Jpn. J. Infect. Dis., 64, 316-321, 2011

Original Article

Assessment of Lymph Node Tuberculosis in Two Provinces in Turkey Dursun Tatar, Gunes Senol1*, Serpil Alptekin2, and Ebru Gunes3

Department of Respiratory Medicine and 1Department of Infectious Diseases and Clinical Microbiology, Dr. Suat Seren Chest Diseases and Surgery Training and Research Hospital, Izmir; 2Esrefpasa Tuberculosis Dispensary, Izmir; and 3Van Tuberculosis Dispensary, Van, Turkey (Received November 12, 2010. Accepted May 31, 2011) SUMMARY: The aim of this study was to evaluate lymph node tuberculosis (LNT) cases in two provinces in Turkey with different demographic and socioeconomic characteristics. A total of 109 LNT cases were reviewed retrospectively. The cases were analyzed and compared for symptoms, findings, age, vaccination status, and diagnostic procedures. Socioeconomic conditions were also assessed for the two provinces. A palpable cervical node was considered a significant predictor for all LNT. Mediastinal lymph node involvement was found to be common in cases of pulmonary manifestation of LNT. Female patients were predominantly from the Van Province, while older patients were found to be from Izmir Province. LNT should be suspected in lymphadenitis patients of all age-groups especially in young adolescents with cervical lymph node enlargements. In the presence of mediastinal lymphadenopathy, pulmonary tuberculosis should be investigated. and soon afterward, more reliable data on TB epidemiology became available. Turkey is a country that has a moderate level of TB with an incidence of 26/100,000, and the incidence of LNT is lower than that of pleural TB according to a 2007 report of the agency. The incidence of pulmonary TB in Van and Izmir Provinces was reported to be 29 and 16.4 per 100,000, while the mortality rates for TB were found to be 0.6z and 4.6z, respectively. TB is responsible for 0.08z and 0.27z of all deaths in Van and Izmir, respectively. The incidence of extrapulmonary TB was found to be 30.7/100,000 in Izmir and 29.6/100,000 in Van, but LNT incidences are low for both provinces. Of the extrapulmonary TB cases in Turkey, extrathoracic LNT is reported in 26.7z and intrathoracic LNT in 5.5z of the cases. However, separate data for each province have not been published (7,8). In the present study, records of the clinical features and the socioeconomic and demographic characteristics of patients with LNT were retrospectively investigated in regional dispensaries. Data were evaluated for comparing two different provinces for LNT.

INTRODUCTION Tuberculosis (TB) is a major health concern in developing countries. Its association with low socioeconomic status is well established. Of the 22 highest burden countries that account for 80z of the global TB cases, 17 are classified as low-income countries (1). The World Health Organization (WHO) estimates that 98z of the 3 million annual TB deaths and 95z of the 8.4 million new TB cases occur in developing countries (2). As a consequence of increasing immigration and human immunodeficiency virus (HIV) prevalence, TB is also reemerging as a health care problem in developed countries. However, recent data from the United States suggest that, in the incidence of TB, socioeconomic factors act independently of the HIV epidemic (3). In practically all TB infections, regional lymph nodes are involved. TB bacilli enter the lymphatic system and bloodstream to reach the extrapulmonary organs. Lympho-hematogenous spread is limited by the immune system and more than 90z of infected people recover, as indicated by tuberculin skin test (TST) (4,5). Lymph node tuberculosis (LNT) occurs either by reactivation of dormant TB bacilli in lymph nodes or, very rarely, by direct exposure to infection (6). LNT is a common cause of lymphadenopathy in areas in which TB is endemic. In countries with a low prevalence of TB, LNT is the most common extrapulmonary form. In contrast, in high prevalence areas, the LNT incidence is second to that of TB pleuritis. In Turkey, the National Reference Laboratory and Epidemiology Unit was established through cooperation with the Refik Saydam National Public Health Agency and Japan International Cooperation Agency in 1999,

