GYNECOLOGIC
ONCOLOGY
46,
309-312 (1992)
Fine-Needle Aspiration Cytology in Patients with Gynecologic Malignancies’ MASASHI IMACHI, M.D., NAOKI TSUKAMOTO, M.D., TOSHIYUKI SHIGEMATSU,M.D., TOSHIAKI SAITO, M.D., TOSHIHARU KAMURA, M.D., AND HITOO NAKANO, M.D. Department of Gynecology
and Obstetrics, Faculty of Medicine,
Kyushu University,
Maidashi 3-1-1,
Higashi-ku,
Fukuoka 812, Japan
Received August 5, 1991
Four hundred five fine-needle aspiration (FNA) cytologies were obtained from 352 sites in 287 patients with gynecologic malignancies. The majority of specimens were aspirated for recurrent or metastatic disease. The most common clinical diagnosis was cervical carcinoma (128 cases)followed by ovarian carcinoma (80 cases)and others (79 cases). The sites of FNA were lymph node (134 cases), retroperitoneal Iymphocyst (57 cases), pelvic mass (52 cases), subcutaneous mass (34 cases), and others (75 cases). Of335 sites excluding inadequate specimens, 162 (48.4%) revealed malignant cells. There was no difference in the accuracy of FNA between diseases in the superficial sites and those in the deep sites. However, the sensitivity for local recurrent diseases was lower than that for primary or metastatic d&apes (86.4% vs 100.0%). The overall accuracy of FNA cytology was 95.2%, and it was satisfactory for the diagnosis of gynecologic malignant diseases.FNA should be repeated if the initial FNA specimen is inadequate for diagnosis. When distinct malignant cells are observedin FNA cytology, a biopsy may be omitted. o MJZ Academic Press, Inc.
INTRODUCTION
Fine-needle aspiration (FNA) was first described by Guthrie [l] in 1921, followed by Stewart [2] in 1933, and Martin and Ellis [3] in 1934. A FNA cytology is technically much easier to perform and less harmful to the patient than an open biopsy. FNA cytologies have thus been performed in organs for which open biopsies are anatomically difficult to perform. To date, FNA cytology has been used for various nongynecologic sites such as the breast, prostate, lung, liver, and pancreas [4-71. For gynecologic sites, FNA cytology was first utilized for the diagnosis of primary ovarian tumor [g-11]. However, its usage has remained contro’ Supported in part by a Grant-in-Aid (62480349) for general scientific research from the Ministry of Education, Science and Culture of Japan.
versial because of the possibility of tumor spread [12, 131. A few studies have focused on the usage of FNA cytology to diagnose palpable mass and intraabdominal mass in gynecologic malignancies [ 14-181. This report is the result of a study of FNA cytologies from 352 sites in 287 patients with gynecologic malignancies. MATERIALS
AND METHODS
From January 1982 through March 1991, 405 FNAs were obtained from 352 sites in 287 patients with gynecologic malignancies at Kyushu University Hospital. The clinical giagnoses of these patients are listed in Table 1. Table 2 shows the FNA sites. The most frequent site was lymph nodes followed by retroperitoneal lymphocysts, pelvic masses, subcutaneous masses, and others. FNA cytologies df superficial masseswere obtained using a 22-gauge needle attached to a lo-cc or 20-cc plastic syringe under local anesthesia. Deep aspirates were performed with a 23-gauge or 22-gauge spinal needle, guided by ultrasonography. The needle was inserted into the mass and negative pressure was generated by retracting the plunger. The suction was released before the needle was withdrawn. The aspirated material within the needle was spread over clean glass slides which were immediately fixed in alcohol and stained by the standard Papanicolaou technique. If any amount of fluid was aspirated, the fluid was centrifuged and the sediment was smeared between two slides. The specimens were divided into three groups: positive, benign, and inadequate. The positive specimens revealed some malignant cells. The benign specimens contained some normal cells derived from the target organ, but no malignant cells. When there were no cells derived from the target organ, the specimen was evaluated as inadequate FNA. FNA was repeated in patients whose spec-
309 0090-82%X/92 $4.00
Copyright 0 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.
