Thyroid Nodules & Thyroid Cancer

Meet the patient

A 42-year-old woman felt a lump on the left of her neck while fastening a necklace. It is painless, moves when she swallows, and she is otherwise well — clinically and biochemically euthyroid. Her GP ordered a thyroid function test and an ultrasound "to be sure".[1][2]

Roughly one in two of us is carrying a thyroid nodule on ultrasound, and only 5 to 10 percent of clinically significant nodules are malignant. So the single question that matters at her first visit is the risk-stratification fork: which nodule gets a needle, which gets watched, and which is cancer. Hold that fork and the entire topic slots into place.[1]

Two cell lineages, four cancers — the embryology that decides everything

The cell a thyroid cancer comes from dictates its marker, its spread, its treatment and its prognosis — so name the lineage first. Follicular epithelial cells (the iodine-trapping, thyroglobulin-making majority) give rise to the differentiated cancers — papillary and follicular — and to their undifferentiated end-point, anaplastic. Parafollicular C cells (neural-crest-derived, calcitonin-secreting) give rise to medullary carcinoma.[1]

The differentiated cancers share three therapeutic assets — they produce thyroglobulin, retain the sodium-iodide symporter (so they take up radioactive iodine), and respond to TSH suppression with levothyroxine. Medullary cancer has none of these: it makes calcitonin and CEA, ignores iodine, and is driven by RET. That single embryological fact is why the management of medullary cancer is its own algorithm.[1]

Malignant nodules divide along that cell of origin — and examiners expect you to separate them on behaviour, marker and driver:[1]

[1]

The numbers that frame the whole topic — most are reassuring, one (anaplastic) is not:[1]

[1]

How common, and why a nodule is usually good news

Thyroid nodules are among the commonest of all clinical findings, and almost all of them are benign. Palpable prevalence is 5 to 10 percent of adults; ultrasound prevalence reaches 30 to 50 percent and climbs past 60 percent in older cohorts; autopsy finds nodules in about half of all glands. With a pre-test probability this firmly benign, the discipline is not diagnosis — it is selective triage of the malignant minority.[1][2]

Thyroid cancer is the commonest endocrine malignancy, but the steep rise in incidence is a story of detection, not disease. Over three decades the age-adjusted rate has climbed almost entirely on the back of small papillary carcinomas picked up by wider neck ultrasound, carotid Doppler, CT and PET — a wave of over-diagnosis. Mortality has stayed flat, confirming most newly "found" cancers were always there and always indolent. Women are affected two to four times more often; differentiated cancer peaks at 25 to 65, anaplastic after 60.[1]

The risk factors split into the ones that change the histological mix and the ones that change your threshold to biopsy:[1]

[1]

Iodine status shifts the histological mix — the one epidemiological fact worth a viva line. Endemic-goitre regions lean toward follicular and anaplastic carcinoma; iodine-sufficient (or supplemented) populations toward papillary. Population iodination has been shown to push the mix toward papillary and away from the more aggressive subtypes.[2]

The bedside round — what raises the malignancy odds

Examination in a thyroid nodule rarely finds the cancer; it finds the features that change your pre-test probability. A focused thyroid consult takes five minutes and sets the whole work-up. Take a structured history — age and sex, childhood neck radiation, family history of thyroid cancer or MEN-2, tempo of growth, compressive or voice symptoms — and ask specifically about flushing, diarrhoea and hypertension (MEN-2) and hypercalcaemia (parathyroid).[1]

Examine from the front, palpate from behind while the patient swallows. The thyroid (and any nodule in it) rises with deglutition, which separates it from a fixed non-thyroid neck mass. Document the nodule's size, consistency (soft, firm, hard), surface (smooth, irregular), mobility (mobile, fixed), tenderness, then systematically palpate the cervical node chains (levels I to VI), assessing for tracheal deviation, stridor, and — by voice quality or flexible laryngoscopy — vocal-cord function.[1]

The high-yield clinical risk features that push a nodule toward the needle:[1]

• History of childhood neck irradiation, family history of thyroid cancer or MEN-2, age under 20 or over 70, male sex (cancer is rarer in men but more often aggressive when present).
• Growth — rapid enlargement of a previously stable nodule is ominous, and a sentinel clue to anaplastic transformation.
• Consistency — a firm or fixed nodule adherent to trachea or strap muscles.
• Voice — hoarseness, sign of recurrent laryngeal nerve involvement and one of the strongest single predictors of malignancy.
• Nodes — palpable cervical lymphadenopathy, especially levels II to VI.
• Compression — dysphagia, dyspnoea or stridor from a large or rapidly growing mass.[1]

