- Email: [email protected]
A diagnostic approach to hyperferritinemia with a non-elevated transferrin saturation
Review
A diagnostic approach to hyperferritinemia with a non-elevated transferrin saturation Paul C. Adams1,⇑, James C. Barton2 1
Department of Medicine, University Hospital, University of Western Ontario, London, Ontario, Canada; 2Southern Iron Disorders Center and University of Alabama at Birmingham, Birmingham, AL, USA
Elevated serum ferritin concentrations are common in clinical practice. In this review, we provide an approach to interpreting the serum ferritin elevation in relationship to other clinical parameters including the patient history, transferrin saturation, serum concentrations of alanine, and aspartate aminotransferases (ALT, AST), testing for HFE mutations, liver imaging, liver biopsy, and liver iron concentration. We used observations from a large series of patients with hepatic iron overload documented by liver iron concentration measurement from two referral practices as a gold standard to guide the interpretation of the predictive values of non-invasive iron tests. Three case studies illustrate common problems in interpreting iron blood tests. Ó 2011 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved.
Introduction Elevated serum ferritin concentrations are common. The cause of an elevated ferritin concentration is usually increased ferritin synthesis (including acquired/genetic conditions with or without iron overload) or increased release of ferritin from damaged cells (Table 1). A clue to the cause of the elevation is often the accompanying clinical setting. For example, an isolated elevation of serum ferritin <1000 lg/L with a normal transferrin saturation (TS) in the context of daily alcohol consumption or obesity is a common presentation. Another group of patients with an isolated elevation of serum ferritin <1000 lg/L and normal TS values have hyporegenerative anemia untreated with transfusion (usually due to renal insufficiency, chronic disease, or marrow failure), malignancy, or chronic inflammation of diverse etiologies. In either situation, iron overload is usually not present and thus diagnosis and management of the disorders that contribute to hyperferritinemia is indicated. The diagnostic approach to elevated serum ferritin concentrations in patients who also have elevated serum concentrations of liver enzymes is more complicated. Genetic hemochromatosis is often considered when evaluating an elevated ferritin. Regardless of this, most patients
Keywords: Haemochromatosis; Iron overload. Received 10 December 2010; received in revised form 26 January 2011; accepted 4 February 2011 ⇑ Corresponding author. Address: Department of Medicine, University Hospital, 339 Windermere Rd., London, Ontario, Canada N6A 5A5. Fax: +1 519 663 3549. E-mail address: [email protected] (P.C. Adams).
without an elevated TS have hyperferritinemia related to inflammation, chronic alcohol consumption, cell damage, or metabolic abnormalities (e.g., obesity). Reference concentrations of serum ferritin and TS vary across laboratories due to differences in analytical techniques and reference populations; age and sex are also important determinants of these measures. A large, multi-ethnic, multi-racial screening study of iron overload in North Americans recruited in primary care settings defined that serum ferritin concentrations greater than 300 lg/L in men and greater than 200 lg/L in women were elevated [1]. By definition, TS values greater than 50% in men and greater than 45% in women were defined as elevated. Analysis of race/ethnicity groups within this large cohort revealed that mean SF and percentages of participants with elevated SF were significantly greater in blacks, Asians, and Pacific Islanders than in whites. Mean TS values were greatest in Asians and Pacific Islanders, intermediate in whites, and lowest in African Americans. This indicates that ‘‘elevated’’ serum ferritin or TS concentrations are probably not always the indicators of disease, although the biological basis and clinical significance of these race/ethnicity differences in iron phenotypes are incompletely understood. Using transferrin saturation to interpret elevated serum ferritin level Serum ferritin concentrations are often measured to investigate fatigue, possible liver disease, anemia, malignancy, or other conditions. Many clinicians interpret the serum ferritin concentration together with TS to determine if iron overload, iron deficiency, or subnormal iron mobilization for erythropoiesis is present. Thus, many clinicians rely on inferences made from TS measurements as an aid to the diagnosis of a variety of abnormalities [2,3]. In a large population screening study, we have demonstrated that there is significant biological variability in serum TS [4]. In 64,230 participants, a non-fasting TS >45% in women and >50% in men had a sensitivity of 75% and specificity of 95% in the detection of C282Y homozygotes. This variability decreases the reliability and specificity of TS as a test to clarify the diagnostic significance of elevated serum ferritin concentrations. Withinperson biological variability in iron test values has been reported previously; diurnal fluctuations in test values have been described primarily for serum iron measurements [4,5]. TS is a calculated value determined as the quotient of the serum iron level divided by one of the following: total iron-binding capacity; unsaturated
Journal of Hepatology 2011 vol. 55 j 453–458
Review Table 1. Mechanism of hyperferritinemia in selected disorders.
Increased apoferritin (or L ferritin) synthesis/secretion: chronic ethanol ingestion; malignancy (malignant histiocytosis; carcinomas of lung, breast, ovary, and kidney; lymphoma; liposarcoma); Gaucher disease; reactive histiocytosis; hereditary hyperferritinemia-cataract syndrome Increased ferritin release from injured cells: hepatic steatosis and steatohepatitis; chronic viral hepatitis; massive liver necrosis due to sepsis, acute hepatitis, or toxic injury; autoimmune disorders; acute and chronic infections; acute myocardial infarction; splenic infarct Increased ferritin synthesis due to iron overload: HFE and other types of hemochromatosis; heritable and acquired anemias associated with ineffective erythropoiesis; increased iron absorption from supplemental iron or ingestion of traditional beer (subSaharan African Natives); transfusion iron overload; parenteral iron overload; aceruloplasminemia
Relationship between transferrin saturation and liver iron concentration Elevated TS is an important prerequisite for iron loading of hepatocytes typical of hemochromatosis due to mutations in genes that encode HFE, hemojuvelin, transferrin receptor 2, and hepcidin. Elevated TS depends partly on increased iron absorption, because food iron deprivation causes TS to decrease [7] and supplemental iron increases TS. Increased release of iron from macrophages into plasma also contributes to elevated plasma iron and TS concentrations. Overall, the elevated TS that may be observed in relatively young persons suggests that iron absorption is increased. Acute infections, menses, recent blood donation may temporarily reduce TS concentrations in persons with excessive hepatocyte iron deposition [8]. Due to this variability, we recommend testing Caucasian patients with serum ferritin >1000 lg/L for the HFE C282Y mutation. We investigated the relationship between an elevated TS and liver iron concentration further. Using observations from two tertiary referral centers (London, Ontario and Birmingham, Alabama), we tabulated data from patients who underwent measurement of TS and liver iron concentration (Fig. 1). In the Alabama samples, liver iron concentrations were measured in patients with iron overload initially determined by liver iron 454
histochemical grading criteria. In the London, Ontario samples, liver iron was measured on a more diverse group, because many of the patients participated in a study of MRI and liver iron [9]. Patients from both locations were separated into two groups: HFE C282Y homozygotes and other cases. Diagnoses in these patients included hemochromatosis, non-HFE iron overload, juvenile hemochromatosis, alcoholic liver disease, fatty liver, hepatitis C, and hepatitis B. Because most patients do not undergo routine liver biopsy, this sample does not represent the general population. We analyzed these data using receiver operator characteristic (ROC) curves; we calculated positive likelihood ratios for predicting hepatic iron overload (LIC >40 lmol/g) (Table 2). From these results, we conclude that: (1) TS predicts hepatic iron overload better in C282Y homozygotes than in non-homozygotes; and (2) some patients with a normal TS and an elevated serum ferritin concentration have iron overload proven by liver biopsy. Assessment of patients with an elevated ferritin concentration and a non-elevated transferrin saturation In this setting, it is useful to measure serum concentrations of AST, ALT, alkaline phosphatase (AP), gamma-glutamyl transpeptidase (GGT), HBsAg, and anti-HCV; assess alcohol history and body mass index; and perform abdominal ultrasonography. If a patient has elevated serum concentrations of hepatic enzymes, the next question is whether the elevated serum ferritin level is the sequel of hepatocyte necrosis or, less likely, whether iron overload caused the liver damage [10,11]. Typical patients with hyporegenerative anemia have normal serum concentrations of hepatic enzymes, and anemia is
900
Liver iron concentration (µmol/g)
iron-binding capacity + serum iron concentration; or serum transferrin concentration multiplied by a constant. Liver disease sometimes causes low serum transferrin concentrations that may lead to increased TS without iron overload. The higher variability of TS than the serum iron level may be due to the fact that measurement of TS is a two-step, rather than a single-step, test. It has been reported that most HFE C282Y homozygotes have persistently elevated TS, and that false positive tests in non-homozygotes would likely return to normal on a second test [6]. This was the previous rationale for using two TS tests (first test random, second test fasting) for the presumptive diagnosis of hemochromatosis before proceeding to more specific diagnostic tests, including liver biopsy or DNA-based testing. Regardless, this and other studies failed to confirm that fasting iron tests had greater specificity for the diagnosis of HFE C282Y-linked hemochromatosis than iron tests drawn without regard to fasting [2,3]. In addition, variability of TS measurements in this study was similar in homozygotes and non-homozygotes [4]. The second fasting value was as likely to increase as to decrease, and regression to the mean is the most likely explanation for this phenomenon. Fasting as a condition for testing adds a level of complexity and inconvenience to a screening program. In addition, any biochemical test with such wide biological variation is unlikely to be an ideal screening test.
