Endocrinology

The pathologist slides a section of adrenal cortex under the scope, dotted with small yellow-brown nodules; the resident blurts out that it's metastatic melanoma, but the attending shakes her head and says it's lipofuscin. The color looks the same, but the story is entirely different. Endocrine pathology looks at first like memorizing a clue to match a disease, but what you actually need to do is understand why each clue looks the way it does — once you understand it, the pairing forms in your mind on its own, with no rote memorization required.

The entire set of changes in the adrenal cortex resolves once you understand one thing: cortisol is the boss, ACTH is the sidekick — when cortisol runs high, the brain shuts ACTH off, and the normal cortex depends on ACTH to be fed. So when a tumor secretes large amounts of cortisol on its own authority, the brain senses the excess in the blood and suppresses ACTH; the normal cortex nearby, which lives off ACTH, receives no nourishment and starves into atrophy, leaving only the tumor end to keep growing. The same logic applies when a patient takes exogenous steroids long-term: the body mistakenly believes cortisol is already sufficient and likewise withholds ACTH, so both sides of the cortex starve into atrophy together. That is why iatrogenic Cushing syndrome means atrophy, not hyperplasia — this is not a pairing to memorize, but the inevitable result of starving with nobody left to feed it.

The exam loves asking what the adrenal glands look like in Cushing syndrome from long-term steroid use — follow this causal chain and the answer is always atrophy. The yellow-brown pigment in micronodular hyperplasia is lipofuscin, not melanin, and an adenoma secretes autonomously and suppresses the contralateral side into atrophy — both follow the same logic. That is also why, when sorting the causes of Cushing syndrome, you must first check whether ACTH is high or suppressed: if suppressed, the problem lies in the adrenal gland's own autonomous secretion, most commonly an adenoma; if ACTH is still elevated, the source lies in pituitary Cushing disease or ectopic secretion such as small cell lung cancer, and ectopic sources often come with markedly elevated ACTH, hypokalemia, and hyperpigmentation. Sort out first whether ACTH is high or suppressed, and the location of the tumor reveals itself.

Waterhouse-Friderichsen syndrome is another kind of emergency. Meningococcal sepsis releases massive amounts of endotoxin, triggering disseminated intravascular coagulation; small vessels throughout the body thrombose and hemorrhage indiscriminately. The sinusoidal vessels of the adrenal medulla are especially fragile and collapse first, so the hemorrhage spreads from the central medulla outward to the cortex. Exam questions love reversing the direction, claiming the cortex collapses first, to deceive you — but as long as you remember that the most fragile center fails first, the direction can never be wrong. Once the cortex is ruined, soaking in blood, the cortisol supply is cut off instantly, producing the adrenal crisis of sudden-onset shock, hypotension, hypoglycemia, hyponatremia, and hyperkalemia.

Pheochromocytoma tests whether you can tell its two cell types apart. The chief cells are the true protagonist — they actually secrete the catecholamines and stain for chromogranin and synaptophysin; S100 stains the sustentacular cells that wrap around the periphery of the cell nests and hold the stage, not the protagonist. Diagnosis relies on plasma or urinary metanephrines, because the conversion of catecholamines into metanephrines is continuous and therefore more stable than measuring the catecholamines themselves, which fluctuate. Medication must start with an α-blocker before a β-blocker, because blocking β first leaves the vasculature with unopposed α-mediated constriction, and blood pressure surges into a crisis. The old rule of 10s is now outdated; with genetic testing now widespread, roughly 30–40% turn out to be hereditary, linked to RET, VHL, NF1, and SDHx, so genetic referral is recommended for all; malignancy is not judged by appearance but by whether metastasis is present.

The final two trailers also follow their own logic. Prolactinoma is the most common functional pituitary adenoma; once an adenoma grows past one centimeter and presses up against the optic chiasm, it produces bitemporal hemianopsia, and if it hemorrhages and undergoes necrosis internally, that is pituitary apoplexy. Craniopharyngioma arises from Rathke pouch remnants, so it is a suprasellar cystic tumor containing keratinized squamous epithelium and motor-oil-like cyst fluid that also calcifies. The two types of diabetes in the islets form a beautiful contrast: type 1 is immune lymphocytes storming straight into the islets and killing off β cells until almost none remain, hence an absolute deficiency that demands insulin and predisposes to ketoacidosis; type 2 is the amyloid protein amylin accumulating in the islets, underlain by insulin resistance plus relative deficiency, appearing alongside obesity and metabolic syndrome. One is destroyed by the immune system, the other clogged by protein — this origin not only explains what the slide looks like, but foreshadows the direction of treatment, and connects onward to the insulin-resistance throughline in the later chapter on blood lipids. The detective logic running through the whole volume of pathology is the same: look at one clue, first ask why it looks that way, fill in the whole story, and the diagnosis will surface on its own.

Think of the thyroid as the body's accelerator: secrete more and you press it deep, whole-body metabolism speeds up, so you lose weight, cannot tolerate heat, and your heart races; secrete less and you ease off, metabolism slows, so you gain weight, feel cold, your heart slows, and you feel sleepy. A young woman lost eight kilograms, has palpitations and heat intolerance, trembling hands and slightly bulging eyes, and thinks it's just stress — in truth this gland has floored the accelerator. Understand this axis of fast-versus-slow metabolism, and you can deduce most of the symptoms yourself, with no need to memorize them one by one.

To understand how to interpret these tests, you first need to understand who actually does the work in the thyroid. T4, the main hormone the thyroid secretes, is in fact the reserve currency, not very potent; it must be stripped of one iodine atom by deiodinase in the liver and kidney to become T3, three to four times more potent, before it truly acts. Once this conversion chain clicks, several scenarios become easy to understand: in critical illness or starvation, when the body wants to conserve energy, deiodinase instead converts T4 into inactive reverse T3, so T3 falls and reverse T3 rises — this is sick euthyroid syndrome, and supplementing thyroid hormone at this point actually does more harm than good. Interpretation itself needs only two axes, TSH plus free T4; the key is that TSH is the report the pituitary hands back based on feedback from thyroid hormone: when the gland itself is the problem, high thyroid hormone suppresses TSH and low thyroid hormone drives TSH up, so the two must move in opposite directions; but if the problem lies in the pituitary itself, it cannot even file an accurate report, and TSH may fail to rise, or may even be low or falsely normal — this is exactly why looking at TSH alone misses central disease, and free T4 must always be checked alongside it.

