Malignancy in Infracentimetric Thyroid Nodules 67Clinical Research

67

Baskent University, Departmentof Endocrinology and MetabolicDisease, Ankara, Turkey.

Address correspondence toDr. Mustafa Sahin, BaskentUniversitesi Tip Fakultesi,Endokrinoloji ve MetabolizmaBilim Dali, 5. sokak, Bahcelievler,Ankara, Turkey. E-mail:drsahinmustafa@yahoo.com

Endocrine Pathology, vol. 17,no. 1, 67–74, Spring 2006© Copyright 2006 by HumanaPress Inc. All rights of anynature whatsoever reserved.1046-3976/1559-0097 (Online)/06/17:67–74/$30.00

Ultrasound-Guided Fine-Needle Aspiration Biopsyand Ultrasonographic Featuresof Infracentimetric Nodulesin Patients with Nodular Goiter:Correlation with Pathological Findings

Mustafa Sahin, MD, Aysegul Sengul, MD, Zeynep Berki, MD,Neslihan B. Tutuncu, MD, and Nilgun D. Guvener, MD

AbstractWe evaluated the usefulness of ultrasound-guided fine-needle-aspiration biopsy (US-FNAB)for infracentimetric nodules. In addition, we used sonography to assess the risk of malig-nancy of thyroid nodules, and we evaluated the extent of disease in infracentimetriccancers. The cytopathological results of 472 US-FNABs from 207 nodular goiter patients(170 women, 37 men; mean age, 51.5 ± 13.1 yr) seen between 1999 and 2004 werecategorized into five groups: inadequate, benign, suspicious, follicular neoplasm, andmalignant. There were 145 infracentimetric nodules and 327 supracentimetric nodules.All patients underwent surgery. Final histopathological results correlated with cytologicresults. The sensitivity, specificity, positive predictive value, negative predictive value, andaccuracy of US-FNAB for infracentrimetric nodules were 96.3%, 71.2%, 44.8%, 98.8%,and 76.1%; and for supracentimetric nodules, these values were 98.1%, 63.1%, 35.6%,99.4, and 69.1%, respectively. There were no significant differences betweeninfracentimetric and supracentimetric nodules. More thyroid cancer could be detected ininfracentimetric nodules that were hypoechoic or had fine calcification on ultrasonogra-phy (which may be helpful in discriminating which nodules are appropriate for FNAB)than in supracentimetric nodules. However, logistic regression analyses showed that nosingle variable was predictive of malignancy in infracentimetric nodules. The malignancyrate in infracentimetric nodules was 21.4%. In this subgroup, 4 of 31 patients (12.9%)had multifocal tumors at surgery, 3 of 31 had extrathyroidal invasion, and 1 had ametastasis to the lung. In addition, at surgery, 11 of 55 tumors (20%) larger than 1 cmwere multifocal. In conclusion, small tumor size does not guarantee a low risk of thyroidcancer, and US-FNAB may be useful tool for diagnosing malignant infracentimetric nodules.Key Words: Thyroid cancer; infracentimetric nodules; ultrasound-guided fine-needleaspiration; biopsy.

Introduction

Fine-needle-aspiration biopsy (FNAB)is considered the most reliable test fordiagnosing malignant thyroid nodules[1,2]. Detecting nonpalpable thyroid

nodules in the general population isincreasing owing to more widespread useof ultrasound (US) technology [3]. US pro-vides good information about the location,number, size, echo structure, and echo-genicity of thyroid nodules. Controversy

68 Endocrine Pathology Volume 17, Number 1 Spring 2006

exists over whether nonpalpable thyroidlesions should be assessed by FNAB [4–6].Some authors have reported that US-guided fine-needle-aspiration biopsy (US-FNAB) may be valuable in detectingmalignancy in nonpalpable thyroid nod-ules [7,8]. Others have reported poor effi-cacy and low diagnostic accuracy withUS-FNAB for this purpose [9]. Little isknown about the biological behavior ofinfracentimetric thyroid nodules and theextent of disease in these lesions. Recently,Papini and colleagues found that a combi-nation of sonographic features was usefulin predicting malignancy in nonpalpablethyroid nodules [10].

