Correlation of fine needle aspiration cytology, smear and culture in tuberculous lymphadenitis: a prospective study.G Nataraj, S Kurup, A Pandit, P Mehta
Seth G.S Medical College and K.E.M. Hospital, Parel, Mumbai - 400012, India. , India
BACKGROUND AND AIM: Bacteriological studies are necessary to confirm the diagnosis of tuberculous lymphadenitis, as cytological appearances mimic other granulomatous lesions. The objective was to assess the diagnostic role of culture of fine needle aspiration done on clinically suspected cases of tuberculous lymphadenitis and to determine the prevalence of drug resistance in M. tuberculosis isolates. SETTING AND DESIGN: A prospective, double-blind study over a period of one year in a tertiary care hospital. MATERIAL AND METHODS: Fine needle aspiration cytology and culture were done on 250 patients with clinical suspicion of tuberculous lymphadenitis. STATISTICAL ANALYSIS: Data was statistically analysed using chi square test. Sensitivity, specificity, positive predictive and negative predictive values and likelihood ratio were also calculated. RESULT: Of the 161 cytologically or microbiologically proven cases of tuberculous lymphadenitis, cytological changes consistent with tuberculosis were observed in 133 patients, out of which mycobacteria were isolated in 102 aspirates. Mycobacteria were also isolated from 28 aspirates cytologically missed as tuberculous lymphadenitis. Of the 130-mycobacterial isolates, 5 were non-tuberculous mycobacteria. Culture positivity was significantly higher (P<0.001) than smear positivity. Drug susceptibility studies showed resistance to one or more drugs in 61% of isolated strains with maximum resistance to isoniazid (16% primary and 48% secondary) and minimum to ethambutol (4% primary and 12% secondary). CONCLUSION: Culture for mycobacteria should be carried out on all aspirates from patients suspected with tuberculous lymphadenitis.
Keywords: Adolescent, Adult, Antitubercular Agents, administration &dosage,pharmacology,Biopsy, Needle, Child, Double-Blind Method, Drug Resistance, Microbial, Female, Human, India, Male, Microbial Sensitivity Tests, Middle Age, Mycobacterium tuberculosis, isolation &purification,Predictive Value of Tests, Probability, Prospective Studies, Sensitivity and Specificity, Tissue Culture, Tuberculosis, Lymph Node, diagnosis,drug therapy,microbiology,
Tuberculosis is still a major public health problem in developing countries with a high mortality rate. Lymphadenitis is the most common form of extra-pulmonary tuberculosis. Fine needle aspiration diagnosis of pulmonary and extra pulmonary tuberculosis, is becoming increasingly popular as a diagnostic tool because of its simplicity, rapidity, and performance-friendly nature., Mycobacteria are slow growing and hence culture is not routinely done in all laboratories. Few studies have tried to correlate the cytological findings with microbiological results for the presence of acid-fast bacilli in smears, and culture for mycobacteria., Also, the increasing incidence of resistance in M. tuberculosis isolates is a cause for concern. There are very few reports on the resistance patterns of M. tuberculosis isolates in tuberculous lymphadenitis. The present study was undertaken to determine the role of culture on fine needle aspirates in clinically suspected cases of tuberculous lymphadenitis and to determine the prevalence of drug resistance in M. tuberculosis isolates.
Two hundred and fifty consecutive clinically suspected cases of tuberculous lymphadenitis over a period of one year were enrolled in this prospective study. Patients with enlarged cervical and/or axillary lymph node(s) and with a history suggestive of tuberculosis were included after taking an informed consent. Relevant clinical details were recorded. Fine needle aspiration was performed aseptically with sterile 24 G needle and 10 ml syringe.
A part of the aspirate was used for preparing three smears that were stained by Hematoxylin and Eosin (H & E) for cytology. Two smears were prepared and stained by the Ziehl Neelsen acid fast staining method and the findings were recorded under oil immersion objective. The remaining portion was used for microbiological studies. The cytologists and microbiologists were blinded till the analysis.
Two independent observers recorded the cytological findings for the presence or absence of granulomas, Langerhan’s giant cells, plasma cells, lymphocytes, macrophages, neutrophils and necrosis. The cytological criteria8 for diagnosis of tuberculous lymphadenitis were defined as epithelioid cell granulomas with or without multinucleate giant cells and caseation necrosis.
Each sample was also inoculated on two slants of Lowenstein Jensen medium and incubated at 370C for at least eight weeks. The cultures were read every day for the first week and then weekly thereafter. All the positive cultures were first confirmed for their acid fastness and subsequently identified by using the following criteria: rate of growth, pigment production, growth on MacConkey’s agar, Niacin accumulation, Nitrate reduction, semi quantitative catalase test and para nitro benzoic acid (PNBA) test.9,10 Five aspirates that were considered insufficient to prepare smears were directly inoculated onto Lowenstein Jensen slants.
The diagnosis of tuberculous lymphadenitis was made when the following criteria were met: the presence of epithelioid cell granulomas with necrosis and/or smear positivity for acid fast bacilli and/or positive cultures for mycobacteria.
