Blood transfusion

Gaining the most accurate information on the rate of iron loading from transfusion therapy is important in assisting selection of the best chelation therapy for each patient. A unit processed from 420 ml of donor blood contains approximately 200 mg of iron, or 0.47 mg/ml of whole donor blood. For red cell preparations with variable haematocrits, the iron in mg/ml of blood can therefore be estimated from 1.16 x the haematocrit of the transfused blood product. In cases where organisational systems or other difficulties prevent such estimations from being calculated, a rough approximation can be made based on the assumption that 200 mg of iron is contained in each donor unit. Irrespective of whether the blood used is packed, semi-packed or diluted in additive solution, if the whole unit is given, this will approximate to 200 mg of iron intake. According to the recommended transfusion scheme for thalassaemia major, the equivalent of 100–200 ml of pure red blood cell (RBC) per kg body weight per year are transfused. This is equivalent to 116-232 mg of iron/kg body weight/year, or 0.32-0.64 mg/kg/day. Regular blood transfusion therapy therefore increases iron stores to many times the norm unless chelation treatment is provided. If chelation therapy is not given, Table 1 shows how iron will accumulate in the body each year, or each day.

Table 1

Iron loading rates in the absence of chelation.

Increased gastrointestinal absorption of iron

In transfusion-dependent thalassaemia (TDT), the contribution of iron absorbed from the diet is small compared with blood transfusion. Normal intestinal iron absorption is about 1-2 mg/day. In patients with thalassaemia who do not receive any transfusion, iron absorption increases several-fold. It has been estimated that iron absorption exceeds iron loss when expansion of red cell precursors in the bone marrow exceeds five times that of healthy individuals. Transfusion regimens aimed at keeping the pre-transfusion haemoglobin above 90 g/l have been shown to prevent such expansion (Cazzola et al., 1997). In individuals who are poorly transfused, absorption rises to 3-5 mg/day or more, representing an additional 1-2 g of iron loading per year.

Mechanisms of iron toxicity

Iron is highly reactive and easily alternates between two states – iron3+ and iron2+ – in a process which results in the gain and loss of electrons, and the generation of harmful free radicals (atoms or molecules with unpaired electrons). These can damage lipid membranes, organelles and DNA, causing cell death and the generation of fibrosis. In health, iron is ‘kept safe’ by binding to molecules such as transferrin, but in iron overload the capacity to bind iron is exceeded both within cells and in the plasma compartment. The resulting ‘free iron’, either within cells or in plasma, damages many tissues in the body or is fatal unless treated by iron chelation therapy. Free iron also increases the risk of infections and neoplasia.

Fig 1

A summary of the mechanisms for toxic effects of iron overload. Pathological mechanisms and consequences of iron overload. In iron overload resulting from repeated blood transfusions or long-term increased iron absorption, iron that is not bound to naturally (more...)

Distribution and consequences of transfusional iron overload

In the absence of iron overload, uptake of iron into cells is controlled by the interaction of transferrin with its receptors – mainly on red cell precursors, hepatocytes and dividing cells. In iron overload, transferrin becomes saturated and iron species that are not bound to transferrin are present in plasma (plasma non-transferrin bound iron, or NTBI). The distribution of NTBI uptake is fundamentally different from transferrin uptake, and is thought to involve calcium channels. Organ damage in transfusional iron overload reflects the pattern of tissue iron uptake from NTBI. Some tissue are spared from iron loading through this mechanism (such as skeletal muscle), while others such myocardial muscle, endocrine tissue and hepatocytes take up NTBI rapidly. This iron is then stored as ferritin or haemosiderin which are visible by magnetic resonance imaging (MRI). The myocardial iron overload can induce heart failure from cardiomyopathy in patients without chelation in as early as the second decade of life. Iron overload also causes pituitary damage, leading to hypogonadism, growth retardation and delayed puberty.

Endocrine complications, namely diabetes mellitus, hypothyroidism and hypoparathyroidism are also seen. Liver disease with fibrosis and eventually cirrhosis and hepatocellular carcinoma, particularly if concomitant chronic viral hepatitis is present, are also serious complications.

Fig 2

The main routes of iron turnover and uptake are shown by solid black arrows on the right panel: 20 mg of iron is delivered daily to the erythron in health. This increases several fold in untransfused thalassaemias but can be inhibited by hypertransfusion. (more...)

Why measure serum ferritin?

