Acute interstitial nephritis and tubulointerstitial disease

• Acute interstitial nephritis (AIN), also called acute tubulointerstitial nephritis, is an inflammatory disease of the renal interstitium and tubules that leaves the glomeruli largely untouched. It is one of the common causes of acute kidney injury, found in up to 10% to 30% of kidney biopsies performed for acute kidney disease when infection and drug-related disease are common.^[2]
• In most cases the trigger is a drug. Over 95% of cases result from infection or an allergic drug reaction, not from glomerular disease, so the first question is always "what new drug was started?"^[1]
• The classic teaching triad of fever, rash and eosinophilia is present in less than 10% of medication-induced cases; the more reliable clues are sterile pyuria, white-cell casts, a modest rise in creatinine and a temporal link to a drug.^[1]
• Stopping the offending drug is the core of treatment, and renal function usually recovers within 6 to 8 weeks. Corticosteroids may speed recovery, but the optimal dose, route, timing and duration remain unsettled, and biopsy before starting them is advised.^[1][2]
• Longstanding analgesic abuse produces a different, chronic form of tubulointerstitial disease: analgesic nephropathy with renal papillary necrosis, bland sediment and often non-nephrotic proteinuria, usually in middle-aged women with years of painkiller use.^[3][4]

By the end of this page you should be able to:

1. Define acute interstitial nephritis (AIN, ATIN) and explain how it differs from acute tubular injury, pyelonephritis and glomerulonephritis.
2. List the major drug classes and non-drug conditions that cause AIN, and describe the latency of drug reactions.
3. Interpret the urine sediment in suspected AIN, including what eosinophiluria does and does not tell you.
4. Plan investigations, including the role and timing of renal biopsy.
5. Apply an evidence-based management approach: remove the trigger, support kidney function, use corticosteroids appropriately, and know when to involve nephrology.
6. Recognise and manage analgesic nephropathy: its pathogenesis, presentation, CT diagnosis and prognosis.
7. Outline chronic tubulointerstitial nephritis and its causes (drugs, heavy metals, obstruction, genetic and metabolic disease).

The tubulointerstitial compartment does most of the kidney's work: reabsorption, secretion and concentration. A disease that attacks this compartment therefore produces tubular dysfunction (concentrating and acid-base defects) rather than heavy proteinuria or red-cell casts, although both can spill in.

Tubulointerstitial disease is traditionally divided by time course and mechanism:

• Acute tubulointerstitial nephritis (ATIN). Onset over days to months, with interstitial oedema and inflammation; usually reversible if the cause is removed early. Drug allergy and infection are the dominant triggers.^[1]
• Chronic tubulointerstitial nephritis (CTIN). Insidious, partially irreversible, with tubular atrophy and interstitial fibrosis on biopsy; causes include long-term analgesic, lithium, calcineurin-inhibitor and heavy-metal exposure, chronic obstruction and inherited or metabolic disease.^[1]
• Acute tubular injury (acute tubular necrosis). Direct toxic or ischaemic damage to tubular epithelium rather than immune-mediated interstitial inflammation; the sediment and course differ, although both can follow drugs.
• Pyelonephritis. Infection extending into the parenchyma; the urine is not sterile and bacterial culture is positive, which is not usual in drug-related AIN unless a concurrent infection is triggering it.

A useful clinical habit is to classify a rising creatinine by compartment (prerenal, glomerular, tubular, interstitial, post-renal). ATIN belongs to the interstitial group, and its separation from ATN matters because ATN is usually managed supportively while AIN may justify immunosuppression.