Don't mix up the two antibodies, but rather than memorizing them by rote, remember what they actually do. Hashimoto's TPO antibody grinds the thyroid down bit by bit, and once the gland collapses it heads toward hypothyroidism; Graves' TRAb stimulates the TSH receptor, effectively tricking the gland into working nonstop, hence hyperthyroidism. Graves' exophthalmos is caused by that same TRAb attacking retro-orbital tissue, so it has no direct relationship with the level of thyroid function in the blood — even after function is pushed back to normal, the exophthalmos can still progress, and smoking only pours more oil on the fire. Subacute de Quervain thyroiditis, by contrast, is the gland being damaged by inflammation and leaking out its stored hormone, so on one hand it is painful and swollen with an elevated ESR, while on the other its radioactive iodine uptake is instead low — exactly the opposite of the elevated uptake seen in Graves' disease.

Nodule evaluation begins by drawing TSH and getting an ultrasound. Fine-needle aspiration is the preoperative gold standard for distinguishing benign from malignant because it directly examines the appearance of the nucleus; the ground-glass nuclei, nuclear grooves, and psammoma bodies of papillary carcinoma can be recognized right there under the needle. But the distinction between benign and malignant follicular tumors lies in whether the capsule or vessels have been breached, something that simply cannot be seen from aspirated cells alone — the whole gland must be removed and examined pathologically, which is why fine-needle aspiration cannot distinguish follicular carcinoma. The route of metastasis is likewise decided by the tumor's nature: papillary carcinoma favors the lymphatics and carries the best prognosis; follicular carcinoma travels hematogenously to bone and lung; medullary carcinoma arises from calcitonin-secreting C cells, deposits amyloid in the stroma, and links to RET and multiple endocrine neoplasia type 2; anaplastic carcinoma is a rapidly enlarging mass in the elderly with a very poor prognosis. The surveillance marker follows the tissue of origin: after total resection of well-differentiated papillary or follicular carcinoma, follow thyroglobulin; for medullary carcinoma, follow calcitonin plus carcinoembryonic antigen. The single most important risk factor is a history of head-and-neck radiation exposure.

The treatment of thyroid storm has to throw everything at once because it must choke off the entire pathway simultaneously. Give PTU first — it inhibits synthesis while also blocking peripheral conversion of T4 to T3, which is why it is preferred over methimazole in storm; about an hour later, give the iodine solution to suppress hormone release, and this sequence is absolutely critical, because iodine is itself the raw material for thyroid hormone synthesis — give it before the antithyroid drug and you are delivering the raw material before sealing the furnace, pouring oil on the fire instead. Propranolol blocks sympathetic tone while conveniently inhibiting conversion, and hydrocortisone replaces the relative adrenal insufficiency. Precisely because iodine is a raw material, amiodarone, which carries a huge iodine load, must never be used when storm is complicated by atrial fibrillation. Incidentally, amiodarone's own iodine load can either trigger overproduction in a pre-existing nodule, causing hyperthyroidism, or directly damage the follicles and let hormone leak out — so its thyroid complications run in both directions. The three ordinary paths for treating hyperthyroidism each have their own logic too: among antithyroid drugs, methimazole is first-choice, switching to PTU only in the first trimester of pregnancy or in storm; radioactive iodine is definitive treatment but can worsen exophthalmos; surgery is reserved for a large goiter, compression, or suspected malignancy. One trap worth calling out specifically: the "radiotherapy" for hyperthyroidism refers to radioactive iodine, not the stereotactic radiation used for brain lesions. And one more life-saving reminder: if someone on an antithyroid drug develops fever plus sore throat, stop the drug first and check the blood count, because this could be agranulocytosis — never dismiss it as a cold. Finally, the emergencies at the two extremes: myxedema coma is severe hypothyroidism, and hydrocortisone must be given before thyroid hormone replacement, otherwise rapidly driving up metabolism can precipitate adrenal crisis — the same logic applies to the management of Sheehan syndrome. The whole chapter really comes down to one chain: once you have worked through T4-to-T3 conversion and negative feedback, the symptoms, the interpretation, and the order of medications can all be deduced from there.

Two beds in the emergency department: a young man breathing deep and fast, breath carrying the whiff of rotten apples, pH 7.1; next to him, an elderly woman with dry, papery skin, blood glucose 700, confused. Both are hyperglycemic emergencies, but one is drowning in acid, the other parched by osmolarity, and the source of the difference is just one thing — exactly how much insulin remains.

To understand this fork, first understand that glucose homeostasis relies on a pair of antagonistic hormones: insulin from β cells pulls glucose into cells and suppresses hepatic glucose production — it lowers glucose; glucagon from α cells forces the liver to release glucose — it raises glucose. So glucagon is actually a rescue drug for hypoglycemia, not a glucose-lowering drug, and this direction is the one most often tested in reverse. The classification also comes down to the underlying nature: type 1 is absolute deficiency from the immune system wiping out β cells, hence insulin is mandatory and ketoacidosis is common; type 2 is insulin resistance plus relatively insufficient secretion. Because the lesion in type 2 is resistance, not immune destruction, even if the patient later needs insulin to control glucose, the underlying etiology has not changed — it remains type 2 and does not thereby become type 1. This is a trap that is often slipped in.

Back to those two beds. In type 1, insulin is deficient all the way down, the body mistakenly believes it is starving, breaks down fat frantically, and fat metabolism produces massive amounts of ketones, which are acidic — so DKA drowns in metabolic acidosis, while the body desperately blows off CO₂ through deep, fast Kussmaul breathing to compensate, making it metabolic acidosis with respiratory compensation (a falling PaCO₂), not respiratory acidosis. HHS, by contrast, still has a little insulin left — too little to hold down blood glucose, but just enough to block ketone production, so it is almost never acidic; but blood glucose spikes past 600, and the glucose drags water frantically into the urine, causing violent osmotic diuresis, severe dehydration, and osmolarity blowing past 320. One is acidic, the other osmotic, and both are decided entirely by how much insulin remains.

Two electrolyte details are the most dangerous, and both can be fully explained by mechanism. Sodium: excessively high glucose pulls water from the cells into the vasculature and dilutes the serum sodium, so the measured sodium is diluted and appears low; after correction it often becomes hypernatremia, which is what actually reflects the true severity of dehydration — so the measured sodium in HHS cannot be used to judge dehydration. Potassium is even more treacherous: insulin deficiency plus acidosis drives potassium from inside the cells into the blood, so the drawn potassium looks normal or even elevated, when in fact the total body potassium has long been deficient. If you see a potassium of 5.0 at this point and feel safe starting insulin, the moment insulin goes in it will drive potassium straight back into the cells, and serum potassium will crash into fatal hypokalemia and trigger arrhythmia. Once you understand this, the order of management follows naturally: fluids first, check potassium, then insulin — begin with large-volume normal saline to restore volume and flush out the glycosuria, then check potassium; below 3.3, replace potassium first and hold insulin; once potassium is adequate, start low-dose continuous IV insulin, and once glucose falls to about 200, add glucose and continue infusing until the ketoacidosis clears. Oral glucose-lowering drugs must never be substituted throughout. Bicarbonate is not routinely replaced, and is only cautiously considered when pH falls below 6.9, because correcting it too early can instead cause paradoxical central acidification and worsen hypokalemia.