Serum ferritin (SF) generally correlates with body iron stores, and is relatively easy and inexpensive to determine repeatedly. Serum ferritin is most useful in identifying trends. A decreasing trend in SF is good evidence of decreasing body iron burden but absence of a decreasing trend does not exclude a decreasing iron burden. However, an increasing SF trend implies an increasing iron burden but may also be due to inflammation or tissue damage, so clinical judgment must be used to interpret these trends.

Long term control of SF is also a useful guide to the risk of complications from iron overload in thalassaemia major (TM); many studies have shown an association between the control of serum ferritin and prognosis (Borgna-Pignatti et al., 2004; Davis et al., 2004; Gabutti & Piga, 1996; Olivieri et al., 1994). Studies have identified a significantly lower risk of cardiac disease and death in at least two-thirds of cases where serum ferritin levels have been maintained below 2,500 μg/l (with deferoxamine – desferrioxamine, DFO) over a period of a decade or more (Olivieri et al., 1994). Observations with larger patient numbers show that maintenance of an even lower serum ferritin of 1,000 μg/l may be associated with additional clinical advantages (Borgna-Pignatti et al., 2004) (see Table 2).

Table 2

Use of serum ferritin for monitoring chelation treatment.

What are the limitations of serum ferritin measurements?

Most SF assays were developed mainly for detecting iron deficiency, and the linear range of the assay at high SF values needs to be known. SF must be performed in a laboratory that has established how to dilute samples with high values, to give readings within the linear range of the assay. SF measures do not always predict body iron or trends in body iron accurately. In TM, variation in body iron stores accounts for only 57% of the variability in serum ferritin (Brittenham et al., 1993). This variability is in part because inflammation increases serum ferritin, and partly because the distribution of liver iron between macrophages (Kupffer cells) and hepatocytes in the liver has a major impact on serum ferritin. A sudden increase in serum ferritin should prompt a search for hepatitis, other infections, or inflammatory conditions. A lack of fall in SF with chelation does not therefore necessarily prove that the patient is a ‘non responder’ to the chelation regime.

As outlined above, this can be because inflammation may have falsely raised SF, or because the relationship between body iron and SF is not always linear, particularly in the context of inflammation or tissue damage (Adamkiewicz et al., 2009), and body iron can fall considerably from a high starting point (e.g. liver iron concentration >30 mg/g dry weight) before a change in ferritin is clear. Below 3000 μg/l SF values are influenced mainly by iron stores in the macrophage system, whereas above 3000 μg/l they are determined increasingly by ferritin leakage from hepatocytes (Davis et al., 2004; Worwood et al., 1980). Day-to-day variations are particularly marked at these levels. The relationship between serum ferritin and body iron stores may also vary depending on the chelator used (Ang et al., 2010) and by duration of chelation therapy (Fischer et al., 2003).

Why monitoring liver iron concentration?

- To identify whether body iron is adequately controlled.

Adequate control of LIC is linked to the risk of hepatic damage as well as the risk of extrahepatic damage. Normal LIC values are up to 1.8 mg/g dry weight (wt), with levels of up to 7 mg/g dry wt seen in some non-thalassaemic populations without apparent adverse effects. Sustained high LIC (above 15-20 mg/g dry wt) have been lined to worsening prognosis, liver fibrosis progression (Angelucci et al., 1997) or liver function abnormalities (Jensen et al., 2003). In the absence of prior iron chelation therapy, the risk of myocardial iron loading increases with the number of blood units transfused and hence with iron overload (Jensen et al., 2003; Buja & Roberts, 1971). However, the relationship between LIC and extra-hepatic iron is complicated by chelation therapy as iron tends to be accumulate initially in the liver and later in the heart but also is removed more rapidly from the liver than the heart by chelation therapy (Noetzli et al., 2008; Anderson et al., 2004). Thus, in patients receiving chelation therapy, whilst high LIC increases the risk of cardiac iron overload, the measurement of LIC will not predict myocardial iron and hence cardiac risk reliably, and myocardial iron may be found in some patients despite currently well controlled LIC.

- To determine iron balance: is body iron increasing or decreasing on current therapy?

