The presence of a thyroid nodule raises the question of malignancy, although fewer than 5% of nodules are actually malignant [2, 3]. The development and application of radionuclide scanning, sensitive ultrasonography, and needle biopsy have been helpful in distinguishing benign from malignant nodules and in selecting patients for surgery [2, 3, 8, 9]. During the last decade, confidence in fine-needle aspiration biopsy as a reliable test has grown considerably, and it has emerged as the most direct, accurate diagnostic procedure in the management of nodular thyroid disease, gaining worldwide acceptance [10-14].

Considerable data on fine-needle aspiration biopsy have been reported during the past few years [15-45]. We examine the utility and limitations of fine-needle aspiration biopsy in the evaluation and management of thyroid nodules by reviewing studies reported in the literature from 1982 to 1991. Relying on these data and our own 12 years of experience with more than 11 000 biopsies, we hope to offer a reasoned approach to the management of thyroid nodules.

Fine-Needle Aspiration Biopsy

Procedure

Fine-needle aspiration biopsy, an office procedure, is relatively simple and is described in detail elsewhere [10, 13, 17]. In brief, after careful thyroid palpation, the nodule or nodules to undergo biopsy are identified. The procedure is usually done with the patient in the supine position; the patient's neck is flexed backward, allowing maximal exposure of the thyroid lesion. The skin is cleaned with alcohol; usually, no local anesthesia is necessary. A 23- or 25-gauge needle attached to a 10-mL disposable syringe is used. Following the recommendation of Lowhagen and colleagues [10], most aspirators use a mechanical syringe-holder (Cameco-syringe pistol, Precision Dynamics Corporation, Burbank, California), but some prefer to manipulate the syringe directly with the fingers. Use of the syringe-holder allows one hand to be free for nodule identification and continuous palpation. With one hand holding the pistol-grip holder and the other hand palpating the nodule, the aspirator places the needle into the nodule, applies suction, and moves the needle back and forth within the nodule. Suction is then released and the needle is withdrawn; aspirated material is expelled on glass slides and prepared in a manner similar to that for blood smears. Slides are either air-dried and stained using the May-Grunwald-Giemsa technique or immediately wet-fixed in 95% ethyl alcohol and stained using a modified Papanicolaou method [17]. Most cytologists trained in the United States prefer Papanicolaou staining because it is more similar to other cytologic preparations.

Two to four aspirations are usually done. Hamburger and Hamburger [26] found that as the number of aspirations increased, false-negative results decreased, and they suggested doing a minimum of six aspirations. If cystic fluid is obtained, it may be concentrated by filtration (8-micron Nucleopore filter; Nucleopore Corp., Pleasanton, California) and stained using the same technique. After the procedure is completed, the patient is observed for a few minutes and then allowed to depart. Serious complications have not been reported; minor pain and local hematomas are transient and tolerable.

Cytodiagnosis

Cytodiagnoses can be divided into four categories [10, 29, 31, 40]: benign (negative), suspicious (indeterminate), malignant (positive), and nondiagnostic (unsatisfactory) (Table 1). A satisfactory smear contains five or six groups of well-preserved cells; each group consists of at least 10 to 15 cells [11, 17]. Patients with a "benign" cytodiagnosis do not have malignancy and may have a normal thyroid, a colloid nodule, lymphocytic thyroiditis, subacute thyroiditis, or other benign conditions. Colloid is often, but not invariably, present, and foam cells indicating degeneration are often noted [17, 22, 32]. Patients with a suspicious (indeterminate) cytodiagnosis have specimens showing hypercellularity and a pattern suggestive of follicular- or Hurthle-cell neoplasms or atypical features suggestive of, but not diagnostic for, malignancy [44]. Hypercellularity may represent non-neoplastic lesions, and Hurthle-cell changes may be seen in patients with lymphocytic thyroiditis. Patients with a malignant cytodiagnosis have cytologic findings indicating the presence of malignant cells consistent with primary or metastatic thyroid carcinoma [17, 20]. In patients with nondiagnostic (unsatisfactory) cytologic results, specimens are found to be inadequate for proper cytopathologic interpretation, usually because of the presence of cystic fluid or hemorrhagic material. Aspirates with too few cells are regarded as nondiagnostic and not as negative for malignancy. A repeat aspiration may provide diagnostic smears in up to 50% of cases [17]. It is clear that proper cytologic interpretation requires specific training in cytology and considerable familiarity with the pathology of thyroid diseases.

Results

Top Results Author & Article Info References

Data presented in Table 2 are from seven recent, large series of patients who underwent fine-needle aspiration [17, 19, 24, 30, 35, 40, 46]. Altogether, 18 183 specimens were obtained and examined cytologically. In the seven series, the rate of benign cytologic results ranged from 53% to 90% (average, 69%), and the rate of malignant cytologic findings ranged from 1% to 10% (average, 3.5%). The rate of suspicious or indeterminate cytologic results ranged from 5% to 23% (average, 10%). The rate of nondiagnostic cytologic results varied from 2% to 21% (average, 17%). Caruso and Mazzaferri [11] recently reported similar average rates from 10 series including 9119 patients: benign, 74%; malignant, 4%; and inadequate or suspicious, 22%. With the exception of one study [24], it appears that in most institutions, the rate of nondiagnostic results varies from 15% to 20%. Although the number of nondiagnostic aspirates decreases with experience, it is clear that even with repetitive aspirations, a residual 10% of the smears remain unsatisfactory [12, 17]. However, in more than two thirds of all patients (approximately 75% to 80%), a definitive diagnosis of either benign or malignant disease is suggested by the biopsy results.

Yield of Cancer

Yield is defined as the ratio of the total number of patients with carcinoma to the total number of cases operated. Detection of malignancy in the nodular thyroid on the basis of clinical evaluation, thyroid scanning, and ultrasonography resulted in a 10% to 15% incidence of malignancy at surgery [2, 3, 9, 12, 14]. For example, Ashcraft and Van Herle [8], in a review of 22 series, reported that the use of thyroid scanning resulted in detection of malignancy in 16% of cold nodules. The application of fine-needle aspiration biopsy resulted in an increased frequency of carcinoma in excised thyroid nodules. As shown in Table 3, 3144 of 18 183 patients (17%) were referred to surgery and 995 patients (32%) proved to have malignancy. The yield of cancer varied from 17% to 51% in the series summarized in Table 3. These results correlate well with the findings of recent studies in which fine-needle aspiration biopsy increased the yield of thyroid cancer to 20% to 50% [11-14]. The increased yield of thyroid cancer is not attributable to the presence of small malignant tumors found incidentally in the thyroid gland at the time of surgery.

False-Negative Diagnosis

False-negative errors are worrisome because they imply missed malignant lesions. False-negative diagnoses may occur because of sampling error [17, 46] or interpretive mistakes [17, 26, 46]. Regardless of the cause, for fine-needle aspiration biopsy to be considered a useful and reliable diagnostic technique, false-negative results must be acceptably low. It is essential that this problem be carefully reviewed.

The false-negative rate is defined as the percentage of patients with "benign" cytologic findings who are confirmed to have malignant lesions of the thyroid. The true frequency of false-negative results is available only in series in which all patients screened by fine-needle aspiration had subsequent surgery with histologic review [8, 12, 14, 47]. Reported false-negative rates in the seven series summarized in Table 3 ranged from 1.3% to 11.5%, with an average rate of 5.2%. However, when data from the series are combined, only 10% of patients with benign cytologic findings underwent thyroid surgery. Caruso and Mazzaferri [11] found an identical false-negative rate of 5% based on pooled data from 10 series (range, 1% to 6%) in which only 14% of nodules were excised. Campbell and Pillsbury [14], analyzing combined data from 912 patients with benign cytologic results who had a histologic examination, found a false-negative rate between 0.5% and 11.5%, with a pooled rate of 2.4%. Ashcraft and Van Herle [8] noted that false-negative results varied in reported series from 2% to 50% and that among 1330 patients, all of whom had a histologic examination, the false-negative rate was 1.7%. It is clear that unless all patients undergoing fine-needle aspiration biopsy also have a histologic examination, the true false-negative rate will not be known. Most authorities agree that the true false-negative rate is below the 5% rate if all patients undergo thyroid surgery. However, based on our review of the data, a false-negative rate between 1% and 5% appears to be unavoidable.

In addition to histologic (operative) verification, another way of evaluating missed malignancies is by following patients with negative cytologic results. In two recent studies, investigators searched for thyroid cancer in such patients. Boey and colleagues [47] followed 365 patients with benign cytologic results for a mean of 30 months and found only two cancers; the overall false-negative rate was 2.1%. Grant and colleagues [48] studied 439 patients who underwent fine-needle aspiration and had benign cytologic results but who did not undergo thyroid surgery; during a follow-up of 6.1 years, malignancy occurred in 3 patients, for a calculated false-negative rate of 0.7%.

False-Positive Diagnosis

False-positive reports do not constitute a major problem with fine-needle aspiration. The false-positive rate is the percentage of patients with "malignant" fine-needle aspirates who are found to have benign lesions at surgery. The false-positive rate varied from 0% to 7.7% (average, 2.9%) (Table 3). These results are similar to the 0.5% rate reported earlier by Ashcraft and Van Herle [8], the 1.2% rate reported by Campbell and Pillsbury [14] on the basis of a literature survey, and the rate of less than 6% reported by Caruso and Mazzaferri [11]. Overall, 1 in 40 patients with a "benign" fine-needle aspiration biopsy specimen may have a malignant lesion, and 1 in 100 patients with a "malignant" fine-needle aspiration biopsy specimen may actually have benign disease [14].

Sensitivity and Specificity

Two values used to estimate the accuracy of fine-needle aspiration biopsy for malignancy are sensitivity (true-positive results divided by true-positive plus false-negative results) and specificity (true-negative results divided by true-negative plus false-positive results) [25, 31, 35]. The estimation of sensitivity and specificity depends on how the suspicious category is handled [2, 17, 48, 49]. If suspicious cytologic results are considered as positive, then sensitivity will increase and specificity will decrease. On the other hand, if the suspicious results are considered as negative, specificity will increase and sensitivity will decrease. Table 3 shows a sensitivity of 65% to 98% and a specificity of 72% to 100%. These variations reflect how the cytologist interpreted the suspicious smear and how the suspicious cytologic findings were finally entered into the calculations. Nonetheless, an overall sensitivity of 83% and a specificity of 92% in the series summarized in Table 3 translates into a diagnostic accuracy greater than 95% [14]. These data suggest that fine-needle aspiration is slightly more specific than sensitive in detecting thyroid cancer and confirm it as a reliable diagnostic test.

Advantages

Fine-needle aspiration biopsy is safe, simple, and accurate. It is done in an outpatient setting, and repeated aspirations may be done. The procedure should be done by clinicians with a special interest in thyroid diseases. Cytologic interpretation should be done by a pathologist with a special interest and expertise in cytodiagnosis [10, 17, 32]. The procedure has no serious disadvantages.

Limitations

Fine-needle aspiration has two major limitations: nondiagnostic results and suspicious or indeterminate results. Inadequate, or nondiagnostic, smears often occur in the setting of cystic or vascular lesions [50]. Examination of cystic fluid, biopsy of the cyst capsule, or repeat biopsy may produce satisfactory results. Clearly, the success of fine-needle aspiration depends on careful aspiration and smear preparation as well as on cytologic evaluation. Hall and colleagues [46], reviewing the experience at the University of California at Los Angeles (UCLA), found that the frequency of inadequate smears was 15% for hospital-based physicians and 32% for community-based clinicians. On the other hand, the number of years of experience with the procedure did not influence results. These findings imply that the acquisition and maintenance of aspiration skills are more important than the total number of years of experience in influencing the rate of nondiagnostic results.

The second limitation of fine-needle aspiration is the "indeterminate" or "suspicious" result [8, 11, 13, 17, 51, 52]. We previously reviewed this problem and found that in most series, between 20% and 30% of aspirates show a "suspicious" result [44] because of the cytopathologic difficulty in differentiating cellular benign Hurthle- or follicular-cell neoplasms from their malignant counterparts [17]. Approximately 25% of nodules with indeterminate cytologic findings are found to be malignant at surgery. Different ways to minimize suspicious cytologic findings have been suggested. In general, the use of large-needle biopsies [26] or the quantitative analysis of nuclear size and shape or DNA content of thyroid nuclei [53] has not significantly influenced results of aspiration cytology. Radioisotope scanning of the suspicious nodule has also been suggested [9]. The UCLA group reported that 29% of suspicious nodules were "hot" on scans and, therefore, that patients with such nodules should not undergo surgery. However, our experience seems to contradict this finding. We found that 83% of nodules diagnosed as suspicious were hypofunctional or "cold" on radioisotope imaging, and, surprisingly, among patients with functioning nodules, two had a malignant lesion [44].

