Tuberculosis (TB) is an infectious disease caused by various strains of mycobacterium, mainly by Mycobacterium tuberculosis in humans, it has been present in humans since antiquity [1]. TB is a chronic, progressive infectious with a period of latency following initial infectious. It occurs most commonly in the lungs. Pulmonary symptoms include productive cough, chest pain, and dyspnea. Tuberculosis is the leading infectious cause of morbidity and mortality in adults worldwide, killing about 2 million people every year [2]. Sub- Saharan Africa is disproportionately burdened by tuberculosis and multidrug resistant tuberculosis infection. Extra-pulmonary tuberculosis is common in patients with HIV/AIDS [3].

Tuberculous Lymphadenitis (TL) is a chronic specific granulomatous inflammation with caseation necrosis of the lymph node. Suspected TL is difficult to diagnose and remains expensive when applying surgical procedures and becomes easier to diagnose and less traumatic when applying fine needle aspiration and cytology technique [4].

In Rwanda, the number of people infected with tuberculosis is increasing. In 2009, the Ministry of Health through the National Reference Laboratory in collaboration with University of Maryland, School of Medicine (UMSOM) introduced a new technique; the Fine Needle Aspiration and Cytology (FNAC) to help in the diagnosis of suspected TL one form of the extra-pulmonary tuberculosis. The technique applies a fine gauge needle which is inserted into a nodule to withdraw material and cells. The cells are studied under microscope [5,6]. The fine needle aspiration and cytology technique is quick, safe, cheap and effective for the diagnosis of TL. The technique is less traumatic than an open surgical biopsy, and significant complications are usually rare, depending on the body site [7]. We aimed to evaluate the usefulness of the fine needle aspirates to diagnose TL in patients from a low-income country.

Fine needle aspiration and cytology technique: The technique uses a needle gauge of 23-25 G that is inserted into the lymph node and a certain amount of material is withdrawn. The cells are studied under the microscope by a cytologist or a pathologist. This method is commonly used to determine whether a nodule in the mass is benign or malignant [8]. In this technique, a thin, hollow needle is inserted into the mass of a lymph node to extract cells that, after being stained, will be examined under a microscope.

Inclusion and exclusion criteria: All patients presenting with any mass including lymph node, axial or inguinal mass was included and complying with TB signs and symptoms such as night sweat, fatigue, weight loss and or presenting other signs that is likely to be TB suspect were excluded any case without the above-mentioned signs and symptoms especially absence of lymph node, inguinal or axial mass nodes.

Specimens collection, laboratory techniques and results interpretation: We collected lymph node fluid from suspected patients and made smears on separate slides. A total number of 138 specimens were collected from the lymph nodes of suspected patients to have TL. Specimens were collected from out patients attending district hospitals in TB department, seeking for TB treatment and presenting any enlarged mass including lymph nodes, inguinal or axial masses. Triplicate smears were prepared from lymph node aspirates for each specimen and from each patient. All smears were air-dried and stained with hot Ziehl Neelsen (ZN) stain for Acid-Alcohol Fast Bacilli (AAFBs) examination by a laboratory technologist. To exclude any bacterial infection, if pus was present on site of sample collection, Gram stain was performed while Papanicolaou stain for cytology was used to detect features of TL, malignant cells and other pathologies. All laboratory procedures were performed as described previously elsewhere [9,10].

Triplicate smears were labelled and stained as follow: Slide # 1, slide # 2, and slide # 3. On the slide number one, we applied hot ZN method to identify Acid-Alcohol Fast Bacilli (AAFBs). With the slide number two, we used Papanicolaou staining technique to identify tuberculous lymphadenitis features and other pathology [11].

Finally, slide number three was stained with Gram stain technique to exclude other bacterial infections if pus was present at the time of sample collection. The ZN and Gram stained slides were screened under microscope at 100xS by a laboratory technologist while the slides stained by Papanicolaou technique were examined under microscope at 10xS and 40xS by a pathologist.

The majority of the patients tested were aged between 10 to 40 years old. The age group of 10 to 25 years represent 27.5% while the age group of 26 to 40 years represent 30.4%.

Ziehl Neelsen stain results: A total number of 138 fine needle aspiration and cytology (FNAC) have been collected, processed and stained using ZN staining technique. Among 138 specimens, ZN stain yields positive results in 14 cases which represent 10.1% for Acid Alcohol Fast Bacilli (AAFBs). In 116 (84.1%) cases a ZN stain was done and yielded negative results for AAFBs. In 8 (5.8%) cases either ZN was not done or it could not be ascertained whether a ZN stain was done as shown in Figure 1.

Figure 1: Results of smear microscopy by Ziehl Neelsen staining technique

Gram stain results: Out of 138 specimens collected, only 73 (52.9%) specimens contained pus at the site specimen collection and were analysed after Gram stain was done. Among 73 specimens where pus was present, the Gram stain results showed only 2 cases (2.7%) that were positive and the two samples contained Gram positive cocci and Gram negative bacilli respectively. In 71 (97.3%) cases, the Gram stains yield a negative result as shown in Figure 2.

Figure 2: Results of smear microscopy by Gram staining technique

Fine needle aspiration and cytology (FNAC) on lymph node results: Among 138 FNAC specimens collected, 25 (18.1%) cases are TL. A total number of 19 (13.8%) cases show reactive node or inflammatory features in the lymph node. Inadequate samples for any diagnostic interpretation were observed in 33 (23.9%) cases while 20 (14.5%) cases were indeterminate.