The ordinary glucose-lowering drugs likewise come alive once you grasp their mechanisms. Metformin mainly suppresses hepatic glucose production and increases sensitivity; it barely stimulates insulin release on its own, so it very rarely causes hypoglycemia and is first-choice, but long-term use impairs B12 absorption, and it must be held around iodinated contrast studies to prevent lactic acidosis. Sulfonylureas and meglitinides directly force β cells to release insulin regardless of whether glucose is actually high, so they carry the highest hypoglycemia risk. GLP-1 receptor agonists and SGLT-2 inhibitors act in a glucose-dependent manner or by excreting glucose, so their own hypoglycemia risk is low, and both can also produce weight loss, each with its own cardiovascular or cardiorenal protection; DPP-4 inhibitors merely prolong endogenous GLP-1, with a mild effect and weight neutrality — do not mistake them for a weight-loss drug. Remember that SGLT-2 inhibitors can cause euglycemic ketoacidosis and must be held before surgery. And one more direction not to reverse: thiazide diuretics raise blood glucose but do not cause hypoglycemia.

The key to chronic complications is likewise early interception. The earliest signal of nephropathy is microalbuminuria, at which point creatinine is still normal, so ACEI or ARB should be started for renal protection before creatinine rises — don't wait until the numbers have already gone bad. When osteomyelitis is suspected in a diabetic foot, a surface swab is often contaminated by skin flora and does not match the deep tissue, so bone biopsy is the gold standard for both diagnosis and antibiotic selection. Glycemic targets must be individualized — forcing an elderly patient with multiple comorbidities down to a strict target instead risks hypoglycemia, so it should be relaxed to an HbA1c of 8 to 8.5%. Finally, hypoglycemia itself: the most common cause of fasting hypoglycemia is actually a glucose-lowering drug, not an insulinoma; to tell them apart, check C-peptide — an insulinoma is the body over-secreting on its own, so insulin and C-peptide rise together, whereas exogenously injected insulin suppresses the body's own secretion, so C-peptide is instead low. Hold onto this one throughline of how much insulin remains, and every number in DKA and HHS stops being rote memorization and becomes two endings of the same story.

The pituitary gland is only pea-sized, yet it splits into two parts with very different personalities. The anterior lobe is the true factory, synthesizing six hormones on its own; the posterior lobe is merely a warehouse, storing the antidiuretic hormone and oxytocin that the hypothalamus has already made, then releasing them. Within the anterior lobe there is one special case most worth remembering: prolactin is the only hormone under primarily inhibitory control, with hypothalamic dopamine keeping its foot on the brake at all times. Understand this brake, and two things click into place. First, once a tumor compresses the pituitary stalk, dopamine cannot get through, the brake loosens, and prolactin rises mildly — so high prolactin does not necessarily come from a prolactinoma. Second, precisely because prolactin is suppressed by dopamine, prolactinoma is the only pituitary adenoma for which medication is first-choice: the dopamine agonist cabergoline both lowers prolactin and shrinks the tumor, exactly the opposite of the surgery-first approach for other secretory adenomas.

Next comes a frequently tested choice: when acromegaly is suspected, why not simply draw growth hormone instead of ordering IGF-1? Because growth hormone is secreted in intermittent pulses, and a single blood draw might catch exactly a low point, so one normal value cannot rule anything out; growth hormone, on reaching the liver, stimulates the production of IGF-1, and IGF-1 has a long half-life and a stable concentration, effectively integrating a whole day's worth of growth hormone, which is why it screens accurately. Also remember that the effector molecule that actually drives tissue overgrowth is IGF-1 — this is why a child whose epiphyses have not closed grows into gigantism, while an adult whose epiphyses have already closed can only grow at the extremities and in soft tissue, producing jaw protrusion, widened tooth gaps, and enlarged hands and feet. Confirmation uses the oral glucose suppression test: in a normal person, glucose suppresses growth hormone, and failure to suppress it in acromegaly confirms the diagnosis, after which MRI localizes the tumor. First-choice treatment is transsphenoidal surgery, with a somatostatin analog used for residual disease, and the leading cause of death is cardiovascular disease from cardiac hypertrophy.

Distinguishing the three brothers of polyuria is likewise a chain of cause and effect. The job of antidiuretic hormone is to make the renal collecting duct reabsorb water, so the moment something goes wrong with it, massive urination follows. But there are three sources of this excess urination: central means insufficient hormone secretion, nephrogenic means the kidney does not respond to the hormone, and primary polydipsia means the person is simply drinking too much water. Distinguishing them takes two steps, and each step settles one thing: water deprivation first — primary polydipsia, because the kidney is otherwise normal, obediently concentrates the urine, so it is sorted out immediately, but both central and nephrogenic causes still fail to concentrate urine, so at this point the two still cannot be told apart; then comes desmopressin replacement, and this is what reveals the truth — the central case, which was hormone-deficient, immediately concentrates its urine, while the nephrogenic case, whose kidney simply ignores the hormone, still does not rise even after replacement. The most common drug cause of the nephrogenic form is lithium. Be careful not to confuse this with the syndrome of inappropriate antidiuretic hormone secretion, which is the opposite story — excess hormone causing water retention and hyponatremia.

Sheehan syndrome is a good example of stringing several clues together: three weeks postpartum, milk not coming in, blood pressure almost too low to register, with a history of massive hemorrhage during delivery. The mechanism is that during pregnancy, prolactin cells proliferate massively and the anterior lobe swells, but the blood supply does not increase proportionally, making it especially vulnerable to ischemia; once postpartum hemorrhage triggers shock, the anterior lobe undergoes ischemic necrosis. The key is that what gets damaged is the ischemia-sensitive anterior lobe, while antidiuretic hormone lives in the posterior lobe, whose blood supply is independent and is usually spared — so diabetes insipidus is rarely seen in Sheehan syndrome, and seeing "Sheehan plus diabetes insipidus" paired together is usually a trap. The order in which anterior pituitary hormones are lost is almost exactly the order symptoms appear; the first to go is prolactin, so the first symptom is the inability to lactate. In acute decompensation, remember to replace glucocorticoid before thyroid hormone, because driving up metabolism first without cortisol support can precipitate adrenal crisis — the same logic as myxedema coma.