M. tuberculosis isolates from smear positive cases were studied for their drug susceptibility pattern to primary anti-tuberculous drugs (streptomycin, iso nicotinic acid hydrazide, ethambutol and rifampicin) using the proportion method as per standard protocol.11
Data was statistically analysed using the chi square test. Specificity, sensitivity, positive predictive value (PPV), negative predictive value (NPV) and likelihood ratio of smear and cytology were compared.
Aspirates from 250 patients were examined. The age of the patients ranged from 12 to 55 years, the majority being in the 21-30 age group (42.7%). The male to female ratio was 1:1.3. The majority of the aspirations were from cervical lymph nodes (58%) followed by axillary lymph nodes (42%). Tuberculosis was diagnosed in 161 patients based on cytology and/or culture findings.
Mycobacteria were isolated in 130 cases, of which 125 (96.2%) were identified as M. tuberculosis and five as non-tuberculous mycobacteria (NTM). On further speciation of NTM, two were identified as M. fortuitum and three as M. scrofulaceum. The minimum incubation time for isolation of M. tuberculosis was 21 days and the maximum was 42 days (mean 29 days). One of the five aspirates, which were used for culture studies alone, also grew M. tuberculosis.
Based on the criteria of Das et al cytodiagnosis of tuberculosis was reported in 155 patients, who were categorised into 3 groups [Table - 1]. Based on the present study criteria, only 133 could be classified as tuberculous. Out of these, mycobacteria were isolated in 102 cases. However, mycobacteria were isolated from 28 cases, which were cytologically missed, 11 from group 1 and 17 from suppurative cases.
Of the 133 cytologically diagnosed cases, the highest smear positivity of 77.7% was when the cytological findings were consistent with necrosis, with or without degenerating granulomas. In all cases, culture positivity was significantly higher than smear positivity (P <0.001). The culture positivity varied from 50% in granulomatous lesions to 83.3% in necrotic lesions. When culture was taken as the gold standard, cytology was found to be more sensitive than smear with a higher NPV but smear has the higher specificity, PPV and better accuracy [Table - 2].
Overall, there were 85 smear positive cases. All 85 smear positive cases were culture positive. Of these, M. tuberculosis was isolated from 80 and NTM from five. Drug susceptibilities in these 80 M. tuberculosis strains showed resistance to one or more drugs in 61% [Table - 3]. Of the four primary line drugs tested maximum resistance was observed to INH (16% primary and 48% secondary) and minimum to ethambutol (4% primary and 12% secondary). Acquired multi-drug resistant tuberculosis (MDRTB), i.e. resistance to both INH and rifampicin, was found to be 16% whereas primary MDRTB was only 1%.
The immunological response to M. tuberculosis determines the spectrum of cytological / histopathological patterns seen, the most characteristic feature being granuloma formation., It is also observed that there is an inverse relationship between granulomas and the presence of acid-fast bacilli,, which was the case in our study with the highest smear positivity at 77.7% found in aspirates, which showed absence of granulomas. The overall smear positivity in various reports range from 18% to 75%., In comparison, the overall smear positivity in the present study at 49.4% was on the lower side. This may be due to the fact that we had screened the smears after acid-fast stain and not fluorescent stain.
The culture positivity in our study also varied, depending on the type of tuberculous lesion. The higher smear and culture positivity in necrotic lesions as compared to granulomatous lesions may reflect the role of cell mediated immunity though this finding is not reflected in the study by Ramanathan et al on lymph node biopsies using a battery of media, who had similar culture positivity rates irrespective of the presence or absence of granulomas.
The prevalence of non-tuberculous mycobacteria in our study was 3.85%. M. tuberculosis still appears to be the most common causative agent of lymphadenitis, a finding also seen in other studies done from India. All the isolates reported by Arora and Arora from North India in 1990 were M. tuberculosis. Ramanathan et al had reported a NTM isolation rate of 5.26%.
Nonetheless, the detection of NTM is another reason for culturing all specimens in order to effectively treat and control infections.
There are no reports of studies done to determine the prevalence of drug resistance on M tuberculosis isolates from tuberculous lymphadenitis in India. However, studies done on isolates from respiratory specimens show a very high prevalence of acquired resistance to INH and rifampicin at 68-72% and 49-60% respectively, as compared to 48% and 30% in this study. In our study, both primary and secondary drug resistance was the highest for INH at 16% and 48% respectively and the least for ethambutol at 4% and 12% respectively. The prevalence of drug resistance to INH and rifampicin in isolates from lymph nodes thus reflects the trend seen in pulmonary tuberculosis.
Since culture is more sensitive than smear, the significantly high culture positivity as compared to smear positivity (P<0.001) in all cases is to be expected. More importantly, in 28 cytologically missed cases, culture was positive emphasising the need for incorporating culture as a routine laboratory investigation in all suspected tuberculous lymph nodes. The addition of a more sensitive technique like PCR8 and the availability of improved automated detection systems using enriched media would have helped ascertain the specificity of cytology in culture negative cases.
Missed cytological diagnosis, isolation of non-tuberculous mycobacteria, and prevalence of drug resistance including multi drug-resistant strains justify mycobacteriological studies on all suspected tuberculous lymphadenitis cases, especially in a region of high prevalence.
We would like to acknowledge the support provided by the staff of cytology and microbiology departments and the Dean, Dr. N. A. Kshirsagar.
[Table - 1], [Table - 2], [Table - 3]