Renal blood supply and autoregulation
A fifth of the cardiac output squeezes through two arterioles in series; the pressure between them sets the GFR, and most of the diseases and drugs on this page work by shifting that pressure — until ischaemia lasts long enough to turn a pressure problem into a tissue problem.
What it normally does
The kidneys take about a fifth to a quarter of everything the heart pumps — roughly 1 litre of blood a minute — but extract only a small fraction of the oxygen in it. That blood is there to be filtered, not to feed the tissue [high flow, low whole-organ oxygen extraction].
Explains why the kidney complains first when blood pressure drops, and why the hardest-working tubules in the outer medulla can still starve for oxygen despite that enormous whole-organ flow (acute tubular necrosis).
Each glomerulus hangs between two arterioles in series: one in (afferent), one out (efferent). Pressure inside the tuft — and therefore the filtration rate (GFR) — is set mainly by the balance of their tone. Squeeze the outflow and pressure rises; open the inflow and pressure rises. Squeeze the outflow hard enough, though, and plasma flow falls so far that GFR falls with it.
Explains how the kidney defends filtration when flow falls, why glomerular pressure can be too high in diabetes, and why a drug that relaxes or constricts either arteriole can move GFR within a fortnight.
Between a mean arterial pressure of roughly 80 and 180 mmHg the kidney holds its own blood flow and GFR nearly constant (autoregulation); the exact lower limit is not a fixed number and sits higher in someone with long-standing hypertension. Two mechanisms do the work: the afferent arteriole tightens when stretched (myogenic response), and the macula densa senses the NaCl concentration of fluid arriving at the end of the thick ascending limb and adjusts afferent tone through adenosine (tubuloglomerular feedback). Locally made prostaglandins buffer the afferent arteriole against constriction — a minor influence in a well-filled person, a decisive one once angiotensin II and sympathetic tone are switched on.
Explains why blocking prostaglandins injures kidneys mainly in the patients already leaning on them, why an SGLT2 inhibitor changes glomerular pressure from inside the tubule, and why below the lower limit the kidney falls off the plateau and flow simply follows pressure.
When perfusion falls, angiotensin II constricts both arterioles but the efferent more than the afferent, propping filtration pressure up while total renal blood flow drops. At the same time aldosterone and ADH drive the tubules to grab back sodium and water.
Explains the whole biochemical picture of prerenal failure — concentrated urine, almost no sodium in it, urea rising faster than creatinine — and why blocking angiotensin II in that state can drop filtration abruptly.
What goes wrong
- Prerenal acute kidney injury← from “When perfusion falls, angiotensin II constrict…”
Bleeding, vomiting and diarrhoea, sepsis or a failing heart pull mean arterial pressure below the autoregulatory range. Angiotensin II clamps the efferent arteriole and buys time, but filtration still falls. The tubules are structurally intact and working hard: they reabsorb as much sodium and water as they can, and urea is dragged back with that water while creatinine is not.
Prerenal AKI is a working kidney in a failing circulation. The abnormal numbers are the kidney doing its job, not the kidney breaking.
You would find: Oliguria, dry mucosa, tachycardia, postural drop. Urine concentrated (osmolality above 500 mosmol/kg), urine sodium under 20 mmol/L, fractional excretion of sodium under 1%, urea risen out of proportion to creatinine, bland urinary sediment. FeNa is unreliable in anyone on a diuretic or with CKD — fractional excretion of urea (under about 35%) is the fallback there. Creatinine usually falls within a day or two once volume is restored. On Australian wards the usual triggers are gastroenteritis, heat and dehydration, sepsis and decompensated heart failure.
- Acute tubular necrosis (ischaemic)← from “The kidneys take about a fifth to a quarter of…”
Let the low perfusion run on and the parts of the tubule that do the most work on the least oxygen — the straight (S3) proximal tubule and the thick ascending limb in the outer medulla — are injured and die. Dead cells slough into the lumen, form casts, obstruct flow and let filtrate leak back. Until that epithelium regenerates, the kidney cannot concentrate urine or hold on to sodium, however well the circulation is corrected.
Huge blood flow, tiny oxygen reserve in the outer medulla. Once the urine stops being concentrated, you have usually crossed from a pressure problem to a tissue problem.
You would find: Urine osmolality falls to about that of plasma (around 300 mosmol/kg), urine sodium rises above 40 mmol/L, FeNa above 2%, and muddy brown granular casts appear on microscopy. Fluid no longer restores filtration. Recovery takes days to weeks and often passes through a polyuric phase.
- NSAID kidney injury and the triple whammy← from “Between a mean arterial pressure of roughly 80…”
Locally made prostaglandins hold the afferent arteriole open when angiotensin II and sympathetic tone are trying to shut everything down. An NSAID blocks cyclo-oxygenase and removes that dilator. In a well-filled person, little happens. In someone leaning on those prostaglandins — elderly, volume deplete, heart failure, cirrhosis, CKD — the afferent constricts and GFR falls. Add an ACE inhibitor or ARB, which relaxes the efferent, and a diuretic, which drops the volume, and both arterioles and the preload are hit at once.
