CD55 Deficiency With Budd-Chiari Syndrome Treated by Liver Transplantation and Eculizumab Free

We report on a 25-year-old male patient born to consanguineous healthy parents (Fig 1, Supplemental Fig 4). Since his first year of life, he had suffered from protein-losing enteropathy, chronic diarrhea, episodes of recurrent abdominal pain, and severe hypoproteinemia (albumin <3.0 g/dL [normal range 3.5–5 g/dL], total protein <55 g/l [57–80 g/l], immunoglobulin G [IgG] <4.0g/l [4.7–15g/l]). He had received weekly albumin infusions and had required hospitalization 2 to 5 times per year. Endoscopy showed discontinuous inflammation, erosions, mucosal hyperplasia, and lymphangiectasia, predominantly in the small intestine. Histology findings were nondiagnostic but were interpreted as “Crohn’s-like disease.” Recurrent jejunal stenosis required bowel resections at the age of 2, 5, and 9 years. Fecal calprotectin was elevated. Treatment with azathioprine and aminosalicylate was ineffective but the patient experienced improvement in his symptoms after long-term glucocorticoid therapy. The adverse side effects of therapy were growth retardation and osteopenia.

At the age of 8 years, he developed thrombosis of the vena cava inferior and hepatic veins. He received treatment with enoxaparin that was later discontinued after a work-up for thrombophilia was unremarkable. At the age of 12 years, the patient developed Budd-Chiari syndrome that caused acute liver failure and necessitated urgent LTX. Because of nonfunction of his primary graft, he received a second graft after 16 days. Histology showed severe cholestasis but no signs of rejection or thrombosis in the first graft. A few days after the second LTX, the patient developed a steroid-resistant graft rejection despite adequate tacrolimus levels. The graft was rescued by antithymocyte globulin therapy. Over the course of long-term follow-up, liver function remained stable and serial liver biopsies did not identify evidence of findings consistent with rejection.

Nonetheless, the patient’s intestinal symptoms persisted and prompted a more aggressive immunosuppression regimen (aiming to achieve a tacrolimus level of 6–8 ng/mL and prescribing 7.5–20 mg per day prednisolone, depending on the clinical condition). Only intermittent treatment with prednisolone 20 mg (=0.5 mg/kg) improved symptoms. He was also treated regularly with intravenous immunoglobulin because of the increased risk of infections caused by the combination of continuing enteral protein loss, posttransplant immunosuppressive medication, and the previous administration of antithymocyte globulin.

Immediately after his LTX in 2010, we excluded the conditions listed in the introduction that have been associated with Budd-Chiari syndrome. Flow cytometry revealed an isolated CD55 deficiency but normal CD59 expression (Supplemental Fig 5). We had excluded a diagnosis of classic paroxysmal nocturnal hematuris (deficiency of all glycosylphosphatidylinositol-anchored proteins) by absence of a mutation in the phospatidylinositol glycan class A. Moreover, the patient failed to show any signs of hemolysis. At that time, there was not a clear clinical correlation of isolated CD55 deficiency.

We initiated Sanger-sequencing of coding exons and flanking introns of the patient’s CD55 for further insights (Supplement 1). The comparison of homology using a consensus-sequence of CD55 revealed a novel homozygous splice-site variation (G>A) on position +5 of intron 4, NM_000574.4, c.578 + 5G>A. Seven years later, this area was found to be a splice donor site (SDS), and bioinformatic analysis predicted it to cause a loss or skipping of exon 4 in next-generation sequencing. However, some caution is advised for any in silico predictions, and definitive proof would need further experimental work-up. The corresponding in silico data can be found in Supplement 2.

Immunohistochemical staining of colon biopsies and the liver explant (Supplement 3) using a polyclonal antibody against CD55 showed absent CD55 expression in both specimens (Fig 2) and was in line with the hypothesis that the splice variant leads to a frameshift and truncation of the reading frame by aberrant splicing. Although there is practically no doubt that this patient has complete CD55 deficiency, it cannot be proven by immunohistochemical staining and the antibody used, so some caution should be applied. The immunogen, the specific antibody used, resides in a protein area in which the splicing aberration is located, meaning that, whether there is exon 4 skipping or partial intronic inclusion, this antibody cannot bind to the protein. Indeed, although this proves an aberration in the region, it cannot show whether there might be residual or truncated protein synthesis.