310
IMACHI ET AL.
TABLE 1 Clinical Diagnosis
TABLE 3 Resultsof FNAs, Histologic Findings, and Clinical Findings
Diagnosis
Number of cases
%
Cervical carcinoma Ovarian carcinoma Endometrial carcinoma Vulvar carcinoma Vaginal carcinoma Uterine sarcoma Tubal carcinoma Other malignancies
128 80 26 20 6 5 4 18
44.6 27.9 9.1 7.0 2.1 1.7 1.4 6.3
Total
287
100
imens were inadequate. When cells derived from the target organ were not observed even in repeated aspirates, with a mean of 3.8 aspirates, those lesions were excluded from the data of the present study. When a lesion was stable or regressive for more than 12 months, the lesion was deemed clinically benign. A lesion showing distinct progress within 6 months after FNA cytology was deemed clinically malignant, even when there was no histological evidence of malignancy. All FNAs were performed in association with a biopsy, surgery, or a complete clinical follow-up. FNA cytology was considered to be false positive (FP) when cytologically malignant cells were observed despite the lesion being histologically benign, or showing a clinically benign course. FNA cytology was deemed false negative (FN) when cytologically malignant cells were not observed despite the lesion being histologically malignant, or showing a clinically malignant course. A true positive FNA (TP) was a positive cytology for a lesion showing histologically malignant cells or a clinically malignant appearance. A true negative FNA (TN) TABLE 2 Sitesof Fine-NeedleAspirations Site
Number of FNAs
Lymph node Retroperitoneal lymphocyst Pelvic mass Subcutaneous mass Vulvar mass Parametrium Vaginal mass Intraabdominal mass Retroperitoneal mass Umbilicus Kidney Liver Others
134 57 52 34 20 15 14 9 5 4 3 2 3
Total
352*( 17)
Number of positive FNAs 69 3 31 22 8 6 5 7 2 4 1 1 3 162
Note. *, including inadequate FNAs; ( ), inadequate FNAs.
Histologic/clinical finding Result of FNA 162 173
Malignant
Benign
160 6
2 167
positive negative
was a negative cytology for a lesion yielding a benign biopsy or showing a clinically benign course. The diagnostic sensitivity and specificity were defined as follows: sensitivity = TP/TP + FN; specificity = TN/TN + FP; predictive value of a positive result = TP/TF + FP; predictive value of a negative result = TN/TN + FN. The x2 test was used for statistical analysis. RESULTS Of the total 405 FNAs, 162 FNAs (40.0%) revealed malignant cells (i.e., positive), 179 (44.2%) were benign, and 64 (15.8%) were inadequate for diagnosis. Repeat FNAs were performed for the 64 lesions which yielded inadequate specimens. Finally, of 352 sites, 162 (46.0%) were positive, 173 (49.1%) were benign, and 17 (4.8%) were inadequate. Table 2 shows the sites of positive FNAs. Lymph nodes were the most frequent positive sites followed by pelvic masses, subcutaneous masses, and others. Of the 162 positive FNAs, histological specimens were obtained for 81 (50.0%) lesions. Fifty-six biopsies and 25 surgeries were performed, and 77 of them yielded malignant histological findings. The remaining 81 lesions with a positive FNA did not have histological confirmation, but malignancy was confirmed by the clinical course. Of the 173 negative FNAs, histological specimens were obtained for only 38 (22.0%) lesions. However, the remaining 135 lesions with a negative FNA were subjected to a complete clinical follow-up. There were four reasons for which biopsies were not performed for some lesions with a positive FNA. That is, a biopsy was anatomically difficult for 37 (34.9%) lesions, reductions surgery was directly performed without a biopsy for 25 (23.6%) lesions, biopsy was omitted because of the distinct clinical findings of malignancy for 23 (21.7%) lesions, and the general condition of the patients was too poor to perform biopsy for 21 (19.8%) lesions. The sites of FNA were divided into two groups; namely, superficial sites including superficial lymph nodes, subcutaneous masses,vulvar massesand vaginal masses,and deep sites including pelvic masses, intraabdominal masses, parametrium, retroperitoneal masses, deep lymph nodes, and others. Biopsy was performed for 49.5% of the lesions in superficial sites, but for only 8.5%
FINE-NEEDLE
ASPIRATION
CYTOLOGY
TABLE4 Sensitivity of FNA for RecurrentLocal Lesionsand Primary/MetastaGcLesions
Recurrent local lesions Primary/metastatic lesions
Sensitivity(%)
Cases
86.4* 100.0*
38/44 1221122
*, P < 0.001.
