Printer Friendly

Granulomatous inflammation--an underestimated cause of false-positive diagnoses in lung fine-needle aspirates: observations from the College of American Pathologists Nongynecologic Cytopathology Interlaboratory Comparison Program.

Although very accurate, lung fine-needle aspiration (FNA) is subject to diagnostic pitfalls like any other diagnostic technique. While the single largest contributing factor to false-negative diagnoses is sampling error resulting from either poor technique or inability to aspirate the lesion because of its small size or of its inaccessibility, (1,2) false-positive diagnoses are often caused by benign lesions that either induce marked reactive atypia and/or exhibit cohesive groups mimicking carcinoma. (3-18) Granulomatous inflammation is one of the most common known sources of false-positive diagnoses in lung FNA. (9,11,14,16,17)

The College of American Pathologists (CAP) Nongynecologic Cytopathology (NGC) Interlaboratory Comparison Program is an educational program consisting of 5 glass slides (case challenges) with accompanying clinical history. The slide sets are circulated 4 times a year to participating laboratories. A diagnostic list/menu is provided from which a general category (unsatisfactory, benign, suspicious, or malignant) must be chosen by participants. A specific reference diagnosis is also chosen by the participant. The data for each challenge are compiled by the CAP and specific feedback is sent to each participant and participating laboratory, allowing comparison of performance to that of other laboratories. The CAP NGC Program includes lung FNA cases whose reference diagnosis is "specific infections/granulomatous inflammation."

We retrospectively reviewed the cumulative data of the CAP NGC Program from 1998 to 2008 from those challenges with the reference diagnosis "specific infections/granulomatous inflammation" to determine the accuracy of FNA diagnosis when granulomatous inflammation was present and to determine the most common misinterpretation patterns.

MATERIALS AND METHODS

The cumulative data from the CAP NGC Program participants' responses to all lung FNA cases of granulomatous inflammation from 1998 to 2008 were obtained through the CAP SCORES (CAP, Northfield, Illinois) computer system.

There were a total of 1092 participant responses, of which 502 were from pathologists available for 44 lung FNA challenges, with a reference diagnosis of "specific infections/granulomatous inflammation." False-positive responses were defined as a "malignant" or "suspicious" general category response, chosen for the cases in which the specific reference diagnosis was "specific infections/granulomatous inflammation." True negative responses were those cases of "specific infections/granulomatous inflammation" for which "benign" was chosen as a general reference category. The specific reference diagnosis was analyzed and compared to other specific reference categories available for lung FNA. False-positive rates by participant type (pathologist versus cytotechnologist) and the general category, reference diagnosis, and preparation type (Papanicolaou-stained smears, modified Giemsa-stained smears, or cytocentrifuged preparations) were analyzed for the pathologists' responses in order to determine the frequency distribution of the pathologists' responses and the rates of false-positive responses for these lesions according to participant and preparation types.

Fisher exact test was used to evaluate the association between the false-positive response rate and slide preparation type. For statistical analysis, the overall false-positive rate was calculated with "suspicious" cases classified as false positives. All tests were run at the .05 significance level. All statistical analyses were performed with SAS version 9.1 (SAS Institute Inc, Cary, North Carolina).

RESULTS

Of the 502 pathologist responses from the cumulative 10-year review of the CAP NGC Program responses to lung FNA challenge cases with a reference diagnosis of "specific infections/granulomatous inflammation," 85.3% (428) were benign (true negative), 11% (55) were malignant, and 3.8% (19) were suspicious (Table 1). There were specific reference category diagnoses provided by 499 pathologists for the category "specific infections/granulomatous inflammation"; the distribution of benign and malignant specific reference diagnoses for the granulomatous inflammation challenges are shown in Tables 2 and 3, respectively. Of those responses that were suspicious or malignant (ie, false-positive responses), non-small cell carcinoma, adenocarcinoma, and squamous carcinoma accounted for 64% of the responses, while small cell carcinoma and carcinoid comprised 13%. Of those responses that were benign, only 68.7% were an exact match to the reference diagnosis of granulomatous inflammation. The remainder of the benign responses was "infectious/inflammatory" and "normal/reactive" categories, with rare answers in a variety of other benign categories.

There was no statistical difference in the false-positive response rate between Papanicolaou-stained smears and modified Giemsa-stained smears (P = .76). There was a significant difference in the performance of cytocentrifuged samples, with no false positives seen in the cytocentrifuged challenges (Table 4). There were no differences between the false-positive rates of cytotechnologists versus pathologists (P = .39) (Table 5).