800 700 600 500 400 300 200 100 0
0
20
40
60
80
100
Transferrin saturation (%) Fig. 1. The relationship between transferrin saturation and hepatic iron overload in C282Y homozygotes (d) and non-homozygotes (s).
Journal of Hepatology 2011 vol. 55 j 453–458
JOURNAL OF HEPATOLOGY rare in persons with hemochromatosis due to HFE C282Y homozygosity who have not been treated with phlebotomy. Obesity or fatty liver detected by ultrasonography suggest the presence of steatohepatitis. Chronic alcohol consumption may be suspected based on the clinical history and an elevated level of GGT. In patients with viral hepatitis B or C, serum ferritin is often elevated but usually does not indicate iron overload. A clinical syndrome has been described in France associated with insulin resistance and features of the metabolic syndrome [12,13]. Patients with dysmetabolic syndrome usually have elevated serum ferritin concentrations and a normal TS. Liver biopsy in such patients usually demonstrates iron in Kupffer cells and steatosis; insulin resistance in such patients may decrease with phlebotomy therapy [14]. Obesity, diabetes mellitus, dyslipidemia, and hypertension are other common although not diagnostic features of this syndrome. Hereditary hyperferritinemia–cataract syndrome (HHCS) is caused by heterogeneous mutations in the iron-responsive element (IRE) of L-ferritin that reduce the binding affinity of iron-responsive proteins (IRPs) to IREs and thereby diminish the negative control of L-ferritin (but not H-ferritin) synthesis. This leads to the constitutive up-regulation of ferritin L-chain synthesis characteristic of HHCS. Serum ferritin concentrations are elevated and relatively constant; TS is normal. Patients with HHCS do not develop iron overload, although common HFE mutations, especially H63D, may occur in HHCS patients coincidentally. L-ferritin deposition in the ocular lens causes bilateral cataracts at an early age [15–17]. It may be possible to have an elevated ferritin on this basis without cataracts [18]. The concurrence of hyperferritinemia and cataract do not establish the diagnosis of HHCS [19]. Although routine genetic testing to verify HHCS is usually not feasible, demonstration of hyperferritinemia and cataract syndrome in two or more members of the same kinship is usually diagnostic [19]. Aceruloplasminemia is a rare genetic syndrome due to mutations in the CP gene that encodes ceruloplasmin. This condition is associated with normal TS, hyperferritinemia, and iron overload; retinal abnormalities; and prominent neurological problems. Anemia is uncommon. Aceruloplasminemia can be distinguished from Wilson disease by the absence of ceruloplasmin in the serum, by the absence of excess urinary copper, and by the absence of excess storage copper in the liver [20,21]. Ferroportin is an iron transport protein and the receptor for hepcidin. ‘‘Loss-of-function’’ mutations of the SLC40A1 gene that encodes the ferroportin result in an uncommon, iron overload disorder characterized by normal TS, elevated serum ferritin concentration, and hepatic iron overload. This condition is inherited as an autosomal dominant trait. Ferroportin normally occurs at the basolateral surfaces of enterocytes, macrophages, hepatocytes, and placental syncytiotrophoblasts. In patients with a pathogenic SLC40A1 mutation, ferroportin multimers consists of both normal Table 2. Receiver operating curve assessment of transferrin saturation to predict hepatic iron overload.
TS for iron overload
Sensitivity (%)
Specificity (%)
>55%* (n = 371)
81
73
>50% (n = 371)
85
69
>45 (n = 371)
89
61
>55% in C282Y homozygotes
92
40
(n = 209) ⁄
Iron overload is defined as a liver iron concentration >40 lmol/g dry weight.
and abnormal molecules, causing a dominant negative effect of the mutant SLC40A1 allele that prevents normal function of the wild-type (normal) ferroportin. Excessive iron is typically deposited in macrophages in the liver, spleen, bone marrow, and lymph nodes; decreased iron mobilization from macrophages decreases serum iron and TS values [22]. ‘‘Gain-of-function’’ SLC40A1 mutations, in contrast, result in iron overload that is usually associated with elevated TS, hyperferritinemia, and parenchymal iron deposition similar to that of HFE hemochromatosis. A trial of clinical observation There are many patients in whom observation and follow-up is more appropriate than an invasive investigation or a trial of phlebotomy. It is informative to observe whether the serum ferritin concentration is stable over time. Stable serum ferritin concentrations are typical of hyperferritinemia in some patients with HFE hemochromatosis, patients of certain race/ethnicity groups, and persons with HHCS. In patients with hepatic steatosis or steatohepatitis or alcohol-induced hyperferritinemia, serum ferritin concentrations typically fluctuate greatly. Progressively increasing serum ferritin concentrations suggest iron overload. During an interval of observation, there are opportunities to repeat the TS. Management of non-iron overload causes of hyperferritinemia Treatment of hyperferritinemia involves management of disorders that results in increased synthesis of apoferritin or cell injury. Interventions, such as alcohol abstinence, improved diabetes control, weight reduction, or lowering triglyceride concentrations with diet and medications may reduce serum ferritin concentrations during the observation period. Treatment of hyporegenerative anemia, causes of chronic inflammation, malignancy, viral hepatitis, or other non-iron overload causes of hyperferritinemia can reduce serum ferritin concentrations and substantiate provisional diagnoses of the cause(s) of hyperferritinemia. Treatment of hyperferritinemia in persons with HHCS is not indicated. MRI scanning to detect iron overload MRI scanning has been used to estimate liver iron concentrations indirectly. Overall, MRI technology for iron quantification is improving, but requires a dedicated radiology team and appropriate machine-specific calibration. MRI scanning is better at detecting and quantifying severe hepatic iron overload than mild iron overload usually associated with mild elevations of serum ferritin concentrations [23,24]. Trial of phlebotomy One diagnostic criterion for iron overload is a trial of phlebotomy; this method was typically used before other methods were devised. It was assumed that a patient had significant iron overload if they could tolerate, e.g., 20 weekly phlebotomies without developing anemia. It has been estimated that a 500 ml phlebotomy removes 0.25 g of mobilizable body iron. If the patient developed anemia after a few phlebotomies, it was assumed that the elevated serum ferritin value was due to inflammation or other non-iron overload cause, rather than to iron overload, because many normal persons have sufficient storage iron to support this level of phlebotomy. Quantitative phlebotomy has great utility
Journal of Hepatology 2011 vol. 55 j 453–458
455
Review today; its few adverse effects include venous access, phlebitis, vasovagal symptoms, anemia, and the costs and potential inconvenience of arranging treatments. The positive aspects of this approach to measuring body iron stores are that the patient feels that something is being done to correct their laboratory abnormalities. Many patients report positive physical and mental effects after phlebotomy. An alternative to arranging weekly phlebotomy therapy is to suggest that the patient be a voluntary blood donor every 2–3 months [25]. Patients with HHCS should not be treated with phlebotomy. Some patients with iron overload due to ‘‘loss-of-function’’ SLC40A1 mutations reconstitute circulating erythrocyte concentrations very slowly after phlebotomy treatments and should be managed with relatively low phlebotomy volumes and long intervals between treatments. The choice of observation, MRI, phlebotomy, or liver biopsy is based on clinical judgment, patient preferences, and local resources. There is not always a preferred pathway and different approaches may suit different patients.