Our main aims in this study were tocorrelate the US findings for infra-centimetric and supracentimetric thyroidnodules with the cytologic and histologicfeatures of these lesions, to assess the preva-lence of cancer in infracentimetric andsupracentimetric thyroid nodules, todetermine the value of US for predictingmalignancy in these nodules. We also com-pared the value of US-FNAB for detectingmalignancy in infracentimetric andsupracentimetric thyroid nodules andassessed the extent of disease in patientswho were found to have thyroid cancerbased on histopathological findings.

Methods

Patients

Between January 2002 and December2005, 207 patients (170 women, 37 men;mean age, 51.5 ± 13.1 yr) with nodulargoiter underwent US-FNAB prior to sur-gery at Baskent University in Ankara, Tur-key. Patients had 472 thyroid lesions,including 145 infracentimetric nodulesand 327 supracentimetric nodules. Eachpatient’s medical records were reviewed,and all US and histopathologic findings for

each nodule were recorded. All nodules0.5 cm or larger were evaluated by US-FNAB. The indications for surgery weresuspected carcinoma (153 patients), mul-tinodular goiter (49 patients), patient’srequest (5 patients). All subjects gaveinformed consent before US-FNAB andsurgery, and the study was approved by thelocal institution’s ethics committee.

Serum levels of thyroid-stimulating hor-mone and free thyroid hormone (FT3 andFT4) were determined with commerciallyavailable IRMA and RIA kits.

Ultrasound Examination,FNAB Technique, and Cytology

A standard US examination of the thy-roid gland was done by a radiologist usinga real-time US scanner unit (SiemensSI-400 or Siemens Elegra; Siemens AG,Munich, Germany) equipped with a7.5-MHz linear transducer. The radiolo-gist then performed US-FNAB on nodulesthat were 0.5 cm or larger. Each of thesenodules was measured in three dimensions,and echo structures (solid, cystic, mixed),echogenicity (hyperechoic, isoechoic,hypoechoic), presence/absence of fine cal-cification, and margin characteristics (well-defined or poorly defined) were recorded.The examiner also assessed for heteroge-neity of the parenchyma and lymph nodeenlargement.

US-FNAB was done using a 21-gaugeneedle and a 10-mL syringe. US guidancewas used to confirm placement of theneedle in the nodule. In partially cysticnodules, only the solid portion wassampled. Each aspirate was smeared on aslide, immediately fixed in 95% ethanol,and stained using the Papanicolaoumethod.

Cytologic diagnoses were made by anexperienced pathologist who classified eachspecimen as being in one of five catego-

Malignancy in Infracentimetric Thyroid Nodules 69

ries: benign, suspicious of follicular neo-plasm, suspicious of cancer, consistent withcarcinoma, and insufficient for cytologicdiagnosis. A sample was considered suffi-cient if the slide showed six or more clus-ters of follicular cells and each clustercontained at least 10 follicular cells [11].

Surgery and Surgical Specimens

All 207 patients underwent surgery.If multiple nodules were detected on US,the patient underwent total or near-totalthyroidectomy. Individuals with suspiciouslymph node enlargement had the node(s)dissected.

Surgical specimens were cut and thenfixed for macroscopic examination. In eachcase, multiple 5-mm slices of the wholethyroid gland were examined to accuratelydetermine the dimensions of the malignantlesions and assess whether they were mul-tifocal. Each nodule specimen was fixed inbuffered formalin, embedded in paraffin,and stained with hematoxylin and eosin forhistologic study. For each neoplastic lesion,the tumor size, nodal involvement, and anymetastases (local or distant) were recorded.

Statistical Analyses

For each nodule, the US-FNAB resultswere compared with histopathologicalfindings in the surgical specimen. Sensi-tivity, specificity, positive predictive value(PPV), negative predictive value (NPV),and accuracy of US-FNAB were deter-mined for the infracentimetric nodules andfor the supracentimetric nodules. Statisti-cal analyses were done using SPSS software(Statistical Package for the Social Sciences,version 11.0, SSPS Inc, Chicago, IL, USA).Frequency distributions of the US featuresof thyroid cancer in the different groupswere compared using the chi-square andFisher exact tests. Correlations between

variables and histopathologic results wereanalyzed using the Pearson correlationcoefficient. Relative risk of malignancy wasevaluated by logistic regression analysis.p values less than 0.05 were considered sta-tistically significant.

Results

Characteristics of Study Population

A total of 207 nodular goiter patients(170 women, 37 men; mean age, 51.5 ±13.1 yr) was evaluated at the Endocrinol-ogy and Metabolism Department ofBaskent University in Ankara, Turkey.US-FNAB was performed prior to surgeryon 472 nodules in these patients. None ofthe patients had had any radiation therapyto the head or neck. Study patients werefrom an iodine-deficient area of Turkey [12].