COMMENT

Lung FNA is a well-established method for the investigation of various lung lesions and has been shown to have a high degree of accuracy in experienced hands. Because the positive predictive value of lung FNA ranges from 96% to 100%, (19) a patient with a lung fine-needle aspirate diagnosed as malignant has a high probability of harboring a malignancy. Nevertheless, like any other diagnostic tests, lung FNA is subject to diagnostic pitfalls that can lead to occasional false-positive diagnoses; the false-positive rate of lung FNA has historically been cited as being less than 1%. (19)

Causes of false-positive diagnoses in lung FNA include granulomatous inflammation, thermal/irradiation/chemotherapy effects, lung abscess, pulmonary infarct, diffuse alveolar damage, organizing pneumonia, and pulmonary fibrosis. (1,3-17) All these conditions are associated with reactive atypia of the surrounding alveolar, bronchial, or squamous metaplastic epithelium; this atypia may lead to a malignant diagnosis. Atypia of surrounding cellular elements in the presence of a lung mass or a clinical impression of malignancy may further bias the observer to render a false-positive interpretation in clinical practice. An example of mild cellular atypia seen at the periphery of a granuloma is shown in Figure 1, an image from a challenge case in the CAP NGC Program; this case was misinterpreted as malignant by some participants.

Another factor contributing to false-positive diagnoses is the presence of cohesive cellular groups, which may lead to an erroneous interpretation of carcinoma. This factor is known to play a significant role in the frequent misdiagnoses of pulmonary hamartoma (18) and in rare cases of pulmonary sequestration (15); in both lesions, the culprit is the abundant benign cohesive epithelium often misinterpreted as well-differentiated adenocarcinoma. Analogously, it appears that the perception of cohesive elements (not from an epithelial component, but rather from the granulomata themselves) may lead to the false-positive responses in the context of granulomatous inflammation in lung FNA (Figure 2, A and B). This factor may be more significant than the induced reactive atypia as a cause of false-positive diagnoses in the context of granulomatous inflammation, at least in this retrospective series. All the cases circulated in the CAP NGC Inter-laboratory Comparison Program go through a rigorous review by 2 cytopathologists, who agree that the cases represent classic examples of a specific entity (in this instance, granulomatous inflammation), before acceptance in the program; it is unlikely that the circulated cases demonstrate any significant confounding reactive atypia.

The key diagnostic feature of granulomata in FNA material is the absence of a monophasic population in the 3-dimensional clusters; granulomata are formed of at least 2 cell types (ie, epithelioid and spindled histiocytes) and are often admixed with lymphocytes and/or multinucleated giant cells. Although multinucleated macrophages can be present, they are often absent and are not essential for a diagnosis of granulomatous inflammation. Lymphocytes are usually also seen in the background; however, acute inflammation may be prominent depending on the etiology of the granulomatous inflammation.

Interestingly, in this retrospective study, even among those who recognized the challenge as harboring a benign diagnosis, only 68.7% identified granulomatous inflammation specifically as such. Failure to recognize granulomatous inflammation may lead to potential unnecessary investigations (including repeated procedures); indeed, a specific cytologic diagnosis of granulomatous inflammation will usually trigger a workup that differs from that of a nonspecific benign diagnosis.

Although the false-positive rate has been historically cited as being approximately 1% in lung FNA, (19) the rate of false-positive response due to granulomatous inflammation is 14.8% in this retrospective series. It is important to remember that the CAP NGC Program is not equivalent to clinical practice. Participants in the CAP NGC Program might not interpret challenge cases with the same frame of mind as they might in actual patient cases. Although purely educational, the CAP NGC Program may put unusual pressure on the participant to perform, in particular to be more cautious not to miss a malignant diagnosis (test bias). Program participants expecting malignant findings may have been bolder in their diagnoses in the context of an educational program as compared to real-life practice; this has been speculated as being a factor in a previous CAP NGC Program related to liver, kidney, pancreas, and salivary gland fine-needle aspirates in which a significant percentage of responses included a variety of neoplastic diagnoses when confronted with normal cellular elements. (20) The slides used in the CAP NGC Program originate from different laboratories, resulting in varied staining techniques and preparations, which may affect individual performance as well. Participants come from a variety of backgrounds and experience, including cytotechnologists and pathologists with different training experience and practices. In the CAP NGC Program, cytotechnologists and pathologists must arrive at a diagnosis based upon a single slide and history, without the aid of ancillary morphologic studies, additional clinical consultation, or radiologic studies. It is artificial and represents a strictly morphologic approach to cytology. It is understood that the evaluation of these slides does not mimic daily practice and represents an educational challenge. While the CAP NGC Program does not directly mimic clinical practice, the diagnosis of granulomatous inflammation is usually based upon initial morphologic impressions. In clinical practice, other sources of useful information, (eg, review of imaging studies, consultation with clinicians) are available to help resolve challenging cases as compared to the limited clinical information provided in the context of the CAP NGC Program.

