Prospective study of combined use of bronchial aspirates and biopsy specimens in diagnosis and typing of centrally located lung tumoursBMJ 1995; 310 doi: http://dx.doi.org/10.1136/bmj.310.6980.624 (Published 11 March 1995) Cite this as: BMJ 1995;310:624
- Eric Piaton,
- Marie-Helene Grillet-Ravigneaux,
- Bernard Saugier,
- Helene Pellet
- Clinical Cytopathology Laboratory, Lyon Grange-Blanche Medicine Faculty, Domaine Rockefeller, 69373 Lyons, France E Piaton, pathologist M-H Grillet-Ravigneaux, statistician H Pellet, professor of histology. Internal Medicine Unit, General Hospital, BP 364, 39016 Lons-Le-Saunier, France B Saugier, oncologist.
- Correspondence to: Dr E Piaton, Laboratoire d'Histologie, Faculte de Medecine, Lyon Grange-Blanche, 8 avenue Rockefeller, 69373 Lyons, Cedex 08, France.
- Accepted 30 December 1994
Objective: To determine the diagnostic accuracy of examining bronchial secretions in pulmonary cytopathology and whether cytology and histopathology can complement each other in routine practice among lung specialists.
Design: A prospective study comparing 1225 cytological and biopsy results, conducted during 1987-93. Tumours were confirmed by histopathology, imaging techniques, or clinical outcome and imaging techniques combined.
Setting: 11 lung or internal medicine units, France.
Subjects: 1128 patients (874 men; 254 women) aged 65.3 (SD 13.7) years who underwent fibreoptic bronchoscopy for various pulmonary symptoms.
Results: Exact concordance between cytological and biopsy results was obtained in 1036/1187 (87.3%) satisfactory specimens. In all 574 lung tumours were diagnosed. One case (0.08%) was a false positive cytological diagnosis in a patient with tuberculosis. Patients with lung cancer were more likely to have positive cytological results than positive biopsy results (P<0.001). Agreement in tumour typing was observed in 375/424 (88.4%) cases, when non-small cell carcinomas, small cell carcinomas and undifferentiated carcinomas were separated. In the 11 patients with squamous cell carcinomas in situ, eight (72.7%) of the carcinomas were diagnosed cytologically as squamous cell. Unsatisfactory material was obtained in only 20 (1.6%) and 19 (1.6%) cases by cytology and biopsy respectively. Examinations had to be repeated in 86 (7.6%) patients.
Conclusions: Examination of bronchial secretions complements histopathology in both diagnosing and typing lung tumours and could be performed more systematically in patients undergoing fibreoptic bronchoscopy.
In this study an accuracy of 88.4% was achieved in typing lung tumours by examining bronchial secretions
Repeated examinations for confirmation or to resolve apparent discrepancies led to confirmation of previous findings in two thirds of cases and to the recognition of malignancy in a third of cases
Simple aspiration of bronchial secretions is of value in diagnosing and typing lung tumours
Lung tumours are the most common cause of death from cancer in men and according to recent studies are now emerging as one of the most common causes of death from cancer in women.1 Usually for appropriate management of lung tumours a tissue diagnosis and typing must be obtained and a clear distinction between small cell carcinomas and non-small cell carcinomas must be made preoperatively or before any aggressive treatment. Histopathological investigation remains the mainstay of the diagnosis. Recent reports based on the recommendations of the pathology committee of the International Association for the Study of Lung Cancer show, however, that an approach that combines cytological examination and biopsy helps the search for mixed small cell and large cell carcinomas and may have important prognostic implications.2 Moreover, bronchial biopsies cannot be performed in inaccessible sites or in patients at risk of haemorrhage, and alternative methods for obtaining tissue diagnosis (cytopathology) are sometimes needed.
Of the specimens that can be obtained by bronchoscopy, washings, brushing, and fine needle aspirations may complement biopsies in the diagnosis of lung cancer.3 4 Traditionally, the bronchial secretions aspirated during fibreoptic bronchoscopy are considered to have little diagnostic value—as they are thought to contain mainly mucoid debris—and are discarded.4 Thus malignancy is often assessed by biopsies alone, whereas true carcinomatous cells can be readily recognised and typed with bronchial secretions.3 5 The aim of our study was to determine whether the simple aspiration of bronchial secretions can complement biopsies in the diagnosis and typing of lung tumours.
Patients and methods
During 1987-93 we obtained 1225 combined bronchial aspirates and biopsy specimens during fibreoptic bronchoscopy for various pulmonary symptoms in 1128 patients (874 men, 254 women; mean age 65.26 (SD 123.67) years). Patients were seen at 11 independent lung or internal medicine units. A questionnaire documenting clinical data was completed by the lung specialist at the first examination and at each control stage.
