Thyroid cancers

Overview

• Epidemiology (US 2025): ~44,000 new cases, ~2,300 deaths; incidence has been decreasing over the last decade. F:M ~3:1. 5-yr relative survival ~98%.

Classification

• Cancers of follicular epithelial cells (produce thyroid hormone):
  • Differentiated thyroid cancer (DTC): papillary (PTC, ~85%), follicular (FTC, ~10%), oncocytic carcinoma (formerly Hurthle cell, ~3%). Best prognosis.
  • Poorly differentiated thyroid cancer.
  • Undifferentiated: anaplastic thyroid carcinoma (ATC, <2%).
• Cancer of parafollicular (C) cells (produce calcitonin): medullary thyroid cancer (MTC, ~3%).
• Other (rare): lymphoma (Hashimoto background), sarcoma (primary squamous cell carcinoma is now a subtype of anaplastic thyroid carcinoma in WHO 5th ed).

Risk factors & genetics

• Radiation (childhood neck RT, e.g., Hodgkin treatment; nuclear accident) is the major DTC/PTC risk factor, but >90% of DTC patients have no RT history. Iodine status: low → FTC; high → PTC.
• Hereditary syndromes (~5% of DTC): Gardner syndrome (familial adenomatous polyposis, APC), Cowden syndrome (PTEN; follicular thyroid + breast + endometrial), Pendred, Werner, and Carney complex. Familial PTC is more aggressive than sporadic (multifocal, bilateral, ↑ recurrence).
• Recurrence pattern: PTC recurs locally; FTC and Hurthle cell recur distantly (bone and lungs).
• Molecular: PTC, BRAF V600E (~50%), RET fusions (~10%, esp. radiation-associated), TERT (aggressive); FTC, RAS mutations, PAX8-PPARG fusions; MTC, RET (germline in MEN2A/2B, somatic in ~50% of sporadic); ATC, BRAF V600E, TP53, TERT.
• Poor-prognosis features (DTC): age ≥55 yrs (AJCC 8th ed), male sex, poorly differentiated, tumor >1.5 cm, and extrathyroidal extension (ETE). 10-yr OS ~90%.

Differentiated thyroid cancer (PTC, FTC, Hurthle)

• Nodule workup: TSH first. Low TSH → I-123 scan (hot nodule = unlikely cancer). Normal/high TSH → ultrasound. FNA is the best test for a solitary thyroid nodule.
• US suspicious features: hypoechoic, microcalcifications, taller-than-wide, irregular margins, extrathyroidal extension.
• FNA (Bethesda I to VI): indeterminate (III/IV) → molecular testing (Afirma, ThyroSeq); Bethesda VI (malignant, ~99%) → surgery. FTC cannot be diagnosed by FNA (capsular/vascular invasion needs a surgical specimen).

• Lobectomy: tumors 1 to 4 cm, low risk, no ETE/LN+ (also reasonable in patients <45 y/o with small tumors, avoiding the higher complication rate of total thyroidectomy).
• Total thyroidectomy: >4 cm, bilateral, ETE, LN+, prior RT, distant mets.
• Central (level VI) neck dissection: therapeutic for clinically positive central nodes; prophylactic central dissection may be considered in selected cN0 advanced (T3/T4) PTC, generally not needed in FTC.

• Given ×1 to ablate normal thyroid remnant and destroy microscopic deposits; risk-stratified by ATA criteria. RAI is risk-adapted: recommend for high-risk (gross ETE, distant mets, large-volume N1b/bulky nodal); consider for intermediate-risk (N1, microscopic ETE, vascular invasion, elevated post-op Tg); generally omit in low-risk. No single feature (isolated >4 cm, one Tg value, limited vascular invasion) independently mandates RAI.
• RAI AEs: nasolacrimal duct stenosis, cytopenias, ↑ risk of secondary malignancy (MDS), ↓ fertility; contraindicated in lactation; avoid pregnancy for 6 mo after RAI. IV iodinated contrast is not absolutely contraindicated: use when it improves staging (invasive primary, bulky nodal disease) and delay RAI ~1 to 2 mo for iodine clearance (check urinary iodine if unsure).

