XI Respiratory Tract

Nongynecological Cytology Practice Guidlines

prepared by the American Society of Cytopathology, Cytopathology Practice Committee.

Adopted by the ASC Executive Board, March 2, 2004

XI Respiratory Tract Supplementary Information


XI.A. Epidemiology and Public Health

XI.A.1. Incidence and Mortality

Lung cancer is now the most common fatal malignancy in both sexes and its incidence is second only to prostate cancer in men and breast cancer in women.103 In the year 2001, 169,500 new cases of lung cancer, with 157,400 deaths were expected (90,100 men; 67,300 women.)103 Cancer of the lung occurs at an average age of 60 years, and is rare before 40 years of age.104

XI.A.2. Risk Factors for Lung Cancer

XI.A.2.a Tobacco Smoke

Lung cancer is strongly associated with tobacco smoking, particularly cigarettes.104 There is also an increased risk associated with secondary smoke or “passive smoking”.105,106,107,108 It is estimated that 87% of lung cancers occur in tobacco smokers.109 The risk pertains to all major types of lung cancer, particularly squamous and small cell types, which rarely occur in patients who have never smoked, and also adenocarcinomas, although they are less strongly associated with smoking.110,111,112,113,114 Tobacco smoking is also associated with a wide-variety of other cancers, including head and neck, gastrointestinal, and bladder cancers, as well as other non-malignant conditions, such as chronic obstructive pulmonary disease.111,115

Although many continue to smoke despite the well-known risks of tobacco usage, the good news is that smoking rates have been decreasing, particularly among males.104 A downward trend in smoking should be followed by a decrease in overall lung cancer mortality116 which, if current trends continue, may in fact have begun in 1997. For those who have quit smoking for 10 years or more, the lung cancer rate approaches that of persons who never smoked.111,116,117 There also has been a shift in the histologic type and location of lung tumors. In men and women, the relative proportion of adenocarcinomas, particularly bronchioloalveolar carcinoma, and peripherally located tumors has increased markedly.110,118 These tumors are less strongly associated with smoking.

XI.A.2.b. Other Co-Factors

The pathogenesis of lung cancer is probably multifactorial, including environmental and genetic factors119, such as radiation, air pollution120, diet121, and occupational exposure to certain metals (e.g., arsenic, nickel, chromate, cadmium) and chemicals (e.g., chloromethylesters, acrylonitrile, vinyl chloride)104. Radon is perhaps the second most common cause of lung cancer.122,123 Genetic or familial cases of lung cancer also occur.124,125,126,127 Lung cancer is the most common malignancy associated with asbestos exposure. Asbestos potentiates the ill effects of tobacco smoking, and in those exposed to asbestos, the risk of lung cancer in smokers is substantially increased compared with non-smokers.128,129,130,131 Pre-existing pulmonary disease, such as interstitial fibrosis and emphysema, may predispose to lung cancer.132,133 Viruses may also be associated with the development of some lung cancers.110,134

The actual mechanism by which cells become malignant is thought to involve an accumulation of genetic alterations, including oncogenes, such as c-myc in small cell carcinomas and K-ras in adenocarcinomas, with loss or inactivation of tumor suppressor genes, such as p53, and alterations of the short arm of chromosome 3.135,136 It is estimated that by the time a tumor has become clinically apparent, some 10 to 20 genetic mutations have occurred.137

XI.A.3 Screening

The most common routine screening tests for lung cancer are chest radiograph (x-ray) and sputum cytology. However, these procedures are no longer recommended for routine screening of asymptomatic persons, because they lack sufficient sensitivity and specificity.

The accuracy of the chest x-ray is limited by the capabilities of the technology, suboptimal technique, and observer variation among radiologists. By the time lung cancer is suspected on chest x-ray, micrometastases often have occurred, limiting the effectiveness of early detection. Furthermore, the yield of screening chest radiography is low, largely due to the low prevalence of lung cancer in asymptomatic individuals, even those at high risk.