Bronchial secretions were usually obtained from the bronchial tree during bronchoscopy. In 98 cases additional material was obtained by instilling isotonic saline into a selected zone and reaspirating. All samples were preserved in 50% ethyl alcohol and posted to the cytopathology laboratory with clinical, radiological, and bronchoscopic information. Specimens were centrifuged (1500 rpm, 5 min). Three to five slides were prepared from the cell concentrate and stained according to a hypochromic Papanicolaou procedure.6 All slides were evaluated by the same pathologist. Malignant cells were typed at high power field and recorded as squamous cell carcinoma-type cells, adenocarcinomatous cells, poorly differentiated carcinomatous cells, large cell carcinoma-type cells, small cell carcinoma-type cells, or undifferentiated carcinomatous cells.
Usually one to four biopsy specimens were taken from abnormal areas within bronchoscopic range. Tissue specimens were fixed in Bouin's solution, posted to the corresponding histopathology laboratories, and processed routinely for paraffin embedding. Independent pathologists examined the specimens using the World Health Organisation's classification.7
For both cytology and histopathology, only specimens with unequivocal malignant features were considered to be positive. No “suspicious” or “atypical” case was included in the cancer cases. The data were coded and computerised. The X2 test on paired series was used to compare categorical variables, and a probability level of 0.05 was regarded as significant.
Twenty bronchial aspirate samples were unsatisfactory owing to poor cellularity or the presence of too much blood. Histological examination was impossible in 19 cases owing to poorly preserved or non-representative material.
Table I shows a comparison of cytological and biopsy results. If unsatisfactory specimens are ignored, exact concordance was obtained in 87.3% (1036/1187) of cases. Malignancy was diagnosed by cytology in 424 out of 479 (88.5%) positive results from biopsy specimens and in 519 out of 574 (90.4%) tumours proved by both histopathology and subsequent follow up. Patients with lung cancer were more likely to have positive cytological results than positive biopsy results (X2ps=11.13; P<0.001). All the tumours diagnosed by cytology or biopsy, or both, were confirmed by histopathology (repeated biopsy or surgery specimens, when patients could undergo an operation), computed tomography, or the combination of imaging techniques and clinical outcome, except in one case of a false positive cytological result in a patient with tuberculosis.
Table II shows the agreement between cytological and biopsy typing. Agreement occurred in specimens from 377 men and 47 women (mean age 65.21 (SD 11.11) years). When lung tumours were separated into non-small cell carcinomas, small cell carcinomas, and undifferentiated carcinomas, exact concordance in typing was obtained in 375 of the 424 (88.44%) cases. Eight of 11 (72.7%) carcinomas in situ were diagnosed by cytology as squamous cell carcinomas.
In the group of histologically proved small cell carcinomas exact typing was obtained by cytology in 51 out of 68 (75%) cases, and undifferentiated features were shown in 13 of the 68 (19.1%) cases. In six cases a small cell carcinoma component was shown by one method and a non-small cell carcinoma by the other; these cases were considered to be combined carcinomas. Undifferentiated components were found by biopsy in 10 out of 328 (3.0%) cytologically proved non-small cell carcinomas and by cytology in 10 out of 310 (3.2%) histologically proved non-small cell carcinomas. These cases were considered to be lung carcinomas showing anaplastic features.
DIAGNOSIS BY CYTOLOGY ALONE
In 95 cases lung cancer was diagnosed by cytology alone (81 men, 14 women; mean age 67.82 (10.61) years). Table III shows the comparison between the cytological and biopsy results. Biopsy specimens showed non-specific findings in 70 of the 82 (85.4%) cases. No significant differences occurred in the number of biopsies taken or in the bronchoscopic presentation of these cases. Except in one case (0.08%) of a false positive cytological result, malignancy was confirmed by repeated radiography and computed tomography.
DIAGNOSIS BY BIOPSY ALONE
In 55 cases lung cancer was diagnosed by biopsy alone (40 men, 15 women; mean age 64.30 (13.46) years). Table IV shows the comparison between the cytological and biopsy results. Non-specific cytological disorders such as inflammatory cells and metaplastic and hyperplastic changes of bronchial epithelial cells were found in 33 out of 55 (60.0%) cases, irrespective of the histopathological type of the tumours. Conversely, modifications of the squamous epithelium indicating potentially irreversible changes (keratinising metaplasia, moderate or severe nuclear atypias, cytological findings suggestive of squamous cell carcinoma) were found in nine out of 21 (42.9%) cases with squamous cell carcinoma on biopsy examination, and in two out of three (66.6%) cases with squamous cell carcinomas in situ.