Genetic testing HFE genotyping is often used to test patients with an elevated serum ferritin level. Some HFE C282Y homozygotes, and less commonly HFE H63D homozygotes and compound heterozygotes (HFE C282Y/H63D) develop iron overload secondary to a genetic abnormality in iron absorption and metabolism [1]. An increasing number of pathogenic mutations in other iron related genes (hemojuvelin, hepcidin, ferroportin, transferrin receptor 2) have
been identified, but there is no widely available commercial testing, if any, for these mutations [26]. In many of the rare cases of non-HFE iron overload, the genes most likely involved must be deduced from clinical features in the patient and his/her family members. Even after the probable causative gene is identified by these provisional methods, sequencing of the gene, rather than mutation-specific analysis, is often required. In many cases, there are no abnormal findings, because there are heritable forms of iron overload for which no explanatory gene or mutation has yet been reported. Furthermore, it is becoming increasingly difficult to differentiate between genetic polymorphisms and mutations with functional and clinically significant implications. For a clinician, it can be helpful to test siblings and children of index patients for iron overload (TS, serum ferritin level) in this setting to screen for an unusual iron overload disorder. The study of genomics is rapidly advancing and large scale high-throughput genetic sequencing is becoming widely available (exome and NextGen sequencing, total genome sequencing). The storage and interpretation of these huge volumes of data, their costs, and the genetic counseling that would be provided to the patient and their family as an accompaniment to such testing are a great future challenge, although whether such maneuvers will be feasible diagnostic tools for many patients is unproven.
Liver biopsy Percutaneous liver biopsy is a well-established method for evaluation of iron overload [27] and concomitant liver disease (Fig. 2).
Fig. 2. A liver biopsy in an 88 year C282Y homozygous woman with severe iron overload but without liver fibrosis. (PERLS Prussian Blue stain).
456
Journal of Hepatology 2011 vol. 55 j 453–458
JOURNAL OF HEPATOLOGY The safety of the procedure has been well-documented [28] and accepted by hepatologists, although other physicians and some patients are reluctant to accept or recommend the procedure. In patients who have isolated hyperferritinemia, liver biopsy is often used to exclude iron overload. Liver biopsy is also useful for staging steatohepatitis, alcoholic liver disease, or viral hepatitis, or to diagnose liver disease that cannot be identified with non-invasive methods. A liver biopsy is not routinely recommended to assess iron overload in patients with elevations of serum ferritin <1000 lg/L, because the risks of the procedure may outweigh the possibility that a diagnosis will be made that would change management recommendations [29,30]. Case study 1 A 39 year-old woman was referred to liver clinic because she had serum ferritin of 1822 lg/L. She is a native of the Philippine Islands with no known European ancestors. She denies any alcohol use and has no history of liver diseases or hepatitis. Initial investigations included AST 72 IU/L, ALT 123 IU/L, bilirubin 9.7 lmol/L, hemoglobin 138 g/L, MCV 92, platelets 314 103/ lL, INR 1.0, albumin 45 g/L, HBsAg negative, anti-HCV negative, TS 47%, and repeat serum ferritin 2913 lg/L. She has a body mass index of 32 kg/m2 and an abdominal ultrasound suggesting fatty infiltration of the liver without splenomegaly or ascites. HFE mutation analyses did not detect C282Y or H63D. Clinical commentary Serum ferritin concentration >1000 lg/L is an indication for further investigations because this level of ferritin has been associated with increased risk of hepatic fibrosis. HFE-linked hemochromatosis is an unlikely diagnosis in a Filipino, although European colonization could account for HFE C282Y in that geographic region. A clinical diagnosis of steatohepatitis seems likely, although the ferritin elevation is more elevated than usually expected. Although the TS is within the reference range, with what degree of certainty does this exclude iron overload? Options to consider at this point include: liver biopsy with hepatic iron concentration; MRI imaging with indirect estimate of liver iron concentration; trial of phlebotomy; and further observation and monitoring. The patient has already had a period of observation and the serum ferritin level is rising. Thus, we would not consider this option. A trial of phlebotomy is commonly done in community practice. If the serum ferritin elevation is not due to iron overload, the patient will not tolerate the treatment, and will probably develop anemia after 3–4 units of blood are removed by phlebotomy. MRI scanning to estimate liver iron concentration is feasible and yields good estimates of liver iron content in centers with a special interest in this technique. This would be a reasonable option, if available. In this case, a liver biopsy was performed; the specimen revealed severe steatohepatitis without iron overload or fibrosis. Liver iron concentration was 17 lmol/g (reference 0–40 lmol/g). This case illustrates that serum ferritin elevation may be extreme in persons with severe liver inflammation (non-alcoholic steatohepatitis = NASH). Her normal TS may have been a clue to the absence of iron overload, but there can be marked biological variability in TS values. In patients without HHCS, serum ferritin can be composed of different ferritin moieties including glycosylated ferritin, a protein secreted by macrophages, and/or tissue ferritins or ferritin light- or heavy-chains released from
damaged cells. Isoferritin analysis is not a practical diagnostic tool to differentiate serum ferritin of iron overload from that of inflammation or HHCS. The ferritin iron concentration has also been studied as a diagnostic test, but was tedious and never reached widespread clinical use despite initially promising results. The correlations of concomitant markers of inflammation such as C-reactive protein (CRP) or erythrocyte sedimentation rate (ESR) with serum ferritin of inflammation have been inconsistent. Nonetheless, markedly elevated values of CRP or ESR suggest that hyperferritinemia is probably due to inflammation or infection. Phlebotomy has been studied as a treatment for steatohepatitis [14,31] and can reduce serum ferritin concentrations; a beneficial effect on liver fibrosis or long-term outcomes has not been clearly established. In this context, persons with HFE C282Y may benefit more from phlebotomy than those without this common HFE mutation [32]. Case study 2 A 55 year-old Scottish woman complained of fatigue. She was told by a friend that fatigue is a symptom of iron overload in some patients, and she asked her physician to perform iron tests. Her serum ferritin level was 4400 lg/L. She has no known liver diseases or family history of liver disease. She lives with her identical twin sister. She consumes 2–3 alcoholic drinks daily. Two weeks after the initial test, her serum ferritin level was 2307 lg/L; TS was 53%; AST, ALT, and GGT concentrations were 223, 172, and 582 IU/L, respectively; and HBsAg and anti-HCV were negative. Abdominal ultrasound was normal. Her twin sister volunteered for blood testing that showed a serum ferritin level of 217 lg/L and TS 29%. Testing for the C282Y and H63D mutations of the HFE gene was normal in both twins. Clinical commentary We have screened approximately 30,000 patients for hemochromatosis in our community, and it interests us that a member of the general public suggests iron overload as a cause of fatigue rather than iron deficiency. In this unusual case, the demonstration that an identical twin did not have an elevated serum ferritin concentration was a helpful diagnostic tool to exclude HFE hemochromatosis or other types of iron overload in the index twin, although there were other risk factors (alcohol). Elevations of AST > ALT are more typical of alcoholic liver disease than hemochromatosis. Elevated GGT values are also typical of chronic alcohol consumption. The TS in this index twin was within the reference range, whereas it is expected that a patient with iron overload and serum ferritin of 4400 lg/L would have a markedly elevated value of TS. A liver biopsy was performed in the index twin that demonstrated severe alcoholic hepatitis without visible iron staining. Liver iron concentration was normal (17 lmol/g; reference 0–40 lmol/g). The patient admitted that she was depressed and had abused alcohol daily. She stopped drinking after the initial serum ferritin of 4400 lmol/g was detected; after two weeks of alcohol abstinence, serum ferritin decreased to 2307 lg/L. After 4 months of abstinence, the serum ferritin level decreased to 257 lg/L. Case 3 An asymptomatic 55 year-old Caucasian physician was referred for evaluation of serum ferritin 660 lg/L. He was obese and drank two glasses of wine daily. TS was 40%, and serum concentration
Journal of Hepatology 2011 vol. 55 j 453–458
457
Review of ALT and AST were 58 and 37 IU/L, respectively. Testing for HFE C282Y and H63D mutations showed that he is a H63D heterozygote. The patient was told not to worry that the serum ferritin level was likely due to inflammation due to fatty liver or to stimulation of ferritin synthesis induced by regular alcohol intake. He decided to become a regular blood donor; after donation of six units, his serum ferritin was 77 lg/L. This is a common clinical scenario, and mild elevations of serum ferritin (300–1000 lg/L) are frequent in clinical practice. The usual risk factors are daily alcohol consumption and fatty liver. The cause of the elevation is often not clearly established, or may be multifactorial, and invasive testing with liver biopsy is not recommended in such patients. It is paradoxical that if the serum ferritin level is elevated due to inflammation that it decreases with phlebotomy, although phlebotomy may have anti-inflammatory effects, including the removal of relatively small amounts of iron that participate in causing tissue injury.