A significant reduction in the mortality rate is the gold standard for any cancer-screening test. A 20-year follow-up study of the Mayo Lung Project (MLP) reinforces the conclusion published in 1986 that screening for lung cancer by frequent chest x-rays does not save lives.138 Perhaps more importantly, these latest findings suggest that aggressive screening for lung cancer could do more harm than good, by detection of lesions that will never cause serious illness or death. In the absence of screening, it is likely that these incidental lesions would never be found. Such over-diagnosis can lead to unnecessary anxiety or, more seriously, to expensive and risky biopsies or surgery. These findings may diminish enthusiasm for newer, ever more sensitive screening technology.

Sputum cytology is an even less effective screening test than chest x-ray, largely due to its low sensitivity. In screening trials, the sensitivity of sputum cytology ranges as low as 10% to 20%. When combined with chest x-ray, only one in four cancers are detected by sputum cytology alone; most of those were squamous cell carcinomas found at a favorable stage. Mass screening to detect lung cancer with tests that lack a high sensitivity is inefficient. Moreover, the spectrum of lung cancer type has shifted over the last two decades. Squamous cell carcinoma used to be the most common type of lung cancer and is usually centrally located where it is more readily detected by sputum cytology. Now, adenocarcinoma is the most common type and it is usually peripherally located where it is less readily detected by sputum cytology.139,140 Another disadvantage of sputum cytology is that even when a cancer is detected, another method must be used to localize the lesion. Still, attempts to refine the use of sputum cytology for early cancer detection continue.141

At present lung cancer is mainly treated in advanced stages. Identifying molecular markers that could detect the disease while still confined to the bronchial epithelium would potentially allow cures with local therapies. Preliminary data suggest that over-expression of a tumor-associated antigen, heterogeneous nuclear ribonucleoprotein (hnRNP), is found in normal appearing lung epithelium in patients with lung cancer. If biomarker-based screening were found to be accurate, this could lead to identifying disease at a much earlier and more treatable stage. Aerosol delivery of therapeutic agents is a possible solution to the problem of treating multiple lesions in the tracheobronchial tree.

In summary, there is no compelling evidence that screening for lung cancer using chest x-ray or sputum cytology can reduce lung cancer mortality. No major medical organization currently recommends routine screening of either the general population or of smokers for lung cancer with either chest x-rays or sputum cytology.142,143,144,145,146,147 There are intensive efforts to improve lung cancer screening with newer technologies (e.g., low-radiation-dose [spiral] computed tomography) and molecular techniques which, although promising, have not been validated in large controlled studies.148,149 Counseling people against the use of tobacco products is currently the best method to reduce lung cancer mortality.

XI.B. Specimen Collection and Submission

The diagnostic reliability of respiratory cytology depends upon many factors that affect sensitivity and specificity. Time and method of specimen collection, number of samples submitted, location of the tumor, and tumor type affect sensitivity. Specificity is affected by state of preservation, lack of clinical history, background necrosis and inflammation, as well as cellular differentiation.150

XI.B.1. Sputum

Sputum has been a mainstay for the diagnosis of lung cancer with the first reports by Hampeln in 1876 and 1887, Menetrier in 1886 and Betschardt in 1895. Earlier reports of malignant cells in sputum by Walshe in 1843 and Beale in 1860 represented exfoliated cells from upper airway tumors.151

Sputum is most easily obtained in patients who are symptomatic and have a productive cough. Three to five sputum specimens will most likely detect malignant cells.152 Pooled morning samples are the optimal specimens in which to detect cancer cells in symptomatic patients.150

Induced sputum, collected by specially trained individuals, may be necessary in patients with a lung mass who are not producing sputum. Nebulizing solution stimulates secretions in the respiratory tract. Nebulized solutions are varied and may include: 15% nebulized saline, 15% saline with 20% propylene glycol, or heated (115_F) 3-8% saline.150 Hypertonic solutions are more successful in inducing sputum, but less well tolerated.153 While cytologists are most familiar with induced sputum specimens used for diagnosis of cancers, in the last decade they have been used for assessing the inflammatory components of patients with asthma, sarcoidosis and chronic obstructive pulmonary disease.154,155 Induced sputum specimens are also used for the investigation of opportunistic organisms in immunosuppressed patients.156