Hypogonadism can be cracked with just one mental shortcut: look at FSH and LH. The gonads are driven to work by these two hormones, so when the gonad itself fails and cannot produce sex hormone, negative feedback is released and FSH and LH instead surge upward — this is primary hypogonadism, and postmenopause is the most typical example: ovarian failure means FSH stays persistently high and estradiol is low, so don't remember this backwards as FSH falling. Conversely, if the pituitary or hypothalamus upstream has failed, there isn't even a driving signal to send, so FSH and LH are low — this is secondary hypogonadism, and Sheehan syndrome and Kallmann syndrome belong here. As for multiple endocrine neoplasia, type 1 arises from mutation of the tumor-suppressor gene menin, with a combination of three P's — pituitary, parathyroid, pancreas; type 2 arises from mutation of the proto-oncogene RET, carrying medullary thyroid carcinoma plus pheochromocytoma. So medullary carcinoma and pheochromocytoma belong to RET-driven type 2, not type 1, and carriers of the mutation can undergo prophylactic total thyroidectomy. Hold onto these few mechanisms — the anterior lobe that makes and the posterior lobe that stores, prolactin's dopamine brake, and growth hormone's pulsatile secretion — and every question in this small universe of the sella turcica can be worked through to the end.

A middle-aged man comes in for a health check: waist circumference 94, blood pressure 132 over 86, fasting glucose 105, triglycerides 180, HDL 38 — each value only slightly over the line, and he thinks nothing of it, but put the five together and he has already developed metabolic syndrome, a storm centered on insulin resistance that burns slowly but deep. Taiwan's diagnosis is abdominal obesity at a waist circumference of 90 for men and 80 for women, high triglycerides at 150, low HDL at 40 for men and 50 for women, blood pressure of 130 over 85, fasting glucose of 100 — meeting three of the five criteria counts.

The thing most worth working out in this whole chapter is why this set of diagnostic criteria deliberately leaves out LDL. It is because the core is insulin resistance, and the lipid change insulin resistance causes is not more LDL, but high triglycerides plus low HDL, plus LDL turning into small, dense particles — a change in quality. In other words, the problem lies in the quality of the lipoprotein, not the quantity of LDL, so the diagnosis captures only triglycerides and HDL. Also remember that Taiwan's ruler for obesity is stricter than the West's: a body mass index of 27 counts as obese, and 24 to 26.9 is overweight, because at the same index, Asians have higher body fat and higher cardiovascular risk; exam questions love giving a calculated value of 25.7 and asking you to judge it as overweight, not obese — don't get the arithmetic right only to file it under the wrong category.

Lipid-lowering drugs come alive as soon as you follow the pathway of cholesterol synthesis. The rate-limiting step of synthesis is HMG-CoA reductase, and this is exactly the enzyme statins inhibit — a reductase, not an oxidase; the moment intrahepatic cholesterol drops, hepatocytes restock by placing more LDL receptors on their surface, pulling LDL out of the blood, so lipids fall, the cardiovascular evidence is the strongest, and it is first-choice, with side effects to watch in muscle and liver enzymes. Following the pathway makes the other drugs easy to understand too: ezetimibe blocks the intestinal-absorption stop, bile acid resins interrupt the enterohepatic circulation of bile acids, fibrates activate PPARα to increase lipolytic enzyme activity and so mainly lower triglycerides and can prevent pancreatitis, and niacin's facial flushing is caused by prostaglandins, which is why aspirin can prevent it. The logic of the grapefruit juice question is likewise simple: the statins affected are the ones metabolized through CYP3A4 — grapefruit juice inhibits this enzyme, raising blood concentration and increasing muscle toxicity; pravastatin and rosuvastatin, which do not go through this pathway, are unaffected, and are preferred when a patient is on multiple medications.

PCSK9 is a player you have to think about backwards, but as long as you remember its identity you won't get it wrong: it is the demolition worker of the LDL receptor, secreted by the liver, and it seizes the receptor and drags it into the lysosome to be degraded, so the receptor cannot be recycled. So the stronger the demolition worker's activity, the fewer the receptors, and the higher the blood LDL. Conversely, if a mutation disables this demolition worker, fewer receptors get torn down, surface receptors increase, LDL is cleared more thoroughly, and cardiovascular risk is instead reduced. Exam questions love asking whether a loss-of-function mutation raises or lowers LDL — the answer is lowers, because what's broken is the demolition worker, and the outcome is actually better; what a PCSK9 inhibitor does is mimic this disablement, which is why it produces the largest reduction.

Fat itself is also an endocrine organ. Leptin is secreted by fat, and normally it reports to the hypothalamus that energy is sufficient and intake should decrease; but in obesity, fat is abundant and leptin is high too, yet the brain has become resistant to it, so leptin is abundant but useless. Adiponectin is the opposite: it is anti-inflammatory and increases insulin sensitivity, yet it instead falls in obesity, so it is scarce but precious, and inversely correlated with metabolic syndrome. As for treatment, the exam threshold for weight-loss surgery still follows a body mass index of 40, or 35 with a comorbidity; it lowers mortality and induces remission of type 2 diabetes, partly through an incretin mechanism rather than simply through weight loss. Among weight-loss drugs, the largest reduction comes from tirzepatide, a dual GIP and GLP-1 agonist, followed by GLP-1 agonists; note that DPP-4 inhibitors are merely weight-neutral and are not approved for weight loss — don't use one as a weight-loss drug. Finally, two commonly tested metabolic abnormalities each follow their own logic: although starvation in anorexia nervosa causes hypotension, bradycardia, and hypoglycemia, chronic stress instead drives cortisol up, and slow gut motility means constipation predominates rather than diarrhea — elevated cortisol and constipation are two frequently tested directions. Hold onto insulin resistance as the core of this whole chapter, and everything from the diagnostic criteria to the medications and the surgical threshold strings together into the same causal chain, with no need for scattered rote memorization.

A 58-year-old woman is brought to the clinic for low back ache, constipation, and increasing forgetfulness, assuming it is just menopause. But her blood work shows calcium at twelve point four, phosphate at two point one, chloride mildly elevated — and the parathyroid hormone that high calcium should have suppressed is still sitting above the upper limit of normal. That is the parathyroid gland acting on its own authority. The calcium-phosphate axis is like four hands locked in a tug-of-war: parathyroid hormone resorbs bone, retains calcium, drives phosphate into the urine, and activates vitamin D; active vitamin D lets the gut absorb calcium and phosphate together; calcitonin suppresses osteoclasts but plays a minor role; FGF-23 forces the kidney to waste phosphate when phosphate runs too high.