Key Points Approximately 90% of patients with hyperferritinemia encountered in routine medical practice do not have iron overload Serum ferritin levels greater than the upper reference limit are often encountered in well persons of subSaharan Native African descent or Asian/Pacific Islander descent Mutation analysis to detect common HFE mutations is commercially available, and such testing is most useful in evaluating European whites with hyperferritinemia. Hyperferritinemia remains unexplained after evaluation in many patients
Conflict of interest The authors declared that they do not have anything to disclose regarding funding or conflict of interest with respect to this manuscript. References [1] Adams PC, Reboussin DM, Barton JC, McLaren CE, Eckfeldt JH, McLaren GD, et al. Hemochromatosis and iron-overload screening in a racially diverse population. N Engl J Med 2005;352:1769–1778. [2] McGrath J, Deugnier Y, Moirand R, Jouanolle A, Chakrabarti S, Adams P. A nomogram to predict C282Y hemochromatosis. J Lab Clin Med 2002;140:6–8. [3] Barton J, Acton R, Dawkins F, Adams P, Lovato L, Leiendecker-Foster C, et al. Initial screening transferrin saturation values, serum ferritin concentrations, and HFE genotypes in whites and blacks in the hemochromatosis and iron overload screening study. Genet Test 2005;9:231–241. [4] Adams PC, Reboussin DM, Press RD, Barton JC, Acton RT, Moses GC, et al. Biological variability of transferrin saturation and unsaturated iron binding capacity. Am J Med 2007;120:999e1–999e7. [5] Adams P, Zaccaro D, Moses G, Eckfeldt J, Leiendecker-Foster C, McLaren C, et al. Comparison of the unsaturated iron binding capacity with transferrin saturation as a screening test to detect C282Y homozygotes for hemochromatosis in 101, 168 participants in the HEIRS study. Clin Chem 2005;51:1048–1051. [6] Edwards CQ, Griffen LM, Kaplan J, Kushner JP. Twenty-four hour variation of transferrin saturation in treated and untreated haemochromatosis homozygotes. J Intern Med 1989;226:373–379.
458
[7] Milder MS, Cook JD, Finch CA. Influence of food iron absorption on the plasma iron level in idiopathic hemochromatosis. Acta Haematol 1978;60:65–75. [8] Barton JC, Bertoli LF, Janich MR, Arthur MW, Alford TJ. Normal transferrin saturation in hemochromatosis. Hosp Pract 1991;26:46–48. [9] Beaton M, Adams PC. Transferrin saturation as a predictor of hepatic iron overload. Liver Int 2011;31:272–273. [10] Adams PC, Passmore L, Chakrabarti S, Reboussin D, Acton R, Barton J, et al. Liver diseases in the hemochromatosis and iron overload screening study. Clin Gastroenterol Hepatol 2006;4:918–923. [11] Wong K, Adams PC. The diversity of liver diseases associated with an elevated serum ferritin. Can J Gastroenterol 2007;20:467–470. [12] Mendler MH, Turlin B, Moirand R, Jouanolle AM, Sapey T, Guyader D, et al. Insulin resistance-associated hepatic iron overload. Gastroenterology 1999;117:1155–1163. [13] Moirand R, Mortaji A, Loreal O, Paillard F, Brissot P, Deugnier Y. A new syndrome of liver iron overload with normal transferrin saturation. Lancet 1997;349:95–97. [14] Valenti L, Moscatiello S, Vanni E, Fracanzani AL, Bugianesi E, Fargion S, et al. Venesection for non-alcoholic fatty liver disease unresponsive to lifestyle counselling – a propensity score-adjusted observational study. QJM 2010;104:141–149. [15] Cazzola M, Bergamaschi G, Tonon L, Arbustini E, Grasso M, Vercesi E, et al. Hereditary hyperferritinemia–cataract syndrome: relationship between phenotypes and specific mutations in the iron-responsive element of ferritin light-chain mRNA. Blood 1997;90:814–821. [16] Girelli D, Corrocher R, Biscelglia L, Olivieri O, Zelante L, Panozzo G, et al. Hereditary hyperferritinemia–cataract syndrome caused by a 29-base pair deletion in the iron responsive element of ferritin L-subunit gene. Blood 1997;90:2084–2088. [17] Wong K, Barbin Y, Chakrabarti S, Adams PC. A point mutation in the ironresponsive element of the L-ferritin in a family with hereditary hyperferritinemia–cataract syndrome. Can J Gastroenterol 2005;19:253–255. [18] Kannengiesser C, Jouanolle AM, Hetet G, Mosser A, Muzeau F, Henry D, et al. A new missense mutation in the L-ferritin coding sequence associated with elevated levels of glycosylated ferritin in serum and absence of iron overload. Haematologica 2009;94:335–339. [19] Hetet G, Devaux I, Soufir N, Grandchamp B, Beaumont C. Molecular analyses of patients with hyperferritinemia and normal serum iron values reveal both L ferritin IRE and 3 new ferroportin (slc11A3) mutations. Blood 2003;102:1904–1910. [20] Xu X, Pin S, Gathinji M, Fuchs R, Harris Z. Aceruloplasminemia: an inherited neurodegenerative disease with impairment of iron homeostasis. Ann NY Acad Sci 2004;1012:299–305. [21] Hellman N, Schaefer M, Gehrke S, Stegen P, Hoffman W, Gitlin J, et al. Hepatic iron overload in aceruloplasminemia. Gut 2003;47:858–860. [22] Pietrangelo A, Montosi G, Totaro A, Garuti C, Conte D, Cassanelli S, et al. Hereditary hemochromatosis in adults without pathogenic mutations in the hemochromatosis gene. N Engl J Med 1999;341:725–732. [23] St. Pierre T, Clark P, Chua-anusorn W, Fleming A, Jeffrey G, Olynyk J, et al. Noninvasive measurement and imaging of liver iron concentrations using proton magnetic resonance. Blood 2005;105:855–861. [24] Gandon Y, Olivie D, Guyader D, Aube C, Oberti F, Sebille V, et al. Non-invasive assessment of hepatic iron stores by MRI. Lancet 2004;363:357–360. [25] Adams PC, Barton JC. How I treat hemochromatosis. Blood 2010;116:317–325. [26] Pietrangelo A. Hereditary hemochromatosis: a new look at an old disease. N Eng J Med 2004;350:2383–2397. [27] Brissot P, Bourel M, Herry D, Verger J-P, Messner M, Beaumont C, et al. Assessment of liver iron content in 271 patients: a reevaluation of direct and indirect methods. Gastroenterology 1981;80:557–565. [28] McGill DB, Rakela J, Zinsmeister AR, Ott BJ. A 21-year experience with major hemorrhage after percutaneous liver biopsy. Gastroenterology 1990;99:1396–1400. [29] EASL clinical practice guidelines for HFE hemochromatosis. J Hepatol 2010;53:3–22. [30] Bacon B, Adams P, Kowdley K, BPowell L, Tavill A. Diagnosis and management of hemochromatosis: AASLD Practice Guidelines. Hepatology, in press. [31] Valenti L, Fracanzani AL, Dongiovanni P, Bugianesi E, Marchesini G, Manzini P, et al. Iron depletion by phlebotomy improves insulin resistance in patients with nonalcoholic fatty liver disease and hyperferritinemia: evidence from a case–control study. Am J Gastroenterol 2007;102:1251–1258. [32] George DK, Goldwurm S, MacDonald GA, Cowley LL, Walker NI, Ward PJ, et al. Increased hepatic iron concentration in nonalcoholic steatohepatitis is associated with increased fibrosis. Gastroenterology 1998;114:311–318.
