Journal of Surgical Oncology 44: 15-19 (1990)
Fine-Needle Aspiration Cytology of Abdominal Masses AlAY KUMAK KtIANNA, MU, MAHENDRA KUMAK MISRA, MS, A N U R A D H A KHANNA, VIJAY KUMAR MISKA, IMS, AN[) SUSt1II.A KHANNA, MI) from thc Insrirurr oi Medical Scipnccs, 13anaras \ jindu University, V a r a n a i , M i d
MO,
An investigation of the role of blind fine-needle aspiration cytology (FNAC) in the assessment of palpable abdoniinal masses was carried out on 196 patients: 124 hepatic, 30 retroperitoneal, and 42 other masses. All of the smears were stained either by Papanicolaou stain or by hematoxylin and eosin stain. The results of FNAC were confirmed by further investigations in all cases. FNAC correctly diagnosed 166 (84.6%) cases. Twenty (10.2%) reports were false negative, and 10 (5.1%) smears were unsatisfactory for any diagnosis. Of 124 hepatic masses, the correct diagnosis was obtained in 106 (85.4%), false-negative reports in 14 ( 1 1.2%), and unsatisfactory smears in 4 (3.2%). There were no falsepositive reports. For all of the lesions, the sensitivity of FNAC was 87.8%; specificity, 100%; positive predictive value, 100%; and negative predictive value. 52.4%. Five (2.5%))patients had considerable pain after the procedure. KEY WORDS:Papanicolaou stain, hepatic, retroperitoneal masses, blind procedure
INTRODUCTION ,Many studies have documented that aspiration cytology of the abdominal masses is a reliable and safe method of diagnosis [ 1-51. More advance studies, such as immunoperoxidase staining [ 61, electron microscopy [7], histological sectioning of the aspirate [8,9], ultrasound-guided cytology [ 101, CT-guided cytology [ 1 11, and biochemical analysis of the aspirate [ 12-14], are being used to improve diagnostic accuracy. In the centers where these advance studies are available, one can be quite sure of the diagnosis and one can conclusively diagnose the different malignancies as per the specific type [31. However. many centers that carry out aspiration cytology do not yet have advanced facilities, but they can still achieve acceptable results. Although fine-needle aspiration cytology (FNAC) has been used for almost all organs, there were techniquerelated complications with abdominal masses such as bilious peritonitis [.IS]. pancreatitis [ 161. and death 1171. These complications, however, can be well avoided by using the thin needle and by avoiding repeated aspiration [ 3 ] . The present study attempts to define the accuracy and safety 0 1 FNAC as a blind procedure with only the Q 1990 Wiley-Liss, Inc.
necessary basic technique for diagnosis. Because this study is an initial one from this medical center, all of the results of FNAC were confirmed by further investigations. and in all cases, as far as possible, a histological diagnosis was also obtained.
MATERIALS AND METHODS From 1982 to 1987, 196 patients admitted to the University Hospital, Banaras Hindu University (Varanasi. India), were subjected to FNAC. The mean age was 32 years (range, 8-70 years), with a male to female ratio of 2.5: 1 . Detailed clinical examination was carried out, and a clinical diagnosis was made. After a preliminary workup, FNAC was carried out with a 10 or 20 ml syringe and a 21 or 22 gauge needle; no anesthesia was used. The aspirate was quickly spread on the glass slide and was either wet fixed by the spray or was put in a 95%- alcohol
Accepted for publication 1)ecember 6 . 1080. Address reprint requests to Ajay Kuinar Khanna, 3313. Kobir Sagar. Durgakund. Varanasi 221 010. India.
16
Khanna et al.
TABLE I. Accuracv of FNAC in Abdominal Masses
TABLE 11. Diagnostic Accuracy in Liver Masses
Correct diapnosis
False positive
False negative
Unsatisfactory
Liver. 124 Retroperitoneurn, 30 Ovary, I6 Gallbladder, 10 Stomach, 10 Colon, 6
106 (85.4)" 22 (73.3) 16 (100) 8 (80) 8 (80) 6 (100)
-
14 (11.2) 4 (13.3)
4 (3.2) 4 (13.3)
Total, I96
166 (84.6)
Site. No. of
-
-
-
-
2 (20) -
2 (20)
-
-
20 (10.2)
10 (5.1)
"Values in parentheses are percentages
FNAC Final diagnosis. No. of Secondary liver, 96 Hepatocellular carcinoma, 14 Cirrhosis of the liver, 8 Amoebic liver abscess, 4 Infected hydatid, 2 Total. 124
and solvent ether ( I :1) solution in a jar. If the amount of aspirate was large, then it was centrifuged and a smear was made. The slides were stained by either Papanicolaou or hematoxylin and eosin. On the basis of clinical diagnosis and aspiration cytology the patients were further investigated. After the cytological report was made, histological confirmation was obtained by laparoscopic biopsy, laparotomy, or blind Tru-cut needle biopsy. Four cases of amoebic liver abscess and two cases of cold abscess, which were diagnosed as inflammatory exudate on cytology, underwent thick-needle aspiration. Statistical analysis was carried out using a binary 2 X 2 table [181.
RESULTS One hundred twenty-four lumps were detected clinically in the liver, 30 in retroperitoneum, 16 in ovaries, 10 in gallbladder, 10 in stomach, and 6 in colon. Aspirate satisfactory for cytological diagnosis was obtained in 182 cases. Only one pass was tried. The accuracy of FNAC is shown in Table I. Liver FNAC of liver masses was accurate in 85.4% of cases. It correctly diagnosed 82 of 96 secondary instances in liver and 12 of 14 cases of hepatocellular carcinoma (Table 11). In the group of secondary lumps, 78 were adenocarcinoma, 2 were melanoma, and 2 were hepatocellular carcinoma. Cirrhosis of the liver was correctly diagnosed in six of eight cases. Six smears gave the diagnosis of inflammatory exudate, of which ultimately four were amoebic liver abscess and two were infected hydatid cysts of the liver. Of the 124 cases in this group, 5 1 underwent laparotonly, 50 had laparoscopy and 20 had Tru-cut needle biopsy. Of 51 patients who needed laparotomy, 24 patients had secondary lumps in liver, 6 patients had hepatoma, 14 had false-negative results, 4 had unsatisfactory smears. 2 had hydatid cysts of the liver, and 1 patient with amoebic liver abscess failed to resolve after repeated aspiration. Laparotomy was carried out in eight
Correct diagnosis
False negative
Unsatisfactory
82 (85.4)" 12 (85.7)
12 (12.5) 2 (14.2)
2 (2.08)
6 (75) 4 (100)
-
-
-
2 (100)
-
-
106 (85.4)
14 ( I 1.2)
4 (3.2)
2 (25)
"Values in parentheses are percentages.
