The International Journal of Mycobacteriology

: 2015  |  Volume : 4  |  Issue : 3  |  Page : 196--201

Tuberculosis lymphadenitis in a southeastern region in Tunisia: Epidemiology, clinical features, diagnosis and treatment

Salma Smaoui1, Mohamed Amine Mezghanni2, Bousaima Hammami3, Neila Zalila4, Chema Marouane1, Sana Kammoun1, Abdelmonoom Ghorbel5, Mounir Ben Jemaa3, Férièle Messadi-Akrout1,  
1 Regional Laboratory of Hygiene; Faculty of Pharmacy, Monastir, Tunisia
2 Regional Laboratory of Hygiene, Tunisia
3 Department of Infectious Diseases, CHU Hedi Chaker, Sfax, Tunisia
4 Department of Care and Basic Health, Sfax, Tunisia
5 Department Oto-Rhino-Laryngology, CHU Habib Bourguiba, Sfax, Tunisia

Correspondence Address:
Salma Smaoui
Regional Laboratory of Hygiene; Faculty of Pharmacy, Monastir


Aim: To evaluate patients’ profiles, demographics, clinical and therapeutic approaches and strategies in patients with tuberculous lymphadenitis (TBG). Patients and methods: A retrospective study of all TBG-confirmed cases admitted in a tuberculosis-specific health care facility between 1 January 2009 and 16 June 2013. Results: A total of 181 clinical files were examined. Mean age was 32 years old; the female/male ratio was 1.78 to 1. Raw milk consumption was noted in 1/3 of patients. Most cases involved the head and neck region (83.4%), nodes involvement, including axillary (12 cases), and mediastinal (9 cases). Clinical symptoms were present in only 55.2%. Tuberculin skin test (TST) was conducted with 82.6% positive responses. Diagnostics confirmation was done with anatomical pathology in most of the patients; only 56 of them had any microbiology analysis done. Demonstration of acid-fast bacilli in microscopy from either fine-needle aspirates or biopsies was done in 17.5% of cases, and cultures yielded positive results in 27%. Treatment duration was varied. Paradoxical reactions were noted in 12% and persistent lymphadenopathy after treatment completion was noted in 10% of cases. Conclusions: TBG remains a disease of interest. Today, its diagnosis and management is still a problem despite its increasing worldwide incidence, and especially in this study area. Disease control should be strengthened in this country.

How to cite this article:
Smaoui S, Mezghanni MA, Hammami B, Zalila N, Marouane C, Kammoun S, Ghorbel A, Jemaa MB, Messadi-Akrout F. Tuberculosis lymphadenitis in a southeastern region in Tunisia: Epidemiology, clinical features, diagnosis and treatment.Int J Mycobacteriol 2015;4:196-201

How to cite this URL:
Smaoui S, Mezghanni MA, Hammami B, Zalila N, Marouane C, Kammoun S, Ghorbel A, Jemaa MB, Messadi-Akrout F. Tuberculosis lymphadenitis in a southeastern region in Tunisia: Epidemiology, clinical features, diagnosis and treatment. Int J Mycobacteriol [serial online] 2015 [cited 2020 Aug 14 ];4:196-201
Available from:

Full Text


Tuberculosis (TB) is one of the most frequently infectious diseases in the world. The World Health Organization (WHO) considers it as a major public health problem and counted, in 2013, more than 9 million new cases and 1.3 million deaths worldwide [1]. TB epidemiology has radically changed since the advent of the HIV pandemic [2] , which would result in the increased incidence of extrapulmonary tuberculosis (EPTB) at the expense of the conventional pulmonary form [1],[3]. In Tunisia, the EPTB/TB ratio increased from 19.6% in 1996 to 32.6% in 2007 while demonstrating a significant decline in the overall incidence during the same period, and this increase is strongly associated with lymph node TB [4].

Tuberculous lymphadenitis (TBLA) is the most common extrapulmonary manifestation of TB. Its proportion varies between 1.5% and 35% by country [5],[6],[7]. In Tunisia, the rate is estimated at 23%, and thus is the second most common form after pulmonary tuberculosis (PTB) [4].

