What it is:This appendix isn't one disease, it's a filing system. The kidney can only fail in a handful of structural ways (kill tubule cells, choke off blood flow, block the plumbing, attack the filter, inflame the interstitium, or clot/occlude the vessels) and almost every nephrotoxic drug you'll ever see maps cleanly onto one of those six buckets.
The core problem:A rising creatinine on its own tells you almost nothing about which drug did it or how. Dozens of drugs can cause it, through completely different mechanisms, on completely different timelines, and needing completely different responses. Treat them all the same ("just hold the med and recheck labs") and you'll miss the ones that need an immediate stop versus the ones that just need a dose adjustment or better hydration.
What you do about it:Sort the injury into one of the six structural-functional patterns first, since the pattern tells you the timeline, the workup, and the urgency. Then layer on the two things pharmacists actually get paged about: preventing it before it happens (aminoglycoside dosing, contrast prep) and recognizing the drugs most likely to cause the pattern you're worried about.
Every drug in this appendix is nephrotoxic through one of two broad routes: it's directly toxicto a specific kidney structure (tubule cell, glomerulus, interstitium, vessel), or it changes the physics of filtrationwithout killing anything (squeezing off blood flow, crystallizing in the tubule, obstructing the outflow tract). Direct toxicity tends to be dose- and duration-dependent and often reversible if you catch it early. Physics-based injury tends to be fast, related to a specific trigger (a contrast bolus, a volume-depleted patient starting an ACE inhibitor), and often preventable with the right prep. Ask "is this drug poisoning a structure, or is it changing the pressure and flow" before you go looking for the mechanism, it narrows things down immediately.
The handbook's master table sorts every drug-induced kidney injury into six structural-functional categories. Learn the categories before the drug lists, since the category is what tells you what to expect clinically.
| Pattern | What's actually damaged | Classic tempo |
|---|---|---|
| Tubular epithelial cell damage | Acute tubular injury/necrosis (direct toxin kills tubule cells) or osmotic nephropathy (tubule cells swell and vacuolate from an osmotic load) | Days, dose- and duration-dependent |
| Hemodynamically mediated injury | Nothing structurally, at first. Glomerular blood flow and filtration pressure drop because a drug pulls out one of the two "vasodilator vs vasoconstrictor" balance points that keep GFR stable | Hours to days, often tied to a trigger (volume depletion, a new dose) |
| Obstructive nephropathy | Crystals, stones, or calcium deposits physically block tubules or the collecting system | Hours (crystal) to weeks (stones, nephrocalcinosis) |
| Glomerular disease | The filtration barrier itself (podocytes, basement membrane) is immunologically or directly injured | Weeks to months |
| Tubulointerstitial disease | Allergic or chronic inflammatory infiltrate in the interstitium, or ischemic necrosis of the renal papilla | Acute allergic form: days to weeks after exposure. Chronic form: months to years |
| Vasculitis, thrombotic microangiopathy, thrombosis, cholesterol emboli | The renal vasculature itself: inflamed, clotted, or physically plugged with cholesterol debris | Variable, cholesterol emboli classically days after an arterial procedure |
Notice cyclosporine and tacrolimus appear under tubular injury, hemodynamic injury, andchronic interstitial disease. Calcineurin inhibitors are vasoconstrictive to the afferent arteriole (hemodynamic), directly tubulotoxic at high levels (tubular), and cause chronic fibrosis with prolonged exposure (interstitial). NSAIDs show up under hemodynamic injury, minimal change disease, membranous disease, AND acute interstitial nephritis. A single drug rarely has one single kidney story, match the pattern to what the labs and timeline actually show in your specific patient, don't assume from the drug name alone.
Before the drug lists, the one-sentence mechanism behind each pattern, because the mechanism is what makes the drug list make sense instead of feeling like rote memorization.