Everyone forgets the atypical presentations that bite. Medullary cancer may announce itself as the calcitonin syndrome (flushing, secretory diarrhoea) or as part of MEN-2 — screen for phaeochromocytoma, primary hyperparathyroidism and the MEN-2B marfanoid phenotype with mucosal neuromas. Anaplastic cancer presents dramatically as a rapidly enlarging, hard, painful mass in an older patient with systemic decline. Thyroid lymphoma arises almost exclusively in long-standing Hashimoto thyroiditis. Metastases to the thyroid (kidney, breast, lung, melanoma) are rarer still but worth a thought in the right patient.[1]

Read the ultrasound like a radiologist — the suspicious five

High-resolution ultrasound is the single most useful test in thyroid nodule medicine, and the ACR TI-RADS score is how it speaks. The American College of Radiology Thyroid Imaging, Reporting and Data System assigns points across five feature categories and converts the sum into a risk level (TR1 to TR5) that, combined with nodule size, decides whether to biopsy, follow up or ignore.[3]

The five feature categories — Composition, Echogenicity, Shape, Margin, Echogenic foci — sum to give the TR level. Memorise the categories first (the "CE-SME" five), then the high-scorer inside each.[3]

Here is the memory device that earns marks — the VTEP cluster: four features score three points each and are the heaviest hitters, Very hypoechoic, Taller-than-wide, Extra-thyroidal extension, Punctate echogenic foci. Two more score two points — Solid composition and Irregular or lobulated margin. Say it as one line: VTEP scores three, SI scores two. Any single VTEP feature almost lands a nodule at TR3 on its own.[3]

The benign-versus-suspicious face-off, with the discriminator that decides the needle:[3]

[3]

One-line discriminator: taller-than-wide plus punctate echogenic foci is papillary carcinoma until proven otherwise. Cysts and spongiform nodules, at the other extreme, are benign in over 97 percent of cases — leave them alone.[3]

The FNA thresholds are a clean number rule — the 1-1.5-2.5 rule. TR1 and TR2 are never biopsied. TR3 is biopsied at at least 2.5 cm (follow-up at least 1.5 cm). TR4 at at least 1.5 cm (follow-up at least 1 cm). TR5 at at least 1 cm (follow-up at least 0.5 cm). Smaller TR3 to TR5 nodules may still be biopsied for high-risk clinical features, suspicious nodes, or extra-thyroidal extension — and a suspicious cervical node (round, cystic, lost hilum, punctate foci) is an indication to biopsy regardless of nodule size.[3]

Reveal — name the four VTEP features and why each scores three

Very hypoechoic (darker than strap muscle), Taller-than-wide (anteroposterior more than transverse on transverse view — it grows against the plane of least resistance, a mark of malignancy), Extra-thyroidal extension (capsular breach), and Punctate echogenic foci (psammoma bodies, the calcified hallmark of papillary cancer). Each is high-yield enough that the ACR weights it at three points — one VTEP feature alone puts a nodule on the suspicious side of the ledger.[3]

TSH first, ultrasound second, FNA third — the investigation ladder

The investigation of a thyroid nodule is a strict four-step sequence — and step one is the one juniors skip. Serum TSH first, then high-resolution ultrasound with TI-RADS, then fine-needle aspiration reported by Bethesda, then ancillary tests (calcitonin, cross-sectional imaging, molecular panels) as indicated.[1][2]

Step 1 — Serum TSH

Serum TSH is the first test in every thyroid nodule, full stop. A normal or raised TSH prompts ultrasound risk-stratification. A suppressed TSH with raised free T4 and T3 points to an autonomous (hot) nodule — which carries a much lower malignancy risk — and the next step is radionuclide scintigraphy (technetium-99m pertechnetate or iodine-123) to confirm a tracer-avid nodule, which generally does not need FNA. Within the normal range, a TSH in the upper half slightly raises the probability of malignancy.[1]

The classic trap: a hot nodule suppresses TSH and is low-risk for cancer. Do not rush the FNA on a suppressed-TSH nodule — confirm autonomous function with scintigraphy first. The needle is for the cold nodule. Reach for it the wrong way round and you biopsy a benign toxic adenoma while the real decision (whether it needs antithyroid drugs, radioiodine or surgery) goes unaddressed.[1]

Everyone forgets: a pure cyst (anechoic with comet-tail artefact) is almost always benign — aspirate it only if symptomatic, do not send it for cytology that will come back nondiagnostic. And in a child, a pregnant patient, or anyone with palpable cervical nodes, the FNA threshold drops — biopsied earlier and more liberally.[1]

