Comparative study of cytology and immunocytochemistry with trucut biopsy and immunohistochemistry in diagnosis of localized lung lesions: A prospective study.

BACKGROUND: Percutaneous lung biopsy is now a common procedure in pulmonary medicine, and several different techniques are in use. The most common has been the use of a fine needle under computed tomography (CT) guidance combined with the trucut needle for histology. AIM: To evaluate the efficacy of fine needle aspiration cytology (FNAC) and immunocytochemistry in comparison with trucut biopsy and immunohistochemistry in patients with localized intrathoracic lesions suspicious for malignancy.
MATERIALS AND METHODS: Eighty patients with localized mass lesions in the lung on imaging (chest radiograph/CT) were selected for this study over a period of 1 year. FNAC was carried out by a 22 G spinal needle after localization of the mass in the CT scan followed by guided trucut biopsy. Immunocytochemistry and immunohistochemistry were performed as and when required.
RESULTS: The mean age of our study population was 57.6 years and the M:F ratio was 4.2:1. Majority of the lesions were peripheral and in the right lung. Adenocarcinoma was most prevalent (49%), followed by squamous cell carcinoma and small cell carcinoma. Cyto-histopathological concordance was seen in 60 cases (75%). The highest rate of concordance was seen in small cell carcinoma (83.3%). The overall sensitivity of FNAC in distinguishing malignant lung lesions from benign lesions was 84.2% and the specificity was 100%. The sensitivity of cytology in diagnosing small cell carcinoma was 83.3% and of non-small cell carcinoma was 65.38%. Immunocytochemistry was carried out in 34 cases, all of which were followed by immunohistochemistry. Cyto-histopathological concordance was noted in 31 of these cases (91.2%). We used the standard panel of four markers (cytokeratin-7, thyroid transcription factor-1, p63 and CD56) for all selected cases.
CONCLUSION: Cytology along with immunocytochemistry is highly effective in diagnosing and categorizing lung neoplasms, producing comparable results to trucut biopsy and immunohistochemistry.

Introduction

Lung cancer is the leading cause of cancer-related death all over the world, with an average 5-year survival rate of 16.8% despite advancement in chemoradiation.[1] The diagnostic evaluation of patients with suspected lung cancer includes tissue diagnosis and a complete staging work-up, including evaluation of metastases. The modality selected to diagnose a suspected lung cancer is based on the size and location of the primary tumor in the lung, the presence of potential metastatic spread and the anticipated treatment plan (Epidermal Growth Factor Receptor/EGFR and Anaplastic Lymphoma Kinase/Alk mutation study).

Percutaneous lung biopsy is now a common procedure in pulmonary medicine, and several different techniques are in use. The most common has been the use of a fine needle, yielding an aspirate for cytological evaluation combined with the trucut needle, where material may be obtained for histology. The indications for these techniques overlap with the more recently developed transbronchial needle aspiration (TBNA).

Haaga and Alfidi[2] reported computed tomography (CT)-guided biopsy in the 1970's, but Menetrier[3] probably performed the first cutting needle biopsy for histological diagnosis.

The diagnostic accuracy has been reported as being >80% for benign disease and >90% for malignant disease.[4]

Cytology with immunocytochemistry is, by its virtue, faster and cost-effective and yields almost as good a result when compared with more invasive trucut biopsy and immunohistochemistry. Therefore, the question arises — Is trucut biopsy loosing importance in diagnostic work-up of lung masses?

For these reasons, we have conducted a prospective study on the efficacy of fine needle aspiration cytology (FNAC) and immunocytochemistry in comparison with trucut biopsy and immunohistochemistry in patients with localized intrathoracic lesions suspicious for malignancy.

Materials and Methods

This institution-based prospective, observational study was conducted over a period of 1 year (January 2013 to January 2014). Eighty patients who were admitted with a radiologically visible localized lesion in the lung (either in chest radiograph or CT), were included in this study.

All cases were thoroughly evaluated radiologically to exclude primary malignancy in any site other than lung.