cases of secondary liver instances and two cases of hepatoma mainly because of the therapeutic approach, whereas in the 16 cases of secondary liver involvement and 4 cases of hepatoma it was done for the final diagnosis. Laparoscopy was carried out in 50 cases and a punch biopsy was obtained, whereas in 20 cases a Trucut needle biopsy under laparoscopic vision was done. The blind Tru-cut needle biopsy was obtained in 24 cases of which 20 gave the correct diagnosis. Four cases of amoebic liver abscess and two cases of infected hydatid cyst diagnosed as inflammatory exudate on cytology underwent thick-needle aspiration. Ultrasound of the liver in cases of hydatid cyst revealed the diagnosis, and cysts were explored. Four cases of amoebic liver abscess underwent thick-needle aspiration and of these one patient failed to respond and thus underwent laparotomy. Of the 124 liver aspirates. 4 (3.2%) spreads were scantily cellular and thus inadequate for interpretation, 14 ( 1 1.2%) reports were false negative, and 106 (85.4%) were diagnosed correctly. Of the 94 smears of malignancy, 4 were suggestive of malignancy.
Retroperitoneal Mass From 30 patients with suspected retroperitoneal mass, 4 (13.3%) smears were insufficient for any interpretation. Of the 26 aspirates, the correct diagnosis was obtained in 22 (84.6%), while 4 (15.3%) gave false-negative results (Table 111). Of 22 aspirates, 10 were diagnosed as lymphoma, 2 as retroperitoneal sarcoma, 4 as secondary involvement of testicular tumor of which 3 were from seminoma and 1 seminoteratoma, 3 as metastatic deposits from adenocarcinoma, 2 as inflammatory exudate, and 1 as dermoid cyst. In cases of inflammatory exudate based on the clinical picture and on the site of the lesion, a diagnosis of cold abscess was made. The diagnosis was confirmed by thick-needle aspiration, and the pus was submitted for acid fast bacilli ~ m e a and r culture. A dermoid cyst was diagnosed because a few hairs were aspirated during the procedure and on laparotomy it was
FNAC of Abdominal Masses 1’ABLE 111. Diagnostic Accuracy in Hetroperitoneal Masses FNAC
Con.ect diagnosis
False ncgativc
Lymphoma. 14 Sarcoma. 3 Tcsticular tumor, 3 Secondaries lyinph nodes. 3 ‘fubcrcular lymph nodes. 2 Cold :ihsc.exs. 2 1)ermoid cyst. I
10 (71.4)‘’ 2 (50) 4 (100) 3 (loo)
4 (28.5)
Totnl. 30
22 (73.3)
Final diagnosis. Ko. of
Ensatisfactory -
-
2 (50)
-
2 ( 100)
4 (13.3)
4 (13.3)
7- ( 100) I (100)
17
Overall Results The sensitivity ([he probability that the test will be positive when the disease is present) of FNAC for palpable abdominal masses is 87.8%. The specificity (the probability that the test will be negative when the disease is absent) is 100%. The positive predictive value (the probability that the disease is present when the test is positive) is 100%. The negative predictive value (the probability that the disease is absent when the test is negative) is 52.4%. The overall accuracy rate was 84.6%. The correct diagnosis was achieved in 84.7% of malignant tumors, 85.7% of benign tumors, and 83.3% of inflammatory masses.
“Valucs i n parentheses are percentages.
DISCUSSION In the presence of a palpable abdominal mass, thc usual practice of establishing a diagnosis is radiology, guided-needle biopsy, laparoscopy , or laparotomy. Others These procedures involve hospitalization and anesthesia FNAC gave a IOO% correct diagnosis of ovarian tu- and a nursing staff, a surgeon, a technical staff, and a mors. Of the 16 smears, 12 were positive for malignant pathologist. The procedures are costly, and, at medical cells. Laparotomy was performed in 14 cases. and, his- centers where ultrasound and CT are not available, they tologically. 4 were serous cystadcnoma, I clear cell tu- are traumatic too. With ultrasound and CT one can obmor, 5 serous cystadcnocarcinonia, 3 mucinous cystad- tain an impression about the tumor but cannot make a enocarcinoma, and 1 endodermal sinus tumor. Two cytohistologic diagnosis unless one does a guided-needle patients had laparoscopy. and a biopsy from the omen- biopsy. Biopsy of the abdominal mass often requires a tum revealed adenocarcinoma. In the gallbladder group. team that includes a radiologist and radiographer and out of 10 cases the correct diagnosis was obtained in 8 involves expensive imaging equipment. The procedure is (80%). Needle aspiration was done only in those gall- time consuming. Under these circumstances the tcchbladder cases in which the lump was firm to hard in nique is rather more expensive and one may be justified consistency. Histological diagnosis was achieved in this i n arranging for a cytotechnician and a radiographer so group by laparoscopy and biopsy in 6; the others had that sufficient material froin the right site is aspirated. If this avoids major surgical intervention, then the added laparotomy . Of 10 cases of stomach mass. 8 (80%) smears gave the trouble is justified [2]. Blind percutaneous FNAC is an inexpensive. simple. correct diagnosis and all were malignant. Two (20%) smears were insufficient for any diagnosis. Laparotomy safe, rapid. and reliable method of diagnosis in palpable was done in nine cases, and it revealed growth in the abdominal masses. It docs not require hospitalization body of the stomach in four and in the pyloric antrum in anesthesia or a nursing staff and is carried out in the 5. Palliative gastrcctomy was done in 3 , bypass opcra- day-carc ward. The sensitivity of FNAC depends on the tion in 3 , and biopsy in 3 . One case associated with site of the lesion. the depth of the lesion, the skill of the nodular hcpatonicgaly underwent laparoscopy and a bi- physician performing the procedure, and the cxperiencc opsy was obtained from the liver nodule. Of six smears of the pathologist [ 191. The reported sensitivity of FNAC from the colon, four were malignant and two were of for all abdominal malignancies varies between 70 and inflammatory origin. Of these six cases, five underwent 95% 120).The greater strength of the cytodiagnosis is the laparotomy . Three had adenocarcinoma of the colon, one almost complete lack of false-positive results; the prehad amoebic typhlitis, and one had an appcndicular ab- dictive valuc of a positive test is almost 100% [ 2 11. Strict scess. One patient had associated hepatoniegaly , and lap- criteria of malignancy must be applied to prevent falscpositive results. Most of the reported false-positive diaroscopy confirmed the diagnosis. agnoses have occurred with renal cytology where B Complications chronic inflammatory condition may produce marked There were n o major complications. Minor complica- changes in cellular pathology [22]. For dcaling with lions were encountered in five (2.5%) cases, consisting these problems and to be more accurate, newer techof pain, tenderness, and bruising at the site. niques have been applied such as ultrasound-guided needetermined t o be a rctroperitoncal dernioid cyst. In this group 24 patients underwent laparotomy, and 4 had laparoscopy .