Step 2 — High-resolution ultrasound with ACR TI-RADS

Ultrasound is where the fork is decided. Apply ACR TI-RADS (above) to every nodule over the FNA threshold; cysts and spongiform nodules are usually benign and watched; solid hypoechoic nodules with VTEP features get the needle. Cross-sectional imaging is reserved for large nodules, suspected invasion, retrosternal extension, bulky nodes, or anaplastic cancer — and iodinated-contrast CT is avoided in any patient who may receive radioactive iodine, because the iodine load blocks RAI uptake for weeks; use MRI instead.[1][3]

Step 3 — Fine-needle aspiration cytology (Bethesda System)

Fine-needle aspiration under ultrasound guidance is the cornerstone of nodule triage. Cytology is reported by The Bethesda System for Reporting Thyroid Cytopathology (TBSRTC) — six categories, each carrying an implied malignancy risk and a linked management recommendation. Learn it as a ladder you climb, not a list you recite.[4]

Every thyroid FNA lands on one rung of the six-tier Bethesda ladder:[4]

[4]

The 2017 Bethesda revision nudged the implied risks upward, and the reason is a single reclassification worth naming: NIFTP — non-invasive follicular thyroid neoplasm with papillary-like nuclear features. A tumour previously called encapsulated follicular-variant papillary carcinoma is now regarded as very-low-risk and non-malignant, managed by lobectomy alone. That one change lowered apparent cancer rates across Bethesda III, IV and V without changing a single outcome.[4]

For the troublesome indeterminate zone (Bethesda III and IV), molecular testing refines risk and can spare surgery: gene-expression classifiers (Afirma GSC) and next-generation sequencing panels (ThyroSeq, ThyGeNEXT and ThyraMIR) sort nodules into low-risk (observe) and high-risk (operate) groups, cutting diagnostic lobectomies by about half.[1]

Step 4 — Ancillary tests

The tests you reach for when cytology is not enough, or when the cancer type demands them:[1]

• Serum calcitonin — order if medullary cancer is suspected (family history, cytology suggesting MTC, or as a baseline screen). A basal calcitonin over 100 pg/mL (or a stimulated rise after pentagastrin or calcium) is highly suggestive; CEA is a tumour-burden marker in MTC and is also raised in anaplastic cancer.[1]
• Cross-sectional CT or MRI of neck and chest — for large nodules, suspected local invasion, retrosternal extension, bulky nodal disease, or anaplastic cancer to map extent and plan surgery. Avoid iodinated contrast if RAI is planned; MRI is the alternative.[1]
• Vocal-cord assessment — flexible laryngoscopy before any thyroid surgery if the patient is hoarse.
• Baseline thyroglobulin — measured before definitive treatment of differentiated cancer; it cannot distinguish benign from malignant, but is the non-secretion reference after thyroidectomy and RAI.
• RET genetic testing and family screening — for any medullary cancer patient and first-degree relatives; positive carriers proceed to prophylactic thyroidectomy timed to the mutation risk class.[1]

Why follicular cancer hides from the needle — the molecular pathways

Each thyroid cancer runs a distinct, therapeutically actionable molecular pathway — and one of them explains why FNA cannot diagnose follicular cancer. Most nodules are benign (colloid nodule, simple or haemorrhagic cyst, follicular adenoma, multinodular goitre, focal thyroiditis). The malignant ones follow well-characterised molecular routes.[1]

Papillary carcinoma (PTC) is driven by the MAPK pathway — most often BRAF V600E (about 45 percent) and RET-PTC rearrangements (about 15 percent, classically after radiation), with RAS in a smaller fraction. The cascade produces the diagnostic nuclear morphology — orphan Annie-eye (ground-glass) nuclei, nuclear grooves and intranuclear pseudoinclusions — and the laminated calcifications called psammoma bodies. PTC spreads lymphatically to cervical nodes (central level VI, then lateral III to IV) and only rarely haematogenously to lung and bone.[1]

Follicular carcinoma (FTC) runs a different pathway — RAS mutations and the PAX8-PPAR-gamma fusion. Its hallmark is capsular and vascular invasion, which is exactly why it cannot be diagnosed on FNA: the needle samples cells, not architecture. FTC spreads haematogenously to bone and lung, characteristically sparing nodes early. The Hürthle-cell (oncocytic) variant is more aggressive and less avid for radioactive iodine.[1]