AIN is a common reason to biopsy. In one systematic review, biopsy-proven AIN represented up to 10% to 30% of biopsies performed for acute kidney disease, reflecting how often drugs and infections attack the interstitium and how easily the diagnosis is missed clinically.^[2] The percentage varies by region and by how often renal biopsy is used; where biopsy is rare, AIN is under-recognised. Over 95% of cases result from infection or an allergic drug reaction, so a meticulous drug and infection history is more useful than chasing autoantibodies.^[1]

No dedicated AIN guideline exists. Kidney Disease: Improving Global Outcomes (KDIGO) does not have a guideline restricted to AIN: its acute kidney injury guideline is the 2012 edition, with a 2026 update (AKI and acute kidney disease) in public review at the time of writing, and its glomerular-diseases guideline covers glomerulonephritides rather than tubulointerstitial disease. Management below therefore follows major reviews, professional manuals and the published case-series and trial literature.^[6]

More than 120 drugs have been implicated. The most frequently reported groups come from professional nephrology manuals and are summarised below; the specific agent is less important than the temporal relationship, so the list should be used as a prompt rather than an exclusion test.^[1]

Risk factors worth naming:

• A new or recently changed drug, especially an antibiotic, a proton-pump inhibitor, an NSAID or allopurinol.
• Older age and pre-existing chronic kidney disease, both of which increase drug nephrotoxicity.
• Autoimmune disease and atopy (IgG4-related disease, sarcoidosis, Sjogren syndrome).
• Occupational or environmental exposure to heavy metals (lead, cadmium, mercury) and aristolochic acid-containing herbal products.

Latency is variable and is itself diagnostic. Symptom onset can lag for several weeks after the first exposure, or appear within 3 to 5 days after a second exposure, and extremes range from 1 day with rifampin to 18 months with an NSAID.^[1] This is why stopping the drug is the single most useful therapeutic act and why the whole medication list, including over-the-counter preparations and herbal products, must be reviewed.

The pathogenesis is immunological in most drug-induced cases. The drug or one of its metabolites acts as a hapten on tubular or interstitial proteins, triggering an allergic, predominantly T-cell-mediated interstitial infiltrate with oedema. Electron and immunofluorescence studies rarely show a pathognomonic pattern. In infection-related and systemic forms (sarcoidosis, IgG4 disease, lupus), the mechanism is the same final common pathway of immune-mediated interstitial inflammation.

The interstitial oedema and cellular infiltrate raise intrarenal pressure, reduce peritubular blood flow and disturb tubular transport, so the creatinine rises while tubular defects (polyuria, salt wasting, renal tubular acidosis, Fanconi-like syndromes in severe disease) dominate the picture.

"ATIN should be proven by biopsy before corticosteroids are started" is the conservative advice from professional manuals, because empiric steroids are least justified when infection, myeloma or lymphoma is on the differential.^[1]

On biopsy the earliest finding is interstitial oedema, followed by interstitial infiltration with lymphocytes, plasma cells, eosinophils and a few neutrophils. In severe cases inflammatory cells invade the space between tubular basement membrane cells (tubulitis); non-caseating granulomas occur with beta-lactam antibiotics, sulfonamides, mycobacteria and fungi, and storiform (basket-weave) fibrosis with IgG4-related disease. Glomeruli are usually normal and immunofluorescence or electron microscopy is seldom pathognomonic.^[1] A biopsy should be reported with appropriate scoring for chronicity (tubular atrophy and interstitial fibrosis) because the balance of active inflammation against scarring predicts steroid responsiveness.

Patients are often asymptomatic until renal impairment becomes significant, and the presentation is dominated by the clinical features of acute kidney injury (nausea, anorexia, malaise, reduced urine output or polyuria) plus clues from the systemic allergic response. A careful history includes:

• a new or discontinued drug in the preceding weeks, including supplements and herbal products;
• fever, rash or both (characteristic early features of medication-induced AIN);
• flank or abdominal pain, weight loss;
• recent infection or a systemic autoimmune illness;
• muscle pains or headache explaining long-term analgesic ingestion.