18
Khanna et al.
dle FNAC [lo], CT-guided cytology [ 113. histological seem very promising to identify lymphomas in very section of the aspirate IS], immunoperoxidase staining small samples. We diagnosed 10 of 14 abdominal lym[6], electron microscopy [7], and biochemical and other phomas. There were four cases of sarcoma of which two analyses of the aspirate [12-141. All of these new pro- were correctly diagnosed. It has been observed that cedures require sophisticated laboratories, instruments, smears from sarcomas contain mostly isolated malignant and chemicals, and such are not available at all medical cells whereas in carcinomas the cells are mostly present centers. Although these procedures improve the accuracy in clumps. A case of retroperitoneal dermoid cyst on of the FNAC technique, the level of accuracy of blind aspiration yielded brownish fluid along with a few hairs. FNAC with simpler staining procedure is acceptable. The aspirate was centrifuged, and the smear showed a Guided FNAC is more useful especially if the tumor is large number of epithelial cells without any evidence of malignancy. nonpalpable and is visualized with ultrasound or CT. The most common diagnostic problem to be solved by Aspiration cytology of other masses, excluding those FNAC of the liver is to distinguish between primary and from the liver and retroperitoneum, was positive in secondary tumors [4]. The peculiar cell morphology of 90.4%. The high accuracy in stomach and colon malighepatocellular carcinoma has been shown to mimic nor- nancies may have been due to loss of cohesion among mal liver cells [23,24]. The metastases to the liver de- malignant cells, which is the reason for their exfoliation. rived from different primary tumors have special features In ovarian masses, aspiration from the tumor yielded and hence can be identified [25.26]. The sensitivity of thick mucinous fluid. In cases of cystadenocarcinoma of FNAC in liver tumors varies from 86 to 94% [25,27,28]. the ovaries, the smears were rich in cell populations with The lesser the degree of the differentiation of metastatic aggregates and anisonucleosis, and the cells showed a adenocarcinoma is, the higher is the diagnostic accuracy tendency toward an acinar arrangement in a mucinous of the aspiration cytology. The inability to strike a focal background. The diagnostic accuracy of ovarian tumors tumor deposit or necrosis within the tumor is the main varies from 90 to 95% [30]. We could diagnose all 16 cause of false-negative results, especially in liver tu- tumors of the ovaries correctly by FNAC. mors, and that is why the guided-needle biopsy is helpful The main problems of FNAC of abdominal masses are in deep-seated tumors [4]. Fine-needle biopsy of the liver internal bleeding and the risk of bowel perforation. There is a useful means of rendering a diagnosis in patients are occasional case reports of the fatal outcome of FNAC with “liver defects” detected by imaging techniques, involving acute pancreatitis, hemorrhage in the pancreand it is useful in verifying the presence of liver me- atic abscess, haematoma formation and biliary peritonitis tastases in patients with known carcinoma. It determines [ 15-17,311. Hemorrhagic complications are so rare with the nature of the primary tumor in patients with unknown the fine needle that a bleeding diathesis is an indication primary, and it identifies the type of metastatic carci- for caution rather than for desisting [2 11. Implantation of noma [5]. The presence of a pathologist during the as- cancer cells along the needle track is another disadvanpiration procedure provides a good rapport between the tage, but it occurs rarely. Some studies have shown a pathologist and radiologist and leads not only to better similar survival rate in matched groups of fine-needle clinical and radiological correlation but also to sugges- and surgically biopsied patients with breast cancer [32] tion of additional biopsy sites and sampling for special and renal cancer [33]. There is thought to be the possiprocedures such as cultures, Gram staining of abscess. bility of seeding along the needle track in ovarian tumors, and electron microscopy [5]. Although the facilities of a but there has been no documentation. Puncture of a hyradiologist and electron microscopy were not available in datid cyst has long been considered a contraindication, our study, the results were in the acceptable range. We but the anaphylaxis after such puncture has been reported achieved a correct diagnosis in 106 of 124 (85.4%) liver after using a thicker needle, more than 1 mm diameter. masses; there were 14 (1 1.2%) false-negative reports, FNAC procedure in superficial lesions of the liver, kidand 4 (3.2%) smears were unsatisfactory for any diag- ney, and spleen is not painful, but in deep seated lesions nosis. Although we were not able to locate the primary it is often associated with severe pain. Small biliary leaks site by FNAC, it gave an adequate diagnosis of the sec- and minute hematomas are common and are probably ondary tumors, which prompted us to search for the pri- responsible for pain [34]. In our series five patients had mary sites. We also encountered six patients with lesions considerable pain and tenderness requiring analgesia. of inflammatory origin in 124 cases of liver masses: 4 Four of these had retroperitoneal lesions, and one had were amoebic liver abscesses and 2 were infected hy- hepatocellular carcinoma. In conclusion, FNAC of abdominal masses is a safe datid cysts. Few cytologists are confident in diagnosing lymphoma and reliable procedure, and blind percutaneous FNAC and reticulosis, and the sensitivity of the technique is as gives a sufficiently correct diagnosis. The absence of low as 40% [29]. Immunocytochemical marker studies false-positive results with this procedure is probably its
FNAC of Abdominal Masses
greatest strength. Thus one can plan appropriate treatment based on the positive diagnosis, but in cases of negative diagnosis further investigations are mandatory.
17. 18.
