Place · Level 3
Thyroid Nodules · this is how the organ is built
甲状腺是几百万个小球堆出来的 · 大脑几十年一直催它长 · 长出来的包九成安静不动 · 神经和甲状旁腺就贴在它背面 · 关键是分层不是别查
Story path
- 1An organ built from tiny reservoirsAn organ built from tiny reservoirs
- 2The brain keeps telling it to growThe brain keeps telling it to grow
- 3Why most of them don't moveWhy most of them don't move
- 4Two delicate things on its backTwo delicate things on its back
- 5Stratify — not biopsy on sightStratify — not biopsy on sight
- 6Nodules & iodine · backwardsNodules & iodine · backwards
Chapter 1
An organ built from tiny reservoirs
An organ built from tiny reservoirs
An ultrasound found a thyroid nodule. The most likely explanation isn't that something bad grew in you — it's that the machine finally resolved what this organ has always looked like.
Cut a thyroid open and put it under a microscope and you don't see a solid piece of meat. You see millions of tiny spheres packed together. Each sphere is a follicle, a fraction of a millimeter across, with a structure as simple as a toy: a ring of epithelial cells holding hands to form a shell, with a pool of thick colloid stored in the middle.
That colloid is a warehouse of finished hormone. The thyroid is the only endocrine gland in the body that stores its finished product outside its cells — it builds a large molecule called thyroglobulin, hangs iodine on it, and seals it inside the sphere; when hormone is needed, it chews the molecule back and cuts thyroid hormone loose into the blood. That warehouse holds enough for two to three months (Zimmermann 2009).
So structurally, this organ is a heap of tiny reservoirs. Its building block isn't a fiber or a tube. It's a sphere. Everything downstream follows from that.
Now put that next to the probe. Follicles live at the fraction-of-a-millimeter scale; when a patch of follicles grows a little harder than its neighbors, the bulge it makes is one, two, a few millimeters across. And the high-frequency ultrasound probe in today's check-up rooms happens to resolve exactly that scale.
One German group drew the line bluntly: earlier work with a 7.5 MHz probe reported nodules in 33% of people; a later group scanned 635 check-up patients with a sharper 13 MHz probe and found nodules in 68% (Guth 2009). Those are two different groups in two different studies, not one group rescanned, so strictly it can't rule out the groups simply differing. But the direction is plain: once the ruler gets as fine as the organ's own building block, you start seeing the organ's grain.
That isn't more disease. That's resolution catching up with anatomy.
The same logic holds on the autopsy table: when pathologists section a thyroid finely enough, they find plenty of tiny foci in people never diagnosed in life — and that detection rate has not risen in sixty years (Harach 1985; Furuya-Kanamori 2016). The numbers are on the page below; the conclusion is one line — the finer you look, the more you find.
Hold onto it. The next five scenes are all its corollaries: finding it doesn't mean it was ever going to hurt you.
Cut a thyroid open and put it under a microscope and you don't see a solid piece of meat. You see millions of tiny spheres packed together. Each sphere is a follicle, a fraction of a millimeter across, with a structure as simple as a toy: a ring of epithelial cells holding hands to form a shell, with a pool of thick colloid stored in the middle.
That colloid is a warehouse of finished hormone. The thyroid is the only endocrine gland in the body that stores its finished product outside its cells — it builds a large molecule called thyroglobulin, hangs iodine on it, and seals it inside the sphere; when hormone is needed, it chews the molecule back and cuts thyroid hormone loose into the blood. That warehouse holds enough for two to three months (Zimmermann 2009).
So structurally, this organ is a heap of tiny reservoirs. Its building block isn't a fiber or a tube. It's a sphere. Everything downstream follows from that.
Now put that next to the probe. Follicles live at the fraction-of-a-millimeter scale; when a patch of follicles grows a little harder than its neighbors, the bulge it makes is one, two, a few millimeters across. And the high-frequency ultrasound probe in today's check-up rooms happens to resolve exactly that scale.
One German group drew the line bluntly: earlier work with a 7.5 MHz probe reported nodules in 33% of people; a later group scanned 635 check-up patients with a sharper 13 MHz probe and found nodules in 68% (Guth 2009). Those are two different groups in two different studies, not one group rescanned, so strictly it can't rule out the groups simply differing. But the direction is plain: once the ruler gets as fine as the organ's own building block, you start seeing the organ's grain.
That isn't more disease. That's resolution catching up with anatomy.
The same logic holds on the autopsy table: when pathologists section a thyroid finely enough, they find plenty of tiny foci in people never diagnosed in life — and that detection rate has not risen in sixty years (Harach 1985; Furuya-Kanamori 2016). The numbers are on the page below; the conclusion is one line — the finer you look, the more you find.
Hold onto it. The next five scenes are all its corollaries: finding it doesn't mean it was ever going to hurt you.