of lesions in deep sites. The incidence of reduction surgery without biopsy was higher for lesions in deep sites than for lesions in superficial sites (28.8% vs 7.8%). However, there was no difference between the accuracy of FNA for the superficial sites and for the deep sites. Table 3 summarizes the results of the FNAs, histologic findings, and clinical findings. The sensitivity for 335 sites was 96.4%) the specificity was 98.8%) the predictive value of a positive result was 98.8%, and the predictive value of a negative result was 96.5%. The sensitivities for parametrium (60.0%), retroperitoneal masses (66.7%), and vaginal masses (83.3%) were lower than those for other organs (99.2%). There were six (1.8%) false negative FNAs. The sites of those lesions were three parametriums, one vulvar mass, one pelvic mass, and one vaginal mass. On the other hand, there were two false-positive FNAs. One site was a small massin the cul-de-sac of a patient with ovarian carcinoma before a second-look operation, and the other site was left inguinal lymph node in a patient with ovarian carcinoma. In these two patients, the histological findings were negative for malignancy, and there was no clinical evidence of disease. In both specimens, we had recognized some histiocytes as malignant cells, and re-evaluation yielded no malignant findings. Of the total of 166 malignant sites, 122 sites were primary lesions or metastatic lesions, and 44 sites were local recurrent lesions. The sensitivity for local recurrent lesions was statistically lower than that for primary or metastatic lesions (86.4% vs 100%) (P < 0.001) (Table 4). Complications occurred with 4 of the 405 FNAs, and all were intraabdominal masses or pelvic masses. The complications consisted of 2 aspirations of the intestines, 1 aspiration of the bladder, and 1 aspiration of an artery. These 4 complications did not cause hemorrhage, abscess, or fistula, and they did not disrupt the treatment of the patients. DISCUSSION Fine-needle aspiration cytology has been widely utilized for the diagnosis of malignant disease of the lymph nodes, breast, prostate, liver, pancreas, and other organs, and its diagnostic accuracy has been high [4-71. In gynecologic sites, FNA cytology was first utilized for the diagnosis of
IN GYNECOLOGIC
CANCER
311
primary ovarian tumors [8-111. However, in gynecologic malignancies, FNA cytology has not been widely used for the diagnosis of persistent or recurrent disease. Sevin et al. [15] presented four indications for FNA cytology in gynecologic malignancies: (1) primary diagnosis of pelvic masses, (2) superficial masses in patients with prior gynecologic malignancies, (3) persistent or recurrent disease of the pelvis following radiation therapy, and (4) the follow-up of patients during and after chemotherapy. In our institution, FNA cytology of primary pelvic tumors has not been performed because of the possibility of tumor spread [12,13]. In this study, FNA cytology was performed mainly for metastatic or recurrent diseases. Only a few cases of FNA for primary vulvar diseases were included in our study. Of 335 sites, 162 (48.4%) were positive for malignancy, a result almost equal to the findings of Belinson et al. [16] and Moriarty et al. [17]. Our overall accuracy for 335 lesions was 95.2%, which is comparable to the accuracies reported by Sevin et al. [15], Moriarty et uf. [17], and Layfield et al. [18]. Moriarty et al. [17] divided the aspiration sites into superficial sites and deep sites. Superficial sites included superficial lymph nodes, parametrial/paravaginal masses, subcutaneous masses, and vulvar masses,while deep sites included pelvic masses, intraabdominal masses, ovarian masses, deep lymph nodes, retroperitoneal masses, and others. They reported that no significant difference was found in the false-negative rate between superficial sites and deep sites. Similarly, we divided the sites of FNAs into two groups on the basis of palpability, i.e., superficial and deep. In our study, parametrial masseswere classified as deep sites. The accuracies for superficial sites