Thyroiddiseases are common problem in surgical practice and even now been manage atprimary care level. However, the problem comes in managing thyroid nodules. Inthe era of technology, the treatment of thyroid problem especially in a groupof well differentiated tumour give a good prognosis and overall survival. Inview of this, the early detection of thyroid nodule is imperative for earlyevaluation and treatment.
The initial evaluation should always include completehistory and physical examination focusing in features suggestive of malignancy.Tissue biopsy and imaging modalities should thoroughly be done to improve thepre-operative diagnosis and provide good treatment options for patients withthyroid nodules. In Malaysia,only a few studies focused on thyroid nodule and risk associated with thyroidmalignancies. Sothi et al reported only 31 cases of thyroid malignancy reportedbetween 1985-1989 with highest incidence in 5th and 6thdecade (Sothy et al., 1991).
The demographic data in this study isslightly different with mean age of 44.36 years (SD14.69).
In present study,the age of patient ranged from 16-85 years old comparable to other previousstudy. The peak incidence is at 4th -6th decade of life.Another study done by Htwe et al describe in Kelantan between 1994–2004, 28.1%of 1,480 thyroid lesions were neoplastic, the incidence of cancer was 3.
5 per100,000 admission and thyroid cancer made up 4.9% of all cancers seen inhospital admission (Htwe, 2012).Many literatureshave stated thyroid incidence are more common in female. Epidemiologicalstudies indicated approximately 5% of women and 1% of men resident in iodinesufficient area have palpable thyroid nodule. This is a bit different with ourstudy whereby women with thyroid nodule are 10 times higher compare to man(Table 1). From final histopathology report, majority are benign (68.
4%) andmalignant case is 31.6% (Figure 3). 52.5% were solitary nodule followed by33.2% multinodular and diffuse goitre is 14.3%. the proportion of disease issimilar with other study.(Sinna andEzzat, 2012).
Gharib et al suggested that the most sensitive testavailable to detect thyroid lesions is High-resolution ultrasound.(Gharib et al.,2010).Some author even suggested ultrasound is an independent predictor formalignancy in indeterminate thyroid nodule (Cheung et al.).
USG features suggestive of malignant growth including presence of solidcomponent, hypoechogenicity, microcalcifications, increase vascularity,lobulated or irregular margins, infiltrative margins, taller-than-wide shape ontransverse view.(Haugen et al.,2016) Laura et alconcluded the presence of suspicious US features did not discriminate malignantfrom benign nodules. The risk of thyroid cancer of nodule more than 4 cm withno suspicious US features was 20 %.
The false negative rate of benign cytologywas 10.4 %, and the absence of suspicious US features did not reliably excludemalignancy. At minimum, thyroid lobectomy should be strongly considered for allnodules > 4 cm (Wharry et al.
, 2014) In our study, 55% of initial assessment from USG isbenign/low risk, followed by suspicious/intermediate risk (15%) and thecategory high risk/high suspicion ultrasound finding is 13.3%. The rest was notdone pre-operatively. Among reasons are patient was undergoing CT scan of neckrather than USG, presence of ulcer, and patient refusal.
From benign USGfinding, 25 of 160 (15%) of the final histology were malignant. It was slightlyhigher as compare to ATA which reported in low suspicious thyroid ultrasoundthe expected risk of malignancy was 5-10%. (Haugen et al., 2016).
Expected risk for intermediate/ suspiciousgroup is 10-20%. In this study, it is 19/45(42.2%) and for high suspiciousgroup, we found that 32/40(80%) were malignant and within the expected riskbetween 70-90% in other literature.(Haugen et al., 2016) the difference of result is due to differentgrading/ grouping system used by radiologist to stratify risk of malignancy.
Some author has suggested a standardize reporting method as demonstrate by AACE/AME/ETA:In 2007, theNational Cancer Institute (NCI) hosted the NCI ‘Thyroid Fine-Needle AspirationState of the Science Conference’ in Bethesda, and subsequently published ‘TheBethesda System for Reporting Thyroid Cytopathology’ in 2008.(Edmund S.Cibas and Syed Z. Ali, 2009) Its aim was to establish comprehensiveguidelines regarding terminology and morphological criteria in reportingthyroid FNAC. It delineates the reporting of ‘suspicious for malignancy’ as anaspirate containing some features of malignancy, but lacking definitivediagnostic changes. Many studies have proved that FNAC is a diagnostic tool forthyroid nodule due to its sensitivity and specificity. However, we want toexamine the agreement between cytology and histology with regards to localdata.
The complete sensitivity in our study was moderate (63%) as the positivepredictive value of malignancy is 61%, a specificity of 81% and negativepredictive value of 82%. Although its lower than other study, it still withinaccepted range (table 10). The wide difference may due to number of cases,classification of cytology, and diagnostic category (Popoveniucand Jonklaas, 2012)Positivepredictive value (PPV) is a correctly malignant histopathology from pre-operative malignant cytology express as a percentage of all cytological diagnosedmalignancies. The higher percentage is a good indicator to demonstrate thereliability of the test itself. In our study, PPV is 61% which lower as compareto other study but still within accepted range. A study in Egypt reported PPVof 94.6%(Sinna andEzzat, 2012) whereby other study is 98.
6% (Sangalli et al., 2006). One of the reason the low PPV value in thisstudy is due to large number of inadequate sample from histology which turn outto be malignant post-operatively.Inadequate samples may be because of largeareas of cystic degeneration or necrosis and sclerotic or calcified lesions. To increase the PPV value, other measureshould be implied such as to repeat FNAC or use an adjunct such as USG guidedFNAC to minimised inadequate sample in pre-operative cytology.Benign cytologyin whom malignant lesions are later confirmed on histopathology is known as falsenegative rate (FNR).
