Transbronchial Needle Aspiration Biopsy in the Diagnosis of Pulmonary Neoplasms Nalini Bhat, M.D., Pradeep Bhagat, M.D., Eugene S. Pearlman, M.D., Gregory Kane, M.D., William Figueroa, M.D.,Vaidehi Kannan, M.D., and Tilde S. Kline, M.D.
To evaluate the usefulness of transbronchial needle aspiration biopsy (TBNAB) in the primary diagnosis of lung cancer, we undertook a prospective study of this technique, assessing specimens from 124 patients over a 3-yr period. Cytologic andlor histologic material obtained by an alternate technique was available f o r 105 (84.7%) of the patients. The diagnostic sensitivity of TBNAB was comparable with that of other bronchoscopic techniques; with the addition of TBNAB, the overall sensitivity of bronchoscopy was increased from !i4 % to 72% {P< 0,001). Because of its usefulness f o r diagnosis of submucosal tumors and its safety, we conclude that TBNAB is a useful addition to current methods for the diagnosis of suspicious pulmonary lesions. Diagn Cytopathol 1990;6: 14-11. Key Words: Lung; Cytology;Carcinoma;Sensitivity; Specificity;
Diagnostic efficiency With the mounting incidence of lung cancer, the elficient evaluation of suspicious pulmonary lesions has become increasingly important. The analysis of bronchiall wash (TBW), brush (TBBr), and biopsy material (TBBx) obtained from fiberoptic bronchoscopy allows the diagnosis of lung neoplasms to be made with a high degree of accuracy.' Percutaneous fine-needle aspiration biopsy (PNAB) has been used successfully for almo.;ct two decades, particularly for peripheral and metastatic lesions. In recent years, transbronchial needle aspiration biopsy (TBNAB), pioneered by Wang et a1.' for the diagnosis and staging of mediastinal and paratracheal lesions, has been coming into more widespread use. We undertook this study to evaluate the role of TBNAB in comparison to established techniques for primary diagnosis of lung cancer, specifically with respect to sensitivity,
Received December 7, 1988. Accepted June 10, 1989. From the Departments of Pathology and Medicine, Lankenaiu Hospital, Philadelphia, PA. Address reprint requests to T.S. Kline, M.D., Department of' Pathology, Lankenau Hospital, Lancaster Avenue West of City Line, Philadelphia,PA 19151.
14
Diagnostic Cytopathology,Vol6, No I
specificity, diagnostic efficiency, and predictive value. We also sought those special situations where only TBNAB may provide diagnostic information.
Materials and Methods One-hundred and twenty-four patients were examined by a pulmonologist using flexible bronchoscopy and TBNAB at Lankenau Hospital. Over a 3-yr period, 128 TBNAB specimens were obtained. All patients had a suspicious clinical history and/or chest x-ray. The location of the lesion was classified into three categories: central if adjacent to a bronchus, peribronchial if more than 5 cm from a bronchial hilum, and peripheral if pleural-based or immediately subpleural. If at the time of bronchoscopy, mucosal irregularity or granularity was noted, a needle was introduced through the bronchoscope directly into the area. When mucosal alteration could not be seen in a site that was radiographically suspicious, an effort was made to pass a needle into the area through the bronchoscope under fluoroscopic guidance. All aspirations were performed as previously d e ~ c r i b e d .The ~ smears were fixed immediately in 95% ethanol and stained according to the method of Papanicolaou. The TBNAB specimens were reviewed for all major aspiration biopsy cytology criteria of malignancy: cellularity, dyshesion, monomorphism, nuclear membrane irregularity, anisonucleosis, and macronucleoli.3 The material also was evaluated for features important in the analysis of exfoliative cytologic specimens, including alteration of the nuclear/cytoplasmic ratio, hyperchromatism, and background. Interpretation of TBNAB material was done with four categories: benign, suspicious of carcinoma, carcinoma, and unsatisfactory. The specimen was interpreted as benign when it was composed of normal respiratory mucosal cells and inflammatory cells. Specimens were interpreted as suspicious when abnormal cells were sparse 0
