ASM Journal go to: http://journals.asm.org/site/subscriptions/
Downloaded from http://jcm.asm.org/ on July 8, 2014 by UNIVERSITY OF SHEFFIELD LIBRARY
Marion Safont, Emmanouil Angelakis, Hervé Richet, Hubert Lepidi, Pierre-Edouard Fournier, Michel Drancourt and Didier Raoult J. Clin. Microbiol. 2014, 52(4):1161. DOI: 10.1128/JCM.03491-13. Published Ahead of Print 29 January 2014.
Bacterial Lymphadenitis at a Major Referral Hospital in France from 2008 to 2012 Marion Safont, Emmanouil Angelakis, Hervé Richet, Hubert Lepidi, Pierre-Edouard Fournier, Michel Drancourt, Didier Raoult
Lymph node enlargement is a common medical problem, and in a large number of patients, the causes of lymphadenopathy remain undiagnosed. We report a thorough microbiological analysis of 1,688 lymph node biopsy specimens collected in our bartonellosis reference center. We studied lymph node biopsy samples from patients with suspected regional infectious lymph node enlargement from January 2008 to December 2012. To evaluate a useful strategy for the diagnosis of infectious lymphadenitis, specimens were cultured and subjected to molecular assays. Histologic analysis was done when possible. A total of 642 (38%) biopsy specimens were infected with a bacterial agent, and quantitative PCR (qPCR) was significantly better than 16S rRNA gene PCR (rrs) for the detection of Bartonella henselae (P ⴝ 0.05), Mycobacterium tuberculosis (P ⴝ 0.05), and Mycobacterium avium (P ⴝ 0.007). Molecular assays were significantly better than bacterial cultures for the diagnosis of Francisella tularensis (P ⴝ 0.017) but were less effective for detecting M. tuberculosis (P ⴝ 0.004) and M. avium (P ⴝ 0.001). Histologic analysis was done for 412 lymph nodes, and 20% of these were compatible with an infectious lymphadenitis, whereas a neoplasm was found in 29% of these lymph nodes. M. tuberculosis was detected significantly more in female than in male patients (P ⴝ 0.01), and patients with cat scratch disease (CSD) were younger than patients with M. tuberculosis, Tropheryma whipplei, and F. tularensis. Negative rrs PCR does not exclude the diagnosis of infectious lymphadenitis. Histologic analysis of lymph node biopsy specimens is critical, as a diagnosis of infectious lymphadenitis does not preclude other concurrent diseases.
E
nlarged lymph nodes are a common symptom in a number of infectious, autoimmune, and malignant diseases (1). The clinical symptoms are often similar or nonspecific, and several etiological agents could be responsible for enlarged lymph nodes (1). A variety of bacterial, viral, fungal, and protozoal agents may serve as a focal point for subsequent clinical investigation (2). Bartonella henselae, the primary causative agent of cat scratch disease (CSD), is the most common organism responsible for infectious lymphadenopathy in adults and children (2, 3). CSD is a seasonal disease (4), and live B. henselae was recently found at the primary inoculation sites of patients after a cat scratch (5). Mycobacteria commonly cause lymphadenopathy, and peripheral tuberculous lymphadenitis accounts for approximately 10% of tuberculosis cases in the United States (6). The incidence of lymphadenitis caused by nontuberculous mycobacteria has increased in recent decades, primarily in children (7). Staphylococcus aureus and Streptococcus pyogenes are responsible for approximately 40 to 80% of cases of acute unilateral cervical lymphadenitis in children (8). Francisella tularensis, the causative agent of tularemia, and Tropheryma whipplei, the causative agent of Whipple’s disease, are less common sources of infectious lymphadenopathy (2, 3). Excisional biopsy specimens are routinely performed to obtain lymph node material for appropriate laboratory studies and histologic examination, and such biopsy specimens are usually required for a definitive diagnosis (9). Cultures are routinely ordered on lymph nodes biopsy specimens, but there are recognized limitations to this technique. Moreover, some etiological agents are not easily isolated using conventional methods and require specialized cell culture systems. Recently, we proposed a molecular strategy based on real-time quantitative PCR (qPCR) for the detection of the most commonly encountered causative agents of infectious lymphadenopathy (3). This
April 2014 Volume 52 Number 4
qPCR assay is significantly more sensitive than 16S rRNA gene PCR amplification and sequencing for the detection of B. henselae, F. tularensis, and mycobacteria (3). 16S rRNA gene PCR is still important for the detection of other new or emerging bacterial agents (10). In this study, we analyzed a large collection of lymph nodes using molecular techniques and bacterial cultures. Our objective was to evaluate a useful strategy for the diagnosis of infectious lymphadenitis. MATERIALS AND METHODS Samples. We studied lymph node biopsy samples from patients with suspected regional infectious lymph node enlargement that were sent to our laboratory from January 2008 to December 2012. As a national reference center for rickettsioses and bartonellosis, we routinely receive lymph node biopsy specimens from patients with suspected CSD throughout France. Moreover, we receive lymph node biopsy specimens from hospitalized patients and outpatients with suspected regional infectious lymphadenitis of our hospital. We received a fragment or the entire lymph node. Samples were sent to our laboratory in sterile conditions, frozen (⫺20°C) or on dry ice (⫺80°C). In our laboratory, all samples were handled under sterile conditions at ⫺20°C to avoid cross-contaminations and were analyzed within 12 h upon receipt. Lymph nodes fixed in formalin were also received for histologic analysis for some patients. Ordinarily, a standardized
Received 16 December 2013 Returned for modification 10 January 2014 Accepted 18 January 2014 Published ahead of print 29 January 2014 Editor: D. J. Diekema Address correspondence to Didier Raoult, [email protected]. M.S. and E.A. contributed equally to this work. Copyright © 2014, American Society for Microbiology. All Rights Reserved. doi:10.1128/JCM.03491-13
Journal of Clinical Microbiology
p. 1161–1167
jcm.asm.org
1161
Downloaded from http://jcm.asm.org/ on July 8, 2014 by UNIVERSITY OF SHEFFIELD LIBRARY
Unité de Recherche sur les Maladies Infectieuses et Tropicales Emergentes, URMITE CNRS-IRD 198 UMR 6236, Université de la Méditerranée, Faculté de Médecine et de Pharmacie, Marseille, France
Safont et al.
TABLE 1 Primers and probesa Primer or probe name
Gene
Target
Human beta actin Human beta actin Human beta actin 16S-23S rRNA Bartonella 16S-23S rRNA Bartonella 16S-23S rRNA Bartonella pap31 B. henselae pap31 B. henselae pap31 B. henselae yopP B. quintana yopP B. quintana yopP B. quintana hsp60 B. alsatica hsp60 B. alsatica B. alsatica hsp60 WiSP family protein-encoding gene T. whipplei WiSP family protein-encoding gene T. whipplei WiSP family protein-encoding gene T. whipplei nhAd F. tularensis nhAd F. tularensis nhAd F. tularensis Hypothetical protein-encoding C. burnetii gene IS30a 3R Hypothetical protein-encoding C. burnetii gene IS30a F3-R3 P Hypothetical protein-encoding C. burnetii gene ITSd 16S-23S rRNA Mycobacteria ITSr 16S-23S rRNA Mycobacteria Tuberculosis probe 16S-23S rRNA M. tuberculosis Avium probe 16S-23S rRNA M. avium
PCR amplification and sequencing MycoF MycoR 536F Rp2 a
rpob rpob 16S RNA 16S RNA
Mycobacteria Mycobacteria Mycobacteria Mycobacteria
Sequence CATGCCATCCTGCGTCTGGA CCGTGGCCATCTCTTGCTCG FAM-CGGGAAATCGTGCGTGACATTAAG GATGCCGGGGAAGGTTTTC GCCTGGGAGGACTTGAACCT FAM-GCGCGCGCTTGATAAGCGTG TATGCCTTATGTTGCTGGTGGT ACCACCGCCAAGAGTGAAAC FAM-CAAGCAGCAGATGATGCAGAAATCGC TAAACCTCGGGGGAAGCAGA TTTCGTCCTCAACCCCATCA FAM- CGTTGCCGACAAGACGTCCTTGC TGCTAACGCTATGGAAAAAGTTG CCACGATCAAACTGCATTCC FAM-GTCGAAGAAGCAAAAACGGCTGAAACC TGAGGATGTATCTGTGTATGGGACA TCCTGTTACAAGCAGTACAAAACAAA FAM- GAGAGATGGGGTGCAGGACAGGG GGCCTTGGGGGTAGCTTATT CCCTAGTGCAATTACCCAAGTCC FAM-GCAGCTGGAGTTGCCGTGATGG CGCTGACCTACAGAAATATGTCC
172
104
151
134
108
100
120
164
11 11 11 This study This study This study This study This study This study 13 13 13 19 19 19 28 28 28 This study This study This study 29
GGGGTAAGTAAATAATACCTTCTGG
29
FAM-CATGAAGCGATTTATCAATACGTGTATGC
29
GGGTGGGGTGTGGTGTTTGA CAAGGCATCCACCATGCGC FAM- GCTAGCCGGCAGCGTATCCAT FAM- GGCCGGCGTTCATCGAAAT -Mgb
144
14 14 14 14
GGCAAGGTCACCCCGAAGGG AGCGGCTGCTGGGTGATCATC CAGCAGCCGCGGTAATAC ACGGCTACCTTGTTACGACTT
751
14 14 10 10
994
FAM, 6-carboxyfluorescein.