and deep sites were almost equal. Bonfiglio et al. [19] reported that in the FNA cytology of retroperitoneal lymph nodes (47 cases), which were classified as deep sites in our study, only 4.3% of FNAs were false negative and there were no false positive FNAs. Moriarty et al. [17] re-evaluated the negative FNA cytology specimens and divided them into three categories: (1) adequate specimen and cell population appropriate for the site, defined as more than two cell groupings on two separate slides or sheets of lymphocytes if a lymph node was the target; (2) inadequate specimen, defined as less than two cell groupings on two separate slides; and (3) adequate specimen but wrong cell type for alleged biopsy site. Re-evaluation revealed inadequate specimens to be 22% (71/317) and inappropriate specimens 3% (8/317). Excluding these inadequate and inappropriate specimens, the sensitivity increased from 73 to 91%, and the predictive value of negative results also increased from 70 to 84%. Therefore, they recommended repeating aspirations for inadequate and inappropriate specimens. Also, in a study by Belinson et al. [16], 5.6% of 90 FNAs of gynecologic cancer were insufficient for evaluation. In their study, the specimens were considered to be adequate
312
IMACHI
when some normal cells from the target organ were observed, and they excluded specimens without normal cells from the target organ, namely, inadequate specimens. As a result, in our study, the sensitivity increased from 90.9% for 352 FNAs including inadequate specimens to 96.4% for 335 FNAs after excluding the inadequate specimens. Of the total 405 FNAs, 64 specimens (15.8%) were inadequate for diagnosis. Repeat FNA, with a mean of 2.6 aspirations, was performed for the patients with an inadequate FNA, and in consequence only 17 sites at which 3.8 aspirations were performed had inadequate FNAs. Therefore, FNA should be repeated for lesions yielding inadequate aspirations. Flint et al. [20] noted that negative aspirates must be judged more cautiously, especially if the aspirate is scanty. Layfield et al. [X3] reported that all primary lesions (13/13) were correctly identified by FNA, but only 63% (12/19) of secondary malignancies were correctly diagnosed. This was because the histological findings for primary malignancies were of greater cellularity with less scirrhous stroma than recurrent tumors. They also reported that the marked fibrotic response after radiation therapy caused a higher false-negative rate for recurrent and irradiated lesions. Therefore, in our study, the sensitivity for local recurrent lesions was compared with that for primary or metastatic lesions. Although there were only a few primary lesions in our study, the sensitivity for local recurrent disease was statistically lower than that for primary or metastatic disease (86.4% vs 100.0%). The difference in these two accuracies might be caused by the differences in cellularities between the two groups. In this study, 65.4% (106/162) of the positive FNA lesions were not biopsied. Reduction surgery without biopsy was performed directly for 25 lesions. Therefore, histological evaluation was performed for 81 lesions and 77 revealed malignant findings. Histologically, 2 cases were true negative and 2 were false negative. Two histologically false-negative specimens showed only necrotic and inflammatory tissue in both patients. Malignancy was confirmed on the basis of clinical evidence in the remaining 81 cases. The rate of false negatives was 1.8%, while that of false positives was 0.6%. These accuracies seem to be sufficient for the diagnosis of gynecologic malignant diseases. Therefore, when clearly malignant cells are observed in a FNA cytology specimen, biopsy may be omitted. Especially in patients with a poor general condition and in patients whose lesions are anatomically difficult to biopsy, the results of fine-needle aspiration cytology can play a very important role in deciding on further treatment.
ET AL
REFERENCES 1. Guthrie, C. G. Gland puncture as a diagnostic measure, Johns Hopkins Hosp. Bull. 366, 269 (1921).