Medullary carcinoma (MTC) arises from parafollicular C cells and is defined by a gain-of-function RET proto-oncogene mutation — germline in the familial forms (MEN-2A, MEN-2B and familial MTC, together about 25 percent), somatic in roughly half of sporadic cases. The RET M918T mutation (exon 16) is the hallmark of MEN-2B and the most aggressive phenotype. C cells make calcitonin (a sensitive marker) and CEA (a burden marker), and the stroma typically contains amyloid. MTC produces no thyroglobulin and concentrates no iodine.[1]

Anaplastic carcinoma (ATC) is undifferentiated — pleomorphic spindle, giant and squamoid cells — and is thought usually to arise by dedifferentiation of a pre-existing papillary or follicular cancer, accumulating TP53, TERT promoter and beta-catenin (CTNNB1) mutations that override cell-cycle control (BRAF V600E is often co-present, inherited from the antecedent PTC). It is among the most aggressive of all human cancers, invading trachea, oesophagus, great vessels and skin, with a doubling time measured in weeks.[1]

Etymology for viva gold: psammoma is from the Greek psammos, "sand" — the "sand-body" calcification of papillary cancer. Orphan Annie-eye nuclei are named after the blank-eyed orphan of Harold Gray's comic strip, because the optically clear nuclei look like her empty eyes. Both words outlived their origins because the pathology they describe is unchanged.[1]

Reveal — why can follicular carcinoma NOT be diagnosed on FNA?

Because FTC is defined by capsular and vascular invasion — architectural features. FNA samples cells, not the tumour's relationship to its capsule and vessels, so Bethesda IV (follicular neoplasm) triggers diagnostic lobectomy for histology. Papillary cancer, in contrast, is diagnosable cytologically because its nuclear features (orphan Annie-eye nuclei, grooves, pseudoinclusions) are visible in a single aspirated cell.[1][4]

The differential — and the one question that matters

The differential of a thyroid nodule is broad; the role of imaging and FNA is to triage among them, not to list them. Most are benign; a minority are malignant; a critical few are not thyroid at all.[1]

The differential in three columns — benign, malignant, and the neck mass that is not thyroid:[1]

[1]

The clinically distinguishing question is always the same: does this nodule carry enough risk to warrant FNA? The two earlier sections — TI-RADS and Bethesda — are the entire apparatus built to answer it.[1]

When the airway is the emergency — resuscitation and peri-operative prep

Most thyroid nodule and cancer work is outpatient. Acute resuscitation is rare, and lives in three situations.[1]

1. Airway compromise from a large or anaplastic mass — stridor, respiratory distress, or rapid tracheal compression needs urgent airway assessment (awake fibreoptic intubation may be needed), multidisciplinary oncology and palliative input, and may demand emergency tracheostomy or debulking radiotherapy.
2. Neck haematoma after thyroid surgery — a surgical emergency: open the wound at the bedside to release the clot and relieve airway compression, then return to theatre for haemostasis.
3. Bilateral recurrent laryngeal nerve injury — post-extubation stridor may demand re-intubation and sometimes tracheostomy.[1]

Peri-operative preparation is high-yield exam material, and the medullary-cancer step is the one that kills patients when missed. Document vocal-cord function by laryngoscopy before surgery; check serum calcium, phosphate, vitamin D and PTH pre-operatively to anticipate post-operative hypoparathyroidism; render the patient biochemically euthyroid before elective surgery (a hyperthyroid patient risks thyroid storm); exclude phaeochromocytoma first in medullary cancer (24-hour urinary metanephrines or plasma free metanephrines) to avoid a lethal intra-operative hypertensive crisis; group and save, and counsel on the risks of tracheostomy, permanent hypocalcaemia and voice change.[1]

Surgery — lobectomy, total, and the neck

Surgery is the central curative modality for every thyroid cancer, and the extent is chosen on histology and risk. The framework is the 2015 ATA guidelines, and the range runs from diagnostic lobectomy to total thyroidectomy with neck dissection.[1]

The four surgical options, who gets each, and why:[1]

[1]

The classic trap: operating on a medullary cancer patient with an undiagnosed phaeochromocytoma provokes a fatal intra-operative hypertensive crisis. Always exclude phaeo first — it is the single most examinable peri-operative step in the whole topic.[1]

