Increased prevalence and severity of cardiac iron overload in transfusion dependent diamond-blackfan anemia syndrome caused by small versus large ribosomal subunit mutations Free

Diamond-Blackfan anemia syndrome (DBAS) is an inherited bone marrow failure syndrome characterized by erythroid hypoplasia and frequent transfusion-dependence. Secondary iron overload (IO) is a major cause of morbidity and mortality, but data on its prevalence and risk factors are limited. We previously showed more severe erythroid failure and earlier age of presentation with anemia in DBAS caused by small (RPS) versus large (RPL) ribosomal subunit mutations (Iskander, 2021). This study aimed to define the frequency and severity of hepatic and cardiac IO and their relationship to genotype in a UK cohort of 53 transfusion-dependent patients with DBAS.

Methods Retrospective data were collected from electronic patient records and interviews. Patients who discontinued transfusions due to spontaneous recovery were included if transfusion duration was ≥ 18 months. Genotype was determined using a targeted sequencing panel and variant pathogenicity defined as previously described (Gerrard, 2013). Liver iron concentration (LIC) and cardiac iron were measured by the gold standard non-invasive modality: MRI Ferriscan® and MRI T2*, respectively. Liver biopsy was used to grade fibrosis using Ishak scores from 0-6. Data were analysed using SPSS v29.0.2. Groups were compared using Mann-Whitney-U or Pearson chi-squared tests and relationships between variables was measured using linear or logistic regression coefficients. Kaplan-Meier analysis was used to calculate median iron-free survival. Two-tailed testing was used and the threshold for significance was p<0.05.

Results Median age at analysis was 17.3 yrs (range 2.5-62.6). Female: male ratio was 1.2 and ethnicity distribution matched UK Census data. A heterozygous pathogenic variant was identified in 49/53 (92.4%) of cases: RPS in 33 (62.2%) and RPL in 16 (30.2%). 51 (96.2%) of patients were treated with chelation at first MRI assessment: oral deferasirox in 37 (69.8%) and subcutaneous desferrioxamine in 14 (26.4%).

Ferrriscan® data were available for 51/53 (96.2%) patients: 44/51 (86.2%) and 36/51 (70.5%) of patients developed moderate (LIC=7-10mg/g dry weight) or severe (>15) hepatic IO at a median age of 5.5 and 8.5 yrs, respectively. The earliest age of hepatic IO was 1.0 yrs and lifetime mean LIC was 9.6 mg/g dry weight (range 2.2–29.9). Survival with no or mild hepatic IO (LIC ≤7 mg/g dry weight) was short (median 1.5 yrs post transfusion dependence). Liver fibrosis (Ishak ≥1) occurred in 34/36 (94.4%) patients who underwent histological analysis. Higher mean LIC significantly predicted mean Ishak score (p<0.001) and advanced fibrosis (Ishak ≥4) only arose in patients with severe hepatic IO.

Cardiac MRI T2* data were available for 44/53(83%) patients. Cardiac IO (T2* <20ms) occurred in 11/44 (25%) and life-threatening cardiac IO (T2*<10ms) in 2/44 (4.5%) at median ages of 9.4 and 24.5 yrs, respectively. The earliest age at which cardiac IO was detected was 2.28 yrs and the average lifetime mean cardiac MRI T2* was 31.25ms.

Mean LIC was significantly increased in patients with interruptions to chelation due to non-compliance or toxicity (p=0.035). Sex, ethnicity, transfusion volume (ml/kg/yr), transfusion duration and RPL/RPS genotype did not predict mean LIC. Cardiac IO developed in 0/15 RPL and 10/29 (34.4%) RPS patients, therefore was significantly associated with RPS genotype (p=0.026). Furthermore, RPS patients had more severe cardiac loading than RPL patients (p=0.039), defined as lower lifetime mean MRI T2*, not confounded by follow-up period/transfusion burden. No other variables were significantly associated with cardiac IO, and the effect of genotype on cardiac IO was preserved on multivariate analysis.

Conclusion IO is the second leading cause of mortality in non-transplanted DBAS (Vlachos, 2018), but disease-specific risk factors are unknown. Leveraging retrospective data from a large UK DBAS cohort, we show, for the first time, that prevalence and severity of cardiac IO are increased in patients with RPS genotype. We hypothesise that more severe depletion of bone marrow erythroid precursors in RPS versus RPL DBAS (Iskander, 2021) leads to lower usage of transferrin iron (Porter, 2014), hence increased susceptibility to cardiac IO. Cardiac IO arises in 25% of transfusion-dependent patients with DBAS, while hepatic IO occurs in 86%, despite aggressive chelation strategies.