2. Stewart, F. W. The diagnosis of tumors by aspiration, Am. J. Purhol. (Suppl), 801-812 (1933). 3. Martin, H. E., and Ellis, F. G. Aspiration biopsy, Surg. Gynecol. Obstet. 59, 578-589 (1934).
4. Bibbo, M., and Zuspam, F.P. Fine needle aspiration of the breast in an obstetrics and gynecologic hospital, Am. J. Obstet. Gynecol. 121, 525-526 (1975). 5. Franz&t, S., Giertz, G., and Zajicek, J. Cytologic diagnosis of prostatic tumors by transrectal aspiration biopsy, Br. J. Ural. 32, 193-196 (1960). 6. Hajdu, S. I., and Melamed, M. R. The diagnostic value of aspiration smears, Am. J. Clin. Puthol. 59, 350-356 (1973). 7 Zajicek, J. Aspiration biopsy cytology. I. Cytology of supradiaphragmatic organs, in Monographs in clinical cytology (G. L. Wied, Ed.), S. Karger Base1 (1974). 8. Kjellgren O., Angstom, T., Bergman, F., and Wiklund, D. E. Fineneedle aspiration biopsy in diagnosis and classification of ovarian carcinoma, Cancer 28, 967-976 (1971). 9. Ramzy, I., Delaney, M. Fine needle aspiration of ovarian masses. I. Correlative cytologic and histologic study of celomic epithelial neoplasms, Actu Cytol 23, 97-104 (1979). 10. Ramzy, I., Delaney, M., and Rose, P. Fine needle aspiration of ovarianmasses. II. Correlative cytologic and histologic study of nonneoplastic cysts and noncelomic epithelial neoplasms, Actu Cytol. 23, 185-193 (1979). 11. Ganjei, P., and Nadji, M., Aspiration cytology of ovarian neoplasms: A review, Actu Cytol. 28, 329-332 (1984). 12. Christopherson, W. W. Cytologic detection and diagnosis of cancer: Its contribution and limitations, Cancer 51, 1201-1208 (1983). 13. Hajdu, S. I., and Melamed, M. R. Limitations of aspiration cytology in the diagnosis of primary neoplasms, Actu Cytol. 28, 337-345 (1984). 14. Nordqvist, S. R. B., Sevin, B. U., Nadji, M., Greening, S. E., and Ng, A. B. P. Fine-needle aspiration cytology in gynecologic oncology. I. Diagnostic accuracy, Obstet. Gynecol. 54,719-724 (1979). 15. Sevin, B. U., Greening, S. E., Nadji, M., Ng, A. B. Averette, H. E., and Nordqvist, S. R. B. Fine needle aspiration cytology in gynecologic oncology, I. Clinical aspects, Acfu Cytol. 23, 277-281 (1979). 16. Belinson, J. L., Lynn, J. M., Papillo, J. L., Lee, K., and Korson, R. Fine-needle aspiration cytology in the management of gynecologic cancer, Am. J. Obstet. Gynecol. 139, 148-153 (1981). 17. Moriarty, A. T., Glant, M. D., and Stehman, F. B. The role of fine needle aspiration cytology in the management of gynecologic malignancies, Actu Cytol. 30, 59-64 (1986). 18. Layfield, L. J., Heaps, J. M., and Berek J. S. Fine-needle aspiration cytology accuracy with palpable gynecologic neopiasms, Gynecol. Oncol. 40, 70-73 (1991). 19. Bonfiglio, T. A., Macintosh, P. K., Patter, S. F., Cafer, D. J., Woodworth, F. E., and Kim, C. W. Fine needle aspiration cytopathology of retroperitoneal lymph nodes in the evaluation of metastatic disease, Actu Cytol. 23, 126-130 (1979). 20. Flint, A., Terhart, K., Murad, T. M., and Taylor, P. T. Confirmation of metastases by fine needle aspiration biopsy in patients with gynecologic malignancies, Gynecol. Oncol. 14,382-391 (1982).