Informed written consent was obtained from each patient and the lesion was localized on CT scan. Images were obtained of the region of interest by using a section thickness of 3 mm. The slice of interest (entry and depth from the skin surface) was marked with an axial laser beam localizer. FNAC was performed with a 22 G spinal needle for confirmation of the needle track as well as to obtain a cytologic sample. At least two passes were made in each case, with care taken to biopsy the periphery as well as the center of the lesion. Smears were made and were fixed in ethanol for Papanicolaou and dry-fixed to stain in May-Grünwald-Giemsa (MGG). Then, local skin disinfection was carried out and 2% lignocaine was infiltrated along the track. A 2-mm incision was made with a scalpel of size 11. A trucut biopsy needle (BARD, 18” needle with 23 mm throw) was inserted through the cut up to the pre-determined distance and the biopsy material was obtained. Imprint smears were made from the tissue core, which was then transferred to formalin for histopathology. Two to three passes were made in all cases. Local dressing was performed. In each case, the patient was monitored for any complication by erect chest radiograph/repeat CT after 1 h. Immunohistochemistry and immunocytochemistry were carried out as and when required. For immunocytochemistry, we used cold acetone-fixed smear and for immunohistochemistry, a peroxidase-based study was performed in an automated platform.

Results

A total of 80 cases in the age range of 22-82 years were evaluated. The mean age was 57.6 years (57.6 ± 15.3 years) and the M:F ratio was 4.2:1. 93.2% of the cases were above 50 years of age. The earliest age of onset was a 24-year-old male and a 34-year-old female, both with a diagnosis of adenocarcinoma. A total of 52 (65%) patients had a past or present addiction of tobacco. Majority of the cases were in the right lung (right:left = 2.1:1). Majority were peripheral lung lesions (87.8%).

Adenocarcinoma predominated among the cases (49%), followed by squamous cell carcinoma (SCC, 16%) and small cell carcinoma (8%). No malignancy was found in 13% of the patients while the rest comprised of metastatic deposit, lymphoma, plasmacytoma, synovial sarcoma, primitive neuro-ectodermal tumor (PNET), adenoid cystic and langerhan cell histiocytosis (LCH).

Cyto-histopathological concordance was seen in 60 cases (75%) [Table 1]. The highest rate of concordance was in small cell carcinoma (83.3%). 46.6% of SCCs were rightly diagnosed in cytology, whereas 65.78% of adenocarcinoma (including poorly differentiated types) were positively diagnosed in cytology. Plasmacytoma, non Hodgkin lymphoma (NHL), synovial sarcoma and benign hyperplasia were all concordant in cytology.

Table 1

Diagnostic break up according to histology (n = 80)

Three cases of poorly differntiated squamous cell carcinoma (PD SCC), eight cases of PD adeno and one small cell carcinoma yielded either inadequate or only hemorrhagic/inflammatory cells in cytology. Seven cases of adenocarcinoma were wrongly interpreted as SCC in cytology as the Papanicolaou-stained smear showed orangeophilia.

Histology was taken as the gold standard when we only consider benign and malignancy to be accurately diagnosed by FNAC [Table 2]:

Table 2

Cyto-histo correlation in our study

When we consider small cell versus non-small cell cancer:

Sensitivity of cytology in diagnosing small cell: 83.3% and non-small cell: 65.38%

Immunocytochemistry was performed in 34 cases. Among them, adenocarcinoma was predominant (n = 23), followed by SCC (n = 4), small cell carcinoma (n = 4) and one case each of spindle cell sarcoma, metastatic adenocarcinoma and non-Hodgkins lymphoma. All of these cases were followed by trucut biopsy and immunohistochemistry. Immunocyto — immunohisto concordance was noted in 31 cases among these (91.2%) [Table 3].

Table 3

Immunocytochemistry — immunohistochemistry correlation (n = 34)

One case of small cell carcinoma cytology yielded predominantly necrotic material, but CD56 staining by immunocytochemistry picked up few cells with intense membrane positivity. Trucut biopsy and immunohistochemistry confirmed the case to be small cell carcinoma (TTF-1 +ve, p63 −ve, synaptophysin +ve, CD56 +ve) [Figure 1].

Figure 1

(a) Cytology smear showing clusters of pleomorphic cells with molding and smudging, small cell carcinoma (MGG, ×400). (b) Immunocytochemistry smear showing CD56-positive cells in a necrotic background (IHC, ×400). (c) Trucut biopsy showing solid sheets of cells with smudging (H and E, ×400). (d) Immunohistochemistry showing positivity with chromogranin-A (IHC, ×400)

Similarly, in one case of SCC, immunocytochemistry by p63 was helpful to delineate degenerated squamous cells among granulomatous inflammation in cytology smears.