• Levothyroxine TSH suppression (TSH stimulates thyroid growth): target <0.1 in high-risk, <0.5 in intermediate, normal in low-risk; suppression for ~5 yrs improved median OS (15 vs 11 yrs). AEs of suppression: ↑ coronary artery disease, atrial fibrillation, and osteoporosis.
• Thyroglobulin (Tg) is an excellent tumor marker once RAI has ablated normal tissue; follow with Tg + anti-Tg antibodies and neck US. At 12 months, if Tg is undetectable, recheck after two doses of rh-TSH; Use response- and risk-adapted surveillance; consider FDG-PET in higher-risk patients with rising or elevated Tg (often >10 ng/mL) and negative iodine imaging, TSH stimulation not routinely required.
• RAI can be used for metastatic disease (CR ~45%) or in recurrence if the last RAI was >1 yr ago.

• Definition (any of): RAI scan negative or progression of RAI-avid tumors; response unlikely once cumulative activity reaches ~600 mCi (relative threshold, not an absolute ceiling); or structural progression/recurrence within 6 to 12 mo of properly administered RAI. Observation is reasonable for slow-growing tumors; cytotoxic chemotherapy is not routine in DTC but may be considered as salvage for rapidly progressive RAI-refractory disease when TKIs fail, preferably on trial.
• Lenvatinib (VEGFR 1 to 3, FGFR 1 to 4, PDGFR-α, RET, KIT; SELECT trial): ORR 65%, mPFS 18.3 vs 3.6 mo placebo; preferred 1L TKI. FDA approved 2015. AEs: HTN (68%), diarrhea (59%), fatigue (59%), anorexia (50%), weight loss (46%).
• Sorafenib (BRAF, RET, VEGFR 1 to 3, c-KIT; DECISION): mPFS 10.8 vs 5.8 mo, ORR ~12%; less active than lenvatinib.
• Cabozantinib (COSMIC-311): post-VEGFR-TKI (after sorafenib/lenvatinib), 2L ORR ~15%, mPFS ~11 mo; FDA approved Sep 2021.
• BRAF V600E+ (PTC): dabrafenib + trametinib (on-label after prior therapy via the Jun 2022 tumor-agnostic BRAF V600E approval) or selumetinib for re-differentiation to restore I-131 uptake.
• NTRK fusion: larotrectinib (thyroid ORR ~71%; ~86% in PTC/FTC), entrectinib (ORR ~60%).
• RET fusion: selpercatinib (LIBRETTO-001, 1L ORR ~100%/CR 12%, 2L ORR ~80%), pralsetinib (2L ORR ~89%). FDA-approved thyroid label.
• ALK fusion (rare): crizotinib, ceritinib, alectinib, brigatinib, entrectinib.
• Pembrolizumab: TMB-high (≥10 mut/Mb), MSI-H, or dMMR tumors.

Medullary thyroid cancer (MTC)