[FIGURE 1 OMITTED]

[FIGURE 2 OMITTED]

In daily practice, granulomatous inflammation is usually not seen as often as malignant lesions and the frequency of encountering granulomatous lung inflammation varies with the geographic location, the population, and the prevalence of its causative agents. The infrequent encounter of granulomatous inflammation in clinical practice may contribute to the difficulty of recognizing this entity prospectively. Importantly, neither slide preparation type (with the exception of cytocentrifuged preparations, which most likely represent a small sample bias) nor participant type (ie, pathologist versus cytotechnologist) significantly affected the recognition of granulomatous inflammation in lung FNA.

Awareness that granulomatous inflammation is an important and potential source of false-positive diagnoses in lung FNA is crucial to avoid a false-positive diagnosis. Cytotechnologists and pathologists should be aware that the atypia associated with and the cohesive nature of granulomatous inflammation lead to false-positive responses in an educational interlaboratory comparison program, and by extrapolation, may lead to false-positive diagnoses in clinical practice.

References

(1.) Cagle PT, Kovach M, Ramzy I. Causes of false results in transthoracic fine needle lung aspirates. Acta Cytol. 1993;37(1):16-20.

(2.) Zakowski MF, Gatscha RM, Zaman MB. Negative predictive value of pulmonary fine needle aspiration cytology. Acta Cytol. 1992;36(3):283-286.

(3.) Sinner WN. Transthoracic needle biopsy of small peripheral malignant lung lesions. Invest Radiol. 1973;8(5):305-314.

(4.) Sinner WN. Pulmonary neoplasms diagnosed with transthoracic needle biopsy. Cancer. 1979;43(4):1533-1540.

(5.) Westcott JL. Direct percutaneous needle aspiration of localized pulmonary lesions: results in 422 patients. Radiology. 1980;137(1):31-35.

(6.) Michel RP, Lushpihan A, Ahmed MN. Pathologic findings of transthoracic needle aspiration in the diagnosis of localized pulmonary lesions. Cancer. 1983; 51(9):1663-1672.

(7.) Kato H, Konaka C, Kawate N, et al. Percutaneous fine-needle cytology for lung cancer diagnosis. Diagn Cytopathol. 1986;2(4):277-283.

(8.) Zaman MB, Hajdu SI, Melamed MR, et al. Transthoracic aspiration cytology of pulmonary lesions. Semin Diagn Pathol. 1986;3:176-187.

(9.) Pitman MB, Szyfelbein WM, Niles J, Fienberg R. Clinical utility of fine needle aspiration biopsy in the diagnosis of Wegener's granulomatosis. Acta Cytol. 1992;36(2):222-229.

(10.) Johnston WW. Type II pneumocytes in cytologic specimens: a diagnostic dilemma. Am J Clin Pathol. 1992;97(5):608-609.

(11.) Naryshkin S, Young NA. Respiratory cytology: a review of non-neoplastic mimics of malignancy. Diagn Cytopathol. 1993;9(1):89-97.

(12.) Hayes MMM, Zhang DY, Brown W. Transthoracic fine-needle aspiration biopsy cytology of pulmonary neoplasms. Diagn Cytopathol. 1994;10(4):315-319.

(13.) Bocking A, Klose KC, Kyll HJ, Hauptmann S. Cytologic versus histologic evaluation of needle biopsy of the lung, hilum and mediastinum: sensitivity, specificity and typing accuracy. Acta Cytol. 1995;39(3):463-471.

(14.) Mourad WA, Vallieres E, Power RF, Hirji M. Fine-needle aspiration cytology of bronchogenic granulomatosis: a potential diagnostic pitfall. Diagn Cytopathol. 1996;14(3):263-267.

(15.) Matthew S, Erozan YS. Pulmonary sequestration--a diagnostic pitfall: a case report. Diagn Cytopathol. 1997;16(4):353-357.

(16.) Crapanzano JP, Zakowski MF. Diagnostic dilemmas in pulmonary cytology. Cancer. 2001;93(6):364-375.

(17.) Williamson JD, Murphree SS, Wills-Frank L. Atypical squamous cells as a diagnostic pitfall in pulmonary Wegener's granulomatosis: a case report. Acta Cytol. 2002;46(3):571-576.

(18.) Hughes JH, Young NA, Wilbur DC, Renshaw AA, Mody DR. Fine-needle aspiration of pulmonary hamartoma: a common source of false-positive diagnoses in the College of American Pathologists Interlaboratory Comparison Program in Nongynecologic Cytology. Arch Pathol Lab Med. 2005;129(1):19-22.

(19.) Zarbo RJ, Fenoglio-Preiser CM. Interinstitutional database for comparison of performance in lung fine-needle aspiration cytology: College of American Pathologists Q-Probe study of 5264 cases with histologic correlation. Arch Pathol Lab Med. 1992;116(5):463-470.