REFERENCES 1. Alpern GA, Dekker A: Fine needle aspiration cytology of the
19.
pancreas: An analysis of its use in 52 patients. Acta Cytol 29: 873-878. 1985. 2. Lever JV, Trott PA, Web AJ: Fine needle aspiration cytology. J Clin Path01 38:l-1 I , 1985. 3. Pilotti S , Rilke F, Alasio L, Garbagnati F: The role of fine needle aspiration in the assessnient of renal masses. Acta Cytol 32:l-10,
20. 21.
1988. 4. Pilotti S, Rilke F, Claren R. Milella M, Lombardi L: Conclusive
5. 6. 7. X.
9.
10. 11.
12.
13. 14.
15. 16.
diagnosis of hepatic and pancreatic malignancies by fine needle aspiration. Acta Cytol 32:27-38, 1988. Pinto MM. Avila NA, Heller C1. Criscuolo EM: Fine needle aspiration of the liver. Acta Cytol 32:15-21, 1988. Nadji M: The potential value of immunoperoxidase techniques in diagnostic cytology. Acta Cytol 24:442-447. 1980. Akhtar M. Ali A, Owen EW: Application of electron microscopy in the interpretation of fine needle aspiration biopsies. Cancer 48:2458-2463, 1981. Kasugai H . Yanianioto R. Tatsuta M, Okano Y, Okuda S , Kishignami Y, Kitamura T, Wada A, Taniura H: Value of heparinized fine needle aspiration biopsy in liver malignancy. Am J Roentgenol I44:243 -244, 1985. Tatsuta M. Yamamoto R, Kasugai H, Okano Y, Noguchi S, Okuda S, Wada A, Tamura H: Cytohistologic diagnosis of neoplasms of the liver by ultrasonically guided fine needle aspiration biopsy. Cancer 54:1682-1686, 1984. Hidvegi D, Nieman HL, James W: Percutaneous transperitoneal aspiration of pancreas guided by ultrasound. Acta Cytol 23: 181184, 1979. Luning M. Kursawe R. Schopke W , Lorenz D. Menzel A, Hoppe E. Meyer R: CT-guided percutaneous fine needle biopsy of the pancreas. Eur J Radiol 5:104-108, 1985. Kimura K. Yamanaka T. Sakai H, Ido K , Seki H, Yoshida Y: Biochemical and cytological analyses of cytic fluid aspirated by percutaneous puncture under ultrasonic guidance in cystic diseases of the pancreas. Gastroenterol Jpn 17:4-9, 1982. Nosher JL: Aspiration with amylase determination in differentiating pancreatic from renal cysts. J Urol 123:94-95, 1980. Tatsuta M. Yamamoto R, Yamamura H, Okuda S , Tamura H: Cytologic examination and CEA measurements in aspirated pancreatic material collected by percutaneous fine needle aspiration biopsy under ultrasonic guidance for the diagnosis of pancreatic carcinoma. Cancer 52693-698, 1983. Schulz TB: Fine needle biopsy of the liver complicated with bile peritonitis. Acta Med Scand 199:141-142, 1976. Evans WK, Ho CS, McCloughlin MJ, Tao LC: Fatal necrotising
22.
23.
24. 25. 26. 27. 28. 29.
30.
3I . 32. 33. 34.
19
pancreatitis following fine needle aspiration. Radiol 141:61-62, I98 I . Riska H, Friman C: Fatality after fine needle aspiration biopsy of the liver. Br Med J 1517, 1975. Griner PF, Mayewski RJ, Mushlin AI, Greenland P: Selection and interpretation of diagnostic tests and procedures. Ann Intern Med 94553-589, 1981. Bognel C, Rougier P, Leclere J. Duvillard P, Charpentier P, Prade M: Fine needle aspiration of the liver and pancrease with ultrasound guidance. Acta Cytol 32:22-26, 1988. Yeh HC: Percutaneous fine needle aspiration biopsy of intraabdominal lesions with ultrasound guidance. Am J Gastroenterol 75:148-152, 1981. Leese WR: Guided abdominal fine needle puncture. In Russell RCG (ed): “Recent Advances in Surgery.” New York: Churchill Livingstone, 1986, Vol 12. 57-69. Holm HH, Als 0, Ganimelgaard J: Percutaneous aspiration and biopsy procedures under ultrasound visualization. In Taylor KJW (ed): “Clinics in Diagnostic Ultrasound: Diagnostic Ultrasound in gastrointestinal disease.” New York: Churchill Livingstone. 1978. Vol. 1, 137-149. Carney CN: Clinical cytology of the liver. Acta Cytol 19:244250. 1975. Tao LC. McLoughlin MJ, Evans WK, Donat EE: Cytologic diagnosis of hepatocellular carcinoma by fine needle aspiration biopsy. Cancer 53547-552, 1984. Johansen P. Svendsen KN: Scan-guided fine needle aspiration biopsy in malignant hepatic disease. Acta Cytol 22:292-296, 1978. Tao LC, Donat EE, Ho C, McLoughlin MJ: Percutaneous fine needle aspiration biopsy of liver. Cytodiagnosis of hepatic cancer. Acta Cytol 23:287-291, 1979. Schwerk W, Schimitz-Moormann P: Ultrasonically guided fine needle biopsies in neoplastic liver disease. Cancer 48: 1469-1477, 1981. Whitlatch S , Nunez C, Pitlik DA: Fine needle aspiration biopsy of the liver: A study of 102 consecutive cases. Acta Cytol 28:719725, 1984. Fermcci JT, Wittenberg J , Mueller PR, Simeone JF, Harbin WP. Kirkpatrick RH. Taft PD: Diagnosis of abdominal malignancy by radiologic fine needle aspiration biopsy. Am J Roentienol -134: 323-330, 1980. Kjellorgren 0, Angstran T, Bergman F, Wicklund DB: Fine needle aspiration biopsy in diagnosis and classification of ovarian cancer. Cancer 28:967-976, 197 I . Lundquist A: Fine needle aspiration biopsy of the liver. Acta Med Scand [Suppl] 520:l-28, 1971. Berg JW, Robbins GE: A late look at the safety of aspiration biopsy. Cancer 15:826-827. 1962. Van Schree T, Franzen S, Lundquist A: Renal adenocarcinoma. Scand J Urol 1265-269, 1967. Livraghi T, Damascelli B, Lombardi C, Spagridi 1: Risk in fine needle abdominal biopsy. J Clin Ultrasound I1:77-81. 1983.
