Unchanneling cardiac iron in humans

In this issue of Blood, Fernandes et al show that treatment of thalassemia patients with the calcium channel blocker amlodipine significantly reduces cardiac iron loading compared with placebo (P < .02).¹ 

Thirty subjects with cardiac iron >0.59 mg/g dry weight heart tissue based on magnetic resonance imaging (MRI) were stratified to the iron “reduction” group and 29 with cardiac iron ≤0.59 mg/g cardiac tissue were placed in the “prevention” group. After 12 months, the cardiac iron in the “reduction” group was 0.26 mg/g lower in the 15 subjects randomized to amlodipine plus standard iron chelation therapy compared with a 0.01 mg/g increase in the 15 subjects treated with placebo plus standard chelation (P < .02). Each of the subjects on amlodipine had a decrease in iron, and the overall average decrease was 21%. Amlodipine caused no change in liver iron concentration (LIC) or left ventricular ejection fraction.

So the question is, why is this small clinical trial of such pivotal importance in this day and age of massive multicenter prospective randomized studies?

The answer, in our opinion, is that this clinical study tells us that iron entry into the heart through l-type calcium channels (see figure), a mechanism that has been clearly demonstrated in vitro,^2,3  seems to be actually occurring in humans. As an added bonus, we have a possible new adjunctive treatment of iron cardiomyopathy.

The whole field of thalassemia and management of transfusional iron overload has made quantum advances in the past decade or 2. In the mid-70s, the median survival for thalassemia major was around 15 years of age, with death being secondary to cardiac complications of iron overload. By the early 2000s, mortality from iron cardiomyopathy had dropped by over 70%, thanks to effective chelators and the ability to routinely monitor organ iron concentration by MRI.⁴  It is clear from serial monitoring by MRI that iron loading and unloading occurs at different rates in different tissues. The liver loads iron very quickly, and in fact the LIC is highly correlated with total body iron. However, there is a very poor correlation between the LIC and the pancreatic, pituitary, or cardiac iron concentrations, indicating that other factors control iron trafficking in these organs.⁵  The explanation lies in the very elegant iron regulatory biochemistry that has been elucidated over the past 2 decades. Normally, iron circulates bound to TF and enters cells by receptor-mediated endocytosis via the TF receptor, TfR1. When LCI (Fe⁺²) increases, TfR1 transcription decreases, preventing cellular iron overload. However, when total iron body increases dramatically, as is the case in patients on chronic transfusion, the TF-binding ability is rapidly exceeded and circulating non-TF bound iron (NTBI) appears in the plasma. NTBI rises considerably when the TF saturation reaches ∼60%, and a highly reactive Fe⁺² subspecies of NTBI called LPI increases concomitantly. This is when problems begin. Humans do not normally have circulating Fe⁺². LPI can then enter cells through ion transporters that are normally designed to carry divalent cations like zinc and calcium. For the most part, these ion transporters are not regulated by intracellular iron concentration and iron loading proceeds, even though cytosolic iron levels may be very high. Various levels of expression of divalent metal transporters (DMT1, ZIP14) and of the iron exporter, ferroportin have been demonstrated in the liver, pancreas, erythroid precursors, and the pituitary of animal models, and the proportion of transporters that have their transcription regulated by iron is organ specific. For example, there are few iron response elements (IRE)-containing DMT1 messenger RNA splice variants in the pancreas, whereas all DMT1 splice variants in the brain contain an IRE. Thus, the pancreas loads Fe⁺² in the face of high intracellular iron, whereas the brain does not.⁵  The heart also contains l-type and t-type calcium channels,³  both being quite capable of transporting Fe⁺² into the heart, which of course brings us to the l-type calcium channel blocker, amlodipine (see figure).

This influx of Fe⁺² into the heart via calcium channels has several consequences.^3,6  Fe⁺² competes with the entry of Ca⁺² into the heart via calcium channels. This can have a significant effect on cardiac excitation contraction coupling. The elevated Fe⁺² in the cardiomyocytes increases the production of intracellular ferritin, the protein that binds toxic Fe⁺² and converts it to the nonreactive Fe⁺³. LCI in the cytosol is highly toxic and produces reactive oxygen species that poison the ryanodine receptors in the sarcoplasmic reticulum that are critical for myocyte function.³  Lastly, Fe⁺² is transported into myocardial mitochondria leading to mitochondrial iron overload and dysfunction that can further affect myocardial contractility. Interestingly, severe mitochondrial iron loading can cause relative depletion of cytosolic iron, resulting in a positive feedback loop that increases iron flux into the myocyte.⁷  This probably explains the clinical observation that cardiac function and cardiac iron levels remain stable for years, but once loading starts, cardiac iron loading and dysfunction progress very rapidly over a few months, often leading to death.

The problem is that all of this seemingly coherent and elegant iron trafficking story, including the idea that the l-type calcium channel blocker amlodipine, as well as other calcium channel blockers can modulate cardiac iron trafficking, is derived from in vitro studies in cell culture and from knockout mice. Certainly, serial measurements showing differential rates of loading and unloading during chelation in the liver, endocrine organs, and heart in humans are consistent with the elegant biochemistry worked out in the laboratory.⁵  However, some hand waving is required to explain this organ-specific iron trafficking in humans, because mice are not humans. Fernandes et al have shown that the effect of calcium channel blockade amlodipine on cardiac iron loading predicted in animals also resulted in reduced cardiac iron in humans. As the authors point out, more clinical studies are needed, and certainly biochemical studies need to continue because all calcium channel blockers do not have the same effect in vitro,³ but at least the “channels” for more progress on both clinical and biochemical fronts are now open.