1990 WILEY-LISS, INC.
TBNAB AND PULMONARY NEOPLASMS
or poorly preserved or when the aspirate lacked some of the criteria of malignancy. Material was regarded as unsatisfactory if blood or inflammatory cells obscured underlying cells. Sensitivity was calculated as the number of specimens from patients with carcinoma that were correctly diagnosed as positive by TBNAB (true-positives) divided by the number of specimens from patients who had carcinoma and underwent TBNAB. Sensitivity is therefore a measure of how well TBNAB diagnosed malignancy. Specificity was calculated as the number of specimens from patients with benign conditions correctly diagnosed as negative (true-negatives) divided by the total number of specimens from patients with benign conditions who underwent TBNAB. The predictive value of a positive test (i.e., the probability of a patient having a malignant condition in the presence of a positive test result) was calculated using the above parameters and assuming a prevalence rate of 50% for carcinoma in those patients coming to bronchoscopy. The predictive value of a nonpositive test was derived similarly. Diagnostic efficiency was calculated as the ratio of the sum of true-positives and true-negatives to the total number of specimens from patients undergoing TBNAB (excluding specimens deemed unsatisfactory). The statistical significance of the increase in sensitivity resulting from adding TBNAB to other bronchoscopic techniques was evaluated using the McNemar test for related samples4
Table XI. Statistical Parameters for TBNAB Parameter
Percentage
Sensitivity Specificity Diagnostic efficiency* Predictive value positivet Predictive value nonpositivet
56 74 69 100
70
*Calculation drops unsatisfactory specimens. tCalculation assumes the prevalence rate of tumor to be 50%.
suspicious were eventually confirmed as carcinoma. Nineteen specimens interpreted as benign were from patients subsequently diagnosed as having carcinoma. The sensitivity, specificity, diagnostic efficiency, and predictive values for TBNAB are presented in Table 11. For purposes of comparison, the sensitivities and diagnostic efficiencies of other bronchoscopic techniques and PNAB are presented in Tables 111 and IV, respectively. None of the patients who underwent TBNAB had any medical complication as a result of the p r ~ c e d u r e . ~
Discussion
Of the 128 TBNAB specimens obtained between 1985 and 1988, 89 were from 85 patients ultimately diagnosed as having carcinoma; the remaining 39 were from patients who were finally diagnosed as having benign conditions. The diagnosis was made by the clinician on the basis of the clinical and radiographic course of the patient in conjunction with results of cytologic and/or histologic examination. Material from another bronchoscopic technique (bronchial washings, brushings, or biopsy) or from percutaneous needle aspiration, surgery, or autopsy was available in 105 (84.7%) of the 124 patients. The results of TBNAB are presented in Table I. Ten percent of the specimens (13 of 128) were rejected as unsatisfactory. Sixteen of the 17 cases diagnosed as
TBNAB in conjunction with fiberoptic bronchoscopy augments diagnosis of pulmonary neoplasms, particularly those in the submucosal and peribronchial areas. The figures for sensitivity and specificity presented in Table I1 reflect that, of the 89 specimens from patients with carcinoma, 50 were correctly diagnosed. There were also 39 specimens from patients with benign conditions; of these, 29 were interpreted as benign. We made every effort to be conservative and, with respect to the cytologic criteria, rigidly consistent in our diagnosis of carcinoma. This approach is reflected in the relatively high proportion (1 3.3%) of specimens diagnosed as suspicious. Interestingly, all but one of these patients was ultimately diagnosed as having malignant disease. The predictive values in Table I1 indicate that a positive result by TBNAB is essentially conclusive. Predictive value is, of course, dependent on the value chosen for the prevalence of disease. With the results enumerated above, a change in assumed prevalence of carcinoma in the population undergoing bronchoscopy for possible malignancy from 50% to 70% changes the predictive value of a negative test from 70% to 50%. The predictive
Table I. Results of TBNAB
Table 1x1. Sensitivity of Bronchoscopic Techniques and PNAB
Results
Cytologic diagnosis
Carcinoma Suspicious Benign Unsatisfactory Total
Specimens 50 17 48 13 128
Ultimate diagnosis of carcinoma
50 16 19 4 89
Technique
TBNAB TBBr TBW TBBx PNAB
Number of specimens
Percentage
50189 34/66 12/33 3 1/60 11/14
Diagnostic Cytopathology, Vol6, No I
56
51.5 36 50.2 78.6
15
BHAT ET AL. Table IV. PNAB Technique
TBNAB TBBr TBW TBBx PNAB
Diagnostic Efliciency of Bronchoscopic Techniques and Number of specimens
Percentage
791115 61 192 22/33 61/91 13/16
69 66 67 67 81.3
Calculations drop unsatisfactory specimens from consideration.