Fine-Needle Aspiration Cytology of Abdominal Masses AlAY KUMAK KtIANNA, MU, MAHENDRA KUMAK MISRA, MS, A N U R A D H A KHANNA, VIJAY KUMAR MISKA, IMS, AN[) SUSt1II.A KHANNA, MI) from thc Insrirurr oi Medical Scipnccs, 13anaras \ jindu University, V a r a n a i , M i d
MO,
An investigation of the role of blind fine-needle aspiration cytology (FNAC) in the assessment of palpable abdoniinal masses was carried out on 196 patients: 124 hepatic, 30 retroperitoneal, and 42 other masses. All of the smears were stained either by Papanicolaou stain or by hematoxylin and eosin stain. The results of FNAC were confirmed by further investigations in all cases. FNAC correctly diagnosed 166 (84.6%) cases. Twenty (10.2%) reports were false negative, and 10 (5.1%) smears were unsatisfactory for any diagnosis. Of 124 hepatic masses, the correct diagnosis was obtained in 106 (85.4%), false-negative reports in 14 ( 1 1.2%), and unsatisfactory smears in 4 (3.2%). There were no falsepositive reports. For all of the lesions, the sensitivity of FNAC was 87.8%; specificity, 100%; positive predictive value, 100%; and negative predictive value. 52.4%. Five (2.5%))patients had considerable pain after the procedure. KEY WORDS:Papanicolaou stain, hepatic, retroperitoneal masses, blind procedure
INTRODUCTION ,Many studies have documented that aspiration cytology of the abdominal masses is a reliable and safe method of diagnosis [ 1-51. More advance studies, such as immunoperoxidase staining [ 61, electron microscopy [7], histological sectioning of the aspirate [8,9], ultrasound-guided cytology [ 101, CT-guided cytology [ 1 11, and biochemical analysis of the aspirate [ 12-14], are being used to improve diagnostic accuracy. In the centers where these advance studies are available, one can be quite sure of the diagnosis and one can conclusively diagnose the different malignancies as per the specific type [31. However. many centers that carry out aspiration cytology do not yet have advanced facilities, but they can still achieve acceptable results. Although fine-needle aspiration cytology (FNAC) has been used for almost all organs, there were techniquerelated complications with abdominal masses such as bilious peritonitis [.IS]. pancreatitis [ 161. and death 1171. These complications, however, can be well avoided by using the thin needle and by avoiding repeated aspiration [ 3 ] . The present study attempts to define the accuracy and safety 0 1 FNAC as a blind procedure with only the Q 1990 Wiley-Liss, Inc.
necessary basic technique for diagnosis. Because this study is an initial one from this medical center, all of the results of FNAC were confirmed by further investigations. and in all cases, as far as possible, a histological diagnosis was also obtained.
MATERIALS AND METHODS From 1982 to 1987, 196 patients admitted to the University Hospital, Banaras Hindu University (Varanasi. India), were subjected to FNAC. The mean age was 32 years (range, 8-70 years), with a male to female ratio of 2.5: 1 . Detailed clinical examination was carried out, and a clinical diagnosis was made. After a preliminary workup, FNAC was carried out with a 10 or 20 ml syringe and a 21 or 22 gauge needle; no anesthesia was used. The aspirate was quickly spread on the glass slide and was either wet fixed by the spray or was put in a 95%- alcohol
Accepted for publication 1)ecember 6 . 1080. Address reprint requests to Ajay Kuinar Khanna, 3313. Kobir Sagar. Durgakund. Varanasi 221 010. India.
16
Khanna et al.
TABLE I. Accuracv of FNAC in Abdominal Masses
TABLE 11. Diagnostic Accuracy in Liver Masses
Correct diapnosis
False positive
False negative
Unsatisfactory
Liver. 124 Retroperitoneurn, 30 Ovary, I6 Gallbladder, 10 Stomach, 10 Colon, 6
106 (85.4)" 22 (73.3) 16 (100) 8 (80) 8 (80) 6 (100)
-
14 (11.2) 4 (13.3)
4 (3.2) 4 (13.3)
Total, I96
166 (84.6)
Site. No. of
-
-
-
-
2 (20) -
2 (20)
-
-
20 (10.2)
10 (5.1)
"Values in parentheses are percentages
FNAC Final diagnosis. No. of Secondary liver, 96 Hepatocellular carcinoma, 14 Cirrhosis of the liver, 8 Amoebic liver abscess, 4 Infected hydatid, 2 Total. 124
and solvent ether ( I :1) solution in a jar. If the amount of aspirate was large, then it was centrifuged and a smear was made. The slides were stained by either Papanicolaou or hematoxylin and eosin. On the basis of clinical diagnosis and aspiration cytology the patients were further investigated. After the cytological report was made, histological confirmation was obtained by laparoscopic biopsy, laparotomy, or blind Tru-cut needle biopsy. Four cases of amoebic liver abscess and two cases of cold abscess, which were diagnosed as inflammatory exudate on cytology, underwent thick-needle aspiration. Statistical analysis was carried out using a binary 2 X 2 table [181.
RESULTS One hundred twenty-four lumps were detected clinically in the liver, 30 in retroperitoneum, 16 in ovaries, 10 in gallbladder, 10 in stomach, and 6 in colon. Aspirate satisfactory for cytological diagnosis was obtained in 182 cases. Only one pass was tried. The accuracy of FNAC is shown in Table I. Liver FNAC of liver masses was accurate in 85.4% of cases. It correctly diagnosed 82 of 96 secondary instances in liver and 12 of 14 cases of hepatocellular carcinoma (Table 11). In the group of secondary lumps, 78 were adenocarcinoma, 2 were melanoma, and 2 were hepatocellular carcinoma. Cirrhosis of the liver was correctly diagnosed in six of eight cases. Six smears gave the diagnosis of inflammatory exudate, of which ultimately four were amoebic liver abscess and two were infected hydatid cysts of the liver. Of the 124 cases in this group, 5 1 underwent laparotonly, 50 had laparoscopy and 20 had Tru-cut needle biopsy. Of 51 patients who needed laparotomy, 24 patients had secondary lumps in liver, 6 patients had hepatoma, 14 had false-negative results, 4 had unsatisfactory smears. 2 had hydatid cysts of the liver, and 1 patient with amoebic liver abscess failed to resolve after repeated aspiration. Laparotomy was carried out in eight
Correct diagnosis
False negative
Unsatisfactory
82 (85.4)" 12 (85.7)
12 (12.5) 2 (14.2)
2 (2.08)
6 (75) 4 (100)
-
-
-
2 (100)
-
-
106 (85.4)
14 ( I 1.2)
4 (3.2)
2 (25)
"Values in parentheses are percentages.
