Evaluation of Imprint Smears of Bronchoscopic Biopsy in Lung Tumors: A Cytohistological Correlation.

BACKGROUND: There are several methods for obtaining samples in patients of lung tumors, of which bronchoscopic biopsy is the most common. In most of Indian scenario, however, histopathology diagnosis is time taking. AIMS AND
OBJECTIVES: To evaluate imprint cytology as a tool for rapid diagnosis of lung carcinoma and its histopathological correlation. STUDY TYPE: Prospective study on accuracy of a diagnostic test.
MATERIALS AND METHODS: A total of 175 cases were included in the study, and all of them were subjected to brochoscopic biopsy. Imprint smears were prepared from all the bronchoscopy specimens obtained from 175 cases. Imprint smears were stained with Leishman-Giemsa cocktail and Pap stain, and histopathology sections were stained with hematoxlin and eosin. Histopathological findings were confirmed by immunohistochemistry.
RESULTS: Sensitivity and specificity of imprint cytology was 84.9% and 72.4%, respectively when compared to histopathology as standard.
CONCLUSION: Imprint cytology can be used as a preliminary tool for diagnosis in lung tumor bronchoscopic biopsies.

INTRODUCTION

Lung cancer is currently the most frequently diagnosed cancer in the world and the most common cause of cancer-related mortality. The major risk factor for developing lung cancer is tobacco use. Over the coming decades, changes in smoking habits will greatly influence lung cancer incidence and mortality as well as the prevalence of various histologic types of lung cancer.[1]

Several methods are there for getting histopathological specimens. Among various bronchoscopic techniques, bronchial biopsy has the highest sensitivity for endobronchial malignant lesions.[2] However, as with any histopathology, it takes time to arrive at a diagnosis. Several researchers have emphasized about the role of imprint cytology in combination with histopathology to improve diagnostic accuracy and lessen reporting time to the patients. It can be used to give rapid, preliminary diagnosis of lung cancer.[3456789] Bronchial biopsy has also been used as the gold standard diagnostic test to assess the efficacy of other cytologic techniques.[10] Some studies in this topic are present worldwide, but there are only few studies in India, and to the best of our knowledge, none from our state or eastern India. Keeping these in mind, the current study was undertaken to assess the sensitivity and specificity of imprint smears of bronchial biopsy in diagnosing lung masses, histopathological diagnosis being taken as gold standard for confirmation.

MATERIALS AND METHODS

This study was a prospective one, conducted in the Department of Pathology and department of Pulmonology of a teaching institute, after prior approval from institutional ethics committee. The specimens for cytological and histological examination were collected from the indoor and outdoor patients of the Pulmonology Department in whom a provisional diagnosis of lung carcinoma was made according to clinical and radiological findings. A total of such 175 patients were included in the study. Patients who were not willing to undergo the procedure or were not medically fit for the procedure were excluded. The samples were obtained by flexible fiberoptic bronchoscopy done by the pulmonologist. Imprint smears were prepared from the bronchial biopsy in all the 175 cases. Imprint smears were prepared by placing forceps biopsy specimens on a glass slide by gentle touching and rolling over the surface. Care was taken to avoid crushing the specimen. Few of the smears were air dried and stained with Leishman-Giemsa (LG) cocktail stain. Others were fixed in absolute alcohol and stained with Pap stain. Bronchial biopsies were fixed in 10% neutral buffered formalin, and formalin fixed paraffin embedded sections were prepared, which were subsequently stained with hematoxylin and eosin (H and E). Cytology and histopathology slides were viewed independently by different pathologists. Immunohistochemical confirmation was done. Data were analyzed using Microsoft Excel (Microsoft Redmond, Washington, USA) and GraphPad Prism 5(GraphPad Prism softwares CA, USA).