MATERIALS AND METHODS Setting: In the Turkish healthcare systems, dispensaries are the official locations where TB drugs are distributed to patients free of charge and where patients are followed up for therapy and treatment outcomes. Two regional dispensaries that were separated by the farthest distance possible in the country were included in the study. The first was one of the eight central dispensaries in the western city of Izmir, the third largest city in Turkey. The second dispensary was in Van in Eastern Turkey. The populations of these two provinces have different demographic and socioeconomic characteristics (Table 1 and Fig. 1). Study subjects: Records of LNT patients who had visited the dispensaries between 2000 and 2005 were

*Corresponding author: Mailing address: 1703 S. No: 42 D. 3 Karsiyaka, Izmir, Turkey. Tel: ＋90 2324333333, E-mail: drshenol＠yahoo.com 316

Fig. 1. Various variables of socioeconomic status.

lymph nodes). TB was diagnosed in the mediastinal nodes by testing specimens from a bronchoscopic fineneedle biopsy. Mediastinoscopy and video-assisted thoracoscopic surgery (VATS) were performed for mediastinal nodes. One aliquot of the FNA sample was fixed in 95z ethyl alcohol and stained by Giemsa for cytological analysis. A second aliquot was used for microscopic detection using Ziehl-Neelsen staining and culturing on Lowenstein-Jensen medium and/or radioactive culture (BACTEC; Becton Dickinson, Sparks, Md., USA) in cases in which bacteriological examinations were available. The same procedure was followed for samples from lymph node excisions. Histological analyses were performed on samples fixed in formalin. Definition of socioeconomic status: Socioeconomic status was defined at household and regional levels, and was determined on the basis of the following parameters: (i) human resources, (ii) education, and (iii) access to community infrastructures and facilities (Fig. 1). Statistics: SPSS 15 software was used for statistical analysis. Distribution of data was evaluated using the Kolmogorov-Smirnov test. Analysis of parametric variables was performed using the t test. Kruskal-Wallis and Mann-Whitney tests were applied for nonparametric variables. Chi-square tests were used for analyzing categorical data. Analysis of patient records was performed according to the Helsinki declaration of 2008.

Table 1. Demographic features of two provinces Age group (y)

Izmir

Van

Total

º15

Total Male Female

797,900 410,100 387,800

415,900 217,800 198,100

20,220,000 10,453,000 9,767,000

15–64

Total Male Female

2,243,600 1,089,400 1,154,200

441,300 220,300 221,000

43,724,625 22,143,735 21,580,890

À65

Total Male Female

229,500 99,500 130,000

20,240 9,940 10,300

3,859,300 1,750,000 2,109,300

reexamined retrospectively. A total of 109 LNT cases from Van and Izmir Provinces were included in the study. Patients were divided into two groups: LNT and LNT associated with pulmonary manifestations (LNTPM). Methods: Medical history records and physical examination details were reviewed for all patients. Definitive diagnosis was established by histopathological, bacteriological, and radiological methods. TST was performed in 44 patients. PPD test tuberculin (BBMCIPD, Sofia, Bulgaria; 5 TU/0.1 ml) was applied intradermally (0.1 ml, or 5 TU). Skin responses were evaluated 72 h after application, with a transverse diameter of induration Æ10 mm being judged as a positive reaction. In order to detect pulmonary manifestations, all subjects underwent chest radiography, and were asked to provide sputum samples for analysis. Only 9 patients could give the sputum sample. Patients showing cavernous or disseminated patterns on the radiographs and patients with smear test (i.e., Ziehl-Neelsen staining) results that were positive for the presence of acid-fast bacilli in sputum samples were excluded. Histological and/or microbiological analyses for the presence of mycobacteria in lymph node tissue were conducted in all patients (except 2 patients who were diagnosed on the basis of just clinical and radiological findings). Fine-needle aspiration (FNA) and lymph node excision and incision were used to investigate superficial lymph nodes (i.e., cervical, axillary, or inguinal

