Vanishing Bile Duct Syndrome Caused by a Treatment Regimen Including Avacopan for Microscopic Polyangiitis

Abstract

An 81-year-old man was treated with prednisolone, avacopan, and rituximab for microscopic polyangiitis and sulfamethoxazole/trimethoprim (SMX/TMP) and vonoprazan for prophylaxis. The liver enzyme levels were elevated 42 days after avacopan administration. Avacopan, SMX/TMP, and vonoprazan treatment were discontinued. A liver biopsy revealed vanishing bile duct syndrome. The patient was subsequently treated with ursodeoxycholic acid and glucocorticoid. Although vasculitis remained in remission, the patient ultimately died. Autopsy revealed persistent bile ductopenia and progressive liver cell injury confirmed by cytokeratin 7 positivity, severe cholestasis, and portal fibrosis. Further studies are required to elucidate associated mechanisms and risk factors.

Introduction

Avacopan is a C5a receptor inhibitor recommended for remission induction in recent universal guidelines for anti-neutrophil cytoplasmic antibody-associated vasculitis.

In the ADVOCATE trial (1), avacopan was superior to prednisone tapering in terms of achieving sustained remission at week 52. As for adverse events, the percentage of patients with serious adverse events was 37.3% in the avacopan group and 39.0% in the prednisone group. Nine patients in the avacopan group and six in the prednisone group had serious adverse events involving abnormalities in liver function testing. In subsequent real-world data, the frequency of hepatic-related adverse reactions differed depending on the population. The incidence of hepatic-related adverse reactions was 0% in a Spanish cohort (2), 4.3% in a USA cohort (3), 5.1% in a German cohort (4), and 16.7%, 38.1%, and 40.9% in Japanese cohorts (5-7). Notably, the three Japanese cohorts had a higher frequency than the other cohorts. The mechanism of liver damage, details of the patient's background, and concomitant medications that predispose the patient to liver damage remain unknown.

We herein report an autopsy case of liver injury caused by a treatment regimen that included avacopan for microscopic polyangiitis (MPA).

Case Report

An 81-year-old man presented with a 17-kg weight loss over 5 months, appetite loss, and fatigue. The patient's height, weight, and body mass index were 161.5 cm, 52.9 kg, and 20.3, respectively. He had been diagnosed with interstitial pneumonia 2 years previously when his myeloperoxidase antineutrophil cytoplasmic antibody (MPO-ANCA) level was 31.3 IU/mL without any other symptoms of vasculitis. His recent laboratory findings indicated an estimated glomerular filtration rate (eGFR) of 52 mL/min/1.73 m², proteinuria of 0.488 g/gCr, hematuria with red blood cell casts, C-reactive protein level of 7.27 mg/dL, and MPO-ANCA of 723 IU/mL. A kidney biopsy revealed pauci-immune crescentic glomerulonephritis. Based on the 2022 American College of Rheumatology/European League Against Rheumatism classification criteria, the patient was diagnosed with MPA. The Birmingham Vasculitis Activity Score (BVAS) was 12. The patient had no history of viral hepatitis or other liver disease.

The patient was treated with prednisolone (PSL; 20 mg/day), avacopan (60 mg/day), and 4 cycles of rituximab (375 mg/m²) as induction therapy. Sulfamethoxazole/trimethoprim (SMX/TMP; 200/40 mg) and vonoprazan (20 mg/day) were administered to prevent Pneumocystis pneumonia and peptic ulcers, respectively. The patient was discharged on day 37 when the PSL dose was tapered to 5 mg/day and the BVAS score was 0.

Liver function test results at the first outpatient visit following discharge, 25 days after avacopan initiation, were within the normal range (aspartate aminotransferase, 23 U/L; alanine aminotransferase, 17 U/L; gamma-glutamyl transpeptidase, 28 U/L; alkaline phosphatase, 64 U/L; total bilirubin 0.3 mg/dL). However, the patient was readmitted 42 days after avacopan initiation owing to liver damage (aspartate aminotransferase, 220 U/L; alanine aminotransferase, 274 U/L; gamma-glutamyl transpeptidase, 596 U/L; alkaline phosphatase, 402 U/L; total bilirubin 3.1 mg/dL), indicating a marked change within the preceding 17 days.