Triple whammy: NSAID plus ACE inhibitor/ARB plus diuretic. Take away the afferent dilator, the efferent constrictor and the volume, and little is left holding filtration up.
You would find: Creatinine climbing a few days after an NSAID is started, or after a gastro illness, in an older patient already on an ACE inhibitor and a thiazide or frusemide. Often no symptoms at all — it turns up on routine bloods.
- Glomerular hyperfiltration and progressive chronic kidney disease← from “Each glomerulus hangs between two arterioles i…”
In early type 2 diabetes the filtered glucose load rises and the proximal tubule reabsorbs extra glucose and sodium through SGLT2, so less sodium chloride reaches the macula densa. Tubuloglomerular feedback reads that low NaCl as low flow and dilates the afferent arteriole, while angiotensin II keeps the efferent tight. Inflow wide open, outflow squeezed: pressure inside the glomerulus rises. Years of high pressure push protein through the filter and scar the tuft (glomerulosclerosis).
Diabetes is the most common primary cause of kidney failure among Australians starting kidney replacement therapy (ANZDATA), and Aboriginal and Torres Strait Islander people start kidney replacement therapy at several times the rate of other Australians — higher again in remote central Australia. The lesion is a pressure lesion, so the treatment is a pressure treatment.
You would find: A high-normal or frankly raised eGFR early on, then a rising albumin:creatinine ratio on a first-morning urine (confirmed on repeat samples), then a falling eGFR. Blood pressure creeps up alongside it.
- Renal artery stenosis unmasked by an ACE inhibitor← from “When perfusion falls, angiotensin II constrict…”
A narrowed renal artery — usually atherosclerosis in an older smoker with vascular disease elsewhere, or fibromuscular dysplasia in a younger woman — delivers less pressure to the glomerulus. Filtration is then propped up largely by angiotensin II squeezing the efferent arteriole. Block angiotensin II and the efferent opens, filtration pressure falls, and GFR drops. This matters when both kidneys are affected, or when the stenosis is in a single functioning kidney; with one healthy kidney on the other side, total GFR is usually maintained.
A creatinine rise of up to about 30% after starting an ACE inhibitor is expected and is the drug working. A bigger jump, or a rising potassium, means recheck, look for volume depletion or an NSAID, and consider bilateral renal artery stenosis.
You would find: Creatinine jumping more than about 30% within one to two weeks of starting an ACE inhibitor or ARB. Supporting hints: hypertension resistant to several drugs, an abdominal bruit, recurrent flash pulmonary oedema, vascular disease everywhere else.
What we give, and how it works
Open a drug to see what it binds, what that does to the cell, and what you then see in the patient.
- Binds
- No receptor. Sodium stays in the extracellular fluid, so an isotonic solution expands plasma volume and preload (only about a quarter to a third of the volume given stays intravascular).
- Which does
- More venous return means a bigger stroke volume and higher cardiac output, so MAP rises back towards the autoregulatory range (roughly above 80 mmHg) and autoregulation can grip again.
- So you see
- Urine output returns and creatinine typically falls over 24-48 hours. The response is also informative: creatinine that improves promptly supports a prerenal picture in retrospect. But a creatinine that does not budge does NOT prove the tubules have died — before calling it established ATN, exclude obstruction on ultrasound, an ongoing insult (continuing sepsis, bleeding, nephrotoxin), and a cardiac rather than hypovolaemic cause.
- And the same mechanism causes
- 0.9% saline carries 154 mmol/L of chloride, against a plasma value of about 100. Load enough in and hyperchloraemic metabolic acidosis follows predictably. Physiologically, that extra chloride also reaches the macula densa, which senses luminal NaCl concentration and responds with adenosine-mediated afferent constriction through tubuloglomerular feedback — a plausible mechanism for reduced renal blood flow. Be careful how far you take this: in the large randomised trials the clinical signal is small. SMART (US) found a modest reduction in a composite of death, new renal replacement therapy or persistent renal dysfunction with balanced solutions, with no clear difference in the individual components; PLUS (Australia and New Zealand) found no difference in 90-day mortality or renal replacement therapy. Balanced solutions sit closer to plasma chloride and are a reasonable default, but the biochemical harm of saline is far more certain than any outcome harm.
- Handling
- The first question is whether the tank is empty or the pump is failing. If the low perfusion is a failing heart, the same fluid goes straight into the lungs. Standard practice is clinical reassessment after each bolus rather than a fixed volume ordered in advance. Note that Hartmann and Plasma-Lyte 148 both contain about 5 mmol/L of potassium, so in severe hyperkalaemic AKI that is worth a thought, though at that concentration they are not usually the driver.
Catches people out: Once tubules are necrotic, fluid no longer restores filtration — it largely accumulates in the interstitium and the lungs.
- Binds
- Cyclo-oxygenase (COX-1 and COX-2), the enzyme that makes prostaglandins from arachidonic acid.
