Acute Tubulointerstitial Nephritis, Diagnosis, and Pathology
Author: Andrew Bland, MD, MBA, MS
The short answer to the question of what immune checkpoint inhibitor–associated acute kidney injury (ICI-AKI) represents pathologically is this: it is acute tubulointerstitial nephritis. Not acute tubular necrosis. Not a hemodynamic prerenal process. Not glomerulonephritis in most cases. The dominant lesion — identified in 70–93% of biopsied patients across multiple series — is acute tubulointerstitial nephritis (ATIN), driven by disinhibited cytotoxic T cells infiltrating the renal interstitium and attacking tubular epithelium (1,2,4).
Immune checkpoint inhibitors — monoclonal antibodies targeting CTLA-4, PD-1, or PD-L1 — remove critical inhibitory brakes on T cell activation, enabling durable antitumor immunity. In the kidney, the same T cells unleashed to destroy tumor cells can turn against renal tubular epithelium, producing an inflammatory interstitial nephritis that is immune-mediated, steroid-responsive, and distinguishable from other causes of AKI (1,5).
The incidence of ICI-AKI is 2–5% with monotherapy and rises to approximately 10% with combination anti-CTLA-4 plus anti-PD-1 therapy (2,5). Recognition of ICI-ATIN — its timeline, its urinary signature, its histologic appearance — is now a core nephrology competency.
Under normal circumstances, CTLA-4 and PD-1 serve as coinhibitory receptors that constrain T cell activation and maintain peripheral tolerance. ICIs block these receptors, resulting in enhanced T cell activation and reduced regulatory T cell suppression. In susceptible patients, this immune amplification leads to loss of self-tolerance against renal tubular antigens, generating a CD8+ cytotoxic T cell infiltrate within the kidney interstitium (1,5).
Two major pathways are operative:
First, ICIs may reactivate autoreactive T cells that were previously held in check by checkpoint pathways — cells that recognize renal tubular cell antigens (possibly shared with tumor antigens) and are now capable of mounting a nephritogenic attack.
Second, in patients taking concurrent medications such as proton pump inhibitors or NSAIDs, those drugs may serve as haptens that prime a drug-specific T cell response, which ICI therapy then amplifies beyond the threshold of self-tolerance (1,2). The high frequency of PPI use in cancer patients — and the independent association of PPI use with ICI-ATIN — lends clinical support to this second mechanism (2,6).
The interstitial infiltrate is composed predominantly of CD4+ helper and CD8+ cytotoxic T cells, with macrophages and plasma cells contributing in variable proportions. The relative paucity of eosinophils — compared to classic drug-induced AIN — reflects the T-cell–driven, rather than hypersensitivity-driven, nature of the immune response.
ICI-ATIN develops at a median of 12–16 weeks after ICI initiation, with a broad range extending from two days to 18 or more months (1,2). The peak incidence occurs during weeks 8 through 20.
| Risk Factor | Details |
|---|---|
| Combination ICI therapy | OR ~3 vs monotherapy; strongest independent risk factor (2) |
| Concurrent PPI use | ~2-fold increased risk; robust across multiple series (2,6) |
| Pre-existing CKD | Reduced GFR reserve; smaller interstitial inflammation produces larger proportional Cr rise |
| Age >60 years | General substrate of immune dysregulation |
| Autoimmune disease history | Increased susceptibility to irAE |
| Prior extrarenal irAE | Systemic predisposition to off-target immune toxicity |
| Melanoma / RCC | Somewhat higher rates than NSCLC |
Most commonly asymptomatic or minimally symptomatic, discovered incidentally on routine laboratory monitoring. The patient typically presents without urinary complaints — no dysuria, no gross hematuria, no flank pain, no urgency — and with only vague constitutional symptoms.
Serum creatinine is elevated from baseline, with median elevations of approximately 1.8-fold in large series (2). Urine output is typically preserved; oliguria is unusual and should prompt consideration of alternative or concurrent diagnoses such as obstructive nephropathy, severe ATN, or hemodynamic compromise.
On targeted history, the clinician will identify ICI initiation 2–5 months prior, frequently in combination with a PPI. Extrarenal immune-related adverse events — thyroiditis, pneumonitis, colitis, hepatitis, dermatitis — may be concurrently present or may have preceded the renal presentation (1).