value of a positive test would remain unchanged in this instance because of the absence of false-positives. If we had been less conservative in our diagnosis of malignancy and classified all of our suspicious results as carcinoma, our sensitivity would have increased to 74% from 56%, but at the cost of introducing one false-positive result. The predictive value of a positive test would have changed from 100% to 97% with a prevalence rate of 50% and to 99% with a prevalence rate of 70%. The predictive value of a negative result becomes 79% with a prevalence rate of 50% and 53% with a prevalence rate of 70%. Hence, a diagnosis of carcinoma is nearly conclusive while any other result requires additional work-up. Our sensitivity (56.2%) is consistent with the values (45% and 52%) obtained in other studies.637Although our diagnostic efficiency (6 1.7%) was lower than previously reported results,7s8our study differs from previous ones in important ways. The present study was prospective and therefore included patients with benign lung conditions, which was not the case in most previous ~ t u d i e s .We ~ . ~did not attempt, moreover, to partition our suspicious-formalignancy results retrospectively into positive or negative categories depending on the ultimate clinical diagnosis,*which would have inflated our diagnostic efficiency. As Tables 111 and IV indicate, the sensitivity and diagnostic efficiency of TBNAB is comparable with those of other bronchoscopic techniques. Fourteen specimens from patients ultimately diagnosed as having carcinoma were positive by bronchial washings, brushings, or biopsy but not by TBNAB. However, 16 specimens from patients with carcinoma were interpreted as positive by TBNAB but were not detected by other bronchoscopic techniques. Of these I6 cases, radiographic information was available in 1 1 patients; the suspicious area was peribronchial in six instances and peripheral in the remaining five. Overall, bronchial washings, brushings, and biopsy had a sensitivity of 54% (48 of 89), which increased to 72% (64 of 89) when TBNAB was also utilized. This was statistically significant using the McNemar test ( P < 0.001). Individual cellular details rather than consideration of pattern provided the best diagnostic criteria for specimens obtained by TBNAB. Hyperchromasia, alteration of the nuclear/cytoplasrnic ratio, irregularity of the nuclear 16
Diagnostic Cytopathology, Vol6,No 1
Fig. 1. Transbronchial needle aspiration biopsy of moderately to poorly differentiated carcinoma. Note the cluster of malignant cells with marked nuclear alterations (Papanicolaou stain, x 500).