Cytologic Findings of FNAB

All 472 nodules were aspirated underultrasonographic guidance. The US-FNABresults of 145 infracentimetric noduleswere as follows: 80 (55.1%) were benign,9 (6.2%) were diagnosed as suspicious offollicular neoplasm, 44 (30.3%) were sus-picious of cancer, 5 (3.4%) were consis-tent with carcinoma, and 7 (4.8%) wereinadequate for cytologic diagnosis neo-plasm (Table 1). The US-FNAB results of327 supracentimetric or centimetric nod-ules were as follows: 165 (50.4%) werebenign, 45 (13.8%) were diagnosed as sus-picious of follicular neoplasm, 88 (26.9%)were suspicious of cancer, 16 (4.9%) wereconsistent with carcinoma, and 13 (4 %)were inadequate for cytologic diagnosis(Table 1).

Table 2 shows the diagnostic index ofUS-FNAB according to nodule size. Sen-sitivity, specificity, positive predictive value,negative predictive value, and accuracy of

70 Endocrine Pathology Volume 17, Number 1 Spring 2006

US-FNAB in infracentimetric nodules was96.3%, 71.2%, 44.8%, 98.8%, and76.1%, respectively. Sensitivity, specificity,positive predictive value, negative predic-tive value, and accuracy of US-FNAB incentimetric or supracentimetric noduleswere 98.1%, 63.1%, 35.6%, 99.4%, and69.1%, respectively.

Patients with malignant infracentimetricnodules were older than were patients withbenign infracentimetric nodules (54.6 ±13.3 vs 48.4 ± 11.3 yr, p = 0.005).

There were no significant differences inage between patients with supracentimetricmalignant nodules and patients withbenign supracentimetric nodules (p = 0.440).There was no difference in mean thyroid

stimulating hormone (TSH) values inmalignant and benign supracentimetricand infracentimetric nodules.

Ultrasonographic Findings

The mean size of the benign infra-centimetric nodules was 7.4 ± 1.8 mm, andthe mean size of the malignant infra-centimetric nodules was 6.9 ± 1.7 mm. Themean size of the benign supracentimetricnodules was 22.2 ± 10.9 mm, and themean size of the malignant infracentimetricnodules was 19.2 ± 7.6 mm. There wereno significant differences in nodule size ormalignancy in supracentimetric andinfracentimetric nodules (p = 0.67 and p =0.052 respectively; Table 3).

The rate of malignancy on histo-pathology was significantly higher in infra-centimetric nodules with punctuatecalcification than it was in those withoutpunctuate calcification [55.6% (5/9) vs19.1% (26/136), p < 0.022]. On sono-graphy, the rate of malignancy on histo-pathology was significantly higher incentimetric or supracentimetric noduleswith parenchymal heterogenicity [27.7%(36/130) vs 9.6% (19/197), p < 0.002]than it was in those with parenchymalhomogenicity and lymph node enlarge-ment [37.5% (12/32) vs 14.6% (43/295),p < 0.0001] (Table 3). However, there wereno significant differences in the rates ofmalignancy among nodules with regard tomargin, existence of an incomplete haloaround the nodule, echo structure, orechogenicity (Table 3).

Logistic regression analyses demon-strated no independent risk factors formalignancy in infracentimetric nodules.On histologic analysis, in centimetric orsupracentimetric nodules, parenchymalheterogenicity (OR, 3.25; p < 0.0001) andlymph node enlargement (OR, 3.39; p <0.002) were shown to be independent riskfactors for malignancy.

Table 1. Comparison of Cytologic and Histopathological Results in InfracentimetricNodules and Centimetric or Supracentimetric Nodules

Histopathology

Cytology Benign Carcinoma Total

Infracentimetric nodulesBenign 79 1 80Suspicious 26 18 44Follicular neoplasm 6 3 9Malignant 3 2 5Inadequate 7 7Total 114 31 145

Centimetric or supracentimetric nodulesBenign 164 1 165Suspicious 50 38 88Follicular neoplasm 41 4 45Malignant 5 11 16Inadequate 12 1 13Total 272 55 327

Table 2. Diagnostic Index of US-FNAB According to Nodule Size

Infracentimetric Centimetric or supracentimetric

Sensitivity 26/27 (96.3%) 53/54 (98.1%)Specificity 79/111 (71.2%) 164/266 (63.1%)PPV 26/58 (44.8%) 53/149 (35.6%)NPV 79/80 (98.8%) 164/165 (99.4%)Accuracy 105/138 (76.1%) 217/314 (69.1%)

Nodules with inadequate cytologic results were excluded, and suspicious cytologic resultswere considered malignant. PPV, positive predictive value; NPV, negative predictive value.