LIC is the most reliable indicator of body iron load, which can be derived from the following formula: Total body iron stores in mg iron /kg body wt = 10.6 x the LIC (in mg/g dry wt) (Angelucci et al., 2000). Sequential measurement of LIC is the best way to determine whether body iron is increasing or decreasing with time (iron balance). While serum ferritin is simple, relatively inexpensive and can be repeated frequently, LIC determination should be considered for those patients whose serum ferritin levels deviate from expected trends (i.e. those with suspected co-existing hepatitis, or patients on chelation regimens with variable or uncertain responses), as this may reduce the risk of giving either inadequate or excessive doses of chelation therapy. Since the relationship of SF to iron overload and iron balance has not yet been established, assessment of LIC may be particularly useful when new chelating regimes are being used. At high levels of SF (>4000 μg/l), the relationship to LIC is not linear and patients may show a fall in LIC (negative iron balance) without a clear trend in SF in the first 6-12 months. When a patient fails to show a fall in SF over several months the change in LIC can identify whether the current regime is adequate or need to be modified (increased frequency or adherence, increased dose or change in regime).

Methods for measuring LIC

- Biopsy

Measurement of LIC was initially done by chemical determination on a liver biopsy sample (fresh, fixed or from dewaxing of paraffin-embedded material) (see Table 3). Biopsy is an invasive procedure, but in experienced hands has a low complication rate (Angelucci et al., 1997). Inadequate sample size (less than 4 mg dry wt or less than about a 2.5 cm core length) or uneven distribution of iron, particularly in the presence of cirrhosis (Villeneuve et al., 1996), may give misleading results however. Biopsy also allows the evaluation of liver histology which cannot yet be reliably estimated by non-invasive means. Laboratory standardisation is not trivial and differences between laboratories, for example in wet to dry weight ratios, can mean that results from different laboratories may not be equivalent.

Table 3

Rationale, advantages and disadvantages of (LIC) determination by (MRI) and biopsy.

Myocardial iron estimation: T2* and other tools

The physical principles of iron measurement for the heart by MRI are the same as for the liver (see above), with the additional challenge of measuring a moving object – the myocardium. The T2* (or R2*) techniques have the advantage over T2 or R2 in that they require shorter acquisition times and can be achieved with a single breathhold (Kirk et al., 2010). The utility of myocardial T2* (mT2*) MRI was originally identified on the basis of shortened T2* values <20 ms in patients with decreased left ventricular ejection fraction (LVEF) (Anderson et al., 2001). More recently the relationship between biochemically measured myocardial iron concentration and myocardial T2* has been shown using post-mortem myocardial material (Carpenter et al., 2011). Here, mean myocardial iron causing severe heart failure in 10 patients at post-mortem was 5.98 mg/g dry wt (ranging from 3.2 to 9.5 mg/g); levels that in the liver would not be regarded as harmful. The relationship of myocardial iron concentration (MIC) to T2* is: MIC (mg/g dry wt) = 45 * (T2* ms)-1.22 (Kirk et al., 2009). This relationship is non-linear so small changes in mT2* at values <10 ms may indicate relatively large changes in MIC. The risk of developing heart failure increases with T2* values <10 ms, which are associated with a 160 fold increased risk of heart failure in the next 12 months (Kirk et al., 2009). This risk further increases progressively with T2* values <10 ms, so that the proportion of patients developing heart failure in the next 12 months at T2* of 8-10 ms, 6-8 ms and <6 ms was 18%, 31%, and 52% respectively. These risks were derived from patients whose chelation therapy and adherence was not reported, so this risk may be less in patients taking regular chelation. For example, in a recent prospective study in patients with severe myocardial iron loading (T2* values <10 ms), no patients developed heart failure over a 2 years period while taking deferasirox (DFX) and deferoxamine (DFO) combination therapy (Aydinok et al., 2013).

In centres where the T2* method has been validated, the T2* value may have predictive value in identifying patients at high risk of developing deterioration in LVEF, thus allowing targeted intensification of treatment before heart failure develops. Prompt identification of patients at risk by MRI, timely therapeutic intervention and improved chelation options have contributed to decreasing number of cardiac deaths (Thomas et al., 2010; Voskaridou et al., 2019). T2* monitoring has now been established and validated internationally (Kirk et al., 2010), and is now recommended as part of yearly monitoring of multi-transfused patients at risk of developing myocardial iron loading. However, it is very important that a given centre undertakes procedures to independently validate and calibrate measurements of the method adopted, otherwise inappropriate assessment of heart failure prognosis may result. Table 4 summarizes advantages and disadvantages of using T2* MRI for monitoring cardiac iron overload.

Table 4

MRI T2* method to assess myocardial iron.

