Dear Editor, Diabetes mellitus is one of the leading causes of chronic kidney disease and end-stage kidney disease worldwide. In clinical practice, diabetic kidney disease is usually diagnosed based on the duration of diabetes, albuminuria, associated microvascular complications, and the trajectory of kidney function. However, kidney disease in a patient with diabetes is not necessarily attributable to diabetes, and an atypical clinical course should prompt consideration of an alternative diagnosis. We report a patient initially considered to have diabetic kidney disease in whom kidney biopsy demonstrated pauci-immune crescentic glomerulonephritis (PIGN). A 36-year-old man with a 22-year history of type 2 diabetes mellitus, 9 years of hypertension, and approximately 1 year of chronic kidney disease had been followed at another center. His known baseline serum creatinine was 1.2 mg/dL, and diabetic retinopathy was present. He was hospitalized elsewhere because of acute kidney injury and was referred to our center for consideration of kidney biopsy after 24-hour urinary protein excretion of 3 g/day was detected. At presentation, bilateral pretibial edema, proteinuria, and acute kidney injury were present. The admission serum creatinine was 8.44 mg/dL, blood urea nitrogen was 56 mg/dL, and estimated glomerular filtration rate was 8 mL/min/1.73 m² . Renal replacement therapy was initiated because of uremic symptoms. The patient had undergone metabolic surgery in 2021, and HbA1c during hospitalization was 5.6%. On follow-up, 24-hour urinary protein excretion was 5.8 g/day, albumin excretion was 3.1 g/day, and creatinine clearance was 2.35 mL/min. Serologic testing for anti-PLA2R, antinuclear antibodies, Antı MPO and PR3 ANCA, anti-GBM antibodies, anti-dsDNA, and ENA was negative. C3, C4, and immunoglobulin levels were normal, and serum protein electrophoresis showed no monoclonal band. Urinary RBC count was elevated at 109; however, the sediment remained inactive, with no RBC casts. Although the long-standing diabetes, diabetic retinopathy, proteinuria, and clinical background initially supported diabetic kidney disease, the abrupt and severe decline in kidney function was atypical. Kidney biopsy was therefore performed to clarify the cause of acute kidney injury. Nineteen glomeruli were evaluated. 10 glomeruli are globally sclerotic. Out of the 9 non-sclerotic glomeruli, 7 show a cellular-fibrocellular crescentic morphology. Fibrinoid necrosis is present in 2 glomeruli. Segmental increase in basement membrane rigidity, mild mesangial matrix and cellular increase, and occasional polymorphonuclear leukocyte infiltration are observed in 2 glomeruli. The interstitium shows increased fibrosis involving 30% of the examined cortical area, along with patchy mononuclear inflammatory cell infiltration. The represented vascular structures show subintimal fibrosis that disproportionately narrows the lumen for the patient's age, fibrinoid necrosis of the vessel wall, and adventitial arteritis. No immunofluorescence staining was detected in renal biopsy specimen (Figures A and B). Electron microscopy could not be performed. The findings were considered consistent with pauci-immune crescentic glomerulonephritis Following informed consent, the patient received intravenous pulse methylprednisolone therapy and a single dose of cyclophosphamide. However, renal function did not recover sufficiently to discontinue renal replacement therapy, and the patient remained dialysis-dependent during approximately two months of follow-up. In view of the persistent dialysis dependence and lack of renal recovery, corticosteroid therapy was gradually tapered and discontinued, and no further cyclophosphamide was administered. This case illustrates the diagnostic risk of attributing kidney dysfunction in a patient with long-standing diabetes solely to diabetic kidney disease. The presence of diabetic retinopathy and substantial proteinuria made diabetic kidney disease clinically plausible; however, the rapid loss of kidney function, very low creatinine clearance, oligoanuria, and requirement for renal replacement therapy were not compatible with the expected slowly progressive course. Kidney biopsy established a different diagnosis and directly changed the treatment strategy. Pauci-immune crescentic glomerulonephritis is an important cause of rapidly progressive glomerulonephritis and may lead to irreversible kidney failure when diagnosis is delayed. Notably, ANCA was negative in our patient. Thus, negative serologic testing does not exclude pauci-immune crescentic glomerulonephritis and should not preclude kidney biopsy when the clinical course is atypical or rapidly progressive. In patients with diabetes, a rapid increase in serum creatinine, oligoanuria, nephrotic-range proteinuria, or another unexpected clinical trajectory should raise suspicion for non-diabetic kidney disease. This case emphasizes the central role of kidney biopsy in establishing the correct diagnosis and guiding disease-specific treatment in such patients.