Postbronchoscopy sputum is used in conjunction with bronchial washings and brushings for diagnosis of carcinoma. It may have a higher diagnostic rate than standard sputum collection.157,158

Fresh sputum submitted immediately to the laboratory is preferred. Preservation with 50% to 70% ethanol may be necessary when specimens cannot be directly submitted. Saccamanno fixative consisting of 50% ethanol with polyethylene glycol (Carbowax®) is also used. Proprietary liquid based cytology medium should be added if using liquid based cytology processors. Some laboratories use mucolytic agents in the preparation process for sputum specimens.159,160,161,162

XI.B.2. Bronchial Brushing

Bronchoscopy is performed when patients have respiratory symptoms or a radiologically evident lung mass. During fiberoptic bronchoscopy the operator may wish to sample a lesion by brushing. After brushing, the cellular material may be submitted as follows:

  • The surface of the brush may be rotated on slide(s) and immediately fixed and/or
  • The brush may be submitted in a physiologic transport solution or a proprietary solution used for liquid based techniques or
  • The brush may be discarded after completely removing the adherent cellular material into a container of fixative or transport solution.

XI.B.3. Bronchial Washing

During bronchoscopy small aliquots of balanced saline solution are washed over a directly visualized area and removed immediately by using suction. These washings are usually submitted immediately to the laboratory. If a delay is anticipated, they may be partially fixed in an amount of 50-70% ethanol that is equal to the specimen volume, or in the proprietary transport medium supplied by one of the manufacturers of liquid based processors.

XI.B.4. Bronchoalveolar Lavage

Bronchoalveolar lavage (BAL) is a method of sampling the lower respiratory tract. A bronchoscope is advanced until it is “wedged” into a subsegmental bronchus. Saline or a balanced salt solution that is suitable for use in vivo is introduced and re-aspirated. The volume of fluid and the extent of the lavage depend upon the suspected disease and the training and preference of the bronchoscopist, as well as the patient’s tolerance. Usually 20-60 mL aliquots of fluid are used to separately sample up to three subsegmental areas of the lung. In this manner, 1% of the lung, or 1 to 3 million alveoli are sampled. The samples initially contain respiratory epithelium, but the latter portions of the aliquots are enriched for alveolar components.163  BAL is performed for the detection of microorganisms, interstitial lung disease, transplant rejection, pulmonary hemorrhage, acute inflammatory diseases, disorders in which lymphocytes predominate, and malignancy.150 BAL specimens are submitted fresh because microbiologic studies, immunologic studies or chemical analyses may be requested. When separate aliquots are provided for ancillary studies then the BAL specimen may be fixed as previously described for bronchial washings (section II.C ).

XI.B.5. Transbronchial or Transesophageal Fine Needle Aspiration Biopsy

Transbronchial fine needle aspiration biopsy is performed during the bronchoscopic procedure to sample endobronchial or peribronchial lesions and peritracheal or peribronchial lymph nodes, usually for evaluation of malignancy. Transesophageal FNAB is usually performed to evaluate paraesophageal abnormalities in the chest cavity or mediastinal and lower thoracic lymphadenopathy. A small, sheathed needle is advanced during bronchoscopy or endoscopy, and under fiberoptic visualization is introduced into the lymph node or lesion. Suction is applied while vigorously sampling the site. Suction is released, the needle is re-sheathed and removed from the bronchus. The material is expressed onto slides and direct smears are prepared as described in section II.A.4. Additionally or alternatively, the needle can be rinsed in transport medium and submitted as a liquid based sample. The needle should never be submitted.9

XI.B.6. Pulmonary Microvascular Cytology

Pulmonary microvascular cytology is not commonly used to evaluate pulmonary lymphatic carcinomatosis. A pulmonary artery is catheterized and the catheter is wedged into a small vessel. A blood sample from the wedged pulmonary catheter is collected into a heparinized tube. The heparinized blood sample is processed to separate red cells from any diagnostic cells, most frequently by gradient centrifugation.164 The residual white cell components are evaluated for the presence of carcinoma. Megakaryocytes, which signal an adequate specimen and which are normally seen in the pulmonary bed, must be distinguished from cancer cells.165