RESULTS Van and Izmir account for 1.3z and 5.3z of the population of Turkey, respectively. In Van Province, 52.5z of the population lives in urban areas, whereas 85z of the population resides in cities and towns in Izmir Province. In Turkey, 70.3z of the total population lives in urban settings (10). Between 2000 and 2005, a total of 109 cases were diagnosed as LNT in both provinces. In Izmir Province, 56 (24.8z) patients were enrolled from 225 patients with extrapulmonary disease. According to the study inclusion criteria, these patients were classified into two groups: 48 (21.3z) patients were classified into the LNT group, and 8 (3.5z) patients were classified into the LNT-PM group. In Van, 53 (30.8z) LNT patients 317

Table 2. Demographic data of the patients LNT patient Characteristic

Gender Female Male Age (y) Range Mean±SD TB contact history BCG status No. of TST tested TST1) (＋) Predisposing factors HIV Diabetes mellitus Alcoholism Malignancy Chronic renal failure

LNT-PM patient

Izmir (n ＝ 48) no. (z)

Van (n ＝ 46) no. (z)

P

13 (27) 35 (73.9)

28 (60) 18 (39.1)

0.001 0.001

19–73 41.7±17.6 13 (27) 46 (95.6) 28 (58.3) 24 (85.7)

4–70 31.9±15.1 13 (28.2) 22 (50) 10 (21.7) 4 (40)

0.007 0.9 0.0001 0.0004 0.0001

0 2 0 2 3

0 0 0 0 1

NA

Izmir (n ＝ 8) no. (z)

Van (n ＝ 7) no. (z)

4 4

4 3

27–64 48.8±15.9 2 (25) 8 (100) 4 (50) 4 (100) 0 0 0 1 0

1–72 39±28.3 3 (42.8) 2 (22.2) 2 (28.5) 1 (50) 0 0 0 0 1

P

0.78

0.4 0.06 0.007 0.06 0.33 NA

1): Number of TST tested cases were used as denominator. LNT, lymph node tuberculosis; LNT-PM, LNT associated pulmonary manifestations; NA, low number of data.

were enrolled from 172 patients with extrapulmonary disease; among these, 46 (26.7z) cases were classified into the LNT group and 7 were classified into the LNTPM group. Demographical aspects, BCG vaccination status, TST status, and predisposing factors of these patients are shown in Table 2. There were significantly higher cases of LNT in women in Van while the reverse was true in Izmir ( P ＝ 0.0006). There was no difference in terms of gender in the distribution of LNT-PM cases ( P À 0.05). LNT patients in Izmir were significantly older than those in Van ( P ＝ 0.005), while the ages of the LNT-PM patients in Izmir and Van were not significantly different. Similarly, there was no difference in the ages of LNT and LNT-PM patients from the same region ( P À 0.05). No difference in TB contact history between case groups was detected ( P À 0.05). PPD-testing rates in LNT patients were higher in Izmir than in Van (P ＝ 0.0003), as was the positivity of TST testing ( P ＝ 0.01). The rate of BCG vaccination in both the LNT and LNTPM groups in Izmir was much higher than those in Van ( P º 0.0001 and P ＝ 0.0007, respectively). No differences were detected in terms of predisposing factors between the two groups. The distribution of LNT cases according to age is shown in Fig. 2. Cases were more prevalent in the age range of 21–30 years for both LNT groups ( P º 0.05). Cases were also more prevalent in the early decades in Van. None of the patients had alcohol dependence, although 10 (34.5z) declared a smoking habit with a mean duration of 10 years. Additional disorders were found in 20 (22.2z) LNT patients. Three patients had chronic renal failure; 2 each had diabetes mellitus, malignancy, acute rheumatic fever, psychotic disorders, cardiac disorders, and gastritis; and 1 had rheumatoid arthritis, systemic lupus erythematosus. Three (20z) of the LNT-PM patients also had additional disorders

Fig. 2. Distribution of LNT cases by age and by provinces.