The key that unlocks this topic is: both parathyroid hormone and vitamin D raise serum calcium, so calcium alone cannot tell you which one is responsible — but their effects on phosphate run in opposite directions. Parathyroid hormone drives phosphate away, while vitamin D pulls phosphate in along with calcium, so whenever you see high calcium, your second glance must go to phosphate. Low phosphate points to excess parathyroid hormone; high phosphate is an entirely different story. That woman had primary hyperparathyroidism, most often a single adenoma secreting on its own. Her triad fingerprint can all be reasoned out from the mechanism: the gland resorbs bone furiously while forcing the kidney to retain calcium, so serum calcium is high; parathyroid hormone naturally drives phosphate into the urine, so serum phosphate is low; it also inhibits bicarbonate reabsorption in the proximal tubule, so bicarbonate is lost and chloride rises to compensate, giving high serum chloride and a tendency toward hyperchloremic metabolic acidosis. The cleanest comparison for differentiation is malignancy-associated hypercalcemia, where a tumor's parathyroid-hormone-related protein props up serum calcium, and in that case the body's own true parathyroid hormone is instead suppressed by the high calcium — so telling real from fake comes down to whether that parathyroid hormone level is abnormally elevated or suppressed. Another benign trap that mimics it closely is familial hypocalciuric hypercalcemia: its calcium-sensing receptor is disabled and cannot tell that calcium is already sufficient, so parathyroid hormone goes unsuppressed while the kidney clings to calcium for dear life, leaving urinary calcium low — the exact opposite of primary hyperparathyroidism's high urinary calcium — so urinary calcium alone can tell them apart, and it is benign and needs no surgery.

The classification in chronic kidney disease is also a timeline: the kidney loses its ability to excrete phosphate, and active vitamin D can no longer be made, so low calcium and high phosphate force all four glands into compensatory hyperplasia — this is secondary hyperparathyroidism, with calcium low or normal and phosphate high. After prolonged hyperplasia, the glands become addicted and start secreting autonomously beyond control, and serum calcium flips from low to high — this is tertiary hyperparathyroidism. The direction of calcium's effect on neuromuscular activity must also be kept straight: high calcium suppresses, leaving a person weak, constipated, and drowsy; low calcium instead over-excites the neuromuscular system, causing numbness and cramping in the hands and feet — reversing this is the most common way to lose points on the exam.

Once serum calcium surges past fourteen with symptoms, it is an emergency, and the order of treatment is itself pure logic. Hypercalcemic patients are usually severely dehydrated from polyuria and vomiting, and dehydration in turn keeps the kidney from excreting calcium — a vicious cycle — so the first choice is always a large volume of normal saline, restoring volume while flushing out urinary calcium, the fastest and safest move. Here, thiazides must never be used under any circumstance, because their pharmacologic action is precisely to spare calcium — used clinically to lower urinary calcium and treat kidney stones — so using one in hypercalcemia is pouring oil on the fire. The bisphosphonates that truly treat the root cause work by shutting down the osteoclasts resorbing bone, but they take one to three days to act, too slow to help in the moment; calcitonin instead works within hours but does not last, so the two are often combined, one for the acute rescue, one to hold the line.

On the hypocalcemia side, the first trap is not to be fooled by low albumin. About forty percent of serum calcium is bound to albumin, so when albumin is low, the measured total calcium drops along with it, but the truly active ionized calcium may not actually be low — this is pseudohypocalcemia — so you must correct for it before acting: corrected calcium equals measured calcium plus zero point eight times four minus albumin. Acid-base status also directly affects ionized calcium: alkalosis makes calcium bind albumin more avidly, so ionized calcium falls, which is exactly the mechanism behind the tetany caused by hyperventilation. The evaluation order is to first correct albumin and rule out pseudohypocalcemia, then check parathyroid hormone to distinguish whether the problem is in the gland itself or downstream, and third, check magnesium. This third step is the most commonly overlooked yet most commonly tested: magnesium is required for both the secretion and the action of parathyroid hormone, so when magnesium is low, parathyroid hormone can neither be secreted nor be heard by bone — no matter how much calcium you pour in, it will not come up until magnesium is replaced first.

Finally we arrive at bone itself. Osteoporosis is diagnosed by T-score, with a value of negative two point five or lower defining it. Treatment splits into two camps, and the most elegant point in between is parathyroid hormone's two faces: sustained elevation, as in primary hyperparathyroidism, causes net bone resorption; but made into teriparatide and given briefly and intermittently, it instead causes net bone formation — so teriparatide is a bone-building drug (as are abaloparatide and romosozumab), and it only works when given intermittently, which echoes parathyroid hormone's dual nature from earlier. All the rest — bisphosphonates, denosumab, raloxifene, calcitonin — are antiresorptive. One detail often tested backward is that acute pain relief for a vertebral compression fracture relies on anti-inflammatory analgesics and short-term calcitonin; antiresorptive agents are meant to prevent further fractures, not to relieve acute pain. The whole chapter closes in one sentence: four hands locked in a tug-of-war — see calcium, check phosphate first, and the direction is mostly settled.

Three women with menstrual problems sit down in the clinic one after another: a sixteen-year-old who has never had a period, petite, with a webbed neck; a twenty-eight-year-old with a fuller build, a face full of acne, and a period only once every six months; a thirty-year-old whose cycle was regular for over a decade but has suddenly stopped for three months. The first move in all three stories is the same: a pregnancy test. This is an ironclad rule of gynecologic endocrinology, because the most common cause of secondary amenorrhea is pregnancy, so before drawing a single tube of hormones, you test for pregnancy first, and only once it is excluded do you branch the workup. When branching, keep the definitions straight: primary amenorrhea is defined by no menarche by age fifteen, and secondary amenorrhea is a woman who has previously menstruated stopping for three months or more.

To make sense of these conditions, you first have to understand the logic of the menstrual gear-train, and the most central, most frequently tested piece is estrogen's singular dual nature. When estrogen concentration is not high, it acts as a brake, its negative feedback holding down FSH; but once the concentration surges and stays high long enough, the system is instead ignited, triggering the LH peak and releasing the egg. This is the only positive feedback in the entire cycle, and remembering it unlocks both halves of the cycle: the follicular phase runs on FSH nurturing the follicles while estradiol, still low, holds FSH down, until estradiol spikes just before ovulation and flips into positive feedback to trigger ovulation; afterward the luteal phase hands control to progesterone, which turns the endometrium into the secretory phase suited for implantation, and the corpus luteum survives a fixed span of about fourteen days, regressing into menstruation unless hCG takes over. Beneath these gears lies a division of labor between two cell types: the theca cell is driven by LH to make androgen as raw material, but it cannot aromatize this itself, so it hands the job to the granulosa cell, which is driven by FSH and uses aromatase to convert the androgen into estradiol. So rather than memorizing pairings, remember this production line instead: LH supplies the raw material, FSH handles the processing into estrogen. One counterintuitive point worth noting in passing: GnRH must be secreted in pulses to sustain the gonads, and if it is instead given continuously without interruption, it actually suppresses the gonads — this is exactly the pharmacologic basis for treating fibroids and precocious puberty.