RESULTS

The sensitivity and specificity of imprint smear were 84.9% and 72.4%, respectively. There were 22 false negative and eight false positive malignancy cases by imprint cytology [Table 1]. Seven cases of imprint cytology could not be categorized into specific cytologic subtype, however, malignancy was confirmed. These were categorized as “positive for malignancy.” Out of all 175 cases subjected to bronchoscopy, lung cancer was confirmed in 146 (83.4%) cases by histopathology of bronchial biopsy [Table 2]. Squamous cell carcinoma was found to be the most common lung cancer (40.4%) [Figure 1a], followed by small cell carcinoma (31.5%) [Figure 1b] and adenocarcinoma (24.6%) [Figure 2a]. One case was diagnosed as small round cell tumor, which also correctly diagnosed by imprint cytology [Figure 2b]. Among patients diagnosed with lung cancer, 109 males (96.5%) and 12 females (36.4%) were active smokers or had a long history of smoking. The patients’ age group ranged from 45 to 78 years, and the mean age was 62 years. Out of 175, 116 patients were from a rural background (66.3%).

Table 1

Results of imprint cytology compared to gold standard

		Malignant	Non-malignant
Imprint cytology	Positive for malignancy	124	08
	Negative for malignancy	22	21

Table 2

Lung cancer distribution on final histopathological typing

Morphologic type	Number of cases
Squamous cell carcinoma	59
Small cell carcinoma	46
Adenocarcinoma	36
Poorly differentiated carcinoma	03
Small round cell tumor	01
Large cell neuroendocrine carcinoma	01
Total	146

Figure 1

(a) Squamous cell carcinoma histopathology (Haematoxylin and eosin ×100). (b) Small cell carcinoma histopathology (Haematoxylin and eosin ×400)

Figure 2

(a): Adenocarcinoma, imprint showing 3D clusters (Leishman Giemsa stain ×400). (b) Small round cell tumor imprint (Leishman Giemsa stain ×400)

DISCUSSION

In the current study, we aimed at assessing the sensitivity and specificity of imprint smear cytology of bronchoscopic biopsy specimens in suspected cases of lung carcinoma. The imprint cytology results were compared to histopathological diagnosis, which was used as gold standard. Our study yielded a sensitivity of 84.9% and specificity of 72.4% for imprint smears. Out of the 120 cases detected correctly by imprint cytology as malignant cases, subtyping was done for 113 cases (94.1%), and seven cases were reported as “positive for malignancy.” The reason was observed to be scanty material and presence of obscuring blood and cell clumps.

Wolfgang et al. in 1991 studied the role of different techniques of diagnosis in bronchoscopic biopsy specimens.[5] The imprint of the forceps biopsy specimen yielded the highest number of positive results, and also there was agreement in the final morphologic tumor type in 136 of 158 cases (86.1%) by positive imprint cytology. Bodh et al. also found tumor typing good in imprint cytology.[11] They found that sensitivity and specificity of imprint smear were 81.35% and 78.12%, respectively and were higher than that of brushing. Tumor typing in their study was done in 69% cases. In our study, subcategorization into morphologic types was possible in 113 cases (94.1%). Our study results were in keeping with these studies, and correct morphological tumor typing was better in our study. This may be because of our use of Leishman-Giemsa cocktail stain and Pap stain for cytology, whereas Bodh et al. used Giemsa only.[12] According to the latest guidelines, tumor subcategory has been recommended in cytology also.[13] In studies by Jan et al., Paulose et al., Kawaraya et al., and Goyal et al., imprint cytology showed a high sensitivity, comparable to our study.[3489]

There were 8 false positive cases and 22 false negative cases in our study. False positive cases were most likely because of the abrasive artifacts produced during the imprinting process, which resulted in distorted cytomorphology. Similar problem was encountered by Wolfgang et al. as found in their study.[5] False negative cases might be due to the scanty tumor cells at the surface of specimen. Peritumoral inflammation and hemorrhage might have yielded less tumor cell on glass slides during imprinting. Some tumors such as mucinous adenocarcinoma and lepidic predominant adenocarcinoma have less atypical tumor cells and hence could have been missed oncytology.[1314]

The present study included only bronchoscopy as a modality to obtain biopsy samples. Previous studies also accessed other methods such as transthoracic needle aspiration, core biopsy for obtaining imprints. Peripherally located lesions could not have possibly been sampled as they usually do not produce endobronchial growths.[14]

CONCLUSION

From the present study, it may be concluded that there is a definite role of imprint cytology in bronchoscopic biopsies in preliminary diagnosis of lung cancer, and it can be used over other cytologic methods with a better degree of accuracy. Tumor typing according to morphology is also an advantage in imprint cytology. Hence, use of imprint cytology in bronchoscopic biopsies of lung tumors is recommended for early and preliminary diagnosis.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.