Radiol Oncol 2006; 40(Suppl 1): S77-S85.
Contributions of cytology examination and methods in lung cancer diagnostic
Doprinos citoloških preiskav in metod v diagnostiki raka
Maja Jerše1 and Marjeta Terčelj2
1Institute of Pathology, Medical Faculty, University of Ljubljana, Ljubljana 2Centre of Pulmonary Diseases and Allergy, Clinical Centre, Ljubljana, Slovenia
Background. Lung cancer (LC) is still the leading cause of cancer death according to published data world-wide and confirmed also by the data obtained from the central Cancer Registry of Slovenia. Early detection of LC has an important impact on the long-term survival rate of the patients. In spite of a great advance in imaging technology for a better visualization and early detection of the neoplasms and a variety of screening tests, only cytopathology examination finally define the neoplastic lesion.
Methods. To evaluate the contribution of cytology examination in the diagnosis of LC we studied the cytol-ogy diagnoses, comparing them with histology reports in patients, who underwent the diagnostic procedure under suspicion of the LC during last 2 years.
Results. Of a total 772 patients, in 241 patients cancer was microscopically confirmed. The most frequent diagnoses were adenocarcinoma (36.9%), squamous cell carcinoma (26.6%), and small cell carcinoma (SCLC) (12.9%). There were 22% of neoplasms classified as non-small cell carcinomas (NSCLC). From the clinician point of view considering the therapy it is very important to distinguish NSCLC from SCLC. And in our study the cytology-histology correlation between these two major types of carcinoma was almost 100%. Based only on cytology, 68 (28.2%) patients received microscopic diagnosis of malignoma, and the specimens for this group of patients were obtained mostly from transbronchial or transthoracic fine needle aspiration biopsies.
Conclusions. Cytology is of great diagnostic value, a reliable and relatively non-invasive method for patients. Cytology specimens should be taken in cases where it is not possible to obtain samples for histology.
Key words: lung neoplasms – cytology – pathology; biopsy
S78
Jerše M and Terčelj M / Cytology and lung cancer
Izhodišča. Od vseh vrst karcinomov prav pljučni rak, sodeč po objavljenih podatkih, povzroča v svetovnem merilu največjo umrljivost. To potrjujejo tudi podatki iz Registra raka za Slovenijo. Zgodnje odkritje pljučnega raka je ključnega pomena za dolgoročno preživetje pacientov. Kljub izrednemu tehnološkemu razvoju slikovne diagnostike za zgodnje odkrivanje raka in različnim presejalnim testom je za dokončno potrditev raka še vedno potrebna citopatološka preiskava.
Metode. Z namenom, da bi preučili vrednost in doprinos citoloških preiskav v diagnostiki pljučnega karci-noma, smo preučili citološke izvide in jih primerjali s histološkimi pri bolnikih, ki so bili v zadnjih dveh letih preiskani zaradi suma na pljučno neoplazmo.
Rezultati. Od skupno 772 bolnikov, je bila diagnoza pljučnega raka mikroskopsko potrjena pri 241 bolnikih. Izkazalo se je, da je najpogostejša oblika raka žlezni rak (36,9%), sledita ploščatocelični rak (26,6%) in drobnocelični rak (12,9%). 22% rakov je bilo opredeljenih kot nedrobnocelični rak. S kliničnega in terapevtskega stališča je ključno razlikovati nedrobnocelični od drobnoceličnega raka. V naši raziskavi je bila citološko-histološka korelacija med tema glavnima vrstama rakov skoraj 100%. Malignom je bil zgolj s citološkimi preiskavami potrjen kar pri 68 bolnikih (28,2%), in to predvsem s perbronhialno ali pertorakalno igelno biopsijo.
Zaključki. Citologija ima visok diagnostični pomen, je zanesljiva in sorazmerno neinvazivna metoda, primerna predvsem v primerih, ko tkivnih odvzemkov za histološki pregled zaradi različnih vzrokov ni mogoče zagotoviti.