That twenty-eight-year-old had polycystic ovary syndrome, and the entire story can be traced back to insulin resistance. High insulin simultaneously drives the theca cell to overproduce androgen and suppresses the liver's sex-hormone-binding globulin, raising free androgen even further, resulting in hirsutism, acne, anovulation, and infrequent periods. The critical consequence is this: chronic anovulation means no corpus luteum and no progesterone to close out the cycle, so the endometrium is continuously stimulated by unopposed estrogen, and endometrial hyperplasia — even elevated risk of endometrial cancer — follows. Understanding this turns treatment from rote memorization into logic: giving cyclic progesterone is really supplying the closing step that an anovulatory endometrium is missing, preventing cancer, not merely regulating the cycle. Diagnosis uses the Rotterdam two-out-of-three criteria, and it is worth noting that although obesity and insulin resistance commonly coexist, they are not part of the diagnostic criteria. Treatment depends entirely on what the patient wants: someone obese with no desire for fertility should lose weight first rather than start with ovulation-inducing drugs, and when ovulation induction is truly needed, letrozole is now first-line, with a higher live-birth rate than clomiphene.

That sixteen-year-old — primary amenorrhea is worked out along a single axis: first check the breasts, then the axillary and pubic hair, and finally the uterus. The two most worth understanding are the ones with breast development: complete androgen insensitivity is 46,XY, and the body actually has testosterone, but the receptor that receives androgen is broken, so androgen has "no effect" — no axillary or pubic hair — while the small amount of testosterone still gets aromatized into estrogen, which is enough to support breast development, and because the testes' anti-Müllerian hormone still functions normally, there is no uterus — so this is breasts, no hair, no uterus. Müllerian agenesis is 46,XX, with both the ovaries and androgen entirely normal, so the breasts, axillary hair, and pubic hair are all present — it is only that the Müllerian duct itself never developed, leaving no uterus and no upper vagina, and because this duct shares its origin with the urinary tract, the kidneys must always be screened. As for those with no breast development, it comes down to FSH: Turner syndrome, where the gonads themselves have failed, has a high FSH; Kallmann syndrome, where the central GnRH drive is missing, has a low FSH and is also accompanied by loss of smell.

The last few clinical decisions all share the same underlying logic. Hormone replacement comes down to one question: is there a uterus that needs protecting? If a uterus is present, progesterone must be added alongside estrogen to oppose estrogen's stimulation of the endometrium and prevent endometrial cancer; if the uterus has been removed, estrogen alone is used, since there is no endometrium left to protect, and adding progesterone would only raise the risk of breast cancer. In uterine structural anomalies, because the ovaries are not of Müllerian origin, these patients have normal secondary sexual characteristics and normal ovaries — the problem is only that the menstrual blood has nowhere to go — so in vaginal agenesis, progesterone must never be given to induce a withdrawal bleed, since with no outflow it only causes more pain, and a surgical outflow tract must be created instead. The GnRH agonist used for uterine fibroids relies on the same logic as before — continuous administration instead suppressing the gonads — producing a low-estrogen state that shrinks the fibroid, but it causes bone loss, so it is used only within six months before surgery. Postmenopausal bleeding carries one ironclad rule: treat it as endometrial cancer until proven otherwise, because although the most common cause is actually endometrial atrophy, the cost of missing endometrial cancer is too high, so the first clinical priority is always to rule it out first. Hold onto these few core threads — estrogen's dual nature, the two-cell production line, and the Müllerian origin of the internal genitalia — and gynecologic endocrinology turns from scattered memorization into a single chain of reasoning.

This section is a master key: given any hormone at all, you can work out which receptor it uses and how fast it acts. What strings all of this together is one chain — chemical nature decides solubility, solubility decides where the receptor sits, and where the receptor sits decides how fast or slow the action is.

Lipid-soluble hormones can dissolve straight through the cell membrane and enter the nucleus to alter gene transcription; altering genes takes time and the effect lasts long, so steroids, thyroid hormone, and active vitamin D are all slow and persistent. Water-soluble hormones cannot dissolve through the membrane, so they can only stop at the door and knock on a membrane receptor, relaying the message inside through a second messenger, which is why peptides and amines are all fast and brief. Understanding this cracks two classic traps: although T4 is derived from tyrosine, it is lipid-soluble and uses a nuclear receptor; epinephrine, also derived from tyrosine, is instead water-soluble and uses a membrane receptor. What matters is not the origin, but whether it can dissolve through the membrane. For the segment from hypothalamus to pituitary, just remember that the anterior lobe manufactures and the posterior lobe stores. The hypothalamus releases hormones that drive the anterior lobe, and the anterior lobe manufactures its own trophic hormones to govern the target glands; the posterior lobe manufactures no hormones at all — it is merely a warehouse that stores and releases the antidiuretic hormone and oxytocin that the hypothalamus has already made. So the attributions frequently tested are really an extension of the same logic: GnRH originates from the hypothalamus, and although antidiuretic hormone and oxytocin are released by the posterior lobe, they are synthesized by the hypothalamus.

The adrenal cortex, from outside in, is salt, sugar, sex. The outermost zona glomerulosa makes aldosterone, and what is special is that it can be directly stimulated by hyperkalemia, because excreting potassium is precisely its job — when potassium runs high, of course it must step up. The middle zona fasciculata's cortisol and the innermost zona reticularis's weak androgens are both governed by ACTH through the cAMP–protein kinase A pathway, which is exactly why the classic comparison gets tested: DHEA is regulated by ACTH, not LH. Cortisol peaks in the early morning because it follows the biological clock and the pulses of ACTH; chronic excess erodes bone and produces central obesity and proximal myopathy, making it the hormone least likely to increase bone mass. The medulla behaves like a specialized postganglionic neuron, releasing epinephrine the instant preganglionic sympathetic neurons contact it directly with acetylcholine. The full presentation of 21-hydroxylase deficiency is an elegant diversion chain: once this enzyme is missing, cortisol and aldosterone cannot be made, negative feedback is lost, ACTH skyrockets and drives precursor production relentlessly, but the downstream path is blocked, so with nowhere else to go, the precursors are diverted into the androgen pathway that does not need this enzyme — androgen surges, virilizing female infants' external genitalia, while aldosterone deficiency simultaneously causes salt-wasting: low sodium, high potassium, low blood pressure. For contrast, remember one counterpart, 11-hydroxylase deficiency, whose accumulated deoxycorticosterone has mineralocorticoid activity, causing hypertension instead.

Primary hyperaldosteronism also follows its mechanism logically: autonomous aldosterone secretion causes sodium retention, raising both blood pressure and blood volume, and the expanded volume in turn suppresses renin — this low renin is precisely the watershed separating it from secondary hyperaldosteronism, where renin is high. At the same time, potassium and hydrogen are excessively excreted, producing hypokalemia plus metabolic alkalosis, and hypokalemia in turn drags down insulin secretion, causing glucose intolerance. So its fingerprint is hypertension, hypokalemia, alkalosis, and low renin, screened for with the aldosterone-to-renin ratio. Pheochromocytoma, often tested alongside it, is likewise an adrenal tumor plus hypertension, but because catecholamines are released in bursts, it presents as episodic hypertension plus palpitations, headache, and sweating, whereas hyperaldosteronism presents as sustained hypertension plus hypokalemia.

