PF785 CHANGING CAUSES OF MORTALITY IN TDT DURING THE ERA OF ORAL CHELATION THERAPY FROM 2000 TO 2018

Abstract

Background:Heart disease due to transfusion iron overload has been previously reported as the predominant cause of death in transfusion dependent thalassaemia (TDT), but recent reports suggest that liver disease, malignancy and infection are becoming more frequent. If this is the case, changes in standard treatment may be warranted in order to further improve life expectancy.Aims:We aimed to provide robust evidence of these trends by analysing data from the Cyprus thalassaemia cohort study.Methods:This is a long‐term study of outcomes of TDT in the setting of an endemic Mediterranean population where national treatment policies have been consistent in four thalassaemia centres. The cohort has been described previously (Telfer et al, Haematologica 2006). The study update protocol was agreed between investigators at a series of meetings during 2017. A data spreadsheet was populated at each treatment centre and returned to the study centre. Data were checked, queries resolved and final dataset for each centre validated up to 31stDecember 2018. Inclusion criteria were birth after 1/1/60 and continuous treatment and follow up in the four Greek Cyprus thalassaemia centres. The population included a birth cohort of those born from 1/1/74 onwards who had been carefully tracked from birth as part of the Cyprus thalassaemia control programme. Causes of death were classified cardiac, liver (including hepatocellular carcinoma), malignancy, infection and other. Mortality rates were calculated overall and for periods 1980–89, 1990–99, 2000–09 and 2010–18. Survival analysis replicated the methodology in the 2006 study, and for multivariate analysis of risk factors, the study was divided into two time periods (1980–1999 and 2000–2018) to account for the the introduction of oral chelator therapy in 1999/2000. Risk factors for survival were evaluated as Hazard ratios with 95% CI.Results:538 TDT patients were studied over a total of 19,083 patient years. Median age at 30/12/18 was 40.9 yrs, IQR 38.0‐ 47.6. There were 95 deaths, 47.4% cardiac, 9.5% liver, 8.4% malignancy, 14.7% infection and 20% other. Overall the mortality was 5.0 per 1000 patient years follow up. During recent decades, there have been no overall differences in mortality rate, but causes of death have changed due to a downward trend in cardiac and an increasing trend in liver, malignancy and infection‐ related deaths. In the most recent period (2010–18) there was no difference in mortality rates due to different causes (Figure). Estimated survival during the period 2000–2018 was 94% at age 30 (95% CI: 90 ‐ 99%) and 89% at age 40 (95% CI: 85 ‐ 94%). Survival to age 30 improved 8 percentage points compared with the period 1980–1999. In multivariate analysis, independent predictors of survival between 2000 and 2018 were male sex (HR 2.1 95% CI 1.2–4.1), born after 1974 (HR 0.45, 95% CI 0.20 −0.99) and oral chelation therapy with either deferiprone (DFP): HR 0.11, 95% CI 0.06–0.20; or deferasirox (DFX): HR 0.10, 95% CI 0.03–0.32, compared with desferrioxamine (DFO).Summary/Conclusion:These results demonstrate a shift in mortality to causes not directly related to transfusion iron overload. Nevertheless the strongest protective factor for survival is iron chelation therapy, and our data would suggest that DFX and DFP are equally protective. This may relate to better adherence or improved 24 hr iron control compared to DFO. For clinical practice, it remains important to minimize iron loading of heart and liver and to be vigilant about controlling and treating infection.image