Radioactive iodine and TSH suppression — the differentiated-cancer duo

Radioactive iodine-131 ablation is given after total thyroidectomy to destroy residual thyroid tissue and micrometastases, enable sensitive thyroglobulin surveillance, and reduce recurrence in intermediate- and high-risk differentiated cancer. Patients are prepared either by thyroid hormone withdrawal (to raise endogenous TSH) or by recombinant human TSH (rhTSH, Thyrogen) injections, which avoids iatrogenic hypothyroidism. Activity is risk-stratified — low-activity ablation (around 30 mCi, 1.1 GBq) for low-to-intermediate risk, and higher-activity therapy (100 to 150 mCi) for high-risk disease with nodal metastases, extra-thyroidal extension or distant metastases. RAI has no role in medullary or anaplastic cancer — those cells do not concentrate iodine.[1]

After total thyroidectomy every patient needs lifelong levothyroxine — typically 1.6 micrograms per kilogram per day as a starting adult dose, titrated to a risk-adapted TSH target. The 2015 ATA guidance stratifies suppression intensity:[1]

• High-risk cancer — TSH below 0.1 mU/L indefinitely, balancing against bone and cardiac toxicity.
• Intermediate-risk — TSH 0.1 to 0.5 mU/L for several years, then relaxed.
• Low-risk — TSH in the low-normal range (0.5 to 2.0 mU/L); aggressive suppression is unnecessary and harmful.[1]

Everyone forgets that over-suppression is itself a harm. Excess TSH suppression causes osteoporosis (especially post-menopausal women) and atrial fibrillation (especially older patients) — a compelling reason to relax suppression once recurrence risk has fallen. Differentiated thyroid cancer is usually indolent; over-treating a low-risk patient with maximal surgery, RAI and full suppression is a real, preventable harm.[1]

The 2015 ATA guidelines formalised dynamic risk stratification — an initial staging estimate (TNM-8 and the ATA response-to-therapy categories) updated at each follow-up by thyroglobulin trends, neck ultrasound and, when needed, cross-sectional imaging or diagnostic iodine scans. Excellent responders (undetectable stimulated thyroglobulin, normal ultrasound) move to lighter surveillance; biochemical incomplete (rising thyroglobulin) or structural incomplete (imaging-positive) responders escalate to further local or systemic therapy.[1]

The trials that tamed advanced disease

For RAI-refractory and medullary disease, multitargeted tyrosine-kinase inhibitors have transformed a near-hopeless situation into a chronic-disease conversation. Three phase III trials anchor the modern systemic pathway.[5]

For RAI-refractory differentiated cancer, lenvatinib (24 mg daily) and sorafenib (400 mg twice daily) both improve progression-free survival; lenvatinib produced a striking 18.3 versus 3.6 month median PFS in the SELECT trial.[5]

[1] [5]

For medullary cancer, RET- and VEGFR-directed TKIs are the systemic mainstay. Vandetanib (300 mg daily) and cabozantinib (140 mg daily) both prolong progression-free survival; cabozantinib's effect is independent of RET mutation status.[6][7]

[1] [6] [1] [7]

Two newer targets complete the modern picture. For anaplastic cancer with a BRAF V600E mutation (about half of cases), dabrafenib plus trametinib (BRAF plus MEK inhibition) has produced remarkable responses in a disease that was until recently uniformly fatal within weeks. For RET-mutant medullary cancer, the highly selective RET inhibitors selpercatinib and pralsetinib are emerging as first-line targeted therapy.[2]

The four cancers up close — know the subtype

Papillary thyroid carcinoma (PTC)

PTC is 80 to 85 percent of thyroid cancers, peaks at 25 to 65, is two to three times commoner in women, and is the cancer most strongly linked to childhood neck radiation. Variants include the classic type, the follicular variant (partly reclassified as NIFTP when encapsulated and non-invasive), the aggressive tall-cell, columnar-cell and hobnail variants, and the diffuse sclerosing variant. Standard treatment is total thyroidectomy plus or minus prophylactic central neck dissection, then selective RAI ablation and risk-adapted TSH suppression — and active surveillance of low-risk papillary microcarcinoma (T1a, no invasion) is now a legitimate first choice.[1]

Follicular thyroid carcinoma (FTC)

FTC is 5 to 10 percent, is commoner in iodine-deficient regions, presents slightly later than PTC, and spreads haematogenously to bone and lung. It is defined by capsular and vascular invasion, so it cannot be diagnosed on FNA — Bethesda IV triggers diagnostic lobectomy to inspect architecture. It divides into minimally invasive (excellent prognosis) and widely invasive (worse). The Hürthle-cell (oncocytic) variant is more aggressive, less RAI-avid, and managed with total thyroidectomy plus RAI and lifelong thyroglobulin surveillance.[1]

Medullary thyroid carcinoma (MTC)