Tazobactam/piperacillin (TAZ/PIPC) was initiated for suspected biliary tract infection, and avacopan was discontinued on the first day of readmission. Endoscopic retrograde cholangiopancreatography and computed tomography did not reveal biliary obstruction or gallstones, but diffuse wall thickening of the bile duct and gallbladder was observed. A liver biopsy showed mixed cellular infiltration limited to the portal area and ductopenia in ＞50% of the portal area (Fig. 1A, B). Consequently, the patient was diagnosed with vanishing bile duct syndrome (VBDS).

Figure 1.

Histology. (A, C) Hematoxylin and Eosin staining. (B, D) Cytokeratin 7 (CK7) staining. (A, B) A liver biopsy performed on day 8 revealed mixed cellular infiltration limited in the portal area and ductopenia in more than 50% of the portal tracts. (C, D) Autopsy revealed persistent bile ductopenia, liver cell injury showing CK 7 positivity on immunohistochemistry, severe cholestasis, and portal fibrosis.

The Roussel-Uclaf Causality Assessment Method score (8) was 7 points for avacopan, SMT/TMP, and rituximab and 6 points for vonoprazan. The drug-induced lymphocyte stimulation tests for avacopan, SMX/TMP, and vonoprazan were negative. On day 11 of the second admission, SMX/TMP and vonoprazan treatment were discontinued. The patient was treated with ursodeoxycholic acid (600 mg/day) and methylprednisolone (1,000 mg/day) for 3 days, followed by 60 mg/day of mPSL. Vancomycin and ampicillin for catheter-related bloodstream infections, ganciclovir for cytomegalovirus reactivation, and TAZ/PIPC and azactam for bacteremia caused by metallo-β-lactamase-producing Klebsiella oxytoca were administered during treatment (Fig. 2).

Figure 2.

Clinical course during the second hospitalization. The date of readmission was marked as day 1. Prednisolone was administered orally until day 11, whereas methylprednisolone was administered intravenously after day 12. PSL: Prednisolone, mPSL: methylprednisolone, TAZ/PIPC: tazobactam/piperacillin, VCM: vancomycin, ABPC: ampicillin, AZT: azactam, SMX: sulfamethoxazole, TMP: trimethoprim, UDCA: ursodeoxycholic acid, GCV: ganciclovir, AST: aspartate aminotransferase, ALT: alanine aminotransferase, GGT: gamma glutamyl transferase, ALP: alkaline phosphatase, T-Bil: total bilirubin

Although vasculitis remained in remission, the patient succumbed to septic shock on day 78 of second admission. Autopsy revealed persistent bile ductopenia and progressive liver cell injury, confirmed by cytokeratin 7 positivity, severe cholestasis, and portal fibrosis (Fig. 1C, D).

Discussion

We herein report an autopsy case of a patient with MPA whose ductopenia persisted and whose liver cell injury, cholestasis, and portal fibrosis progressed despite discontinuation of the culprit drugs avacopan, SMX/TMP, and vonoprazan.

The incidence of liver injury in avacopan-treated cohorts is shown in Table 1 (2-7). The age, sex, ANCA type, eGFR, and immunosuppressive therapy details in the table reflect the characteristics of the overall cohort. The three Japanese cohorts (5-7) showed a higher incidence of liver injury and a higher average age than those of the Spanish (2), USA (3), and German (4) cohorts. Frequency by sex varied among the three Japanese cohorts. The eGFR values in the Japanese patients were not lower than those reported in other studies. No clear differences were identified between Japan and other countries regarding the use of immunosuppressive and molecular-targeted agents. Based on the currently available literature, the elevated incidence of avacopan-related liver injury in Japan may be attributed to racial or age-related factors.

Table 1.

Frequency of Liver Injury Induced by Avacopan.