Sterile pyuria (>5 WBC/hpf in the absence of bacteriuria) is present in 60–80% of affected patients (1,3,7). WBC casts are the most specific urinary finding for ATIN and carry the highest diagnostic weight. They form when leukocytes become embedded within a protein matrix within the tubular lumen. Their presence confirms tubulointerstitial origin of the pyuria. Absence does not exclude ICI-ATIN; WBC casts may be absent in up to 40% of confirmed cases (1,3).
Eosinophiluria (>1% on Hansel stain) is less reliably present in ICI-ATIN than in classic drug-induced AIN. When prominent (>5 eosinophils/hpf), it should raise suspicion for concomitant PPI-induced or antibiotic-induced AIN rather than pure ICI-ATIN (1,6).
Characteristically subnephrotic (<2 g/g creatinine). The Cortazar single-center series reported a median UPCR of 0.48 g/g (3). The proteinuria is a mixture of low-molecular-weight tubular proteins (beta-2 microglobulin, retinol-binding protein) and albumin.
Typically absent or minimal. Prominent hematuria (>10–15 RBCs/hpf), dysmorphic red cells, or RBC casts suggest glomerular pathology. RBC casts are NOT a feature of ATIN and should not be attributed to this diagnosis (1,4).
FENa is typically >2%, reflecting inappropriate sodium wasting from inflamed tubules (contrasting with prerenal azotemia where FENa <1%). Urine osmolality <300 mOsm/kg is consistent with tubular concentrating defect. Elevated urine beta-2 microglobulin provides additional evidence of proximal tubular dysfunction (1,5).
Most common cause of AKI in hospitalized patients. Volume depletion produces concentrated urine (Uosm >600) with low FENa (<1%). Should resolve with volume replacement over 24–48 hours; failure to improve strengthens the case for intrinsic renal pathology.
From concurrent platinum-based chemotherapy, aminoglycosides, or iodinated contrast. Produces granular or pigmented waxy casts — not WBC casts — and presents within days of nephrotoxic exposure rather than weeks to months after ICI initiation.
Must be excluded with renal ultrasound in any patient with AKI and pelvic or retroperitoneal malignancy. May present with non-dilated collecting systems in early or partial obstruction.
May produce pyuria and WBC casts indistinguishable from ATIN on urinalysis alone. Positive urine culture (>100,000 CFU/mL) establishes this diagnosis. Sterile pyuria strongly favors ATIN over infection.
Suggested by heavy proteinuria (>2 g/day), prominent hematuria with dysmorphic red cells or RBC casts, or hypoalbuminemia. Cortazar 2020 identified concurrent glomerular pathology in 27% of biopsied patients — including pauci-immune crescentic GN, MCD, membranous, and IgA nephropathy (2).
The Cortazar/MD Anderson approach advocates for early biopsy in all stage 2 or higher AKI before high-dose steroid initiation (2,4). The Brigham approach recommends proceeding empirically with corticosteroids in classic ATIN presentations, reserving biopsy for atypical features (1,6). The SITC 2021 guideline supports biopsy when the result will change management (1).
Biopsy is most strongly indicated when: the diagnosis is genuinely uncertain, GN cannot be excluded, AKI is severe enough to require aggressive immunosuppression, or the patient has failed 2 weeks of appropriate steroid therapy.
Dense inflammatory infiltrate within the renal interstitium, expanding the interstitial space between tubules. The cellular composition is predominantly mononuclear — lymphocytes predominate. Key findings:
| Grade | Interstitial Involvement | Tubular Architecture | Prognosis |
|---|---|---|---|
| Mild | <10% | Preserved | Excellent recovery |
| Moderate | 10–50% | Focal necrosis | Good recovery expected |
| Severe | >50% | Extensive necrosis, early fibrosis | Incomplete recovery likely |
Characteristically sparse or negative. Immunoglobulin deposits (IgG, IgM, IgA) and complement (C3, C1q) are absent or trace. No linear or granular immune complex deposits. This is consistent with T-cell–mediated pathogenesis (1,4,5).
The absence of linear IgG along the TBM excludes anti-TBM nephritis. Absence of granular mesangial IgA excludes IgA nephropathy. In concurrent membranous nephropathy: granular IgG + C3 subepithelially, PLA2R-negative. In lupus-like nephritis: full-house pattern (IgG, IgM, IgA, C3, C1q).
