Intended for healthcare professionals

Practice Rational Imaging

Incidental thyroid nodule

BMJ 2009; 338 doi: (Published 04 March 2009) Cite this as: BMJ 2009;338:b611
  1. Chirag N Patel, specialist registrar in radiology 1,
  2. Georgina Gerrard, consultant clinical oncologist2,
  3. Andrew F Scarsbrook, consultant radiologist and nuclear medicine physician1
  1. 1Departments of Radiology and Nuclear Medicine, St James’s University Hospital, Leeds Teaching Hospital, Leeds LS9 7TF
  2. 2Department of Clinical Oncology, St James’s University Hospital, Leeds Teaching Hospital, Leeds LS9 7TF
  1. Correspondence to: A F Scarsbrook andrew.scarsbrook{at}

    Thyroid nodules may be coincidentally detected as a result of neck imaging, and most are benign; the authors discuss how best to decide which nodules are most likely to be malignant and require further evaluation

    Learning points

    • Thyroid nodules are very common, and the overwhelming majority are benign

    • Thyroid nodules may be coincidentally detected as a result of the increasing use of neck imaging

    • No imaging modality can reliably differentiate between benign and malignant thyroid nodules, but ultrasonography is the best technique for evaluating nodules and triaging those that require histological evaluation

    • Ultrasound guided fine needle aspiration cytology is accurate, cheap, and safe in most cases, but local resources vary

    • Metabolically active thyroid nodules detected on fused positron emission tomography and computed tomography (PET-CT) are associated with a 1 in 3 chance of malignancy and should be further evaluated

    • PET-CT may have a future role in reducing unnecessary surgery in patients with indeterminate cytology because of its high negative predictive value

    The patient

    A 78 year old man presented with haemoptysis and was found to have a bronchial tumour at bronchoscopy. He was referred for a 18fluorine fluorodeoxyglucose PET-CT scan (which combines positron emission tomography and computed tomography) for accurate staging before surgical treatment; this scan confirmed operable disease. An incidental solitary thyroid nodule was detected, with the scan showing increased uptake of 18fluorine fluorodeoxyglucose within the nodule (fig 1).


    Fig 1 Half body PET-CT scan. The axial CT scan (top left), PET scan (top right), and fused PET-CT scan (bottom left) show abnormal focal uptake of 18fluorine fluorodeoxyglucose in the left lobe of the thyroid gland (indicated by red “crosshairs”). The maximum intensity projection of the PET scan (bottom right) shows abnormal focal uptake of 18fluorine fluorodeoxyglucose in the thyroid gland (long arrow) and two adjacent foci of abnormal 18fluorine fluorodeoxyglucose activity in the right lung in keeping with the primary lung carcinoma and an adjacent right hilar node (short arrow)

    What is the next investigation?

    Incidental thyroid nodules are usually detected as a consequence of the increasing use and higher resolution of neck imaging with ultrasound, computed tomography, positron emission tomography, and magnetic resonance imaging (MRI). Published data from autopsy series suggest that about 50% of the adult population have thyroid nodules, although data from ultrasound screening studies suggest that the prevalence may be as high as 67%1—substantially higher than the 4-8% of cases that are clinically palpable.2

    The overwhelming majority of thyroid nodules are benign and unlikely to be of any clinical significance. A much smaller percentage will represent thyroid carcinoma of which the commonest type (papillary carcinoma) has an excellent prognosis with a 30 year survival rate of 95%.3 The incidence of malignant disease in non-palpable nodules—that is, those likely to be detected on cross sectional imaging—is controversial. Although studies suggest similar rates of malignant thyroid disease for non-palpable and palpable nodules,4 the clinical significance of detecting early microcarcinomas that have an indolent course and excellent prognosis remains debatable.

    No evidence based consensus has been reached on how best to manage incidental non-palpable thyroid nodules. No single imaging modality can differentiate benign and malignant disease accurately, and pathological evaluation of all nodules would not be cost effective and could cause unnecessary distress to most patients. Therefore a rational approach about who to investigate further is crucial for identifying the small number of patients with thyroid malignancy without overinvestigating the much larger number with benign disease.

    Clinical assessment, on the basis of history and/or clinical examination, is an important first step to identifying which patients might be at a higher risk of malignancy and therefore require imaging (box). The relative strengths and weaknesses of each imaging technique are summarised in the table.

    Clinical features suggestive of thyroid malignancy1:

    • Age (<20 years or >60 years )

    • Symptoms of dysphagia or dysphonia (local invasion)

    • Previous history of neck irradiation

    • Prior or family history of thyroid carcinoma

    • Firm, hard, or immobile nodules

    • Presence of cervical lymphadenopathy

    • Rapidly growing nodules

    Comparison of imaging modalities for evaluating thyroid nodules

    View this table:

    Thyroid function tests

    Thyroid stimulating hormone can be used as a screening test in patients with a solitary thyroid nodule for detecting hyperfunctioning adenomas that cause a suppressed thyroid simulating hormone level. Hyperfunctioning nodules account for about 10% of solitary thyroid nodules and are almost always benign.3 Nearly all patients with thyroid cancer are euthyroid. If any clinical risk factors are present and the level of thyroid simulating hormone is normal, further imaging evaluation may be advisable.


    High resolution ultrasonography is highly sensitive for detecting thyroid nodules. Although ultrasonography cannot definitely distinguish between benign and malignant nodules because of an overlap in ultrasound characteristics, certain features favour benign or malignant aetiology. These features may be used to triage patients into those who require no further investigation and can be reassured and those who should undergo tissue sampling for further evaluation.