- Which does
- Less PGE2 and PGI2 in the kidney, so the afferent arteriole loses its dilator buffer and constricts under angiotensin II and sympathetic tone that are now unopposed.
- So you see
- In a well-filled patient, usually nothing you can measure. In someone volume deplete, elderly, or in heart failure, cirrhosis or CKD, GFR falls and creatinine rises within days. Sodium and water are also retained — same prostaglandin loss, acting on the tubule — so blood pressure rises and ankles swell.
- And the same mechanism causes
- Acute kidney injury from a throttled afferent arteriole. The analgesia and the kidney injury come from the same enzyme block, which is why no systemically absorbed NSAID, selective or not, can be assumed free of it (topical preparations reach much lower systemic levels).
- Handling
- Usual practice is to withhold during any acute illness with poor intake, and to avoid combining with an ACE inhibitor or ARB plus a diuretic in an older patient. Where an NSAID is genuinely needed in someone at risk, that means the shortest course at the lowest effective dose with creatinine checked — a judgement for the prescriber, not a rule to apply from a page.
Catches people out: Celecoxib spares the stomach because gastric mucosal protection is mostly COX-1. It does not spare the kidney: renal prostaglandins come from both isoforms, with COX-2 constitutively expressed at the macula densa and in the medulla. Celecoxib also carries cardiovascular risk, so it is not a free pass in the patients most likely to have a vulnerable afferent arteriole.
- Binds
- ACE inhibitors block angiotensin converting enzyme, so less angiotensin II is made. ARBs block the AT1 receptor on vascular smooth muscle (and in the adrenal cortex), so the angiotensin II that exists cannot act.
- Which does
- Efferent arteriolar smooth muscle relaxes and less aldosterone is released.
- So you see
- Pressure inside the glomerulus falls. Albuminuria drops and eGFR declines more slowly over years. A creatinine rise of up to about 30% — with the corresponding dip in eGFR — over the first week or two, which then plateaus, is the drug working rather than the drug harming.
- And the same mechanism causes
- In a kidney whose filtration is being held up largely by efferent constriction — bilateral renal artery stenosis, stenosis in a single functioning kidney, or anyone volume deplete — the same relaxation drops filtration pressure and causes acute kidney injury.
- Handling
- Creatinine and potassium are rechecked one to two weeks after starting or increasing the dose. Sick day rules matter: the drug is held during vomiting, diarrhoea or any illness with poor oral intake, and restarted once the patient is eating and drinking normally.
Catches people out: Less angiotensin II means less aldosterone means less potassium excreted, so hyperkalaemia comes from the same mechanism. The dry cough does not: ACE (kininase II) also degrades bradykinin, so blocking it lets bradykinin accumulate in the airway. ARBs leave bradykinin degradation intact, which is why an intolerable cough is a reason to swap.
- Binds
- The sodium-glucose cotransporter 2 (SGLT2) in the early proximal tubule (S1/S2), which normally reabsorbs the bulk of filtered glucose along with sodium.
- Which does
- Glucose and sodium are left in the tubular fluid, so more sodium chloride reaches the macula densa. The macula densa senses that higher NaCl concentration and, through tubuloglomerular feedback, releases adenosine, which constricts the afferent arteriole.
- So you see
- Glomerular pressure falls, hyperfiltration is relieved, albuminuria falls and eGFR decline slows. Glucose lost in the urine takes water with it, so weight and blood pressure fall a little too.
- And the same mechanism causes
- A drop of a few mL/min/1.73m2 in eGFR over the first weeks. That is the afferent arteriole doing exactly what the drug intends; it plateaus and reverses on stopping. Stopping the drug because of it throws away the benefit.
- Handling
- A baseline eGFR, a warning to the patient about the early dip, and a recheck rather than a reaction to it — that is the usual pattern.
Catches people out: Sugar in the urine feeds yeast, so genital mycotic infection (thrush) is among the commonest adverse effects. Shifting fuel from glucose to fat also raises ketones, and euglycaemic diabetic ketoacidosis can occur with a near-normal glucose — which is why these drugs are withheld during acute illness, prolonged fasting and around surgery (Australian Diabetes Society advice is to withhold for about three days before elective surgery), and why ketones, not glucose, are the test when an unwell patient on one is acidotic.
Two arterioles, one number. Prerenal AKI: urine osmolality above 500 mosmol/kg, urine sodium under 20 mmol/L, FeNa under 1%, urea up out of proportion to creatinine, fixed by fluid. Established ATN: urine osmolality around 300, urine sodium above 40 mmol/L, FeNa above 2%, muddy brown granular casts, fluid will not fix it. Treat both sets of indices as supportive, not diagnostic — diuretics and CKD blunt them. And the triple whammy in one line: the NSAID shuts the afferent, the ACE inhibitor opens the efferent, the diuretic empties the tank.
Now test whether it stuck
Reading this through is not the same as being able to reconstruct it. Every question in the bank is free, with a full debrief on each option.