Value of Different Bronchoscopic Sampling Techniques in Diagnosis of Sarcoidosis

A Prospective Study of 151 Patients

Abstract

Background: 

The exact position of routine bronchoscopic sampling techniques in diagnostic workup of sarcoidosis is often debated. Herein, we ascertain the role of transbronchial needle aspiration (TBNA), endobronchial ultrasound-guided TBNA (EBUS-TBNA), transbronchial lung biopsy (TBLB), and endobronchial biopsy (EBB) in diagnosis of sarcoidosis.

Methods: 

Consecutive patients with suspected sarcoidosis who underwent fiberoptic bronchoscopy were studied. TBLB, EBB, TBNA, or EBUS-TBNA was performed as indicated in a standardized manner. A diagnosis of sarcoidosis was established based on the finding of non-necrotizing granulomas or on clinical grounds at 6-month follow-up. Individual and cumulative yield of various procedures and their correlation with clinicoradiologic parameters was analyzed.

Results: 

Of the 164 patients studied, 151 were finally diagnosed as sarcoidosis. Granulomas were demonstrated in 127 (84.2%) patients. Diagnostic yield of TBLB, EBB, TBNA, and EBUS-TBNA was 68.7%, 49.6%, 22.43%, and 57.1%, respectively. Cumulative yields of various procedures were: EBB+TBLB 81.4%; TBLB+TBNA 73.7%; TBNA+EBB 62.9%; TBLB+EBB+TBNA 86.9%; and TBLB+EBB+EBUS-TBNA 86.4%. In those with visible mucosal abnormalities, TBLB+EBB conferred the highest diagnostic yield (92.8%). Clinical findings or radiologic stage had no impact on diagnostic yield.

Conclusions: 

TBLB is an important tool in bronchoscopic diagnosis of sarcoidosis. If endobronchial abnormalities are seen during bronchoscopy, TBLB with EBB gives the best results otherwise TBLB combined with conventional TBNA and EBB or EBUS-TBNA are required to maximize the diagnostic yield.

Sarcoidosis is a multisystem granulomatous disorder of unknown etiology, most commonly affecting the lungs and intrathoracic lymph nodes. The diagnosis of sarcoidosis is made on a combination of compatible clinicoradiographic findings and demonstration of typical non-necrotizing epithelioid cell granulomas in the absence of a competing diagnosis such as tuberculosis, fungal disease, or malignancy.¹ In view of its remarkable clinical similarities to other granulomatous conditions, histopathologic confirmation in patients with suspected sarcoidosis is mandatory in countries with high prevalence of tuberculosis. As lung and hilar/mediastinal lymph nodes are the most affected sites in sarcoidosis, sampling through the flexible bronchoscope is the preferred mode of obtaining tissue for confirmation of diagnosis.²

Diagnostic procedures like transbronchial lung biopsy (TBLB), transbronchial needle aspiration (TBNA), and endobronchial biopsy (EBB) are routinely used to obtain samples for the pathologic confirmation of pulmonary sarcoidosis. Bronchoalveolar lavage is used mainly for research purposes or for exclusion of tuberculosis. Real-time convex probe endobronchial ultrasound-guided TBNA (EBUS-TBNA) has shown immense potential, however, it is costly, labor intensive, and still has limited availability, especially in low-income and middle-income countries.³ Each of the commonly used procedures like TBNA, EBUS-TBNA, TBLB, and EBB have limitations in terms of safety and efficiency besides the implication on both time and cost.

The exact positioning of various bronchoscopic procedures in diagnostic workup of sarcoidosis remains unclear, and only few studies have explored the comparative and additive yields of different sampling procedures. We had previously reported the diagnostic yield of TBLB, EBB, and TBNA in patients with sarcoidosis.^4,5 Herein, we studied the individual and cumulative yield of all the 3 (TBLB, EBB, and TBNA) bronchoscopic sampling techniques, in an attempt to ascertain the best possible strategy for maximizing the efficiency of flexible bronchoscopy in diagnosis of sarcoidosis.