Background · what autopsies find
This page is background, not this story's argument. The argument is in the main text: this organ is natively a heap of millimeter-scale spheres. Autopsy data only confirms the same thing from another direction.Finland (Harach 1985): 101 consecutive autopsies, thyroids subserially sectioned at 2-3 mm. Fifty-two occult papillary carcinoma foci were found across 36 glands — a 35.6% detection rate, the highest reported anywhere at the time. Tumor diameter ranged from 0.15 mm to 14 mm, with 67% under 1 mm. The authors titled the paper accordingly: in Finland, this is a normal finding.
Six-decade pooled analysis (Furuya-Kanamori 2016): 35 studies, 12,834 autopsies. Stratified by examination intensity:
Whole gland examined: 11.2% (95% CI 6.7-16.1%)Partial sampling only: 4.1% (95% CI 3.0-5.4%)
The same pooled analysis found that the autopsy detection rate has not risen over six decades.
Why is Finland 35.6% and the pooled figure only 11.2%? Because Finland's section interval (2-3 mm) was far finer than most studies. Notice the coincidence: the blade's interval is in millimeters, and the probe's resolution is in millimeters — both rulers are graduated at exactly the scale at which a patch of follicles bulges into a lump, so they are seeing the same thing.
In other words, a detection rate is a number about your method, not about the disease. And the reason a method can move it at all is the anatomical fact in the main text: this organ's grain sits at the millimeter scale to begin with.
Chapter 2
The brain keeps telling it to grow
The brain keeps telling it to grow
Your thyroid hasn't taken a day off since you were born — and something has been prodding it the whole time. After a few decades of prodding, an organ that grew unevenly is the most ordinary outcome there is.
The thyroid doesn't decide its own workload. The pituitary, at the base of your brain, does. The pituitary constantly tastes the thyroid hormone level in your blood, and when it runs low it releases thyroid-stimulating hormone (thyroid-stimulating hormone: A pituitary hormone that prods the thyroid to work — it rises when the thyroid is underactive.).
TSH drifts down the bloodstream to your neck and parks on the receiving ports on the surface of follicle cells. That docking does three things at once:
Tells the cell to grab iodine out of the bloodTells the cell to chew the colloid back out of the sphere and cut hormone loose into the bloodTells the cell to get taller and to multiply — that is, to proliferate
The first two are today's work. The third is the long-run account: TSH isn't only a throttle, it is also a growth signal. When blood hormone is adequate, the pituitary lifts its foot; when it isn't, the foot stays down.
When iodine is short, that foot never comes up. Not enough raw material, not enough hormone; not enough hormone, so the pituitary keeps releasing TSH; TSH keeps coming, so follicle cells are driven to proliferate year after year, and the whole gland enlarges — that is goiter (Zimmermann 2009).
The key is the next step. The same dose of TSH lands on millions of follicles, and they do not respond alike. Follicles carry innate differences: some are sensitive to TSH and grow eagerly; others barely bother. Over one year the difference is invisible. Over ten years, thirty years, the leaders pull further and further ahead — from a speck only a microscope could see into a lump you can feel and scan.
That is where a nodule comes from. Nothing burrowed in from outside. It is your own follicles, grown uneven after decades of prodding.
So two facts that used to be puzzling need no statistics to explain:
Why are nodules so common? Because everyone's thyroid is being prodded, without a break, for life.Why do they multiply with age? Because the longer the prodding runs, the wider the gap opens.
A nodule you can feel or scan has, as its default explanation, this organ's normal way of working — not something in there killing you.
The thyroid doesn't decide its own workload. The pituitary, at the base of your brain, does. The pituitary constantly tastes the thyroid hormone level in your blood, and when it runs low it releases thyroid-stimulating hormone (thyroid-stimulating hormone: A pituitary hormone that prods the thyroid to work — it rises when the thyroid is underactive.).
TSH drifts down the bloodstream to your neck and parks on the receiving ports on the surface of follicle cells. That docking does three things at once:
Tells the cell to grab iodine out of the bloodTells the cell to chew the colloid back out of the sphere and cut hormone loose into the bloodTells the cell to get taller and to multiply — that is, to proliferate
The first two are today's work. The third is the long-run account: TSH isn't only a throttle, it is also a growth signal. When blood hormone is adequate, the pituitary lifts its foot; when it isn't, the foot stays down.
When iodine is short, that foot never comes up. Not enough raw material, not enough hormone; not enough hormone, so the pituitary keeps releasing TSH; TSH keeps coming, so follicle cells are driven to proliferate year after year, and the whole gland enlarges — that is goiter (Zimmermann 2009).
The key is the next step. The same dose of TSH lands on millions of follicles, and they do not respond alike. Follicles carry innate differences: some are sensitive to TSH and grow eagerly; others barely bother. Over one year the difference is invisible. Over ten years, thirty years, the leaders pull further and further ahead — from a speck only a microscope could see into a lump you can feel and scan.
That is where a nodule comes from. Nothing burrowed in from outside. It is your own follicles, grown uneven after decades of prodding.