questionnaire with the gender and age of the patients and the suspected infection was received for each lymph node. Strategy. The quality of DNA extraction of samples was verified by a housekeeping gene encoding beta-actin (11). All samples were tested by qPCR for the presence of F. tularensis, T. whipplei, and Coxiella burnetii at the species level and for Bartonella species and mycobacteria at the genus level. A second qPCR assay for characterization of the species level was used in cases of positive results for Bartonella sp. or mycobacteria. Independently, all lymph nodes were screened for the presence of bacteria by PCR amplification and sequencing, targeting the 16S rRNA gene (rrs). Lymph nodes were cultured for epidemiological reasons when F. tularensis or T. whipplei were suspected based on a patient’s clinical manifestation or when molecular assays were positive for these agents. Similarly, due to the very low isolation rate previously published, lymph nodes infected with Bartonella species were not systematically cultured (2, 12). Lymph nodes were routinely cultured when mycobacteria were suspected or when molecular assays were positive. Moreover, lymph nodes were cultured in cases of negative molecular assays and poor-quality DNA extraction. Patients were classified as definitely having an infected lymph node if
1162
jcm.asm.org
there was direct evidence of infection using culture or molecular assays or when a specific agent was detected by histopathology. Molecular assays. The total genomic DNA was extracted from samples using a QIAamp tissue kit (Qiagen, Hilden, Germany). The samples were handled under sterile conditions to avoid cross-contamination. The genomic DNA was stored at 4°C and used as a template in the qPCR assays. For the detection of Bartonella species, we used primers and probes targeting the 16S-23S RNA intergenic region (ITS), for B. henselae, we used primers and probes targeting the pap31 gene, and for F. tularensis, we used primers and probes targeting the yqaB gene, as previously described (Table 1) (3). For lymph nodes positive only at the genus level for Bartonella (ITS positive; pap31 negative), the samples were tested by qPCR for Bartonella quintana and Bartonella alsatica, as previously described (13). We used primers and probes targeting the ITS region for the detection of mycobacteria (14). In cases of a positive result, characterization at the species level was performed using a probe for M. tuberculosis sp. complex and another probe for M. avium complex, as previously described (14). For the nymph nodes positive at the genus level but negative for the M. tuberculosis sp. complex and M. avium complex, PCR amplification and
Journal of Clinical Microbiology
Downloaded from http://jcm.asm.org/ on July 8, 2014 by UNIVERSITY OF SHEFFIELD LIBRARY
qPCR Actine_F Actine_R Actine_P Barto ITS3 F Barto ITS3 R Barto ITS3 P PAP 246F PAP 396R PAP246/396_MBP B qui 11580F B qui 11580R B qui 11580P B_alsa_hsp60_F B_alsa_hsp60_R B_alsa_hsp60_P T_whi2_F T_whi2_R T_whi2_P F_tul1728_F F_tul1728_R F_tul1728_P IS30a 3F
Amplicon (no. of Reference bases) or source
Diagnosis of Bacterial Lymphadenitis
RESULTS
We tested 1,688 lymph node biopsy specimens from 1,688 patients in 5 years. The median age ⫾ intraquartile range (IR) was 30 ⫾ 21 years (range, 2 days to 95 years), and 965 (57%) were males and 723 (43%) were females. In total, 642 (38%) lymph nodes were classified as infected by a bacterial agent, and 389 (60%) were from male and 253 (40%) were from female patients. The median age ⫾ IR of the patients with infected lymph nodes was 28 ⫾ 22 years (range, 2 months to 90 years). Molecular assays. By qPCR, we found 340 lymph nodes (53%) infected with B. henselae. One lymph node was ITS positive and pap31 negative and was identified as B. alsatica (19). M. tuberculosis was detected in 42 (6%) patients, M. avium in 27 (4%) patients, Mycobacterium genavense in 2 (0.3%) patients, and Mycobacterium bolletti in 2 (0.3%) patients. We detected 23 lymph nodes (4%) infected with F. tularensis and 7 (1%) infected with T. whipplei. A total of 98% of Bartonella species-, 44% of mycobac-
April 2014 Volume 52 Number 4
terium-, 100% of F. tularensis-, and 100% of T. whipplei-infected lymph nodes were detected by qPCR. The amplification of the rrs gene was positive for 324 patients (50%; Table 2), and B. henselae was the most frequently amplified bacterium (189 cases, 58%). Other common bacteria identified by rrs PCR were staphylococci (52 cases, including 32 S. aureus), streptococci (15 cases, including 5 Streptococcus pyogenes), F. tularensis (18 cases), mycobacteria (6 cases), and Pseudomonas aeruginosa (5 cases). rrs PCR amplification and sequencing compared to qPCR was positive for 56% of Bartonella species-, 8% of mycobacterium-, and 78% of F. tularensis-positive lymph nodes. No lymph nodes infected by T. whipplei were found. qPCR was significantly better than rrs PCR amplification and sequencing for the detection of B. henselae (P ⫽ 0.05) and mycobacteria (P ⫽ 0.01), whereas no difference was found for the detection of F. tularensis (P ⫽ 0.2) (Table 2). Moreover, qPCR was significantly better than rrs PCR amplification and sequencing for the detection of M. tuberculosis (P ⫽ 0.05) and M. avium (P ⫽ 0.007), whereas no difference was found for the other mycobacteria. Culture. Bacteria were isolated from 182 (28%) specimens, and mycobacteria were the most frequently recovered organisms (62 cases, 34%). Other common bacteria recovered by culture were S. aureus (24 lymph nodes, 13%) and Staphylococcus epidermidis (19 lymph nodes, 10%). B. henselae was cultured and successfully passaged from a single lymph node. We isolated 1 isolate of Francisella tularensis and 1 isolate of T. whipplei by the shell vial cell culture. The anaerobic bacteria cultured from lymph nodes included Propionibacterium acnes (33 cases, 18%), Fusobacterium necrophorum (1 sample, 0.5%), and Fusobacterium naviforme (1 sample, 0.5%). Culture was positive for 83% of mycobacterium-, 58% of staphylococcus-, and 33% of streptococcus-positive lymph nodes. qPCR was significantly better than the culture for the diagnosis of F. tularensis (P ⫽ 0.017) (Table 2). The culture was significantly better than rrs PCR amplification and sequencing for P. acnes (P ⫽ 0.003) and S. epidermidis (P ⫽ 0.003). Moreover, the culture was significantly better than qPCR for the detection of M. tuberculosis (P ⫽ 0.004) and M. avium (P ⫽ 0.001). No significant difference was found between the culture and molecular assays for the other bacterial agents. Histologic analysis. Histologic analysis was done for 412 lymph nodes (Table 3). Most of these lymph nodes (194, 20%) were inflammatory, compatible with an infectious lymphadenitis, whereas a neoplasm was found in 119 lymph nodes (29%). Histologic analysis was not conclusive for 49 lymph nodes (12%) and was normal for 59 lymph nodes (14%). By bacteriological assays, 27 (14%) lymph nodes compatible with an infectious lymphadenitis were positive, including 13 CSD cases (positive PCR), 6 M. tuberculosis cases (3 positive by PCR and culture and 3 positive only by PCR), 4 S. aureus cases (both positive PCR and culture), 3 T. whipplei (positive PCR), 1 Pseudomonas aeruginosa (positive culture), and 1 Streptococcus pyogenes (both positive PCR and culture). Five lymph nodes (4%) with neoplasm were also positive for an infectious agent (4 cases were B. henselae positive by PCR and 1 S. aureus case was positive by PCR and culture). The malignancy for the lymph nodes infected by CSD was determined as a nonHodgkin lymphoma. Bacteriological assays showed that 4 lymph nodes had CSD not conclusive by histologic analysis; one was also infected by an Actinomyces sp. (positive by both PCR and culture), one was infected by Stenotrophomonas maltophilia (positive by
jcm.asm.org 1163
Downloaded from http://jcm.asm.org/ on July 8, 2014 by UNIVERSITY OF SHEFFIELD LIBRARY
sequencing targeting the ITS region were performed as previously described. The PCR amplification and sequencing targeting the rrs gene were performed using the methods previously described (15). All the sequences were compared with those available in the GenBank, EMBL, and DJB databases using the gapped BLASTN 2.0.5 program through the National Center for Biotechnology Information server. The quality of DNA extraction was verified by reverse transcription (RT)-PCR for a housekeeping gene encoding beta-actin (11). Culture. For the isolation of B. henselae or T. whipplei, samples were cultured in human embryonic lung (HEL) fibroblasts to obtain a confluent monolayer. For the culture, we used the centrifugation shell vial technique (Sterilin, Feltham, England; 3.7 ml) with 12-mm round coverslips seeded with 1 ml of medium containing 50,000 cells and incubation in a 5% CO2 incubator at 37°C for 3 days. The cultures were surveyed, and the bacterial growth was assessed every 7 days on coverslips directly inside the shell vial using Gimenez and immunofluorescence staining (11). When the staining method was positive, the isolates were identified using PCR and sequencing as described above. For the isolation of common bacteria, the samples were placed on 5% sheep blood, chocolate, Mueller-Hinton, Trypticase soy, and MacConkey agar media (bioMérieux, France) and incubated at 37°C in aerobic, microaerophilic, and anaerobic atmospheres. The isolates were identified using matrix-assisted laser desorption ionization–time of flight (MALDI-TOF) as previously described (16). For the isolation of mycobacteria, the samples were cultured in MGIT tubes (Becton, Dickinson, Pont-De-Claix, France) or on a homemade 5% sheep blood agar (bioMérieux, La Balme-les-Grottes, France) for 2 to 45 days at 32°C or 37°C, as previously described (17). Gram staining was used to ensure the absence of any contaminant organisms in the culture, and standard identification of mycobacteria was performed using MALDI-TOF, as previously described (18). For the isolation of F. tularensis, samples were cultured in chocolate agar. Histologic analysis. Lymph nodes fixed in formalin were tested for histologic analysis. Stains used included Gram, hematoxylin and eosin, periodic acid-Schiff, Ziehl-Neelsen, and Warthin-Starry. Statistical analysis. When the distribution was not normal, one-way analysis of variance (ANOVA) and the Kruskal-Wallis test were used to compare the lymph node groups. Linear regression with forward selection was used to estimate the association of bacterial agents of lymphadenitis with sex and age. The analyses were performed using SPSS v20.0 (IBM, Paris, France) and XLSTAT v12 (Addinsoft, Paris, France) software. For data comparison between different assays, the Student t test or 2 test was performed by using EpiInfo version 6.0 software (Centers for Disease Control and Prevention, Atlanta, GA, USA). A P value of ⬍0.05 was considered significant.
Safont et al.