XI.B.7. Percutaneous Thoracic Fine Needle Biopsy

Percutaneous thoracic FNAB is performed for evaluation of any pulmonary abnormality, but is usually used for evaluation of suspected malignancy. The diagnostic sensitivity ranges from 75-95% and the specificity from 95-100%.166 Percutaneous transthoracic FNAB is usually radiologically guided. A variety of imaging modalities are used including computerized axial tomography (CT) scan, fluoroscopic guidance and ultrasound guidance. The lesion is entered with a hollow needle containing a stylet. The stylet is removed and the lesion is aspirated.

Pneumothorax is the most significant complication, but of those patients experiencing pneumothorax, 5-10% require treatment; most cases of pneumothorax resolve without intervention. Hemoptysis occurs in up to 8% of patients. Rare complications of air embolism do occur. Contraindications for performing the procedure may include an uncooperative patient unable to remain still during the procedure, anticoagulation therapy or bleeding diathesis, poor lung function, pulmonary hypertension, or a suspected vascular lesion.166

Because lung FNAB is associated with possible serious complications, a cytologist is often requested to perform intraprocedural adequacy evaluation to avoid multiple thoracic punctures. Assessment of adequate pulmonary specimens includes evaluation of: overall cellularity, architecture, presence of malignant features, presence of inflammatory features, and cellular elements that explain the lesion that is identified radiologically. Once adequacy has been established, the cytologist may direct triage of remaining material or subsequently obtained material for appropriate ancillary studies such as cultures, flow cytometry or immunologic studies.
See section IX.

Post procedural x-rays are often obtained by the attending clinician to detect pneumothorax.166

XI.B.8. Pleural Fluid

Normally, not more than 15 mL of pleural fluid exist in the pleural space, although liters of blood plasma are filtered daily through the semipermeable membrane of endothelial capillaries of the mesothelial coverings. Clinically, pleural fluid accumulations of less than 250 mL are undetectable, but radiological imaging can usually detect smaller amounts.167,168 Pleural fluid specimens are usually obtained by thoracentesis or during thoracic surgery. Thoracentesis is usually performed under local anesthesia with radiological guidance. Sterile technique is used to prepare the skin and the thoracic wall is punctured entering the thoracic cavity with a large caliber needle. For small fluid accumulations the entire specimen is submitted for laboratory evaluation. For larger pleural effusions, 50 mL of well-mixed fluid should be sent for cytologic examination; however, the entire specimen is also acceptable.8 All fluid specimens must be transported in accordance with OSHA regulations for biohazardous substances.9

Pleural fluid is collected into a dry container and submitted in the fresh state to the laboratory.7 Adding 3-5 IU heparin/mL to a container prior to obtaining a bloody sample will not adversely affect morphology and will usually prohibit clotting. If delay in transportation to the laboratory is unavoidable, fluids may be kept refrigerated at 4ºC until the specimen can be processed. The chemical composition of pleural fluid is such that it usually maintains cellular integrity for up to a week or more with refrigeration.169 However, some laboratories opt to partially fix specimens at the time of collection with 50% ethanol equal to the volume of the specimen.170 Alternatively, proprietary transport medium supplied by the manufacturers of liquid based processors may be used. Any added fixative should be noted on the requisition.

XI.C. Laboratory Sample Processing

XI.C.a. Sputum

Sputum samples received fresh can be processed by the pick and smear technique.171 Less widely used is the Saccamanno collection technique. In this technique, sputum samples are collected in 50% ethanol and 2% polyethylene glycol (Carbowax®). The sample is homogenized in a blender, concentrated by centrifugation, and smears are prepared from the cell pellet.172 Sputum specimens can be collected in alcohol or proprietary transport medium supplied by the manufacturers of liquid based systems. Alcohol or proprietary transport medium supplied by the manufacturers of liquid based systems may also be added to fresh specimens that are received in the laboratory. Slides can then be prepared using cytocentrifugation, manual or automated filtration, sedimentation and/or cell block.