Malignancy in Infracentimetric Thyroid Nodules 71

Pathological Findings After Surgery

In infracentimetric nodules in patientswith differentiated thyroid cancer (papil-lary and follicular thyroid carcinoma),extrathyroidal extension was observed in9.7% (3/31), metastasis was observed in3.2% (1/31), and multifocal tumors wasobserved in 12.9% (4/31) of patients (Table 4).Of infracentimetric nodules, 25.8% (8/31)had either metastatic or multifocal orextrathyroidal invasion (Table 4). One40-yr-old female patient with lung metastaseshad a Hürthle cell carcinoma on histo-pathology. She received a total of 300 mCiof radioiodine after total thyroidectomy.

In patients with thyroid cancer, incentimetric or supracentimetric nodules,extrathyroidal extension was observed in9.1% (5/55), distant metastases wereobserved in 3.6% (2/55), regional lymphnode metastases were observed in 14.5%(8/55), and multifocal tumors were observedin 20% (11/55) of patients (Table 4).

Among infracentimetric cancers, 28 werepapillary thyroid cancer, 1 was an invasivefollicular variant papillary thyroid cancer,and 2 were Hürthle cell carcinoma. Amongsupracentimetric cancers, 42 were papillarythyroid cancer, 7 were follicular variantpapillary thyroid cancer, 1 was Hürthle cell

Table 3. Characteristics of Patients and Ultrasonographic Findings of Thyroid Nodules According to NoduleSize (Infracentimetric, Centimetric, or Supracentimetric) and the Results of Histopathology

Infracentimetric nodules Centimetric or Supracentimetric nodules

Benign Malignant p Benign Malignant p

Number (n) 114 31 272 55Age (yr ± SD) 48.4 ± 11.3 54.6 ± 13.3 0.005 51.5 ± 13.1 50.1 ± 13.0 0.440Sex, male/female 3/33 1/30 0.152 20/65 14/41 0.067TSH, mU/L 1.35 ± 2.2 1.15 ± 1.2 0.59 0.67 ± 1.13 1.15 ± 2.01 0.012Size of nodule, mean (mm) 7.4 ± 1.8 6.9 ± 1.7 0.67 22.2 ± 10.9 19.2 ± 7.6 0.052Margin

Well-defined 112 29 0.2 264 54 0.53Poorly defined 2 2 8 1

Incomplete haloPresent 7 2 0.61 238 48 0.55Absent 107 29 34 7

Echo structureCystic 6 1 0.79 16 1 0.10Solid 106 29 224 46Mixed 2 1 32 8

EchogenicityIsoechoic 37 4 0.08 135 19 0.12Hyperechoic 8 2 33 8Hypoechoic 69 25 104 28

Fine calcificationPresent 4 5 0.022 79 12 0.17Absent 110 26 193 43

Lymph node enlargementPresent 13 3 0.540 20 12 0.002Absent 101 28 252 43

Parenchymal heterogenicityPresent 50 18 0.115 94 36 0.0001Absent 64 13 178 19

TSH, thyroid stimulating hormone; SD, standard deviation.

72 Endocrine Pathology Volume 17, Number 1 Spring 2006

carcinoma, 4 were follicular thyroid carci-noma, and 1 was medullary thyroid carcinoma.

Discussion

Our results demonstrate that the rate ofdiagnosis or suspicion of cytologic malig-nancy was 33.7% (49/145), and the histo-logically proven rate was 21.4% (31/145)in infracentimetric thyroid nodules. Thereare numerous debates on the clinical sig-nificance of thyroid microcarcinomas.Whether clinically nonpalpable thyroidnodules should be routinely assessed by US-

FNAB is unclear. Leenhardt and co-workers[9] reported that the adequacy rate was85% in 335 nodules (70 operations) 1 cmin diameter or larger and 69% in 115 nod-ules (24 operations) 1 cm in diameter orsmaller. Kim and co-workers [13] con-cluded that US-FNAB is a useful tool fordetermining the treatment plan of non-palpable solid nodules. In our study, allnodules had a final diagnosis by histo-pathology; sensitivities, specificities, andaccuracies between the infracentimetric andsupracentimetric nodules were comparable.