In one case of poorly differentiated adenocarcinoma, the mass was deep seated and close to the mediastinum and we missed the mass in trucut biopsy (yielded only granulation tissue and few scattered atypical cells in the trucut sample). Cytology and immunocytochemistry helped to reach to a conclusive diagnosis in this case (TTF-1 +ve, CK-7 +ve, p63 -ve).

Two cases were diagnosed as SCC in cytology and immunocytochemistry (p63 +ve, TTF-1 −ve, CK-7 +ve) and trucut biopsy revealed adenosquamous carcinoma (CK-7 +ve, Napsin-A +ve, TTF-1 +ve, patchy p63 +ve) [Figure 2].

Figure 2

(a) Cytology smear demonstrating prominent orangeophilia, diagnosed as squamous cell carcinoma on cytology (Pap, ×400). (b) Trucut biopsy shows squamous areas among cells with glandular differentiation: Adenosquamous carcinoma (H and E, ×400). (c) Immunohistochemistry shows strong CK-7 positivity (IHC, ×400). (d) Patchy p63 positivity in squamoid areas (IHC, ×400)

One case of adenocarcinoma diagnosed by cytology proved to be metastatic in origin, where trucut biopsy showed a villous papillary pattern and mucin secretion (CK-7 +ve, p63 −ve, CDX2 +ve, TTF-1 −ve). Clinical follow-up and endoscopy revealed later mass in the pylorus.

The overall findings suggest that when cytology is combined with immunocytochemistry, it has almost the same precision compared with more invasive trucut biopsy [Table 3]. However, newer prognostic tests like EGFR and ALK mutation studies require tissue samples and cytology samples have not been yet approved for these tests.

Lastly, we evaluated the value of different immunohistochemistry markers in appropriately diagnosing lung neoplasms. Adenocarcinomas expressed napsin-A in 78% and TTF-1 in 62%, whereas small cell carcinoma expressed it in 88%. CD56 positivity and TTF-1 positivity were seen in 50% of the cases. SCC expressed p63 positivity in 100% but TTF-1 positivity was only seen in 25% of the cases. Small cell carcinoma expressed any two of three neuroendocrine markers (CD56, synaptophysin, chromogranin-A) in 100% of the cases. We conclude that, at present, a panel of four markers (CK-7, p63, TTF-1, CD56) should be enough to differentiate or categorize lung neoplasms. Newer markers like napsin-A are more useful in categorizing adenocarcinoma and can replace CK-7 in the near future.

Discussion

The American College of Chest Physicians supports the use of transthoracic needle aspiration biopsy (TNAB) as the procedure of choice in patients in whom the benign nature of the solitary pulmonary nodule cannot be established by clinical criteria and in whom surgery (exploratory thoracotomy or video-assisted thoracoscopic surgery) cannot be undertaken.[5]

Sensitivity of TNAB for the diagnosis of malignancy was reported to range from 64% to 100%. Similarly, sensitivity of TNAB for specific benign diagnosis was reported to range from 11.7% to 68%.[67]

We found the mean age of the population of this study to be 57.6 years, with a M:F ratio of 4.2:1 and smoker: Non-smoker ratio of 65%. This is in accordance with Noronha et al.[8]

Non-small cell carcinomas predominated, similar to the experience of other workers.[910] We found adenocarcinoma to be the most common malignancy in our study population.

The overall accuracy of diagnosing small cell versus non-small cell lung cancer at cytology in this study was 83.3%. The high accuracy of distinction between small cell and non-small cell lung cancer in our study matched with other studies.[1112]

We found that cytological diagnosis of non-small cell lung cancer is less reliable (misclassification in 12% of the cases) than cytological diagnosis of small cell lung cancer. Mukherjee et al.[11] found the contrary.[11]

Majority of poorly differntiated SCC were misdiagnosed as non-small cell adeno because it failed to demonstrate keratinization in imprint/FNA smears. Three cases were reported as inflammatory/hemorrhagic as the mass was almost entirely necrotic. One case was misdiagnosed as small cell because cytology showed smudging artefacts.

Adenocarcinomas that were misdiagnosed as SCC in cytology were mostly poorly differentiated in nature; three of them showed areas of focal squamous differentiation in histology. One case of bronchioloalveolar carcinoma (minimal deviation adenocarcinoma, lepidic type) was diagnosed as well-differentiated adenocarcinoma, one mucinous adenocarcinoma was diagnosed as SCC as cells showed orangeophilia and three adenocarcinoma yielded inadequate material.

One small cell carcinoma was misdiagnosed as inflammatory in cytology probably due to the needle aspirating necrotic part of the tumor.

