Reduced serum ceruloplasmin levels in hereditary haemochromatosis


Gaetano Cairo, Istituto Patologia Generale, Università di Milano, Via Mangiagalli 31, 20133 Milano, Italy. E-mail:


As ceruloplasmin (Cp) seems to be involved in iron mobilization, serum Cp levels were measured in 35 patients with hereditary haemochromatosis (HH), 12 with acquired iron overload (AIO) and 36 healthy subjects. Cp was lower in HH patients than in controls (P < 0·001); no difference was found between untreated HH patients and those on a phlebotomy programme (P = 0·07) and between the HH patients carrying the normal and the mutated alleles of the HFE gene (P = 0·8). Cp levels in AIO subjects were significantly higher than in HH patients (P < 0·004) and similar to those of controls (P = 0·2). No differences in albumin, α1 acid glycoprotein and copper serum levels were observed in the three groups.

The recent identification of HFE as the gene responsible for hereditary haemochromatosis (HH), a disease of iron metabolism characterized by iron accumulation that can lead to severe damage of several organs, has provided a means for HH screening and diagnosis. However, the role of HFE in causing the derangement of iron absorption in HH remains to be elucidated (Adams et al, 2000). Furthermore, HH may be caused by defects affecting genes other than HFE (Pietrangelo et al, 1999; Camaschella et al, 2000).

HH is not only characterized by unregulated absorption, but also by a non-homogeneous distribution of iron deposits, with overloaded parenchymal cells and inappropriately low iron levels in reticuloendothelial (RE) cells, as reflected by the high activity of iron regulatory proteins in monocyte–macrophages (Cairo et al, 1997). These findings suggest defective iron retention in the macrophages of HH patients, but the mechanism(s) underlying iron release from the cell is/are still unclear.

The involvement of ceruloplasmin (Cp), a copper-containing protein mainly synthesized by liver cells, in iron mobilization (Askwith & Kaplan, 1998) prompted us to measure serum Cp levels in HH patients.

Patients and methods

Biochemical evaluation Serum levels of iron, copper, ferritin, aminotransferases, bilirubin, albumin, α1 acid glycoprotein, transferrin saturation and hepatitis markers were measured using standard methods. Serum Cp was measured using a turbidimetric technique after reaction with an anti-human Cp antibody. Hepatic iron stores were estimated semiquantitatively (Deugnier et al, 1992) and quantitatively using atomic absorption spectrophotometry (Pietrangelo et al, 1991).

Study population The study plan was approved by the pertinent Ethics Committee. HH was diagnosed using currently accepted criteria (Adams et al, 2000). Homozygosity for the C282Y mutation in the HFE gene, as determined by polymerase chain reaction (Piperno et al, 1998), was found in 25 patients (71%) and a wild-type HFE genotype was present in 10 patients (29%). Exclusion criteria included drug or alcohol abuse, oestrogen treatment, pregnancy, concomitant secondary iron overload, chronic hepatitis B virus (HBV) or hepatitis C virus (HCV) infection and neoplasia. All patients were in good and stable clinical condition and cirrhotic subjects were in Child's class A.

The acquired iron overload (AIO) group included cirrhotic patients in Child's class B or C with sporadic porphyria cutanea tarda (n = 4), major (n = 6) or intermediate (n = 1) β-thalassaemia and sideroblastic anaemia (n = 1) with chronic HCV-related hepatitis. No patient had neoplasia and no woman was pregnant or under oestrogen treatment. Controls were not pregnant or on current drug treatment.

Statistical analysis Comparison between different groups was performed using the Mann–Whitney test. All the analyses were two-sided, and a P-value of < 0·05 was considered statistically significant.


Serum iron, transferrin saturation and serum ferritin were higher (P < 0·001) in both HH and AIO patients than in controls (Table I). Cp levels (Fig 1) were lower (1·38 ± 0·24 versus 1·9 ± 0·52 μmol/l; P < 0·001) in the HH patients than in the controls. Thirteen patients (37%) presented with a Cp concentration below the lower value of the normal range. Cp concentrations of the AIO patients were higher than in the HH patients (1·74 ± 0·42 versus 1·38 ± 0·24 μmol/l; P < 0·004) and similar to the control values (1·74 ± 0·42 versus 1·9 ± 0·52 μmol/l; P = 0·2). Cp levels of patients on a phlebotomy programme were slightly higher than in untreated HH subjects, but the difference did not reach statistical significance (1·25 ± 0·25 versus 1·42 ± 0·22 μmol/l; P = 0·07). The values of both groups were different from those of controls (P < 0·001). Cp levels were similar in C282Y homozygous HH patients and in those without the mutation (1·39 ± 0·29 versus 1·37 ± 0·22 μmol/l; P = 0·8). Furthermore, no differences were found in the levels of serum albumin and α1 acid glycoprotein, indicators of general hepatic protein synthesis and acute phase reaction, respectively, as well as in serum copper levels (Table I).

