The original JAMA abstract (2007) reported no overall effect of iron reduction intervention by phlebotomy. However, pre-planned analyses according to randomization variables at entry, including age and ferritin level, were described in the JAMA paper showing improved outcomes with iron reduction with younger age by quartile for the secondary endpoint (p for interaction =0.004) and also suggested a favorable effect in smokers (p for interaction 0.006). Age analyzed as a continuous variable using the Cox proportional hazards regression model and log relative hazard plots revealed that age interacted nonlinearly with treatment in both primary (p=0.04) and secondary (p\<0.001) outcomes. The Cox model showed improved primary (HR 0.47, 95% CI 0.24-0.90, p=0.02) and secondary (HR 0.41, 95% CI 0.24-0.68, p\<0.001) outcomes in youngest age quartile participants (age 43 to 61) randomized to iron reduction versus control. Thus, an interaction between age and level of body iron may have masked beneficial effects of iron reduction in the overall cohort.
Detailed analysis of the effect of age and ferritin levels published in the American Heart Journal confirmed that iron reduction significantly improved primary and secondary outcomes in youngest age quartile participants, as described above, displayed as Kaplan-Meier plots. Mean follow-up ferritin levels (MFFL) declined with increasing entry age in controls. Older age (p=0.026) and higher ferritin (p\<0.001) at entry predicted poorer compliance with phlebotomy and rising MFFL in iron reduction participants. Iron reduction intervention also produced greater ferritin reduction in younger participants. Improved outcomes with lower MFFL occurred in iron reduction patients for both primary (HR 1.11, 95% CI 1.01-1.23, p=0.028) and secondary (HR 1.10, 95% CI 1.0-1.20, p=0.044) outcomes, and for the entire cohort: primary outcome (HR 1.11, 95% CI 1.01-1.23, p=0.037). Improved outcomes occurred with MFFL below versus above the median of the entire cohort means: primary outcome HR 1.48, 95% CI 1.14-1.92, p=0.003; secondary outcome HR 1.22, 95% CI 0.99-1.50, p=0.067.
Thus, lower iron burden predicted improved outcomes overall and was enhanced with iron reduction by phlebotomy. Controlling iron burden may improve survival, and prevent or delay non-fatal myocardial infarction and stroke. These findings warrant confirmation using further studies.
A possible effect of iron levels on risk of cancer as well as vascular disease was recognized at trial inception. Participants with visceral malignancy within the preceding five years were excluded from this study. However, information was collected prospectively on the occurrence of new visceral malignancy and cause-specific mortality including death due to cancer. As reported in the Journal of the National Cancer Institute, a new visceral malignancy was diagnosed during follow-up in 60 control and 38 iron reduction participants, a 37% (HR 0.63; 95% CI = 0.42 - 0.95, p = 0.026) decrease in risk with iron reduction. Reduced cancer risk with iron reduction was confirmed on time-to-event analysis (HR = 0.65; 95% CI = 0.43 - 0.97, p = 0.036). Reduced risk was observed for several common tumor types. Iron reduction participants had lower cancer - specific mortality and lower all-cause mortality in participants diagnosed with cancer (HR = 0.39; 95% CI = 0.21 - 0.72, p = 0.003 and HR = 0.49; 95% CI = 0.29 - 0.83, p = 0.009 respectively), compared to control participants. MFFL during follow-up in those participants randomized to iron reduction who developed cancer were comparable to levels in control participants (t93 = 0.8, p = 0.428). The MFFL in participants randomized to iron reduction developing cancer was 127 ng/mL, 95% CI = 71.2 - 183.0. The MFFL was significantly lower in participants not developing cancer, 76.4 ng/mL, 95% CI = 71.4 - 81.4, p = 0.017). Participants randomized to iron reduction developing cancer appeared to be relatively non-compliant with intervention.
Analysis of data from the FeAST study continues to delineate interactions between iron status and smoking, lipid levels and statin use, diabetes and race. It has been shown that ferritin levels ranging from about 70 to 79 ng/mL are associated with lower mortality and levels of inflammatory markers. Statin use, while not a randomization variable, has been monitored and shown to relate to lower ferritin levels.
