Revisiting iron overload status and change thresholds as predictors of mortality in transfusion-dependent β-thalassemia: a 10-year cohort study.
Khaled M MusallamSusanna BarellaRaffaella OrigaGiovanbattista Giovanni Battista FerreroRoberto LisiAnnamaria PasanisiFilomena LongoBarbara GianesinGian Luca Forninull nullPublished in: Annals of hematology (2024)
Data on iron overload status and change thresholds that can predict mortality in patients with transfusion-dependent β-thalassemia (TDT) are limited. This was a retrospective cohort study of 912 TDT patients followed for up to 10 years at treatment centers in Italy (median age 32 years, 51.6% female). The crude mortality rate was 2.9%. Following best-predictive threshold identification through receiver operating characteristic curve analyses, data from multivariate Cox-regression models showed that patients with Period Average Serum Ferritin (SF) > 2145 vs ≤ 2145 ng/mL were 7.1-fold (P < 0.001) or with Absolute Change SF > 1330 vs ≤ 1330 ng/mL increase were 21.5-fold (P < 0.001) more likely to die from any cause. Patients with Period Average Liver Iron Concentration (LIC) > 8 vs ≤ 8 mg/g were 20.2-fold (P < 0.001) or with Absolute Change LIC > 1.4 vs ≤ 1.4 mg/g increase were 27.6-fold (P < 0.001) more likely to die from any cause. Patients with Index (first) cardiac T2* (cT2*) < 27 vs ≥ 27 ms were 8.6-fold (P < 0.001) more likely to die from any cause. Similarly, results at varying thresholds were identified for death from cardiovascular disease. These findings should support decisions on iron chelation therapy by establishing treatment targets, including safe iron levels and clinically meaningful changes over time.
Keyphrases
- prognostic factors
- iron deficiency
- cardiovascular disease
- cardiovascular events
- sickle cell disease
- electronic health record
- cardiac surgery
- big data
- multiple sclerosis
- end stage renal disease
- mass spectrometry
- chronic kidney disease
- magnetic resonance imaging
- data analysis
- computed tomography
- peritoneal dialysis
- combination therapy
- stem cells
- coronary artery disease
- heart failure
- machine learning
- magnetic resonance
- ejection fraction
- newly diagnosed
- contrast enhanced
- bone marrow
- deep learning