• Neuroendocrine tumor of parafollicular (C) cells. 10-yr OS ~75 to 85%; if localized, total thyroidectomy is curative (10-yr OS ~95%). 80% sporadic (age ~60s), 20% hereditary (MEN2, age 20 to 30s).
• Markers/hormones: calcitonin (key; → diarrhea, flushing) + CEA; ACTH secretion → Cushing syndrome. Pre-op calcitonin correlates with disease extent. Diagnosis by FNA cytology + serum/washout calcitonin and CEA with calcitonin IHC (RAI scan not useful; PET/CT reserved for staging or recurrent disease).
• MEN2: 2A (MTC, pheochromocytoma, parathyroid hyperplasia; RET codon 634) and 2B (MTC, pheochromocytoma, mucosal/intestinal ganglioneuromas, marfanoid; RET M918T). FMTC: MTC-predominant MEN2A variant; associated RET variants span multiple exons (10, 11, 13, 14, 15), not exon 11 alone. Germline RET in ALL MTC patients → screen for pheochromocytoma (24-h urine metanephrines/catecholamines) before thyroid surgery, offer prophylactic thyroidectomy (within the first year of life, ideally first months, for MEN2B/RET M918T; timing by ATA risk category for other RET genotypes), and refer to genetics.
• Surgery: total thyroidectomy + prophylactic bilateral central (level VI) dissection; lateral neck dissection for imaging-proven involvement, contralateral only if calcitonin >200 pg/mL with a positive ipsilateral neck (not a 1-cm cutoff). Post-op levothyroxine to maintain euthyroid (no role for TSH suppression). Monitor CEA and calcitonin; doubling time predicts poor prognosis. RAI, RT, and chemo have ↓ efficacy. Use MRI to monitor liver mets (more sensitive than CT for small tumors). Metastatic disease: observe if asymptomatic/slow-growing; surgery or RT for symptom control.
• Advanced MTC systemic therapy:
  • Selpercatinib (RET; LIBRETTO-531 phase 3, NEJM 2023): in RET-mutant MTC significantly better PFS than vandetanib/cabozantinib; FDA traditional approval Sep 27, 2024. LIBRETTO-001: 2L ORR ~69%, 1L ORR ~73%. AEs: ↑ LFTs, hyperglycemia, leukopenia.
  • Pralsetinib (RET; ARROW): 2L ORR ~60%, 1L ORR ~66%; US MTC indication withdrawn (2023), so it is no longer FDA approved for MTC.
  • Cabozantinib (MET, VEGFR2, RET; EXAM): ORR ~28%, mPFS 11.2 vs 4.0 mo (HR 0.28); FDA approved 2012. Black box: GI perforation.
  • Vandetanib (VEGFR, RET, EGFR; ZETA): ORR ~45%, mPFS 30.5 vs 19.3 mo (HR 0.46); FDA approved 2011. Black box: QT prolongation (torsades, sudden death), REMS program.
  • Pembrolizumab: TMB-high, MSI-H, or dMMR.

Anaplastic thyroid cancer (ATC)

• Very aggressive, mOS ~3 to 6 mo without targeted therapy; ~age 60s, F = M; up to ~50% carry BRAF V600E. Presents as a rapidly growing mass invading surrounding organs (dysphagia, hoarseness, hemoptysis; asphyxia is the most common cause of death). 20 to 50% present with metastatic disease (most often lung). Differentiate histologically from thyroid large-cell lymphoma.
• Staging: all ATC is stage IV. IVA = T1-T3a N0 M0; IVB = T1-T3a N1 M0 or T3b/T4 any N M0; IVC = any T any N M1 (resectability is a separate assessment, not part of staging).
• Workup: urgent BRAF testing plus comprehensive NGS in all ATC.
• Local control: total thyroidectomy for intrathyroidal disease, tracheostomy if airway compromise; RT (ORR ~80% but ↑ local recurrence; limited concurrent chemoRT data with doxorubicin). No role for RAI or TSH suppression.
• BRAF V600E+ (~30 to 50%): dabrafenib + trametinib (BRAF/MEK; FDA approved May 2018), ORR ~56%, median PFS 6.7 mo, median OS 14.5 mo, 12-mo PFS ~43% and 12-mo OS ~52% (mature 36-pt ROAR cohort); dramatic responses may convert unresectable to resectable.
• NTRK / RET / ALK fusions: targeted agents (larotrectinib, entrectinib, selpercatinib, etc.). Pembrolizumab added to dab/tram (off-label) is emerging.
• No targetable mutation: doxorubicin or paclitaxel ± RT (poor ORR, no OS improvement).

High-yield thyroid pearls

• PTC: psammoma bodies, Orphan Annie nuclei, BRAF V600E; best prognosis; recurs locally.
• FTC: capsular/vascular invasion defines malignancy (FNA cannot diagnose); recurs distantly. Hurthle cell (oncocytic) is often less RAI-avid; RAI-refractory disease is treated like RAI-refractory DTC.
• MTC: calcitonin/CEA markers; check germline RET and screen for pheochromocytoma BEFORE surgery; maintain euthyroid (no TSH suppression); selpercatinib preferred for RET-mutant disease (LIBRETTO-531).
• ATC: test BRAF urgently; dabrafenib + trametinib can be life-saving.
• RAI-refractory DTC: lenvatinib is the preferred 1L multikinase inhibitor only if no actionable fusion; for RET/NTRK/ALK fusion+ disease, the matched selective inhibitor is preferred 1L.
• TSH suppression is risk-adapted in DTC (<0.1 high-risk, 0.1 to 0.5 intermediate, low-normal in low-risk); balance against cardiac and bone toxicity.