cases of secondary liver instances and two cases of hepatoma mainly because of the therapeutic approach, whereas in the 16 cases of secondary liver involvement and 4 cases of hepatoma it was done for the final diagnosis. Laparoscopy was carried out in 50 cases and a punch biopsy was obtained, whereas in 20 cases a Trucut needle biopsy under laparoscopic vision was done. The blind Tru-cut needle biopsy was obtained in 24 cases of which 20 gave the correct diagnosis. Four cases of amoebic liver abscess and two cases of infected hydatid cyst diagnosed as inflammatory exudate on cytology underwent thick-needle aspiration. Ultrasound of the liver in cases of hydatid cyst revealed the diagnosis, and cysts were explored. Four cases of amoebic liver abscess underwent thick-needle aspiration and of these one patient failed to respond and thus underwent laparotomy. Of the 124 liver aspirates. 4 (3.2%) spreads were scantily cellular and thus inadequate for interpretation, 14 ( 1 1.2%) reports were false negative, and 106 (85.4%) were diagnosed correctly. Of the 94 smears of malignancy, 4 were suggestive of malignancy.
Retroperitoneal Mass From 30 patients with suspected retroperitoneal mass, 4 (13.3%) smears were insufficient for any interpretation. Of the 26 aspirates, the correct diagnosis was obtained in 22 (84.6%), while 4 (15.3%) gave false-negative results (Table 111). Of 22 aspirates, 10 were diagnosed as lymphoma, 2 as retroperitoneal sarcoma, 4 as secondary involvement of testicular tumor of which 3 were from seminoma and 1 seminoteratoma, 3 as metastatic deposits from adenocarcinoma, 2 as inflammatory exudate, and 1 as dermoid cyst. In cases of inflammatory exudate based on the clinical picture and on the site of the lesion, a diagnosis of cold abscess was made. The diagnosis was confirmed by thick-needle aspiration, and the pus was submitted for acid fast bacilli ~ m e a and r culture. A dermoid cyst was diagnosed because a few hairs were aspirated during the procedure and on laparotomy it was
FNAC of Abdominal Masses 1’ABLE 111. Diagnostic Accuracy in Hetroperitoneal Masses FNAC
Con.ect diagnosis
False ncgativc
Lymphoma. 14 Sarcoma. 3 Tcsticular tumor, 3 Secondaries lyinph nodes. 3 ‘fubcrcular lymph nodes. 2 Cold :ihsc.exs. 2 1)ermoid cyst. I
10 (71.4)‘’ 2 (50) 4 (100) 3 (loo)
4 (28.5)
Totnl. 30
22 (73.3)
Final diagnosis. Ko. of
Ensatisfactory -
-
2 (50)
-
2 ( 100)
4 (13.3)
4 (13.3)
7- ( 100) I (100)
17
Overall Results The sensitivity ([he probability that the test will be positive when the disease is present) of FNAC for palpable abdominal masses is 87.8%. The specificity (the probability that the test will be negative when the disease is absent) is 100%. The positive predictive value (the probability that the disease is present when the test is positive) is 100%. The negative predictive value (the probability that the disease is absent when the test is negative) is 52.4%. The overall accuracy rate was 84.6%. The correct diagnosis was achieved in 84.7% of malignant tumors, 85.7% of benign tumors, and 83.3% of inflammatory masses.
“Valucs i n parentheses are percentages.
DISCUSSION In the presence of a palpable abdominal mass, thc usual practice of establishing a diagnosis is radiology, guided-needle biopsy, laparoscopy , or laparotomy. Others These procedures involve hospitalization and anesthesia FNAC gave a IOO% correct diagnosis of ovarian tu- and a nursing staff, a surgeon, a technical staff, and a mors. Of the 16 smears, 12 were positive for malignant pathologist. The procedures are costly, and, at medical cells. Laparotomy was performed in 14 cases. and, his- centers where ultrasound and CT are not available, they tologically. 4 were serous cystadcnoma, I clear cell tu- are traumatic too. With ultrasound and CT one can obmor, 5 serous cystadcnocarcinonia, 3 mucinous cystad- tain an impression about the tumor but cannot make a enocarcinoma, and 1 endodermal sinus tumor. Two cytohistologic diagnosis unless one does a guided-needle patients had laparoscopy. and a biopsy from the omen- biopsy. Biopsy of the abdominal mass often requires a tum revealed adenocarcinoma. In the gallbladder group. team that includes a radiologist and radiographer and out of 10 cases the correct diagnosis was obtained in 8 involves expensive imaging equipment. The procedure is (80%). Needle aspiration was done only in those gall- time consuming. Under these circumstances the tcchbladder cases in which the lump was firm to hard in nique is rather more expensive and one may be justified consistency. Histological diagnosis was achieved in this i n arranging for a cytotechnician and a radiographer so group by laparoscopy and biopsy in 6; the others had that sufficient material froin the right site is aspirated. If this avoids major surgical intervention, then the added laparotomy . Of 10 cases of stomach mass. 8 (80%) smears gave the trouble is justified [2]. Blind percutaneous FNAC is an inexpensive. simple. correct diagnosis and all were malignant. Two (20%) smears were insufficient for any diagnosis. Laparotomy safe, rapid. and reliable method of diagnosis in palpable was done in nine cases, and it revealed growth in the abdominal masses. It docs not require hospitalization body of the stomach in four and in the pyloric antrum in anesthesia or a nursing staff and is carried out in the 5. Palliative gastrcctomy was done in 3 , bypass opcra- day-carc ward. The sensitivity of FNAC depends on the tion in 3 , and biopsy in 3 . One case associated with site of the lesion. the depth of the lesion, the skill of the nodular hcpatonicgaly underwent laparoscopy and a bi- physician performing the procedure, and the cxperiencc opsy was obtained from the liver nodule. Of six smears of the pathologist [ 191. The reported sensitivity of FNAC from the colon, four were malignant and two were of for all abdominal malignancies varies between 70 and inflammatory origin. Of these six cases, five underwent 95% 120).The greater strength of the cytodiagnosis is the laparotomy . Three had adenocarcinoma of the colon, one almost complete lack of false-positive results; the prehad amoebic typhlitis, and one had an appcndicular ab- dictive valuc of a positive test is almost 100% [ 2 11. Strict scess. One patient had associated hepatoniegaly , and lap- criteria of malignancy must be applied to prevent falscpositive results. Most of the reported false-positive diaroscopy confirmed the diagnosis. agnoses have occurred with renal cytology where B Complications chronic inflammatory condition may produce marked There were n o major complications. Minor complica- changes in cellular pathology [22]. For dcaling with lions were encountered in five (2.5%) cases, consisting these problems and to be more accurate, newer techof pain, tenderness, and bruising at the site. niques have been applied such as ultrasound-guided needetermined t o be a rctroperitoncal dernioid cyst. In this group 24 patients underwent laparotomy, and 4 had laparoscopy .