Tubular injuryhappens because the proximal tubule does the heaviest lifting of any nephron segment, reabsorbing and concentrating everything that gets filtered, which also means it concentrates whatever toxin is in that filtrate to levels far higher than plasma. That's why aminoglycosides, cisplatin, and amphotericin B, all of which get taken up and trapped inside tubule cells, are tubular poisons first and foremost.
Hemodynamic injuryhappens because GFR in a compromised kidney is often being propped up by exactly two opposing forces: angiotensin II constricting the efferent arteriole (keeping pressure up inside the glomerulus) and prostaglandins dilating the afferent arteriole (keeping blood flowing in). Pull out either one, an ACE inhibitor/ARB removes the efferent constriction, an NSAID removes the afferent dilation, and filtration pressure collapses, especially in a kidney that was relying on that compensation because of volume depletion or preexisting renal artery narrowing.
Obstructive nephropathyhappens because some drugs are poorly soluble in acidic or concentrated urine and simply fall out of solution, forming crystals inside the tubule lumen (acyclovir, sulfonamides, methotrexate) or stones downstream (indinavir, triamterene). Volume status and urine pH are the two variables that determine whether this happens, which is exactly why hydration is the universal prevention strategy.
Glomerular diseasehappens through immune-mediated podocyte injury (minimal change disease, membranous nephropathy) or direct podocyte toxicity (sirolimus, anabolic steroids causing FSGS), rather than a simple dose-response toxin effect, which is why it can show up after a drug has been tolerated for a long time.
Tubulointerstitial diseaseis mostly a hypersensitivity reaction: T-cell mediated inflammation targeting the space between tubules, triggered by a drug acting as a hapten. It is not dose-dependent and can happen after a single dose in a sensitized patient, which is the single biggest thing that separates it from tubular injury.
Vascular injuryhappens through ANCA-associated vasculitis (hydralazine, propylthiouracil, allopurinol), direct endothelial toxicity causing thrombotic microangiopathy (calcineurin inhibitors, gemcitabine, mitomycin C, some VEGF-pathway agents like bevacizumab), or mechanical plugging of small vessels with cholesterol debris dislodged during an arterial procedure and unmasked by anticoagulation.
Before you even open the drug list, ask: did this look like it built up with cumulative dose and duration (tubular injury, chronic interstitial disease), did it happen fast around a specific trigger (hemodynamic, crystal, cholesterol emboli), or did it happen out of proportion to dose in a way that smells allergic (acute interstitial nephritis, minimal change disease, vasculitis)? That single question does most of the pattern-recognition work before you ever need the full table.
The proximal tubule reabsorbs roughly two-thirds of everything filtered, so anything toxic that passes through the glomerulus gets concentrated here before it goes anywhere else. This is the single most commonly tested pattern.
| Drug / class | Why it's tubulotoxic |
|---|---|
| Aminoglycosides(gentamicin, tobramycin) | Actively taken up by proximal tubule cells via megalin-mediated endocytosis and accumulate to levels far exceeding plasma; classic cumulative-dose, cumulative-duration toxicity |
| Radiographic contrast media | Combination of direct tubular toxicity and renal medullary hypoxia from vasoconstriction; risk rises sharply with preexisting CKD, diabetes, and volume depletion |
| Cisplatin, carboplatin | Platinum accumulates in proximal tubule cells and cross-links DNA; cisplatin is far more nephrotoxic than carboplatin |
| Ifosfamide | Chloroacetaldehyde metabolite is directly tubulotoxic, notably at the proximal tubule (can cause a Fanconi-like picture) |
| Amphotericin B | Binds tubular cell membrane cholesterol, creates pores, causes both direct cytotoxicity and severe renal vasoconstriction |
| Cyclosporine, tacrolimus | Direct tubular and vascular toxicity at high trough levels, on top of their separate chronic hemodynamic effect |
| Adefovir, cidofovir, tenofovir | Actively transported into proximal tubule cells; mitochondrial toxicity from nucleotide/nucleoside analogue accumulation |
| Pentamidine | Direct tubular toxicity, can also cause hyperkalemia through a separate collecting-duct mechanism |
| Foscarnet | Crystallizes within the tubule and is also directly cytotoxic; needs aggressive volume expansion during infusion |
| Zoledronate | Direct proximal tubular toxicity, particularly with rapid infusion or in volume-depleted patients |
Mannitol, dextran, sucrose-containing IVIG formulations, and hydroxyethyl starch don't poison the tubule cell, they overload it. The cell takes up the osmotically active solute, water follows, and the cell swells and vacuolates until it can't function. It's a volume/dose problem, not a direct-toxin problem, and it's why IVIG-related renal injury is largely avoidable by choosing a sucrose-free formulation in at-risk patients.