The importance of review of the slides has been cited by Tan et al.[12] We found that the review was especially important in the cases diagnosed as negative for malignancy. We also found that the review of slides by a second expert followed by a consensus diagnosis by both increased the accuracy.

Bocking et al.[13] compared fine needle aspiration biopsy (FNAB) and punch biopsy (PB), and found that the overall sensitivities of the biopsy methods were equal. They concluded that FNAB should be the method of choice in pulmonary and hilar lesions because of the similar diagnostic accuracy. Mediastinal and pleural lesions and presumed mesenchymal tumors should be sampled with PB because the typing accuracy of FNAB is insufficient in these cases. Mondal et al.[14] found FNAC to be superior in categorizing lung neoplasms.

Manhire et al.[15] suggested that percutaneous transthoracic lung biopsy (PTLB) should be considered in the following cases:

New or enlarging solitary nodule or mass on the chest radiograph that is not amenable to diagnosis by bronchoscopy or CT shows it is unlikely to be accessible by bronchoscopy.

Multiple nodules in a patient not known to have malignancy or who has had a prolonged remission or more than one primary malignancy.

Persistent focal infiltrates, either single or multiple, for which no diagnosis has been made by sputum or blood culture, serology or bronchoscopy.

Hilar mass.

There are relative contraindications to PTLB, and the balance of benefit against risk for the procedure should be assessed at a multidisciplinary meeting.

Pre-operative investigations should include coagulation indices like prothrombin time (PT), activated partial thromboplastin time (APTT) and platelet count. Oral anticoagulants should be stopped before a percutaneous lung biopsy.

They found that an erect chest radiograph should be performed 1 h after the biopsy, and is sufficient to detect the majority of post-biopsy pneumothoraces.[15]

Khan et al.[16] showed that endobronchial ultrasonography-guided transbronchial needle aspiration biopsy (EBUS-TNA) has a very high sensitivity (97%) in the diagnosis of central lung masses related to lung cancer. It could be used as a “one stop” as an early minimally invasive tool in the lung cancer diagnostic pathway to enable accurate diagnosis, and a negative result may warrant other invasive tests.

CT-guided trucut biopsy is generally regarded as a safe procedure with limited morbidity and extremely rare mortality. The pneumothorax remains the most frequent complication of TNAB. Tsukada et al.[4] have shown that the lesion size is a significant factor contributing to diagnostic accuracy and that a lesion size >30 mm in diameter has an accuracy as high as 93%. Yeow et al.[17] in a study of 660 consecutive biopsies showed that patients with lesions ≤20 mm have a higher incidence of pneumothorax than those with larger lesions. Wang et al.[18] found a statistically significant correlation between pneumothorax and the factors of smoking (P = 0.015) and position (P < 0.01) and length of the needle in the normal parenchyma (P = 0.011) as well as between hemorrhage and the maximal diameter (P = 0.005) and length of the needle in the normal parenchyma (P < 0.01) and the frequency of needle adjustments (P < 0.01). We encountered only two cases of pneumothorax that resolved by occlusive dressing only.

Bozetti et al.[19] studied the efficacy of cytology and immunocytochemistry to categorize lung neoplasms according to prognostic markers like Ki-67, bcl-2 and p53 and found a concordance rate with histology as high as 95%.

Kulshreshtha et al.[20] used FNA material supplemented by cell block preparation and immunohistochemistry and effectively subcategorized lung neoplasms using IHC markers like CK, chromogranin-A, synaptophysin, TTF-1 and CEA. In our study, we used a panel of TTF-1, CK-7, CD56 and p63 in most of the cases. Several other authors used a combination of markers in cytology specimens and showed efficacy comparable to trucut.[21222324] Newer markers like napsin-A are more specific for lung adenocarcinomas and likely to prove more effective in the near future.[25]

Conclusion

In this study, we evaluated the demographics of lung cancer in our institution and compared cytology, immunocytochemistry, trucut and immunohistochemical parameters of lung cancer. We found adenocarcinoma to be the predominant histological type. We also found that although cytology is highly accurate in distinguishing true negative cases (benign lesions), a needle biopsy is often required to complement the cytology when the material is inadequate or the typing of malignant lesion is critical for treatment. Immunocyto chemistry and histochemistry is an invaluable tool to categorize when morphology is equivocal. We found immunocytochemistry to be highly effective in categorizing lung neoplasms, producing comparable results to immunohistochemistry.