Fever and urticarial rash are characteristic early manifestations, but the classically described triad of fever, rash and eosinophilia is present in less than 10% of patients with medication-induced AIN, so its absence never excludes the diagnosis.^[1] Abdominal pain, weight loss and bilateral renal enlargement can mimic renal cancer or polycystic kidney disease, and peripheral oedema and hypertension are uncommon unless renal failure has occurred.^[1]

• Blood pressure and volume status.
• Skin: urticarial rash, purpura or signs of vasculitis.
• Eyes: uveitis in TINU syndrome (an AIN with associated uveitis, most often in adolescents and young adults).
• General signs of acute kidney injury: uraemic fetor, hypertension if advanced, volume overload.
• Lymphadenopathy or hepatosplenomegaly, which point away from simple drug-induced AIN toward lymphoma, sarcoidosis or infection.

The differential is usually one of compartment, mechanism and trigger. Key separations:

• Acute tubular injury: ischaemia or nephrotoxins, muddy-brown granular casts and tubular epithelial cells rather than white-cell casts; tends to follow hypoperfusion or aminoglycosides.
• Pyelonephritis: fever and flank pain with a positive urine culture and bacteria in the sediment; a urine culture is positive in pyelonephritis and should be negative in drug-induced AIN.
• Glomerulonephritis/nephritic syndrome: haematuria with dysmorphic red cells and red-cell casts, hypertension, nephritic-range abnormalities; consider post-infectious, IgA or ANCA disease.
• Interstitial nephritis from systemic infection or sarcoidosis: look for extrarenal signs and serology.
• Chronic interstitial nephritis (including analgesic nephropathy): insidious, shrunken kidneys, tubular defects; this is the subject of the last section.
• Myeloma kidney and obstruction: both can present as non-glomerular renal failure and are easy to miss.

• Urine dipstick and microscopy. An active sediment with red cells, white cells and white-cell casts together with sterile pyuria (white cells without bacteria on culture) is typical of ATIN. Marked haematuria with dysmorphic red cells is uncommon in isolated ATIN. Eosinophiluria has traditionally been thought to suggest ATIN, but the presence or absence of urinary eosinophils is not diagnostically useful and should not drive decisions.^[1]
• Protein quantification. Proteinuria is usually minimal, conventionally under 1 g per day; the exception is NSAID-induced AIN, in which proteinuria can reach the nephrotic range (3.5 g per day) through combined ATIN-glomerular injury. Nephrotic-range proteinuria should make you think of NSAIDs, ampicillin, rifampin, interferon alfa or ranitidine as the trigger.^[1]
• Renal function: creatinine, electrolytes, bicarbonate, blood urea; expect a non-anion-gap metabolic acidosis and hypokalaemia from tubular defects when tubulitis is marked.
• Full blood count, C-reactive protein and eosinophil count; the eosinophil count (a peripheral eosinophilia) is a supportive but not predictive finding.
• Liver profile and total immunoglobulin G, including IgG4 subsets, and complement (C3, C4) where IgG4-related disease or hypocomplementaemic interstitial nephritis is possible.
• Infection serology targeted to the history (hepatitis B and C, HIV) and blood cultures if infection is plausible.

Renal ultrasonography may show enlarged, echogenic kidneys from interstitial inflammation and oedema in ATIN and is useful to exclude obstruction, stones and chronic scarring. Ultrasonography should precede biopsy for safety. In analgesic nephropathy, non-contrast CT is the diagnostic test of choice (see below).

Renal biopsy is the definitive (reference-standard) diagnostic test, distinguishes ATIN from ATN, identifies the infiltrate and scores fibrosis. A reasonable position, stated in professional manuals, is that AIN should be proven by biopsy before corticosteroids are started, and that empiric steroids are reserved for severe cases where a biopsy is unsafe.^[1] In practice the biopsy also guides prognosis: a diffuse rather than patchy interstitial infiltrate, significant fibrosis, delayed response to prednisone, or acute kidney injury lasting more than 3 weeks all suggest less reversibility.^[1]

There are no numerical diagnostic criteria for AIN in the sense of a score that establishes the diagnosis; the diagnosis is clinical (sterile pyuria with white-cell casts, a rising creatinine, a temporal drug link) and is confirmed histologically. Severity should be described by a combination of:

• rate of rise of creatinine and urine output;
• presence of dangerous complications (hyperkalaemia, pulmonary oedema, severe acidosis);
• degree of tubular dysfunction (concentrating defect, tubular acidosis, Fanconi syndrome);
• biopsy findings: extent of tubulitis, distribution of the infiltrate (diffuse versus patchy), and the amount of interstitial fibrosis and tubular atrophy;
• whether there is nephrotic-range proteinuria, which in NSAID-induced disease predicts a mixed ATIN-glomerular picture and a slower course.