18
Khanna et al.
dle FNAC [lo], CT-guided cytology [ 113. histological seem very promising to identify lymphomas in very section of the aspirate IS], immunoperoxidase staining small samples. We diagnosed 10 of 14 abdominal lym[6], electron microscopy [7], and biochemical and other phomas. There were four cases of sarcoma of which two analyses of the aspirate [12-141. All of these new pro- were correctly diagnosed. It has been observed that cedures require sophisticated laboratories, instruments, smears from sarcomas contain mostly isolated malignant and chemicals, and such are not available at all medical cells whereas in carcinomas the cells are mostly present centers. Although these procedures improve the accuracy in clumps. A case of retroperitoneal dermoid cyst on of the FNAC technique, the level of accuracy of blind aspiration yielded brownish fluid along with a few hairs. FNAC with simpler staining procedure is acceptable. The aspirate was centrifuged, and the smear showed a Guided FNAC is more useful especially if the tumor is large number of epithelial cells without any evidence of malignancy. nonpalpable and is visualized with ultrasound or CT. The most common diagnostic problem to be solved by Aspiration cytology of other masses, excluding those FNAC of the liver is to distinguish between primary and from the liver and retroperitoneum, was positive in secondary tumors [4]. The peculiar cell morphology of 90.4%. The high accuracy in stomach and colon malighepatocellular carcinoma has been shown to mimic nor- nancies may have been due to loss of cohesion among mal liver cells [23,24]. The metastases to the liver de- malignant cells, which is the reason for their exfoliation. rived from different primary tumors have special features In ovarian masses, aspiration from the tumor yielded and hence can be identified [25.26]. The sensitivity of thick mucinous fluid. In cases of cystadenocarcinoma of FNAC in liver tumors varies from 86 to 94% [25,27,28]. the ovaries, the smears were rich in cell populations with The lesser the degree of the differentiation of metastatic aggregates and anisonucleosis, and the cells showed a adenocarcinoma is, the higher is the diagnostic accuracy tendency toward an acinar arrangement in a mucinous of the aspiration cytology. The inability to strike a focal background. The diagnostic accuracy of ovarian tumors tumor deposit or necrosis within the tumor is the main varies from 90 to 95% [30]. We could diagnose all 16 cause of false-negative results, especially in liver tu- tumors of the ovaries correctly by FNAC. mors, and that is why the guided-needle biopsy is helpful The main problems of FNAC of abdominal masses are in deep-seated tumors [4]. Fine-needle biopsy of the liver internal bleeding and the risk of bowel perforation. There is a useful means of rendering a diagnosis in patients are occasional case reports of the fatal outcome of FNAC with “liver defects” detected by imaging techniques, involving acute pancreatitis, hemorrhage in the pancreand it is useful in verifying the presence of liver me- atic abscess, haematoma formation and biliary peritonitis tastases in patients with known carcinoma. It determines [ 15-17,311. Hemorrhagic complications are so rare with the nature of the primary tumor in patients with unknown the fine needle that a bleeding diathesis is an indication primary, and it identifies the type of metastatic carci- for caution rather than for desisting [2 11. Implantation of noma [5]. The presence of a pathologist during the as- cancer cells along the needle track is another disadvanpiration procedure provides a good rapport between the tage, but it occurs rarely. Some studies have shown a pathologist and radiologist and leads not only to better similar survival rate in matched groups of fine-needle clinical and radiological correlation but also to sugges- and surgically biopsied patients with breast cancer [32] tion of additional biopsy sites and sampling for special and renal cancer [33]. There is thought to be the possiprocedures such as cultures, Gram staining of abscess. bility of seeding along the needle track in ovarian tumors, and electron microscopy [5]. Although the facilities of a but there has been no documentation. Puncture of a hyradiologist and electron microscopy were not available in datid cyst has long been considered a contraindication, our study, the results were in the acceptable range. We but the anaphylaxis after such puncture has been reported achieved a correct diagnosis in 106 of 124 (85.4%) liver after using a thicker needle, more than 1 mm diameter. masses; there were 14 (1 1.2%) false-negative reports, FNAC procedure in superficial lesions of the liver, kidand 4 (3.2%) smears were unsatisfactory for any diag- ney, and spleen is not painful, but in deep seated lesions nosis. Although we were not able to locate the primary it is often associated with severe pain. Small biliary leaks site by FNAC, it gave an adequate diagnosis of the sec- and minute hematomas are common and are probably ondary tumors, which prompted us to search for the pri- responsible for pain [34]. In our series five patients had mary sites. We also encountered six patients with lesions considerable pain and tenderness requiring analgesia. of inflammatory origin in 124 cases of liver masses: 4 Four of these had retroperitoneal lesions, and one had were amoebic liver abscesses and 2 were infected hy- hepatocellular carcinoma. In conclusion, FNAC of abdominal masses is a safe datid cysts. Few cytologists are confident in diagnosing lymphoma and reliable procedure, and blind percutaneous FNAC and reticulosis, and the sensitivity of the technique is as gives a sufficiently correct diagnosis. The absence of low as 40% [29]. Immunocytochemical marker studies false-positive results with this procedure is probably its
FNAC of Abdominal Masses
greatest strength. Thus one can plan appropriate treatment based on the positive diagnosis, but in cases of negative diagnosis further investigations are mandatory.
17. 18.