(malignancy, thyroid disorder, and chronic renal failure). The most frequent complaint of LNT patients was the occurrence of a palpable mass due to lymphadenopathy (76.6z), followed by night sweats (13.3z), and fatigue (7.7z). Although a palpable mass was the most common symptom in Izmir patients, fatigue, cough, and weight loss were also found with remarkable frequency. In the Van group, a palpable mass was the major symptom, and the incidence of this symptom in the Van and Izmir groups showed a statistically significant difference ( P º 0.001). A definitive LNT diagnosis was established on the basis of histopathological findings in 92 (97.8z) cases. The disease was diagnosed in 2 cases by clinico-radiological and bacteriological methods. Thirteen LNT-PM cases (86.6z) were diagnosed histopathologically. One case was diagnosed bacteriologically and another case was diagnosed solely on the basis of clinical findings. Mediastinal lymph node biopsies were performed by VATS in 4 (3.7z) of LNT cases in Izmir. Twelve patients in the Izmir LNT group were also diagnosed 318

Table 3. Clinical symptoms and diagnostic procedures of the LNT patients LNT patient

Symptoms Palpable mass Weakness Night sweat Cough Weight loss Fever Lack of appetite Sputum expectoration Histopathological diagnosis Excision Incision Fine needle Bacteriological tests performed Positive Clinico-radiological diagnosis

Izmir (n ＝ 48) no. (z)

Van (n ＝ 46) no. (z)

26 (54.1) 5 (10.4) 4 (8.3) 5 (10.4) 4 (8.3) 2 (4.1) 1 (2.1) 1 (2.1) 47 22* 2* 12* 13 1 1

44 2 10 1 2 5 2 1 45 37 3 5 13 0 1

(95.6) (4.5) (21.7) (2.1) (4.3) (10.8) (4.3) (2.1)

LNT-PM patient P

0.0001 0.43 0.085 0.86 0.67 0.41 NA NA 1.0 NA 1.0 NA

Izmir (n ＝ 8) no. (z)

Van (n ＝ 7) no. (z)

2 6 1 4 2 1 1 1 8 6 0 2 6 1 0

5 0 2 2 0 0 1 0 5 4 0 1 5 1 1

(25) (75) (12.5) (50) (25) (12.5) (12.5) (12.5)

(71.4) (28.5) (28.5)

(14.2)

P

0.13 0.007 NA 0.60 NA NA NA NA 1.0

1.0 NA

*Total number of cases are unknown. Abbreviations are in Table 2.

Fig. 4. Localizations of lymphadenopathies in total LNT-PM patients.

Fig. 3. Localizations of lymphadenopathies in total LNT patients.

these 4 cases. Hilar enlargement was encountered in 10 (66.6z) cases; chronic fibrotic changes, in 2 (13.3z) cases, and pulmonary infiltration, in 3 (20z) LNT-PM cases. Significant variables predicting LNT in the study included an age of 20–30 years, palpable cervical node, BCG vaccination, positive TST, and mediastinal lymph node involvement in LNT-PM cases. Variables considered to be indicators of socioeconomic status were not generally associated with TB infection, but prevalence of TB infection tended to be higher in wealthier categories.

histopathologically; however, the method of biopsy retrieval was not found in the patient records. Symptoms and diagnostic features of the cases are shown in Table 3. A total of 99 lymphadenopathies (51 in Izmir and 48 in Van) were performed in the 94 LNT cases. In 5 cases, multiple nodes were encountered. Lymphadenopathies were most frequently (56.5z) localized to the cervical region ( P º 0.001). There was no significant difference in lymph node localization between the Van and Izmir LNT groups ( P À 0.05). Cervical and mediastinal nodes were detected at the same frequency in LNT-PM patients, and there was no significant difference in their occurrence between the two provinces. The localization, number, and proportions of the lymphadenopathies are shown in Figs. 3 and 4. In addition, 1 case of larynx TB and 1 case of bone and parotid TB were detected in the Izmir and Van LNT groups, respectively. Three cases were associated with pulmonary TB in the LNT-PM group. Chest radiograms were found to be normal in 90 (95.7z) of the LNT cases. Chronic fibrotic pulmonary changes were detected in 4 cases, with hilar enlargement being observed in 2 (mediastinal lymphadenopathy) of