Nothing dies here, at least not at first. GFR falls because the pressure gradient across the glomerulus that depends on drugs is disrupted, and it typically reverses if you catch it and act before the hypoperfusion becomes sustained enough to cause secondary tubular injury.
| Drug / class | Mechanism |
|---|---|
| ACE inhibitors, ARBs | Block angiotensin II-mediated efferent arteriolar constriction; drops intraglomerular pressure, especially dangerous with bilateral renal artery stenosis or volume depletion |
| NSAIDs | Block prostaglandin-mediated afferent arteriolar dilation; the kidney loses its ability to compensate for reduced effective circulating volume |
| SGLT-2 inhibitors | Restore tubuloglomerular feedback and reduce intraglomerular pressure by increasing distal sodium delivery; usually a modest, expected early eGFR dip rather than true injury, but can be exaggerated by concurrent volume depletion |
| Cyclosporine, tacrolimus | Direct afferent arteriolar vasoconstriction, dose- and level-related |
| Chimeric antigen receptor (CAR) T-cell therapy | Cytokine release syndrome drives systemic hypotension/capillary leak, causing prerenal hemodynamic injury on top of any direct effect |
| OKT3, high-dose interleukin-2 | Cytokine-release-mediated capillary leak and hypotension, same hemodynamic final pathway as CAR-T |
Stacking these three removes both compensatory mechanisms (efferent constriction AND afferent dilation) while also shrinking effective circulating volume. This combination is one of the highest-yield medication-reconciliation catches in an AKI workup, and it's a trap because each drug alone is usually fine, it's the combination plus a volume-depleted patient that tips things over.
Three distinct sub-patterns, all sharing one prevention strategy: keep the urine dilute and flowing so nothing has the chance to fall out of solution.
| Sub-pattern | Drugs | Why |
|---|---|---|
| Crystal nephropathy | Acyclovir, sulfonamides, indinavir, atazanavir, foscarnet, methotrexate, ascorbic acid, ethylene glycol, orlistat, ciprofloxacin | Poorly soluble at physiologic or acidic urine pH; precipitate directly inside the tubule lumen, worse with rapid IV push and volume depletion |
| Nephrolithiasis | Sulfonamides, triamterene, indinavir, atazanavir | Same solubility problem but forming discrete stones in the collecting system rather than diffuse intratubular crystals |
| Nephrocalcinosis | Oral sodium phosphate bowel prep solution | Acute phosphate load precipitates as calcium phosphate in the renal tubules; largely why oral sodium phosphate prep has fallen out of favor |
IV acyclovir crystallizing in the tubules is largely a rate-and-hydration problem: give it slowly (over at least an hour) with the patient well hydrated, and the risk drops substantially. This is a classic "the pharmacist catches it on order verification" scenario, not something that needs a dose change so much as an administration-technique fix.