XI.C.b. Bronchial Brushing

Bronchial brushings may be collected from multiple sites in the respiratory tree during a single procedure. Bronchial brushings may be received as direct smears, fluid-filled containers with brushes enclosed, or liquid specimens in which the sample was removed from the brush. The type of sample received dictates processing. By convention, slides prepared from bronchial brushings are stained using a Papanicolaou stain. However, if submitted as air-dried preparations, they are amenable to Romanowsky staining. The number of slides, the variety of specimens, and the site(s) of brushing should be documented.

XI.C.c. Bronchial Washing

Bronchial washings are recovered under bronchoscopic visualization from a number of sites. They are usually low volume, fresh specimens. If a delay is expected in transportation to the laboratory, bronchial washings may be received preserved in 50-70% ethanol equal to specimen volume, or in the proprietary transport medium supplied by the manufacturers of liquid based processors. Depending upon their volume and cellularity, they may first need to be centrifuged (III.B.2 ) with the re-suspended pellet being used for direct smears, preparation of slides by using automated liquid-based methods, filtration or subsequent cytocentrifugation. A cell block can also be prepared from the centrifuged sample. Specimens of low cellularity and low volume may be cytocentrifuged directly or processed using automated liquid based systems. Specimens submitted in commercially available preservative products should be processed as indicated by the manufacturer. Cell blocks can be prepared from these liquid based processed specimens. Most bronchial washing preparations are stained by the Papanicolaou technique.

XI.C.d. Brochoalveolar Lavage

Bronchoalveolar lavage (BAL) specimens are often received fresh and may be submitted as separate aliquots or as a single specimen. Volumes vary depending upon the preference of the bronchoscopist, the disease process and the site(s) of abnormality. Additional microbiologic studies, immunologic studies or chemical analyses may be requested on BAL specimens, and therefore, these specimens should not be fixed, unless those studies are requested separately and separate aliquots are provided for the ancillary studies a priori. Depending upon the local laboratory practice, BAL specimens may first be sampled by microbiologic laboratories, or for cell counting prior to submission to cytology. BALs should be processed as detailed in III.B.2. Slide preparations are often stained by the Papanicolaou stain, Romanowsky stain (to detect organisms) and a variety of special stains for the evaluation of specimens from immunosuppressed patients. Coordination and communication among multiple laboratory sections is important in the analysis of BAL specimens.

XI.C.e. Pleural Fluid

Pleural fluid specimens should be accessioned, processed and stained as outlined in section III, depending upon their volume, presence of fixation, and gross characteristics.

XI.C.f. Fine Needle Biopsy

Fine needle biopsies of the respiratory tract may be obtained by percutaneous transthoracic fine needle aspiration, or by endoscopically or bronchoscopically directed fine needle aspiration. The material may be received as direct smears, needle rinses in transport medium or entire specimens submitted in a proprietary transport medium supplied by the manufacturers of liquid based processors. These specimens should be processed and stained as indicated in section III.

XI.C.g. Pulmonary Microvascular Cytology

These specimens are rarely used for evaluation of pulmonary lymphatic carcinomatosis. Density gradient centrifugation is used to concentrate the nucleated cells from a heparinized sample of pulmonary artery blood.164 The buffy coat is used for smears or other preparations. Megakaryocytes indicate that the specimen is adequate for evaluation.165 Papanicolaou or Romanowsky stains may be used for morphologic assessment. Ancillary immunologic stains may be necessary to separate megakaryocytes from malignant cells.