PATIENTS AND METHODS

This was a prospective observational study conducted between July 2010 and October 2011 at a tertiary care research institute in North India. The study protocol was approved by the Institute Ethics Committee (No. 8420 PG-lTrg-l0/300) and a written informed consent was taken from all participants.

Inclusion and Exclusion Criteria

Consecutive patients presenting with clinicoradiologic features suggestive of sarcoidosis and an indication for bronchoscopy were eligible for inclusion in the study. The baseline diagnostic evaluation included clinical history, physical examination, laboratory tests (tuberculin skin test, complete blood count, coagulation profile, liver and renal function tests, angiotensin-converting enzyme levels, and HIV serology), chest radiograph, and computed tomography (CT) of the chest. Patients with any of the following were excluded: pregnancy, hypoxemia (SpO₂<90%) on room air, poor lung function (forced expiratory volume in first second<1 L), patients with deranged clotting profile (prothrombin time >3 s above control; activated partial thromboplastin time >10 s above control, platelet count <50,000/μL), patients already initiated on glucocorticoids, diagnosis of sarcoidosis possible on minimally invasive techniques, such as skin biopsy or peripheral lymph node biopsy, and failure to provide informed consent.

Bronchoscopy Preparation

All patients were nebulized with 4% lignocaine solution immediately before the procedure. Topical 2% lignocaine gel was applied in the nasal cavity and 2% lignocaine was instilled over the vocal cords and the airways. No sedation was used during or before the bronchoscopy except in those undergoing EBUS-TBNA who received intravenous midazolam and pentazocine for sedation and analgesia, respectively.

Bronchoscopy Procedure

Flexible bronchoscopy was performed in a well-equipped bronchoscopy suite in an outpatient setting. The procedure was performed in the supine position by transnasal (or oral) route by experienced faculty or by fellows under direct supervision, using a fiberoptic bronchoscope (BF-1T20 or BF-TE2; Olympus, Japan; FB-19TV; Pentax, Japan). All abnormalities in the bronchial tree (erythema, plaques, nodules, or cobblestone mucosa) were noted.

EBB (at least 4 specimens) was performed with the standard biopsy forceps (FB15C; Olympus) from the abnormal looking areas if any; or from secondary carinal areas. Transbronchial lung biopsies (at least 4 specimens) were obtained without fluoroscopy using the standard biopsy forceps (FB19C; Olympus) from the site of maximal involvement on CT chest or from right lower lobe, if CT chest was normal. TBNA was carried out on right paratracheal and subcarinal lymph nodes. Lymph nodes were considered enlarged and amenable for TBNA when the short axis nodal diameter on CT chest was ≥1 cm. Three passes were attempted in each lymph node station using the conventional 21 G TBNA needle (Smooth Shot; Olympus) without an onsite cytopathologist. In some patients, EBUS-TBNA was performed using an EBUS bronchoscope (BF-UC 180F; Olympus) with a compatible ultrasound processor (EU-ME1; Olympus) at an ultrasonic frequency of 7.5 MHz. In those who underwent TBLB, a chest radiograph and ultrasound of the chest was performed 2 hours after the procedure to look for any iatrogenic pneumothorax. Complications if any were carefully recorded.

Specimen Preparation

Endobronchial and TBLB specimens were fixed with formalin and transported to the histopathology laboratory. TBNA specimen was aspirated from the needle onto a slide using air in the syringe, and the smear was made with the help of another slide which was immediately fixed in a 95% alcohol jar. Biopsy and TBNA slides were stained with Ziehl-Neelsen stain for mycobacteria, besides the stains routinely used for morphologic evaluation. The TBNA aspirate was also sent for culture of mycobacteria (mycobacterial growth indicator tube technique) and fungi.

Endpoints

The primary endpoint was the detection of granuloma on any of the bronchoscopy procedure. The secondary endpoint was the rate of serious adverse events (pneumothorax, pulmonary hemorrhage of at least 100 mL, mediastinal infection, and death) related to the diagnostic procedure. The TBNA was defined as successful if there was preponderance of benign lymphocytes on aspirate. Adequacy of the tissue sample on TBLB and EBB was independently reported by the pathologist (A.B.). The diagnostic yield of the procedure/s (individually or in combination) was defined as the number of patients with detected granulomas by a particular procedure/s divided by the number of patients undergoing that procedure/s among the patients with a final diagnosis of sarcoidosis.