Reference	Country	Number of Facilities	n	Liver Injury (%)	Age (years)	Male (%)	ANCA Type (%)	eGFR (mL/min/1.73 m2)	Immunosuppressants/Molecular-targeted drugs (%)
(2)	Spain	10	29	0	56.0 (46.5-67.5)	41.4	MPO; 72.4 PR3; 27.6	28 (13.5-38.5)	Induction. RTX 48.2, CY 6.8 RTX+CY 41.3 Maintenance. RTX 82.7, AZA 6.89, MMF 3.4
(3)	USA	12	92	4.3	59±17	36.0	MPO; 72 PR3; 27 Negative;1	32.9±26.7	RTX 48, CY 2 RTX+CY; 47, MTX; 1
(4)	German	5	39	5.1	64 (51-72)	54.0	MPO; 43.6 PR3; 56.4	37±36	Induction. RTX 62, CY 38 Maintenance. RTX 78, MMF 11, AZA 7, MTX 4
(5)	Japan	3	36	16.7	74.5 (68.5-79.5)a 76.5 (65.3-82.0)b	38.9	MPO; 83.3 PR3; 8.3 No data; 8.3	57.94 (42.45-69.56)a 55.64 (20.62-75.55)b	Induction. RTX 63.9, CYC 2.8, CYC+RTX 5.6 Maintenance. AZA 16.7, RTX 5.6, RTX+AZA 2.8
(7)	Japan	1	21	38.1	77 (66-81)	52.4	MPO; 85.7 PR3; 9.5 Negative; 4.8	43 (35-51)	Induction. RTX 76.2, AZA 14.3 Maintenance. RTX 71.4, AZA 14.3, MIZ 4.8
(6)	Japan	2	22	40.9	68.0±11.1	37.4	MPO; 68.2 PR3; 31.8	Not mentioned	Concomitant at the onset of liver injury. RTX 31.8, CY 13.6, AZA 22.7, MTX 9.1

ANCA: anti-neutrophil cytoplasmic antibody, eGFR: estimated glomerular filtration rate, MPO: myeloperoxidase, PR3: proteinase 3, RTX: rituximab, CY: cyclophosphamide, AZA: azathioprine, UDCA: ursodeoxycholic acid, MMF: mycophenolate mofetil, MTX: methotrexate, MIZ: mizoribine

^aLiver dysfunction

^bNo liver dysfunction

Another possibility is a potential association with CYP3A4 expression. CYP3A4 is the major enzyme responsible for the clearance of avacopan and the formation and clearance of the major circulating metabolite M1, a monohydroxylated product of avacopan. In vitro CYP inhibition studies have demonstrated that avacopan and its major metabolite M1 are time-dependent CYP3A4 inhibitors (9). The major metabolizing enzyme of vonoprazan is CYP3A4. The concomitant use of avacopan and vonoprazan might have contributed to hepatotoxicity in this case by slowing the metabolism of these two drugs. Japanese individuals have a high frequency of CYP3A53*3, a mutation that causes 85% loss of CYP3A5 activity (10); CYP3A4 may compensate for metabolism in such cases (11). This background may also affect the metabolism of drugs by CYP3A4 and may be a reason for the high frequency of liver injury from avacopan in Japanese patients.

Several cases of VBDS have been reported (12-14). In C5aR1-deficient mice, disease was aggravated during resolution of intoxication, as evidenced by increased liver necrosis and serum alanine aminotransferase. Furthermore, decreased compensatory hepatocyte proliferation and increased caspase-3 activation in the surrounding necrotic cores were detectable in C5aR1-deficient mice (15). Liver regeneration is impaired after partial hepatectomy in C5a-deficient mice (16) Avacopan may delay the recovery from liver injury caused by SMT/TM, or the overlapping use of drugs capable of inducing VBDS may contribute to the severity of the condition.

Table 2 summarizes the previously reported cases of avacopan-associated liver injury (5-7,17-23). The onset of liver injury after avacopan administration ranged from 25 to 189 days. All types of liver injury patterns, including hepatocellular, cholestatic, and mixed, have been reported. VBDS was confirmed by a liver biopsy in three patients (6,20,22), in addition to the present case, and the case reported by Mori et al. also resulted in death (6). Given the potentially fatal nature of avacopan-induced liver injury, regular monitoring of the liver function is essential, particularly during the first six months after starting avacopan. The details of concomitant medications other than immunosuppressive therapy have not been fully disclosed in some reports, highlighting the need for further investigations.