There is no disease-specific guideline for AIN, and the evidence base is dominated by case series and retrospective cohorts. A 2025 systematic review of pharmacotherapy in 1,205 patients across 23 studies found corticosteroids were the main treatment in most series, but with marked variation in dose, route, timing and duration, and no clear consensus regimen; eight of the included studies showed better renal function with steroids and eight did not.^[2] Shared practice is therefore to (1) withdraw the offending agent early, (2) support kidney function, and (3) reserve corticosteroids for patients who fail to recover, or for severe or immune-mediated disease after biopsy. Clinicians should weigh the known harms of glucocorticoids (hyperglycaemia, hypertension, infection, bone loss) against uncertain benefit in each case.^[2]

• Stop the suspected drug as soon as AIN is reasonably suspected, because withdrawal alone often leads to recovery over days to weeks.^[1]
• Avoid further nephrotoxins, including NSAIDs, intravenous contrast and aminoglycosides, and correct hypovolaemia or overload.
• Treat concurrent infections and control blood pressure, glucose and anaemia.
• Provide renal replacement therapy when indicated (dangerous hyperkalaemia, refractory fluid overload, severe acidosis or uraemic complications), and remember that many patients need only transient dialysis.^[1]

Professional manuals cite typical regimens such as prednisone 1 mg per kg orally once daily with gradual taper over 4 to 6 weeks, or an initial methylprednisolone pulse followed by oral prednisolone. Reported series in the 2025 systematic review most often used oral prednisolone 1 mg per kg per day for a period before tapering over several weeks, or 2 to 3 days of intravenous methylprednisolone before stepping down to oral prednisolone; four studies comparing intravenous with oral methylprednisolone found no superiority of either.^[1][2]

Two pragmatic rules are more evidence-based than any fixed dose:

1. Steroids are likely most useful when given early. Professional manuals report that corticosteroids are most effective when given within 2 weeks of stopping the causative medication. Steroid benefit is generally less in NSAID-induced AIN than in other drug-induced forms.^[1]
2. Biopsy first if it is safe. Because ATIN can be produced by infection, myeloma, lymphoma or granulomatous disease, histological proof is preferred before corticosteroid treatment.^[1]

Steroid-resistant or steroid-intolerant cases (an uncommon situation) have been treated in small series with mycophenolate mofetil, cyclophosphamide, azathioprine, cyclosporine or infliximab; mycophenolate is the most studied, but none of these can be routinely recommended on current evidence and should only be used under nephrology supervision after biopsy.^[1][2]

• TINU (tubulointerstitial nephritis with uveitis) is managed with drug withdrawal where a trigger exists, and corticosteroids when renal involvement is significant or vision-threatening; ophthalmology co-management is mandatory.
• Autoimmune and granulomatous causes (sarcoidosis, Sjogren syndrome, IgG4-related disease, anti-tubular basement membrane disease) are treated according to the underlying systemic condition plus corticosteroids; nephrology should direct therapy.

• Indications are the usual ones for acute kidney injury: refractory hyperkalaemia, severe metabolic acidosis, fluid overload with pulmonary oedema, or severe uraemic symptoms. Many patients with AIN need dialysis only briefly.
• Avoid NSAIDs and, during episodes of acute kidney injury, avoid agents that worsen renal perfusion.
• Patients with glomerular overlap (NSAID cases) need the nephrotic-syndrome supportive measures appropriate to their proteinuria.

