Computed tomography-guided fine-needle aspiration biopsy

The therapeutic and prognostic evaluation of malignant neoplasias depends largely upon a precise morphologic diagnosis. Several papers have focused on the importance of fine-needle aspiration under computed tomography guidance in the diagnosis of unresectable neoplasms and in the investigation of metastases. The objective of this study is to evaluate the diagnostic accuracy, the sensitivity, and the negative predictive value obtained with the technique. Fine-needle aspiration cytology (FNAC) was performed on 207 patients, with a total of 210 cases, from 1991 to 1994, under computed tomography (CT) guidance. There were 128 (61.8 percent) males and 79 (38.2 percent) females with a mean age of 41 years (range 1 to 91 years). Lung and Iiver were the most frequent anatomic sites. The analysis 'of this material disclosed 41 cases with cytological diagnosis of negative for malignancy (19.52 percent), and in 131 (62.38 percent), the diagnosis was positive. It was possible to define the cytologic lineage in 54 percent of the cases. The diagnosis in 14 (6.67 percent) cases was suspicious for malignancy, and in 24 (11.43 percent) cases the material was insufficient for the cytologic diagnosis. Of the 210 cases, 106 showed histological diagnosis and/ or clinicai follow-up. Forty-seven (44.3 percent) had histological diagnosis before the FNAC and 50 cases (56.7 percent) histological diagnosis after the procedure. The comparison between cytological and histological diagnosis showed a sensibility of 80.4 percent, specificity of 100 percent, positive predictive value of 100 percent and negative of 16.7 percent. The efficiency of the test was 81.1 percent. This study showed that FNAC, under computed tomography guidance, is a sensitive and specific technique for the diagnosis of deep-seated lesions.


INTRODUCTION F
ine-needle aspiration biopsy (FNAB) is an outpatient lnethod used for the evaluation of palpable masses.The assessment of deep-seated, non-palpable lesions requires imaging methods, computed tomography (CT) in particular.FNAB application is important for the diagnosis of advanced and unresectable neoplasias as available results have confirmed the effe,ctiveness and safety of this technique,3 which is indicated for the investigation of benign or malignant, primary or metastatic neoplasias.A precise morphologic diagnosis is essential for the planning of treatment and the prognosis of malignant neoplasias.Several reports have emphasized the high specificity and sensitivity of CT-guided aspirative cytology.2.9.,o The present report presents the results obtained with CT-guided FNAB and compares them to those obtained by histology of biopsy material and surgical specitnens, and in terms of the clinicaI follow-up of the patients in order to establish the diagnostic accuracy of the method.

Table 2 Distribution of FNAB by anatomic sites
The 210 aspirative biopsies were perfonned on 207 patients, 128 males (6 1.8 percent) and 79 felnales (38.2 percent).Patient age ranged froln 1 to 9 I years (Inean 4 I years), with a Iarger number of cases situated in the seventh decade of life (Table 1).The procedures were perfonned from July 199 I to June 1994.Table 2 snows their distriblltion by year.The biopsies were obtained froln different Iocations, with the lung and Iiver being the Inost frequent (Figs.I, 2 and 3).Table 2 shows their distribution by 10cation.1344 1994.AlI cases were submitted to the classical FNAB technique Ó using a Franzen type cytoaspiration gun coupled to a 10 lnl disposable syringe with.a disposable needle comlnonly used for spinal anesthesia, 0.7 mm in outer dialneter (22 G).The procedures were carried out under the guidan.ce of a tomography apparatus (CT MAX, General Electric).
The Slnears were fixed in 95 percent alcohol and stained by the Papanicolaou and H&E techniques.Airdried slides were then submitted to Giemsa staining.In some cases, part of the material obtained was fixed in .1Opercent formalin and embedded in paraffin for the preparation of a cell block, and the slides were stained by the H&E technique.
The positive and negative cytologic diagnoses were considered to be true or false in relation to the histopathologic diagnoses and clinicaI folIow-up of the patients.On the basis of these data, statistical indices of clinical-Iaboratory importance were calculated, such as specificity, sensitivity, positive and negative predictive value, and efficiency."xCases with a suspicious or inconclusive cytologic diagnosis, and cases with no histopathologic examination', were excluded froln the statistical analysis.
The patients were divided into cases with a histopathologic diagnosis obtained before the aspirative biopsy, and cases with an anatomopathologic report and/ or clinicaI follow-up after the cytologic examination.The diagnosis for the presence of neoplastic cells was negative in 41 cases (19.52 percent) and positive in 131 (62.38 percent).The diagnosis of the celllineage was possible in 54 percent of the latter cases.A sllspicious diagnosis (atypical cells) was lnade in 14 cases (6.67 percent), and in 24 cases (11.43 percent) the InateriaI was considered insufficient.Of the 207 patients stlldied, 106 had an anatomopathologic exalnination and/or clinicaI follow-up.A diagnosis preceding the aspirative biopsy was available in 47 cases (44.3'percent), and a histopathologic examination folIowing the biopsy was obtained in 50 cases (56.7 percent).
Comparison of the cytologic and histopathologic diagnoses revealed 82 cases of true-positive results (77.3 percent), 4 cases with true-negative results' (3.8 percent), no false-positive case, and 20 cases w,ith false-negative results (18.9 percent) (Table 3).The test showed 80.4 surgery. 2 The method permits a precise Inorphologic diagnosis, reducing the need for diagnostic thoracotolnies in 31-35 percent of cases. 7. 12FNAB offers several advantages over other diagnostic Inethods, especially in thoracic lesions: it is rapid, does not require general anesthesia or prolonged hospitalization, and is less costly and 1110retoIerabIe to patients compared to surgery OI' mediastinoscopy.15  In cases of hepatic Iesions, the Inajor advantage of FNAB compared to standard-needle biopsy is its higher success rate in detecting malignant neoplasias.The discrepancy between these two lnethods with respect to tumor detection is reIated to thereIatively atraulnatic nature .ofFNAB,which permits lnultiple punctures ofthe Jesion, whereas the number of procedures (or attelnpts) with a The present study showed that CTguided FNAB is a sensitive and highIy specific technique for the diagnosis of deepseated lesions, with resuIts silnilar to those obtained in previous investigations.2 . 9The method has proved to be reproducibIe at severaI medicaI centers and is being appIied for the diagnosis of neoplasias, especially unresectabIe lesions or lesions of difficuIt surgicaI access, and for the confirmation of metastases.AnaIysis of the results revealed that approximately 44 percent of patients with an anatolnopathoIogic examination (47 cases) presented a previous diagnosis, with FNAB being used to evaIuate metastases.
Because of the inherent characteristics of FNAB, which uses slnalI-gauge needIes (22 G), the method is safe, without the risks of more invasive methods.Furthermore, it pennits the use of appropriate treatment by offering a precise typing of the neoplasias in a significant number of cases, with the consequent avoidance of risky major