Despite a renewed interest and a resurgence of cases in recent years, TBLA remains a disease whose many factors are largely unknown [7]. The diagnostic management remains difficult [6],[7],[8],[9],[10],[11]. The treatment of TBLA is based on an essentially medical approach and rarely surgical excision biopsy of affected sites with a large range of technical, protocols and duration of treatment [11],[12],[13]. In Tunisia, there is little information on the diagnosis and therapy of TBLA. To better understand this disease, this study was performed to evaluate patients’ profiles, demographics, clinical and therapeutic approaches, and strategies in patients with TBLA in a population with low prevalence for HIV infection.

 Patients and methods

This is a descriptive retrospective single-center study of all cases of TBLA admitted to the TB clinic (TBC) in Sfax between 1 January 2009 and 16 June 2013. The TBC is a center specializing in the treatment of TB which receives all tuberculous patients diagnosed in the hospital and private structures in the region of Sfax for additional support and monitoring of TB. The diagnosis of TBLA stood on microbiological, anatomopathological, radio-clinical and/or therapeutic criteria. Besides an unconfirmed diagnosis or incomplete files, no other exclusions criteria were retained. The data were collected through a form provided for this purpose. Additional data were collected from the records of patients who were admitted in the university hospitals of Sfax. Epi Info™ Version 7.1.2 was used for the acquisition and analysis of data.



A total of 203 cases of TBLA were identified during the study period, of which 18 were excluded for missing data and 4 unconfirmed diagnosis of TBLA. So, the total number of considered cases is 181. The mean (±SD) age was 32 ± 18.6 years, with a range from 3 months to 81 years. The age group “20–30 years” was the most affected (33%) [1]. The female/male ratio was 1.78:1. The majority of patients resided in the urban area of Sfax city (60.8%). The epidemiological characteristics of this group are summarized in [Table 1]. {Table 1}

Only 10 patients (5.5%) had a past medical history of TB, including 5 with at least one previous episode of TBLA. No patient was HIV positive. Diabetes mellitus was documented only in one patient, and raw milk consumption was found in 33.3% of cases. Two patients had congenital immunodeficiency: one case with IL12 receptor deficiency and one case with chronic granulomatous disease; both patients were aged less than 5 years at the moment of diagnosis, and it was presumed that BCG vaccination was the main etiology behind both cases.

Clinical symptoms

One hundred patients (55.2%) had enlarged lymphadenopathy (ADP). The majority of ADP cases were cervical (83.4%), and unilateral (78.4%), sizing between 1 and 3 cm (54.6%) ([Table 2]). Patients with multiple ADP all expressed at least one sign of TB impregnation. ADPs were discharging in 18.2% of cases and firm in 23.7% of cases. About 1/3 of this population (32.1%) presented significant weight loss. The prolonged daytime fever (>38 °C) was reported in 21% of cases. Asthenia and night sweats were reported in 12.1% and 11.6% of cases, respectively. Acute intra or perinodular ADP was noted in 13 cases (7.2%), chronic cough was found in the history of 5 patients, of whom only 1 had hemoptysis (due to PTB). Ten cases of active PTB were found; the diagnosis was established in parallel while investigating for LNT. Concomitant EPTB was found in 13 patients (miliary TB = 5; pleural TB = 2, abdominal TB = 1, skin TB = 1, other = 4). Tuberculosis skin Test (TST) was positive in 82.6% of these patients (>10 mm). {Table 2}


Of a total of 111 X-ray chest scans, 24.3% showed bronchial syndrome. An ultrasound scan, predominantly cervical (86.9%) and abdominal (4.2%), was performed at least once on 46 of these patients (25.4%). A positive imagery concurring with TBLA was noted in 35 patients (76%). The majority of positive results were based on a hypoechoic image with a hyperechoic center for lymph node abscess or necrosis, heterogeneous hypoechoic image to newly contaminated nodes, and calcification of one or more nodes in 7 cases (20%; n = 35). CT scan was positive in 13 out of 20 conducted cases.

Pathological and microbiological diagnosis

One hundred seventy-four patients (96.1%) had a pathological-based exam. The excision biopsy was the sampling means in 72% of cases, fine needle aspiration (FNA) technique was used in 10% of cases, and 18% had both methods. The sensitivity of pathological examination was 94.2% and 62.5%, respectively, for biopsies and FNA. Histological analysis revealed the presence of epithelioid granulomas with or without caseous necrosis in 95.2% of cases and giant Langerhans cells in 64.9% of cases. Of the 4 patients excluded from the study to be proven non-TBLA, 3 underwent histological examination and only 1 had a suggestive result of TBLA. The specificity of these tests would go up to 99.4%.