membrane, and the presence of macronucleoli played the major role in the diagnosis of carcinoma (Fig. 1). By contrast, cellularity, dyshesion, and monomorphism were not useful for interpretation. Of the 124 patients in this study, 16 (12.9%) underwent PNAB. Fourteen of these 16 patients were ultimately diagnosed as having carcinoma. Tumor cells were identified in aspirates from 11 of these 14 cases without any false-positive results (sensitivity, 78.6%; diagnostic efficiency, 81.3%) and as suspicious of carcinoma in one additional case. Eight of the 11 cases diagnosed as positive and the one case diagnosed as suspicious PNAB had all been diagnosed as benign by TBNAB. Radiographic information was available in five of these nine patients; in three instances, the suspicious lesion was peripheral in location. It is of interest that, in one patient, the diagnosis by PNAB was negative while TBNAB yielded a positive diagnosis confirmed by bronchial biopsy. Chest x-ray in this instance indicated a central lesion that was less than 2 cm in diameter. We believe that our data support the addition of TBNAB to the more commonly used bronchoscopic techniques for the primary diagnosis of suspected pulmonary neoplasms. Patients with central submucosal lesions are prime candidates for this safe procedure. For effective selection of the suitable modality in a case, a team approach among pathologist, radiologist, and ciinician is necessary. In selected patients, all modalities should be performed at the time of bronchoscopy. An unequivocally positive report avoids a second bronchoscopy or thoracotomy.
References 1. Kuale PA, Bode FR, Kini S. Diagnostic accuracy in lung cancer: comparison of techniques used in fiberoptic branchoscopy. Chest 1976; 69:75 2-9. 2. Wang KP, Terry PB, Marsh BR. Bronchoscopic needle aspiration biopsy of paratracheal tumors. Am Rev Respir Dis 1978;118:17-21.
TBNAB A N D PULMONARY NEOPLASMS 3. Kline TS. Handbook of fine needle aspiration biopsy cytology. 2nd ed. New York: Churchill-Livingstone, 1988:34, 11-3.
4, McNemar Q. Note on the sampling of the correlated samples. Psychometrika 1947;12:153-7.
between
5. Herman PG, Hessel SJ. The diagnostic accuracy and complications of closed lung biopsies. Radiology 1977;125:11-4.
6. Shure D, Fedullo FP. Transbronchial needle aspiration of peripheral masses. Am Rev Respir Dis 1983;128:1090-2. 7. Schenk DA, Bower JM. Transbronchial needle aspiration in the diagnosis of bronchogenic carcinoma. Chest 1987;92:83-5. 8. Morsley RJ, Miller RE, Amy RWM, King EG. Bronchial submucosal needle aspiration performed through the fiberoptic bronchoscope. Acta Cytol 1984;28:211-7.
Diagnostic Cytopaihology,Vol6, No 1
17
To evaluate the usefulness of transbronchial needle aspiration biopsy (TBNAB) in the primary diagnosis of lung cancer, we undertook a prospective study of this technique, assessing specimens from 124 patients over a 3-yr period. Cytologic andlor histologic material obtained by an alternate technique was available f o r 105 (84.7%) of the patients. The diagnostic sensitivity of TBNAB was comparable with that of other bronchoscopic techniques; with the addition of TBNAB, the overall sensitivity of bronchoscopy was increased from !i4 % to 72% {P< 0,001). Because of its usefulness f o r diagnosis of submucosal tumors and its safety, we conclude that TBNAB is a useful addition to current methods for the diagnosis of suspicious pulmonary lesions. Diagn Cytopathol 1990;6: 14-11. Key Words: Lung; Cytology;Carcinoma;Sensitivity; Specificity;
Diagnostic efficiency With the mounting incidence of lung cancer, the elficient evaluation of suspicious pulmonary lesions has become increasingly important. The analysis of bronchiall wash (TBW), brush (TBBr), and biopsy material (TBBx) obtained from fiberoptic bronchoscopy allows the diagnosis of lung neoplasms to be made with a high degree of accuracy.' Percutaneous fine-needle aspiration biopsy (PNAB) has been used successfully for almo.;ct two decades, particularly for peripheral and metastatic lesions. In recent years, transbronchial needle aspiration biopsy (TBNAB), pioneered by Wang et a1.' for the diagnosis and staging of mediastinal and paratracheal lesions, has been coming into more widespread use. We undertook this study to evaluate the role of TBNAB in comparison to established techniques for primary diagnosis of lung cancer, specifically with respect to sensitivity,
Received December 7, 1988. Accepted June 10, 1989. From the Departments of Pathology and Medicine, Lankenaiu Hospital, Philadelphia, PA. Address reprint requests to T.S. Kline, M.D., Department of' Pathology, Lankenau Hospital, Lancaster Avenue West of City Line, Philadelphia,PA 19151.
