Time-course analysis of serum hepcidin, iron and cytokines in a C282Y homozygous patient with Schnitzler's syndrome treated with IL-1 receptor antagonist

It is currently unknown if the increase of the hepatic ironregulatory hormone hepcidin during inflammation in man dependson an intact HFE-protein. Here we describe the temporal relationshipof serum hepcidin, serum iron and cytokines in a patient withHFE-related (C282Y homozygous) hereditary hemochromatosis whowas treated for an auto-inflammatory condition, i.e. variantSchnitzler’s syndrome, with the potent anti-inflammatorycytokine inter-leukin-1 receptor antagonist (IL-1ra, anakinra).The patient had bouts of fever with peaking serum IL-6 concentrationsfollowed by peaking serum hepcidin levels, while serum ironwas low. Upon treatment, these peaks disappeared and hepcidinlevels became non-detectable, consistent with HFE deficiency.In conclusion, this in vivo human model: i) supports the importanceof an HFE-independent IL-6-hepcidin axis in the developmentof hypoferremia and anemia of inflammation; and ii) suggeststhat chronic inflammation protects patients with HFE-relatedhereditary hemochromatosis from iron accumulation.

Key words: hepcidin, iron, hemochromatosis, inflammation, interleukin-1.

Introduction

TOP

Introduction

One of the critical host defense strategies after infectionis to withhold iron from the invading pathogen by redistributionof iron out of the circulation into macrophages and decreaseof iron absorption by enterocytes.^1,2 This inflammation-associatedreaction is mediated primarily by the hepatocellular peptidehormone hepcidin.^3,4 Hepatic hepcidin production is under thecontrol of various physiological stimuli, i.e it is down-regulatedby low body iron status and erythropoietic activity and increasedby pro-inflammation cytokines, particularly interleukin -6 (IL-6).^5,6

Patients with HFE-gene related (C282Y-homozygous) hereditaryhemochromatosis (HH) have an impaired hepcidin gene expressionwith decreased, although not uniformly low, serum hepcidin levels.⁷Experiments in HFE-knock out (KO) mice show that, in spite ofthis positive regulatory action of HFE-protein on basal hepcidinlevels, HFE-protein is not required for hepcidin upregulationduring inflammation.^5,8,9 One study, however, showed a bluntedhepcidin response to inflammatory stimuli in the KO-mice.¹⁰Thus, with respect to hepcidin regulation at least two distinguishablepathways are operational: one which is HFE-protein dependentand one engaged after pro-inflammatory stimuli. To the bestof our knowledge, these mouse data have not yet been confirmedby observations in humans. Here we describe the temporal relationshipsof serum hepcidin, serum iron and inflammatory cytokines, ina C282Y homozygous HH patient who was treated for an auto-inflammatorycondition, i.e. variant Schnitzler’s syndrome,¹¹ withthe potent anti-inflammatory cytokine interleukin-1 receptorantagonist (anakinra).¹² We demonstrate the existence of anHFE-independent cytokine-hepcidin axis in the development ofhypoferremia and anemia of inflammation. Our observations furthersuggest that chronic inflammation countermands iron accumulationin C282Y-homozygotes.

Case description
The patient, a 53-year old man, started having complaints in1991, at the age of 36 years. From that time he experiencedperiodically malaise, myalgia, spiking fever, arthralgia, painfulswelling and redness of eye lids. At first examination his spleenwas slightly enlarged and laboratory analysis showed a markedIgG type- ${kappa}$ monoclonal gammopathy of 21 g/L. Under the suspicionof malignant lymphoma, a splenectomy was performed with biopsyof liver and intra-abdominal lymph nodes. No malignancy couldbe detected. Unexpectedly, on histology spleen and liver showedsigns of hemosiderosis. More specifically, in the absence offibrosis and cirrhosis, Perls’ stain was generalized andstrongly positive in hepatocytes and also present in some fociof Kupffer cells. At that time ferritin was 1540 µg/L(normal <280 µg/L) and the transferrin saturation (TS)45%. In the late 1990s, HFE-gene analysis disclosed homozygosityfor the C282Y mutation in the patient and his brother. Thisconfirmed the diagnosis of HFE-related HH.

