Diagnostic Cytology
Fig. 13.120 Metastatic transitional cell carcinoma. Bronchial brushing
(Papanicolaou x OI).
Table 13.18 summarizes our experience with such patients.
Metastatic neoplasms have represented 15% of the total patient
population examined by FNA of the lung in our laboratory. With
reference to the type of neoplasm, tissue, or organ system of ori-
gin, malignant melanoma was seen in 26.8%; neoplasms from
the urinary and male genital tract in 17.2%; adenocarcinoma
from the breast in 14.9%; neoplasms, usually squamous cell
carcinomas, from the female genital tract in 12.6%; adenocarci-
nomas from the gastrointestinal tract in 10.0%; and neoplasms
from the bones and soft tissues in 8.4%. In terms of histologic
type, the six most common metastatic neoplasms were malignant
melanoma, adenocarcinoma of the breast, adenocarcinoma of
the colon, transitional cell carcinoma of the bladder, squamous
cell carcinoma of the cervix, and adenocarcinoma of the kidney.
The large number of melanomas in this series is a specific exam-
ple of the significant role that needle aspiration may have in
the therapeutic management of patients with cancer metastatic
to the lung. An example of metastatic transitional carcinoma as
seen in an FNA of the lung is depicted in Fig. 13.120.
A marked difference is noted between the cytologic presenta-
tion of neoplasms metastatic to the lung in FNAs and in speci-
mens of sputum and bronchial material. This difference lies
mainly in the amount of cellular material available for study.
Most metastatic masses when aspirated yield large numbers of
tumor cells and tissue fragments. Extensive necrosis and inflam-
mation may also be present and may partially obscure the
tumor. Once a cancer diagnosis is established, the major diag-
nostic decision is determination of metastatic cancer versus pri-
mary lung cancer. Paraffin sections of cell blocks become useful
when special stains and immunocytochemistry are needed for
differential studies. Tissue fragments fixed in glutaraldehyde are
extremely valuable for electron microscopy.
Other metastases reported in the literature as diagnosed by
FNA have included lymphoma,488 malignant schwannoma,489,490
adenoid cystic carcinoma,491,492 ameloblastoma,493 malignant
fibrous histiocytoma,494-496 thymoma,497 dermatofibrosarcoma
protuberans,498 synovial sarcoma,499 giant cell tumor of bone,500
malignant phyllodes tumor,503 and tibial adamantinoma.504
In a 1986 study, Kim and associates reported the morpho-
logic features of cells aspirated from 17 sarcomas metastatic
to the lung.505 The group included five malignant fibrous his-
tiocytomas, three fibrosarcomas, three leiomyosarcomas, three
endometrial stromal sarcomas, one osteosarcoma, and two
poorly differentiated sarcomas.
In their report of metastases to the lung from osteogenic sar-
coma as found in fine-needle aspirates from four cases, Dodd
and associates stressed the diagnostic importance of the pres-
ence of matrix and malignant osteoblastlike cells.506
Diagnostic Accuracy
In the writing on clinical laboratory tests, when expressing pre-
dictive accuracy, it has become fashionable to use the terms
as introduced by Galen and Gambino.507
Sensitivity is the mathematic expression of the ability of a clini-
cal test to detect disease in a diseased population. Specificity is
the mathematic expression for the ability of a test to refrain from
falsely diagnosing disease in a nondiseased population.508 Sensi-
tivity is a measure of the percentage of known diseased patients
with positive test results among all the diseased patients evalu-
ated. The specificity is the percentage of patients with negative
test results among all known patients without disease tested.
Expressed mathematically,
Sensitivity = TP/(TP + FN)
Specificity = TN/(TN + FP),
where TP = true positives; FN = false negatives; TN = true negatives,
and FP = false positives.
In diagnostic cytology, these equations become one means of
expressing the false-negative and false-positive rates. Although
useful, they are only a partial expression of the broader concept
of diagnostic accuracy in cytology. Such accuracy embraces not
only the assessment of presence or absence of cancer in a given
specimen but also prediction of cancer differentiation and iden-
tification of benign disease states. This section of the chapter dis-
cusses these various facets of accuracy when applied to sputum,
bronchial material, and FNA. The complicated and difficult
"gray" area where cancer can neither be diagnosed conclusively
nor precluded was discussed in a preceding section.
Sensitivity and Specificity of Specimens of Sputum
and Bronchial Material
Studies in the literature document the level of accuracy that may
be achieved in the detection and classification of lung neoplasms
through the use of sputum, bronchial washings, and bronchial
brushings.30,31,509-533 Some of these are summarized in Table
13.19. In 1964, Koss and his associates273 reported on their study
of 149 patients who had histologically proven lung cancer and
on whom three or more satisfactory sputum specimens had been
examined, the largest series that had been published up until that
time. The overall accuracy (sensitivity) of cytology in detecting
the presence of the tumor was 89% when three or more cyto-
logic specimens of sputum were examined. Their specificity was
99.8%. Two false-positive diagnoses were recorded.
The importance of examining several specimens was also
studied by Erozan and Frost in 1970.534 Among their patients
with lung cancer, one bronchoscopic examination yielded diag-
nostic cytologic results in 61%, whereas one sputum specimen
yielded diagnostic cytologic results in only 42%. Diagnostic
yield, however, increased to 82% with three sputum examina-
tions and to 91% with five. Bedrossian and his associates in
previous page 349 ComprehensiveCytopathology 1104p 2008 read online next page 351 ComprehensiveCytopathology 1104p 2008 read online Home Toggle text on/off