• Permanent kidney damage from delayed diagnosis or continuation of the offending drug, progressing to chronic kidney disease. ATIN lasting more than 3 weeks, or diffuse fibrosis, predicts incomplete recovery.^[1]
• Hyperkalaemia, severe acidosis and fluid overload, which may require renal replacement therapy; transient dialysis is needed in a substantial proportion of drug-induced cases.
• Nephrotic-range proteinuria in NSAID-, ampicillin-, rifampin-, interferon- and ranitidine-related AIN, with the thrombotic and infection risks that accompany heavy proteinuria.
• Systemic allergic reactions (rash, fever, eosinophilia) and, in severe hypersensitivity, widespread organ involvement.
• Relapse on re-exposure to the same drug; patients need lifelong avoidance and clear documentation of the trigger.

With prompt withdrawal of the trigger, renal function usually recovers within 6 to 8 weeks, although some residual scarring is common and the baseline creatinine may remain above the pre-illness level.^[1] Outcomes are worse when the diagnosis is delayed, when there is diffuse and severe interstitial fibrosis, when drug exposure continues, and when the cause is an NSAID, which is less steroid-responsive than other drug-induced forms.^[1] The recent evidence for corticosteroids is mixed, so expectations should be tempered and anchored to histology rather than to habit.^[2]

• Avoid known triggers in anyone with a prior episode (drug-specific lifelong avoidance), and record the reaction prominently.
• Review complete medication and herbal products lists before starting nephrotoxic combinations, particularly in older adults and those with chronic kidney disease.
• Prefer prevention over treatment for metabolic causes: maintain urate control and hydration to reduce tumour lysis risk, and monitor high-risk regimens with serial creatinine.
• Vaccination and infection-control measures reduce infection-related AIN; for granulomatous causes, treat the systemic disease early.

Pregnancy. AIN in pregnancy is usually drug- or infection-related. Corticosteroids and the trigger drug's safety dominate decision-making; drug withdrawal and support remain first-line, and management should be multidisciplinary. There are no AIN-specific pregnancy guidelines beyond general AKI and CKD practice.

Children and adolescents. Drug-induced AIN occurs (beta-lactams, NSAIDs), but TINU and infection-related forms are more common; nephrotic-range proteinuria with NSAID exposure is well described. Management in children follows the same sequence with paediatric nephrology support.

Older adults. Polypharmacy, proton-pump inhibitors, NSAIDs and allopurinol make drug-induced AIN common; lower renal reserve and comorbidity mean recovery is less complete.

• The diagnosis rests on clinical recognition plus basic tools: careful drug history, urinary microscopy for sterile pyuria and white-cell casts, creatinine and urea, and ultrasound.
• Biopsy may be unavailable; clinical diagnosis and a documented response (or lack of response) to drug withdrawal are reasonable alternatives, with referral where possible.
• Stop the suspected drug early; ensure hydration, potassium control and blood-pressure management; add loop diuretics for overload.
• Reserve corticosteroids for biopsy-proven or strongly suspected immune-mediated disease when infection has been reasonably excluded, and for severe cases; empiric steroids are most defensible where progressive renal failure is present and biopsy is unavailable.
• If dialysis is limited, focus on treating dangerous hyperkalaemia and acidosis, restricting potassium, and using whatever renal replacement is available.

Chronic tubulointerstitial nephritis is the common final response of the kidney to a wide variety of agents: analgesics, lithium, chemotherapy (cisplatin, nitrosoureas), calcineurin inhibitors (ciclosporin, tacrolimus), chronic obstruction, heavy metals and chronic infection. It is usually asymptomatic until picked up on routine testing or during evaluation of chronic kidney disease, and urinalysis shows low-grade proteinuria with sterile pyuria and evidence of tubular dysfunction. Renal biopsy shows variable interstitial infiltration, tubular atrophy and fibrosis. Because it is preventable, and sometimes partly reversible, recognising the exposure matters as much as treating the renal failure.^[1]