Unlike tubular injury, this pattern isn't about cumulative dose, it's about the drug triggering an immune or direct podocyte injury that can show up after months of otherwise uneventful therapy.
| Lesion | Drugs |
|---|---|
| Minimal change disease | NSAIDs, COX-2 inhibitors, lithium, pamidronate, interferon-α and β |
| Membranous nephropathy | NSAIDs, penicillamine, captopril |
| Focal segmental glomerulosclerosis (FSGS) | Pamidronate, interferon-α and β, lithium, sirolimus, anabolic steroids, tyrosine kinase inhibitors |
Unlike tubular or hemodynamic injury, which usually present as a rising creatinine with a relatively bland urinalysis, glomerular disease classically presents with nephrotic-range proteinuria (frequently with edema and hypoalbuminemia). If a patient on long-term lithium, an NSAID, or a bisphosphonate develops new significant edema and a big jump in urine protein, think glomerular pattern, not tubular.
Acute allergic interstitial nephritis (AIN) is the pattern you'll be asked about the most on rotations, because it's common, drug-induced, and its own dedicated table is coming up in full below. The chronic and papillary forms round out the interstitial category.
| Sub-pattern | Drugs | Key feature |
|---|---|---|
| Acute allergic interstitial nephritis | β-lactams, ciprofloxacin, NSAIDs/COX-2 inhibitors, proton pump inhibitors, loop diuretics, immune checkpoint inhibitors | Idiosyncratic, not dose-dependent, can follow a single exposure in a sensitized patient (see full drug list and presentation below) |
| Chronic interstitial nephritis | Cyclosporine, lithium, aristolochic acid, combination analgesics | Slow, fibrotic, often insidious over months to years; frequently only picked up on a slowly rising baseline creatinine |
| Papillary necrosis | NSAIDs, combined phenacetin/aspirin/caffeine analgesic products | Ischemic necrosis of the renal papilla from chronic medullary hypoperfusion; classically presents with flank pain and sloughed tissue in the urine |
The least common pattern on this list but the one most likely to be missed, since it doesn't look like a typical "nephrotoxic drug" picture at all.
| Sub-pattern | Drugs |
|---|---|
| Vasculitis / thrombosis / TMA | Hydralazine, propylthiouracil, allopurinol, penicillamine, gemcitabine, mitomycin C, methamphetamines, cyclosporine, tacrolimus, adalimumab, bevacizumab |
| Cholesterol emboli | Warfarin, thrombolytic agents |
Warfarin and thrombolytics don't cause cholesterol emboli directly, they unmask it. An arterial procedure (cath, aortic manipulation) dislodges plaque, and anticoagulation/thrombolysis removes the clot that would otherwise seal the embolized fragment in place, letting showers of cholesterol crystals lodge in small renal vessels. Look for the triad of a recent arterial procedure, livedo reticularis or blue toes, and eosinophilia alongside the rising creatinine, that combination is what separates this from a garden-variety hemodynamic or contrast-related AKI.
Aminoglycosides are the drug this appendix gets built around clinically, mostly because the risk factors are modifiable and pharmacists own most of them.
| Category | Risk factors |
|---|---|
| Dosing-related | Large total cumulative dose, prolonged therapy, trough concentration >2 mg/L (4.2-4.3 μmol/L), recent prior aminoglycoside course |
| Synergistic combinations | Cyclosporine, amphotericin B, vancomycin, diuretics, iodinated contrast, cisplatin, NSAIDs, all combined with an aminoglycoside |
| Patient-related | Preexisting kidney disease, diabetes, older age, poor nutrition, shock, gram-negative bacteremia, liver disease, hypoalbuminemia, obstructive jaundice, dehydration, hypotension, potassium or magnesium deficiency |
An elevated trough (>2 mg/L) means the drug never cleared between doses and tubule cells stayed saturated the whole interval, that sustained exposure is what drives toxicity. This is exactly why extended-interval (once-daily) aminoglycoside dosing exists: it achieves a high peak for bactericidal effect while giving the tubule cells a real trough-free window to recover, which lowers nephrotoxicity risk compared to traditional multiple-daily dosing at the same total dose.
When you see an aminoglycoside ordered, scan the active med list for vancomycin, amphotericin B, cyclosporine, a loop diuretic, or a recent contrast study. None of these individually is a reason to avoid the aminoglycoside, but stacking them raises the bar for how carefully you should be dosing and monitoring, and it's the kind of catch that's easy to make on chart review and easy to miss if you're only thinking about the aminoglycoside in isolation.