Certain considerations will reduce the potential for missed malignancies. The pathologist must require an adequate smear for diagnosis and avoid "overdiagnosis" of inadequate smears. Aspirates from multiple sites of the nodule or nodules are more likely to provide representative cytologic results than are several smears from a single aspirate. In evaluating multinodular goiters, dominant or growing nodules should undergo biopsy, although some suggest that all accessible nodules should undergo biopsy [26]. Sampling error is greater with larger nodules (> 4 cm), whereas smaller nodules (< 1 cm) may be missed altogether. In these circumstances, biopsy with ultrasonic guidance improves yield and accuracy [11]. Our opinion is that patients with suspicious (indeterminate) cytologic results should undergo thyroid exploration. A cyst that recurs after repeated aspirations should be removed. Tetracycline hydrochloride or thyroxine-suppression therapy in patients with recurrent cysts is not recommended [50, 54].

Nodule Management

The advantages and disadvantages of various management schemes in patients with thyroid nodules have been examined in several recent reviews [11-14, 55]. Considerations and comparisons have been made regarding the diagnostic precision and the cost effectiveness of different management approaches. Several studies convincingly suggest that fine-needle aspiration biopsy is pivotal in the diagnostic management of nodular thyroid disease, improves care, and decreases cost.

Effect on the Management of Thyroid Nodules

Many studies have shown that fine-needle aspiration biopsy has had a significant effect on the number of thyroid operations and on the incidence of carcinoma in patients undergoing surgery [56-65]. Miller and colleagues [56] reported that the availability of biopsy decreased the percentage of patients undergoing thyroid surgery from 48% to 24%, simultaneously increasing the yield of thyroid cancer from 12% to 14%. Wetzig and Giddings [58] reviewed the utility of fine-needle aspiration biopsy and concluded that in their practice, 50% of patients with solitary thyroid nodules could have been spared surgery by fine-needle aspiration biopsy. Summarizing the Mayo Clinic experience, Hamberger and colleagues [60] reported that after the introduction of fine-needle aspiration, the percentage of patients who had surgery decreased from 67% to 43%, and the percentage of carcinomas in operated patients doubled from 14% to 29% (Figure 1). More recently, Caplan and colleagues [40] found that the introduction of fine-needle aspiration in their practice decreased the percentage of those requiring thyroidectomy from 61% to 33% and that malignancy found at surgery doubled (from 18% to 39%).

View larger version (28K): [in this window] [in a new window]   Figure 1. The effect of fine-needle aspiration biopsy (introduced in 1980) on the total number of thyroid operations and on the percentage of thyroid carcinomas among surgical specimens at the Mayo Medical Center from 1979 through 1990.

Block and colleagues [66] reported that the frequency of finding carcinoma in patients selected for surgery can be increased from between 15% and 30% to approximately 85% by the use of fine-needle aspiration. They also noted, however, that this improved selectivity was accompanied by a simultaneous increase in the total number of thyroidectomies. Our experience [60] and that of others [11, 26, 40, 57] are different. It is true that technical difficulties in evaluating patients with suspicious cytologic findings lead to an increased number of operations. Analyzing our data (Gharib H. Unpublished data), we found that fine-needle aspiration resulted in a 12% increase in the frequency of surgery for benign Hurthle- or follicular-cell neoplasms but that, overall, 25% fewer thyroid operations for benign goiters were done (see Figure 1).

Cost of Care

The issue of cost effectiveness has been addressed by the UCLA [9] and Mayo [60] groups. The economic effect of fine-needle aspiration biopsy appears to be considerable and results in a significant reduction of the overall medical and surgical costs per patient. The present charges for various tests and techniques used in the diagnostic evaluation and the cumulative costs for different diagnostic protocols at our institution are shown in Table 4. As a test, fine-needle aspiration biopsy costs less than either a radioisotope scan or a high-resolution sonogram. If work-up includes a scan, an ultrasonogram, and fine-needle aspiration, then the total cost of nodule evaluation will be approximately $1140. Use of only the scan (cost, $589) or fine-needle aspiration (cost, $541) decreases the total cost per evaluation by half. Ridgway [67] recommended fine-needle aspiration biopsy of all nodules found hypofunctional by scanning. In this scheme, work-up would cost $859 (see Table 4). However, because cytologic examination offers direct and specific information, only the plan with fine-needle aspiration as the first diagnostic test is cost effective. We concur with Caruso and Mazzaferri [11] that "radionuclide scintigraphy is not (the) first choice in evaluating thyroid nodules."

