Should I have half my thyroid removed or all of it?

This used to be an easy question with an unexamined answer: take the whole gland, because more surgery must be safer against cancer. That answer has not survived the evidence, and the decision is now genuinely open for most people who face it.

Survival is the part that is settled#

No randomised trial compares the two operations, and the reason is instructive: survival in this disease is so good that a trial would need enormous numbers and decades of follow-up. The best available substitute is propensity matching, which pairs patients who had each operation and were otherwise alike.

A single institution matched 918 pairs from a database of 6,259 patients, restricted to intrathyroidal, node-negative cancers under 4 cm.

None of the differences reached significance, and the direction of the small differences favours the smaller operation rather than the larger one. For this group of patients, taking the whole gland buys nothing measurable in survival.

The fear that drives people to the bigger operation#

The usual objection to starting small is that the pathology might come back worse than expected, the rest of the gland has to come out, and the patient ends up having two operations instead of one — with two sets of risks to the nerves and the parathyroid glands.

That has now been measured. A meta-analysis pooled twelve studies covering 43,362 patients, of whom 6,810 had a completion thyroidectomy after an initial lobectomy, and compared them against patients who had a total thyroidectomy from the outset.

The two-stage route was not worse on any of these, and on transient hypocalcaemia it was better. The likely reason is that a completion operation only disturbs one side, while a total thyroidectomy puts both sets of parathyroid glands at risk in a single sitting.

The reason to choose it that does not hold#

Most people who ask for the smaller operation are trying to avoid a daily tablet for the rest of their life. Half a thyroid is supposed to be enough to keep hormone levels normal. Often it is not.

That is close to a coin toss. It does not make the smaller operation pointless — the other half of that group did avoid medication, and a lobectomy patient who becomes hypothyroid usually needs a lower dose than someone with no thyroid at all. But it means the tablet cannot be treated as the thing lobectomy reliably prevents, and a decision made mainly on that basis is being made on a fifty per cent proposition.

If pregnancy is possible, this matters more#

The people having these operations in Korea are overwhelmingly women in their thirties and forties, so this is not a niche consideration. A matched cohort compared 463 pregnancies in women with a previous hemithyroidectomy against 2,293 matched pregnancies in women who had never had thyroid surgery.

• Prior hemithyroidectomy was associated with 4.2 times the odds of gestational hypothyroidism (95 percent confidence interval 2.9 to 6.3).
• The authors also report a modestly higher later-gestation hazard of preterm birth or pregnancy loss.
• Their conclusion is for closer thyroid monitoring during pregnancy rather than against the operation.

Half a thyroid can meet ordinary demand and fail to meet the higher demand of pregnancy. That is an argument for knowing this in advance and having thyroid function checked early, not an argument for removing the whole gland — which guarantees the medication rather than risking it.

What the larger operation costs#

A meta-analysis of 64 studies found the greatest risk of perioperative hypoparathyroidism falls on patients having total thyroidectomy with lateral neck dissection. Taking everything means both sides of the neck are operated on, which is where the parathyroid glands and both recurrent laryngeal nerves are.

Against that, total thyroidectomy makes follow-up simpler: thyroglobulin becomes an interpretable tumour marker and radioactive iodine remains available. For higher-risk disease those are real advantages. For an intrathyroidal cancer under 4 cm with no involved nodes, they are advantages against a problem that mostly does not arise.

What this does not say#

1. It applies to low- and intermediate-risk, node-negative, intrathyroidal cancer under 4 cm. Extrathyroidal extension, involved lymph nodes, aggressive histology or a larger tumour change the answer.
2. The survival comparison is propensity-matched observational data from one institution, not a randomised trial. Matching corrects for what was recorded, and surgeons choose lobectomy for patients who look favourable in ways a database may not capture.
3. The completion-thyroidectomy meta-analysis pools observational studies, and patients who go on to need completion are selected by their pathology rather than at random.
4. The 46.6 percent hypothyroidism figure comes from one tertiary centre of 507 patients. Reported rates after hemithyroidectomy vary widely with how hypothyroidism is defined and how long patients are followed.
5. The pregnancy cohort is Israeli. Nothing about it is specific to Korean patients beyond the fact that the population having these operations is similar.
6. This is a question about which operation, and it comes after the question of whether to operate at all — which for a small screen-detected papillary cancer is the one worth asking first.

The defensible summary is that for this group the whole gland buys no survival, that starting with half does not cost safety if the rest has to follow, and that the medication most people are trying to avoid arrives anyway about half the time. The smaller operation is the better default and it is not the escape it is usually sold as.