Ključne besede: pljuča novotvorbe – citologija – patologija; biopsija
Introduction
Lung cancer is still the leading cause of cancer death worlwide mainly for men. And this is confirmed also by the data ob-tained from the central Cancer Registry of Slovenia. The incidence of lung cancer in men remains in the first place, in women lung cancer appears after the cancers of the breast, skin, rectum and corpus uteri. Early detection of lung cancer, and screening tests especially in high-risk individuals rise significantly the 5-year survival rate and have as such an important impact on the outcome. It would be necessary to evaluate the most successful screening test for an early and accurate staging of the cancer. The screening test must provide benefits,
Correspondence to: Maja Jerše, MD, Institute of Pathology, Medical Faculty, University of Ljubljana, Korytkova 2, 1000 Ljubljana, Slovenia; Phone: + 386 543 71 52; Fax: + 386 543 71 01; E-mail address: maja. jerse@mf.uni-lj.si
and must be capable of detecting the dis-ease at a point at which its course can be altered through treatment- because early detection can reduce lung cancer mortal-ity. On the other hand, test should not be dangerous, harmful, nor time-consuming, nor should it have numerous false-posi-tive results that cause necessitate invasive follow-up tests. The tests should with a high specifity predict and confirm also the benign lesions which might on chest-x-ray, or computed tomography (CT) look like cancerous lesions, and thus avoid surgical risk for a patients- such as thoracotomy.
In diagnostic work with patients sus-pected of having lung cancer, it is some-times difficult to obtain adequate biopsy material for examination by pathologist. And cytology can be very helpful and diagnostic reliable method in such cases. From our experience cytology is of high diagnostic contribution in confirming the lung cancer.
Radiol Oncol 2006; 40(Suppl 1): S77-S85.
Jerše M and Terčelj M / Cytology and lung cancer
S79
Table 1. The 241 (31.2%; 173 men, 68 women) of total 772 patients who received diagnosis of lung cancer based on cytology examination, with specification- subtyping of carcinoma
Subtype of carcinoma                                      No.                                      %
Adenocarcinoma                                               89                                     36.9
Squamous cell carcinoma                                 64                                      26.6
Small cell carcinoma                                         31                                      12.9
Non-small cell carcinoma                                 53                                      22.0
Carcinoid                                                           4                                       1.6
Methods
Into the 2-year study (2004-2005) were in-cluded all the patients with a suspicion of lung cancer, visible as a nodule or an infiltrate on x-ray or CT. The patients underwent diagnostic procedure with bronchoscopy or transthoracic fine-needle aspiration biopsy. In cases, when the tumour was visible dur-ing bronchoscopy, specimens were taken for histology, and also cytology, by means of bronchial brushings, smears or bronchial lavage. When the tumour was not visible at bronchoscopy, transbronchial sampling was performed: for histology and for cytol-ogy. Transbronchial fine needle aspiration (TBFNA) for cytology sampling were defined and obtained at different sites, as: parat-racheal, tracheobronchial, hilar, bronchial, peripheral. In some cases, when a pulmo-nary nodule cannot be reached by bronchos-copy, samples were taken percutaneously (by percutaneous transthoracic fine-needle aspiration biopsy – PTTFNAB), for cytol-ogy examination only. All the samples for the histological and cytological examination were processed in a routine fashion. For the study we examined all the cytology spec-imens taken by bronchial/transbronchial or PTTFNAB procedures, but we excluded aspirates, sputum and pleural fluids speci-mens. We evaluated the cytologic material and classified as: neoplastic, nonneoplastic disease, suspicious for malignancy, unclas-sified, unadequate material for diagnosis.
All the neoplastic lesions were further differ-entiated into subgroups according to WHO classification if possible. The diagnoses ob-tained from cytologic material were com-pared with histologic diagnoses.
Results
A total of 772 patients were examined (72% men, 28% women). The diagnosis of lung cancer was confirmed in 31,2% patients. The patients’ age ranged from 24-87 in men (mean age 64.8 years), and 41-88 years in women (mean age 61.8 years). There were 13 men and 5 women aged 50 years or less, which represents 7.5 % of the men and 7.2% of the women.