MTC is 2 to 5 percent of thyroid cancers but a disproportionate share of thyroid-cancer deaths, because it is biologically more aggressive and does not respond to RAI or TSH suppression. About 25 percent are hereditary (MEN-2A, MEN-2B, familial MTC) from germline RET mutations; sporadic cases carry somatic RET mutations in roughly half. Every new MTC patient must have RET genetic testing and family screening — identifying an asymptomatic carrier allows prophylactic thyroidectomy (infancy for MEN-2B, early childhood for MEN-2A and FMTC).[1]

Surgery is the only cure: total thyroidectomy plus routine central (level VI) and ipsilateral lateral (levels II to V) neck dissection, with contralateral lateral dissection if nodal disease is present. Exclude phaeochromocytoma first. Surveillance is with serial calcitonin and CEA — doubling times are strong prognostic markers, and a calcitonin doubling time under six months is ominous. For progressive metastatic disease, vandetanib, cabozantinib, or selective selpercatinib (RET-mutant) prolong PFS.[6][7]

Anaplastic thyroid carcinoma (ATC)

ATC is only 1 to 2 percent of thyroid cancers but causes a third to a half of all thyroid-cancer deaths. It arises in older patients (over 60) as a rapidly enlarging, hard, painful, fixed mass, often with stridor, dysphagia, hoarseness and systemic decline. Histology shows undifferentiated pleomorphic spindle and giant cells; TP53, TERT promoter and beta-catenin mutations are characteristic, with BRAF V600E often co-present (dedifferentiation from a prior papillary cancer). All ATC are stage IV at diagnosis. Treatment is multimodal where resectable — surgery (rarely curative, often debulking), external-beam radiotherapy and chemotherapy (taxanes) — and dabrafenib plus trametinib for the BRAF-mutant subgroup.[1]

TNM-8 — the staging that moved the age cliff from 45 to 55

The 8th edition of the AJCC and UICC TNM staging made three changes that materially improve staging accuracy for differentiated thyroid cancer. First, the prognostic age cut-off moved from 45 to 55 years (the age gradient below 55 is much smaller than once thought). Second, minimal extra-thyroidal extension was removed from T3, leaving only tumour size over 4 cm (T3a) or gross strap-muscle invasion (T3b). Third, N1 disease no longer auto-stages an older patient to stage I, because nodal metastasis in well-differentiated PTC confers a much smaller survival penalty than historically assumed.[1]

Staging by cancer type — the table the examiner wants reproduced verbatim:[1]

[1]

How thyroid-cancer patients come to harm — complications and the classic traps

Surgical complications

Surgery for thyroid cancer is safe in high-volume hands but carries three classic complications every student must know.[1]

• Recurrent laryngeal nerve injury — unilateral causes hoarseness; bilateral causes stridor and may need tracheostomy. Transient in 5 to 10 percent, permanent in 1 to 2 percent with intra-operative nerve monitoring in expert hands.
• Hypoparathyroidism — transient hypocalcaemia in up to 25 percent after total thyroidectomy with central dissection; permanent in 1 to 4 percent. Managed with oral calcium and calcitriol; severe cases need intravenous calcium gluconate. Always check calcium and PTH post-operatively.
• Neck haematoma — rare (under 1 percent) but life-threatening; open the wound at the bedside to release the clot and decompress the airway, then return to theatre.
• Superior laryngeal nerve injury — loss of high-pitch voice and aspiration of liquids.
• Chyle leak, seroma, infection, hypertrophic scar — less common.[1]

Treatment-related complications

• Radioactive iodine — sialadenitis and dry mouth, transient neck pain, nausea, small secondary-malignancy risk (salivary and leukaemia), and pregnancy and breastfeeding restrictions (avoid pregnancy for 6 to 12 months).
• TSH suppression — osteoporosis (post-menopausal women) and atrial fibrillation (older patients); justify only in high-risk disease.
• TKI therapy — hypertension, diarrhoea, fatigue, hand-foot syndrome, QTc prolongation (vandetanib), proteinuria and bleeding risk.[1]

The classic pitfalls

The recurring errors that cost marks — and patients:[1]

• Operating on a medullary cancer patient with an undiagnosed phaeochromocytoma — provokes a fatal intra-operative hypertensive crisis. Always exclude phaeo first.
• Diagnosing follicular carcinoma on FNA — impossible; it needs histology of capsular or vascular invasion. Bethesda IV is a diagnostic lobectomy indication.
• Misreading a thyroglobulin level after RAI — interpret with anti-thyroglobulin antibody status (antibodies cause false-low readings) and only after thyroidectomy and ablation.
• Giving RAI for medullary or anaplastic cancer — useless; those cells do not concentrate iodine.
• Treating iodinated-contrast CT as routine in differentiated cancer — iodine loading blocks subsequent RAI for weeks; use MRI if cross-sectional imaging is needed before ablation.
• Over-suppressing TSH in a low-risk older patient — precipitates atrial fibrillation and osteoporosis without benefit.[1]

