Serum Ferritin Levels and Morbidity in β-Thalassemia Intermedia: A 10-Year Cohort Study

An association between iron overload and morbidity in patients with β-thalassemia intermedia (TI) has been established. However, available studies relied on cross-sectional analysis (measures of prevalence rather than incidence) without a clear chronological relationship between risk factors and outcomes.

We conducted a retrospective cohort study of TI patients treated at five comprehensive care centers in Italy, Lebanon, Oman, Iran and Egypt. We included all patients attending the centers since 01 January 2000 and regularly followed until 31 December 2009 or death. For each patient, we retrieved serum ferritin (SF) and total hemoglobin (Hb) levels for every year during the 10-year follow up period, and calculated the average SF and total Hb levels during the study period (10-year indices). Moreover, we retrieved data on the incidence of nine pre-defined morbidities during the study period. Data on splenectomy status was also retrieved. For this preliminary study, we analyzed data from the first 52 recruited patients. None of the patients were regularly transfused or iron chelated during the study period.

The mean age of patients at study entry was 24.1 ± 11.3 years with 48.1% of patients being males. Thirty (57.7%) patients were splenectomized. The mean SF-index was 714.3 ± 518.7 ng/ml (range: 38.4 to 1926.2 ng/ml) and the mean Hb-index was 8.9 ± 1.2 g/dl (range: 7–12 g/dl). None of the patients died during the study period. Thirty-six patients experienced a morbidity during the study period, giving a 10-year cumulative incidence of 69.2% (19 [36.5%] had a single morbidity while 17 [32.7%] had multiple morbidities). The most common morbidity was osteoporosis (48.1%) followed by extramedullary hematopoiesis (19.2%), liver disease (17.3%), hypothyroidism (9.6%), diabetes mellitus (7.7%), hypogonadism (7.7%), thrombosis (5.8%), pulmonary hypertension (1.9%), and hypoparathyroidism (1.9%). The mean SF-index was higher in patients who developed morbidity than those who did not (877.5 vs. 346.9 ng/ml, p<0.001). There was also a positive correlation between SF-index and the number of developed morbidities (r=0.507, p<0.001). On logistic regression analysis, the association between SF-index and the development of morbidity remained significant upon adjustment for age, sex, splenectomy and Hb-index (p=0.015). SF-index had a strong predictive power for morbidity (Area Under the Receiver Operating Characteristic Curve=0.816 ± 0.059, p<0.001). We also evaluated effect estimates for the development of morbidity for three SF-index thresholds, of clinical interest and representing quartile cut-offs. The relative risk of morbidity for patients with a SF-index of >300 ng/ml compared with ≤300 ng/ml was 8.00 (95% CI: 1.92 to 33.38); while it was 14.00 (95% CI: 2.73 to 71.86) for patients with a SF-index of >600 ng/ml compared with ≤600 ng/ml. The effect estimate for the >1000 ng/ml threshold was non-estimable (all patients with a SF-index of >1000 ng/ml developed a morbidity).

There exists a clear association between elevations in SF level and an increased risk of morbidity in TI patients, further supporting the need for iron chelation therapy in this patient population. The relative risk becomes considerably high for increases beyond 300 ng/ml.

Musallam:Novartis Pharmaceuticals: Honoraria. Cappellini:Novartis Pharmaceuticals: Speakers Bureau. Taher:Novartis: Honoraria, Research Funding.