Improved survival in thalassemia major patients on switching from desferrioxamine to combined chelation therapy with desferrioxamine and deferiprone

Life expectancy in thalassemia major (TM) is reduced, mainly as a result of deaths from transfusion iron-overload and cardiac dysfunction. Early deaths are preventable by effective long-term iron chelation therapy, but 50% are unable to adhere adequately to standard treatment consisting of sub-cutaneous desferrioxamine (DFO) infusions administered at least five times per week.1,2 In a previous report on a cohort of TM patients treated in Cyprus, we noted a significantly lower incidence of fatal events co-inciding with follow-up and treatment after 1999,3 an observation also made in the UK.4 The oral chelating agent deferiprone (DFP) was licensed in the European Union in 1999. Since effectiveness of DFP as monotherapy was thought to be insufficient, and there had been reports of more efficient chelation when DFP was used in combination with DFO,5 a protocol of combination chelation therapy (CCT) was introduced in Cyprus to optimize the benefit from DFP. In the previous study, we suggested that use of CCT was the explanation for the decrease in mortality, but could not show this statistically due to insufficient data. With a further two years of follow-up, we have readdressed this question.

Indications for CCT were: moderate or high iron stores (ferritin <1500 μg/L), cardiac dysfunction (symptomatically or on echocardiography) or decreased myocardial T2* (<20 milliseconds). The protocol for CCT was similar to that described by Origa et al.:6 DFP 75–100 mg/kg per day seven days per week, and DFO 30–50 mg/kg over 12–24 hours 2–5 infusions per week depending on ferritin levels. DFO was generally stopped if ferritin was consistently <500 μg/L, but DFP continued.

The patient cohorts and methods of follow-up were described in the original publication.3 The cohort born after 1973 has undergone an additional analysis as a birth cohort (these patients were prospectively registered at diagnosis and consist of all new diagnoses in Cyprus). Cox Proportional Hazards was used to assess factors impacting on survival with CCT modelled using time of starting as a time-dependent covariate. We exclude those treated for less than six months.

Table 1.Demographic data for Cyprus patients.

Table 2.Hazard ratios for prediction of survival in multivariate analysis.

Demographic data are shown in Table 1. The full cohort includes 5 more than in the previous publication: these had been erroneously omitted. Sixty-five patients died between 1/1/80 and 31/12/06. Thirty-two (49% of deaths) were from cardiac causes. Other causes of death were: infection 9 (14%), accident 6 (9%), other 6 (9%), not known 5 (8%), malignancy 4 (6%), liver disease 3 (5%). Within the Birth Cohort, 18 (6%) patients died, 7 from cardiac causes.

Three-hundred and four (55.9%) patients switched to CCT and 269 (49.4%) continued for at least six months. Indications for switching were: high serum ferritin (66.6%), deteriorating left ventricular function on echocardiography or clinically (25.2%), low myocardial T2* (2.8%), and intolerable side effects of frequent DFO infusions (5.4%).

Results of multivariable analysis are shown in Table 2. For CCT, the hazard ratio of 0.14 equates to 7.4-fold improved survival for each year on therapy. In the birth cohort, CCT was the only independent factor associated with survival. For cardiac deaths, the effect of CTT could not be calculated since there were no events in those initiated on therapy. Making an arbitrary assumption that the patient with the median duration of CCT had in fact died of a cardiac cause, CCT is associated with a 4-fold increased chance of survival from cardiac death.

Seventy-five patients (24.7%) stopped CCT. This high incidence of discontinuations was consistent with most other long-term studies in CCT.6–9 Reasons for stopping were: agranulocytosis (5%), recurrent neutropenia (2.9%), gastrointestinal disturbance (5.6%), non-compliance with DFO (3.3%), pregnancy (2.6%), arthralgia (1.6%), allergy (0.7%), weight gain (0.7%), increased liver enzymes (0.3%), other (2%). For those treated for at least six months, the median (standard deviation) therapy duration was 3.59 (1.96) years. There was one death in a patient during treatment with CCT. This was due to E. coli sepsis and was not associated with neutropenia or agranulocytosis.

This study provides good evidence that CCT is effective in controlling body iron stores in moderately to heavily iron loaded patients and prevents iron overload related deaths. It is not clear whether equal benefits would be obtained with DFP monotherapy, as shown in another study,10 or with the new oral iron chelating drug deferasirox, where the efficacy in terms of survival has not yet been established. We observed a high rate of discontinuations and of severe but manageable adverse events with CCT, highlighting the need for supervision from a specialist center. Management of the risk of agranulocytosis requires weekly full blood counts, and periodically reminding patients about this potential risk.

Acknowledgments

the authors would like to acknowledge the cooperation of The Ministry of Health of Cyprus for supporting the work, The Cyprus Institute of Neurology and Genetics for undertaking DNA analysis, the present and past staff in the thalassemia clinics, laboratories and blood banks in the thalassemia centers and associated hospitals in Nicosia, Limassol, Larnaca and Paphos. They would also like to acknowledge the co-operation of the thalassemic patients and families and voluntary groups, including the Pancyprian Thalassaemia Society and the Thalassaemia International Federation.

Footnotes

• Funding: SC, MH, MS, AK and MA received travel grants to attend scientific meetings from Apopharma and Novartis. PT has received travel grants to attend scientific meetings and honoraria for lectures and advisory work from Apopharma and Novartis.