So two facts that used to be puzzling need no statistics to explain:
Why are nodules so common? Because everyone's thyroid is being prodded, without a break, for life.Why do they multiply with age? Because the longer the prodding runs, the wider the gap opens.
A nodule you can feel or scan has, as its default explanation, this organ's normal way of working — not something in there killing you.
Mechanism · why many, not one
The word *multi* in the diagnosis multinodular goiter is a direct consequence of the heterogeneity in the main text. If every follicle answers thyroid-stimulating hormone: A pituitary hormone that prods the thyroid to work — it rises when the thyroid is underactive. differently, then after decades of prodding what surfaces is never one lump but a batch of them, in assorted sizes.Autonomous nodules, also called hot nodules: as they grow, some nodules simply stop taking orders from TSH and make hormone at their own pace. This is especially common in regions with long-standing iodine deficiency — decades of prodding accumulate a set of unsupervised workshops.
Those workshops produce a counterintuitive consequence: when iodine is restored to a long-deficient population, a transient wave of hyperthyroidism appears — iodine-induced hyperthyroidism. The unsupervised nodules suddenly receive ample raw material, run flat out, and overproduce hormone (Zimmermann 2009). This is part of the right-hand half of scene 6's U-shaped curve; remember it, you'll need it there.
So could you run it backwards — suppress TSH and shrink the nodule? It's a natural thought: if TSH is the growth signal, cut the signal. That road was tried and rejected. Guidelines explicitly advise against levothyroxine TSH suppression for benign nodules in iodine-sufficient populations, because the benefit doesn't cover the cost — to hold down a lump that most likely won't harm you, you soak your whole body in mild thyrotoxicosis for years, buying arrhythmia and bone-loss risk (ATA 2015 / Haugen 2016).
Mechanistically coherent is not the same as clinically worth it. This site will keep coming back to that.
Chapter 3
Why most of them don't move
Why most of them don't move
The small things inside the thyroid mostly sit still for a lifetime. That isn't reassurance — people actually watched, and measured, year after year.
Start with why they can be so quiet. The last scene said a nodule is just follicles that grew eagerly; and the cancer that arises from follicular epithelium is classified, in pathology, as differentiated thyroid cancer. Differentiated means it still looks a lot like the follicle cell it came from and still does the old job — still takes up iodine, still makes thyroglobulin, still listens to thyroid-stimulating hormone: A pituitary hormone that prods the thyroid to work — it rises when the thyroid is underactive.. A cell that keeps showing up for its day job divides slowly and isn't much good at escaping. It is a different species from the cancer that has entirely forgotten who it is (that one is in scene 6, and it is savage).
This isn't armchair reasoning; someone spent decades testing it. From 1993, Kuma Hospital in Japan did something brave: it offered patients with low-risk papillary microcarcinoma an extra option — don't cut, just watch with periodic ultrasound. This is active surveillance. The low-risk bar is explicit: no more than 1 cm across, not growing outside the capsule, no lymph node metastasis, and not sitting against the trachea or the recurrent laryngeal nerve.
1,235 people were watched this way for years, and the results were (Ito 2014):
Tumors that grew by 3 mm or more: 4.9% at 5 years; 8.0% at 10 yearsNew lymph node metastases at 5 / 10 years: 1.7% / 3.8%Across the whole program, nobody developed distant metastasis and nobody died of thyroid carcinoma
Over ten years, more than nine in ten didn't budge. And for the under-one-in-ten that did grow, operating once it grew was still in time — the group converted to delayed surgery had no life-threatening recurrences.
Read this part slowly: active surveillance isn't a gamble and it isn't giving up on treatment. It is saving the operation for when it's needed.
And the reason it's possible is, again, anatomy. The thyroid sits right under the skin at the front of your neck, where an ultrasound sees it clearly through one layer of tissue, remeasurable whenever you like. It's one of the few organs in the body that can be watched this closely — the pancreas and the ovaries, buried deep in the abdomen, get no such courtesy.
Start with why they can be so quiet. The last scene said a nodule is just follicles that grew eagerly; and the cancer that arises from follicular epithelium is classified, in pathology, as differentiated thyroid cancer. Differentiated means it still looks a lot like the follicle cell it came from and still does the old job — still takes up iodine, still makes thyroglobulin, still listens to thyroid-stimulating hormone: A pituitary hormone that prods the thyroid to work — it rises when the thyroid is underactive.. A cell that keeps showing up for its day job divides slowly and isn't much good at escaping. It is a different species from the cancer that has entirely forgotten who it is (that one is in scene 6, and it is savage).
This isn't armchair reasoning; someone spent decades testing it. From 1993, Kuma Hospital in Japan did something brave: it offered patients with low-risk papillary microcarcinoma an extra option — don't cut, just watch with periodic ultrasound. This is active surveillance. The low-risk bar is explicit: no more than 1 cm across, not growing outside the capsule, no lymph node metastasis, and not sitting against the trachea or the recurrent laryngeal nerve.