    Features such as solid composition, microcalcification, irregular margins, and marked hypoechogenicity are more predictive of malignancy than size criteria alone.4 The presence of at least one malignant feature has an overall sensitivity of 83% and specificity of 74%.5 However, features such as a predominantly cystic mass with well defined margins and no calcification are much more typical of benign disease.1

    Fine needle aspiration cytology

    Fine needle aspiration cytology is a highly accurate and cost effective method of evaluating thyroid nodules that show suspicious features on an ultrasound scan. Cytology has a high negative predictive value (95-98%), but up to 15% of specimens may be indeterminate depending on local expertise and other factors.6 Using fine needle aspiration cytology selectively when investigating thyroid nodules with suspicious ultrasound features can reduce the requirement for excision by at least 25% and doubles the yield of malignancy in those excised.7

    Although fine needle aspiration cytology is safe and easily performed without major complication, potential minor risks include local haemorrhage and inadvertent tracheal puncture, although these are rare. The procedure is contraindicated in patients who have a bleeding diathesis or who are unable to tolerate the procedure. Local expertise and availability of cytologists vary widely, and core biopsy for histological evaluation may be required at some institutions. Pathological examination must be used selectively, and published guidelines advise that fine needle aspiration cytology should be reserved for thyroid nodules over 1 cm in size and with suspicious ultrasound features as no direct evidence exists that early diagnosis of microcarcinomas (<1 cm) improves life expectancy.1 6

    Radionuclide scintigraphy

    Thyroid scintigraphy allows the functional status of a thyroid nodule to be assessed. It is useful in patients with a suppressed thyroid stimulating hormone level for confirming the presence of a hyperfunctioning nodule, in which malignancy is extremely rare. In contrast, solitary hypofunctioning nodules have a 10-15% incidence of malignancy.3 Scintigraphy is generally not helpful for determining which nodules are more likely to be malignant in euthyroid patients. Functional imaging may be useful in patients with indeterminate cytology as the demonstration of tracer uptake in a thyroid nodule virtually excludes malignancy.

    Positron emission tomography-computed tomography

    PET-CT is a new hybrid imaging technique that combines functional (positron emission tomography) and anatomical (computed tomography) images to localise sites of increased metabolic activity more accurately in patients with malignancy. The most commonly used radiotracer, 18fluorine fluorodeoxyglucose (a glucose analogue), is a sensitive marker of increased glucose metabolism, which occurs in a variety of malignant tumours. PET-CT scanning is commonly used in patients with potentially operable lung and oesophageal cancer and in the evaluation of patients with lymphoma.

    Focal 18fluorine fluorodeoxyglucose uptake in a thyroid nodule is seen as a coincidental finding in some patients, and, although not specific, it is associated with an increased risk of thyroid malignancy (up to 35% of solitary thyroid nodules showing uptake are malignant), and further evaluation with ultrasonography and fine needle aspiration cytology is essential.8 Although the positive predictive value is relatively low, PET-CT scanning has a reported negative predictive value of 100%.9 This could potentially help to reduce, by up to two thirds, the number of unnecessary thyroidectomies performed for benign disease in patients with indeterminate cytology.9

    In the UK the use of PET-CT scanning is limited by availability and strict referral criteria, but the cost savings from reducing unnecessary surgery would offset the high cost of the investigation. Once the technique becomes more widely available, this may be a valuable future use.

    Computed tomography and magnetic resonance imaging

    Cross sectional imaging with CT or MRI is not particular sensitive for detecting intrathyroidal lesions and cannot reliably distinguish between benign and malignant nodules. As a result, each has a limited role in the evaluation of incidentally discovered thyroid nodules but may be valuable in the staging of patients with proved thyroid malignancy. Although cross sectional imaging is not part of routine staging for thyroid cancer, it is useful for assessing substernal thyroid masses and for evaluating local extension, cervical lymphadenopathy, and distant metastatic disease.


    Before surgery for the patient’s lung cancer, ultrasonography confirmed the presence of a solitary thyroid nodule that had irregular margins and peripheral calcification (fig 2). Ultrasound guided fine needle aspiration cytology was performed at the same time, and cytology was suspicious for malignancy. The patient went on to have curative resection of his lung carcinoma and thyroidectomy with radioiodine ablation for follicular thyroid carcinoma. He has remained disease-free to date, with no evidence of recurrence of either tumour.


    Fig 2 Axial image from a high resolution ultrasound scan of the thyroid showing the thyroid nodule with peripheral calcification and posterior acoustic shadowing (arrow). This is an unusual appearance of thyroid malignancy and is more usually seen in benign nodules. Microcalcification in a nodule (not seen in this case) would be much more suspicious.


    Cite this as: BMJ 2009;338:b611


    • This series provides an update on the best use of different imaging methods for common or important clinical presentations. The series advisers are Fergus Gleeson, consultant radiologist, Churchill Hospital, Oxford, and Kamini Patel, consultant radiologist, Homerton University Hospital, London.

    • Contributors: AFS had the original idea. AFS and GG selected the patient. CNP and AFS searched the literature. CNP wrote the paper. AFS and GG reviewed and edited the paper and reassessed the literature. AFS is the guarantor.

    • Funding: None.

    • Competing interests: None declared.

    • Provenance and peer review: Commissioned; externally peer reviewed.

    • Patient consent not required (patient anonymised, dead, or hypothetical).


    View Abstract