Many studies have shown that inciden-tally detected thyroid papillary carcinomastake an indolent course, but some reportshave shown that small thyroid carcinomasmay lead to death due to local or distantmetastases [14]. Yokozawa and colleagues[15] reported that 15.9% of all carcino-mas smaller than 1 cm had extrathyroidalinvasion. Recently, Papini and colleagues[10] found that 33.3% (4/12) of thyroidcarcinomas smaller than 1 cm had extra-thyroidal invasion and 25.0% (3/12) ofmicrocarcinomas had lymph node inva-sion. Nam-Goong and colleagues [16]found extrathyroid invasion and/orregional metastases were present in 69%of 36 occult differentiated thyroid cancers.In our study, after pathological examina-tion, 8 of 31 patients (25.8%) had eithermetastatic or multifocal or extrathyroidalinvasion. These findings suggest that lowrisk may not be guaranteed by smallernodule size.

Nam-Goong and co-workers [16] alsofound some ultrasonographic characteristicssignificantly associated with subsequentlyconfirmed diagnoses of malignancy. In ourstudy, malignancy was observed in infra-centimetric nodules more frequently if thenodule had microcalcification than if thenodule did not have calcification. How-ever, in infracentimetric nodules, othersonographic characteristics were not asso-

Table 4. Comparison of Thyroid Cancers for Lymph Node Metastases, Multifocality,Extrathyroid Invasion, and Distant Metastases According to Size

Lymph node metastases

Negative Positive Total

Infracentimetric 31 (100%) 1(3.2%) 31CarcinomasSupracentimetric 47 (85.5%) 8 (14.5%) 55, p = 0.023CarcinomasTotal 78 (90.7%) 9 (10.5%) 86

Extrathyroid invasion

Negative Positive Total

Infracentimetric 28 (90.3%) 3 (9.7%) 31CarcinomasSupracentimetric 50 (90.9%) 5 (9.1%) 55CarcinomasTotal 78 (90.7%) 8 (9.3%) 86, p = 0.604

Metastasis

Negative Positive Total

Infracentimetric 30 (96.8%) 1 (3.2%) 31CarcinomasSupracentimetric 53 (96.4%) 2 (3.6%) 55CarcinomasTotal 83 (96.5%) 3 (3.5%) 86, p = 0.706

Multifocality

Negative Positive Total

Infracentimetric 27 (87.1%) 4 (12.9%) 31CarcinomasSupracentimetric 44 (80%) 11 (20%) 55CarcinomasTotal 71 (82.6%) 15 (17.4%) 86, p = 0.301

Malignancy in Infracentimetric Thyroid Nodules 73

ciated with a subsequently confirmeddiagnosis of malignancy.

In conclusion, although there remainsdebate concerning the value of routine thy-roid ultrasonography for detecting nodules,pathological analyses of guided aspirationof sonographically detected nonpalpablenodules demonstrated a high rate of can-cers in our study. Moreover, in our study, asignificant number of patients withmicropapillary carcinomas had locallyadvanced disease at the time of surgery.Papillary microcarcinoma was associatedwith a 1% disease-related–mortality rate,a 5% lymph-node–recurrence rate, and a2.5% distant-metastasis rate in one study[17]. This suggests that small size, per se,cannot guarantee low risk in incidentallyfound thyroid cancers. But previous studieshave shown that papillary microcarcinomahas an excellent prognosis after surgery[18]. These lesions are frequently detectedas incidental findings at autopsy or in sur-gical specimens [19–21]. Ito and colleaguesreported preliminary observational datathat papillary microcarcinomas are not fre-quently apparent, although they are patho-logically multifocal and involve lymphnodes in high incidence [22].

Long-term observation of patients withlocally advanced, small thyroid cancers andtheir risk-stratification is required todetermine an appropriate scheme fordiagnostic and/or therapeutic approachesfor this population.

References

1. Gharib H, Goellner JR. Fine-needle aspira-tion biopsy of the thyroid: an appraisal. AnnIntern Med 118:282–289, 1993.

2. Gharib H. Fine-needle aspiration biopsy ofthyroid nodules: advantages, limitations, andeffect. Mayo Clin Proc 69:44–49, 1994.