TABLE 2 Results of molecular assays, culture, and histologic analysis for the tested lymph nodes No. (%) of positive lymph nodesc by: Species
No. of positive lymph nodes
Specific PCR
16S rRNA
Culture
Histologic analysis
CSD
Bartonella henselae Bartonella alsatica
340 1
332 (97.6) 1 (100)
189 (56)a 1 (100)
1 (0.3) 0
21 (6) 0
Mycobacteria
Mycobacterium tuberculosis Mycobacterium avium Mycobacterium genavense Mycobacterium bolletti
42 27 2 2
19 (45) 12 (44) 1 (50) 1 (50)
1 (2)a 4 (15)a 1 (50) 0
36 (86)b 22 (82)b 2 (100) 2 (100)
6 (14) 0 0 0
Staphylococci
Staphylococcus aureus Staphylococcus epidermidis Other coagulase-negative Staphylococcus
45 33 11
ND ND ND
32 (71) 18 (55) 2 (18)
24 (53) 19 (58) 9 (82)
5 (11) 0 0
Streptococci
Streptococcus pyogenes Streptococcus b Streptococcus c/d
Propionibacterium acnes Francisella tularensis Tropheryma whipplei Pseudomonas aeruginosa Escherichia coli Stenotrophomonas maltophilia Klebsiella pneumoniae Serratia marcescens Fusobacterium sp. Enterococcus sp. Prevotella sp. Actinomyces Fungi sp. Polymicrobial Other
6 2 10 34 23 7 6 3 3 2 2 3 2 3 2 2 10 19
ND ND ND ND 23 (100) 7 (100) ND ND ND ND ND ND ND ND ND ND ND ND
5 (83) 2 (100) 8 (80) 2 (6) 18 (78) 0 5 (83) 2 (67) 1 1 (50) 2 (100) 1 (33) 1 (50) 3 (100) 1 (50) 1 (50) 9 (90) 14 (74)
2 (33) 0 4 (40) 33 (97)b 1 (4)b 1 (14) 4 (67) 2 (67) 2 (67) 2 (100) 1 (50) 2 (67) 1 (50) 0 2 (100) 1 (50) 2 (20) 7 (37)
1 (17) 0 0 0 0 3 (43) 1 (17) 0 1 (33) 0 0 0 0 0 1 (50) 0 1 (10) 0
Total
642
399 (62)
324 (50)
182 (28)
P ⬍ 0.05, RT-PCR versus 16S rRNA. b P ⬍ 0.05, culture versus molecular assays. c ND, not done. d Exactly what type of histologic findings: CSD, necrotizing lymphadenitis with numerous microabcesses surrounded by inflammatory granulomas; tuberculosis, areas of central granular caseation surrounded by inflammatory granulomas; Whipple’s disease, enlarged lymph node sinuses with numerous foamy macrophages containing PAS-positive particles in their cytoplasm; S. aureus, S. pyogenes, P. aeruginosa, Stenotrophomonas, acute nonspecific lymphadenitis; Actinomyces, abscess formation with tangled mass of filaments surrounded by radiating organisms. a
both PCR and culture), and one was polymicrobial (positive by both PCR and culture). Regression analysis. M. tuberculosis was detected significantly more in female than in male patients (P ⫽ 0.01). No significant difference was found between males and females for the other
TABLE 3 Histologic analysis result for the lymph nodes fixed in formalin
Histologic findings
No. of positive lymph nodes by histologic analysis
No. (%) of positive bacteriological assays
Neoplasm Inflammatory lymphadenitis Normal Not specific
119 194 59 49
5 (3) 27 (14) 0 8 (16)
Total
412
49 (12)
1164
jcm.asm.org
bacterial agents. An association with age was found for Bartonella species, M. tuberculosis, T. whipplei, M. avium, S. epidermidis, and F. tularensis. The patients with CSD were significantly younger (median ⫾ IR, 27 ⫾ 17 years; range, 2 years to 90 years) than those in the non-CSD group (median ⫾ IR, 37 ⫾ 25 years; range, 2 months to 85 years) (P ⫽ 0.001) (Table 4). The patients with lymph nodes infected with M. tuberculosis were significantly older (median ⫾ IR, 47 ⫾ 21 years; range, 1.5 to 85 years) than the non-M. tuberculosis patients (median ⫾ IR, 30 ⫾ 21 years; range, 0.2 months to 90 years) (P ⫽ 0.001). The patients with lymph nodes infected with M. avium were significantly younger than the non-M. avium complex patients (median ⫾ IR, 6 ⫾ 15 years; range, 1.5 to 70 years) versus 33 ⫾ 21 years (range, 2 months to 90 years) (P ⫽ 0.001). The lymph nodes infected with F. tularensis were from significantly older patients (median ⫾ IR, 44 ⫾ 18 years; range, 14 to 73 years) than the non-F. tularensis patients (median ⫾ IR, 31 ⫾ 21 years; range, 2 months to 90 years) (P ⫽ 0.005). The lymph nodes infected with T. whipplei were from pa-
Journal of Clinical Microbiology
Downloaded from http://jcm.asm.org/ on July 8, 2014 by UNIVERSITY OF SHEFFIELD LIBRARY
Diagnosisd
Diagnosis of Bacterial Lymphadenitis
TABLE 4 Demographic data of patient groups Age (median ⫾ intraquartile range) in yrs Infection?
CSD
M. tuberculosis
M. avium
T. whipplei
F. tularensis
S. epidermidis
Yes No P value
26.9 ⫾ 17.3 36.5 ⫾ 24.6 0.001
46.6 ⫾ 20.8 30.4 ⫾ 21.2 0.001
6.3 ⫾ 15 32.6 ⫾ 21.1 0.001
490.5 ⫾ 12.3 31.2 ⫾ 20.8 0.04
43.9 ⫾ 18.1 31 ⫾ 21.5 0.005
40.1 ⫾ 20.3 30 ⫾ 21.7 0.001
DISCUSSION
In this extensive study on lymph nodes, we used molecular assays and cultures for the diagnosis of infectious lymphadenitis. Moreover, lymph nodes were tested by histologic analysis when possible. All of our molecular and culture assays have previously been evaluated (3, 11, 16, 17). Furthermore, we routinely include many negative and positive controls in each assay, and the quality of DNA extraction was verified for all samples. However, a limitation of our study was that we did not test for fungi or viruses that may also represent causes of lymph node enlargement. Moreover, we did not obtain histologic data for all patients because most of the samples sent to our center were frozen or too small. In our series of lymph nodes, neoplasms represented the 7% of cases of suspected CSD. However, this percentage could be higher, as we tested only 24% of the lymph nodes by histology analysis. In addition, a limitation was that we could analyze our results by whether a fraction or the whole of the lymph node was obtained, as the entire lymph node would be more likely to yield a pathogen. To the best of our knowledge, our study testing lymph nodes is the largest described to date (Table 5). However, comparisons between studies are difficult, and the context of the laboratory biases, eras, and different patient populations inherent to each
particular study should be considered. In our study, B. henselae was found primarily in younger patients and was the most common cause of infectious lymphadenitis. Our results confirmed previous studies of our laboratory (2, 3). A bias in our study is that our laboratory is a reference center for Q fever, rickettsioses, and bartonellosis. This was the reason that all lymph nodes were evaluated by qPCR specifically for B. henselae, F. tularensis, T. whipplei, and C. burnetii. Moreover, the higher percentage of positive B. henselae results in our studies was because specimens were from patients with suspected CSD. Similarly, Ridder et al. found B. henselae as the most common cause of infectious lymphadenitis in patients with unclear masses in the head and neck (20). Before the discovery of B. henselae, which explains its absence from the studies during the 1980s (2) and the era of molecular assays, Mycobacterium spp. was the most common cause of infectious lymphadenitis (21–23). In our study, M. tuberculosis was the third most common agent of infectious lymphadenitis after S. aureus. A similar percentage for M. tuberculosis was reported by Doberneck (21). In contrast, Anthony et al. (22) and Freidig et al. (24) found a lower percentage for M. tuberculosis. Moreover, Rosado et al. (25) found two cases of Yersinia species lymphadenitis in an inguinal and a mesenteric lymph node. Mycobacterial infection was identified in 11% of the positive lymph nodes. Compared to our previous study (2), we detected more mycobacterial agents, most likely because lymph nodes are now routinely cultured regardless of the suspected diagnosis. The qPCR was better for the diagnosis of mycobacterial infection than rrs gene amplification. However, our molecular assays were less effective compared to culture for the diagnosis of mycobacterial infection. Despite the long incubation time needed for the isolation of mycobacterial strains, culture remains critical, as it presented better sensitivity for the detection
TABLE 5 Previous studies of bacterial lymphadenitis Diagnostic method used
Commonly detected agents of infectious lymphadenitis (% positive cases)a
Culture
PCR
1st agent
2nd agent
3rd agent
4th agent
S. aureus (4) B. henselae (1) S. epidermidis M. tuberculosis (3) S. pyogenes (2) M. tuberculosis (3) Mycobacteria (7)
ND ND P. acnes M. kansasii (0.7) M. tuberculosis (1) Streptococci (0.5) S. aureus (2)
S. aureus (4) B. henselae (20) S. aureus (7)
ND T. gondii (0.4) S. viridans M. avium complex (2) S. aureus (1) B. henselae (0.7) Coagulase-negative staphylococci (3) F. tularensis (3) Yersiniosis (7) Coagulase-negative staphylococci (7)
Author (yr) (reference)
No. of patients
Doberneck (1983) (21) Anthony et al. (1983) (22) Roberts et al. (1984) (23) Freidig et al. (1986) (24) Ridder et al. (2002) (20) Chau et al. (2003) (30) Rolain et al. (2006) (2)
169 228 163 419 454 423 786
Yes
No
Yes Yes Yes Yes Yes
No No Yes No Yes
M. tuberculosis (6) M. tuberculosis (3) Mycobacteria (17) Histoplasma capsulatum (4) B. henselae (13) T. gondii (4) B. henselae (31)
Angelakis et al. (2009) (3) Rosado et al. (2011) (25) This study
491 368 1,688
No Yes Yes
Yes Yes Yes
B. henselae (30) M. avium complex (60) B. henselae (33)
a
M. tuberculosis (1) F. tularensis (3) M. tuberculosis (7)
ND, not described.
April 2014 Volume 52 Number 4
jcm.asm.org 1165
Downloaded from http://jcm.asm.org/ on July 8, 2014 by UNIVERSITY OF SHEFFIELD LIBRARY
tients significantly older (median ⫾ IR, 50 ⫾ 12 years; range, 35 to 66 years) than the non-T. whipplei patients (median ⫾ IR, 31 ⫾ 21 years; range, 2 months to 90 years) (P ⫽ 0.04). The patients with lymph nodes infected with S. epidermidis were significantly older (median ⫾ IR, 40 ⫾ 20 years; range, 2 months to 79 years) than the non-S. epidermidis patients (median ⫾ IR, 30 ⫾ 22 years; range, 3 months to 90 years) (P ⫽ 0.001). No significant difference was found for the other bacterial agents.
Safont et al.
22.
REFERENCES
23.
1. Fijten GH, Blijham GH. 1988. Unexplained lymphadenopathy in family practice. An evaluation of the probability of malignant causes and the effectiveness of physicians’ workup. J. Fam. Pract. 27:373–376. 2. Rolain JM, Lepidi H, Zanaret M, Triglia JM, Michel G, Thomas PA, Texereau M, Stein A, Romaru A, Eb F, Raoult D. 2006. Lymph node
1166
jcm.asm.org
3. 4. 5. 6. 7. 8. 9. 10. 11.
12.
13. 14.
15.
16.
17. 18.
19. 20. 21.