1,235 people were watched this way for years, and the results were (Ito 2014):
Tumors that grew by 3 mm or more: 4.9% at 5 years; 8.0% at 10 yearsNew lymph node metastases at 5 / 10 years: 1.7% / 3.8%Across the whole program, nobody developed distant metastasis and nobody died of thyroid carcinoma
Over ten years, more than nine in ten didn't budge. And for the under-one-in-ten that did grow, operating once it grew was still in time — the group converted to delayed surgery had no life-threatening recurrences.
Read this part slowly: active surveillance isn't a gamble and it isn't giving up on treatment. It is saving the operation for when it's needed.
And the reason it's possible is, again, anatomy. The thyroid sits right under the skin at the front of your neck, where an ultrasound sees it clearly through one layer of tissue, remeasurable whenever you like. It's one of the few organs in the body that can be watched this closely — the pancreas and the ovaries, buried deep in the abdomen, get no such courtesy.
Chapter 4
Two delicate things on its back
Two delicate things on its back
Just cut it out for peace of mind. It sounds like a bargain, because it quietly prices the cost at zero. The cost is not zero, and the reason is written in the anatomy: pressed against the back of the thyroid are two structures that matter enormously and are very hard to avoid.
First, to be clear: this scene blames no one. Many people have already had the surgery, and under the consensus of the time their doctors' advice was entirely reasonable. The point isn't that you shouldn't have had it — it's why this bill can never be zero, so that people who haven't reached this fork can weigh both sides on the same scale.
The first is the recurrent laryngeal nerve. It works the opening and closing of your two vocal cords. Its route is absurd: it runs down through the neck, loops under the great vessels at the top of the chest, then turns back upward (recurrent means exactly that — it comes back). And the stretch where it climbs back up runs in the groove between the trachea and the esophagus, with the back of the thyroid pressed against the outside of that groove. Which means: to lift the thyroid out of that position, you must cut right beside this nerve. Bruise it and that vocal cord stops moving properly and the voice goes hoarse; injure both sides and breathing is compromised.
The second is the parathyroid glands. Four of them, each the size of a grain of rice, stuck to the back of the thyroid. Their job is not small: the calcium level in your blood is theirs to guard. The trouble is that they are tiny, they look much like the surrounding fat, and their blood comes through fine vessels branching off the thyroid's own supply. So the surgical risk isn't only cutting one out — more often it's severing those fine vessels while dissecting the thyroid away. The gland is still there; it's just starving. When it stops working, blood calcium falls and your hands and feet start to go numb and cramp.
Those two facts are the anatomical explanation of the bill. Why are the signature complications of thyroid surgery a hoarse voice and numb hands? Not because surgeons are careless — because the back of this organ is built that way. Every dissection happens right next to those two things.
One more unavoidable line item: after part or all of the thyroid is removed, the remaining tissue may not supply the hormone you need, so a daily tablet makes up the difference. That isn't a complication, it's arithmetic — you removed part of scene 1's warehouse yourself.
The numbers are on the page below. One note first: those complication rates come from Kuma Hospital in Japan, one of the highest-volume, best centers in the world for this operation. An ordinary hospital only looks worse (Adam 2017).
First, to be clear: this scene blames no one. Many people have already had the surgery, and under the consensus of the time their doctors' advice was entirely reasonable. The point isn't that you shouldn't have had it — it's why this bill can never be zero, so that people who haven't reached this fork can weigh both sides on the same scale.
The first is the recurrent laryngeal nerve. It works the opening and closing of your two vocal cords. Its route is absurd: it runs down through the neck, loops under the great vessels at the top of the chest, then turns back upward (recurrent means exactly that — it comes back). And the stretch where it climbs back up runs in the groove between the trachea and the esophagus, with the back of the thyroid pressed against the outside of that groove. Which means: to lift the thyroid out of that position, you must cut right beside this nerve. Bruise it and that vocal cord stops moving properly and the voice goes hoarse; injure both sides and breathing is compromised.
The second is the parathyroid glands. Four of them, each the size of a grain of rice, stuck to the back of the thyroid. Their job is not small: the calcium level in your blood is theirs to guard. The trouble is that they are tiny, they look much like the surrounding fat, and their blood comes through fine vessels branching off the thyroid's own supply. So the surgical risk isn't only cutting one out — more often it's severing those fine vessels while dissecting the thyroid away. The gland is still there; it's just starving. When it stops working, blood calcium falls and your hands and feet start to go numb and cramp.
Those two facts are the anatomical explanation of the bill. Why are the signature complications of thyroid surgery a hoarse voice and numb hands? Not because surgeons are careless — because the back of this organ is built that way. Every dissection happens right next to those two things.
One more unavoidable line item: after part or all of the thyroid is removed, the remaining tissue may not supply the hormone you need, so a daily tablet makes up the difference. That isn't a complication, it's arithmetic — you removed part of scene 1's warehouse yourself.
The numbers are on the page below. One note first: those complication rates come from Kuma Hospital in Japan, one of the highest-volume, best centers in the world for this operation. An ordinary hospital only looks worse (Adam 2017).