Cardiac function

Sequential monitoring of LVEF has been shown to identify patients at high risk of developing clinical heart failure (Davis et al., 2004; Davis, O’Sullivan & Porter, 2001). When LVEF fell below reference values, there was a 35-fold increased risk of clinical heart failure and death, with a median interval to progression of 3.5 years, allowing time for intensification of chelation therapy. This approach required a reproducible method for determination of LVEF (such as MUGA (Multigated Acquisition) scan or MRI), while echocardiography was generally too operator-dependent for this purpose. Furthermore, there is a clear need to identify high risk patients before there is a decline in LVEF. Myocardial T2* by MRI can achieve this and has additional predictive value (see above). However, as only a subset of patients with T2* values between 10 and 20 ms, or even with T2* less than 10 ms have abnormal heart function, sequential measurement of LVEF can identify the subset of patients who have developed decompensation of LV function and are therefore at exceptionally high risk and require very intensive chelation therapy (see below).

Monitoring of other organ function and iron-mediated damage

The monitoring of organ function as a marker of damage from iron overload is discussed more fully in other chapters. In general, by the time diabetes, hypothyroidism, hypoparathyroidism or hypogonadotropic hypogonadism (HH) have been identified, irreversible damage has set in and the focus then becomes replacing hormones. These are late effects and the primary aim of chelation therapy is to prevent such damage. Iron overloaded patients should be monitored for evidence of HH (growth and sexual development and biochemical markers of HH), diabetes mellitus (yearly oral glucose tolerance test (OGTT)), hypothyroidism and hypoparathyroidism. There has been recent interest in using MRI as a way of identifying iron-mediated damage to the endocrine system. Early work in this area showed good correlation between MRI findings (loss of pituitary volume) and biochemical markers of pituitary damage (Chatterjee et al., 1998). With improved MRI imaging, other endocrine organs have also been evaluated (Wood, 2007). It is of interest, that there is generally a close correlation between iron deposition in the heart and deposition in endocrine tissues, such as those of the pituitary and pancreas (Noetzli et al., 2009; Au et al., 2008). This supports the notion of shared uptake mechanisms for NTBI in heart and endocrine systems and supports clinical observations of shared risks in cardiac and endocrine systems once iron begins to escape from the liver.

24h Urinary iron estimation

Measurement of the urinary iron excretion has been used in assessing the effect on iron excretion by deferoxamine (about half of total iron excreted in urine) (Pippard, Callender & Finch, 1982) or deferiprone (over 80% of iron excreted in urine), but is not useful in patients treated with deferasirox, as nearly all the iron is excreted in faeces. Urine iron has also been used to compare effects of combination and monotherapy regimes containing deferiprone (DFP) (Aydinok et al., 2012; Mourad et al., 2003). The inherent variability in daily iron excretion necessitates repeated determinations and this is not widely used in routine monitoring

Plasma non-transferrin bound iron and labile plasma iron

As plasma iron that is not bound to transferrin (NTBI), is considered to be the main route through which iron is distributed to liver and extrahepatic targets of iron-overloaded thalassaemia major patients, levels of NTBI might be expected to correlate with the risk of damage to these tissues. Assays may estimate NTBI directly using a chelation capture method followed by high performance liquid chromatography (HPLC) (Singh, Hider & Porter, 1990) or by colorimetric analysis (Gosriwatana et al., 1999) or indirectly by exploiting the impact of labile iron species to oxidised fluorochrome, such as in the labile plasma iron (LPI) assay (Zanninelli, Breuer & Cabantchik, 2009; Cabantchik et al., 2005). A potential advantage of the LPI assay is that it is better suited to measurements when iron chelators are present in the plasma (Zanninelli, Breuer & Cabantchik, 2009). Whilst some loose associations of NTBI (Piga et al., 2009) or LPI (Wood et al., 2011) with some markers of cardiac iron or response to chelation have been found by some investigators, thus far measurements have not been sufficiently strongly predictive of cardiac risk to be recommended for routine clinical practice. This is partly because NTBI and LPI are highly labile, rapidly returning or even rebounding (Porter et al., 1996) after an iron chelator has been cleared (Zanninelli, Breuer & Cabantchik, 2009). Although NTBI correlates loosely with iron overload, it is affected by other factors such as ineffective erythropoiesis, the phase of transfusion cycle, and the rate of blood transfusion (Porter et al., 2011) adding to the complexity of interpreting levels (Hod et al., 2010). It is also not clear which methods identify the iron species that are most strongly inked to myocardial iron uptake. Therefore, although the measurement of NTBI (or LPI) has proved a useful tool for evaluating how chelators interact with plasma iron pools, its value as a guide to routine treatment or prognosis has yet to be clearly demonstrated.

Box

Summary, Recommendations and Grade of Evidence.