14
Diagnostic Cytopathology,Vol6, No I
specificity, diagnostic efficiency, and predictive value. We also sought those special situations where only TBNAB may provide diagnostic information.
Materials and Methods One-hundred and twenty-four patients were examined by a pulmonologist using flexible bronchoscopy and TBNAB at Lankenau Hospital. Over a 3-yr period, 128 TBNAB specimens were obtained. All patients had a suspicious clinical history and/or chest x-ray. The location of the lesion was classified into three categories: central if adjacent to a bronchus, peribronchial if more than 5 cm from a bronchial hilum, and peripheral if pleural-based or immediately subpleural. If at the time of bronchoscopy, mucosal irregularity or granularity was noted, a needle was introduced through the bronchoscope directly into the area. When mucosal alteration could not be seen in a site that was radiographically suspicious, an effort was made to pass a needle into the area through the bronchoscope under fluoroscopic guidance. All aspirations were performed as previously d e ~ c r i b e d .The ~ smears were fixed immediately in 95% ethanol and stained according to the method of Papanicolaou. The TBNAB specimens were reviewed for all major aspiration biopsy cytology criteria of malignancy: cellularity, dyshesion, monomorphism, nuclear membrane irregularity, anisonucleosis, and macronucleoli.3 The material also was evaluated for features important in the analysis of exfoliative cytologic specimens, including alteration of the nuclear/cytoplasmic ratio, hyperchromatism, and background. Interpretation of TBNAB material was done with four categories: benign, suspicious of carcinoma, carcinoma, and unsatisfactory. The specimen was interpreted as benign when it was composed of normal respiratory mucosal cells and inflammatory cells. Specimens were interpreted as suspicious when abnormal cells were sparse 0
1990 WILEY-LISS, INC.
TBNAB AND PULMONARY NEOPLASMS
or poorly preserved or when the aspirate lacked some of the criteria of malignancy. Material was regarded as unsatisfactory if blood or inflammatory cells obscured underlying cells. Sensitivity was calculated as the number of specimens from patients with carcinoma that were correctly diagnosed as positive by TBNAB (true-positives) divided by the number of specimens from patients who had carcinoma and underwent TBNAB. Sensitivity is therefore a measure of how well TBNAB diagnosed malignancy. Specificity was calculated as the number of specimens from patients with benign conditions correctly diagnosed as negative (true-negatives) divided by the total number of specimens from patients with benign conditions who underwent TBNAB. The predictive value of a positive test (i.e., the probability of a patient having a malignant condition in the presence of a positive test result) was calculated using the above parameters and assuming a prevalence rate of 50% for carcinoma in those patients coming to bronchoscopy. The predictive value of a nonpositive test was derived similarly. Diagnostic efficiency was calculated as the ratio of the sum of true-positives and true-negatives to the total number of specimens from patients undergoing TBNAB (excluding specimens deemed unsatisfactory). The statistical significance of the increase in sensitivity resulting from adding TBNAB to other bronchoscopic techniques was evaluated using the McNemar test for related samples4
Table XI. Statistical Parameters for TBNAB Parameter
Percentage
Sensitivity Specificity Diagnostic efficiency* Predictive value positivet Predictive value nonpositivet
56 74 69 100
70
*Calculation drops unsatisfactory specimens. tCalculation assumes the prevalence rate of tumor to be 50%.