Of total 241 patients with neoplastic disease, cytology examination subclassi-fied the cancers (Table 1) as adenocarci-noma (AD) in 89 patients (in 25.5% in men, 12.3% in women), squamous cell carcinoma (SC) in 64 patients (20% in men, 5.4% in women), small cell carcinoma (SCLC) in 31 patients (10.8% in men, 3.1% in women), and carcinoids in 4 patients. A diagnosis of non- small cell lung carcinoma (NSCLC) was based in 53 patients (16.9% in men, 3.8% in women) because on cytological basis it wasn’t possible to subclassify them any further. In women and in men the most frequent carcinoma was AD (confirmed in 51.4% of all women and in 34% of all men). In this group classified as AD were includ-Radiol Oncol 2006; 40(Suppl 1): S77-S85.
S80
Jerše M and Terčelj M / Cytology and lung cancer
Figure 1a. Cytology smear obtained by PTTFNAB demonstrated atypical plasma cells, consistent with plasmacytoma (MGG, x 400), radiologically and clini-cally mimmicking carcinoma with local invasion into the rib.
Figure 1b. Immunocytochemistry on cytospin prepara-tions revealed monoclonal expression of kappa-light immunoglobulin chain (alkaline phosphatase, x 600).
ed also metastatic AD from breast (4 cases), colon (3 cases), kidney (2 cases).
On the second place was SC (in 26.8% of all men and in 22% of all women). SCLC was confirmed in aproximatelly 13%, in both men as well in women. 5% (12 cases – 9 AD and 3 SC) were uncorrectly classified in cor-relation to histological diagnoses. From the group with cytology diagnosis as NSCLCs,
aprox. 40% were subclassified by histo-logical examination (AD 14 cases, 7 SC), the remainder were further histologically classi-fied only after surgical lung resection.
Based only on cytology, 68 patients (28.2%) received microscopic diagnosis of lung carcinoma. In all these cases histol-ogy was negative or material inadequate or specimens for histology examination not
Table 2. Cytologic-histologic correlation of diagnoses obtained from a total of 772 patients examined under sus-picion of lung cancer
Type of C-H correlation	No.	%
C neg - H neg	493	63.9
C pos - H pos	165	21.3
C pos - H neg	68	8.8
C susp -H pos	19	2.5
C neg - H pos	10	1.3
C pos - H susp	8	1.0
C susp - H neg	4	0.5
C uncl - H pos	3	0.4
C susp - H susp	2	0.3
Abbreviations: C = cytology; H = histology; neg = negative for malignancy; pos = positive for malignancy (cancer); susp = suspicious for malignancy; uncl = unclassified
Radiol Oncol 2006; 40(Suppl 1): S77-S85.
Jerše M and Terčelj M / Cytology and lung cancer
S81
Figure 2a. Metastasis of skin basal cell carcinoma-BCC (MGG, x 400), clinically presented as primary lung carcinoma with solitary nodule. Cytology specimen, obtained by PTTFNAB.
taken because the lesions weren’t possible to be reached by bronchoscopy forceps. In some cases a broad spectrum of im-munocytochemical stainings was used to elucidate the lesion (Figures 1a, 1b), in others only very detailed additional clinical data were clue to final diagnosis (Figures 2a, 2b). This high percentage (of 28.2%) therefore prove that cytology specimens are of great diagnostic value. In such cases the most useful type of cytology specimens were taken by TBFNA and PTTFNAB. In all these cases (except in two patients) the diagnoses of malignancy were confirmed and further histologically subclassified by the specimens of resected lung. In the re-mainder two cases (0.8%) the cytology was false-positive and misinterpretated due to heavy reactive inflammation changes and atypical adenomatoid hyperplasia which were disclosed after wedge lung resection.
Of all the diagnoses given (for the whole material of 772 cases) by cytological exami-nation only 3.6% (28 cases) were unclassi-fied or diagnosed as ‘suspicious for malig-nancy’. Of these cases histology confirmed the malignoma in 22 cases (78,6%).
In Table 2 are summarized observations from cytologic-histologic correlation.
Figure 2b. The same tumour, BCC, confirmed by his-tology examination after wedge resection of the lung (HE, x 100).