DISCUSSION Effective treatment regimens and improved living standards have helped to decrease the incidence of TB in most geographic areas in the world. However, the spread of HIV infection and resistance to drugs for treating TB have limited the control of the disease, as a result of which TB has reemerged in some regions. Although the occurrence of pulmonary TB infection fluctuates, the incidence of extrapulmonary TB has continued to increase (9). The incidence of LNT in Turkey ranges from 6.8– 319

43z, and LNT is the second most common form of extrapulmonary TB after TB pleuritis according to the literature (7,8,11,12). In addition, the incidence of LNT has been reported to be between 0.06–5.4z of all TB cases in Turkey (13–15). In our study, LNT was found in 21.3z and 26.7z of the extrapulmonary cases in Izmir and Van, respectively. In agreement with previous studies, our findings showed that although in some areas of lower socioeconomic status, the incidence of TB was higher in males, LNT was more frequent in females (14,16,17). Dundapat et al. have suggested that this phenomenon occurs as a consequence of male-dominated communities, where women experience poorer living conditions, and because young females generally notice differences in their appearance earlier than males (18). In regions of high prevalence, people are more commonly exposed to TB bacilli and show signs of disease at an earlier age. LNT is an early post-primary manifestation of TB. While Van Province represents the characteristics of developing countries, Izmir has the socioeconomic conditions of more developed countries, and therefore, it was expected that the patient features in these two provinces would be different. Thus, poorer socioeconomic conditions could explain the high incidence of cases at younger ages in Van Province, while older age (with associated diseases of old age) promoted reactivation of TB in Izmir. It is also thought that one of the reasons underlying the lower incidence of TB cases in Van is the reporting errors associated with issues of compliance to the national TB program (8). Although BCG vaccination is mandatory in Turkey, it cannot be performed regularly in every district. TST values are not useful for the diagnosis of TB due to the moderate prevalence of TB and of BCG vaccination. Studies have shown that TST positivity is between 24– 77z in the worldwide population, while it is 56–69z in TB patients in Turkey (19). Therefore, TST is a valuable, but non-specific, test for the assessment of TB patients in Turkey. The BCG vaccination rate was higher in Izmir Province than in Van Province while TST positivity was similar in the two provinces. While this could be seen as a normal result of the vaccination ratio, it could also indicate enhanced immune responses to TB infection and/or BCG due to better living standards. It is difficult to distinguish between these possibilities. In the literature, correlations between TST positivity and prior history of contact with TB patients were reported, which could help in the diagnosis of LNT, especially in pediatric patients. However, in an adult group, TST has less diagnostic value (13). In areas of high prevalence, high rates of BCG vaccination can lead to a high rate of false-positive TSTs in the healthy population. False-positive reactions to the TST have also been clearly identified in patients previously infected with nontuberculous mycobacteria. This renders interpretation of positive TSTs difficult, especially in situation where symptoms of an active TB infection are not present (19,20). However, false-negative reactions to the TST can also occur since a proportion of individuals that harbor latent TB does not react to the test. This is particularly the case with infants or HIV-infected patients who have been recently exposed to communicable TB. Another notable group of false-negative TSTs