suspicious were eventually confirmed as carcinoma. Nineteen specimens interpreted as benign were from patients subsequently diagnosed as having carcinoma. The sensitivity, specificity, diagnostic efficiency, and predictive values for TBNAB are presented in Table 11. For purposes of comparison, the sensitivities and diagnostic efficiencies of other bronchoscopic techniques and PNAB are presented in Tables 111 and IV, respectively. None of the patients who underwent TBNAB had any medical complication as a result of the p r ~ c e d u r e . ~
Discussion
Of the 128 TBNAB specimens obtained between 1985 and 1988, 89 were from 85 patients ultimately diagnosed as having carcinoma; the remaining 39 were from patients who were finally diagnosed as having benign conditions. The diagnosis was made by the clinician on the basis of the clinical and radiographic course of the patient in conjunction with results of cytologic and/or histologic examination. Material from another bronchoscopic technique (bronchial washings, brushings, or biopsy) or from percutaneous needle aspiration, surgery, or autopsy was available in 105 (84.7%) of the 124 patients. The results of TBNAB are presented in Table I. Ten percent of the specimens (13 of 128) were rejected as unsatisfactory. Sixteen of the 17 cases diagnosed as
TBNAB in conjunction with fiberoptic bronchoscopy augments diagnosis of pulmonary neoplasms, particularly those in the submucosal and peribronchial areas. The figures for sensitivity and specificity presented in Table I1 reflect that, of the 89 specimens from patients with carcinoma, 50 were correctly diagnosed. There were also 39 specimens from patients with benign conditions; of these, 29 were interpreted as benign. We made every effort to be conservative and, with respect to the cytologic criteria, rigidly consistent in our diagnosis of carcinoma. This approach is reflected in the relatively high proportion (1 3.3%) of specimens diagnosed as suspicious. Interestingly, all but one of these patients was ultimately diagnosed as having malignant disease. The predictive values in Table I1 indicate that a positive result by TBNAB is essentially conclusive. Predictive value is, of course, dependent on the value chosen for the prevalence of disease. With the results enumerated above, a change in assumed prevalence of carcinoma in the population undergoing bronchoscopy for possible malignancy from 50% to 70% changes the predictive value of a negative test from 70% to 50%. The predictive
Table I. Results of TBNAB
Table 1x1. Sensitivity of Bronchoscopic Techniques and PNAB
Results
Cytologic diagnosis
Carcinoma Suspicious Benign Unsatisfactory Total
Specimens 50 17 48 13 128
Ultimate diagnosis of carcinoma
50 16 19 4 89
Technique
TBNAB TBBr TBW TBBx PNAB
Number of specimens
Percentage
50189 34/66 12/33 3 1/60 11/14
Diagnostic Cytopathology, Vol6, No I
56
51.5 36 50.2 78.6
15
BHAT ET AL. Table IV. PNAB Technique
TBNAB TBBr TBW TBBx PNAB
Diagnostic Efliciency of Bronchoscopic Techniques and Number of specimens
Percentage
791115 61 192 22/33 61/91 13/16
69 66 67 67 81.3
Calculations drop unsatisfactory specimens from consideration.
value of a positive test would remain unchanged in this instance because of the absence of false-positives. If we had been less conservative in our diagnosis of malignancy and classified all of our suspicious results as carcinoma, our sensitivity would have increased to 74% from 56%, but at the cost of introducing one false-positive result. The predictive value of a positive test would have changed from 100% to 97% with a prevalence rate of 50% and to 99% with a prevalence rate of 70%. The predictive value of a negative result becomes 79% with a prevalence rate of 50% and 53% with a prevalence rate of 70%. Hence, a diagnosis of carcinoma is nearly conclusive while any other result requires additional work-up. Our sensitivity (56.2%) is consistent with the values (45% and 52%) obtained in other studies.637Although our diagnostic efficiency (6 1.7%) was lower than previously reported results,7s8our study differs from previous ones in important ways. The present study was prospective and therefore included patients with benign lung conditions, which was not the case in most previous ~ t u d i e s .We ~ . ~did not attempt, moreover, to partition our suspicious-formalignancy results retrospectively into positive or negative categories depending on the ultimate clinical diagnosis,*which would have inflated our diagnostic efficiency. As Tables 111 and IV indicate, the sensitivity and diagnostic efficiency of TBNAB is comparable with those of other bronchoscopic techniques. Fourteen specimens from patients ultimately diagnosed as having carcinoma were positive by bronchial washings, brushings, or biopsy but not by TBNAB. However, 16 specimens from patients with carcinoma were interpreted as positive by TBNAB but were not detected by other bronchoscopic techniques. Of these I6 cases, radiographic information was available in 1 1 patients; the suspicious area was peribronchial in six instances and peripheral in the remaining five. Overall, bronchial washings, brushings, and biopsy had a sensitivity of 54% (48 of 89), which increased to 72% (64 of 89) when TBNAB was also utilized. This was statistically significant using the McNemar test ( P < 0.001). Individual cellular details rather than consideration of pattern provided the best diagnostic criteria for specimens obtained by TBNAB. Hyperchromasia, alteration of the nuclear/cytoplasrnic ratio, irregularity of the nuclear 16
Diagnostic Cytopathology, Vol6,No 1
Fig. 1. Transbronchial needle aspiration biopsy of moderately to poorly differentiated carcinoma. Note the cluster of malignant cells with marked nuclear alterations (Papanicolaou stain, x 500).
