Can cytomorphology of granulomas distinguish sarcoidosis from tuberculosis? Retrospective study of endobronchial ultrasound guided transbronchial needle aspirate of 49 granulomatous lymph nodes

INTRODUCTION

Sampling of mediastinal lymph nodes by endobronchial ultrasound guided transbronchial needle aspiration (EBUS-TBNA) and endoscopic ultrasound guided fine-needle aspiration (EUS FNA) have been validated for staging in patients with lung cancers. EBUS-TBNA has also been successfully used to diagnose centrally located lung tumors.[123] In Western countries, it has also emerged as a safe tool to establish granulomatous pathology in suspected cases of sarcoidosis.[4] In countries with high prevalence of tuberculosis (TB), mediastinal lymphadenitis is most often due to TB while sarcoidosis is a less common cause.[5] The differential diagnosis of TB and sarcoidosis is very challenging in these regions. The clinic-radiologic features of two are overlapping.[6] Both can present with non-caseating granulomas (NCG). Results of culture and Zeihl-Neelsen (ZN) staining are variable with low sensitivity in non-necrotic lesions. The smear positivity (ZN stain) ranges from 18% to 75% in various studies and is lower in non-necrotic lesions. The culture positivity though higher than smear positivity is also variable and is reported as approximately 50% in granulomatous lesions.[7] Most patients with granulomatus lesions are put on therauptic anti-tubercular therapy (ATT) trials. A diagnosis of sarcoidosis is sometimes only made after failure of response to repeated courses of ATT.

We carried out a pilot study to explore if cytomorphologic features of granulomas could help in differentiation between sarcoidosis and TB. Final diagnosis in these patients was based on clinical, microbiologic and follow-up studies.

MATERIALS AND METHODS

EBUS-TBNA smears of 49 consecutive patients with a final cytologic diagnosis of granulomatous lymphadenitis were reviewed. Both Papanicolaou and May Grunwald Giemsa stained smears were available. The smears were reviewed by three trained cytopathologist without knowledge of clinical or microbiological data.

Features looked for were the presence of granulomas, size and number of granulomas, loose or compact granulomas, presence or absence of necrosis and inflammatory exudate.

Based on these features granulomas were classified as caseating granulomas (CG) or NCG. The latter (NCG) were further divided into two categories:

NCG I – The number of granulomas was less and the size was small. They were made of few epithelioid cells that were well spaced out. The epithelioid cells appeared to be falling away from margins of granulomas [Figure 1].

Close

Figure 1

Small granulomas made of few epithelioid cells. The epithelioid cells are well spaced out

Export to PPT

NCG II – The number of granulomas was more and they were composed of numerous tightly packed epithelioid cells [Figure 2].

Close

Figure 2

Large granulomas, with numerous epithelioid cells, packed tightly

Export to PPT

Presence or absence of giant cells was not used as a criterion in this classification.

A case was labeled as TB if CG or NCG I were observed while cases with NCG II type of granulomas were labeled as sarcoid. This hypothesis was based on anecdotal observations by us while reporting mediastinal granulomatous lymphadenopathy (LAP). Cytological categories so enunciated were correlated with the final diagnosis in each case. A final diagnosis of TB was considered if acid fast bacilli (AFB) could be demonstrated on ZN staining of smears or culture for Mycobacterium tuberculosis (MTB) was positive or the patient showed therauptic response to ATT. A final diagnosis of sarcoidosis was considered in patients where no AFB was demonstrated in smears and/or culture; and therauptic response to steroids or hydroxychloroquine was seen on follow-up.

RESULTS

Table 1 depicts the type of granulomas in these cases. NCG lymphadenitis were observed in 34 cases while in 14 cases caseation was observed.

Table 1 Distribution of granulomas and AFB positivity on EBUS-TBNA

The NCG were further classified into category I or II. 4 cases could not be classified further due to scant material and were labeled as NCG-Not otherwise specified (NOS)

Amongst the CG AFB was demonstrated by ZN staining in eight of the 13 cases. It was not done in one case due to scant material. The case with only necrosis showed strong AFB positivity. Amongst the NCG, ZN staining was negative in 32 cases and not done in 2 cases. Culture for MTB using material obtained by EBUS was performed in 10 cases with CG and was positive in 3 cases. In two of the 3 cases, smears were negative. MTB culture was performed in 26 of the 34 cases with NCG and was positive in 2 cases. In another 2 patients, AFB was demonstrated in bronchioloalveolar lavage (BAL) specimen.

Follow-up and correlation of data

Based upon clinical and radiologic features, EBUS-TBNA results (where helpful in picking up AFB or CG) and therauptic follow-up for at least 6 months, final diagnosis are given in Table 2.

Table 2 Correlation of cytomorphologic features of granulomas with final diagnoses

Eight patients did not complete 6 months follow-up. As is evident from Table 2, cytomorphologic classification of NCG did not correlate with final diagnosis of patients. Of the cases showing NCG I pattern (which was considered consistent with TB in our study), almost 55.5% patients turned out to be sarcoid.