are those in which there is significant induration but which are erroneously interpreted as negative by health professionals (21). Thus, although TST positivity was high in the present study, it was not considered to be a strong indicator for LNT. There is significant variability in the literature on the occurrence of clinical signs and symptoms of LNT. Jha et al. have noted that weakness, night sweats, and cough occur in 17.8z, 10.3z, and 10.3z of the patients, respectively (16). The most frequent complaint was cervical mass (46z). Dundapat et al. reported weight loss in 85z and fever in 40z of patients (18). The literature supports our findings that LNT patients declare their symptoms according to their perception of their own body and health and their ability to express themselves, which is based on their socioeconomic and educational levels. According to WHO TB guidelines published in 1997, pulmonary parenchymal invasion should be limited to extrapulmonary TB cases. Pulmonary TB refers to a disease involving the lung parenchyma. Therefore, tuberculous intrathoracic lymphadenopathy (mediastinal and/or hilar) or tuberculous pleural effusion, without radiographic abnormalities in the lungs, constitutes a case of extrapulmonary TB. A patient with both pulmonary and extrapulmonary TB is considered to have pulmonary TB (22). Priel et al. have found that 28.8z of LNT cases have pulmonary TB (23). In Turkey, particularly in studies originating from hospitals that treat patients with chest diseases, mediastinal LNT together with pulmonary TB is seen at rates as high as 7.5–20z (4,5). We found that 3 cases in the LNT-PM group with mediastinal lymphadenopathy were diagnosed before pulmonary TB. Interpretation of results can be difficult due to the two-stage nature of TB, which is characterized by an infection and a disease stage. Generally, studies do not clearly differentiate between TB infection and TB disease, and it is not clear how socioeconomic conditions are associated with the risk of infection and the risk of developing disease, or both (20). In most previous studies (24–26), data suggest a strong influence of both household crowding and socioeconomic status on the risk of TB infection. Crowding and socioeconomic status perhaps underlie different causes of TB infection: (i) infections at the household level, which occur through overcrowding in poor households; and (ii) infection at the community level. Wealthier households likely reflect a more urban-type setting, characterized by a greater population density and a higher chance of human interaction, which is likely to foster TB transmission. This effect has been demonstrated in previous studies, which show that in settings with a high prevalence of TB, especially densely populated settings, extensive TB transmission can occur via complex social networks that are likely to be as important as household contact in facilitating transmission (27,28). The limitations of the present study that should be acknowledged include probable reporting errors in Van Province and a relatively low patient number. In summary, LNT occurs as a frequent extrapulmonary form of TB among females, and young adolescents have a significantly higher susceptibility, especially in 320

provinces of lower socioeconomic status. Symptoms are variable and people either gender and any age are under risk in regions of moderate to high TB prevalence. The TST is not helpful for diagnosis. In the presence of mediastinal lymphadenopathy, pulmonary TB should be suspected. LNT should be the primary consideration in the differential diagnosis of lymphadenopathies, particularly in cervical cases.

12. Yoon, H.J., Song, Y.G., Park, W.H., et al. (2004): Clinical manifestations and diagnosis of extrapulmonary tuberculosis. Yonsei Med. J., 45, 453–461. 13. Karagoz, T., Senol, T. and Bekci, T.T. (2001): [Tuberculous lymphadenitis]. J. Toraks, 2, 74–79. 14. Aksel, N., Tavusbay, N.A. and Ozsoz, A. (2005): Lymph node tuberculosis cases. Arch. Lung, 1, 30–33 (text in Turkish with English summary). 15. Rieder, H.L., Snider, D.E. and Cauthen, G.M. (1990): Extrapulmonary tuberculosis in the United States. Am. Rev. Respir. Dis., 141, 347–351 16. Jha, B.C., Dass, A., Nagarkar, N.M., et al. (2001): Cervical tuberculous lymphadenopathy: changing clinical pattern and concepts in management. Postgrad. Med. J., 77, 185–187. 17. Geldmacher, H., Taube, C., Kroeger, C, et al. (2002): Assessment of lymph node tuberculosis in Northern Germany. Chest, 121, 1177–1182. 18. Dundapat, M.C., Mishra, B.M., Dash, S.P., et al. (1990): Peripheral lymph node tuberculosis: a review of 80 cases. Br. J. Surg., 77, 911–912. 19. Senol, G., Erer, O.F., Yalcin, Y.A, et al. (2007): Humoral immune response against 38-kDa and 16-kDa mycobacterial antigens in tuberculosis. Eur. Respir. J., 29, 143–148. 20. Farhadi, A., Parhizagar, A., Rahman, S., et al. (1999): Tuberculosis: the most common cause of fever of undetermined origin in Shiraz. Arch. Iran Med., 4, 1. 21. Boccia, D., Hargreaves, J., Ayles, H., et al. (2009): Tuberculosis infection in Zambia: the association with relative wealth. Am. J. Med. Hyg., 80, 1004–1011. 22. Maher, D., Chaulet, P., Spinaci, S., et al. (1997): Treatment of Tuberculosis: Guidelines for National Programmes. 2nd ed. World Health Organization, Geneva. 23. Priel, I.E., Katz, A.N. and Dolex, E. (1994): Tuberculous lymphadenitis in a general hospital. Harefuah, 127, 438–440. 24. Clark, M., Riben, P. and Nowgesic, E. (2002): The association of housing density, isolation and tuberculosis in Canadian First Nations communities. Int. J. Epidemiol., 31, 940–945. 25. Barker, R.D., Millard, F.J., Malatsi, J., et al. (2006): Traditional healers, treatment delay, performance status and death from TB in rural South Africa. Int. J. Tuberc. Lung Dis., 10, 670–675. 26. Elender, F., Bentham, G. and Langford, I. (1998): Tuberculosis mortality in England and Wales during 1982–1992: its association with poverty, ethnicity and AIDS. Soc. Sci. Med., 46, 673–681. 27. Klovdahl, A.S., Graviss, E.A., Yaganehdoost, A., et al. (2001): Networks and tuberculosis: an undetected community outbreak involving public places. Soc. Sci. Med., 52, 681–694. 28. McElroy, P.D., Rothenberg, R.B., Varghese, R., et al. (2003): A network-informed approach to investigating a tuberculosis outbreak: implications for enhancing contact investigations. Int. J. Tuberc. Lung Dis., 7, S486–S493.