A total number of 25 (18.1%) cases show evidence of other pathology outside tuberculosis (including lymphoma, other malignancy benign masses and abscess). The category of "other pathologies" included 3 cases of breast carcinomas, 8 cases of lymphomas, 9 other malignancies, 2 abscesses, 3 benign lesions including (probable salivary gland cysts, and epidermal inclusion cysts as shown in Table 1.

Table 1: Overall Fine Needle Aspirate and Cytology (FNAC) results.

Cytology features of Tuberculous Lymphadenitis: To confirm a patient as having tuberculosis, the presence of Acid Fast Bacilli (AFBs) on smear is compulsory. However, on the FNA cytology we considered some features that are associated with tuberculosis on cytology interpretation. Among those features, include: caseous necrosis, presence of langerhans giant cells and granulomas on a necrotic background.

A diagnosis of tuberculous lymphadenitis was made when smears showed all the three characteristics of tuberculous lymphadenitis (caseous necrosis, presence of Langerhans giant cells and granulomas on a necrotic background). All the above features if present on smear, we considered a case to be supportive or consistent with TL. A case was considered as highly consistent with tuberculosis or definitive for tuberculosis if all those features were present and if we observed AAFBs on the smear.

The sample was qualified as suspicious if the two of the three features were present. Finally, a negative slide was considered if there were no features for TL when there was any feature of the three characteristics mentioned above. A total number of 10 (7.2%) cases were supportive for TL while 15 (10.9%) cases were suspicious for TL. There were no features for tuberculosis in 113 (81.9%) cases as shown in Table 2. Similar results were found elsewhere in the world with in the studies conducted to diagnose TL [12-15].

Table 2: Cytology features of tuberculous lymphadenitis.

Papanicolaou results (vs) Ziehl Neelsen results: In 6 (4.3%) cases, Papanicolaou stain was supportive to the definitive diagnosis of tuberculosis in the presence of a Ziehl Neelsen positive result. In 19 (13.8%) cases the diagnosis is TL through cytology examination by Papanicolaou stain was raised by cytological features on Papanicolaou stain in the presence of a negative ZN result. A total number of 105 (76.0%) cases on Papanicolaou stain did not show any features of tuberculous lymphadenitis and the ZN was also negative. Only 8 (5.7%) cases were positive with Ziehl Neelsen while negative on the basis of Papanicolou stain features as shown in Table 3.

Table 3: Papanicolaou results compared to Ziehl Neelsen (ZN) results.

The age distribution among tested patients shows that the active age is more represented 10 to 40 years (57.9%). The fine needle aspiration and cytology is important technique, and is useful for patient who cannot provide sputum. The fine needle aspiration and cytology is helpful in the diagnosis of one form of extra-pulmonary tuberculosis, tuberculous lymphadenitis.

The ZN staining technique reveals 10.1% while Papanicolaou shows 18.1% of positive results taken separately. The gram stain results help to exclude or confirm bacterial infection. When we compared the FNAC technique to the ZN staining technique, the FNAC is a complementary technique. The cytology results complement ZN results. The FNAC provides more information while compared to ZN staining technique since it reveals other pathologies such as breast carcinomas, lymphomas, benign and malignant lesions, abscesses, and other malignancies as shown in Table 4. Only the presence of Acid Fast Bacilli (AFBs) confirms tuberculosis, however, features observed on the fine needle aspirates and cytology that are associated with tuberculosis are helpful to diagnose tuberculosis. Those features include: caseous necrosis, presence of langerhans giant cells and granulomas on a necrotic background as described previously [16]. To qualify a smear as TL, or to diagnosis tuberculosis while using fine needle Aspirate and cytology, one should consider the following four features: (1) Tuberculosis: Acid Alcohol Fast Bacilli (AAFBs) must be present, (2) Consistent with tuberculosis: AAFBs may not be seen but multiple features of tuberculosis at least three of them, (3) Suggestive or supportive for tuberculosis: No AAFBs seen but few features at least two features of tuberculosis, (4) Negative for tuberculosis: No AAFBs seen and no features of tuberculosis.

Table 4: Features qualification supporting tuberculosis diagnosis.
Other pathology diagnosed by FNAC: The cytology analysis was also able to diagnose other pathologies other than tuberculosis such as malignant or benign cells and reactive nodes. The cytology provides more information from one single sample. For the patient who cannot provide sputum, the FNAC is indicated as cheap and quick technique to be used in the country with low resources-income. Increased number of inadequate samples (23.9%) might be due to inadequate training of nurses and/or clinicians who collected samples and performed the smearing technique (Key: AAFBs: Alcohol Acid Fast Bacilli).

Study limitation: Bacterial culture technique using Lowenstein Jensen (LJ) is considered as gold standard even though it takes several weeks to months to get results [17,18]. In the present study, solid culture using Lowenstein Jensen (LJ) was not performed on all samples which prevent us to compute sensitivity, specificity, positive and negative predictive values for all samples.

In low-income countries fine needle aspiration and cytology is useful to diagnose TL instead of biopsy applying surgical procedures. The technique is less expensive, quick, safe and effective while compared to the biopsy that requires expensive material, equipment, and more time and pathologist specialist for the interpretation. It can be recommended to the district hospital testing package as routine technique in low settings. However, more training is needed to reduce the number of inadequate samples on collection time which might interfere with results interpretation. We recommend further study with large sample size with emphasis on sensitivity, specificity positive and negative predictive to assess the performance of FNAC against Smear microscopy using either Ziehl Neelsen or Fluorescence Microscopy in the diagnosis of Tubercoulous Lymphadenitis.

University of Maryland School of Medicine, Ministry of Health/ Rwanda, National Reference Laboratory, King Faysal Hospital, University Teaching Hospital/Kigali, University Teaching Hospital/ Butare, District hospitals (all).