The mycobacterial research involved only 56 patients: 36 biopsies, 21 FNA samples and 2 fresh discharges directly from the fistula. Ziehl Neelsen stains of biopsy were positive in 5/36 (13.9%), and 5/21 for the FNA (23.8%). Culture concerned only 35 samples (of which 11 were FNA). Culture media included Lowenstein Jensen (LJ) for every sample and MGIT for only 14 samples. The positivity rate on LJ medium was 22.9%, against 14.9% in MGIT ([Table 3]). Five bacterial strains have been identified: 3 Mycobacterium bovis , 1 M. bovis BCG and 1 Mycobacterium tuberculosis. These strains showed no resistance to the first-line antitubercular drugs (isoniazid, rifampicin, ethambutol and streptomycin). No polymerase chain reaction (PCR) or otherwise genomic identification was performed in this series.{Table 3}

Treatment and evolution

The majority of these patients (79%) were treated according to the latest recommendation of the WHO: classic regimen based on a quadruple therapy (isoniazid + rifampicin + ethambutol + pyrazinamide) for 2 months followed by dual therapy (isoniazid + rifampicin) for a period of 4–30 months; 60% received treatment in combined single pharmaceutical form. The mean duration of treatment was 8.7 months ± 4 months. Treatment duration stretched from 2 to 32 months. Eleven cases of drug-induced liver toxicity (cytolysis) and 4 cases of acute rash have been reported. Eighteen patients (12%) had developed at least a clear sign of a paradoxal response (PR), 38 (25.3%) had a partial improvement in their symptoms and 94 (62.7%) responded very well to treatment. All PR patients underwent additional treatment with corticosteroids. An improvement was noted in 15 patients at the end of treatment and 3 have kept enlarged or abnormal ADP.

The majority of patients (53%) improved at the end of TB treatment (regression in the size of ADP and improvement of general health condition with no signs suggesting active TB infection). Approximately 10% of patients have kept their enlarged ADP, and 3 were considered treatment failure (persistence of positive direct examination after treatment completion). Five patients died after or during their treatments: 1 patient following TB complications, 1 case of severe hepatitis, and 3 others from associated diseases.


Demography, epidemiology and risk factors

TBLA progresses rapidly compared with other forms in Tunisia in recent years; the estimated incidence increased from 2.3 cases per 100,000 inhabitants in 1995 to 5.8 cases in 2010 [4]. This phenomenon is also observed globally: Menon et al. reported a net increase from 0 to 19 cases of TBLA, while having a global decrease in the TB cases count [14]. The subjects are rather young and come from high TB prevalence countries (Southeast Asia) [7],[13],[14]. Other reports confirm high rates among immigrants from these regions in developed countries [13],[14]. This study has identified a relatively young age (20–30 years) in line with what is reported in the literature in developing countries [6],[7],[10],[14]. Female to male ratio varies from 1.1:1 [14] to 1.8:1 [7],[8] , or even 2.1:1 [15]. Fontanilla et al. relate a world average of 1.4:1 [16]. The ratio in the present study of 1.78:1 is similar to that found in these countries. Purohit et al. and Oloya et al. explained this dominance by unfavorable socio-economic conditions and certain cultural and social practices and habits in these countries [17].

Consumption of milk and raw milk products appears to play a key role in the transmission of mycobacteria. M. bovis has long been known for its ability to spread to the digestive tract or nasopharyngeal via raw milk consumption, which is believed to be the main cause behind TBLA [18]. Preventive measures to limit bovine TBLA in developed countries during the past century have effectively decreased human infection with M. bovis , 92–98% of the isolates belong to the species M. tuberculosis [16],[18],[19]. In Europe, M. bovis is no longer responsible for EPTB, against 36.4% in Ethiopia, a country with a strong cultural tradition of raw milk consumption [19]. Cheese consumption found in Hispanic communities in the United States has had the effect of raising the ratio of M. bovis/M. tuberculosis from 2% to more than 29.4% [18]. In this series, milk consumption affects 1/3 of the study population, and although microbiological identification concerned only 5 strains, 3 of them resulted in M. bovis , which suggests a possible wide endemicity. According to the latest report of basic healthcare management administration, M. bovis is involved in 30–50% of all documented cases of TBLA.