(20.) Young NA, Mody DR, Davey DD. Misinterpretation of normal cellular elements in fine-needle aspiration biopsy specimens: observations from the College of American Pathologists Interlaboratory Comparaison Program in Non Gynecologic Cytopathology. Arch Pathol Lab Med. 2002;126(6):670-675.

Manon Auger, MD; Ann T. Moriarty, MD; Rodolfo Laucirica, MD; Rhona Souers, MS; Beth Anne Chmara, BS, CT(ASCP); Lisa A. Fatheree, BS, SCT(ASCP); David C. Wilbur, MD

Accepted for publication May 12, 2010.

From the Department of Pathology, McGill University Health Center and McGill University, Montreal, Quebec, Canada (Dr Auger); the Department of Cytology, AmeriPath Indiana, Indianapolis (Dr Moriarty); the Department of Pathology, Baylor College of Medicine, Houston, Texas (Dr Laucirica); Staff Biostatistics (Ms Souers) and the Department of Cytology Surveys (Mses Chmara and Fatheree), College of American Pathologists, Northfield, Illinois; and the Department of Pathology, Massachusetts General Hospital, Boston (Dr Wilbur).

The authors have no relevant financial interest in the products or companies described in this article.

Reprints: Manon Auger, MD, Department of Pathology, McGill University, 3775 University St, Room 105, Montreal, QC H3A 2B4, Canada (e-mail: manon.auger@mcgill.ca).
Table 1. Distribution of Pathologist General
Diagnosis Responses (N = 502)

General Diagnosis Frequency (%)

 Benign 428 (85.3)
 Malignant 55 (11.0)
 Suspicious 19 (3.8)

Table 2. Distribution of Specific Reference Diagnoses
by Pathologists Responding With the General Category
"Benign" (N = 425)

 Benign General Diagnosis,
 Reference Diagnosis Frequency (%)

Specific infections/granulomatous
 inflammation 292 (68.7)
Infectious/inflammatory 86 (20.2)
Normal/reactive 26 (6.1)
Benign lesions, NOS 8 (1.9)
Hamartoma 5 (1.2)
Invalid 2 (0.5)
Unsatisfactory 3 (0.7)
Others 3 (0.7)

Abbreviation: NOS, not otherwise specified.

Table 3. Distribution of Specific Reference Diagnoses
by Pathologists Responding With the General Category
"Malignant or Suspicious" (N = 74)

 General Category Response

 Positive, Suspicious,
 Frequency Frequency

Reference Diagnosis (%) (%)

Non-small cell carcinoma 13 (23.6) 4 (21.1)
Adenocarcinoma (including
 bronchiolalveolar) 14 (25.5) 2 (10.5)
Squamous cell carcinoma 11 (20.0) 5 (26.3)
Small cell undifferentiated
 carcinoma 5 (9.1) 1 (5.3)
Carcinoid 0 (0.0) 4 (21.1)
Sarcoma, NOS 2 (3.6) 1 (5.3)
Metastatic carcinoma, NOS 4 (7.3) 0 (0)
Unsatisfactory 0 (0) 1 (5.3)
Lymphoma/hematopoietic
 malignancy 2 (3.6) 0 (0)
Malignant lymphoma (mixed type) 1 (1.8) 0 (0.0)
Other malignancy 1 (1.8) 0 (0)
Invalid 2 (3.6) 1 (5.3)

Total 55 (100) 19 (100)

Abbreviation: NOS, not otherwise specified.

Table 4. False-Positive Rate by Preparation Type (N = 494) (a)

 Preparation Type
 Smears
 Papanicolaou Modified Giemsa Cytocentrifuged
 (N = 323) (N = 139) Specimen (N = 32)

False-positive
 rate, % 15.5 14.4 0.0

(a) Fisher exact test: P = .03 (when comparing the 3 techniques) and
P = .76 when comparing conventional versus Diff-Quik (Siemens
[formerly Dade Behring], Newark, Delaware) only.

Table 5. False-Positive Rate by Participant (N = 851) (a)

 Participant Type
 Cytotechnologist Pathologist
 (N = 349) (N = 502)

False-positive rate, % 16.9 14.8

(a) P value = .39.
COPYRIGHT 2010 College of American Pathologists
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2010 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:CAP Laboratory Improvement Programs
Author:Auger, Manon; Moriarty, Ann T.; Laucirica, Rodolfo; Souers, Rhona; Chmara, Beth Anne; Fatheree, Lisa
Publication:Archives of Pathology & Laboratory Medicine
Article Type:Clinical report
Date:Dec 1, 2010
Words:2723
Previous Article:Attributes of antiangiogenic factor plasminogen kringle 5 in glomerulonephritis.
Next Article:Sentinel lymph nodes in cutaneous melanoma: handling, examination, and clinical repercussion.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters