Congenital Dyserythropoietic Anemia Type II: An Update from the Congenital Dyseryhtropoietic Anemia Registry of North America (CDAR)

Congenital Dyserythropoietic Anemias (CDAs) are a heterogeneous group of rare anemias characterized by ineffective erythropoiesis and hemolysis. The CDA registry in North America (CDAR; NCT02964494) was established to facilitate natural history studies. Participants who do not have a genetic diagnosis are offered family-trio whole genome sequencing (WGS) and elect to provide samples to CDAR biorepository to support collaborative mechanistic studies.

CDA type II (CDA-II), the most common CDA type, is an autosomal recessive disease caused by biallelic SEC23Bvariants. CDA-II presents with hemolytic anemia with suboptimal reticulocytosis, iron overload, splenomegaly, and characteristic binucleation in 10-30% of the bone marrow (BM) erythroblasts. Although CDA-II is a recessive disorder, only one SEC23B variant is found in some patients. Here, we present the genetic and clinical data of CDA-II patients enrolled in CDAR so far.

As of July 2023, 169 individuals (85 patients; 84 unaffected family members) enrolled in CDAR. Nine have a confirmed CDA-II diagnosis. Six patients have biallelic SEC23B pathogenic variants and 3 have only one variant identified. WGS for the 3 patients with a single variant revealed rare deep-intronic candidate variants. Since SEC23Bis ubiquitously expressed, we studied SEC23B protein levels in patient-derived lymphoblastoid cell lines generated for the CDAR biorepository. Patients with a single SEC23B variant exhibited an expression level comparable to those with biallelic mutations (Figure). The candidate deep intronic variants are currently being investigated for pathogenicity.

Seven patients presented with hemolytic anemia during infancy, but the age of diagnosis was delayed by a mean of 12.5 yrs (0.1 to 55 y.o). Blood smears showed mild macrocytosis, poikilocytosis, and several spherocytes congruent with an osmotic gradient ektacytometry resembling mild hereditary spherocytosis. BM examination, when performed, exhibited erythroid hyperplasia, prominent dyserythropoiesis, and binucleation in ≥ 10% of erythroblasts. Anemia severity was variable: 6 patients were transfusion-dependent, 1 received intermittent transfusions, and 2 did not need blood transfusions. Two became transfusion-independent after splenectomy at ages 6 and 7 y.o, and two underwent successful matched-sibling BMT at age 10 and 26 y.o, after adequate chelation (liver iron concentration [LIC]<7 mg/g)

Iron overload, disproportionate to transfusions, was present in 6 patients; 5 received iron chelation. All 5 patients with available MRI results had hepatic hemosiderosis (LIC range 3.4 to 21.8 mg/g) and evidence of myocardial hemosiderosis (T2* range 6.3 to 20 ms). Ferritin values were variable (range 58 to 1631.5 ng/ml) and discordant from LIC and T2* measurements. Notably, ferritin underestimated iron overload burden: 2 patients with normal ferritin (58 and 160 ng/mL) had hepatic hemosiderosis (LIC 6.1 and 3.4 mg/g respectively), and patients with severe hepatic and myocardial hemosiderosis (LIC 16.6-21.8 mg/g; T2* 6.3-7.8 ms) had modest ferritin values 1475-1631 ng/mL. Iron overload was present even in a non-transfusion-dependent patient and worsened in a patient post-splenectomy, suggesting a role for ineffective erythropoiesis. To further investigate this possibility, we measured erythroferrone (ERFE), GDF-15, and hepcidin levels in 2 patients so far. ERFE and GDF-15 levels were remarkably higher than controls, and hepcidin-to-ferritin ratios were significantly lower, suggesting that high ERFE, caused by ineffective erythropoiesis, inappropriately suppressed hepcidin for the body's iron levels.

In summary, CDA-II diagnosis is often delayed despite early onset of anemia. Genetic testing may expedite the diagnosis. The severity of anemia is variable and there are several patients requiring transfusions. The risk of iron overload, including myocardial overload, is high and appears to be related to ineffective erythropoiesis. While frequency of transfusions may improve post-splenectomy, iron overload is not determined by blood transfusions alone and does not resolve post-splenectomy. Serum ferritin underestimated the burden of tissue iron in CDA-II, suggesting that MRI scans should be employed early. Future studies may identify a role for SEC23B deep intronic variants in CDA-II patients with heterozygous variants.