HIV is the most discussed risk factor in the past two decades in TB literature [1],[3],[12]. Over 50% of observed forms of TB in seropositive patients are typical EPTB [12]. This can be explained by the specific deficit in CD4 + T lymphocytes and therefore the absence of specific macrophage activation against TB antigens [12]. The INF-γ response is decreased during HIV infection. The immune response and tissue hypersensitivity are absent or greatly decreased [20] , therefore, the clinical manifestation of TBLA is radically different [21]. No cases of HIV/TBLA co-infection has been retained in this series, which is in agreement with other studies conducted in countries with low HIV prevalence [22].

Congenital immunodeficiency concerned 2 patients in this series, the first having a quantitative deficiency of receptors for IL-12, essential to the generation of an anti-microbial-based response specifically involving activation of PNN and macrophages [20]. This patient developed an opportunistic disseminated infection following vaccination with BCG at birth. The lymph node involvement is described in the literature for this type of hereditary disease [23]. Chronic granulomatous disease was found in the second patient, and is also an etiology of disseminated TB and should be considered in any case of unexplained massive TBLA in children following vaccination with BCG [23].

Two studies reported a statistically significant and inversely proportional relationship between the occurrence of TBLA and diabetes mellitus [13],[24]. However, no explanation to this date has been formulated.

Clinical and paraclinical

TBLA is characterized by an impaired clinical presentation [13]. Its diagnosis is made difficult by the subacute and fickle nature of general signs [7],[8],[12],[16]. The classic TB signs are less frequent in TBLA than in PTB [7]. For example, the rates vary considerably between 19% and 40% for fever [16]. In this study, 44.7% of these patients showed at least one sign; weight loss and prolonged fever were observed in 32% and 21% of cases, respectively. Wahid et al. report weight loss and prolonged fever rates in 24.7% and 34.6% of cases, respectively [25]. Khan reports in his study in Qatar up to 85% of patients with night sweats, 77% of patients with anorexia, 62% of patients with severe weight loss and 40% with cough [26]. Biadglegne et al. provide that the different frequencies between the studies/regions have a relationship with the bacterial load during contamination. These signs prove to be almost always present in immunodeficient patients [19].

All studies, including the present study, agree on the fact that the majority of ADP is one-sided and localized in the head and neck region. Sharma et al. reported in a meta-analysis that 2/3 of TBLA are strictly cervical [8]. Fontanilla et al. report an average of between 45% and 70% based on 4 trials with 85% unilateral ADP between 1 and 3 cm [16]. The deep sites (mesenteric and mediastinal) are the least observed in practice and compromise immune impaired patients. However, the limited number of confirmed cases, coupled with a low sample survey and a real difficulty in spotting deep ADP in practical routine checks does not reflect the true extent of the involvement of other nodal sites [12]. In studies that included imaging and CT scans as a means of diagnosis, the rate of mediastinal ADP was more importantly up to 26.7% [15]. Spontaneous fistula to the skin was noted in 18% of cases. This result is significantly greater than the average of 4–11% reported in the literature [16].

Pathological and microbiological diagnosis

Histology is today the most used medium in the diagnosis of TBLA, not only in this series, but also in everyday practice [7,9,10,13,27]. However, the WHO considers the identification of the bacteria by direct examination or culture methods as the only guarantee for positive TB infection [1]. FNA has the advantage of being much more convenient to clinicians, cheaper and less burdensome for the patients, and is by far the ideal choice in developing countries with a high prevalence of TB [7],[15],[16]. Compared with other cases, biopsy was much used in this study. [15] The pathological pattern for TB infection in biopsy samples is the rearrangement of the normal structure of the lymph node tissue into tuberculous granulomas, with or without caseous necrosis and Langerhans-type cells [7]. The main characteristic sought in the FNAC is the presence of epithelioid cells in clusters with or without necrotic material [7]. The sensitivity of the pathological examination differs between samples, operating protocols and nature of required materials [13]. It has been noted in this series there is an improved sensitivity for biopsy against the FNA (94.2% against 62.5%), which is consistent with several other studies. The literature reports widely varying rates for FNA, ranging from 9% [27] up to 90% [13] , with most studies between 55% and 85% [13],[16]. The sensitivity of biopsy is significantly higher with rates ranging from 84% to 100% [9],[13],[16],[27].