Table I.  Haemoglobin (Hb), serum iron (SI), transferrin saturation (TS), serum ferritin (SF), albumin, α1 acid glycoprotein (α1-GP) and copper (Cu) levels in controls and patients with hereditary haemochromatosis (HH) or acquired iron overload.
 ControlHH patientsAcquired iron
(n = 36: 24 M, 12 W)
(n = 10: 7 M, 3 W)
(n = 25: 19 M, 6 W)
(n = 12: 9 M, 3 W)
  • *

    P < 0·001 versus control values.

  • Figures indicate mean ± SD. M, men; W, women.

Age (years)47 (17)51 (12)53 (13)32 (11)
Hb (g/dl)13 (2)14 (2)13 (1)10 (3)
SI (μmol/l)16 (4)28 (7)*22 (9)*30 (7)*
TS0·18 (0·07)0·86 (0·17)*0·45 (0·22)*0·77 (0·29)*
SF (μg/l)42 (26)1428 (986)*124 (91)*1661 (2263)*
Albumin (g/l)48 (3)46 (3)47 (5)42 (4)
α1-GP (μmol/l)15 (3)14 (2)14 (3)13 (3)
Cu (μmol/l)23 (6)22 (7)23 (8)25 (10)
Figure 1.

Reduction of serum ceruloplasmin levels in HH. Serum ceruloplasmin levels in patients with hereditary haemochromatosis (HH) compared with controls (CTR) and patients with acquired iron overload (AIO). Individual values are given as SI units, the conversion factor was 0·0662. Horizontal bars indicate the mean and vertical ones indicate the S.D. The horizontal line shows the minimal normal value (P < 0·001 between HH and CTR groups).


Recent genetic and clinical findings have confirmed the interaction of copper and iron metabolism, particularly the role that Cp plays in iron homeostasis (Askwith & Kaplan, 1998). This is the first report concerning Cp levels in patients with HH and with AIO not caused by mutations in the Cp gene. The significant decrease of Cp in HH but not in AIO suggests that hypoceruloplasminaemia is specifically associated with the genetic form of haemochromatosis, although low Cp levels were unrelated to the presence of the C282Y-HFE mutation. This is by no means a contradiction because HH is not necessarily associated with a mutated HFE gene and the diagnosis of the disease is based on clinical findings (Adams et al, 2000). In particular, in Italy the C282Y mutation is present in only 64% of patients with HH (Piperno et al, 1998) and other genetic defects may result in the HH phenotype (Pietrangelo et al, 1999; Camaschella et al, 2000). The limited characterization of the study populations in the preliminary series of Borda et al (1981) reporting hyperceruloplasminaemia in HH patients does not enable a comparison of their findings with the present ones.

Among the factors that may cause hypoceruloplasminaemia, we have been able to rule out the possibility that the liver damage caused by iron overload might lead to a general impairment of hepatic protein synthesis and Cp production. Similarly, the absence of any difference in serum copper and α1 acid glycoprotein between HH patients and controls rules out the possibility that differences in Cp levels were secondary to copper deficiency (Gitlin et al, 1992) or inflammatory conditions (Gitlin, 1988) respectively. Moreover, the finding of low Cp levels in HH cirrhotic patients in Child's class A but not in AIO patients with more severe liver damage (i.e. in Child's class B and C) challenges a relationship between reduction of Cp and cirrhosis. In agreement with this interpretation, Cp levels < 1·32 μmol/l have been found in only 0·6% of a large group of patients with liver disease (Cauza et al, 1997), whereas Cp concentrations were below this value in 37% of our HH patients. The slight increase in Cp levels in treated versus untreated HH patients, although not significantly different, might suggest that Cp concentration is influenced by iron deposition. However, that hypoceruloplasminaemia in HH patients was not caused by excess iron is strongly indicated by the normal Cp levels found in AIO patients, whose iron overload was similar or greater, as well as by the finding of unaltered Cp in patients with iron deficiency anaemia (data not reported).

In HH patients, distinctive iron metabolism abnormalities include increased intestinal absorption, low RE system retention and hepatic overload. As Cp is a circulating protein, it is difficult to understand its role in these different processes, but the present results suggest that low Cp levels may play a role in determining the specific pattern of iron distribution in HH.


We thank M. D. Cappellini for providing sera of thalassaemic patients and F. Bamonti for help in measuring copper levels.

This work was supported by grants from Ministero dell'Università e della Ricerca Scientifica e Tecnologica (Cofinanziamento 1999–2000), IRCCS Ospedale Maggiore (Competition Project 1999), Cariplo Foundation and Consiglio Nazionale delle Ricerche (CNR).