However, the clinical relevance of these findings remained uncertain until 2017, when both Ozen et al and Kurolap and colleagues reported on patients with biallelic loss-of-function mutations in CD55 leading to CD55 deficiency. These patients manifested the signs and symptoms of patients with CHAPLE syndrome^1,4  and encouraged us to do further studies to establish a diagnosis in our patient. We excluded other rare Mendelian diseases by whole-exome sequencing (WES) (Supplement 2). The only conclusive likely pathogenic variant was the novel homozygous variant in CD55, NM_000574.4, c.578 + 5G>A (OMIM 125240) with a highly conserved splice and Genomic Evolutionary Rate Profiling Score 5.5. Thus, we concluded that this variant is likely pathogenic and led to the clinical picture of CHAPLE syndrome (according to the clinically augmented ACMG and ESHG criteria) in our patient.⁶  Both clinically unaffected parents were tested and found to be heterozygous carriers for that variant. No testing was performed for the healthy brother (Supplemental Fig 4).

The peer-reviewed reports on CHAPLE syndrome and the confirmation of CD55 deficiency in our patient allowed us to obtain approval for off-label-use of the complement inhibitor, eculizumab (Soliris). In March 2018, after providing informed consent and obtaining vaccination against encapsulated bacteria, the patient began treatment with intravenous eculizumab (Soliris), 1200 mg as induction dose every 14 days and 1200 mg subsequently as a maintenance dose every 4 weeks until the present time.

Within 1 month of treatment, the patient’s abdominal symptoms had resolved completely and albumin and immunoglobulin levels had normalized (Fig 3). Liver function remained completely normal. Immunosuppressive therapy was reduced (targeting tacrolimus levels at 4–6 ng/mL and reducing prednisolone to 5 mg per day). No signs of new thrombotic events were detected. After 56 months, the patient is still in complete remission without any adverse events. His quality of life has improved significantly, and he was able to graduate from high school. He is currently studying at a university.

We report a successful LTX of a boy with Budd-Chiari syndrome because of CD55 deficiency caused by a novel homozygous splice-site mutation in CD55 c.578 + 5G>A x⁶, NM_000574.4. Cosegregation and lack of CD55 protein expression confirmed the pathogenicity of the variant.

The first report on CD55-associated CHAPLE syndrome by Ozen et al, and the first account of successful treatment with eculizumab in vivo by Kurolap et al in 2017,^1,7  allowed us to obtain approval for off-label-use of eculizumab years after diagnosis. The success of therapy with eculizumab is likewise reported by Kurolap et al, in 3 patients with 18 months of follow-up⁸  and by Ozen and colleagues in 16 patients with 20 months of follow-up.⁹  We now report a 56-month long-term follow-up with excellent response and with no adverse side effects.

CD55 deficiency is a very rare, but most likely underdiagnosed disease. Only 33 patients have been described to date.^1,4,8–11  Interestingly, 13 of the 33 patients experienced thrombotic events, and 4 of these 13 patients were diagnosed with Budd-Chiari syndrome (Supplemental Table 1).

In patients with Budd-Chiari syndrome, we recommend seeking other findings of CD55 deficiency, such as protein-losing enteropathy or hyperactivation of the complement system and suggest immunophenotyping and genetic testing for CD55 deficiency. Even if LTX has already taken place, retrospective testing should be performed since eculizumab appears to improve long-term outcome. In addition, CD55 deficiency should be considered in the differential diagnosis of protein-losing enteropathy, because it is usually the first clinical finding of significance. Early detection and treatment may prevent later life-threatening thrombotic events.

We thank the patient and his family for their support; and Cologne Center for Genomics for conducting the whole-exome sequencing.

CHAPLE

complement system hyperactivation, angiopathic thrombosis, early-onset protein-losing enteropathy

IgG

immunoglobulin G

LTX

liver transplantation

SDS

splice donor site

WES

whole-exome sequencing