PART TWO
Diagnostic Cytology
Table 13.29 Comparison between Fine Needle Aspiration (FNA) and
Histologic Diagnosis in Primary and Secondary Neoplasms
of Lungs
Diagnostic
category
Number
of patients
Percentage of
patients
FNA and tissue
diagnostic for cancer
246
85.4
FNA negative for
cancer and tissue
diagnostic for cancer
40
13.9
FNA positive for
cancer and tissue
negative for cancer
2
0.7
Total
288
100.0
Reproduced with permission from Johnston WW: Cytologic correlations. In: Dail
DH, Hammar SP (eds)
Pulmonary Pathology.
New York: Springer-Verlag, 1987.
Fine-Needle Aspiration and Non-Neoplastic Lesions
of the Lung
Table 13.30 is illustrative of the extent to which FNAs that have
been diagnosed as negative for cancer contribute useful infor-
mation. As was previously noted in Table 13.29, needle aspi-
ration in our laboratory detected cancer in more than 85% of
histologically confirmed tumors. The primary usefulness, then,
of the negative aspirate, is that it gives a reasonable level of con-
fidence that the patient does not have cancer. Unfortunately, it
does not conclusively preclude cancer.
Our series included 636 patients in whom no malignant neo-
plasms were seen and in whom no subsequent studies revealed
malignant neoplasms (see Table 13.30). This patient group was
further divided into the following diagnostic categories: nega-
tive for cancer and without cellular evidence of inflammation
or infectious agent, negative for cancer with nonspecific inflam-
mation, and negative for cancer with evidence of an infectious
organism or specific type of inflammatory process. In two
patients, coin lesions considered negative by aspiration biopsy
were shown on resection to be hamartomas.550-554 On review of
the preparations, isolated fragments of normal-appearing carti-
lage provided evidence of the hamartomas. Kumar and associates
have reported a bronchogenic cyst diagnosed by FNA.555
The third category, comprising 57 patients, is of particular
interest here because of the highly specific diagnostic informa-
tion gained from the lung aspiration. In many of those cases
in which an infectious organism was identified, it was immedi-
ately apparent on the toluidine blue-stained specimen. When
the presence of such an organism was noticed, the radiologist
would then make a second needle pass into the lung and submit
the aspirate for culture. The morphology of these organisms has
been described in a prior section of this chapter.
We previously emphasized the importance of cytologic meth-
ods in the evaluation of patients without lung cancer. Of patients in
this category, 9.0% benefited from the FNA diagnosis, which either
detected an infectious organism or recognized a morphologic
manifestation of a specific type of inflammation. Other laborato-
ries have reported favorable experiences with aspiration in the diag-
noses of non-neoplastic and infectious diseases of the lungs.556-563
In a review of 73 consecutive cases of benign lung disease
in which both fine-needle aspiration and core biopsy had been
Table 13.30 Distribution of Fine Needle Aspiration Patients without Lung
Neoplasm
Diagnostic
Number
Percentage of
category
of patients
patients
Negative for cancer and
without inflammation
489
76.9
Negative for cancer
90
14.1
and inflammation,
nonspecific
Inflammation, specific
Bacteria (7)
Tuberculosis (1)
Nocardiosis (2)
Blastomycosis (3)
Cryptococcosis (6)
Histoplasmosis (3)
Candidiasis (2)
Aspergillosis (8)
Phycomycosis (3)
Granuloma (18)
Abscess (3)
Oxalate crystals (1)
57
9.0
Total
636
100.0
Reproduced with permission from Johnston WW: Cytologic correlations. In: Dail
DH, Hammar SP (eds)
Pulmonary Pathology.
New York: Springer-Verlag, 1987.
performed, Greif and associates found that a specific diagnosis
had been made in 16.7% of the aspirates and in 81.7% of the
core biopsies. They concluded that core biopsy is a safe proce-
dure and was able to provide sufficient histologic material for a
specific diagnosis of peripheral lung disease.564
Correlation between Fine-Needle Aspiration
and Histology
A number of investigations have studied the correlations
between FNA predictive of tumor classification and subsequent
histologic diagnosis of the malignant tumor that had been aspi-
rated. Dahlgren in 1967 reported correlations in the range of
67-81%.565 Sinner recorded ranges of 90-100%.566 In all of the
aforementioned studies, the highest level of predictive accu-
racy was for squamous cell carcinoma. Taft and associates, in
1980, reported 70% correlation;567 Poe and Tobin also in 1980
recorded an overall agreement level of 71%.568 Thornbury and
associates reported in 1981 an overall cytologichistologic cor-
relation level of 86%.378 Bonfiglio's later study reported that
cytologic diagnosis of adenocarcinoma correlated with tissue
diagnosis in 87% of cases, squamous cell carcinoma in 77%,
and large-cell carcinoma in 62%.569
In a retrospective investigation, Kyeom and associates com-
pared in 296 patients the usefulness of transthoracic FNA,
core biopsy, and a combination of the two. They found that
in the malignant tumor group, FNA was diagnostically help-
ful in 91.7% of the patients, and the core biopsy in 87.8%. The
358
previous page 355 ComprehensiveCytopathology 1104p 2008 read online next page 357 ComprehensiveCytopathology 1104p 2008 read online Home Toggle text on/off