24.
biopsy specimens and diagnosis of cat-scratch disease. Emerg. Infect. Dis. 12:1338 –1344. http://dx.doi.org/10.3201/eid1209.060122. Angelakis E, Roux V, Raoult D, Rolain JM. 2009. Real-time PCR strategy and detection of bacterial agents of lymphadenitis. Eur. J. Clin. Microbiol. Infect. Dis. 28:1363–1368. http://dx.doi.org/10.1007/s10096-009-0793-6. Sanguinetti-Morelli D, Angelakis E, Richet H, Davoust B, Rolain JM, Raoult D. 2011. Seasonality of cat-scratch disease, France, 1999 –2009. Emerg. Infect. Dis. 17:705–707. http://dx.doi.org/10.3201/eid1704.100825. Angelakis E, Edouard S, La Scola B, Raoult D. 2010. Bartonella henselae in skin biopsy specimens of patients with cat-scratch disease. Emerg. Infect. Dis. 16:1963–1965. http://dx.doi.org/10.3201/eid1612.100647. Fontanilla JM, Barnes A, von Reyn CF. 2011. Current diagnosis and management of peripheral tuberculous lymphadenitis. Clin. Infect. Dis. 53:555–562. http://dx.doi.org/10.1093/cid/cir454. Donald PR. 2010. The chemotherapy of tuberculous lymphadenopathy in children. Tuberculosis 90:213–224. http://dx.doi.org/10.1016/j.tube.2010 .05.001. Kelly CS, Kelly RE, Jr. 1998. Lymphadenopathy in children. Pediatr. Clin. North Am. 45:875– 888. http://dx.doi.org/10.1016/S0031-3955(05)70051-1. Grossman M, Shiramizu B. 1994. Evaluation of lymphadenopathy in children. Curr. Opin. Pediatr. 6:68 –76. http://dx.doi.org/10.1097/00008480 -199402000-00012. Weisburg WG, Barns SM, Pelletier DA, Lane DJ. 1991. 16S ribosomal DNA amplification for phylogenetic study. J. Bacteriol. 173:697–703. Angelakis E, Richet H, Rolain JM, La Scola B, Raoult D. 2012. Comparison of real-time quantitative PCR and culture for the diagnosis of emerging Rickettsioses. PLoS Negl. Trop. Dis. 6:e1540. http://dx.doi.org /10.1371/journal.pntd.0001540. Rolain JM, Gouriet F, Enea M, Aboud M, Raoult D. 2003. Detection by immunofluorescence assay of Bartonella henselae in lymph nodes from patients with cat scratch disease. Clin. Diagn. Lab. Immunol. 10:686 – 691. http://dx.doi.org/10.1128/CDLI.10.4.686-691.2003. Angelakis E, Rolain JM, Raoult D, Brouqui P. 2011. Bartonella quintana in head louse nits. FEMS Immunol. Med. Microbiol. 62:244 –246. http: //dx.doi.org/10.1111/j.1574-695X.2011.00804.x. Bruijnesteijn Van Coppenraet ES, Lindeboom JA, Prins JM, Peeters MF, Claas EC, Kuijper EJ. 2004. Real-time PCR assay using fine-needle aspirates and tissue biopsy specimens for rapid diagnosis of mycobacterial lymphadenitis in children. J. Clin. Microbiol. 42:2644 –2650. http://dx.doi .org/10.1128/JCM.42.6.2644-2650.2004. Drancourt M, Bollet C, Carlioz A, Martelin R, Gayral JP, Raoult D. 2000. 16S ribosomal DNA sequence analysis of a large collection of environmental and clinical unidentifiable bacterial isolates. J. Clin. Microbiol. 38:3623–3630. Seng P, Drancourt M, Gouriet F, La Scola B, Fournier PE, Rolain JM, Raoult D. 2009. Ongoing revolution in bacteriology: routine identification of bacteria by matrix-assisted laser desorption ionization time-offlight mass spectrometry. Clin. Infect. Dis. 49:543–551. http://dx.doi.org /10.1086/600885. Drancourt M, Raoult D. 2007. Cost-effectiveness of blood agar for isolation of mycobacteria. PLoS Negl. Trop. Dis. 1:e83. http://dx.doi.org/10 .1371/journal.pntd.0000083. El Khéchine A, Couderc C, Flaudrops C, Raoult D, Drancourt M. 2011. Matrix-assisted laser desorption/ionization time-of-flight mass spectrometry identification of mycobacteria in routine clinical practice. PLoS One 6:e24720. http://dx.doi.org/10.1371/journal.pone.0024720. Angelakis E, Lepidi H, Canel A, Rispal P, Perraudeau F, Barre I, Rolain JM, Raoult D. 2008. Human case of Bartonella alsatica lymphadenitis. Emerg. Infect. Dis. 14:1951–1953. http://dx.doi.org/10.3201/eid1412.080757. Ridder GJ, Boedeker CC, Technau-Ihling K, Grunow R, Sander A. 2002. Role of cat-scratch disease in lymphadenopathy in the head and neck. Clin. Infect. Dis. 35:643– 649. http://dx.doi.org/10.1086/342058. Doberneck RC. 1983. The diagnostic yield of lymph node biopsy. Arch. Surg. 118:1203–1205. http://dx.doi.org/10.1001/archsurg.1983.01390100067017. Anthony PP, Knowles SA. 1983. Lymphadenopathy as a primary presenting sign: a clinicopathological study of 228 cases. Br. J. Surg. 70:412– 414. http://dx.doi.org/10.1002/bjs.1800700708. Roberts FJ, Linsey S. 1984. The value of microbial cultures in diagnostic lymph-node biopsy. J. Infect. Dis. 149:162–165. http://dx.doi.org/10.1093 /infdis/149.2.162. Freidig EE, McClure SP, Wilson WR, Banks PM, Washington JA. 1986. Clinical-histologic-microbiologic analysis of 419 lymph node biopsy specimens. Rev. Infect. Dis. 8:322–328. http://dx.doi.org/10.1093/clinids/8.3.322.
Journal of Clinical Microbiology
Downloaded from http://jcm.asm.org/ on July 8, 2014 by UNIVERSITY OF SHEFFIELD LIBRARY
of these agents. To increase the sensitivity for the diagnosis of mycobacterial lymphadenitis, a combination of various procedures with different sensitivities has been proposed (6). In our study, we combined the molecular assays with culture, and on the basis of these results, we will try to design a more effective new qPCR assay specifically for the detection of mycobacteria. Fastidious organisms were the cause of infectious adenitis, and we found that qPCR was the referent assay for the diagnosis of these agents. The rrs amplification and sequencing was less effective for the detection of B. henselae and F. tularensis than specific qPCR, whereas the detection rate for T. whipplei was low. Culture was not effective for the diagnosis of fastidious organisms, because their isolation is often difficult and depends on the number of microorganisms in the cells, which should be as high as possible and on the centrifugation step, which enhances the adhesion of bacteria that are freed from their intracellular location in the cells in culture (26). In previous studies, we observed bacteria by direct immunofluorescence in many lymph nodes that were negative by culture, which suggests that bacteria in lymph nodes are not viable (12). Moreover, in our previously published series of lymph nodes, we found a very low isolation rate for B. henselae from lymph nodes (2). Based on these results, we do not routinely culture lymph nodes positive for B. henselae. The isolation of fastidious organisms from samples remains critical for the description of new species, enabling the genetic descriptions, physiological analyses, improvement in diagnostic tools, and antibiotic susceptibility testing of bacteria (11, 27). Histologic analysis of lymph nodes is critical, as we found four lymph nodes with CSD and a malignancy. Ridder et al. also reported 2 cases with squamous cell carcinoma and 2 cases with malignant B-cell lymphoma in patients with high antibody titers to B. henselae (20). In this extensive lymph node study, we showed that lymph node analysis by the combination of rrs PCR amplification and sequencing with specific qPCR and culture is critical for the diagnosis of bacterial lymphadenitis. Staphylococci, streptococci, and other aerobic and anaerobic bacteria were detected as agents of infectious adenitis. rrs PCR remains critical for the detection of these bacteria (10). Although the efficacy of rrs PCR for the detection of nonfastidious agents varied and was dependent on the bacteria species, no significant difference was found between culture and the rrs PCR. To increase the sensitivity of culture, lymph node biopsy specimens should be sampled before treatment early in the course of the disease and should be inoculated as soon as possible. However, we cannot provide data about the patient treatment at the time of the lymph node biopsy. Physicians should be aware that in cases of negative rrs PCR, an infection by fastidious organisms and mycobacteria is not excluded given the low sensitivity of eubacterial PCR. CSD and M. avium infection occur mostly in young patients, whereas patients with M. tuberculosis are significantly older. Moreover, we reinforced the need that lymph node excision and histologic analysis are critical for accurate diagnosis, as neoplasms could be clinically misdiagnosed as CSD when only molecular assays are used.
Diagnosis of Bacterial Lymphadenitis
April 2014 Volume 52 Number 4
28. Edouard S, Fenollar F, Raoult D. 2012. The rise of Tropheryma whipplei: a 12-year retrospective study of PCR diagnoses in our reference center. J. Clin. Microbiol. 50:3917–3920. http://dx.doi.org/10.1128/JCM.01517-12. 29. Eldin C, Angelakis E, Renvoise A, Raoult D. 2013. Coxiella burnetii DNA, but not viable bacteria, in dairy products in France. Am. J. Trop. Med. Hyg. 88:765–769. http://dx.doi.org/10.4269/ajtmh.12-0212. 30. Chau I, Kelleher MT, Cunningham D, Norman AR, Wotherspoon A, Trott P, Rhys-Evans P, Querci Della RG, Brown G, Allen M, Waters JS, Haque S, Murray T, Bishop L. 2003. Rapid access multidisciplinary lymph node diagnostic clinic: analysis of 550 patients. Br. J. Cancer 88: 354 –361. http://dx.doi.org/10.1038/sj.bjc.6600738.
jcm.asm.org 1167
Downloaded from http://jcm.asm.org/ on July 8, 2014 by UNIVERSITY OF SHEFFIELD LIBRARY
25. Rosado FG, Stratton CW, Mosse CA. 2011. Clinicopathologic correlation of epidemiologic and histopathologic features of pediatric bacterial lymphadenitis. Arch. Pathol. Lab. Med. 135:1490 –1493. http://dx.doi.org /10.5858/arpa.2010-0581-OA. 26. Gouriet F, Fenollar F, Patrice JY, Drancourt M, Raoult D. 2005. Use of shell-vial cell culture assay for isolation of bacteria from clinical specimens: 13 years of experience. J. Clin. Microbiol. 43:4993–5002. http://dx .doi.org/10.1128/JCM.43.10.4993-5002.2005. 27. La Scola B, Raoult D. 1997. Laboratory diagnosis of rickettsioses: current approaches to the diagnosis of old and new rickettsial diseases. J. Clin. Microbiol. 35:2715–2727.
Updated information and services can be found at: http://jcm.asm.org/content/52/4/1161 These include: REFERENCES
CONTENT ALERTS
This article cites 30 articles, 11 of which can be accessed free at: http://jcm.asm.org/content/52/4/1161#ref-list-1 Receive: RSS Feeds, eTOCs, free email alerts (when new articles cite this article), more»
Information about commercial reprint orders: http://journals.asm.org/site/misc/reprints.xhtml To subscribe to to another ASM Journal go to: http://journals.asm.org/site/subscriptions/
Downloaded from http://jcm.asm.org/ on July 8, 2014 by UNIVERSITY OF SHEFFIELD LIBRARY
Marion Safont, Emmanouil Angelakis, Hervé Richet, Hubert Lepidi, Pierre-Edouard Fournier, Michel Drancourt and Didier Raoult J. Clin. Microbiol. 2014, 52(4):1161. DOI: 10.1128/JCM.03491-13. Published Ahead of Print 29 January 2014.
Bacterial Lymphadenitis at a Major Referral Hospital in France from 2008 to 2012 Marion Safont, Emmanouil Angelakis, Hervé Richet, Hubert Lepidi, Pierre-Edouard Fournier, Michel Drancourt, Didier Raoult
Lymph node enlargement is a common medical problem, and in a large number of patients, the causes of lymphadenopathy remain undiagnosed. We report a thorough microbiological analysis of 1,688 lymph node biopsy specimens collected in our bartonellosis reference center. We studied lymph node biopsy samples from patients with suspected regional infectious lymph node enlargement from January 2008 to December 2012. To evaluate a useful strategy for the diagnosis of infectious lymphadenitis, specimens were cultured and subjected to molecular assays. Histologic analysis was done when possible. A total of 642 (38%) biopsy specimens were infected with a bacterial agent, and quantitative PCR (qPCR) was significantly better than 16S rRNA gene PCR (rrs) for the detection of Bartonella henselae (P ⴝ 0.05), Mycobacterium tuberculosis (P ⴝ 0.05), and Mycobacterium avium (P ⴝ 0.007). Molecular assays were significantly better than bacterial cultures for the diagnosis of Francisella tularensis (P ⴝ 0.017) but were less effective for detecting M. tuberculosis (P ⴝ 0.004) and M. avium (P ⴝ 0.001). Histologic analysis was done for 412 lymph nodes, and 20% of these were compatible with an infectious lymphadenitis, whereas a neoplasm was found in 29% of these lymph nodes. M. tuberculosis was detected significantly more in female than in male patients (P ⴝ 0.01), and patients with cat scratch disease (CSD) were younger than patients with M. tuberculosis, Tropheryma whipplei, and F. tularensis. Negative rrs PCR does not exclude the diagnosis of infectious lymphadenitis. Histologic analysis of lymph node biopsy specimens is critical, as a diagnosis of infectious lymphadenitis does not preclude other concurrent diseases.
E
nlarged lymph nodes are a common symptom in a number of infectious, autoimmune, and malignant diseases (1). The clinical symptoms are often similar or nonspecific, and several etiological agents could be responsible for enlarged lymph nodes (1). A variety of bacterial, viral, fungal, and protozoal agents may serve as a focal point for subsequent clinical investigation (2). Bartonella henselae, the primary causative agent of cat scratch disease (CSD), is the most common organism responsible for infectious lymphadenopathy in adults and children (2, 3). CSD is a seasonal disease (4), and live B. henselae was recently found at the primary inoculation sites of patients after a cat scratch (5). Mycobacteria commonly cause lymphadenopathy, and peripheral tuberculous lymphadenitis accounts for approximately 10% of tuberculosis cases in the United States (6). The incidence of lymphadenitis caused by nontuberculous mycobacteria has increased in recent decades, primarily in children (7). Staphylococcus aureus and Streptococcus pyogenes are responsible for approximately 40 to 80% of cases of acute unilateral cervical lymphadenitis in children (8). Francisella tularensis, the causative agent of tularemia, and Tropheryma whipplei, the causative agent of Whipple’s disease, are less common sources of infectious lymphadenopathy (2, 3). Excisional biopsy specimens are routinely performed to obtain lymph node material for appropriate laboratory studies and histologic examination, and such biopsy specimens are usually required for a definitive diagnosis (9). Cultures are routinely ordered on lymph nodes biopsy specimens, but there are recognized limitations to this technique. Moreover, some etiological agents are not easily isolated using conventional methods and require specialized cell culture systems. Recently, we proposed a molecular strategy based on real-time quantitative PCR (qPCR) for the detection of the most commonly encountered causative agents of infectious lymphadenopathy (3). This
April 2014 Volume 52 Number 4
qPCR assay is significantly more sensitive than 16S rRNA gene PCR amplification and sequencing for the detection of B. henselae, F. tularensis, and mycobacteria (3). 16S rRNA gene PCR is still important for the detection of other new or emerging bacterial agents (10). In this study, we analyzed a large collection of lymph nodes using molecular techniques and bacterial cultures. Our objective was to evaluate a useful strategy for the diagnosis of infectious lymphadenitis. MATERIALS AND METHODS Samples. We studied lymph node biopsy samples from patients with suspected regional infectious lymph node enlargement that were sent to our laboratory from January 2008 to December 2012. As a national reference center for rickettsioses and bartonellosis, we routinely receive lymph node biopsy specimens from patients with suspected CSD throughout France. Moreover, we receive lymph node biopsy specimens from hospitalized patients and outpatients with suspected regional infectious lymphadenitis of our hospital. We received a fragment or the entire lymph node. Samples were sent to our laboratory in sterile conditions, frozen (⫺20°C) or on dry ice (⫺80°C). In our laboratory, all samples were handled under sterile conditions at ⫺20°C to avoid cross-contaminations and were analyzed within 12 h upon receipt. Lymph nodes fixed in formalin were also received for histologic analysis for some patients. Ordinarily, a standardized
Received 16 December 2013 Returned for modification 10 January 2014 Accepted 18 January 2014 Published ahead of print 29 January 2014 Editor: D. J. Diekema Address correspondence to Didier Raoult, [email protected]. M.S. and E.A. contributed equally to this work. Copyright © 2014, American Society for Microbiology. All Rights Reserved. doi:10.1128/JCM.03491-13
Journal of Clinical Microbiology
p. 1161–1167
jcm.asm.org
1161
Downloaded from http://jcm.asm.org/ on July 8, 2014 by UNIVERSITY OF SHEFFIELD LIBRARY
Unité de Recherche sur les Maladies Infectieuses et Tropicales Emergentes, URMITE CNRS-IRD 198 UMR 6236, Université de la Méditerranée, Faculté de Médecine et de Pharmacie, Marseille, France
Safont et al.
TABLE 1 Primers and probesa Primer or probe name
Gene
Target
Human beta actin Human beta actin Human beta actin 16S-23S rRNA Bartonella 16S-23S rRNA Bartonella 16S-23S rRNA Bartonella pap31 B. henselae pap31 B. henselae pap31 B. henselae yopP B. quintana yopP B. quintana yopP B. quintana hsp60 B. alsatica hsp60 B. alsatica B. alsatica hsp60 WiSP family protein-encoding gene T. whipplei WiSP family protein-encoding gene T. whipplei WiSP family protein-encoding gene T. whipplei nhAd F. tularensis nhAd F. tularensis nhAd F. tularensis Hypothetical protein-encoding C. burnetii gene IS30a 3R Hypothetical protein-encoding C. burnetii gene IS30a F3-R3 P Hypothetical protein-encoding C. burnetii gene ITSd 16S-23S rRNA Mycobacteria ITSr 16S-23S rRNA Mycobacteria Tuberculosis probe 16S-23S rRNA M. tuberculosis Avium probe 16S-23S rRNA M. avium
PCR amplification and sequencing MycoF MycoR 536F Rp2 a
rpob rpob 16S RNA 16S RNA
Mycobacteria Mycobacteria Mycobacteria Mycobacteria
Sequence CATGCCATCCTGCGTCTGGA CCGTGGCCATCTCTTGCTCG FAM-CGGGAAATCGTGCGTGACATTAAG GATGCCGGGGAAGGTTTTC GCCTGGGAGGACTTGAACCT FAM-GCGCGCGCTTGATAAGCGTG TATGCCTTATGTTGCTGGTGGT ACCACCGCCAAGAGTGAAAC FAM-CAAGCAGCAGATGATGCAGAAATCGC TAAACCTCGGGGGAAGCAGA TTTCGTCCTCAACCCCATCA FAM- CGTTGCCGACAAGACGTCCTTGC TGCTAACGCTATGGAAAAAGTTG CCACGATCAAACTGCATTCC FAM-GTCGAAGAAGCAAAAACGGCTGAAACC TGAGGATGTATCTGTGTATGGGACA TCCTGTTACAAGCAGTACAAAACAAA FAM- GAGAGATGGGGTGCAGGACAGGG GGCCTTGGGGGTAGCTTATT CCCTAGTGCAATTACCCAAGTCC FAM-GCAGCTGGAGTTGCCGTGATGG CGCTGACCTACAGAAATATGTCC
172
104
151
134
108
100
120
164
11 11 11 This study This study This study This study This study This study 13 13 13 19 19 19 28 28 28 This study This study This study 29
GGGGTAAGTAAATAATACCTTCTGG
29
FAM-CATGAAGCGATTTATCAATACGTGTATGC
29
GGGTGGGGTGTGGTGTTTGA CAAGGCATCCACCATGCGC FAM- GCTAGCCGGCAGCGTATCCAT FAM- GGCCGGCGTTCATCGAAAT -Mgb
144
14 14 14 14
GGCAAGGTCACCCCGAAGGG AGCGGCTGCTGGGTGATCATC CAGCAGCCGCGGTAATAC ACGGCTACCTTGTTACGACTT
751
14 14 10 10
994
FAM, 6-carboxyfluorescein.
questionnaire with the gender and age of the patients and the suspected infection was received for each lymph node. Strategy. The quality of DNA extraction of samples was verified by a housekeeping gene encoding beta-actin (11). All samples were tested by qPCR for the presence of F. tularensis, T. whipplei, and Coxiella burnetii at the species level and for Bartonella species and mycobacteria at the genus level. A second qPCR assay for characterization of the species level was used in cases of positive results for Bartonella sp. or mycobacteria. Independently, all lymph nodes were screened for the presence of bacteria by PCR amplification and sequencing, targeting the 16S rRNA gene (rrs). Lymph nodes were cultured for epidemiological reasons when F. tularensis or T. whipplei were suspected based on a patient’s clinical manifestation or when molecular assays were positive for these agents. Similarly, due to the very low isolation rate previously published, lymph nodes infected with Bartonella species were not systematically cultured (2, 12). Lymph nodes were routinely cultured when mycobacteria were suspected or when molecular assays were positive. Moreover, lymph nodes were cultured in cases of negative molecular assays and poor-quality DNA extraction. Patients were classified as definitely having an infected lymph node if
1162
jcm.asm.org
there was direct evidence of infection using culture or molecular assays or when a specific agent was detected by histopathology. Molecular assays. The total genomic DNA was extracted from samples using a QIAamp tissue kit (Qiagen, Hilden, Germany). The samples were handled under sterile conditions to avoid cross-contamination. The genomic DNA was stored at 4°C and used as a template in the qPCR assays. For the detection of Bartonella species, we used primers and probes targeting the 16S-23S RNA intergenic region (ITS), for B. henselae, we used primers and probes targeting the pap31 gene, and for F. tularensis, we used primers and probes targeting the yqaB gene, as previously described (Table 1) (3). For lymph nodes positive only at the genus level for Bartonella (ITS positive; pap31 negative), the samples were tested by qPCR for Bartonella quintana and Bartonella alsatica, as previously described (13). We used primers and probes targeting the ITS region for the detection of mycobacteria (14). In cases of a positive result, characterization at the species level was performed using a probe for M. tuberculosis sp. complex and another probe for M. avium complex, as previously described (14). For the nymph nodes positive at the genus level but negative for the M. tuberculosis sp. complex and M. avium complex, PCR amplification and
Journal of Clinical Microbiology
Downloaded from http://jcm.asm.org/ on July 8, 2014 by UNIVERSITY OF SHEFFIELD LIBRARY
qPCR Actine_F Actine_R Actine_P Barto ITS3 F Barto ITS3 R Barto ITS3 P PAP 246F PAP 396R PAP246/396_MBP B qui 11580F B qui 11580R B qui 11580P B_alsa_hsp60_F B_alsa_hsp60_R B_alsa_hsp60_P T_whi2_F T_whi2_R T_whi2_P F_tul1728_F F_tul1728_R F_tul1728_P IS30a 3F
Amplicon (no. of Reference bases) or source
Diagnosis of Bacterial Lymphadenitis
RESULTS
We tested 1,688 lymph node biopsy specimens from 1,688 patients in 5 years. The median age ⫾ intraquartile range (IR) was 30 ⫾ 21 years (range, 2 days to 95 years), and 965 (57%) were males and 723 (43%) were females. In total, 642 (38%) lymph nodes were classified as infected by a bacterial agent, and 389 (60%) were from male and 253 (40%) were from female patients. The median age ⫾ IR of the patients with infected lymph nodes was 28 ⫾ 22 years (range, 2 months to 90 years). Molecular assays. By qPCR, we found 340 lymph nodes (53%) infected with B. henselae. One lymph node was ITS positive and pap31 negative and was identified as B. alsatica (19). M. tuberculosis was detected in 42 (6%) patients, M. avium in 27 (4%) patients, Mycobacterium genavense in 2 (0.3%) patients, and Mycobacterium bolletti in 2 (0.3%) patients. We detected 23 lymph nodes (4%) infected with F. tularensis and 7 (1%) infected with T. whipplei. A total of 98% of Bartonella species-, 44% of mycobac-
April 2014 Volume 52 Number 4
terium-, 100% of F. tularensis-, and 100% of T. whipplei-infected lymph nodes were detected by qPCR. The amplification of the rrs gene was positive for 324 patients (50%; Table 2), and B. henselae was the most frequently amplified bacterium (189 cases, 58%). Other common bacteria identified by rrs PCR were staphylococci (52 cases, including 32 S. aureus), streptococci (15 cases, including 5 Streptococcus pyogenes), F. tularensis (18 cases), mycobacteria (6 cases), and Pseudomonas aeruginosa (5 cases). rrs PCR amplification and sequencing compared to qPCR was positive for 56% of Bartonella species-, 8% of mycobacterium-, and 78% of F. tularensis-positive lymph nodes. No lymph nodes infected by T. whipplei were found. qPCR was significantly better than rrs PCR amplification and sequencing for the detection of B. henselae (P ⫽ 0.05) and mycobacteria (P ⫽ 0.01), whereas no difference was found for the detection of F. tularensis (P ⫽ 0.2) (Table 2). Moreover, qPCR was significantly better than rrs PCR amplification and sequencing for the detection of M. tuberculosis (P ⫽ 0.05) and M. avium (P ⫽ 0.007), whereas no difference was found for the other mycobacteria. Culture. Bacteria were isolated from 182 (28%) specimens, and mycobacteria were the most frequently recovered organisms (62 cases, 34%). Other common bacteria recovered by culture were S. aureus (24 lymph nodes, 13%) and Staphylococcus epidermidis (19 lymph nodes, 10%). B. henselae was cultured and successfully passaged from a single lymph node. We isolated 1 isolate of Francisella tularensis and 1 isolate of T. whipplei by the shell vial cell culture. The anaerobic bacteria cultured from lymph nodes included Propionibacterium acnes (33 cases, 18%), Fusobacterium necrophorum (1 sample, 0.5%), and Fusobacterium naviforme (1 sample, 0.5%). Culture was positive for 83% of mycobacterium-, 58% of staphylococcus-, and 33% of streptococcus-positive lymph nodes. qPCR was significantly better than the culture for the diagnosis of F. tularensis (P ⫽ 0.017) (Table 2). The culture was significantly better than rrs PCR amplification and sequencing for P. acnes (P ⫽ 0.003) and S. epidermidis (P ⫽ 0.003). Moreover, the culture was significantly better than qPCR for the detection of M. tuberculosis (P ⫽ 0.004) and M. avium (P ⫽ 0.001). No significant difference was found between the culture and molecular assays for the other bacterial agents. Histologic analysis. Histologic analysis was done for 412 lymph nodes (Table 3). Most of these lymph nodes (194, 20%) were inflammatory, compatible with an infectious lymphadenitis, whereas a neoplasm was found in 119 lymph nodes (29%). Histologic analysis was not conclusive for 49 lymph nodes (12%) and was normal for 59 lymph nodes (14%). By bacteriological assays, 27 (14%) lymph nodes compatible with an infectious lymphadenitis were positive, including 13 CSD cases (positive PCR), 6 M. tuberculosis cases (3 positive by PCR and culture and 3 positive only by PCR), 4 S. aureus cases (both positive PCR and culture), 3 T. whipplei (positive PCR), 1 Pseudomonas aeruginosa (positive culture), and 1 Streptococcus pyogenes (both positive PCR and culture). Five lymph nodes (4%) with neoplasm were also positive for an infectious agent (4 cases were B. henselae positive by PCR and 1 S. aureus case was positive by PCR and culture). The malignancy for the lymph nodes infected by CSD was determined as a nonHodgkin lymphoma. Bacteriological assays showed that 4 lymph nodes had CSD not conclusive by histologic analysis; one was also infected by an Actinomyces sp. (positive by both PCR and culture), one was infected by Stenotrophomonas maltophilia (positive by
jcm.asm.org 1163
Downloaded from http://jcm.asm.org/ on July 8, 2014 by UNIVERSITY OF SHEFFIELD LIBRARY
sequencing targeting the ITS region were performed as previously described. The PCR amplification and sequencing targeting the rrs gene were performed using the methods previously described (15). All the sequences were compared with those available in the GenBank, EMBL, and DJB databases using the gapped BLASTN 2.0.5 program through the National Center for Biotechnology Information server. The quality of DNA extraction was verified by reverse transcription (RT)-PCR for a housekeeping gene encoding beta-actin (11). Culture. For the isolation of B. henselae or T. whipplei, samples were cultured in human embryonic lung (HEL) fibroblasts to obtain a confluent monolayer. For the culture, we used the centrifugation shell vial technique (Sterilin, Feltham, England; 3.7 ml) with 12-mm round coverslips seeded with 1 ml of medium containing 50,000 cells and incubation in a 5% CO2 incubator at 37°C for 3 days. The cultures were surveyed, and the bacterial growth was assessed every 7 days on coverslips directly inside the shell vial using Gimenez and immunofluorescence staining (11). When the staining method was positive, the isolates were identified using PCR and sequencing as described above. For the isolation of common bacteria, the samples were placed on 5% sheep blood, chocolate, Mueller-Hinton, Trypticase soy, and MacConkey agar media (bioMérieux, France) and incubated at 37°C in aerobic, microaerophilic, and anaerobic atmospheres. The isolates were identified using matrix-assisted laser desorption ionization–time of flight (MALDI-TOF) as previously described (16). For the isolation of mycobacteria, the samples were cultured in MGIT tubes (Becton, Dickinson, Pont-De-Claix, France) or on a homemade 5% sheep blood agar (bioMérieux, La Balme-les-Grottes, France) for 2 to 45 days at 32°C or 37°C, as previously described (17). Gram staining was used to ensure the absence of any contaminant organisms in the culture, and standard identification of mycobacteria was performed using MALDI-TOF, as previously described (18). For the isolation of F. tularensis, samples were cultured in chocolate agar. Histologic analysis. Lymph nodes fixed in formalin were tested for histologic analysis. Stains used included Gram, hematoxylin and eosin, periodic acid-Schiff, Ziehl-Neelsen, and Warthin-Starry. Statistical analysis. When the distribution was not normal, one-way analysis of variance (ANOVA) and the Kruskal-Wallis test were used to compare the lymph node groups. Linear regression with forward selection was used to estimate the association of bacterial agents of lymphadenitis with sex and age. The analyses were performed using SPSS v20.0 (IBM, Paris, France) and XLSTAT v12 (Addinsoft, Paris, France) software. For data comparison between different assays, the Student t test or 2 test was performed by using EpiInfo version 6.0 software (Centers for Disease Control and Prevention, Atlanta, GA, USA). A P value of ⬍0.05 was considered significant.
Safont et al.
TABLE 2 Results of molecular assays, culture, and histologic analysis for the tested lymph nodes No. (%) of positive lymph nodesc by: Species
No. of positive lymph nodes
Specific PCR
16S rRNA
Culture
Histologic analysis
CSD
Bartonella henselae Bartonella alsatica
340 1
332 (97.6) 1 (100)
189 (56)a 1 (100)
1 (0.3) 0
21 (6) 0
Mycobacteria
Mycobacterium tuberculosis Mycobacterium avium Mycobacterium genavense Mycobacterium bolletti
42 27 2 2
19 (45) 12 (44) 1 (50) 1 (50)
1 (2)a 4 (15)a 1 (50) 0
36 (86)b 22 (82)b 2 (100) 2 (100)
6 (14) 0 0 0
Staphylococci
Staphylococcus aureus Staphylococcus epidermidis Other coagulase-negative Staphylococcus
45 33 11
ND ND ND
32 (71) 18 (55) 2 (18)
24 (53) 19 (58) 9 (82)
5 (11) 0 0
Streptococci
Streptococcus pyogenes Streptococcus b Streptococcus c/d
Propionibacterium acnes Francisella tularensis Tropheryma whipplei Pseudomonas aeruginosa Escherichia coli Stenotrophomonas maltophilia Klebsiella pneumoniae Serratia marcescens Fusobacterium sp. Enterococcus sp. Prevotella sp. Actinomyces Fungi sp. Polymicrobial Other
6 2 10 34 23 7 6 3 3 2 2 3 2 3 2 2 10 19
ND ND ND ND 23 (100) 7 (100) ND ND ND ND ND ND ND ND ND ND ND ND
5 (83) 2 (100) 8 (80) 2 (6) 18 (78) 0 5 (83) 2 (67) 1 1 (50) 2 (100) 1 (33) 1 (50) 3 (100) 1 (50) 1 (50) 9 (90) 14 (74)
2 (33) 0 4 (40) 33 (97)b 1 (4)b 1 (14) 4 (67) 2 (67) 2 (67) 2 (100) 1 (50) 2 (67) 1 (50) 0 2 (100) 1 (50) 2 (20) 7 (37)
1 (17) 0 0 0 0 3 (43) 1 (17) 0 1 (33) 0 0 0 0 0 1 (50) 0 1 (10) 0
Total
642
399 (62)
324 (50)
182 (28)
P ⬍ 0.05, RT-PCR versus 16S rRNA. b P ⬍ 0.05, culture versus molecular assays. c ND, not done. d Exactly what type of histologic findings: CSD, necrotizing lymphadenitis with numerous microabcesses surrounded by inflammatory granulomas; tuberculosis, areas of central granular caseation surrounded by inflammatory granulomas; Whipple’s disease, enlarged lymph node sinuses with numerous foamy macrophages containing PAS-positive particles in their cytoplasm; S. aureus, S. pyogenes, P. aeruginosa, Stenotrophomonas, acute nonspecific lymphadenitis; Actinomyces, abscess formation with tangled mass of filaments surrounded by radiating organisms. a
both PCR and culture), and one was polymicrobial (positive by both PCR and culture). Regression analysis. M. tuberculosis was detected significantly more in female than in male patients (P ⫽ 0.01). No significant difference was found between males and females for the other
TABLE 3 Histologic analysis result for the lymph nodes fixed in formalin
Histologic findings
No. of positive lymph nodes by histologic analysis
No. (%) of positive bacteriological assays
Neoplasm Inflammatory lymphadenitis Normal Not specific
119 194 59 49
5 (3) 27 (14) 0 8 (16)
Total
412
49 (12)
1164
jcm.asm.org
bacterial agents. An association with age was found for Bartonella species, M. tuberculosis, T. whipplei, M. avium, S. epidermidis, and F. tularensis. The patients with CSD were significantly younger (median ⫾ IR, 27 ⫾ 17 years; range, 2 years to 90 years) than those in the non-CSD group (median ⫾ IR, 37 ⫾ 25 years; range, 2 months to 85 years) (P ⫽ 0.001) (Table 4). The patients with lymph nodes infected with M. tuberculosis were significantly older (median ⫾ IR, 47 ⫾ 21 years; range, 1.5 to 85 years) than the non-M. tuberculosis patients (median ⫾ IR, 30 ⫾ 21 years; range, 0.2 months to 90 years) (P ⫽ 0.001). The patients with lymph nodes infected with M. avium were significantly younger than the non-M. avium complex patients (median ⫾ IR, 6 ⫾ 15 years; range, 1.5 to 70 years) versus 33 ⫾ 21 years (range, 2 months to 90 years) (P ⫽ 0.001). The lymph nodes infected with F. tularensis were from significantly older patients (median ⫾ IR, 44 ⫾ 18 years; range, 14 to 73 years) than the non-F. tularensis patients (median ⫾ IR, 31 ⫾ 21 years; range, 2 months to 90 years) (P ⫽ 0.005). The lymph nodes infected with T. whipplei were from pa-
Journal of Clinical Microbiology
Downloaded from http://jcm.asm.org/ on July 8, 2014 by UNIVERSITY OF SHEFFIELD LIBRARY
Diagnosisd
Diagnosis of Bacterial Lymphadenitis
TABLE 4 Demographic data of patient groups Age (median ⫾ intraquartile range) in yrs Infection?
CSD
M. tuberculosis
M. avium
T. whipplei
F. tularensis
S. epidermidis
Yes No P value
26.9 ⫾ 17.3 36.5 ⫾ 24.6 0.001
46.6 ⫾ 20.8 30.4 ⫾ 21.2 0.001
6.3 ⫾ 15 32.6 ⫾ 21.1 0.001
490.5 ⫾ 12.3 31.2 ⫾ 20.8 0.04
43.9 ⫾ 18.1 31 ⫾ 21.5 0.005
40.1 ⫾ 20.3 30 ⫾ 21.7 0.001
DISCUSSION
In this extensive study on lymph nodes, we used molecular assays and cultures for the diagnosis of infectious lymphadenitis. Moreover, lymph nodes were tested by histologic analysis when possible. All of our molecular and culture assays have previously been evaluated (3, 11, 16, 17). Furthermore, we routinely include many negative and positive controls in each assay, and the quality of DNA extraction was verified for all samples. However, a limitation of our study was that we did not test for fungi or viruses that may also represent causes of lymph node enlargement. Moreover, we did not obtain histologic data for all patients because most of the samples sent to our center were frozen or too small. In our series of lymph nodes, neoplasms represented the 7% of cases of suspected CSD. However, this percentage could be higher, as we tested only 24% of the lymph nodes by histology analysis. In addition, a limitation was that we could analyze our results by whether a fraction or the whole of the lymph node was obtained, as the entire lymph node would be more likely to yield a pathogen. To the best of our knowledge, our study testing lymph nodes is the largest described to date (Table 5). However, comparisons between studies are difficult, and the context of the laboratory biases, eras, and different patient populations inherent to each
particular study should be considered. In our study, B. henselae was found primarily in younger patients and was the most common cause of infectious lymphadenitis. Our results confirmed previous studies of our laboratory (2, 3). A bias in our study is that our laboratory is a reference center for Q fever, rickettsioses, and bartonellosis. This was the reason that all lymph nodes were evaluated by qPCR specifically for B. henselae, F. tularensis, T. whipplei, and C. burnetii. Moreover, the higher percentage of positive B. henselae results in our studies was because specimens were from patients with suspected CSD. Similarly, Ridder et al. found B. henselae as the most common cause of infectious lymphadenitis in patients with unclear masses in the head and neck (20). Before the discovery of B. henselae, which explains its absence from the studies during the 1980s (2) and the era of molecular assays, Mycobacterium spp. was the most common cause of infectious lymphadenitis (21–23). In our study, M. tuberculosis was the third most common agent of infectious lymphadenitis after S. aureus. A similar percentage for M. tuberculosis was reported by Doberneck (21). In contrast, Anthony et al. (22) and Freidig et al. (24) found a lower percentage for M. tuberculosis. Moreover, Rosado et al. (25) found two cases of Yersinia species lymphadenitis in an inguinal and a mesenteric lymph node. Mycobacterial infection was identified in 11% of the positive lymph nodes. Compared to our previous study (2), we detected more mycobacterial agents, most likely because lymph nodes are now routinely cultured regardless of the suspected diagnosis. The qPCR was better for the diagnosis of mycobacterial infection than rrs gene amplification. However, our molecular assays were less effective compared to culture for the diagnosis of mycobacterial infection. Despite the long incubation time needed for the isolation of mycobacterial strains, culture remains critical, as it presented better sensitivity for the detection
TABLE 5 Previous studies of bacterial lymphadenitis Diagnostic method used
Commonly detected agents of infectious lymphadenitis (% positive cases)a
Culture
PCR
1st agent
2nd agent
3rd agent
4th agent
S. aureus (4) B. henselae (1) S. epidermidis M. tuberculosis (3) S. pyogenes (2) M. tuberculosis (3) Mycobacteria (7)
ND ND P. acnes M. kansasii (0.7) M. tuberculosis (1) Streptococci (0.5) S. aureus (2)
S. aureus (4) B. henselae (20) S. aureus (7)
ND T. gondii (0.4) S. viridans M. avium complex (2) S. aureus (1) B. henselae (0.7) Coagulase-negative staphylococci (3) F. tularensis (3) Yersiniosis (7) Coagulase-negative staphylococci (7)
Author (yr) (reference)
No. of patients
Doberneck (1983) (21) Anthony et al. (1983) (22) Roberts et al. (1984) (23) Freidig et al. (1986) (24) Ridder et al. (2002) (20) Chau et al. (2003) (30) Rolain et al. (2006) (2)
169 228 163 419 454 423 786
Yes
No
Yes Yes Yes Yes Yes
No No Yes No Yes
M. tuberculosis (6) M. tuberculosis (3) Mycobacteria (17) Histoplasma capsulatum (4) B. henselae (13) T. gondii (4) B. henselae (31)
Angelakis et al. (2009) (3) Rosado et al. (2011) (25) This study
491 368 1,688
No Yes Yes
Yes Yes Yes
B. henselae (30) M. avium complex (60) B. henselae (33)
a
M. tuberculosis (1) F. tularensis (3) M. tuberculosis (7)
ND, not described.
April 2014 Volume 52 Number 4
jcm.asm.org 1165
Downloaded from http://jcm.asm.org/ on July 8, 2014 by UNIVERSITY OF SHEFFIELD LIBRARY
tients significantly older (median ⫾ IR, 50 ⫾ 12 years; range, 35 to 66 years) than the non-T. whipplei patients (median ⫾ IR, 31 ⫾ 21 years; range, 2 months to 90 years) (P ⫽ 0.04). The patients with lymph nodes infected with S. epidermidis were significantly older (median ⫾ IR, 40 ⫾ 20 years; range, 2 months to 79 years) than the non-S. epidermidis patients (median ⫾ IR, 30 ⫾ 22 years; range, 3 months to 90 years) (P ⫽ 0.001). No significant difference was found for the other bacterial agents.
Safont et al.
22.
REFERENCES
23.
1. Fijten GH, Blijham GH. 1988. Unexplained lymphadenopathy in family practice. An evaluation of the probability of malignant causes and the effectiveness of physicians’ workup. J. Fam. Pract. 27:373–376. 2. Rolain JM, Lepidi H, Zanaret M, Triglia JM, Michel G, Thomas PA, Texereau M, Stein A, Romaru A, Eb F, Raoult D. 2006. Lymph node
1166
jcm.asm.org
3. 4. 5. 6. 7. 8. 9. 10. 11.
12.
13. 14.
15.
16.
17. 18.
19. 20. 21.