Discussion
In lung carcinoma, cell typing of tumour is important in determining prognosis and often in influencing therapy. Therefore, it is highly desirable to obtain a correct morphological diagnosis. Clinicians are sometimes reluctant to rely on cytological procedures. But many times it is difficult to obtain adequate biopsy material for ex-amination by the pathologist. That could be due to a tumour location (peripheral location) or practical difficulties with bron-choscopy procedure (patient’s dyspnoe), and in this cases specimens for histology examination may not be representative or even false negative. Cytological examina-tions in such cases are very helpful. Thus, when the tumour is not clearly visible by bronchoscopy, the samples may be taken percutaneously by PTTFNAB, guided by CT or ultrasound (US). US-PTTFNAB makes accurate access to lesions in pleura, pe-ripheral lung, anterior mediastinum, bone and soft tissue and thus unnecessary open biopsy (thoracotomy) can be avoided. By this procedure a false-positive diagnosis of pulmonary malignancy is exceedingly rare, and are estimated to around 1.5%.1
Radiol Oncol 2006; 40(Suppl 1): S77-S85.
S82
Jerše M and Terčelj M / Cytology and lung cancer
The specifity and positive predictive value of PTTFNAB in malignancy are ex-tremely high. However, the rate of definite benign diagnosis by PTTFNAB is low, ranging from 4-14% depending on the preva-lence of lung cancer in the population studied.2,3 So, a negative cytologic report by this procedure may not represent a truly benign diagnosis, but merely a failure to reach a malignant one. A repeat biopsy reduces the rate of false-negative result. The interpretation of the received samples by the cytologist must be careful, to inform the clinicians about the adequacy of the specimen, and even recommend a repeat sampling for additional special stainings or immunocytochemical analysis to avoid misinterpretation of the specimen and to reduce false-negative results. It is impor-tant to clarify that false-negative interpretation of the cytology specimen has many causes. Size and location of the lesion are important. Also, in cases, where the lesions are very hemorrhagic, a bloody aspirate with very few viable malignant cells can be obtained. On the other hand, some lesions are almost entirely necrotic or associated with significant amount of fibrosis, etc. Besides, PTTFNAB, provides the material for bacteriological study in patients with pulmonary infections.
However, it is noteworthy to mention that this method (PTTFNAB) obtained over 21% of all lung cancers for our study.
Concerning the value of cytology, several reports have shown also a good correlation between cytologic and eventual histologic diagnoses.4-7 And our results support those observations. In our series, cytologic-his-tologic correlation in SC and AD is aprox. 85%, in SCLC more than 97%, and in poorly differentiated carcinoma less than 60%. Factors that can affect the correlation between the cytologic and histopathologic examination include beside the adequa-cy (representativity) of sampling also the
Radiol Oncol 2006; 40(Suppl 1): S77-S85.
degree of the cellular differentiation of the tumour and the observer variations in interpretation. Anyway, it is noteworthy, that from the therapeutic implications it is of utmost importance to distinguish with a high diagnostic accuracy NSCLC from SCLC. And in our experience, cytological specimens have almost 100% efficiency of correct diagnosis of SCLC. Probably, the most important finding was that in only one SCLC diagnosis was inconclusive where differential diagnosis included also malignant lymphoma.
However, it is true that cytological speci-mens might sometimes be inconclusive, and in other cases merely give diagnostic indications that need confirmation by his-tology. In our study, in 3.6% of all cases the cytologic diagnosis was inconclusive or ‘malignancy suspected’, and in this cases subsequent histology examination (some-times with open biopsy- mediastinoscopy /thoracotomy) was necessary to evaluate the process. Histology confirmed the ma-lignant process in 78.6% of those cases.
Concerning the value of cytology, dif-ficulties arise (in a very high percentage – 85%) when the clinicians need the speci-fication of a benign process (i.e. inflam-mation, fibrosis, granulomatous lesions, hamartoma, etc) where the cytologic diag-nosis is interpreted only as ‘negative’. In this setting, histology examination is much more specific and valuable.
On the other hand, the diagnosis of lung carcinoma was obtained by cytology specimens alone in 28.2%. In such cases, the majority of cytological specimens were performed by TBFNA and PTTFNAB, and were obviously of high diagnostic value.