A diagnosis of sarcoidosis was made on the following criteria: (a) clinicoradiologic presentation consistent with sarcoidosis; (b) demonstration of non-necrotizing granuloma; (c) absence of other known causes of granulomatous lung disease such as mycobacteria or fungal diseases; and (d) response to steroid therapy. In those without demonstration of granuloma, clinical stability or improvement on follow-up for at least 6 months or response to steroids with no alternative diagnosis at 6-month follow-up was considered diagnostic of sarcoidosis.

Statistical Analysis

Statistical significance was assumed at a P-value of <0.05. The categorical variables were analyzed using χ² test, whereas the continuous variables were analyzed using Mann-Whitney U test.

RESULTS

During the study period, 164 patients with a clinical possibility of sarcoidosis were included. Thirteen patients were excluded, 6 patients had alternative diagnosis on histology (tuberculosis in 4 and 1 each with allergic bronchopulmonary aspergillosis and adenocarcinoma) and 7 were lost to follow-up. A final diagnosis of sarcoidosis was established in 151 patients (127 patients were proven as sarcoidosis by histology or cytology, 19 patients showed rapid improvement on steroids given on clinical possibility of sarcoidosis, and 5 patients had spontaneous resolution of disease consistent with natural history of sarcoidosis). The mean age of study group was 49.3 years (range, 18 to 68 y) and 79 (52.3%) were males. Almost all patients were symptomatic with cough being the most common symptom reported. Nearly one third of patients had extrapulmonary or systemic manifestations; fever, arthralgia, and weight loss being the most common. More than half of the patients were in radiologic stage 1 and had normal spirometry. On bronchoscopy, mucosal abnormalities (mucosal granularity, nodularity, edema, or hyperemia) were seen in 29 (19.2%) patients. Detailed characteristics of the study population are presented in Table 1.

TABLE 1: Clinicoradiologic Details of the Study Population

TBLB, EBB, TBNA, and EBUS-TBNA were performed in 141, 125, 76, and 28 patients, respectively. Adequate tissue was obtained in over 95% patients who underwent TBLB or EBB and adequate aspirate could be obtained in 93% of EBUS-TBNA and 77.6% of TBNA. Individually, TBLB had the highest diagnostic yield (of finding granulomatous inflammation) of 68.7% followed by EBUS-TBNA (57.1%). The yield of TBLB was significantly higher than EBB (P=0.001) or TBNA (P<0.0001) but not EBUS-TBNA (P=0.23) (Table 2).

TABLE 2: Diagnostic Procedures and Their Yield

Diagnostic yield of EBB was significantly higher when performed in patients with visible endobronchial abnormalities compared with normal looking mucosa (75.8% vs. 41.7%, P<0.05). The outcome of bronchoscopic procedures had no significant association with symptoms and spirometry. Also, there was no correlation of diagnostic yield of various procedures with radiologic staging on chest radiograph, even though TBLB was positive in all the patients with stage 3 or 4 disease (Table 3).

TABLE 3: Positivity of Various Procedures Stratified by Chest Radiographic Stage

Analysis for additive yield of various sampling procedures revealed that adding different procedures led to increase in diagnostic yield (Table 4). The best possible diagnostic yield was observed with a combination of TBLB and EBB in patients with endobronchial abnormalities (26/28, 92.8%). Overall TBLB and EBB gave a diagnostic yield of 81.4%. All 3 procedures (TBLB, EBB, and TBNA) were performed in 83 patients (61 with conventional TBNA and 22 with EBUS-TBNA) and a diagnostic yield of over 86% could be achieved in these patients. There was no difference in diagnostic yield with the type of TBNA used when combined with TBLB and EBB (86.9% and 86.4% in the conventional TBNA and EBUS-TBNA groups, respectively) (Table 4). Addition of TBLB significantly improved the yield of other procedures (P=0.003 for TBLB+EBB vs. EBB+TBNA; P=0.0003 for TBLB+EBB+TBNA vs. EBB+TBNA). Even in the small subgroup of EBUS-TBNA, combining TBLB leads to significant increase in the diagnostic yield (20/23 for both procedures vs. 13/23 for EBUS alone, P=0.049). Addition of EBB to EBUS-TBNA was not associated with significant additional yield (19/25 for both procedures vs. 14/25 for EBUS alone, P=0.14).