Journal of Hepatology 2011 vol. 55 j 453–458
A diagnostic approach to hyperferritinemia with a non-elevated transferrin saturation Paul C. Adams1,⇑, James C. Barton2 1
Department of Medicine, University Hospital, University of Western Ontario, London, Ontario, Canada; 2Southern Iron Disorders Center and University of Alabama at Birmingham, Birmingham, AL, USA
Elevated serum ferritin concentrations are common in clinical practice. In this review, we provide an approach to interpreting the serum ferritin elevation in relationship to other clinical parameters including the patient history, transferrin saturation, serum concentrations of alanine, and aspartate aminotransferases (ALT, AST), testing for HFE mutations, liver imaging, liver biopsy, and liver iron concentration. We used observations from a large series of patients with hepatic iron overload documented by liver iron concentration measurement from two referral practices as a gold standard to guide the interpretation of the predictive values of non-invasive iron tests. Three case studies illustrate common problems in interpreting iron blood tests. Ó 2011 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved.
Introduction Elevated serum ferritin concentrations are common. The cause of an elevated ferritin concentration is usually increased ferritin synthesis (including acquired/genetic conditions with or without iron overload) or increased release of ferritin from damaged cells (Table 1). A clue to the cause of the elevation is often the accompanying clinical setting. For example, an isolated elevation of serum ferritin <1000 lg/L with a normal transferrin saturation (TS) in the context of daily alcohol consumption or obesity is a common presentation. Another group of patients with an isolated elevation of serum ferritin <1000 lg/L and normal TS values have hyporegenerative anemia untreated with transfusion (usually due to renal insufficiency, chronic disease, or marrow failure), malignancy, or chronic inflammation of diverse etiologies. In either situation, iron overload is usually not present and thus diagnosis and management of the disorders that contribute to hyperferritinemia is indicated. The diagnostic approach to elevated serum ferritin concentrations in patients who also have elevated serum concentrations of liver enzymes is more complicated. Genetic hemochromatosis is often considered when evaluating an elevated ferritin. Regardless of this, most patients
Keywords: Haemochromatosis; Iron overload. Received 10 December 2010; received in revised form 26 January 2011; accepted 4 February 2011 ⇑ Corresponding author. Address: Department of Medicine, University Hospital, 339 Windermere Rd., London, Ontario, Canada N6A 5A5. Fax: +1 519 663 3549. E-mail address: [email protected] (P.C. Adams).
without an elevated TS have hyperferritinemia related to inflammation, chronic alcohol consumption, cell damage, or metabolic abnormalities (e.g., obesity). Reference concentrations of serum ferritin and TS vary across laboratories due to differences in analytical techniques and reference populations; age and sex are also important determinants of these measures. A large, multi-ethnic, multi-racial screening study of iron overload in North Americans recruited in primary care settings defined that serum ferritin concentrations greater than 300 lg/L in men and greater than 200 lg/L in women were elevated [1]. By definition, TS values greater than 50% in men and greater than 45% in women were defined as elevated. Analysis of race/ethnicity groups within this large cohort revealed that mean SF and percentages of participants with elevated SF were significantly greater in blacks, Asians, and Pacific Islanders than in whites. Mean TS values were greatest in Asians and Pacific Islanders, intermediate in whites, and lowest in African Americans. This indicates that ‘‘elevated’’ serum ferritin or TS concentrations are probably not always the indicators of disease, although the biological basis and clinical significance of these race/ethnicity differences in iron phenotypes are incompletely understood. Using transferrin saturation to interpret elevated serum ferritin level Serum ferritin concentrations are often measured to investigate fatigue, possible liver disease, anemia, malignancy, or other conditions. Many clinicians interpret the serum ferritin concentration together with TS to determine if iron overload, iron deficiency, or subnormal iron mobilization for erythropoiesis is present. Thus, many clinicians rely on inferences made from TS measurements as an aid to the diagnosis of a variety of abnormalities [2,3]. In a large population screening study, we have demonstrated that there is significant biological variability in serum TS [4]. In 64,230 participants, a non-fasting TS >45% in women and >50% in men had a sensitivity of 75% and specificity of 95% in the detection of C282Y homozygotes. This variability decreases the reliability and specificity of TS as a test to clarify the diagnostic significance of elevated serum ferritin concentrations. Withinperson biological variability in iron test values has been reported previously; diurnal fluctuations in test values have been described primarily for serum iron measurements [4,5]. TS is a calculated value determined as the quotient of the serum iron level divided by one of the following: total iron-binding capacity; unsaturated
Journal of Hepatology 2011 vol. 55 j 453–458
Review Table 1. Mechanism of hyperferritinemia in selected disorders.
Increased apoferritin (or L ferritin) synthesis/secretion: chronic ethanol ingestion; malignancy (malignant histiocytosis; carcinomas of lung, breast, ovary, and kidney; lymphoma; liposarcoma); Gaucher disease; reactive histiocytosis; hereditary hyperferritinemia-cataract syndrome Increased ferritin release from injured cells: hepatic steatosis and steatohepatitis; chronic viral hepatitis; massive liver necrosis due to sepsis, acute hepatitis, or toxic injury; autoimmune disorders; acute and chronic infections; acute myocardial infarction; splenic infarct Increased ferritin synthesis due to iron overload: HFE and other types of hemochromatosis; heritable and acquired anemias associated with ineffective erythropoiesis; increased iron absorption from supplemental iron or ingestion of traditional beer (subSaharan African Natives); transfusion iron overload; parenteral iron overload; aceruloplasminemia
Relationship between transferrin saturation and liver iron concentration Elevated TS is an important prerequisite for iron loading of hepatocytes typical of hemochromatosis due to mutations in genes that encode HFE, hemojuvelin, transferrin receptor 2, and hepcidin. Elevated TS depends partly on increased iron absorption, because food iron deprivation causes TS to decrease [7] and supplemental iron increases TS. Increased release of iron from macrophages into plasma also contributes to elevated plasma iron and TS concentrations. Overall, the elevated TS that may be observed in relatively young persons suggests that iron absorption is increased. Acute infections, menses, recent blood donation may temporarily reduce TS concentrations in persons with excessive hepatocyte iron deposition [8]. Due to this variability, we recommend testing Caucasian patients with serum ferritin >1000 lg/L for the HFE C282Y mutation. We investigated the relationship between an elevated TS and liver iron concentration further. Using observations from two tertiary referral centers (London, Ontario and Birmingham, Alabama), we tabulated data from patients who underwent measurement of TS and liver iron concentration (Fig. 1). In the Alabama samples, liver iron concentrations were measured in patients with iron overload initially determined by liver iron 454
histochemical grading criteria. In the London, Ontario samples, liver iron was measured on a more diverse group, because many of the patients participated in a study of MRI and liver iron [9]. Patients from both locations were separated into two groups: HFE C282Y homozygotes and other cases. Diagnoses in these patients included hemochromatosis, non-HFE iron overload, juvenile hemochromatosis, alcoholic liver disease, fatty liver, hepatitis C, and hepatitis B. Because most patients do not undergo routine liver biopsy, this sample does not represent the general population. We analyzed these data using receiver operator characteristic (ROC) curves; we calculated positive likelihood ratios for predicting hepatic iron overload (LIC >40 lmol/g) (Table 2). From these results, we conclude that: (1) TS predicts hepatic iron overload better in C282Y homozygotes than in non-homozygotes; and (2) some patients with a normal TS and an elevated serum ferritin concentration have iron overload proven by liver biopsy. Assessment of patients with an elevated ferritin concentration and a non-elevated transferrin saturation In this setting, it is useful to measure serum concentrations of AST, ALT, alkaline phosphatase (AP), gamma-glutamyl transpeptidase (GGT), HBsAg, and anti-HCV; assess alcohol history and body mass index; and perform abdominal ultrasonography. If a patient has elevated serum concentrations of hepatic enzymes, the next question is whether the elevated serum ferritin level is the sequel of hepatocyte necrosis or, less likely, whether iron overload caused the liver damage [10,11]. Typical patients with hyporegenerative anemia have normal serum concentrations of hepatic enzymes, and anemia is
900
Liver iron concentration (µmol/g)
iron-binding capacity + serum iron concentration; or serum transferrin concentration multiplied by a constant. Liver disease sometimes causes low serum transferrin concentrations that may lead to increased TS without iron overload. The higher variability of TS than the serum iron level may be due to the fact that measurement of TS is a two-step, rather than a single-step, test. It has been reported that most HFE C282Y homozygotes have persistently elevated TS, and that false positive tests in non-homozygotes would likely return to normal on a second test [6]. This was the previous rationale for using two TS tests (first test random, second test fasting) for the presumptive diagnosis of hemochromatosis before proceeding to more specific diagnostic tests, including liver biopsy or DNA-based testing. Regardless, this and other studies failed to confirm that fasting iron tests had greater specificity for the diagnosis of HFE C282Y-linked hemochromatosis than iron tests drawn without regard to fasting [2,3]. In addition, variability of TS measurements in this study was similar in homozygotes and non-homozygotes [4]. The second fasting value was as likely to increase as to decrease, and regression to the mean is the most likely explanation for this phenomenon. Fasting as a condition for testing adds a level of complexity and inconvenience to a screening program. In addition, any biochemical test with such wide biological variation is unlikely to be an ideal screening test.
