Hemolytic uremic syndrome (HUS) is a form of thrombotic microangiopathy (TMA) mainly affecting the kidney and characterized by a triad of microangiopathic hemolytic anemia, thrombocytopenia, and consequent acute kidney injury (AKI) [1]. The International Hemolytic Uremic Syndrome group proposed the 2016 classification of HUS that included several disorders; shiga toxin-induced and pneumococcus-induced HUS, HUS associated with complement dysregulation or mutation of diacylglycerol kinase ε (DGKE), HUS related to cobalamin C defect, and HUS secondary to a heterogeneous group of causes (infections, drugs, cancer, and systemic diseases) [1]. Among them, complement-mediated aHUS is caused by uncontrolled activation of the alternative complement pathway at the endothelial cell surface [2]. Anti-C5 humanized monoclonal IgG antibody—including eculizumab and ravulizumab—recognizes complement protein C5 and blocks the terminal complement cascade [3, 4]. The Kidney Disease: Improving Global Outcomes (KDIGO) Controversies Conference proposed the treatment strategies for aHUS. In the strategies, a complement inhibitor treatment is considered as a first line treatment and is indicated for all patients diagnosed with primary aHUS [5]. However, no reports have compared the pathological kidney findings of a patient with complement-mediated aHUS before and after anti-C5 monoclonal antibody treatment. Here, we report a rare case of complement-mediated aHUS with a complement factor H (CFH) mutation and anti-CFH antibodies who underwent multiple kidney biopsies.

A 53-year-old woman was referred to our hospital by her family doctor after suffering from gastroenteritis for 2 weeks before admission (day 0, symptom onset date), followed by headaches, vomiting, and hypertension 1 week before admission. Her previous health checkup a year prior indicated normal kidney function. However, her father died of unexplained AKI in his thirties. Her height, body weight, and body mass index were 161 cm, 56.3 kg, and 21.6 kg/m2, respectively. Her blood pressure was high (174/85 mmHg), but other vital signs were normal. She had no abnormalities upon physical examination. Laboratory data at admission detected hemolytic anemia with schistocytes, thrombocytopenia, kidney dysfunction, urine abnormalities, and normal coagulation profile including ADAMTS13 activity. We observed that 50% hemolytic complement (CH50) activity was slightly high (57.1 U/mL, normal range 30–53 U/mL). Direct Coombs test, stool culture, pathogenic Escherichia coli immunosera tests, and ADAMTS13 inhibitor tests were negative (Table 1). Laboratory data and imaging detected no background diseases causing secondary TMA.

At her next days of admission, 15 days after the onset of symptoms, her platelet count had decreased to 58,000 /µL, and one session of simple plasma exchange (PE) was performed 16 days after the onset of symptoms (Fig. 1). We calculated the patient’s plasma volume using her body weight and hematocrit value, and fresh frozen plasma equivalent to 1.1 times of her estimated plasma volume was used for PE. After one session of PE, the platelet count had recovered to a level that allowed a kidney biopsy (platelets 119,000/µL); however, severe kidney dysfunction persisted (serum creatinine level 2.89 mg/dL, estimated glomerular filtration rate [eGFR] 14 ml/min/1.73m2), and anemia did not improve (hemoglobin level 9.4 g/dL). Therefore, we performed the first kidney biopsy one month after the onset of symptoms. The kidney biopsy revealed thrombosis in blood vessels, glomerular endocapillary hypercellularity, swelling of endothelial cells, subendothelial exudative lesions in the glomeruli and arterioles, and glomerular basement membrane (GBM) duplication (Fig. 2a and b). Routine immunofluorescence analyses detected slight deposits of immunoglobulin (Ig)-A, IgM, C3c, and fibrinogen along glomerular capillaries, suggesting exudative changes (Supplemental Figure). Electron microscopy revealed subendothelial widening, endothelial cell swelling, and marked endothelial arcade formation, suggesting endothelial cell hypertrophy (Fig. 2c). The eGFR and measured glomerular filtration rate (mGFR) by inulin clearance showed severe kidney dysfunction (14 and 18 mL/min/1.73 m2, respectively). The pathological findings of TMA, the lack of findings of Shiga toxin-producing Escherichia coli infection-mediated HUS, thrombotic thrombocytopenic purpura, or secondary TMA, and the family history of AKI indicated complement-mediated aHUS. After confirming with a positive sheep erythrocyte hemolysis test, we started anti-C5 monoclonal antibody treatment (eculizumab 900 mg/week) 42 days after the onset of symptoms. Laboratory data improved immediately, and her CH50 activity dramatically decreased to below the detection limit (< 4 U/mL), suggesting complete blockage of the complement pathway. She was discharged on day 68. After 4 weeks of initial eculizumab treatment (900 mg/week), the dosage was changed to 1200 mg every 2 weeks. We detected a high titer of anti-CFH antibodies (112.2 AU/mL, normal range + 3SD using healthy control serum was 3.2 + 3.5 AU/mL) using an anti-CFH IgG ELISA kit (#KA1477, Abnova, Taipei City, Taiwan). To reduce her anti-CFH antibodies, glucocorticoids (prednisolone 20 mg/day) were administered from the 73rd day. We performed whole exome sequencing (WES) by next-generation sequencing (NGS) using the Agilent Sure Select V6 + UTR system. The following genes were analyzed: CFH, membrane cofactor protein (MCP), complement factor I (CFI), C3, complement component factor B (CFB), DGKE, thrombomodulin (THBD), and CFH-related protein (CFHR)1–5. This analysis detected two rare variants (minor allele frequency for both variants; MAF < 0.005), a known pathogenic mutation in CFH [c.3572C > T (p. Ser1191 Leu)] that causes complement-mediated aHUS, as well as benign/likely benign variant of complement factor I (CFI) [c.603A > C (p.Arg201Ser)].

