Impact of Compliance, Ferritin and LIC on Long-Term Trends in Myocardial T2* with Deferasirox

Abstract

Background: Previous reports on the effects of various iron chelation regimes, including deferasirox, have shown improvements in myocardial iron over a period of up to 1 year. However no longitudinal analysis of myocardial T2* (mT2*) trends with any chelator regimen beyond 1 year has been reported. Monitoring of mT2* has been part of standard management of transfusional iron overload at UCLH for over 8 years. In this study, we have analysed long-term trends in mT2* up to 4 years for patients entered into pre-registration and extension phases of deferasirox studies, 107 and 108, as well as factors that may determine such trends.Patients and Methods: 33 transfusion dependent patients with thalassaemia or rare anaemias, age range 7–51 years, had mT2*, and T2* LIC assessments repeated with a median interval of 14 months (3.6 occasions per patient) while receiving deferasirox. This allowed analysis of factors affecting trends in mT2* on a total of 104 observation periods (of a median of 371 days). Starting doses of 10–30mg/kg/day were predicated by baseline and 1 year biopsy LIC values according to study protocols. Subsequent dose adjustment depending on adverse events profile, trends of ferritin and serum creatinine, degree of proteinuria, or mT2* and T2* LIC was allowed at the investigators' discretion. LIC was estimated from liver T2* with correction for liver R2. A maximum permissible chelator dose in the final two years of the study was 40mg/kg. Compliance was calculated from the % of intended doses that were missed over the period of observation, assessed from patient interviews conducted on monthly visits. 4 patients were excluded from analysis due to lack of an adequate baseline mT2* pre–treatment. At baseline, 13 patients had myocardial iron loading as evidenced by mT2* <20ms, 5 of whom had severe myocardial siderosis with mT2* < 10ms.Results: Average baseline mT2* of 21.6ms improved significantly to 24.7ms (p= 0.046, n=29) at a median follow-up of 4 years (geometric mean 18.9ms and 20.7ms, respectively). The greatest improvement in mT2* was seen in the first year of treatment (0.16ms/month) and the least in the final year giving a median change over the 4 year period of 1.27ms/year. At 4 years, among patients with baseline T2* values <20ms, 7 patients improved of whom 2 patients normalised mT2*, 3 remained stable (change ≤0.2ms), and 3 showed mT2* deterioration. In 16 patients with baseline mT2* >20ms all remained normal apart from one who received low dosing (10mg/kg) in the first months of the study. Analysis of factors associated with trends in mT2* revealed compliance, LIC, and ferritin as statistically significant. For each observation period, compliance <90% (more than 3 daily doses missed per month) was associated with decrease in mT2*, whereas compliance ≥90% with an increase (p=0.0001). Compliance <90% was strongly associated with lack of response to treatment in the 2nd, 3rd, and 4th year (p <0.05). The proportion of patients (periods of observation) failing to comply ≥ 90% rose from 10% to 31% from 1st to 4th year, which correlated with rise in proportion of patients showing mT2* deterioration from 36% to 53% respectively (r= 0.93, p=0.02). Average serum ferritin <1500μg/L during any observation period was associated with improvement in mT2* (p=0.01) as was a falling ferritin trend (p=0.02). LIC values derived from liver T2* with correction for liver R2 also correlated with trend in mT2* (R =0.28 p=0.04). An LIC <7.3 mg/gdw was associated with stabilisation or improvement in mT2*. (p=0.0005) whereas an LIC > 12 mg/gdw was associated with worsening of mT2* (p=0.02).Conclusions: With long-term administration of deferasirox, we conclude that compliance should be maintained above 90% to optimize the sustained control of mT2*. Longitudinal control of LIC and ferritin may also be important to reducing the risk of myocardial iron loading in patients receiving long-term deferasiox treatment.