3. Ezzat S, Sarti DA, Cain DR, Braunstein GD.Thyroid incidentalomas. Prevalence by pal-

pation and ultrasonography. Arch Intern Med154:1838–1840, 1994.

4. Giuffrida D, Gharib H. Controversies in themanagement of cold, hot, and occult thyroidnodules. Am J Med 99:642–650, 1995.

5. Mazzaferri EL, de los Santos ET, Rofagha-Keyhani S. Solitary thyroid nodule: diagnosisand management. Med Clin North Am72:1177–1211, 1988.

6. Hegedus L. Thyroid ultrasound. EndocrinolMetab Clin North Am 30:339–360, viii–ix,2001.

7. Khurana KK, Richards VI, Chopra PS,Izquierdo R, Rubens D, Mesonero C. The roleof ultrasonography-guided fine-needle aspira-tion biopsy in the management of nonpalpableand palpable thyroid nodules. Thyroid 8:511–515, 1998.

8. Hagag P, Strauss S, Weiss M. Role of ultra-sound-guided fine-needle aspiration biopsy inevaluation of nonpalpable thyroid nodules.Thyroid 8:989–995, 1998.

9. Leenhardt L, Hejblum G, Franc B, et al.Indications and limits of ultrasound-guidedcytology in the management of nonpalpablethyroid nodules. J Clin Endocrinol Metab84:24–28, 1999.

10. Papini E, Guglielmi R, Bianchini A, et al. Riskof malignancy in nonpalpable thyroid nod-ules: predictive value of ultrasound and color-Doppler features. J Clin Endocrinol Metab87:1941–1946, 2002.

11. Baloch ZW, Sack MJ, Yu GH, Livolsi VA,Gupta PK. Fine-needle aspiration of thyroid:an institutional experience. Thyroid 8:565–569, 1998.

12. Erdogan G, Erdogan MF, Emral R, et al.Iodine status and goiter prevalence in Turkeybefore mandatory iodization. J EndocrinolInvest 25(3):224–228, 2002.

13. Kim SJ, Kim EK, Park CS, Chung WY, OhKK, Yoo HS. Ultrasound-guided fine-needleaspiration biopsy in nonpalpable thyroid nod-ules: is it useful in infracentimetric nodules?Yonsei Med J 30:635–640, 2003.

14. Noguchi S, Yamashita H, Murakami N,Nakayama I, Toda M, Kawamoto H. Smallcarcinomas of the thyroid. A long-term fol-low-up of 867 patients. Arch Surg 131:187–191, 1996.

15. Yokozawa T, Fukata S, Kuma K, et al. Thy-roid cancer detected by ultrasound-guidedfine-needle aspiration biopsy. World J Surg20:848–853; discussion 853, 1996.

74 Endocrine Pathology Volume 17, Number 1 Spring 2006

16. Nam-Goong IS, Kim HY, Gong G, et al.Ultrasonography-guided fine-needle aspira-tion of thyroid incidentaloma: correlationwith pathological findings. Clin Endocrinol(Oxf ) 60:21–28, 2004.

17. Chow SM, Law SC, Chan JK, Au SK, Yau S,Lau WH. Papillary microcarcinoma of thethyroid-Prognostic significance of lymph nodemetastasis and multifocality. Cancer 98(1):31–40, 2003.

18. Hay ID, Grant CS, van Heerden JS, GoellnerJR, Ebersold JR, Bergstralh EJ. Papillary thy-roid microcarcinoma: a study of 535 casesobserved in a 50-year period. Surgery 112:1139–1147, 1992.

19. Yamamoto Y, Maeda T, Izumi K, Otsuka H.Occult papillary carcinoma of thyroid. A studyof 408 autopsy cases. Cancer 65:1173–1179,1990.

20. Harach HR, Franssila KO, Wasenius VM.Occult papillary carcinoma of the thyroid:a “normal” finding in Finland. A systematicautopsy study. Cancer 56:531–538, 1985.

21. Piersanti M, Ezzat S, Asa SL. Controversiesin papillary thyroid microcarcinoma of thethyroid. Endocr Pathol 14(3):183–191, 2003.

22. Ito Y, Uruno T, Nakano K, et al. An observa-tion trial without surgical treatment in patientswith papillary microcarcinoma of the thyroid.Thyroid 13(4):381–387, 2003.