Furthermore, it is also important to be able to sort out a tumourous mass in the lung which is a metastasis from primary cancers elsewhere. Conventional morpho-logic evaluation of a patient’s tumours may not accurately define the relationship
Jerše M and Terčelj M / Cytology and lung cancer
S83
between multiple tumours in the same in-dividual and may possibly lead to inappro-priate treatment. Of course, the cytologic-histologic correlation in confirming the metastatic tumours can significantly drop when there is no material for an extra im-munocytochemical stainings. On the other hand, even traditionally used immunohis-tochemistry as a method itself has some limitation. For example, difficullty may be encountered when attempting to differenti-ate squamous cell carcinomas of head and neck and pulmonary origins.8 But the new insights of molecular techniques illustrate the possible practical diagnostic utility us-ing the material obtained by fine needle aspiration biopsy (FNAB) in assessment of primary and metastatic malignancies. It has been shown that that synchronous and metachronous neoplasms are geneti-cally different. Evaluation of primary and metastatic neoplasms may be enhanced by genomic analysis of loss of heterozygosity (LOH). This methodology can be used on a very small amounts of material, even on a cytologic smear. Panels of defined genetic targets are used to asses specific tumours and the relationships between primary and metastatic sides. LOH analysis provides also insight that additional genetic alterations may be observed in the metastatic foci, which are not seen in the primary ori-gin, suggesting genetic progression of the neoplasm. Thus, LOH analysis may play a significant role in evaluating treatment efficacy and is useful and practical adjunct to conventional cytologic diagnosis, as only small numbers of cells from a cytologic smear are required.9,10
Because the patient’s outcome is closely associated with the stage at diagnosis of cancer, there has been persistent interest in testing the methods for early detection of lung cancer. It is wellknown that the current staging test for lung cancer uses the TNM system, where the system relies on
the pathological evaluation of the primary tumour (T), regional nodes (N) and distant metastases (M). The most common sites of metastases are the mediastinal lymph nodes. There are several technologies avail-able for evaluation of mediastinal lymph nodes (LN). LN larger than 1cm in diameter can be detected by CT and are usually pressumed to contain metastatic disease. According to published data CT misses the LN metastasis in 13% and is false positive in 50%.11 Because of this limitation, mediastinoscopy is accepted as ‘golden standard’ but requires general anaesthesia and expertize to perform safely, and there-fore not universally used. Recently, positron emission tomography (PET) with its sensitivity and specifity12 is suggested as non-invasive method for evaluating cancer stage, but cannot 100% distinguish between inflammation and cancer. In the final diag-nosis histological analysis is used to determine the presence/absence of mediastinal LN metastases. However, the presence of micrometastases has a significant impact on long-term survival even among patients with histologically ‘normal’ LN. Namely, survival statistics indicate, that histology is inadequate, because the 5-year survival rate in patients with pathological stage I disease (without histological evidence of LN metastasis) is only 62%, and in patients with metastasis in hilar LN (but not medi-astinal LN) 42%.13 Obviously, histological examination can miss metastasis if not used wih serial sectioning and immunohis-tochemical stainings which are expensive for routine basis. To reduce unnecessary preoperative invasive methods and to in-crease correct preoperative staging, US (as minimally invasive method) with FNAB is used as safe and accurate method for detec-tion LN metastasis, and thus for staging. The molecular characterization of LN tissue is a newly emerging field. Micrometastasis cannot be detected with standard cytologi-Radiol Oncol 2006; 40(Suppl 1): S77-S85.