800 700 600 500 400 300 200 100 0
0
20
40
60
80
100
Transferrin saturation (%) Fig. 1. The relationship between transferrin saturation and hepatic iron overload in C282Y homozygotes (d) and non-homozygotes (s).
Journal of Hepatology 2011 vol. 55 j 453–458
JOURNAL OF HEPATOLOGY rare in persons with hemochromatosis due to HFE C282Y homozygosity who have not been treated with phlebotomy. Obesity or fatty liver detected by ultrasonography suggest the presence of steatohepatitis. Chronic alcohol consumption may be suspected based on the clinical history and an elevated level of GGT. In patients with viral hepatitis B or C, serum ferritin is often elevated but usually does not indicate iron overload. A clinical syndrome has been described in France associated with insulin resistance and features of the metabolic syndrome [12,13]. Patients with dysmetabolic syndrome usually have elevated serum ferritin concentrations and a normal TS. Liver biopsy in such patients usually demonstrates iron in Kupffer cells and steatosis; insulin resistance in such patients may decrease with phlebotomy therapy [14]. Obesity, diabetes mellitus, dyslipidemia, and hypertension are other common although not diagnostic features of this syndrome. Hereditary hyperferritinemia–cataract syndrome (HHCS) is caused by heterogeneous mutations in the iron-responsive element (IRE) of L-ferritin that reduce the binding affinity of iron-responsive proteins (IRPs) to IREs and thereby diminish the negative control of L-ferritin (but not H-ferritin) synthesis. This leads to the constitutive up-regulation of ferritin L-chain synthesis characteristic of HHCS. Serum ferritin concentrations are elevated and relatively constant; TS is normal. Patients with HHCS do not develop iron overload, although common HFE mutations, especially H63D, may occur in HHCS patients coincidentally. L-ferritin deposition in the ocular lens causes bilateral cataracts at an early age [15–17]. It may be possible to have an elevated ferritin on this basis without cataracts [18]. The concurrence of hyperferritinemia and cataract do not establish the diagnosis of HHCS [19]. Although routine genetic testing to verify HHCS is usually not feasible, demonstration of hyperferritinemia and cataract syndrome in two or more members of the same kinship is usually diagnostic [19]. Aceruloplasminemia is a rare genetic syndrome due to mutations in the CP gene that encodes ceruloplasmin. This condition is associated with normal TS, hyperferritinemia, and iron overload; retinal abnormalities; and prominent neurological problems. Anemia is uncommon. Aceruloplasminemia can be distinguished from Wilson disease by the absence of ceruloplasmin in the serum, by the absence of excess urinary copper, and by the absence of excess storage copper in the liver [20,21]. Ferroportin is an iron transport protein and the receptor for hepcidin. ‘‘Loss-of-function’’ mutations of the SLC40A1 gene that encodes the ferroportin result in an uncommon, iron overload disorder characterized by normal TS, elevated serum ferritin concentration, and hepatic iron overload. This condition is inherited as an autosomal dominant trait. Ferroportin normally occurs at the basolateral surfaces of enterocytes, macrophages, hepatocytes, and placental syncytiotrophoblasts. In patients with a pathogenic SLC40A1 mutation, ferroportin multimers consists of both normal Table 2. Receiver operating curve assessment of transferrin saturation to predict hepatic iron overload.
TS for iron overload
Sensitivity (%)
Specificity (%)
>55%* (n = 371)
81
73
>50% (n = 371)
85
69
>45 (n = 371)
89
61
>55% in C282Y homozygotes
92
40
(n = 209) ⁄
Iron overload is defined as a liver iron concentration >40 lmol/g dry weight.
and abnormal molecules, causing a dominant negative effect of the mutant SLC40A1 allele that prevents normal function of the wild-type (normal) ferroportin. Excessive iron is typically deposited in macrophages in the liver, spleen, bone marrow, and lymph nodes; decreased iron mobilization from macrophages decreases serum iron and TS values [22]. ‘‘Gain-of-function’’ SLC40A1 mutations, in contrast, result in iron overload that is usually associated with elevated TS, hyperferritinemia, and parenchymal iron deposition similar to that of HFE hemochromatosis. A trial of clinical observation There are many patients in whom observation and follow-up is more appropriate than an invasive investigation or a trial of phlebotomy. It is informative to observe whether the serum ferritin concentration is stable over time. Stable serum ferritin concentrations are typical of hyperferritinemia in some patients with HFE hemochromatosis, patients of certain race/ethnicity groups, and persons with HHCS. In patients with hepatic steatosis or steatohepatitis or alcohol-induced hyperferritinemia, serum ferritin concentrations typically fluctuate greatly. Progressively increasing serum ferritin concentrations suggest iron overload. During an interval of observation, there are opportunities to repeat the TS. Management of non-iron overload causes of hyperferritinemia Treatment of hyperferritinemia involves management of disorders that results in increased synthesis of apoferritin or cell injury. Interventions, such as alcohol abstinence, improved diabetes control, weight reduction, or lowering triglyceride concentrations with diet and medications may reduce serum ferritin concentrations during the observation period. Treatment of hyporegenerative anemia, causes of chronic inflammation, malignancy, viral hepatitis, or other non-iron overload causes of hyperferritinemia can reduce serum ferritin concentrations and substantiate provisional diagnoses of the cause(s) of hyperferritinemia. Treatment of hyperferritinemia in persons with HHCS is not indicated. MRI scanning to detect iron overload MRI scanning has been used to estimate liver iron concentrations indirectly. Overall, MRI technology for iron quantification is improving, but requires a dedicated radiology team and appropriate machine-specific calibration. MRI scanning is better at detecting and quantifying severe hepatic iron overload than mild iron overload usually associated with mild elevations of serum ferritin concentrations [23,24]. Trial of phlebotomy One diagnostic criterion for iron overload is a trial of phlebotomy; this method was typically used before other methods were devised. It was assumed that a patient had significant iron overload if they could tolerate, e.g., 20 weekly phlebotomies without developing anemia. It has been estimated that a 500 ml phlebotomy removes 0.25 g of mobilizable body iron. If the patient developed anemia after a few phlebotomies, it was assumed that the elevated serum ferritin value was due to inflammation or other non-iron overload cause, rather than to iron overload, because many normal persons have sufficient storage iron to support this level of phlebotomy. Quantitative phlebotomy has great utility
Journal of Hepatology 2011 vol. 55 j 453–458
455
Review today; its few adverse effects include venous access, phlebitis, vasovagal symptoms, anemia, and the costs and potential inconvenience of arranging treatments. The positive aspects of this approach to measuring body iron stores are that the patient feels that something is being done to correct their laboratory abnormalities. Many patients report positive physical and mental effects after phlebotomy. An alternative to arranging weekly phlebotomy therapy is to suggest that the patient be a voluntary blood donor every 2–3 months [25]. Patients with HHCS should not be treated with phlebotomy. Some patients with iron overload due to ‘‘loss-of-function’’ SLC40A1 mutations reconstitute circulating erythrocyte concentrations very slowly after phlebotomy treatments and should be managed with relatively low phlebotomy volumes and long intervals between treatments. The choice of observation, MRI, phlebotomy, or liver biopsy is based on clinical judgment, patient preferences, and local resources. There is not always a preferred pathway and different approaches may suit different patients.