Several reinforcing points of physiology are likewise all cause and effect. Insulin secretion is a depolarization chain: the β cell senses blood glucose through GLUT2, takes it up, and glycolysis generates ATP; the ATP closes the ATP-sensitive potassium channel, the membrane depolarizes, the calcium channel opens, calcium floods in, and vesicles exocytose insulin. Understanding this explains why sulfonylureas cause hypoglycemia — they directly close that same potassium channel, forcing insulin out regardless of whether blood glucose is high. Keep straight that GLUT2 is the sensor in the β cell and liver (human β cells actually rely mainly on GLUT1, but the board answer is GLUT2), while GLUT4 is the transporter insulin recruits to move glucose into muscle and fat. The gonadal axis is symmetric between the sexes: in men, the Leydig cell is driven by LH to make testosterone, and the Sertoli cell is driven by FSH to nurture sperm and secrete inhibin, which feeds back to suppress FSH — exactly mirroring the theca and granulosa cells in women. So in Klinefelter syndrome, the testes themselves have failed, testosterone is low, negative feedback is lost, and FSH and LH rise instead. Growth hormone acts along two lines at once: directly raising blood glucose, promoting lipolysis, and opposing insulin on one hand, while on the other driving the liver to make IGF-1 to promote growth; it is secreted in pulses, peaking during nocturnal deep sleep, so a single random draw has limited value, and diagnosing acromegaly requires showing that growth hormone fails to be suppressed after an oral glucose load. On the appetite side, leptin is secreted by fat and reports to the hypothalamus that energy is sufficient and intake should fall, while ghrelin is secreted by the stomach, rising before meals and falling after, the only gastrointestinal hormone that stimulates appetite — one suppressing, one promoting. One last clinical rule: when the entire pituitary fails, both ACTH and cortisol fall, giving secondary adrenal insufficiency, and when stress such as surgery, infection, or trauma strikes, a stress dose of cortisol must be given, or an adrenal crisis may be forced out; it differs from primary Addison disease in that secondary disease's aldosterone, propped up by the renin system, is usually normal and shows no hyperpigmentation, whereas primary disease shows both hyperpigmentation and salt-wasting because of elevated ACTH. Hold onto this section's three core threads — solubility decides the receptor, the anterior lobe manufactures while the posterior lobe stores, and reasoning backward from negative feedback — and any unfamiliar hormone at all can be worked out.

A 62-year-old man feels a hard mass in the front of his neck, and the fine-needle aspiration report reads ground-glass nuclei, nuclear grooves, psammoma bodies seen — the pathologist knows the direction at a glance. Tumors of the head and neck rarely rely on guesswork; they rely on a handful of features that can name them outright, but these features are not memorized for no reason — there is logic behind every one of them.

Thyroid cancer looks like four types, but it really comes down to only two lines of origin, and once you grasp the origin, everything sorts itself out. One line is follicular epithelium, which gives rise to two well-differentiated cancers, papillary and follicular, while anaplastic carcinoma is the worst-case outcome of follicular epithelium dedifferentiating and regressing; the other line is the calcitonin-secreting C cell, standing entirely on its own, giving rise to medullary carcinoma. Understanding these two lines means the features of each cancer no longer need rote memorization. Why does papillary carcinoma get diagnosed by nuclear features rather than by seeing a papillary architecture? Because its essence is precisely that set of nuclear changes — ground-glass nuclei, nuclear grooves, intranuclear pseudoinclusions — and even the follicular variant still counts as papillary carcinoma, while conversely, a conspicuous, large nucleolus is not one of its features. Why can fine-needle aspiration not distinguish follicular carcinoma from benign disease? Because the difference between benign and malignant lies in whether the tumor has broken through the capsule or invaded blood vessels, which requires examining the whole specimen — cells pulled out by a needle simply cannot show this. Since medullary carcinoma arises from the calcitonin-secreting C cell, it is recognized by calcitonin and stromal amyloid, and it is also tracked using calcitonin plus carcinoembryonic antigen, whereas the follicular-derived papillary and follicular carcinomas are tracked using thyroglobulin. About a quarter of medullary carcinomas are hereditary, driven by RET mutations and linked to multiple endocrine neoplasia type 2; there is one clinical rule of iron here — in these patients, pheochromocytoma must be excluded before thyroid surgery, or intraoperative stimulation under anesthesia can make the adrenal tumor release a surge of catecholamines, sending blood pressure into a crisis.

The parathyroid is worked out all at once using three axes — serum calcium, serum phosphate, and parathyroid hormone — with only one core question: is parathyroid hormone rising actively or passively. In primary disease, the gland secretes on its own accord and pulls calcium up passively, so calcium is high, phosphate is driven away and low, and parathyroid hormone is actively high. Secondary disease originates in chronic renal failure — the kidney loses its ability to excrete phosphate and can no longer make active vitamin D, and the resulting low calcium and high phosphate force all four glands into compensatory hyperplasia, so calcium is low or normal, phosphate is high, and parathyroid hormone is passively high. After prolonged hyperplasia the glands can become addicted and secrete autonomously beyond control, and at that point serum calcium flips from low to high while parathyroid hormone remains stubbornly elevated — this is progression to tertiary disease, and surgery should be considered. Distinguishing this from malignancy-associated hypercalcemia comes back to the same principle: malignancy-associated hypercalcemia is propped up by a tumor's parathyroid-hormone-related protein, while the body's own true parathyroid hormone is instead suppressed by the high calcium, so checking whether that parathyroid hormone level is elevated or suppressed tells them apart; outpatient hypercalcemia is usually primary hyperparathyroidism, while inpatient hypercalcemia is usually malignancy.

The next several groups of tumors each follow one easy-to-grasp thread. Thymic tumors are the anterior mediastinum's four T's; thymoma is most often associated with myasthenia gravis, while thymic carcinoma has squamous cell carcinoma as its most common histologic type. Salivary glands follow one general rule — the smaller the gland, the more likely it is malignant — so the parotid is mostly benign, while the sublingual and minor salivary glands have the highest proportion of malignancy; to distinguish the two common benign tumors, pleomorphic adenoma contains a chondroid stroma, while Warthin tumor is oncocytic epithelium plus lymphoid stroma, with no cartilage, and is associated with smoking. The trap most worth breaking is that the most common malignant salivary gland tumor is mucoepidermoid carcinoma — do not reflexively apply the squamous cell carcinoma that is common elsewhere in the head and neck; adenoid cystic carcinoma, meanwhile, favors perineural invasion, which is why it causes pain, and shows a cribriform pattern. The region from the nasal cavity and nasopharynx to the oral cavity and oropharynx is sorted by virus, and what is most worth remembering is that the correspondence between site and virus has biological meaning: nasopharyngeal carcinoma is associated with EBV, while oropharyngeal squamous cell carcinoma is associated with HPV-16, is p16-positive, responds well to treatment, and therefore carries a better prognosis — an entirely different story from the smoking-, alcohol-, and betel-nut-driven squamous cell carcinoma of the oral cavity. The malignant transformation risk of oral mucosal lesions is also frequently tested: erythroplakia carries roughly a fifty percent transformation risk, far higher than leukoplakia's roughly five percent, so when a question asks which one should more urgently be biopsied, the answer is erythroplakia — do not be misled by leukoplakia simply because it looks more extensive. Finally, three major ocular pathologies worth noting in passing: phthisis bulbi is the end stage of various severe eye injuries, with the globe atrophied and calcified; keratoconus's main pathology is thinning of the corneal stroma, and rupture of Descemet's membrane is seen only during the acute hydrops phase and is not the main cause; ocular melanoma is the most common primary intraocular malignancy in adults, and the iris type progresses slowly with a relatively good prognosis. Grasp the lines of origin and the three-axis reasoning first, then work out the correspondence between virus and site, and the hardest questions in head and neck pathology turn into problems you can name at a glance.