Analgesic nephropathy is chronic tubulointerstitial nephritis from cumulative lifetime use of large amounts of certain analgesics. It occurs almost exclusively with prolonged habitual use of analgesic mixtures, traditionally phenacetin with aspirin or paracetamol/acetaminophen, often with caffeine or codeine, taken for many years before kidney injury develops.^[3][4] Since phenacetin was withdrawn, incidence has fallen dramatically, although the safety of prolonged use of modern non-phenacetin mixtures (aspirin, paracetamol/acetaminophen, NSAIDs, COX-2 inhibitors) remains debated; renal papillary necrosis has been reported with prolonged use of non-phenacetin combinations.^[3][4]

Pathogenesis. The mechanism is not fully settled. Renal papillary necrosis is a feature in practically every case and is attributed to medullary ischaemia (from prostaglandin E2 suppression reducing medullary blood flow) combined with direct cellular toxicity from concentrated analgesic metabolites in the medullary interstitium.^[3][4][5]

Epidemiology and prognosis. Historically analgesic nephropathy accounted for a large fraction of end-stage kidney disease in some regions (up to 10% of cases in Australia), with a clear female predominance (around 50% to 80% of cases) and a peak around the early fifties, with the most commonly affected age range from 30 to 70 years.^[3][4] Renal papillary necrosis is characteristic and the risk of renal papillary carcinoma in long-term users is about 8 cases per 100,000 patients per year.^[5] In one older series, about 12% of patients presented in terminal renal failure; with appropriate management roughly 17% improved, 50% remained stable and 23% deteriorated, with a 6-year cumulative survival of about 70%, and malignant hypertension, persistent proteinuria and small initial renal size predicted worse outcomes.^[5] Renal papillary necrosis is considered a consequence of the chronic toxicity of all NSAIDs (medullary ischaemia from prostaglandin suppression plus direct cellular toxicity).^[5]

Presentation. Patients are often asymptomatic and are picked up by unexpected laboratory abnormalities or during evaluation for another condition. Kidney injury is present with usually non-nephrotic proteinuria and a bland urinary sediment or sterile pyuria. Hypertension, anaemia and impaired urinary concentration are common once renal insufficiency develops. Chronic complaints of musculoskeletal pain, headache, malaise and dyspepsia usually relate to long-term analgesic use itself. Late disease brings flank pain, haematuria and passage of a renal papilla causing upper urinary tract obstruction.^[3][4]

Diagnosis. Diagnosis rests on a history of chronic analgesic use plus non-contrast CT. CT signs of analgesic nephropathy are decreased renal size, bumpy contours (at least 3 indentations in the normally convex outline), and papillary calcifications. Plain pyelography is avoided because of contrast nephrotoxicity. The differential includes diabetes, sickle-cell disease, chronic pyelonephritis, obstruction and other causes of papillary necrosis; renal biopsy is the reference standard but is not always appropriate. Urinalysis shows sterile pyuria, haematuria and low-grade proteinuria, and a protein-to-creatinine ratio is the most useful initial quantitative test.^[3][4]

Management. Stopping the causative analgesic is the first-line treatment. Renal function usually stabilizes, although if injury is advanced it may still progress to chronic kidney disease, and established scarring may not recover. Adequate hydration, treatment of infection and blood-pressure control are important; patients should be warned about greater risk of transitional cell carcinoma of the urinary tract and the need for surveillance. Avoidance of the offending agent prevents progression, and patient education and restriction of over-the-counter analgesic promotion are the cornerstone of prevention.^[3][4] Corticosteroids have no role in established analgesic nephropathy, which is not immune-mediated; the disease is not reversed once fibrosis and papillary necrosis are established.^[3]

Nephrotic syndrome; Acute kidney injury; Chronic kidney disease; Glomerulonephritis and nephritic syndrome; IgA nephropathy; Lupus nephritis; ANCA-associated vasculitis; Renal biopsy and urine sediment (plain-text names; not all yet have MedGrasp topics).