Clinical · the bill, surgery vs surveillance
Same hospital, same low-risk microcarcinomas, immediate surgery next to active surveillance (Oda 2016):Needing lifelong levothyroxine: 66.1% vs 20.7%Transient hypoparathyroidism (numb hands and feet, cramping): 16.7% vs 2.8%Permanent hypoparathyroidism: 1.6% vs 0.08%Transient vocal cord paralysis (voice goes hoarse): 4.1% vs 0.6%Permanent vocal cord paralysis: 0.2% in the surgery group, 0 in surveillance
Map these rows onto the anatomy in the main text: the two parathyroid rows are the four rice grains on the back being injured or starved during dissection; the two vocal-cord rows are the nerve climbing back up the tracheoesophageal groove. The menu of complications is set by anatomy — it isn't random bad luck.
Removing only half isn't free either: after hemithyroidectomy, 26.4% of 535 patients ultimately still needed levothyroxine (Ahn 2019). The remaining lobe doesn't always cover it.
Whether the operator does this often matters, concretely: US data show that the lower a surgeon's annual volume, the higher the patient's complication rate — a relationship that continues up to roughly 26 cases per year (Adam 2017). So the numbers above, from an ultra-high-volume center like Kuma, represent this operation's ceiling performance, not its average.
And one line item that never appears on a consent form. Once the word cancer is attached to you it doesn't peel off: in US data, thyroid cancer survivors report psychological financial hardship at nearly twice the rate of other cancer survivors (46.1% vs 24.0%, Barrows 2020). Insurance, mortgages, work, and the few days before every follow-up scan are all in that account.
ahn-2019-levothyroxine-hemithyroidectomybarrows-2020-thyroid-cancer-financial-burden
Chapter 5
Stratify — not biopsy on sight
Stratify — not biopsy on sight
Those words on your ultrasound report — hypoechoic, microcalcification, taller-than-wide — aren't mystical scoring. Behind each of them is a specific tissue appearance under the microscope. What the radiologist is doing is reading histology through your skin.
The first four scenes told you: this organ is natively a heap of spheres (scene 1), prodded into growing unevenly (scene 2), and what grows mostly keeps doing its day job (scene 3), while acting on it carries an anatomical cost (scene 4). Add those up and you don't get *don't look*. You get: we need a way to guess what the cells inside this lump look like, without cutting it open.
That way is ultrasound. It reads echo — sound goes in, and different things bounce back with different strength. Different tissue appearance, different reflection. So every sonographic feature is an indirect answer to a histological question:
Cystic or spongiform points benign. A lump that looks like a cluster of little bubbles looks that way because that's what it is: swollen follicles, still holding colloid in their centers. Fluid barely blocks sound, so it reads bright. It's scene 1's normal structure, enlarged but not distorted.Solid and hypoechoic (dark on the image) points suspicious. To reflect sound you need interfaces. A dark lump means the cells inside are packed shoulder to shoulder with no colloid and no fluid — the warehouse in the center is gone, and the cells only grow, no longer doing the job of storing hormone. That is what *stopped showing up for work* looks like on ultrasound.Microcalcifications (specks of white) are the most telling one. These pinpoint bright foci often correspond to psammoma bodies under the microscope: concentric little spheres laid down layer by layer as cells that died inside a papillary carcinoma calcify, only tens of micrometers across (Tessler 2017). Benign colloid nodules don't make them. So a scatter of white specks doesn't mean the lump is hard — it means a batch of those cells died in there.Taller-than-wide points suspicious. A benign nodule grows passively: it pushes along the grain of the surrounding tissue, toward least resistance, and gets squashed into a flat, lying-down shape. A malignant one ignores the grain and grows across tissue planes, so on the image it stands up. Its shape isn't merely odd — it isn't obeying the constraints of the tissue around it.Ill-defined or lobulated margins, or growth beyond the gland, points suspicious. A benign nodule usually shuts itself inside a capsule, with a smooth border. A ragged, lobulated margin means it is infiltrating outward with no wall around it.
Once you see this layer, the TI-RADS scoring table (on the page below) stops being a pile of points: it is those five histological questions, asked one at a time, with the answers added up (Tessler 2017).
Only after grading does fine-needle aspiration (FNA) come up. Here's the key: the biopsy threshold depends on grade and size together — the lower the grade, the bigger it has to be. An 8 mm nodule that looks entirely well-behaved does not get biopsied under the rules.
That isn't laziness and it isn't cost-cutting. Back to scenes 1 and 2: biopsying it will most likely find something this organ has always had, and then put you on scene 4's operating table.
So *any nodule means biopsy* is wrong. But so is the reverse: *I have a nodule, so I'll ignore it*. The rules are the rules: follow up what needs following, biopsy what needs biopsy, remove what needs removing. Stratification means sending each nodule down the path it belongs on, not sending all of them down the same one.