A patient who has just received a kidney transplant is started on cyclosporine (CsA) — his gums overgrow, his blood pressure climbs; another patient on tacrolimus (FK506) instead develops new-onset diabetes. Both are immunosuppression, so why do the side effects diverge? The answer is not rote memorization but which station along the signaling relay this drug is stuck at. Grasp that single station for each drug in this section, and reading it becomes as clear as tracing a route on a map.

T-cell activation is a relay race: antigen engages the receptor, calcineurin dephosphorylates NFAT, NFAT enters the nucleus and transcribes IL-2, and IL-2 in turn activates mTOR and drives cell proliferation. Each drug jams a different leg of that relay. Cyclosporine (CsA) and tacrolimus (FK506) bind different immunophilins, yet both jam the calcineurin station and bring IL-2 down; sirolimus also binds FKBP, but what it jams is mTOR downstream instead. Once you understand that the stations differ, it becomes clear why sirolimus can be stacked with the first two rather than being contraindicated — they are not blocking the same link. Corticosteroids act through an entirely different route: they dissolve across the cell membrane, bind intracellular nuclear receptors to alter gene expression, and induce lipocortin, which inhibits phospholipase A2 — shutting down prostaglandins and leukotrienes at the source. That is why they act through nuclear rather than membrane receptors. Their most common long-term side effect is osteoporosis, because they suppress osteoblasts, promote osteoclasts, and impair intestinal calcium absorption all at once — three routes chiseling away at bone simultaneously. And because they normally stand in for the body's own cortisol and let the adrenal gland grow idle, long-term use must never be stopped abruptly, or the adrenal gland will be unable to catch up in time and a crisis follows.

The most worthwhile trade-off to understand among the thyroid drugs is PTU versus methimazole. Both inhibit thyroid peroxidase and block hormone synthesis, but PTU has one extra trick — it also blocks the peripheral conversion of T4 into active T3 — so it is preferred for thyroid storm, where a fast, forceful effect is needed. Ordinary hyperthyroidism instead uses methimazole, once daily and lower in hepatotoxicity, switching to PTU only in the first trimester of pregnancy because methimazole is teratogenic. Bone metabolism drugs split into two camps, and the most elegant contrast among them is still the two-faced nature of parathyroid hormone: bisphosphonates, denosumab, raloxifene, and calcitonin are all antiresorptive, while teriparatide is an anabolic agent (as are abaloparatide and romosozumab) — but it only builds bone if given intermittently and briefly, because sustaining a chronically elevated parathyroid hormone level, the way disease does, instead eats away at bone. The dosing pattern decides whether it is friend or foe.

Most of the exam points among the diabetes drugs can also be explained by mechanism. Sulfonylureas directly shut the β-cell's potassium channel, forcing it to release insulin whether blood sugar is high or not, so they cause hypoglycemia and weight gain; repaglinide works the same way but is short-acting and excreted mainly through bile, so it can actually be used in renal insufficiency. Metformin inhibits hepatic glucose production and increases sensitivity, and combined with alcohol it easily causes lactic acidosis. Remember two extra-glycemic uses beyond blood sugar: GLP-1 agonists are mostly injected (oral semaglutide is the exception) and promote weight loss, while SGLT-2 inhibitors work through glucosuria and diuresis and are now standard therapy for heart failure and chronic kidney disease, though they can cause euglycemic ketoacidosis and must be stopped before surgery. Also remember that oral hypoglycemics can never be used in type 1 diabetes, because type 1 β-cells have already been wiped out and there is no insulin left to coax out. On the reproductive endocrine side, what you need to understand is why the three selective estrogen receptor modulators have such wildly different uses — the key is whether each one agonizes or antagonizes in a given tissue: clomiphene antagonizes the estrogen receptor in the hypothalamus, releasing negative feedback and raising FSH and LH, which is why it is used to induce ovulation; tamoxifen antagonizes in the breast yet agonizes in the uterus, so alongside treating breast cancer it carries a risk of endometrial cancer; raloxifene produces no uterine stimulation at all. For misoprostol, remember it is a prostaglandin E1 analog that both contracts the uterus and protects the gastric mucosa.

Growth hormone–related drugs are chosen by the level of the lesion — the logic is simply to replace or block whatever level is broken. Simple growth hormone deficiency is treated by replacing growth hormone directly. Laron syndrome breaks the growth hormone receptor, so no amount of growth hormone can be received; you must bypass it and replace IGF-1 downstream directly. Acromegaly is growth hormone excess, so replacing it is obviously contraindicated — octreotide is used instead to suppress secretion. Note that octreotide inhibits gut motility, so it worsens constipation rather than treating it. For hematopoietic growth factors, just remember which signal each one travels through: erythropoietin travels through JAK2 and STAT5 and treats renal anemia, but it is contraindicated when hypertension is poorly controlled because it raises thrombosis risk; thrombopoietin agonists act on the thrombopoietin receptor — do not mistake it for the G-CSF receptor. The remaining scattered drugs each carry their own clear chain of cause and effect: spironolactone antagonizes the aldosterone receptor, so it is used preoperatively in primary hyperaldosteronism to correct hypokalemia and hypertension; tolvaptan antagonizes the V2 receptor and clears water without clearing sodium, so it treats hyponatremia, while desmopressin is a V2 agonist that treats central diabetes insipidus and can also release Factor VIII and von Willebrand factor to promote clotting. Most elegant of all is cinacalcet, which tunes the calcium-sensing receptor to be more sensitive, tricking the parathyroid gland into believing serum calcium is high, so it secretes less parathyroid hormone — and both parathyroid hormone and serum calcium fall together, a direction of action exactly opposite to parathyroid hormone itself. The whole section distills into one mental rule: which station the signal is stuck at decides how a drug acts and what side effects it carries — grasp the station, and the entire set of drugs unfolds like an open map.