Total Previous AC Later AC
increases with repeated procedures. 14Among the factors responsible for false-negative results are peritulTIOral infaretion, infIammation and fibrosis.2 A low index of false-negati ve diagnoses has been considered to be the result of intense cooperation and communication between pathologist and radiologist, with a negative result being considered definitive for those cases in which the morphologic and radio.Iogic aspects indicate a benign process. 2  Sensitivity and specificity were within the range reported for aspirative biopsies of thoracic lesions, i.e., 71 to 100 percent and 95 to 100 percent, res pectively.2.5.7.10.lóTheresults revealed that there was no difference in sensitivity, specificity or efficiency between

Table 4 Indices of the diagnostic performance of fine-needle aspirative cytology
Indices thick needle is limited by the potential occurrence of severe cOlTIplications.13  The frequency of insufficient material (11.43 percent) was similar to that obtained in other studies. 2 . 9In the case of superficial palpable lesions of.easieraccess, the index of unsatisfactory material, which includes .insufficient samples, samples that are not representative of the organ and samples represented by blood elements, is variable.The index for mammary lesions ranges from 3 to 36 percent, and the index for thyroid lesions ranges from 8 to 18 percent (4).The proportion of unsatisfactory aspirates decreases with increasing experience with the procedure on the part of the operator. 4t has been suggested that honest communication with clinicians about the quality of the cytologic contribution to the patient management and its limitations could reduce the number of false-positive results. 11Although some lesions are difficult to interpret both in terms of discrilTIinating malignancy and in terms of histogenetic characterization, no false-positive cases occurred in the present study, explaining the high specificity deteeted.
The percentage of false-negative results was 18.86 percent.Thus, a negative result does not exclude the presence of a malignant neoplasia and the examination should be repeated.In the present series, no new aspirative biopsies were taken.However, it is known that positivity  cases with an anatomoclinical diagnosis preceding or following the biopsy.Sensitivity, specificity and predictive value depend on lnany factors in addition to the presence or absence of disease.x The predictive value of a test depends on the prevalence of a given disease.x In the present study, the negative predictive value, VP (-) was low.This is explained by the fact that the study was carried out at a cancer hospital that receives cases screened at general hospitaIs and private offices, with a high prevalence ofthis disease.Thus, there is a reduction in the nUlnber of true-negative results, a variable that directly affects the VP (-) values.The calculation of specificity was not impaired only because there were no false-positive results in this study.However, the high prevalence of cancer in this population (96.2), and the positive predictive value of 100 percent indicate that a patient with a positive result has a high probability (certainty) of really having cancer.
The results obtained in this study, using CT-guiclecl FNAB, in anatolnic parts of the body when t.he access is difficult, recommend the use of this proceclure as an efficient and safe lnethod for the diagnosis of Inalignancy.

Figure 1 -
Figure 1 -Axial tamagraphy sectian af the liver showing the pasitian af the needle in a nadular lesion.

Figure 2 - 1 *
Figure 2 -Axial tomography section of the chest with a patient in ventral decubitus.Note the distai end of the needle positioned inside a nodular lesion of the lung.

Figure 3 -
Figure 3 -Detail of Figure 2, showing the needle positioned inside the nodular lesion.

Table 1
Patient distribution by age range in 207 patients submitted to FNA.