The first four scenes told you: this organ is natively a heap of spheres (scene 1), prodded into growing unevenly (scene 2), and what grows mostly keeps doing its day job (scene 3), while acting on it carries an anatomical cost (scene 4). Add those up and you don't get *don't look*. You get: we need a way to guess what the cells inside this lump look like, without cutting it open.
That way is ultrasound. It reads echo — sound goes in, and different things bounce back with different strength. Different tissue appearance, different reflection. So every sonographic feature is an indirect answer to a histological question:
Cystic or spongiform points benign. A lump that looks like a cluster of little bubbles looks that way because that's what it is: swollen follicles, still holding colloid in their centers. Fluid barely blocks sound, so it reads bright. It's scene 1's normal structure, enlarged but not distorted.Solid and hypoechoic (dark on the image) points suspicious. To reflect sound you need interfaces. A dark lump means the cells inside are packed shoulder to shoulder with no colloid and no fluid — the warehouse in the center is gone, and the cells only grow, no longer doing the job of storing hormone. That is what *stopped showing up for work* looks like on ultrasound.Microcalcifications (specks of white) are the most telling one. These pinpoint bright foci often correspond to psammoma bodies under the microscope: concentric little spheres laid down layer by layer as cells that died inside a papillary carcinoma calcify, only tens of micrometers across (Tessler 2017). Benign colloid nodules don't make them. So a scatter of white specks doesn't mean the lump is hard — it means a batch of those cells died in there.Taller-than-wide points suspicious. A benign nodule grows passively: it pushes along the grain of the surrounding tissue, toward least resistance, and gets squashed into a flat, lying-down shape. A malignant one ignores the grain and grows across tissue planes, so on the image it stands up. Its shape isn't merely odd — it isn't obeying the constraints of the tissue around it.Ill-defined or lobulated margins, or growth beyond the gland, points suspicious. A benign nodule usually shuts itself inside a capsule, with a smooth border. A ragged, lobulated margin means it is infiltrating outward with no wall around it.
Once you see this layer, the TI-RADS scoring table (on the page below) stops being a pile of points: it is those five histological questions, asked one at a time, with the answers added up (Tessler 2017).
Only after grading does fine-needle aspiration (FNA) come up. Here's the key: the biopsy threshold depends on grade and size together — the lower the grade, the bigger it has to be. An 8 mm nodule that looks entirely well-behaved does not get biopsied under the rules.
That isn't laziness and it isn't cost-cutting. Back to scenes 1 and 2: biopsying it will most likely find something this organ has always had, and then put you on scene 4's operating table.
So *any nodule means biopsy* is wrong. But so is the reverse: *I have a nodule, so I'll ignore it*. The rules are the rules: follow up what needs following, biopsy what needs biopsy, remove what needs removing. Stratification means sending each nodule down the path it belongs on, not sending all of them down the same one.
Clinical · TI-RADS and FNA thresholds
TI-RADS stands for Thyroid Imaging Reporting and Data System. Below is the American College of Radiology (ACR) 2017 version (Tessler 2017), placed here so you can read it against your own report. The main text already covered the histology behind each feature; this page only turns them into points.Five scored feature categories:
1. Composition: cystic / spongiform / mixed / solid (solid scores highest)
2. Echogenicity: anechoic / hyper- or isoechoic / hypoechoic / very hypoechoic (darker scores higher)
3. Shape: wider-than-tall / taller-than-wide (standing up scores higher)
4. Margin: smooth / ill-defined / lobulated or irregular / extra-thyroidal extension
5. Echogenic foci: none / large comet-tail / macrocalcification / peripheral rim / punctate echogenic foci (microcalcification)
Sum the five scores to get a level, then combine with maximum diameter to decide the action:
| Level | Points | Malignancy risk | FNA threshold | Follow-up threshold |
|---|---|---|---|---|
| TR1 benign | 0 | ~0.3% | none | none |
| TR2 not suspicious | 1-2 | ~1.5% | none | none |
| TR3 mildly suspicious | 3 | ~4.8% | ≥ 2.5 cm | ≥ 1.5 cm |
| TR4 moderately suspicious | 4-6 | ~9.1% | ≥ 1.5 cm | ≥ 1.0 cm |
| TR5 highly suspicious | ≥ 7 | ~35% | ≥ 1.0 cm | ≥ 0.5 cm |
The thing to catch in this table is the diagonal: at the same size, the more suspicious it looks, the lower the threshold; at the same appearance, the smaller it is, the more you leave it alone. A 1.2 cm TR3 doesn't get biopsied; a 1.2 cm TR5 does. Size alone is never the reason — size plus appearance is.
Two more things worth knowing:
The American Thyroid Association (ATA) 2015 guidelines are a parallel stratification scheme (Haugen 2016) with the same logic and slightly different thresholds. Which one your report uses depends on the hospital.TI-RADS is a communication and triage tool, not a diagnosis. It gives a probability and a next action, not a conclusion. The real conclusion comes from biopsy cytology, plus you and your doctor weighing your age, family history, symptoms, and how much the uncertainty costs you.