In pure ICI-ATIN: no electron-dense deposits; interstitial edema and lymphocytes closely apposed to tubular cells. Normal glomerular capillary loops with intact GBM.
When concurrent GN present, EM becomes diagnostically essential: foot process effacement (MCD/FSGS), subepithelial deposits (membranous), subendothelial deposits (lupus-like), mesangial deposits (IgA), endothelial swelling with platelet-fibrin thrombi (TMA).
The Cortazar 2020 multicenter study (138 patients, 93 biopsied) established the pathologic distribution (2):
| Pattern | Frequency | Key Features |
|---|---|---|
| Pure ATIN | 71% | Interstitial T-cell infiltrate, tubulitis, normal glomeruli |
| ATIN + glomerular disease | 19% | Concurrent GN requiring escalated immunosuppression |
| ATIN + predominant ATI | 10% | Mixed pattern with acute tubular injury |
Glomerular disease patterns associated with ICI include:
PPIs are well-established causes of AIN, producing a histologic pattern indistinguishable from ICI-ATIN on kidney biopsy. Both show dense lymphocytic interstitial infiltration, tubular injury, sterile pyuria, and subnephrotic proteinuria (1,6).
Several noninvasive biomarkers are under active investigation:
| Biomarker | Significance | Status |
|---|---|---|
| Urinary KIM-1 | Proximal tubular cell injury marker; may precede Cr rise | Research stage |
| Urinary IL-6 and TNF-α | Elevated in ICI-ATIN; correlate with histologic severity | Research stage |
| Serum cystatin C | Alternative GFR marker; may detect subclinical decline before Cr | Clinically available |
| Urine beta-2 microglobulin | Proximal tubular dysfunction marker | Limited clinical use |
The gut microbiome has emerged as an area of inquiry: reductions in short-chain fatty acid–producing Bacteroidetes have been observed in patients developing ICI-related irAEs. PPI use itself alters the gut microbiome, potentially explaining part of the PPI-associated ICI-ATIN risk through an indirect immunologic mechanism (5,6).
Three convergent clinical criteria, when simultaneously present, establish a high-probability diagnosis (>90%) without biopsy:
| Criterion | Definition |
|---|---|
| Temporal | Acute Cr elevation 6–20 weeks after ICI initiation, no alternative explanation in 24–48 hours |
| Urinary | Sterile pyuria ± WBC casts, subnephrotic proteinuria <2 g/day |
| Functional | FENa >2% or urine osmolality <300–400 mOsm/kg |
When all three criteria are met, empiric corticosteroid therapy is justified without waiting for biopsy, particularly in stage 2 AKI (1,2,6).
The essential foundation for current understanding of ICI-AKI (2). 138 patients across 12 U.S. centers; 276 controls; 93 biopsied.
| Parameter | Result |
|---|---|
| Median time ICI → AKI | 14 weeks (IQR 6–47) |
| Combination ICI therapy OR | 3.2 |
| PPI use OR | 2.1 |
| Pure ATIN on biopsy | 71% |
| ATIN + glomerular disease | 19% |
| Steroid response (complete + partial) | 87% (44% + 43%) |
| Time to Cr plateau | 14 days |
| Time to Cr ≤1.2x baseline | 28 days |
| Rechallenge success | 77% (23/30) without recurrent AKI |
Immune checkpoint inhibitor–associated acute kidney injury is, in its dominant pathologic form, acute tubulointerstitial nephritis. It is not a hemodynamic problem. It is not a tubulotoxic process in the conventional sense. It responds to immunosuppression, not to fluids, not to nephrotoxin removal, and not to dialysis except as a temporizing bridge in severe cases.
Clinical recognition requires attention to the timeline (weeks 6–20), the urinary signature (sterile pyuria, WBC casts, subnephrotic proteinuria, no RBC casts), and the exclusion of competing diagnoses. When the presentation is atypical, when glomerular features are present, or when the steroid response is absent, kidney biopsy provides the definitive characterization.
The histologic diagnosis — ATIN with dense T-cell infiltration, negative immunofluorescence, and glomerular sparing — is both confirmatory and therapeutic, providing the foundation for a structured corticosteroid course and, ultimately, decisions about ICI rechallenge.
Andrew Bland, MD, MBA, MS | Clinical Mastery Series | Urine Nephrology Now
© 2026. For educational use only.