Prognosis and disposition — the numbers that frame the conversation

Prognosis by cancer type, with the determinants that move it:[1]

[1]

Differentiated thyroid cancer is among the most curable of all solid tumours. Even patients with cervical nodal metastasis do well — reflected in the AJCC-8 change that keeps node-positive disease stage I in patients under 55. Follicular is slightly worse, especially with vascular invasion or distant metastasis. Medullary prognosis is stage- and calcitonin-doubling-time-dependent; MEN-2B and RET M918T carry the worst outcomes. Anaplastic is almost uniformly fatal within months, though the small BRAF-mutant subgroup now benefits dramatically from dabrafenib and trametinib.[1][2]

Disposition is risk-adapted. Most differentiated cancer patients are managed as outpatients with lifelong structured surveillance — six-monthly thyroglobulin and anti-thyroglobulin antibodies, periodic neck ultrasound, and diagnostic whole-body iodine scans when biochemistry and imaging disagree. Medullary surveillance centres on calcitonin and CEA doubling times. Anaplastic demands rapid inpatient multidisciplinary management and early palliative care. Benign nodules are observed safely, and small low-risk papillary microcarcinomas may even be actively surveilled with excellent long-term outcomes.[1][2]

Special populations — when the thresholds move

Examiners use these scenarios heavily. Each one changes presentation, threshold, or both.[1]

• Pregnancy — fine-needle aspiration is safe and indicated for the same ultrasound features as in non-pregnant patients. If a differentiated cancer is found in the first or early second trimester, surgery can usually be deferred to the postpartum unless there are aggressive features (rapid growth, extra-thyroidal extension, nodes); cancers found later in pregnancy are almost always observed through delivery. Radioactive iodine is absolutely contraindicated in pregnancy and breastfeeding.[1]
• Children and adolescents — nodules are less common but carry a higher malignancy risk (around 20 to 25 percent). Paediatric papillary cancer is more often disseminated at presentation (larger primaries, more nodal and pulmonary metastasis) but remains highly curable. Chernobyl fallout demonstrated the strong radiation-PTC link. RAI doses and surveillance are weight-adapted.
• Known MEN-2 and RET carriers — prophylactic thyroidectomy is performed before medullary cancer develops, timed by the ATA risk class of the RET mutation: infancy for MEN-2B (highest-risk codons, e.g. M918T), before age 5 years for high-risk MEN-2A, and individualised for lower-risk mutations. Calcitonin surveillance begins early; phaeochromocytoma screening is lifelong.[1]
• The elderly patient with a rapidly growing neck mass — assume anaplastic until proven otherwise; secure the airway, image, biopsy urgently, and involve the multidisciplinary team and palliative care early.
• Patients on amiodarone — the iodine load can precipitate both hyper- and hypothyroidism and cause nodular change; check thyroid function before and during amiodarone.

Regional deltas and the over-treatment debate

The over-treatment debate is the single most examinable controversy. The steep rise in incidence from increased imaging has led to real concerns about overtreating indolent papillary microcarcinoma. Active surveillance of low-risk T1a tumours, molecular testing to reduce diagnostic lobectomy, and the NIFTP reclassification are all responses. Also live: the extent of surgery (total thyroidectomy versus lobectomy for low-risk unilateral disease); RAI in low-risk disease (increasingly avoided to spare toxicity); TSH-suppression intensity (relaxed for low-risk disease); and prophylactic central neck dissection (the ATA recommends it be considered, not mandated).[1]

The memory devices and exam pearls

[1] [1]

Quick self-test — what is the first test in a thyroid nodule, and why?

Serum TSH. A normal or raised TSH prompts ultrasound (TI-RADS) and selective FNA; a suppressed TSH points to an autonomous (hot) nodule with a low malignancy risk, which is then assessed by radionuclide scintigraphy. Skipping TSH means you cannot interpret the rest of the work-up.[1]

Quick self-test — what MUST you exclude before operating on a medullary thyroid cancer?