This is the one place in the appendix with a graded, actionable protocol rather than just a drug list, and it's built around one idea: give the kidney the least contrast necessary, and hydrate around it so the medulla never goes hypoxic.
| Intervention | Recommendation | Gradea |
|---|---|---|
| Contrast itself | ||
| Contrast volume/dose | Minimize as much as clinically possible | A-1 |
| Contrast type | Use noniodinated contrast studies when feasible | A-2 |
| Contrast osmolality | Use low- or iso-osmolar agents | A-2 |
| Medications | ||
| Concurrent nephrotoxins | Avoid concurrent potentially nephrotoxic drugs (e.g., NSAIDs, aminoglycosides) around the time of the study | A-2 |
| Isotonic sodium chloride (0.9%) | ||
| Standard protocol | Start infusion 3-12 hours before contrast, continue 6-24 hours after. Infuse at 1-1.5 mL/kg/h, titrating post-exposure to keep urine output ≈150 mL/h | A-1 |
| Urgent/unplanned protocol | Start 3 mL/kg/h beginning 1 hour before contrast, then continue at 1 mL/kg/h for 6 hours after | A-1 |
aGrade key:Letter = strength of recommendation (A good, B moderate, C poor evidence). Number = quality of evidence (1 = more than one properly randomized controlled trial, 2 = well-designed clinical trial with randomization, cohort/case-control studies, or dramatic uncontrolled results, 3 = expert opinion or descriptive studies).
The A-1 grade sits specifically on isotonic (0.9%) sodium chloride volume expansion. Other strategies you'll hear about clinically (bicarbonate infusions, N-acetylcysteine) have a much shakier evidence base and aren't the graded, protocolized intervention this appendix builds around. If you're asked "what's the one proven prevention strategy," the answer is adequate isotonic saline hydration timed around the contrast exposure, not a specific additive.
Contrast causes injury partly through renal medullary vasoconstriction and hypoxia. Volume expansion started only at the time of the scan is too late to have corrected that hemodynamic vulnerability; starting hours ahead gives the kidney a genuinely expanded, well-perfused starting point before the vasoconstrictive insult ever hits.
| Parameter | When | Watching for |
|---|---|---|
| Serum creatinine / eGFR | Baseline before any known nephrotoxin, then trended on a schedule matched to the drug's risk (daily for IV aminoglycosides/amphotericin, 48-96 hours after contrast, periodically for chronic exposures like lithium or calcineurin inhibitors) | Any pattern-appropriate rise; a slow creep with a chronic drug is just as important as a sharp jump with an acute one |
| Aminoglycoside trough (or AUC-based level) | Per institutional protocol, typically around the 3rd-4th dose at steady state, or per extended-interval nomogram | Trough >2 mg/L, the specific threshold tied to nephrotoxicity risk |
| Urine output | Continuously during and after contrast administration in an inpatient, and with any acute nephrotoxin exposure | Maintaining ≈150 mL/h during the isotonic saline protocol; oliguria as an early red flag |
| Urinalysis / urine sediment | At first suspicion of AKI in a patient on a known nephrotoxin | Eosinophils and white cell casts (AIN), granular/muddy-brown casts (ATN), crystals (crystal nephropathy), heavy proteinuria (glomerular pattern) |
| Urine protein / albumin | Periodically with chronic exposure to a drug on the glomerular-disease list (lithium, bisphosphonates, interferons, sirolimus) | New or worsening proteinuria, especially nephrotic-range, suggesting a glomerular pattern rather than tubular |
| Electrolytes (K, Mg) | Alongside creatinine with any nephrotoxic-drug course | Hypokalemia/hypomagnesemia both worsen aminoglycoside risk and are common consequences of tubular injury themselves, a two-way relationship worth watching |