Thyroid cancer — treatment notes

26Thyroid cancer

Snapshot. The commonest endocrine cancer — young women, usually a nodule — and three different diseases share the organ: DIFFERENTIATED (papillary ~85% + follicular) — follicular-cell derived, takes up iodine, makes thyroglobulin, excellent prognosis · MEDULLARY (MTC) — parafollicular C-cell tumour: makes CALCITONIN not thyroglobulin, takes up NO iodine, RET-driven (~25% germline — MEN2) · ANAPLASTIC (ATC) — rare, elderly, one of the most lethal human cancers, stage IV by definition, often airway-first. The unique lever in differentiated disease: TSH is a GROWTH FACTOR for the tumour → levothyroxine is dosed to SUPPRESS TSH, not just replace (target set by recurrence risk).

Workup

Nodule → TSH + neck ultrasound → FNA by sonographic risk (Bethesda cytology) · a hot (hyperfunctioning) nodule on uptake scan is almost never cancer — do not FNA it · MEDULLARY suspected → calcitonin + CEA, RET germline testing, and EXCLUDE PHAEOCHROMOCYTOMA (plasma metanephrines) BEFORE any surgery (MEN2 — operating on an unrecognised phaeo is the classic disaster) · advanced/refractory → molecular panel: BRAF V600E · RET fusions/mutations · NTRK fusions (each has a drug) · surveillance in differentiated disease = thyroglobulin + anti-Tg antibodies + neck US.

Treatment by setting

Setting	Treatment
DIFFERENTIATED — localized papillary / follicular	Surgery: lobectomy (small, low-risk, confined) OR total thyroidectomy (larger, nodes, high-risk) 10-yr survival >90–95% — most patients die WITH this cancer, not of it Radioactive iodine (RAI, I-131) ablation for intermediate/high risk after TOTAL thyroidectomy the tumour drinks the iodine — a therapy only differentiated disease can receive; low-risk usually skips it Levothyroxine at a TSH-SUPPRESSION target set by risk — relaxed as the years stay clean a suppressed TSH costs AF + bone density long-term — the target is a standing decision, not set-and-forget Surveillance: thyroglobulin + anti-Tg antibodies + neck ultrasound Tg is only interpretable after total thyroidectomy ± RAI; the antibodies invalidate the assay — always reported together
DIFFERENTIATED — recurrent / metastatic	RAI-avid → repeat RAI · slow asymptomatic disease can simply be WATCHED on suppression RAI-REFRACTORY + progressing → a multikinase VEGFR inhibitor: LENVATINIB (SELECT) or sorafenib (DECISION) start for progression/symptoms, not reflexively · hypertension + proteinuria are the toll Driver-matched: RET fusion → selpercatinib (a selective RET inhibitor) (LIBRETTO-001) · NTRK fusion → larotrectinib / entrectinib
MEDULLARY (MTC)	Total thyroidectomy + central node dissection — surgery is the only cure NO RAI (C cells take no iodine) · NO TSH-suppression benefit — replacement dosing only · phaeochromocytoma excluded FIRST Markers: calcitonin + CEA — their DOUBLING TIMES carry the prognosis Advanced → RET-mutant: selpercatinib (LIBRETTO-001) · alternatives vandetanib / cabozantinib (multikinase inhibitors) vandetanib → QT prolongation · germline RET-positive → family cascade testing (MEN2: phaeo + hyperparathyroidism)
ANAPLASTIC (ATC)	AIRWAY FIRST — then molecular testing the same day: BRAF V600E (~40%) → dabrafenib + trametinib the one lever that has changed this disease — days matter Otherwise: surgery only if truly resectable + chemoradiation (taxane-based) in the fit · early palliative-care involvement for most median survival historically ~6 months — goals-of-care honesty from the first visit

Watch

Levothyroxine over-suppression → palpitations, AF, bone loss — reassess the TSH target yearly. RAI → sialadenitis, dry mouth, a small long-term second-malignancy signal; exclude pregnancy before dosing. Lenvatinib / sorafenib → HYPERTENSION, proteinuria, hand-foot, fatigue — BP diary + urine dip each visit. Vandetanib → QT (ECG + K/Mg). Post-thyroidectomy hypocalcaemia (parathyroids) — perioral tingling → check calcium. A rising thyroglobulin with negative imaging = recurrence hunting, not reassurance.