This page is here to help you read your report, not to grade yourself. Executing the stratification is the clinician's job.
Chapter 6
Nodules & iodine · backwards
Nodules & iodine · backwards
If you have a nodule you must avoid iodine, switch to non-iodized salt, and never eat kelp or nori — this is the most widely repeated thyroid claim on the Chinese internet, and it has the direction backwards.
You already know where it goes wrong, from scene 2: iodine is the raw material, and the direct consequence of short raw material is a pituitary with its foot on the throttle. Not enough iodine, not enough hormone; not enough hormone, thyroid-stimulating hormone: A pituitary hormone that prods the thyroid to work — it rises when the thyroid is underactive. keeps coming; TSH keeps coming, follicles are driven to proliferate, and over years they grow into nodules (Zimmermann 2009). Iodine deficiency is itself a classic cause of nodular goiter — that chain is textbook causation, not correlation.
So a person who already has a nodule and deliberately avoids iodine is doing something deeply ironic: in order to shrink the nodule, they press the throttle down harder with their own hand.
The data agree. Pool 25 studies and over 50,000 people: those with urinary iodine below 100 µg/L have roughly 30% higher nodule risk than iodine-replete people. But look at the other end of the curve too: run the continuous dose-response and the whole relationship is a statistically significant U-shape (P for nonlinearity < 0.001) — risk rises at both the deficient and the excessive end, with the low point around 221 µg/L (Lu 2026, pooled cross-sectional, association not causation).
The right-hand half of the U has a mechanism too, and scene 2's page already planted it: when iodine is abundant, the autonomous nodules that no longer take orders from TSH suddenly receive ample raw material and run flat out (Zimmermann 2009). So neither end is good — it is not *less is safer*, and it is not *more is better*. The full iodine-thyroid U-curve is taught in detail on the iodine island, and the kelp story has a dedicated animation for it; go there rather than have it repeated here.
This story adds only the conclusion: your nodule almost certainly wasn't caused by iodine, and won't be dissolved by avoiding it. It is most likely what scenes 1 and 2 described — your own follicles, prodded for decades, grown uneven, and then lit up by an ultrasound with enough resolution to see them.
There are two genuine exceptions who should discuss iodine with a doctor individually: people who already test positive for Hashimoto's antibodies (details on the Hashimoto island), and people facing iodine contrast imaging or with existing hyperthyroidism.
Finally, be clear where this story lands. It does not tell you don't get checked. It tells you stratify. Anyone with a nodule should follow their own doctor and guideline-based management. This page is education about why the rules look the way they do — it is not medical advice.
See a doctor promptly, don't wait, for any of these: a persistently hoarse voice; new difficulty swallowing or breathing; a nodule that visibly enlarges over a short period; something that feels hard and fixed in place; enlarged lymph nodes in the neck; a childhood history of head-and-neck radiation; or a first-degree relative with thyroid cancer. These are signals that genuinely need pursuing — an entirely different thing from a small nodule an ultrasound stumbled onto.
You already know where it goes wrong, from scene 2: iodine is the raw material, and the direct consequence of short raw material is a pituitary with its foot on the throttle. Not enough iodine, not enough hormone; not enough hormone, thyroid-stimulating hormone: A pituitary hormone that prods the thyroid to work — it rises when the thyroid is underactive. keeps coming; TSH keeps coming, follicles are driven to proliferate, and over years they grow into nodules (Zimmermann 2009). Iodine deficiency is itself a classic cause of nodular goiter — that chain is textbook causation, not correlation.
So a person who already has a nodule and deliberately avoids iodine is doing something deeply ironic: in order to shrink the nodule, they press the throttle down harder with their own hand.
The data agree. Pool 25 studies and over 50,000 people: those with urinary iodine below 100 µg/L have roughly 30% higher nodule risk than iodine-replete people. But look at the other end of the curve too: run the continuous dose-response and the whole relationship is a statistically significant U-shape (P for nonlinearity < 0.001) — risk rises at both the deficient and the excessive end, with the low point around 221 µg/L (Lu 2026, pooled cross-sectional, association not causation).
The right-hand half of the U has a mechanism too, and scene 2's page already planted it: when iodine is abundant, the autonomous nodules that no longer take orders from TSH suddenly receive ample raw material and run flat out (Zimmermann 2009). So neither end is good — it is not *less is safer*, and it is not *more is better*. The full iodine-thyroid U-curve is taught in detail on the iodine island, and the kelp story has a dedicated animation for it; go there rather than have it repeated here.
This story adds only the conclusion: your nodule almost certainly wasn't caused by iodine, and won't be dissolved by avoiding it. It is most likely what scenes 1 and 2 described — your own follicles, prodded for decades, grown uneven, and then lit up by an ultrasound with enough resolution to see them.
There are two genuine exceptions who should discuss iodine with a doctor individually: people who already test positive for Hashimoto's antibodies (details on the Hashimoto island), and people facing iodine contrast imaging or with existing hyperthyroidism.