S84
Jerše M and Terčelj M / Cytology and lung cancer
cal and histological methods. Tumour cell can be detected using molecular techniques such as immunohistochemistry or reverse transcriptase-polymerase chain reaction (RT-PCR).14 ,15 Using this method several markers can be identified, as telomerase and KS1/4, which can detect metastatic disease in a cytopathology-negative LN. 16 ,17 Telomerase can represent a unique target for molecular staging in lung cancer. Evidence of telomerase overexpression was noted in both pathologically positive and negative mediastinal LN from patients with NSCLC after US-FNAB. And according to published studies telomerase expression is absent in all normall LN.18,19 But some new data have put into question the specifity of telomerase as a marker of malignancy20, and in recent publication the effect of EGRF (epidermal growth factor receptor) expression in NSCLC appears promising.21
There are several experiments and at-tempts to identify the specific gene which would be highly expressed in NSCLC. Iwao et al22 showed that lunx is highly expressed in NSCLC but not in normal LN. Another molecular high specifity marker for NSCLC is KS1/4, a gene that encodes a glycopro-tein expressed on epithelial cells.23 Because cytokeratin genes are normally expressed in epithelial but not lymphoid cells, the presence of cytokeratin mRNA in a LN suggests the presence of metastatic cells of epithelial origin. RT-PCR is a highly sensitive technique and is reported to be capable to detect one cancer cell per one milion normal cells.24 Recently, RT-PCR for p53 and K-ras was shown to improve detection of occult LN metastasis.25
There are some controversial publica-tions whether the examination of induced sputum is of use in detecting an early lung cancer. Thunnissen26 in a review analysed the new technical aspects of conventional sputum cytology. Nuclear image analysis, the measurement of heterogeneous nuclear
Radiol Oncol 2006; 40(Suppl 1): S77-S85.
riboprotein A2/B1, p53, and DNA examina-tion appear to be promising for the detec-tion of early lung cancer, especially in a high-risk patients. K-ras analysis could be useful in detection of pulmonary adenocar-cinoma. To date, microsattelite alteration have been detected also on specific cells scraped from a sputum sample, which confirms that chromosomal abnormalities (LOH) are present also in the cytologically abnormal sputum cells. However, this ap-proach is not practical for screening tests.
In conclusion, cytology is of great diag-nostic value, a reliable and relatively non-invasive method for patients. Eventhough, that cytology has sometimes less ‘defined’ diagnosis than histology, cytology speci-mens should be taken in cases where it is not possible to obtain samples for histol-ogy. Also, our results confirm that there is a good correlation between cytology and histology diagnoses, especially concern-ing the malignant lesions. In the prospect, there are also promising some molecular studies for an early detection, evaluation, and staging of lung cancer.
References
1.   Charig MJ, Stutley JE, Padley SP, Hansell DM. The value of negative needle biopsy in suspected operable lung cancer. Clin Radiol 1991; 44: 147-9.
2.   Calhoun P, Feldman PS, Armstrong P, Black WC, Pope TL, Minor GR, et al. The clinical outcome of needle aspirations of the lung when cancer is not diagnosed. Ann Thorac Surg 1986; 41: 592-6.
3.   Khouri NF, Stitik FP, Erozan YS, Erozan YS, Gupta PK, Kim WS, et al. Transthoracic needle aspiration biopsy of benign and malignant lung lesions. AJR Am J Roentgenol 1985; 14 4 : 281-8.
4.   Westcott JL. Direct percutaneous needle aspiration of localized pulmonary lesions: result in 442 patients. Radiology 1980; 137: 31-5.
5.   Horrigan TP, Bergin KT, Snow N. Correlation between needle biopsy of lung tumors and his-topathologic analysis of resected specimens. Chest 1986; 90: 638-40.
Jerše M and Terčelj M / Cytology and lung cancer
S85
6.   Penketh AR, Robinson AA, Barker V, Flower CD. Use of percutaneous needle biopsy in the investi-gation of solitary pulmonary nodules. Thorax 1987; 42: 967-71.
7.   Vazquez MF, Yankelevitz DF. The radiologic ap-pearance of solitary pulmonary nodules and their cytologic-histologic correlation. Semin Ultrasound CT MR 2000; 21: 149-62.
8.   Leong PP, Rezai B, Koch WM, Reed A, Eisele D, Lee DJ, et al. Distinguishing second primary tumors from lung metastases in patients with head and neck squamous cell carcinoma. J Natl Cancer Inst 1998; 90: 972-7.
9.   Huang J, Behrens C, Wistuba I, Gazdar AF, Jagirdar J. Molecular analysis of synchronous and metachronous tumors of the lung: impact on management and prognosis. Ann Diagn Pathol 2001; 5: 321-9.
10. van der Sijp JR, van Meerbeeck JP, Maat AP, Zondervan PE, Sleddens HF, van Geel AN, et al. Determination of the molecular relationship between multiple tumors within one patient is of clinical importance. J Clin Oncol 2002; 20: 1105-14 .