Examinations had to be repeated in 86 of the 1128 patients either to confirm the initial results or to resolve apparent discrepancies between results and clinical data. Examinations had to be repeated once in 76 patients, twice in nine patients, and three times in one patient. Table V shows the resulting 97 cytohistological comparisons.
Fifty three (54.6%) of the new paired results simply confirmed the initial data (42 pairs were concordant negative results, five were concordant positive results, four were negative cytology results with positive biopsy results and two were consecutively negative biopsy results in a patient with a positive cytological result). Of 57 concordant negative results initially recorded, 15 subsequent results (1.2% of all examinations) obtained in 10 patients were then found to be concordant positive. Concordant positive results were also obtained in four cases (three patients) after initial results that showed only positive cytology.
Overall, repeating the examinations confirmed the absence of abnormal cells in 42 cases, confirmed malignancy in 17, allowed documenting of malignancy in 29, and gave no additional information in nine.
DIAGNOSTIC VALUE OF PULMONARY CYTOPATHOLOGY
Before the introduction of the flexible bronchoscope the diagnostic value of examinating bronchial secretions in detecting lung tumours was often lower than 50%: in 1964 Koss reviewed 560 patients with primary bronchial carcinomas in which examination of 288 bronchial aspirates (51.4%) yielded positive results.8 In 1975 Atay and Brandt reported that in a study of 885 patients 79% of centrally located lung tumours could be diagnosed cytologically, whereas only 45% of peripheral tumours could be detected.9 In 1979 Payne et al reported that the diagnostic accuracy of examining bronchial aspirates was 46% (50 of 109 primary bronchial carcinomas).10
Since fibreoptic bronchoscopy was introduced as a diagnostic procedure in 1968, several methods for obtaining satisfactory material have become available. Though examination of sputum can provide evidence of malignancy in many cases of cancer,2 3 4 11 12 specimens collected from fibreoptic bronchoscopy yield a higher rate of positive results. In 1980 Taft et al reported that the sensitivity of bronchial aspirates in diagnosing lung cancers was 75% and could reach 88.7% when the examination was performed by two pathologists.13 In 1991 DiBonito et al reported a sensitivity of 81% with no false positive results.5 The present series shows a sensitivity value of 88.3% if initial biopsy results are considered as the diagnostic reference, and 90.4% if both histopathology and follow up (the gold standard) are taken into account. Data about the cytopathology of bronchial secretions remain rare, and many authors have had difficulties in obtaining representative material from the bronchial tree, thus leading them to discontinue examination of bronchial secretions in favour of other cytological methods.4 14
Bronchial brushings are favoured by most groups for the cytological investigation of proximal lung cancers, and samples that arrive at the laboratory as prepared slides may seem much easier to handle. Standardisation is very difficult to obtain, however, especially in terms of quality of fixation. Many brushings are not satisfactory when they arrive at the laboratory from regional hospitals.
We have shown that, provided that conditions are adequate, the diagnostic accuracy of bronchial aspirates may exceed 80%. Moreover, cases that are difficult to manage—that is, contraindications for biopsy, positive fibreoptic bronchoscopy findings but positive cytological and negative biopsy results, mucosal abnormalities but negative cytological and biopsy results—are clearly resolved by repeating both examinations. Repeating examinations also led to the absence of false negative results for both diagnostic techniques together.
TYPING OF LUNG TUMOURS
According to the WHO's classification7 malignant epithelial tumours are separated into squamous cell carcinomas (30-60% of histopathology cases), small cell carcinomas (about 25%), adenocarcinomas (15-25%), and large cell carcinomas (10-20%). From a therapeutic point of view, lung tumours are better separated into small cell carcinomas, which require intensive chemotherapy and sometimes radiation therapy, and non-small cell carcinomas, which are better treated surgically.15
As far back as 1952 Foot had shown that typing could be correctly obtained by studying sputum in 89.9%, 78.4%, and 83.5% of squamous cell carcinomas, adenocarcinomas, and small cell carcinomas respectively.16 In 1972 Lange and Hoeg showed comparable results.17 In 1985 Truong et al showed that typing of squamous cell and small cell carcinomas could be correctly obtained in more than 80% of cases regardless of the cytological method used (sputum, washings, or brushings).18 In a recent series of 100 cases with confirmation at necropsy, the diagnostic yield of examining bronchial aspirates reached 80%.5 The present series shows that exact concordance can be achieved in 88.4% of cases.
A correct, if not definitive, diagnosis can be obtained in almost all cases of centrally located lung tumours by examining bronchial secretions. Moreover, when a definitive histopathological diagnosis cannot be obtained because of risk of haemorrhagia, severe respiratory failure, or inconclusive sampling diagnostic procedures such as cytological examinations of bronchial aspirates become necessary.