Finally, be clear where this story lands. It does not tell you don't get checked. It tells you stratify. Anyone with a nodule should follow their own doctor and guideline-based management. This page is education about why the rules look the way they do — it is not medical advice.
See a doctor promptly, don't wait, for any of these: a persistently hoarse voice; new difficulty swallowing or breathing; a nodule that visibly enlarges over a short period; something that feels hard and fixed in place; enlarged lymph nodes in the neck; a childhood history of head-and-neck radiation; or a first-degree relative with thyroid cancer. These are signals that genuinely need pursuing — an entirely different thing from a small nodule an ultrasound stumbled onto.
Debunked · six things people say
Nodules grow into cancer, so cut it out while it's smallThe vast majority of nodules are benign, and benign nodules don't turn into cancer — histologically they are two different roads. A benign nodule is scene 1's follicular structure, enlarged, still holding colloid in its centers; a cancer is a follicle cell grown distorted, no longer doing the storage job. They are not two stretches of the same road. As for the ones that really are papillary microcarcinoma, more than nine in ten don't budge in ten years (Ito 2014). The benefit of cutting early is imagined as large and the cost as zero; both are wrong.
Thyroid cancer is the lucky cancer — just take it out and you're fine
This makes two errors at once. It prices the surgical bill at zero (go back to scene 4: the recurrent laryngeal nerve and the four rice grains are pressed against the back of the gland), and it hides the genuinely dangerous minority. Scene 3 said differentiated cancer is mild because it still remembers being a follicle cell; run that backwards and the one that has entirely forgotten is savage — anaplastic carcinoma is only about 1.7% of thyroid cancers, but historical median survival is about 5 months, with 1-year survival of 20% (Bible 2021). Thyroid cancer isn't one disease; it's a group spanning a huge range of differentiation. Calling it lucky is deeply unfair to that minority.
Adding a thyroid ultrasound to the check-up package must be a good thing
Reason from mechanism first: an organ natively grained at the millimeter scale (scene 1), measured with a millimeter-scale ruler in people without symptoms — you will inevitably find a great deal, and many of those people will be pushed toward scene 4's operating table. Which is why in 2017 the US Preventive Services Task Force gave thyroid cancer screening in asymptomatic people a grade D: not recommended, because the harms outweigh the benefits (USPSTF 2017). Read the scope carefully: grade D applies to screening people without symptoms. Having symptoms, having a family history, or being told to get checked by your doctor is an entirely different situation — get checked.
Prunella or seaweed goiter pills dissolve nodules
No reliable evidence supports any herb shrinking nodules. There's a historical misunderstanding worth clearing up: classical formulas treated goiter with seaweed and kombu because ancient goiter was mostly iodine-deficiency goiter — and back then, supplying iodine meant releasing scene 2's throttle, which genuinely worked. But the nodule a check-up finds today is not the same illness at all; the old remedy is answering a different question.
Worth noting: even the pharmaceutical route isn't recommended. Guidelines explicitly advise against levothyroxine thyroid-stimulating hormone: A pituitary hormone that prods the thyroid to work — it rises when the thyroid is underactive. suppression for benign nodules, because the benefit doesn't cover the harm (ATA 2015). If cutting the growth signal doesn't clear the bar even as a drug, an herb certainly doesn't.
Nodules come from anger and bottled-up stress
No evidence — and mechanistically there's no way in. The growth signal for follicle cells is TSH, and TSH is set by the thyroid hormone level in your blood, not by your mood. The real function of this claim isn't to explain a cause; it's to make people blame themselves, converting an anatomical finding that has nothing to do with your personality into a personal failing.
This logic only applies to the thyroid
No. But what you should take away isn't a statistical rule — it's a set of structural conditions. Any organ that meets three criteria at once will see *found* come apart from *going to harm you*:
1. It natively carries a large stock of quiet small lesions (thyroid: follicles prodded by TSH for decades)
2. Most of those lesions still do their day job and grow slowly (differentiated)
3. There is a cheap, non-invasive test whose resolution is finer than the lesions' scale (a 13 MHz probe against millimeter lumps)
With all three in place, you will find a great deal that was never going to surface. That isn't a statistical curiosity; it's the necessary consequence of those three structural conditions.
Prostate is the closest sibling: the same huge reservoir of indolent lesions, the same cheap test that pulls them up, and the same conditional, shared-decision-making recommendation from the US Preventive Services Task Force (USPSTF 2018). See the prostate island for detail.
The pattern has been quantified for thyroid cancer in women in high-income countries: an estimated 50-90% of diagnoses are overdiagnosis (Vaccarella 2016). But that number is only the scoreboard for those three structural conditions — it is not the starting point of the reasoning. The starting point is scene 1: this is simply how the organ is built.
bible-2021-ata-anaplasticuspstf-2017-thyroid-cancer-screeninguspstf-2018-psa-screeningvaccarella-2016-worldwide-overdiagnosis