Clinical and therapeutic course. Eculizumab treatment improved kidney function, hemolytic anemia, and thrombocytopenia in a patient with aHUS who had a CFH mutation and anti-CFH antibodies. Blue, red, and yellow lines indicate the estimated glomerular filtration rate (eGFR) level, hemoglobin level, and platelet counts, respectively. Black rhombuses indicate the measured GFR (mGFR) level. Abbreviations: PE, plasma exchange; Hb, hemoglobin; PLT, platelet count

Pathological changes observed in kidney biopsies of a patient with aHUS pre- and post-eculizumab treatment. The upper and middle panels show representative light microscopic images of glomeruli and arterioles, respectively. Kidney biopsy specimens were stained with periodic acid-methenamine-silver. Scale bar = 50 µm. The lower panels show representative glomerular findings examined using an electron microscope. Scale bar = 2 μm. a–c First kidney biopsy findings before eculizumab treatment (1 month after symptom onset). Light microscopic analyses detected marked endothelial swelling, endocapillary proliferation, and subendothelial exudative lesions in both glomeruli and arterioles. GBM duplication and thrombosis in arterioles were also observed. Electron microscopy analysis detected subendothelial widening, endothelial cell swelling, and marked endothelial arcade formation. d–f Second kidney biopsy findings 3 months after symptom onset (2 months after eculizumab treatment). Acute injuries in glomeruli and arterioles (endothelial swelling, endocapillary proliferation and thrombosis, and endothelial arcade formation) had almost disappeared, but GBM duplication and subendothelial exudative lesions in arterioles remained. g–i Third kidney biopsy findings 17 months after symptom onset (16 months after the initial eculizumab treatment). GBM duplication was decreased, and glomeruli with minor abnormalities had increased. Subendothelial exudative lesions had almost disappeared in the arterioles, and fibrous intimal thickening had increased

After the introduction of eculizumab treatment, her kidney function began to gradually improved (Fig. 1). Three months after symptom onset, the patient’s eGFR and mGFR reached to 24 and 37 mL/min/1.73 m2, respectively; however, her severe kidney dysfunction persisted. She was concerned that her kidney function might not recover any further and requested a histological evaluation to examine the therapeutic effect of the anti-C5 monoclonal antibody treatment. After fully explaining the risks and benefits of a kidney biopsy and obtaining written informed consent, we performed a second kidney biopsy, which revealed scarce evidence of glomerular and arteriolar endothelial cell swelling, endocapillary hypercellularity, and arcade formation. However, GBM duplication and arteriolar subendothelial exudative lesions remained (Fig. 2 d–f). The proportion of glomerular endocapillary hypercellularity decreased from 57 to 16% and GBM duplication decreased from 81 to 66% (Table 2). Minor glomerular abnormality increased slightly from 14 to 22%. In the arterioles, endothelial proliferation decreased from 37 to 0%, and subendothelial exudative lesions decreased from 37 to 21%. Given these improved pathological findings, glucocorticoid treatment was gradually tapered. We reevaluated the anti-CFH antibodies 5 months after symptom onset and found that her anti-CFH antibody titer did not change despite steroid therapy (104.8 AU/mL). However, her CH50 activity remained below the detection limit, suggesting that her complement pathway was completely blocked by the anti-C5 monoclonal antibody treatment. TMA biological parameters and pathological findings were markedly improved despite the presence of anti-CFH antibodies, and the CFH pathogenic mutation necessitated the continuation of anti-C5 monoclonal antibody treatment. Considering these results, we decided that steroid therapy and other immunosuppressive treatments aimed at reducing the anti-CFH antibodies were not important clinically. Therefore, we tapered steroid treatment further, and continued anti-C5 monoclonal antibody treatment without using other immunosuppressive treatments. Continued eculizumab treatment further improved the patient’s kidney function, and 13 months after symptom onset, her eGFR levels were approximately 34 ± 2 mL/min/1.73 m2. Since tissue repair was not complete at her second kidney biopsy, the patient expected further improvement in kidney function. However, her kidney function was slowly recovering, but did not return to the normal level as she hoped. Therefore, she requested pathological reevaluation again. After informed consent concerning the risks of kidney biopsy, a third kidney biopsy was conducted 17 months after symptom onset. Her eGFR and mGFR were 35 and 53 mL/min/1.73 m2, respectively. The remaining GBM duplication had decreased to 36%, and minor glomerular abnormality increased to 58% (Fig. 2 g, and Table 2). In the arterioles, the remaining subendothelial exudative lesions had almost disappeared, and fibrous intimal thickening was increased (Fig. 2 h and Table 2). We considered that the kidney lesions had reached remission level pathologically, and glucocorticoids were discontinued 18 months after symptom onset. We continuously monitored her CH50 activity throughout her clinical course to help judge whether there was complete blockage of the complement pathway. Her CH50 activity persisted below the detection limit, and we tried to extend the eculizumab treatment with a dosing interval of 1200 mg every 3 weeks after 19 months and then 1200 mg every 4 weeks after 34 months. Despite the extension of the treatment interval, non-detection of CH50 activity remained, and her eGFR levels gradually increased and reached a plateau at 42 ± 2 mL/min/1.73 m2. At 76 months from the onset of symptoms, no recurrence of TMA was observed. We will continue anti-C5 monoclonal antibody treatment to prevent TMA recurrence.