The main effect on cost of care has been the elimination of unnecessary thyroid surgery, which significantly decreases cost of management. Van Herle and associates [9] compared the cost effectiveness of several different approaches to nodular evaluation and concluded that among commonly used diagnostic techniques, the most efficient and least costly approach uses fine-needle aspiration biopsy as the first test. Ten years ago, we estimated that fine-needle aspiration biopsy decreases costs by about $400 to $700 per patient evaluated [60], which translates into a significant savings to the health care system. In an analysis done in 1991, Caplan and colleagues [40] found that the mean cost of care per patient decreased approximately $1300 after the introduction of fine-needle aspiration into their practice, which is consistent with our findings.

Nodule Evaluation and Therapy

A simple and reasonable approach to the evaluation of a patient with nodular thyroid is outlined in Figure 2. Because fine-needle aspiration is the screening test, this plan rests on cytologic results. When cytologic results indicate a benign mass, clinical follow-up with palpation and repeat biopsy in 1 year is recommended. Other indications for repeat fine-needle aspiration include an enlarging nodule, a nodule unresponsive to levothyroxine therapy, and recurrent cysts. Repeat biopsy seldom results in the reversal of a benign diagnosis. In the absence of a growing nodule, a small nodular goiter can be followed with or without treatment. Some investigators have suggested treatment with levothyroxine to arrest nodule growth or to shrink nodules [14, 26, 67]. In a double-blind, placebo-controlled clinical study, we were unable to show nodule shrinkage despite documented thyrotropin suppression in a 6-month period [68]. That study [68] and more recent data [69, 70] raise the question of whether routine suppressive therapy for nodular thyroid disease is wise. With suppressive therapy, a nodule that continues to grow or fails to regress warrants surgical excision; a shrinking nodule can be followed with palpation and repeat biopsy. Recent studies suggested that suppressive doses of levothyroxine can be associated with osteoporosis [11, 67, 70]. When suppressive therapy is initiated, this side effect and other disadvantages of subclinical thyrotoxicosis should be important considerations.

View larger version (14K): [in this window] [in a new window]   Figure 2. Algorithm outlining the evaluation and management of nodular thyroid disease when fine-needle aspiration biopsy is used as the initial diagnostic procedure. FNA = fine-needle aspiration.

If cytologic findings are nondiagnostic, repeat fine-needle aspiration should be done. Small cystic lesions (< 4 cm) that collapse after aspiration can be observed, whereas surgical excision is suggested for cysts that either recur or are larger (> 4 cm), because both are associated with an increased risk for cancer [5, 11]. We do not regard the presence of a cellular cystic fluid or fluid-containing foam cells as benign; these are unsatisfactory results and fine-needle aspiration should be repeated. If repeat aspiration still yields nondiagnostic cytologic findings, management should be based on clinical evaluation, including a thyroid scan, with surgical excision of the cold nodule. With rare exceptions, surgical therapy is recommended for a patient with a cytologic diagnosis of malignancy. The suspicious or indeterminate cytologic findings create controversy because some investigators have suggested observation for low-risk patients and surgery for high-risk patients [67]. Other investigators [9, 11] follow fine-needle aspiration with scintigraphy, observing "hot" nodules and excising "warm or cold" nodules. Because of a 30% chance of malignancy in patients with a "suspicious" aspirate, our recommendation has been surgical exploration.

Concern has been expressed that if fine-needle aspiration is used as the initial diagnostic test, hyperfunctioning ("hot") nodules yielding suspicious aspirates will be subjected unnecessarily to surgical excision [9, 67]. Recently, Liel and colleagues [71] reported that in 21 of 22 patients with solitary hot nodules, fine-needle aspiration biopsy yielded benign (colloid) cytologic results. It appears that this concern is unfounded.

Conclusions

In medical centers that have accrued adequate experience with fine-needle aspiration, biopsy has substantially increased diagnostic precision and has significantly affected thyroid practice. Fine-needle aspiration is inexpensive and, because it improves surgical selectivity, cost of care is reduced considerably. In the occasional patient who shows indeterminate clinical signs but has a benign result in a fine-needle aspirate, surgery is a reasonable approach. Total exclusion of false-negative results is virtually impossible; however, efforts should be made to minimize this problem by obtaining adequate aspirates and a proper cytologic diagnosis. Complications are extremely rare. The ease of doing fine-needle aspiration and its diagnostic accuracy mandate continued use of fine-needle aspiration in the diagnostic management of patients with nodular thyroid disease.