membrane, and the presence of macronucleoli played the major role in the diagnosis of carcinoma (Fig. 1). By contrast, cellularity, dyshesion, and monomorphism were not useful for interpretation. Of the 124 patients in this study, 16 (12.9%) underwent PNAB. Fourteen of these 16 patients were ultimately diagnosed as having carcinoma. Tumor cells were identified in aspirates from 11 of these 14 cases without any false-positive results (sensitivity, 78.6%; diagnostic efficiency, 81.3%) and as suspicious of carcinoma in one additional case. Eight of the 11 cases diagnosed as positive and the one case diagnosed as suspicious PNAB had all been diagnosed as benign by TBNAB. Radiographic information was available in five of these nine patients; in three instances, the suspicious lesion was peripheral in location. It is of interest that, in one patient, the diagnosis by PNAB was negative while TBNAB yielded a positive diagnosis confirmed by bronchial biopsy. Chest x-ray in this instance indicated a central lesion that was less than 2 cm in diameter. We believe that our data support the addition of TBNAB to the more commonly used bronchoscopic techniques for the primary diagnosis of suspected pulmonary neoplasms. Patients with central submucosal lesions are prime candidates for this safe procedure. For effective selection of the suitable modality in a case, a team approach among pathologist, radiologist, and ciinician is necessary. In selected patients, all modalities should be performed at the time of bronchoscopy. An unequivocally positive report avoids a second bronchoscopy or thoracotomy.
References 1. Kuale PA, Bode FR, Kini S. Diagnostic accuracy in lung cancer: comparison of techniques used in fiberoptic branchoscopy. Chest 1976; 69:75 2-9. 2. Wang KP, Terry PB, Marsh BR. Bronchoscopic needle aspiration biopsy of paratracheal tumors. Am Rev Respir Dis 1978;118:17-21.
TBNAB A N D PULMONARY NEOPLASMS 3. Kline TS. Handbook of fine needle aspiration biopsy cytology. 2nd ed. New York: Churchill-Livingstone, 1988:34, 11-3.
4, McNemar Q. Note on the sampling of the correlated samples. Psychometrika 1947;12:153-7.
between
5. Herman PG, Hessel SJ. The diagnostic accuracy and complications of closed lung biopsies. Radiology 1977;125:11-4.
6. Shure D, Fedullo FP. Transbronchial needle aspiration of peripheral masses. Am Rev Respir Dis 1983;128:1090-2. 7. Schenk DA, Bower JM. Transbronchial needle aspiration in the diagnosis of bronchogenic carcinoma. Chest 1987;92:83-5. 8. Morsley RJ, Miller RE, Amy RWM, King EG. Bronchial submucosal needle aspiration performed through the fiberoptic bronchoscope. Acta Cytol 1984;28:211-7.
Diagnostic Cytopaihology,Vol6, No 1
17