Immunocytochemistry in respiratory tract cytopathology can be useful in supporting a diagnosis of cancer, subclassifying a tumor, or identification of an infectious agent. It can help in the identification or differentiation of:173

  1. Anaplastic/pleomorphic carcinoma
  2. Adenocarcinoma versus mesothelioma
  3. Neuroendocrine tumors versus other neoplasms
  4. Primary versus metastastic tumor
  5. Unusual primary tumors
  6. Microorganisms (e.g. Pneumocystis carinii)

Adopted by the ASC Executive Board, March 2, 2004


7     Naylor B.  Pleural, Peritoneal, and Pericardial Fluids. In Bibbo M(ed).  Comprehensive Cytopathology. Philadelphia: W.B. Saunders Company, 1997.

8     NCCLS.  Nongynecologic Cytologic Specimens:  Collection and Cytopreparatory Techniques; Approved Guideline. NCCLS document GP 23-A[ISBN 1-56238-380-9]. NCCLS, West Valley Road, Suite 14000, Wayne Pennsylvania 19087-1898 USA, 1999. 

9     Occupational Safety and Health Administration.  Bloodborne pathogens.  Standard 1910.1030.

103     Greenlee RT, Hill-Harmon MB, Murray T, et al: Cancer Statistics, 2001. CA Cancer J Clin 51: 15-36, 2001.

104     DeMay RM: The Art & Science of Cytopathology. Vol 2, Aspiration Cytology. Chapter 21: Lung. ASCP Press, Chicago, 1996.

105     U.S. Environmental Protection Agency: Health Effects of Passive Smoking: Assessment of Lung Cancer in Adults and Respiratory Disorder in Children. Report 6000690006A. U.S. Environmental Protection Agency, Washington, DC, 1990a.

106     U.S. Environmental Protection Agency: Environmental Tobacco Smoke: A Guide to Workplace Policies. Report 400/6-90/004. U.S. Environmental Protection Agency, Washington, DC, 1990b.

107     Alberg AJ, Samet JM.  Epidemiology of lung cancer. Chest. 2003;123(1 Suppl):21S-49S.

108     Brownson RC, Figgs LW, Caisley LE.  Epidemiology of environmental tobacco smoke exposure. Oncogene.  2002;21(48):7341-8.

109     Centers for Disease Control and Prevention, Office of Smoking and Health: Health Benefits of Smoking Cessation: A Report of the Surgeon General. Publication CDC90-8416. U.S. Government Printing Office, Washington, DC, 1990.

110     Auerbach O, Garfinkel L: The Changing Pattern of Lung Carcinoma. Cancer 68: 1973-1977, 1991.

111     Chyou P-H, Nomura AMY, Stemmermann GN: A Prospective Study of the Attributable Risk of Cancer Due to Cigarette Smoking. Am J Public Health 82: 37-40, 1992.

112     Greenberg SD, Fraire AE, Kinner BM, et al: Tumor Cell Type Versus Staging in the Prognosis of Carcinoma of the Lung. Pathol Annul 22(2): 387-405, 1987.

113     Jett JR, Cortese DA, Fontana RS: Lung Cancer: Current Concepts and Prospects. CA Cancer J Clin 33: 74-86, 1983.

114     Morton WE, Treyve EL: Histologic Differences in Occupational Risks of Lung Cancer Incidence. Am J Indust Med 3: 441-457, 1982.

115     Fielding JE: Smoking: Health Effects and Control. N Engl J Med 313: 491-561, 1985.

116     Garfinkel L, Silverberg E: Lung Cancer and Smoking Trends in the United States Over the Past 25 Years. CA Cancer J Clin 41: 137-145, 1991.

117     Hammond EC: Smoking in Relation to the Death Rates of One Million Men and Women. NCI Monogr 19: 127-204, 1966.

118     Linnoila I: Pathology of Non-Small Cell Lung Cancer: New Diagnostic Approaches. Hematol Oncol Clin North Am 4: 1027-1051, 1990.

119     Carney DN: Biology of Small-Cell Lung Cancer. Lancet 339: 843-846, 1992.

120     Whitesell PL, Drage CW: Occupational Lung Cancer. Mayo Clin Proc 68: 183-188, 1993.

121     Fontham ETH: Protective Dietary Factors and Lung Cancer. Int J Epidemiol, suppl 1: 19: S32-42, 1990.

122     Pershagen G, et al: Residential R