TABLE 4: Individual and Additive Diagnostic Yield of Various Bronchoscopic Procedures: Correlation With Mucosal Findings

One patient developed pneumothorax after bronchoscopy which was managed with simple needle aspiration. There was no episode of major hemorrhage.

DISCUSSION

The results of this study reaffirm that flexible bronchoscopy is an easy and safe technique, and conventional bronchoscopy procedures are effective tools for demonstrating granulomatous inflammation in the diagnostic workup of a suspected case of sarcoidosis. TBLB remains the single most useful procedure and when combined with the other 2 procedures [EBB and/or (EBUS) TBNA], it imparted significant improvement to their outcomes.

Flexible bronchoscopy has transformed the diagnosis of sarcoidosis with shift from invasive procedures (like rigid bronchoscopy, mediastinoscopy, and surgical lung biopsy) to a simple day-care procedure. The most commonly used bronchoscopy techniques have included TBLB and EBB.² The technique of TBNA was made popular by Wang et al⁶ by his innovation of using a flexible needle through the flexible bronchoscope. However, 4 decades later TBNA still remains underutilized.⁷ In last decade, the advent of EBUS-TBNA has added excitement to the diagnostic armamentarium by the virtue of sampling mediastinal lymph nodes under direct visualization.⁸ Large volumes of data have proven its efficacy and safety in obtaining tissue diagnosis in sarcoidosis.³ The recently published “Granuloma Trial,” a randomized controlled trial (RCT) found endosonographic TBNA better than conventional biopsies in demonstration of granulomas.⁹ However, the EBUS equipment is expensive and not routinely available especially in developing countries.⁸ This study demonstrates excellent efficiency of conventional bronchoscopic procedures for the diagnosis of sarcoidosis, besides the EBUS.

TBLB has been the bronchoscopic procedure of choice for diagnosis of sarcoidosis.^10–12 TBLB was found to be positive in 68.8% which is comparable with the earlier studies carried out at our institute where the positivity rate ranged from 69% to 72%.^4,13 The TBLB yield has varied between 60% and 97% in several other studies.^14–16 Presence of radiologic stage 2 or 3 and the number of biopsies obtained have been associated with higher yields of TBLB,^14,17 however, in this study none of the symptoms, spirometry or radiologic stage were predictors of a positive biopsy, even though TBLB was positive in all the patients with stage 3 or 4 disease. This is because of the fact that even patients of stage I sarcoidosis with normal chest radiographs invariably demonstrate granuloma on surgical lung biopsy.¹⁸ In a previous study on correlation of FVC with histologic grading also we did not find any correlation between TBLB yield and chest radiographic stage.¹⁹ TBLB, however, is usually not performed in patients with severe restriction (forced expiratory volume in first second<1 L) for the fear of pneumothorax leading to critical hypoxia. Although in this study no such exclusions were required and TBLB was safely performed.

EBB has a lower yield than TBLB that varies between 20% and 62%.^20–22 Visible mucosal abnormalities give higher yields of 54% to 91%, although 20% to 40% of normal-appearing mucosa can also show granuloma on histology.^4,23,24 This was reconfirmed in this study, and in patients with mucosal abnormalities EBB combined with TBLB without any TBNA maximized the diagnostic yield of bronchoscopy to over 90%.