Genetic testing HFE genotyping is often used to test patients with an elevated serum ferritin level. Some HFE C282Y homozygotes, and less commonly HFE H63D homozygotes and compound heterozygotes (HFE C282Y/H63D) develop iron overload secondary to a genetic abnormality in iron absorption and metabolism [1]. An increasing number of pathogenic mutations in other iron related genes (hemojuvelin, hepcidin, ferroportin, transferrin receptor 2) have
been identified, but there is no widely available commercial testing, if any, for these mutations [26]. In many of the rare cases of non-HFE iron overload, the genes most likely involved must be deduced from clinical features in the patient and his/her family members. Even after the probable causative gene is identified by these provisional methods, sequencing of the gene, rather than mutation-specific analysis, is often required. In many cases, there are no abnormal findings, because there are heritable forms of iron overload for which no explanatory gene or mutation has yet been reported. Furthermore, it is becoming increasingly difficult to differentiate between genetic polymorphisms and mutations with functional and clinically significant implications. For a clinician, it can be helpful to test siblings and children of index patients for iron overload (TS, serum ferritin level) in this setting to screen for an unusual iron overload disorder. The study of genomics is rapidly advancing and large scale high-throughput genetic sequencing is becoming widely available (exome and NextGen sequencing, total genome sequencing). The storage and interpretation of these huge volumes of data, their costs, and the genetic counseling that would be provided to the patient and their family as an accompaniment to such testing are a great future challenge, although whether such maneuvers will be feasible diagnostic tools for many patients is unproven.
Liver biopsy Percutaneous liver biopsy is a well-established method for evaluation of iron overload [27] and concomitant liver disease (Fig. 2).
Fig. 2. A liver biopsy in an 88 year C282Y homozygous woman with severe iron overload but without liver fibrosis. (PERLS Prussian Blue stain).
456
Journal of Hepatology 2011 vol. 55 j 453–458
JOURNAL OF HEPATOLOGY The safety of the procedure has been well-documented [28] and accepted by hepatologists, although other physicians and some patients are reluctant to accept or recommend the procedure. In patients who have isolated hyperferritinemia, liver biopsy is often used to exclude iron overload. Liver biopsy is also useful for staging steatohepatitis, alcoholic liver disease, or viral hepatitis, or to diagnose liver disease that cannot be identified with non-invasive methods. A liver biopsy is not routinely recommended to assess iron overload in patients with elevations of serum ferritin <1000 lg/L, because the risks of the procedure may outweigh the possibility that a diagnosis will be made that would change management recommendations [29,30]. Case study 1 A 39 year-old woman was referred to liver clinic because she had serum ferritin of 1822 lg/L. She is a native of the Philippine Islands with no known European ancestors. She denies any alcohol use and has no history of liver diseases or hepatitis. Initial investigations included AST 72 IU/L, ALT 123 IU/L, bilirubin 9.7 lmol/L, hemoglobin 138 g/L, MCV 92, platelets 314 103/ lL, INR 1.0, albumin 45 g/L, HBsAg negative, anti-HCV negative, TS 47%, and repeat serum ferritin 2913 lg/L. She has a body mass index of 32 kg/m2 and an abdominal ultrasound suggesting fatty infiltration of the liver without splenomegaly or ascites. HFE mutation analyses did not detect C282Y or H63D. Clinical commentary Serum ferritin concentration >1000 lg/L is an indication for further investigations because this level of ferritin has been associated with increased risk of hepatic fibrosis. HFE-linked hemochromatosis is an unlikely diagnosis in a Filipino, although European colonization could account for HFE C282Y in that geographic region. A clinical diagnosis of steatohepatitis seems likely, although the ferritin elevation is more elevated than usually expected. Although the TS is within the reference range, with what degree of certainty does this exclude iron overload? Options to consider at this point include: liver biopsy with hepatic iron concentration; MRI imaging with indirect estimate of liver iron concentration; trial of phlebotomy; and further observation and monitoring. The patient has already had a period of observation and the serum ferritin level is rising. Thus, we would not consider this option. A trial of phlebotomy is commonly done in community practice. If the serum ferritin elevation is not due to iron overload, the patient will not tolerate the treatment, and will probably develop anemia after 3–4 units of blood are removed by phlebotomy. MRI scanning to estimate liver iron concentration is feasible and yields good estimates of liver iron content in centers with a special interest in this technique. This would be a reasonable option, if available. In this case, a liver biopsy was performed; the specimen revealed severe steatohepatitis without iron overload or fibrosis. Liver iron concentration was 17 lmol/g (reference 0–40 lmol/g). This case illustrates that serum ferritin elevation may be extreme in persons with severe liver inflammation (non-alcoholic steatohepatitis = NASH). Her normal TS may have been a clue to the absence of iron overload, but there can be marked biological variability in TS values. In patients without HHCS, serum ferritin can be composed of different ferritin moieties including glycosylated ferritin, a protein secreted by macrophages, and/or tissue ferritins or ferritin light- or heavy-chains released from
damaged cells. Isoferritin analysis is not a practical diagnostic tool to differentiate serum ferritin of iron overload from that of inflammation or HHCS. The ferritin iron concentration has also been studied as a diagnostic test, but was tedious and never reached widespread clinical use despite initially promising results. The correlations of concomitant markers of inflammation such as C-reactive protein (CRP) or erythrocyte sedimentation rate (ESR) with serum ferritin of inflammation have been inconsistent. Nonetheless, markedly elevated values of CRP or ESR suggest that hyperferritinemia is probably due to inflammation or infection. Phlebotomy has been studied as a treatment for steatohepatitis [14,31] and can reduce serum ferritin concentrations; a beneficial effect on liver fibrosis or long-term outcomes has not been clearly established. In this context, persons with HFE C282Y may benefit more from phlebotomy than those without this common HFE mutation [32]. Case study 2 A 55 year-old Scottish woman complained of fatigue. She was told by a friend that fatigue is a symptom of iron overload in some patients, and she asked her physician to perform iron tests. Her serum ferritin level was 4400 lg/L. She has no known liver diseases or family history of liver disease. She lives with her identical twin sister. She consumes 2–3 alcoholic drinks daily. Two weeks after the initial test, her serum ferritin level was 2307 lg/L; TS was 53%; AST, ALT, and GGT concentrations were 223, 172, and 582 IU/L, respectively; and HBsAg and anti-HCV were negative. Abdominal ultrasound was normal. Her twin sister volunteered for blood testing that showed a serum ferritin level of 217 lg/L and TS 29%. Testing for the C282Y and H63D mutations of the HFE gene was normal in both twins. Clinical commentary We have screened approximately 30,000 patients for hemochromatosis in our community, and it interests us that a member of the general public suggests iron overload as a cause of fatigue rather than iron deficiency. In this unusual case, the demonstration that an identical twin did not have an elevated serum ferritin concentration was a helpful diagnostic tool to exclude HFE hemochromatosis or other types of iron overload in the index twin, although there were other risk factors (alcohol). Elevations of AST > ALT are more typical of alcoholic liver disease than hemochromatosis. Elevated GGT values are also typical of chronic alcohol consumption. The TS in this index twin was within the reference range, whereas it is expected that a patient with iron overload and serum ferritin of 4400 lg/L would have a markedly elevated value of TS. A liver biopsy was performed in the index twin that demonstrated severe alcoholic hepatitis without visible iron staining. Liver iron concentration was normal (17 lmol/g; reference 0–40 lmol/g). The patient admitted that she was depressed and had abused alcohol daily. She stopped drinking after the initial serum ferritin of 4400 lmol/g was detected; after two weeks of alcohol abstinence, serum ferritin decreased to 2307 lg/L. After 4 months of abstinence, the serum ferritin level decreased to 257 lg/L. Case 3 An asymptomatic 55 year-old Caucasian physician was referred for evaluation of serum ferritin 660 lg/L. He was obese and drank two glasses of wine daily. TS was 40%, and serum concentration
Journal of Hepatology 2011 vol. 55 j 453–458
457
Review of ALT and AST were 58 and 37 IU/L, respectively. Testing for HFE C282Y and H63D mutations showed that he is a H63D heterozygote. The patient was told not to worry that the serum ferritin level was likely due to inflammation due to fatty liver or to stimulation of ferritin synthesis induced by regular alcohol intake. He decided to become a regular blood donor; after donation of six units, his serum ferritin was 77 lg/L. This is a common clinical scenario, and mild elevations of serum ferritin (300–1000 lg/L) are frequent in clinical practice. The usual risk factors are daily alcohol consumption and fatty liver. The cause of the elevation is often not clearly established, or may be multifactorial, and invasive testing with liver biopsy is not recommended in such patients. It is paradoxical that if the serum ferritin level is elevated due to inflammation that it decreases with phlebotomy, although phlebotomy may have anti-inflammatory effects, including the removal of relatively small amounts of iron that participate in causing tissue injury.
