Background Kidneys are major organs targeted by antineutrophilic cytoplasmic antibody (ANCA)-associated vasculitis (AAV). Clinical manifestations, laboratory data, and prognosis of AAV with renal involvement at diagnosis are not elucidated.

Objectives We compared clinical features of AAV with renal involvement with patients without renal involvement.

Methods We conducted an observational study of 104 patients with AAV (12 eosinophilic granulomatosis with polyangiitis, 23 granulomatosis with polyangiitis (GPA), 66 microscopic polyangiitis, 3 renal limited vasculitis) between in 2008 to 2016 in Nagasaki University Hospital. Using medical records, we analyzed the patients' baseline variables, laboratory data, clinical symptoms, and therapeutic outcomes after treatments including episodes of relapses, initiations of dialysis, and death. Renal involvement was defined as the state with estimated glomerular filtration rate <60 mL/min/1.73 m² or microscopic hematuria (2+ or greater) which were not caused by renal diseases except for AAV.

Results Sixty-nine patients had renal involvement. Patients with renal involvement group had higher median age at diagnosis than patients without renal involvement group (75 years vs. 66 years, p<0.001). Patients with renal involvement included fewer GPA patients compared to other AAV types. Patients with renal involvement had lower hemoglobin levels (10.3 g/dL vs. 12.3 g/dL) and lower platelet levels (23.7x10⁴/μL vs. 28.7x10⁴/μL). Patients with renal involvement had higher erythrocyte sedimentation rate (78mm/h vs. 20mm/h), MPO-ANCA titers (116 U/mL vs. 58 U/mL) and urine protein levels (0.81 g/gCr vs. 0.15 g/gCr). Patients with renal involvement had lower C3 levels, but CH50 and C4 levels did not differ between in two groups. There were no differences in treatments including doses of prednisolone and use of methylprednisolone pulse and cyclophosphamide between in two groups. Multivariable regression analysis revealed that age at diagnosis is the most significant explanatory variable to renal involvement. Nineteen percent of patients with renal involvement had initiations of dialysis. Multivariable analysis demonstrated estimated glomerular filtration rate at diagnosis is the most significant explanatory variable to initiations of dialysis (p=0.010). Receiver operating characteristic curve showed the cutoff level of estimated glomerular filtration rate to distinguish initiations of dialysis was 37mL/min/1.73 m² (sensitivity=79%, specificity=70%, area under the curve=0.80). Assessed by a log-rank test, overall survival rate did not differ between in two groups (p=0.29).

Conclusions Patients with renal involvement had higher age at diagnosis. Patients with renal involvement included fewer GPA patients. Patients with renal involvement had lower C3 levels. Estimated glomerular filtration rate at diagnosis is the most significant explanatory variable to initiations of dialysis.

Disclosure of Interest None declared