Direct examination by Ziehl Neelsen staining should be routinely used for each sample and a culture made if possible, or in case of strong suspicion of TB. In this study, the majority of the patients (69%) have not been subject to any microbiological screening. Marrakchi et al. noted 37 microbiological analyses among 50 cases of TBLA and 9 biopsy cultures of 32 samples [9]. The sensitivity varies considerably from one study to another. This result is within the limits of the overall average. The sensitivity of the culture, either from biopsy sampling or FNA, is greater than that reported by Marrakchi et al. (7/26 against 1/9), whether the direct examination was less effective than its counterpart reported in the same study [9].

Overall, the sensitivity of microbiological methods is considered mediocre and disparate. Bayazit et al. estimated the sensitivity of culture between 10% and 69% [28]. The sensitivity of FNA is between 15% and 40% in the meta-analysis of Mazza-Stalder et al. [29] , and 17–82% with Polesky et al. [13].

Few studies have addressed the diagnostic approach using MGIT. The present study found 2 positive cultures from 14 cases (14.9%). Polesky et al. have found 25 positive cultures out of 111 cases (23%) [13]. In a comparative study using MGIT/LJ, Rodrigues et al. have found a sensitivity of 26% (n = 1451) against only 10% for culture on LJ [30]. MGIT medium allowed for a gain in terms of sensitivity and detection limit.

Treatment and evolution

The American Thoracic Society (ATS), Centers for Disease Control and Prevention (CDC) and Infectious Diseases Society of America (IDSA) all recommend treating TBLA for a nominal duration of 6 months: 2 months of daily quadruple anti-tuberculous drugs and 4 months of daily bi-therapy or 3 times a week, with at least 2 major anti-TB drugs prescribed during the course of treatment [16],[31]. There is no clear consensus on a global scale for the treatment of TBLA, including the duration of treatment which varies widely in this population (2–32 months), but is in general consistent with the current practices in the field. The majority of these patients were treated for an intended period of 9 months, which is consistent with the French trend of Lanoix et al. [11] and Clevenberg et al. [22]. Polesky et al. report an average of 7.2 months and a median at 6 months [13]. Using a treatment period of 6 months has not had a significant impact on the relapse rate, which authors put roughly to 3.3% for 6 months and 2.7% for 9 months [28]. Abba et al. prescribed a 9-month regimen in case of non-improved, new or worsening of ADP [32]. Geldmacher et al. prescribed up to 12 months for more than 40% of their study population [15].

Paradoxical reaction is one of the most spectacular phenomena found in the treatment of TB, defined by the deterioration of the general condition during the course of treatment. It is rare in the case of a pure PTB (<3%) [33] , and is 20–30% for the TBLA [34]. The present study reported a lower number (25.3%). Risk factors for the occurrence of PR according to Cho et al. are: a young age, male gender and a mobile inflammatory lymph node appearance (n = 235, p < 0.01) [35]. Gert et al. point out the peripheral ADP and anemia (n = 76; p < 0.05) [33]. The physiology of PR is not yet clear and a precise definition has not been the subject of consensus. TBLA is not a life-threatening illness or requiring urgent care [8],[16]. The relapse rate is low with a well-led regimen; this study shows a single case of relapse during the study period and 5 out of 181 patients have a history of TBLA. Polesky et al. noted a higher rate at 9%, including one case of relapse after 8 years of primary infection [13]. The persistence of ADP after the end of treatment without other clinical symptoms was observed in 10% of cases. It is not an actual failure and is far from trivial; Abba et al. noted that 34% of patients had residual ADP without other clinical signs [36].


TBLA remains a topical disease. Today, its diagnosis and management are problematic and may in some contexts be overlooked despite a growing incidence worldwide in recent years. Disease control should be strengthened in this country.

 Conflict of interest

The authors had no conflicts of interest to declare in relation to this article. All authors agreed to the content of the article and contributed significantly.