11. Wallace MB, Silvestri GA, Sahai AV, Hawes RH, Hoffman BJ, Durkalski V, et al. Endoscopic ultra-sound-guided fine needle aspiration for staging patients with carcinoma of the lung. Ann Thorac Surg 2001; 72: 1861-7.
12. Dwamena BA, Sonnad SS, Angobaldo JO, Wahl RL. Metastases from non-small cell lung cancer: mediastinal staging in the 1990s-meta-analytic comparison of PET and CT. Radiology 1999; 213: 530-6.
13. Mountain CF, Dresler CM. Regional lymph node classification for lung cancer staging. Chest 1997; 111: 1718-23.
14. Salerno CT, Frizelle S, Niehans GA, Ho SB, Jakkula M, Kratzke RA, et al. Detection of occult micrometastases in non-small cell lung carcinoma by reverse transcriptase-polymerase chain reac-tion. Chest 1998; 113: 1526-32.
15. Maruyama R, Sugio K, Mitsudomi T, Saitoh G, Ishida T, Sugimachi K. Relationship between early recurrence and micrometastases in the lymph nodes of patients with stage I non-small-cell lung cancer. J Thorac Cardiovasc Surg 1997; 114: 535-43.
16. Wallace MB, Block M, Hoffman BJ, Hawes RH, Silvestri G, Reed CE, et al. Detection of telomerase expression in mediastinal lymph nodes of patients with lung cancer. Am J Respir Crit Care Med 2003; 167: 1670-5.
17.  Wallace MB, Block MI, Gillanders W, Ravenel J, Hoffman BJ, Reed CE, et al. Accurate molecular detection of non-small cell lung cancer metastases in mediastinal lymph nodes sampled by endo-scopic ultrasound-guided needle aspiration. Chest 2005; 127: 430-7.
18. Marchetti A, Pellegrini C, Buttitta F, Falleni M, Romagnoli S, Felicioni L, et al. Prediction of survival in stage I lung carcinoma patients by tel-omerase function evaluation. Lab Invest 2002; 82: 729-36.
19. Hara H, Yamashita K, Shinada J, Yoshimura H, Kameya T. Clinicopathologic significance of tel-omerase activity and hTERT mRNA expression in non-small cell lung cancer. Lung Cancer 2001; 34: 219-26.
20. Kumaki F, Kawai T, Hiroi S, Shinomiya N, Ozeki Y, Ferrans VJ, et al. Telomerase activity and expression of human telomerase RNA component and human telomerase reverse transcriptase in lung carcinomas. Hum Pathol 2001; 32: 188-95.
21. Tsao MS, Sakurada, A, Cutz JC, Zhu CQ, Kamel-Reid S, Squire J, et al. Erlotinib in lung cancer-mo-lecular and clinical predictors of outcome. N Engl J Med 2005; 353: 133-44.
22. Iwao K, Watanabe T, Fujiwara Y, Takami K, Kodama K, Higashiyama M, et al. Isolation of a novel human lung-specific gene, LUNX, a potential molecular marker for detection of microme-tastasis in non-small-cell lung cancer. Int J Cancer 2001; 91: 433-7.
23. Perez MS, Walker LE. Isolation and characteriza-tion of a cDNA encoding the KS1/4 epithelial carcinoma marker. J Immunol 1989; 14 2: 3662-7.
24. Mori M, Mimori K, Inoue H, Barnard GF, Tsuji K, Nanbara S, et al. Detection of cancer microme-tastases in lymph nodes by reverse transcriptase-polymerase chain reaction. Cancer Res 1995; 55: 3417-20.
25. Hashimoto T, Kobayashi Y, Ishikawa Y, Tsuchiya S, Okumura S, Nakagawa K, et al. Prognostic value of genetically diagnosed lymph node microme-tastasis in non-small cell lung carcinoma cases. Cancer Res 2000; 60: 6472-8.
26. Thunnissen FBJM. Sputum examination for early detection of lung cancer. J Clin Pathol 2003; 56: 805-10.
Radiol Oncol 2006; 40(Suppl 1): S77-S85.