Table 2.

Details of Liver Injury Induced by Avacopan.

Reference	n	Age (years)	Sex	ANCA Type	Induction therapy	Details of liver injury	Time to onset after avacopan initiation	Outcome
Our case	1	81	M	MPO	GC, RTX	VBDS (liver biopsy)	42 days	Death
(23)	1	76	F	MPO	GC, RTX	AST 395 U/L, ALT 624 U/L, ALP 686 U/L, T.Bil 7.5 mg/dL	4 weeks	Improved
(18)	1	10	M	PR3	GC, mPSLpulse, RTX	AST 723 U/L, ALT 1145 U/L, γ-GTP 616 U/L, T.Bil 0.56 mg/dL	31 days	Improved
(19)	1	51	F	PR3	GC, RTX	AST 111 U/L, ALT 477 U/L, T.bil 3.7 mg/dL Liver biopsy; Acute cholestasis	34 days	Prolonged elevation of bilirubin
(20)	1	70	F	MPO	GC, RTX	AST 155 U/L, ALT 357 U/L, ALP 255 U/L, T.Bil 3.6 mg/dL Liver biopsy; VBDS	Approximately 40 days	Prolonged elevation of bilirubin
(22)	1	75	F	MPO	GC, mPSLpulse, RTX	AST 218 U/L, ALT 565 U/L, γ-GTP 654 U/L, T.bil 7.7 mg/dL Liver biopsy; VBDS	45 days	Prolonged elevation of bilirubin
(17)	1	84	F	MPO	GC, RTX	AST 273 U/L, ALT 490 U/L, ALP 325 U /L, T.bil 7.5 mg/dL→Liver biopsy of 66 days later; acute hepatitis like DILI	60 days	Prolonged elevation of liver enzyme
(21)	1	75	M	MPO	GC, mPSLpulse, RTX	AST 203 U/L, ALT 236 U/L, ALP 695 U/L, T.Bil 1.9 mg/dL	9 weeks	Recovered
(7)	8	79 (74-82)	M; 44.4%	MPO; 80% PR3; 10%	GC 100%, mPSLpulse 22.2%, RTX 66.7%, AZA 22.2%	AST or ALT levels exceeding three times the upper limit of normal, without concurrent elevation in bilirubin levels.	Median 1.7 months (IQR 1.5-1.9 months)	Recovered
(6)	9	73.9±6.6	M; 44.4%	MPO	GC 50%, RTX 66.7%	Hepatocellular n=5 Mixed n=3 Cholestatic n=1 Liver biopsy; VBDS n=1, Cholestasis n=1	Median 54 days (range 25-189 days)	Recovered n=8 Death n=1
(5)	6	74.5 (68.5-79.5)	F	MPO	GC 50%, RTX 66.7%	Elevated bilirubin in 2 patients (14.1 mg/dL, 14.9 mg/dL)	Median 58 days (range 31-107 days)	Improved

AST: aspartate aminotransferase, ALT: alanine aminotransferase, GGT: gamma-glutamyl transferase, ALP: alkaline phosphatase, T.Bil: total bilirubin, ANCA: anti-neutrophil cytoplasmic antibody, MPO: myeloperoxidase, PR3: proteinase 3, GC: glucocorticoid, RTX: rituximab, mPSL: methyl prednisolone, CY: cyclophosphamide, AZA: azathioprine, VBDS: vanishing bile duct syndrome

¹The values in the table represent nine cases, including the case of discontinuation due to diarrhea in addition to liver injury described in the paper.

In conclusion, we encountered a case of VBDS caused by a treatment regimen that included avacopan. Two liver biopsies (one at the onset and one at autopsy) revealed the progression of severe liver injury. Further studies are needed to elucidate the underlying mechanisms and risk factors for avacopan-related liver injury.

Author’s disclosure of potential Conflicts of Interest (COI).

N.T. has an advisory agreement with Kissei Pharmaceutical, received lecture fees from Takeda Pharmaceutical, and received a research grant from Zenyaku Kogyo.