1World Health Organization, Global tuberculosis Control. <<, Report 2014.
2J. Burzynski, N.W. Schluger, The epidemiology of tuberculosis in the United States, Semin. Respir. Crit. Care Med. 29 (2008) 492–498.
3H.M. Peto, R.H. Pratt, T.A. Harrington, P.A. LoBue, L.R. Armstrong, Epidemiology of extrapulmonary tuberculosis in the United States, 1993–2006, Clin. Infect. Dis. 49 (2009) 1350– 1357.
4M. Ben Jemaa, La Tuberculose en Tunisie-Problématique. <>.
5K. Mrówka-Kata, G. Namysłowski, E. Czecior, P. Sowa, J. Iwań ska, An updated view on tuberculous lymphadenitis in the context of HIV epidemic as well as multidrug and extensively drug-resistant tuberculosis, Otolaryngologia. Polska. 66 (2012) 176–180.
6D. Beyene, S. Ashenafi, L. Yamuah, A. Aseffa, H. Wiker, H. Engers, et al, Diagnosis of tuberculous lymphadenitis in Ethiopia: correlation with culture, histology and HIV status, Int. J. Tuberc. Lung Dis. 12 (2008) 1030–1036.
7U. Handa, I. Mundi, S. Mohan, Nodal tuberculosis revisited: a review, J. Infect. Dev. Ctries. 6 (2012) 6–12.
8S. Sharma, A. Mohan, Extrapulmonary tuberculosis, Ind. J. Med. Res. 120 (2004) 316–353.
9C. Marrakchi, I. Maaloul, D. Lahiani, B. Hammami, T. Boudawara, M. Zribi, et al, Diagnostic de la tuberculose ganglionnaire périphérique en Tunisie, Med. Mal. Infect. 40 (2010) 119–122.
10P.R. Mohapatra, A.K. Janmeja, Tuberculous lymphadenitis, J. Assoc. Physicians India 57 (2009) 585–590.
11J.P. Lanoix, Y. Douadi, A. Borel, C. Andrejak, Y. El Samad, J.P. Ducroix, et al, Lymph node tuberculosis treatment: from recommendations to practice, Med. Mal. Infect. 41 (2011) 87– 91.
12H. Reuter, R. Wood, H.S. Schaaf, P.R. Donald, Overview of extrapulmonary tuberculosis in adults and children, in: H.S.S. Mbch, M.M. Paed, M.D. Dcm, I.Z. Paed, Alimuddin, et al. (Eds.), Tuberculosis, W.B. Saunders, Edinburgh, 2009, pp. 377–390. Chapter 34.
13A. Polesky, W. Grove, G. Bhatia, Peripheral tuberculous lymphadenitis: epidemiology, diagnosis, treatment, and outcome, Medicine (Baltimore) 84 (2005) 350–362.
14K. Menon, C. Bem, D. Gouldesbrough, D. Strachan, A clinical review of 128 cases of head and neck tuberculosis presenting over a 10-year period in Bradford, UK, J. Laryngol. Otol. 121 (2007) 362.
15H. Geldmacher, C. Taube, C. Kroeger, H. Magnussen, D.K. Kirsten, Assessment of lymph node tuberculosis in Northern Germany, Clin. Rev. Chest 121 (2002) 1177–1182.
16J.M. Fontanilla, A. Barnes, C.F. Von Reyn, Current diagnosis and management of peripheral tuberculous lymphadenitis, Clin. Infect. Dis. 53 (2011) 555–562.
17M.R. Purohit, T. Mustafa, O. Morkve, L. Sviland, Gender differences in the clinical diagnosis of tuberculous lymphadenitis – A hospital-based study from Central India, Int. J. Infect. Dis. 13 (2009) 600–605.
18B. Muller, S. Durr, S. Alonso, J. Hattendorf, C.J. Laisse, S.D. Parsons, et al, Zoonotic Mycobacterium bovis-induced tuberculosis in humans, Emerg. Infect. Dis. 19 (2013) 899–908.
19F. Biadglegne, W. Tesfaye, B. Anagaw, B. Tessema, T. Debebe, A. Mulu, et al, Tuberculosis lymphadenitis in ethiopia, Jpn. J. Infect. Dis. 66 (2013) 263–268.