A phaeochromocytoma (MEN-2). Operating on an unrecognised phaeo precipitates a fatal intra-operative hypertensive crisis. Screen with 24-hour urinary fractionated metanephrines or plasma free metanephrines before any surgery; if positive, the phaeochromocytoma is resected first (usually after alpha-blockade).[1][6]

[1]

The mantra

TSH tells you whether to scan, ultrasound tells you whether to biopsy, Bethesda tells you whether to operate — and a hot nodule stays out of the needle.[1][3][4]

Ward-round test — four stems

Reveal — Stem 1: the suppressed-TSH nodule

A 50-year-old woman has a 2 cm right-lobe nodule found on carotid Doppler. Her TSH is 0.2 mU/L (low) with a raised free T4. The registrar books her for ultrasound-guided FNA. What is wrong with that plan? Model: This is a hot (autonomous) nodule, which suppresses TSH and carries a very low malignancy risk — FNA is the wrong first move. Confirm autonomous function with radionuclide scintigraphy (technetium-99m pertechnetate or iodine-123); a tracer-avid nodule generally does not need FNA. The needle is for the cold nodule. Reach for it the right way round.[1]

Reveal — Stem 2: the cytology that cannot close the loop

A 1.8 cm solid hypoechoic nodule in a 45-year-old woman returns Bethesda IV — follicular neoplasm. The registrar tells the patient the biopsy was "not cancer, so we'll watch it." Correct the plan. Model: Bethesda IV carries a 25 to 40 percent malignancy risk, and follicular carcinoma cannot be diagnosed or excluded on cytology — it is defined by capsular and vascular invasion, an architectural feature FNA cannot sample. The correct next step is diagnostic lobectomy for histology (with molecular testing an option to risk-stratify first). Watching it is the trainee error; this is a surgical indication.[1][4]

Reveal — Stem 3: the nodule with a calcitonin tumour

A 38-year-old man has a thyroid nodule, cervical lymphadenopathy, and intermittent flushing and diarrhoea. Calcitonin is markedly raised. Name the cancer, the syndrome to exclude, and the test that must be done before any operation. Model: This is medullary thyroid carcinoma (parafollicular C-cell origin, calcitonin-secreting, amyloid stroma, RET-driven). Exclude MEN-2 — and before any surgery, exclude phaeochromocytoma with 24-hour urinary fractionated metanephrines or plasma free metanephrines. Operating on an unrecognised phaeo precipitates a fatal intra-operative hypertensive crisis. Then total thyroidectomy with central and lateral neck dissection, plus RET genetic testing and family screening.[1][6]

Reveal — Stem 4: the older patient with the hard, growing mass

A 72-year-old man has a rapidly enlarging, hard, painful neck mass over three weeks, with stridor and hoarseness. What is the diagnosis, what is the stage, and what is the first priority? Model: Anaplastic thyroid carcinoma until proven otherwise — undifferentiated, TP53 and TERT driven, arising in the elderly. It is stage IV by definition at diagnosis. The first priority is the airway — secure it proactively (awake fibreoptic intubation or tracheostomy may be needed), image and biopsy urgently, involve the multidisciplinary team and palliative care on the day of diagnosis, and test for BRAF V600E at once for the subgroup that benefits from dabrafenib plus trametinib.[1]

References

1. [1]Haugen BR, Alexander EK, Bible KC, et al. 2015 American Thyroid Association Management Guidelines for Adult Patients with Thyroid Nodules and Differentiated Thyroid Cancer: The American Thyroid Association Guidelines Task Force on Thyroid Nodules and Differentiated Thyroid Cancer Thyroid, 2016.PMID 26462967
2. [2]Grani G, Sponziello M, Filetti S, Durante C. Thyroid nodules: diagnosis and management Nat Rev Endocrinol, 2024.PMID 39152228
3. [3]Tessler FN, Middleton WD, Grant EG, et al. ACR Thyroid Imaging, Reporting and Data System (TI-RADS): White Paper of the ACR TI-RADS Committee J Am Coll Radiol, 2017.PMID 28372962
4. [4]Cibas ES, Ali SZ. The 2017 Bethesda System for Reporting Thyroid Cytopathology Thyroid, 2017.PMID 29091573
5. [5]Schlumberger M, Tahara M, Wirth LJ, et al. Lenvatinib versus placebo in radioiodine-refractory thyroid cancer N Engl J Med, 2015.PMID 25671254
6. [6]Wells SA Jr, Robinson BG, Gagel RF, et al. Vandetanib in patients with locally advanced or metastatic medullary thyroid cancer: a randomized, double-blind phase III trial J Clin Oncol, 2012.PMID 22025146
7. [7]Elisei R, Schlumberger MJ, Müller SP, et al. Cabozantinib in progressive medullary thyroid cancer J Clin Oncol, 2013.PMID 24002501