TBNA is believed to be a low-yield procedure and concerns regarding its safety, risk of damage to the scope, and the steep learning curves have hindered its widespread use.^7,25 The availability of EBUS-TBNA has significantly changed this scenario. EBUS-TBNA successfully samples mediastinal lymph nodes in >90% of cases and diagnostic yield varies with different diseases.²⁶ In a recent meta-analysis, we found that the diagnostic yield of EBUS-TBNA (79%) was better than the individual diagnostic yield of any of the procedures observed in this study.³ In a small RCT, the diagnostic yield of EBUS-TBNA and TBLB for detecting granulomas in sarcoidosis was 94% and 37%, respectively.²⁷ Subsequently the larger RCT (Granuloma Trial) has also shown that EBUS-guided or EUS-guided mediastinal lymph node sampling could detect granulomas in higher proportion (74%) of patients compared with conventional TBLB and EBB (48%). However, both these studies have not reported on the yield of TBLB and/or EBB added to the TBNA. In addition, the positivity of conventional biopsies were rather low in both these studies (<50%) compared with the accepted yield on around 60% to 70% (vide supra). In fact in 1 study, there was no difference in the diagnostic accuracy of EBUS-TBNA and TBLB (84% vs. 78%, P=0.77) and combined EBUS-TBNA and TBLB procedures yielded a diagnostic accuracy of 100%.²⁸ In another study, EBUS-TBNA was found superior to TBLB only in stage 1 patients.²⁹ Although EBUS-TBNA seems to be the single most useful procedure for demonstrating the granulomatous inflammation in sarcoidosis, yet diagnostic yield is likely to improve when combined with TBLB. This is also borne out by the Granuloma Trial where almost 25% of the study participants required another test for confirmation when only EBUS-TBNA was performed.⁹ In a recent meta-analysis, we found good diagnostic yield (62%) of conventional TBNA in sarcoidosis with practically no complication reported in >900 patients. Moreover, it was also found that cumulative yield of TBNA and TBLB was 83%,³⁰ which was higher than the pooled yield observed with EBUS in the meta-analysis (79%)³ or the Granuloma Trial (74%).⁹ Further, in a randomized trial we have recently shown that routine bronchoscopic techniques (conventional TBNA plus TBLB and EBB) are as good as EBUS-TBNA.³¹

The approach of combining different sampling procedures for optimizing the yield of fiberoptic bronchoscopy was also reported in earlier studies (before the advent of EBUS-TBNA), where combining TBLB and EBB with TBNA or bronchoalveolar lavage were found to give maximal results reaching up to 90%.^32,33 However, the results of these studies are limited by lead-time bias and small sample size.

What is the significance of present study in the current era of EBUS? This study demonstrates that excellent diagnostic yields can be achieved with conventional bronchoscopic techniques, when multiple sampling techniques are used in the same patient. Although EBUS-TBNA was clearly superior to TBNA, the addition of EBB and TBLB to either of them had similar additive effects. In fact in visibly abnormal mucosa, combined TBLB and EBB had the highest yield and there is no further requirement of TBNA. Overall also, a combination of EBB and TBLB can provide diagnostic yields similar if not better than EBUS-TBNA alone.^3,9

Finally, this study is not without limitations. Besides a single-center study, all the procedures, that is, TBLB, EBB, TBNA/EBUS-TBNA were not done in all the patients and the numbers of EBUS-TBNA were rather few. The low numbers of EBUS-TBNA was due to the fact that the equipment was purchased towards the completion of the study. It is possible that the real advantages of EBUS over conventional TBNA and its additive value to EBB or TBLB is influenced by very low numbers of EBUS procedures performed in this study and that too in our initial learning phase. Also, this was an observational study and the diagnostic procedures were performed in the real-world situation of a busy tertiary care hospital catering to large volume of patients derived from a population with limited resources.

In conclusion, in the absence of EBUS multiple conventional bronchoscopic diagnostic procedures are beneficial over 1 diagnostic procedure when sarcoidosis is suspected. Even when EBUS is available, the yield is improved when it is used in combination with TBLB. The use of 1 modality does not provide a diagnostic yield as high as multiple modalities, regardless of the type of TBNA performed. The unavailability of EBUS-TBNA in a resource-constrained setting should not be an obstacle in diagnosing sarcoidosis.