20N. Krishnan, B.D. Robertson, G. Thwaites, The mechanisms and consequences of the extra-pulmonary dissemination of Mycobacterium tuberculosis, Tuberculosis (Edinb) 90 (2010) 361– 366.
21Y.F. Wei, Y.S. Liaw, S.C. Ku, Y.L. Chang, P.C. Yang, Clinical features and predictors of a complicated treatment course in peripheral tuberculous lymphadenitis, J. Formos. Med. Assoc. 107 (2008) 225–231.
22P. Clevenbergh, I. Maitrepierre, G. Simoneau, L. Raskine, J.D. Magnier, M.J. Sanson-Le-Pors, et al, Lymph node tuberculosis in patients from regions with varying burdens of tuberculosis and human immunodeficiency virus (HIV) infection, Presse Med. 39 (2010) e223–e230.
23M. Hodsagi, G. Uhereczky, L. Kiraly, E. Pinter, BCG dissemination in chronic granulomatous disease (CGD), Dev. Biol. Stand. 58 (1986) 339–346.
24O.Y. Gonzalez, L.D. Teeter, B.T. Thanh, J.M. Musser, E.A. Graviss, Extrathoracic tuberculosis lymphadenitis in adult HIV seronegative patients: a population-based analysis in Houston, Texas, USA, Int. J. Tuberc. Lung Dis. 7 (2003) 987–993.
25W. Fazal, R. Habib, I. Ahmad, Extrapulmonary tuberculosis in patients with cervical lymphadenopathy, J. Pak. Med. Assoc. 63 (2013) 1094–1097.
26F.Y. Khan, Clinical pattern of tuberculous adenitis in Qatar: experience with 35 patients, Scand. J. Infect. Dis. 41 (2009) 128–134.
27T. Tachibana, Y. Orita, M. Fujisawa, M. Nakada, Y. Ogawara, Y. Matsuyama, et al, Factors that make it difficult to diagnose cervical tuberculous lymphadenitis, J. Infect. Chemother. 19 (2013) 1015–1020.
28Y.A. Bayazit, N. Bayazit, M. Namiduru, Mycobacterial cervical lymphadenitis, ORL J. Otorhinolaryngol. Relat. Spec. 66 (2004) 275–280.
29J. Mazza-Stalder, L. Nicod, J.P. Janssens, La tuberculose extrapulmonaire, Rev. Mal. Respir. 29 (2012) 566–578.
30C. Rodrigues, S. Shenai, M. Sadani, N. Sukhadia, M. Jani, K. Ajbani, et al, Evaluation of the bactec MGIT 960 TB system for recovery and identification of Mycobacterium tuberculosis complex in a high throughput tertiary care centre, Indian J. Med. Microbiol. 27 (2009) 217–221.
31D.E. Griffith, T. Aksamit, B.A. Brown-Elliott, A. Catanzaro, C. Daley, F. Gordin, et al, An official ATS/IDSA statement: diagnosis, treatment, and prevention of nontuberculous mycobacterial diseases, Am. J. Respir. Crit. Care Med. 175 (2007) 367–416.
32A.A. Abba, A.E. Bamgboye, M. Afzal, R.A. Rahmatullah, Lymphadenopathy in adults, Saudi Med. J. 23 (2002) 282–286.
33G. Geri, A. Passeron, B. Heym, J.B. Arlet, J. Pouchot, L. Capron, et al, Paradoxical reactions during treatment of tuberculosis with extrapulmonary manifestations in HIV-negative patients, Infection 41 (2013) 537–543.
34A. Singh, H. Rahman, V. Kumar, F. Anila, An unusual case of paradoxical enlargement of lymph nodes during treatment of tuberculous lymphadenitis in immunocompetent patient and literature review, Am. J. Case Rep. 14 (2013) 201–204.
35O.H. Cho, K.H. Park, T. Kim, E.H. Song, E.Y. Jang, E.J. Lee, et al, Paradoxical responses in non-HIV-infected patients with peripheral lymph node tuberculosis, J. Infect. 59 (2009) 56–61.
36A.A. Abdullah, Clinical experience with tuberculous lymphadenitis in Central Saudi Arabia, JK-Practitioner 3 (2006) 83–86.