24.
biopsy specimens and diagnosis of cat-scratch disease. Emerg. Infect. Dis. 12:1338 –1344. http://dx.doi.org/10.3201/eid1209.060122. Angelakis E, Roux V, Raoult D, Rolain JM. 2009. Real-time PCR strategy and detection of bacterial agents of lymphadenitis. Eur. J. Clin. Microbiol. Infect. Dis. 28:1363–1368. http://dx.doi.org/10.1007/s10096-009-0793-6. Sanguinetti-Morelli D, Angelakis E, Richet H, Davoust B, Rolain JM, Raoult D. 2011. Seasonality of cat-scratch disease, France, 1999 –2009. Emerg. Infect. Dis. 17:705–707. http://dx.doi.org/10.3201/eid1704.100825. Angelakis E, Edouard S, La Scola B, Raoult D. 2010. Bartonella henselae in skin biopsy specimens of patients with cat-scratch disease. Emerg. Infect. Dis. 16:1963–1965. http://dx.doi.org/10.3201/eid1612.100647. Fontanilla JM, Barnes A, von Reyn CF. 2011. Current diagnosis and management of peripheral tuberculous lymphadenitis. Clin. Infect. Dis. 53:555–562. http://dx.doi.org/10.1093/cid/cir454. Donald PR. 2010. The chemotherapy of tuberculous lymphadenopathy in children. Tuberculosis 90:213–224. http://dx.doi.org/10.1016/j.tube.2010 .05.001. Kelly CS, Kelly RE, Jr. 1998. Lymphadenopathy in children. Pediatr. Clin. North Am. 45:875– 888. http://dx.doi.org/10.1016/S0031-3955(05)70051-1. Grossman M, Shiramizu B. 1994. Evaluation of lymphadenopathy in children. Curr. Opin. Pediatr. 6:68 –76. http://dx.doi.org/10.1097/00008480 -199402000-00012. Weisburg WG, Barns SM, Pelletier DA, Lane DJ. 1991. 16S ribosomal DNA amplification for phylogenetic study. J. Bacteriol. 173:697–703. Angelakis E, Richet H, Rolain JM, La Scola B, Raoult D. 2012. Comparison of real-time quantitative PCR and culture for the diagnosis of emerging Rickettsioses. PLoS Negl. Trop. Dis. 6:e1540. http://dx.doi.org /10.1371/journal.pntd.0001540. Rolain JM, Gouriet F, Enea M, Aboud M, Raoult D. 2003. Detection by immunofluorescence assay of Bartonella henselae in lymph nodes from patients with cat scratch disease. Clin. Diagn. Lab. Immunol. 10:686 – 691. http://dx.doi.org/10.1128/CDLI.10.4.686-691.2003. Angelakis E, Rolain JM, Raoult D, Brouqui P. 2011. Bartonella quintana in head louse nits. FEMS Immunol. Med. Microbiol. 62:244 –246. http: //dx.doi.org/10.1111/j.1574-695X.2011.00804.x. Bruijnesteijn Van Coppenraet ES, Lindeboom JA, Prins JM, Peeters MF, Claas EC, Kuijper EJ. 2004. Real-time PCR assay using fine-needle aspirates and tissue biopsy specimens for rapid diagnosis of mycobacterial lymphadenitis in children. J. Clin. Microbiol. 42:2644 –2650. http://dx.doi .org/10.1128/JCM.42.6.2644-2650.2004. Drancourt M, Bollet C, Carlioz A, Martelin R, Gayral JP, Raoult D. 2000. 16S ribosomal DNA sequence analysis of a large collection of environmental and clinical unidentifiable bacterial isolates. J. Clin. Microbiol. 38:3623–3630. Seng P, Drancourt M, Gouriet F, La Scola B, Fournier PE, Rolain JM, Raoult D. 2009. Ongoing revolution in bacteriology: routine identification of bacteria by matrix-assisted laser desorption ionization time-offlight mass spectrometry. Clin. Infect. Dis. 49:543–551. http://dx.doi.org /10.1086/600885. Drancourt M, Raoult D. 2007. Cost-effectiveness of blood agar for isolation of mycobacteria. PLoS Negl. Trop. Dis. 1:e83. http://dx.doi.org/10 .1371/journal.pntd.0000083. El Khéchine A, Couderc C, Flaudrops C, Raoult D, Drancourt M. 2011. Matrix-assisted laser desorption/ionization time-of-flight mass spectrometry identification of mycobacteria in routine clinical practice. PLoS One 6:e24720. http://dx.doi.org/10.1371/journal.pone.0024720. Angelakis E, Lepidi H, Canel A, Rispal P, Perraudeau F, Barre I, Rolain JM, Raoult D. 2008. Human case of Bartonella alsatica lymphadenitis. Emerg. Infect. Dis. 14:1951–1953. http://dx.doi.org/10.3201/eid1412.080757. Ridder GJ, Boedeker CC, Technau-Ihling K, Grunow R, Sander A. 2002. Role of cat-scratch disease in lymphadenopathy in the head and neck. Clin. Infect. Dis. 35:643– 649. http://dx.doi.org/10.1086/342058. Doberneck RC. 1983. The diagnostic yield of lymph node biopsy. Arch. Surg. 118:1203–1205. http://dx.doi.org/10.1001/archsurg.1983.01390100067017. Anthony PP, Knowles SA. 1983. Lymphadenopathy as a primary presenting sign: a clinicopathological study of 228 cases. Br. J. Surg. 70:412– 414. http://dx.doi.org/10.1002/bjs.1800700708. Roberts FJ, Linsey S. 1984. The value of microbial cultures in diagnostic lymph-node biopsy. J. Infect. Dis. 149:162–165. http://dx.doi.org/10.1093 /infdis/149.2.162. Freidig EE, McClure SP, Wilson WR, Banks PM, Washington JA. 1986. Clinical-histologic-microbiologic analysis of 419 lymph node biopsy specimens. Rev. Infect. Dis. 8:322–328. http://dx.doi.org/10.1093/clinids/8.3.322.
Journal of Clinical Microbiology
Downloaded from http://jcm.asm.org/ on July 8, 2014 by UNIVERSITY OF SHEFFIELD LIBRARY
of these agents. To increase the sensitivity for the diagnosis of mycobacterial lymphadenitis, a combination of various procedures with different sensitivities has been proposed (6). In our study, we combined the molecular assays with culture, and on the basis of these results, we will try to design a more effective new qPCR assay specifically for the detection of mycobacteria. Fastidious organisms were the cause of infectious adenitis, and we found that qPCR was the referent assay for the diagnosis of these agents. The rrs amplification and sequencing was less effective for the detection of B. henselae and F. tularensis than specific qPCR, whereas the detection rate for T. whipplei was low. Culture was not effective for the diagnosis of fastidious organisms, because their isolation is often difficult and depends on the number of microorganisms in the cells, which should be as high as possible and on the centrifugation step, which enhances the adhesion of bacteria that are freed from their intracellular location in the cells in culture (26). In previous studies, we observed bacteria by direct immunofluorescence in many lymph nodes that were negative by culture, which suggests that bacteria in lymph nodes are not viable (12). Moreover, in our previously published series of lymph nodes, we found a very low isolation rate for B. henselae from lymph nodes (2). Based on these results, we do not routinely culture lymph nodes positive for B. henselae. The isolation of fastidious organisms from samples remains critical for the description of new species, enabling the genetic descriptions, physiological analyses, improvement in diagnostic tools, and antibiotic susceptibility testing of bacteria (11, 27). Histologic analysis of lymph nodes is critical, as we found four lymph nodes with CSD and a malignancy. Ridder et al. also reported 2 cases with squamous cell carcinoma and 2 cases with malignant B-cell lymphoma in patients with high antibody titers to B. henselae (20). In this extensive lymph node study, we showed that lymph node analysis by the combination of rrs PCR amplification and sequencing with specific qPCR and culture is critical for the diagnosis of bacterial lymphadenitis. Staphylococci, streptococci, and other aerobic and anaerobic bacteria were detected as agents of infectious adenitis. rrs PCR remains critical for the detection of these bacteria (10). Although the efficacy of rrs PCR for the detection of nonfastidious agents varied and was dependent on the bacteria species, no significant difference was found between culture and the rrs PCR. To increase the sensitivity of culture, lymph node biopsy specimens should be sampled before treatment early in the course of the disease and should be inoculated as soon as possible. However, we cannot provide data about the patient treatment at the time of the lymph node biopsy. Physicians should be aware that in cases of negative rrs PCR, an infection by fastidious organisms and mycobacteria is not excluded given the low sensitivity of eubacterial PCR. CSD and M. avium infection occur mostly in young patients, whereas patients with M. tuberculosis are significantly older. Moreover, we reinforced the need that lymph node excision and histologic analysis are critical for accurate diagnosis, as neoplasms could be clinically misdiagnosed as CSD when only molecular assays are used.
Diagnosis of Bacterial Lymphadenitis
April 2014 Volume 52 Number 4
28. Edouard S, Fenollar F, Raoult D. 2012. The rise of Tropheryma whipplei: a 12-year retrospective study of PCR diagnoses in our reference center. J. Clin. Microbiol. 50:3917–3920. http://dx.doi.org/10.1128/JCM.01517-12. 29. Eldin C, Angelakis E, Renvoise A, Raoult D. 2013. Coxiella burnetii DNA, but not viable bacteria, in dairy products in France. Am. J. Trop. Med. Hyg. 88:765–769. http://dx.doi.org/10.4269/ajtmh.12-0212. 30. Chau I, Kelleher MT, Cunningham D, Norman AR, Wotherspoon A, Trott P, Rhys-Evans P, Querci Della RG, Brown G, Allen M, Waters JS, Haque S, Murray T, Bishop L. 2003. Rapid access multidisciplinary lymph node diagnostic clinic: analysis of 550 patients. Br. J. Cancer 88: 354 –361. http://dx.doi.org/10.1038/sj.bjc.6600738.
jcm.asm.org 1167
Downloaded from http://jcm.asm.org/ on July 8, 2014 by UNIVERSITY OF SHEFFIELD LIBRARY
25. Rosado FG, Stratton CW, Mosse CA. 2011. Clinicopathologic correlation of epidemiologic and histopathologic features of pediatric bacterial lymphadenitis. Arch. Pathol. Lab. Med. 135:1490 –1493. http://dx.doi.org /10.5858/arpa.2010-0581-OA. 26. Gouriet F, Fenollar F, Patrice JY, Drancourt M, Raoult D. 2005. Use of shell-vial cell culture assay for isolation of bacteria from clinical specimens: 13 years of experience. J. Clin. Microbiol. 43:4993–5002. http://dx .doi.org/10.1128/JCM.43.10.4993-5002.2005. 27. La Scola B, Raoult D. 1997. Laboratory diagnosis of rickettsioses: current approaches to the diagnosis of old and new rickettsial diseases. J. Clin. Microbiol. 35:2715–2727.