Several studies have shown conflicting data regardinglarge thyroid nodules with the accuracy of benign cytology. Meko et al. found afalse-negative rate of 17 % (5/30) in nodules >3 cm.(Meko andNorton, 1995). Another published article examined 223patients with nodules >4 cm and identified a false-negative rate of 13 %(9/71) as well as higher incidence of cancer in larger nodules.
(McCoy et al., 2007) Both studies recommended that large thyroidnodules should be considered for surgery, regardless of FNA results. Incontrast, other author reviewed data from 743 ultrasound-guided FNA specimenswith benign cytology.
Twenty percent (145/743) underwent thyroidectomy and onlyone false-negative result (0.7 %) was identified on final pathology.(Porterfield et al.
, 2008) They concluded that with appropriateaspiration and expert cytopathologic interpretation, the false-negative rate ofFNAC is extremely low and that diagnostic resection is unnecessary. However, itis important to note that a most of patients with benign FNAC cytology did notundergo surgery, thereby potentially decreasing the false-negative rate by asignificant margin. Unfortunately, in our study did not mention the size of nodule.However, our entire patient went for thyroidectomy in nodule >1 cm assuggested by many literatures Comparison ofresults of present study with various previous studies is shown in Table 10,demonstrating that thyroid FNAC is a reliable screening test and a valuablemethod of distinguishing neoplastic from non-neoplastic nodules preoperatively.However, owing to some limitations of FNAC, it is recommended that surgicalindications must not depend solely on cytology. Indeed, the results of medicalhistory, physical examination, laboratory tests, and ultrasonography shouldalso be evaluated simultaneously.
Due to high value of false negative and falsepositive sample, the accuracy of this study is slightly lower (75%) as comparedto other study, Al- Sayer et al have accuracy of 92%, Afroze et al (94.5%), buthigher compare to Cusick et al 69%. It can be postulated that accuracy isreducing with increasing number of patient. Chung et al suggested it can bereduced by doing it at different site, repeat FNAC if inadequate sample, USGguided FNAC and more than one cytopathologist involved in reviewing the slide.(Chung-CheCharles Wang et al., 2011)Suspiciouscytology findings account for 3.7–11 % all aspirates.(Baynes et al.
, 2014; Gharib et al., 1993; Raj et al., 2010) Our incidence of suspicious cytology was 8.6%, therefore within range of this reported incidence. Suspicious cytologycarries a risk of malignancy of between 29 and 75 %(Bongiovanni et al.
, 2012; Cibas and Ali, 2009; Gharib et al., 1993) Of our suspicious FNAC results, 16 of 26(69.2 %) were malignant on subsequent histology. This is the expected 60–75 %reported in the Bethesda guidelines. Of these, the diagnosis of papillarythyroid cancer was the most common in 13 cases (81.0 %), followed by follicular(3,18%), anaplastic and micropapillary (1,6%).
One study concluded surgicalremoval of the nodules should be considered strongly as the incidence ofmalignancy in suspicious lesions was high. Mundasad et al also concludedin their study that suspicious and intermediate results prove to be an area ofuncertainty, often resolved by diagnostic surgical resection. (Mundasad et al., 2003)It has beenthought that patients with hyperthyroidism are less likely to have cancer.(Sokal, 1954).
The incidence of incidental thyroid cancerin patients with toxic nodular goiter (TNG) has been estimated to beapproximately 3%.(Kang et al., 2002) In fact, the current recommendation from theATA is that a thyroid scan should be obtained as the initial test in a patientwith a thyroid nodule and a low serum thyroid stimulating hormone (TSH) level.(Cooper et al.
, 2009) If the nodule is found to behyperfunctioning, FNAC is not recommended as its rarely related to cancer. Overthe last 30 years, there has been a dramatic increase in the overall incidenceof thyroid cancer, from 3.6 per 100,000 in 1973 to 8.
7 per 100,000 in 2002.(Davies andWelch, 2006) However, it is unclear if this significantincrease in the incidence of thyroid cancer has also occurred in patients withTNG. Two recent studies have reported higher rates of thyroid cancer, 15.6% and18.3% in patients with TNG, suggesting that the rate of malignancy for TNG isunderestimated.
(Smith et al., 2013a; Smith et al., 2013b) This has resulted in the hypothesis that factorsleading to the increased incidence of the thyroid cancer in the generalpopulation may also be affecting patients with TNG. However thyroid toxicitydid not have significant risk for malignancy in our study as shown by Simple Logistic Regression model the crude ratiois 1.We able todemonstrate that thyroid malignancy was marginally significant with increasingage (in years) with Crude OR 5.22 (p =0.046). Many literatures has described the incidence of thyroid cancer isincreasing in elderly (Raffaelli et al.
, 2010) between 2.5% and 12% of differentiatedthyroid cancer occur in individuals older than 65 years.(Amatoet al., 2013) A retrospective study conducted by Lin et al analyzed 204 thyroidcancer patients aged 60 years and older; 142 (70%) thyroid cancers were welldifferentiated and Fifty-nine (29%) of the thyroid cancers were poorlydifferentiated.(Linet al., 2005) In a more recent retrospective analysis of 1022 patients undergoingthyroidectomy, the rate of malignancy was 68.
7% in a group of 45 year old andolder. well-differentiated thyroidcancer and lymph node metastasis occurred more often in patients younger than50 years while micropapillary carcinoma occurred more often in patients 50years or older(Doet al., 2014)