REFERENCES 1. Alpern GA, Dekker A: Fine needle aspiration cytology of the
19.
pancreas: An analysis of its use in 52 patients. Acta Cytol 29: 873-878. 1985. 2. Lever JV, Trott PA, Web AJ: Fine needle aspiration cytology. J Clin Path01 38:l-1 I , 1985. 3. Pilotti S , Rilke F, Alasio L, Garbagnati F: The role of fine needle aspiration in the assessnient of renal masses. Acta Cytol 32:l-10,
20. 21.
1988. 4. Pilotti S, Rilke F, Claren R. Milella M, Lombardi L: Conclusive
5. 6. 7. X.
9.
10. 11.
12.
13. 14.
15. 16.
diagnosis of hepatic and pancreatic malignancies by fine needle aspiration. Acta Cytol 32:27-38, 1988. Pinto MM. Avila NA, Heller C1. Criscuolo EM: Fine needle aspiration of the liver. Acta Cytol 32:15-21, 1988. Nadji M: The potential value of immunoperoxidase techniques in diagnostic cytology. Acta Cytol 24:442-447. 1980. Akhtar M. Ali A, Owen EW: Application of electron microscopy in the interpretation of fine needle aspiration biopsies. Cancer 48:2458-2463, 1981. Kasugai H . Yanianioto R. Tatsuta M, Okano Y, Okuda S , Kishignami Y, Kitamura T, Wada A, Taniura H: Value of heparinized fine needle aspiration biopsy in liver malignancy. Am J Roentgenol I44:243 -244, 1985. Tatsuta M. Yamamoto R, Kasugai H, Okano Y, Noguchi S, Okuda S, Wada A, Tamura H: Cytohistologic diagnosis of neoplasms of the liver by ultrasonically guided fine needle aspiration biopsy. Cancer 54:1682-1686, 1984. Hidvegi D, Nieman HL, James W: Percutaneous transperitoneal aspiration of pancreas guided by ultrasound. Acta Cytol 23: 181184, 1979. Luning M. Kursawe R. Schopke W , Lorenz D. Menzel A, Hoppe E. Meyer R: CT-guided percutaneous fine needle biopsy of the pancreas. Eur J Radiol 5:104-108, 1985. Kimura K. Yamanaka T. Sakai H, Ido K , Seki H, Yoshida Y: Biochemical and cytological analyses of cytic fluid aspirated by percutaneous puncture under ultrasonic guidance in cystic diseases of the pancreas. Gastroenterol Jpn 17:4-9, 1982. Nosher JL: Aspiration with amylase determination in differentiating pancreatic from renal cysts. J Urol 123:94-95, 1980. Tatsuta M. Yamamoto R, Yamamura H, Okuda S , Tamura H: Cytologic examination and CEA measurements in aspirated pancreatic material collected by percutaneous fine needle aspiration biopsy under ultrasonic guidance for the diagnosis of pancreatic carcinoma. Cancer 52693-698, 1983. Schulz TB: Fine needle biopsy of the liver complicated with bile peritonitis. Acta Med Scand 199:141-142, 1976. Evans WK, Ho CS, McCloughlin MJ, Tao LC: Fatal necrotising
22.
23.
24. 25. 26. 27. 28. 29.
30.
3I . 32. 33. 34.
19
pancreatitis following fine needle aspiration. Radiol 141:61-62, I98 I . Riska H, Friman C: Fatality after fine needle aspiration biopsy of the liver. Br Med J 1517, 1975. Griner PF, Mayewski RJ, Mushlin AI, Greenland P: Selection and interpretation of diagnostic tests and procedures. Ann Intern Med 94553-589, 1981. Bognel C, Rougier P, Leclere J. Duvillard P, Charpentier P, Prade M: Fine needle aspiration of the liver and pancrease with ultrasound guidance. Acta Cytol 32:22-26, 1988. Yeh HC: Percutaneous fine needle aspiration biopsy of intraabdominal lesions with ultrasound guidance. Am J Gastroenterol 75:148-152, 1981. Leese WR: Guided abdominal fine needle puncture. In Russell RCG (ed): “Recent Advances in Surgery.” New York: Churchill Livingstone, 1986, Vol 12. 57-69. Holm HH, Als 0, Ganimelgaard J: Percutaneous aspiration and biopsy procedures under ultrasound visualization. In Taylor KJW (ed): “Clinics in Diagnostic Ultrasound: Diagnostic Ultrasound in gastrointestinal disease.” New York: Churchill Livingstone. 1978. Vol. 1, 137-149. Carney CN: Clinical cytology of the liver. Acta Cytol 19:244250. 1975. Tao LC. McLoughlin MJ, Evans WK, Donat EE: Cytologic diagnosis of hepatocellular carcinoma by fine needle aspiration biopsy. Cancer 53547-552, 1984. Johansen P. Svendsen KN: Scan-guided fine needle aspiration biopsy in malignant hepatic disease. Acta Cytol 22:292-296, 1978. Tao LC, Donat EE, Ho C, McLoughlin MJ: Percutaneous fine needle aspiration biopsy of liver. Cytodiagnosis of hepatic cancer. Acta Cytol 23:287-291, 1979. Schwerk W, Schimitz-Moormann P: Ultrasonically guided fine needle biopsies in neoplastic liver disease. Cancer 48: 1469-1477, 1981. Whitlatch S , Nunez C, Pitlik DA: Fine needle aspiration biopsy of the liver: A study of 102 consecutive cases. Acta Cytol 28:719725, 1984. Fermcci JT, Wittenberg J , Mueller PR, Simeone JF, Harbin WP. Kirkpatrick RH. Taft PD: Diagnosis of abdominal malignancy by radiologic fine needle aspiration biopsy. Am J Roentienol -134: 323-330, 1980. Kjellorgren 0, Angstran T, Bergman F, Wicklund DB: Fine needle aspiration biopsy in diagnosis and classification of ovarian cancer. Cancer 28:967-976, 197 I . Lundquist A: Fine needle aspiration biopsy of the liver. Acta Med Scand [Suppl] 520:l-28, 1971. Berg JW, Robbins GE: A late look at the safety of aspiration biopsy. Cancer 15:826-827. 1962. Van Schree T, Franzen S, Lundquist A: Renal adenocarcinoma. Scand J Urol 1265-269, 1967. Livraghi T, Damascelli B, Lombardi C, Spagridi 1: Risk in fine needle abdominal biopsy. J Clin Ultrasound I1:77-81. 1983.