Acknowledgments Authors thank the staff of both dispensary for their kind help.

Conflict of interest None to declare. REFERENCES 1. World Bank (2000): World Development Indicators. The World Bank, Washington, D.C. 2. World Health Organization (2005): Addressing Poverty in TB Control: Options for National TB Control Programmes. World Health Organization, Geneva. 3. Cantwell, M.F., McKenna, M.T., McCray, E., et al. (1998): Tuberculosis and race/ethnicity in the United States: impact of socioeconomic status. Am. J. Respir. Crit. Care Med., 157, 1016–1020. 4. Tatar, D., Gunes, E.Y., Ozacar, R., et al. (2001): Lymph node tuberculosis. Respir. Dis., 12, 207–211 (text in Turkish with English summary). 5. Powel, D.A. (1994): Tuberculous lymphadenitis. p. 113–120. In Schlossberg, D. (ed.), Tuberculosis. Springer Verlog, New York. 6. Golden, M.P. and Vikram, H.R. (2005): Extrapulmonary tuberculosis: an overview. Am. Fam. Physician, 72, 1761–1768. 7. Akto ¾gu, S., Yorgancƒoglu, A., ǃrak, K., et al. (1996): Clinical spectrum of pulmonary and pleural tuberculosis: a report of 5,480 cases. Eur. Respir. J., 9, 2031–2035. 8. Gumusº lu, F., Ozkara, S. and Ozkan, S. (2007): T äurkiye'de Verem Savasº ƒ, 2007 Raporu. Verem Savasº ƒ Dairesi Basº kanlƒ ²gƒ, Ankara (in Turkish). 9. World Health Organization. Treatment of Tuberculosis Guidelines for National Programmes. 3rd ed. Online at〈http://www. who.int/docstore/gtb/publications/ttgnp/PDF/ 2003.313pdf〉. Accessed 17 May 2006. 10. Turkish Statistical Institute. Online at〈www.tuik.gov.tr〉. Accessed 22 Mach 2007. 11. Tatar, D., Coskunol, I., Aydƒn, M., et al. (2003): The retrospective analysis of cases with extrapulmonary tuberculosis diagnosed at Izmir Esrefpa-sa Tuberculosis Dispensary between 1995–2000. Arch. Lung, 3, 107–112 (text in Turkish with English summary).

321