DISCUSSION

TB is the major cause of mediastinal LAP in India. Until the recent advent of EUS FNA/EBUS-TBNA, patients with mediastinal LAP warranted therauptic ATT trial. However, recently these modalities are being increasingly used to elucidate etiology and exclude alternate diagnosis of lymphoma, metastatic carcinoma and sarcoidosis.[8] In a recent publication of EUS FNA of mediastinal lymph nodes from our center, granulomatous pathology was present in 206 of 269 cases studied. Only 76 of 206 cases could be further categorized as TB based on the presence of necrosis or AFB positivity on ZN staining. Though EUS FNA confirmed the presence of granulomatous lymphadenitis, the dilemma of TB versus sarcoidosis could not be resolved in patients with overlapping clinico-radiologic features from tubercular endemic areas.[9]

Few studies in the literature have evaluated the role of EBUS-TBNA in diagnosis of TB. Most of these do not address the dilemma of TB versus sarcoidosis.

Navani et al.,[10] reported that with the use of EBUS-TBNA, a diagnosis of TB could be established in 146 of the 156 patients with intrathoracic tubercular lymphadenitis. Their study included only patients with a confirmed diagnosis or unequivocal clinical and radiologic response to ATT during follow-up for a minimum period of 6 months. Cytological findings in EBUS-TBNA smears were granulomas with necrosis (68 cases), granulomas without necrosis (58 cases) and only necrosis (8 cases). Further smears were positive for AFB in 27 cases while in 74 cases AFB grew on culture. Smears were positive with negative culture in 8 cases. Thus, microbiologic confirmation was obtained in 82 of the 146 cases. They concluded that granulomas without caseation along with supportive clinical evidence and response to therapy be considered as consistent with final diagnosis of TB. Study of Navani et al., cannot be compared with the present study as the sample selection is different. They included only patients with a confirmed diagnosis of TB and studied cytological and microbiologic findings. Present study included cases with granulomas in EBUS-TBNA smears without consideration of clinical and radiologic features. However like Navani et al., we also found NCG in patients finally diagnosed as TB.

Hassan et al.,[11] demonstrated that EBUS-TBNA performed well in diagnosis of isolated intrathoracic TB. It confirmed TB in 19/24 patients with high pre-test probability of TB.

Fristcher-Ravens et al.,[12] did a prospective study to investigate diagnostic yield of EUS FNA in differentiation between sarcoidosis and TB. They differentiated TB from sarcoidosis predominantly based on “dirty background” in TB and “clean background” in sarcoidosis. In their study, EUS FNA cytology alone led to diagnosis of TB in 24/28 cases. When bacteriological studies were taken into account EUS FNA provided diagnosis in 27/28 cases. However, 3 patients with a “clean background” and cytologic diagnosis of sarcoidosis had positive culture for MTB.

In the present study, we attempted to classify NCG into TB and sarcoidosis based on cytologic features. The hypothesis was based on our observation that two distinct morphologic patterns of granulomatous lymphadenitis were observed in smears. Few small loose granulomas were observed more commonly in TB compared to numerous compact granulomas in sarcoidosis. However, this observation could not be substantiated in the present study.

Out of 49 patients, 27 were confirmed to be TB. Tubercular etiology could be established on EBUS-TBNA (cytology and microbiology) in 17 patients. Another 2 patients had AFB demonstrated in BAL specimen. 8 patients were confirmed as TB and 14 as sarcoidosis based on clinical and radiologic grounds. Sampling of mediastinal lymph nodes by EUS FNA and EBUS-TBNA is been increasingly done nowadays in tubercular endemic regions with clinical suspicion of TB versus sarcoidosis. Though different patterns of granulomas are observed in smears, these do not correlate with the diagnosis of TB and sarcoidosis. Further categorization of granulomatous pathology is not possible in a large subgroup of patients who have NCG with negative ZN staining and culture.

COMPETING INTERESTS STATEMENT BY ALL AUTHORS

The authors declare that they have no competing interests.

AUTHORSHIP STATEMENT BY ALL AUTHORS

All authors of this article declare that we qualify for authorship as defined by ICMJE http://www.icmje.org/#author. Each author has participated sufficiently in the work and takes public responsibility for appropriate portions of the content of this article. GK, AD, PB and KV contributed in conception and design of the study. GK, AD, JA and KV contributed in data acquisition, analysis and interpretation of data. GK, JA, PB and KV drafted and critically analyzed the article. All authors read and approved the final manuscript. Each author acknowledges that this final version was read and approved.

ETHICS STATEMENT BY ALL AUTHORS

This was a retrospective analysis of the data and all the procedures were done after obtaining informed consents from patients; hence, ethics committee had given exemption.

EDITORIAL/PEER-REVIEW STATEMENT

To ensure the integrity and highest quality of CytoJournal publications, the review process of this manuscript was conducted under a double blind model (authors are blinded for reviewers and vice versa) through automatic online system.