Key Points Approximately 90% of patients with hyperferritinemia encountered in routine medical practice do not have iron overload Serum ferritin levels greater than the upper reference limit are often encountered in well persons of subSaharan Native African descent or Asian/Pacific Islander descent Mutation analysis to detect common HFE mutations is commercially available, and such testing is most useful in evaluating European whites with hyperferritinemia. Hyperferritinemia remains unexplained after evaluation in many patients
Conflict of interest The authors declared that they do not have anything to disclose regarding funding or conflict of interest with respect to this manuscript. References [1] Adams PC, Reboussin DM, Barton JC, McLaren CE, Eckfeldt JH, McLaren GD, et al. Hemochromatosis and iron-overload screening in a racially diverse population. N Engl J Med 2005;352:1769–1778. [2] McGrath J, Deugnier Y, Moirand R, Jouanolle A, Chakrabarti S, Adams P. A nomogram to predict C282Y hemochromatosis. J Lab Clin Med 2002;140:6–8. [3] Barton J, Acton R, Dawkins F, Adams P, Lovato L, Leiendecker-Foster C, et al. Initial screening transferrin saturation values, serum ferritin concentrations, and HFE genotypes in whites and blacks in the hemochromatosis and iron overload screening study. Genet Test 2005;9:231–241. [4] Adams PC, Reboussin DM, Press RD, Barton JC, Acton RT, Moses GC, et al. Biological variability of transferrin saturation and unsaturated iron binding capacity. Am J Med 2007;120:999e1–999e7. [5] Adams P, Zaccaro D, Moses G, Eckfeldt J, Leiendecker-Foster C, McLaren C, et al. Comparison of the unsaturated iron binding capacity with transferrin saturation as a screening test to detect C282Y homozygotes for hemochromatosis in 101, 168 participants in the HEIRS study. Clin Chem 2005;51:1048–1051. [6] Edwards CQ, Griffen LM, Kaplan J, Kushner JP. Twenty-four hour variation of transferrin saturation in treated and untreated haemochromatosis homozygotes. J Intern Med 1989;226:373–379.
458
[7] Milder MS, Cook JD, Finch CA. Influence of food iron absorption on the plasma iron level in idiopathic hemochromatosis. Acta Haematol 1978;60:65–75. [8] Barton JC, Bertoli LF, Janich MR, Arthur MW, Alford TJ. Normal transferrin saturation in hemochromatosis. Hosp Pract 1991;26:46–48. [9] Beaton M, Adams PC. Transferrin saturation as a predictor of hepatic iron overload. Liver Int 2011;31:272–273. [10] Adams PC, Passmore L, Chakrabarti S, Reboussin D, Acton R, Barton J, et al. Liver diseases in the hemochromatosis and iron overload screening study. Clin Gastroenterol Hepatol 2006;4:918–923. [11] Wong K, Adams PC. The diversity of liver diseases associated with an elevated serum ferritin. Can J Gastroenterol 2007;20:467–470. [12] Mendler MH, Turlin B, Moirand R, Jouanolle AM, Sapey T, Guyader D, et al. Insulin resistance-associated hepatic iron overload. Gastroenterology 1999;117:1155–1163. [13] Moirand R, Mortaji A, Loreal O, Paillard F, Brissot P, Deugnier Y. A new syndrome of liver iron overload with normal transferrin saturation. Lancet 1997;349:95–97. [14] Valenti L, Moscatiello S, Vanni E, Fracanzani AL, Bugianesi E, Fargion S, et al. Venesection for non-alcoholic fatty liver disease unresponsive to lifestyle counselling – a propensity score-adjusted observational study. QJM 2010;104:141–149. [15] Cazzola M, Bergamaschi G, Tonon L, Arbustini E, Grasso M, Vercesi E, et al. Hereditary hyperferritinemia–cataract syndrome: relationship between phenotypes and specific mutations in the iron-responsive element of ferritin light-chain mRNA. Blood 1997;90:814–821. [16] Girelli D, Corrocher R, Biscelglia L, Olivieri O, Zelante L, Panozzo G, et al. Hereditary hyperferritinemia–cataract syndrome caused by a 29-base pair deletion in the iron responsive element of ferritin L-subunit gene. Blood 1997;90:2084–2088. [17] Wong K, Barbin Y, Chakrabarti S, Adams PC. A point mutation in the ironresponsive element of the L-ferritin in a family with hereditary hyperferritinemia–cataract syndrome. Can J Gastroenterol 2005;19:253–255. [18] Kannengiesser C, Jouanolle AM, Hetet G, Mosser A, Muzeau F, Henry D, et al. A new missense mutation in the L-ferritin coding sequence associated with elevated levels of glycosylated ferritin in serum and absence of iron overload. Haematologica 2009;94:335–339. [19] Hetet G, Devaux I, Soufir N, Grandchamp B, Beaumont C. Molecular analyses of patients with hyperferritinemia and normal serum iron values reveal both L ferritin IRE and 3 new ferroportin (slc11A3) mutations. Blood 2003;102:1904–1910. [20] Xu X, Pin S, Gathinji M, Fuchs R, Harris Z. Aceruloplasminemia: an inherited neurodegenerative disease with impairment of iron homeostasis. Ann NY Acad Sci 2004;1012:299–305. [21] Hellman N, Schaefer M, Gehrke S, Stegen P, Hoffman W, Gitlin J, et al. Hepatic iron overload in aceruloplasminemia. Gut 2003;47:858–860. [22] Pietrangelo A, Montosi G, Totaro A, Garuti C, Conte D, Cassanelli S, et al. Hereditary hemochromatosis in adults without pathogenic mutations in the hemochromatosis gene. N Engl J Med 1999;341:725–732. [23] St. Pierre T, Clark P, Chua-anusorn W, Fleming A, Jeffrey G, Olynyk J, et al. Noninvasive measurement and imaging of liver iron concentrations using proton magnetic resonance. Blood 2005;105:855–861. [24] Gandon Y, Olivie D, Guyader D, Aube C, Oberti F, Sebille V, et al. Non-invasive assessment of hepatic iron stores by MRI. Lancet 2004;363:357–360. [25] Adams PC, Barton JC. How I treat hemochromatosis. Blood 2010;116:317–325. [26] Pietrangelo A. Hereditary hemochromatosis: a new look at an old disease. N Eng J Med 2004;350:2383–2397. [27] Brissot P, Bourel M, Herry D, Verger J-P, Messner M, Beaumont C, et al. Assessment of liver iron content in 271 patients: a reevaluation of direct and indirect methods. Gastroenterology 1981;80:557–565. [28] McGill DB, Rakela J, Zinsmeister AR, Ott BJ. A 21-year experience with major hemorrhage after percutaneous liver biopsy. Gastroenterology 1990;99:1396–1400. [29] EASL clinical practice guidelines for HFE hemochromatosis. J Hepatol 2010;53:3–22. [30] Bacon B, Adams P, Kowdley K, BPowell L, Tavill A. Diagnosis and management of hemochromatosis: AASLD Practice Guidelines. Hepatology, in press. [31] Valenti L, Fracanzani AL, Dongiovanni P, Bugianesi E, Marchesini G, Manzini P, et al. Iron depletion by phlebotomy improves insulin resistance in patients with nonalcoholic fatty liver disease and hyperferritinemia: evidence from a case–control study. Am J Gastroenterol 2007;102:1251–1258. [32] George DK, Goldwurm S, MacDonald GA, Cowley LL, Walker NI, Ward PJ, et al. Increased hepatic iron concentration in nonalcoholic steatohepatitis is associated with increased fibrosis. Gastroenterology 1998;114:311–318.
Journal of Hepatology 2011 vol. 55 j 453–458