Over the following year, the patient was treated with 30 phlebotomiesof 500 mL each to a final ferritin level of 24 µg/L (TSof 9.8%). Because the original physical complaints of arthralgiaand periodic fever persisted, prednisone was started in a maintenancedose of 15 mg/day. Nevertheless, there were various short-lastingbouts of fever, arthralgia, red swollen eyes and auricles. Inthe late 1990s progressive deafness developed. The monoclonalgammopathy remained stable. From 1993 to 2007, the patient receiveda further 15 phlebotomies of 500 mL each. Serum ferritin levels,TS and CRP in this time interval, varied between 36 and 540µg/L, 11.4 and 106%, and 26 and 81 mg/L, respectively.Although these data appeared to be too incomplete to draw conclusionsabout a temporal relationship between inflammation episodesand iron parameters, it is clear that the patient was not alwaysin an inflammatory state.

In 2007, at the age of 52 years, the physical complaints becameprogressively worse, and consisted of malaise, arthralgia, chills,fever, and short lasting generalized urticae. Clinical re-evaluationin 2008 showed, in spite of the invariably high (20–25g/L) monoclonal gammopathy and high number (10–15%) ofbone-marrow plasma cells, no other signs of malignant multiplemyeloma. The bouts of fever were accompanied by IL-6 peaks followedby increased hepcidin-concentrations (Figure 1, panels A andB).

View larger version (12K):
[in this window]
[in a new window]
[Download PPT slide]
Figure 1. Laboratory measurements before and after treatment. Body temperature, CRP and IL-6 levels (panel A), serum hepcidin (panel B), iron and transferrin saturation (panel C) were measured in an iron deplete C282Y homozygous patient with an auto-inflammatory condition, i.e. variant Schnitzler’s syndrome, before, during and after treatment with the potent anti-inflammatory cytokine interleukin-1 receptor antagonist (IL-1ra, anakinra). The period of IL-1ra treatment is indicated in grey.

Based on the occurrence of chills with spiking fever and urticaein the presence of IgG- ${kappa}$ gammopathy, the diagnosis of variantSchnitzler’s syndrome was considered and it was decidedto treat the patient with inter-leukin-1 receptor antagonist(IL-1ra; anakinra; 1^st dose 300 mg i.v., followed by six days100 mg s.c.).^11,12 Serum ferritin levels were around 200 µg/Lin the days before treatment, stayed between 200 and 250 µg/Lduring treatment and decreased thereafter to 140, 80 and 22µg/L, at days 13, 17 and 36, respectively. Immediatelyafter the start of IL-1ra all complaints disappeared. The treatmentabrogated c.q. prevented the pattern of peaking fever in thefollowing days and IL-6 and CRP normalized (Figure 1, panelA). Interestingly, serum hepcidin, clearly detectable in spiteof the HFE-gene defect during the febrile inflammatory episode,decreased rapidly after start of IL-1ra whereas serum iron increased(Figure 1, panels B and C). IL-1β levels remained below20 pg/mL before and after IL-1ra treatment (data not shown).The moderately increased IL-18 levels were not clearly affectedby IL-1ra treatment (data not shown).

Design and Methods

TOP

Design and Methods

Serum iron parameters and CRP were determined by standard laboratoryanalysis. Cytokines and hepcidin were assayed in one run inserum stored at –80°C. Cytokines were measured usinga multiplex Luminex assay.¹³ Hepcidin was measured as describedpreviously.¹⁴ Lower limit of detection was 0.5 nmol/L. Hepcidinrange in healthy volunteers (n=24) 0.5–13.9 nmol/L, median4.2 nmol/L; in untreated C282Y homozygotes (n=22) <0.5–4.8nmol/L, median 1.9 nmol/L; in C282Y homozygotes (n=4) in maintenancephase of phlebotomy (15 samples) <0.5–2.5 nmol/L.⁷The study was carried out in accordance with applicable Dutchregulations concerning IRB review and informed consent.

Results and Discussion

TOP

Results and Discussion

We describe the first human in vivo observation of serum ironparameters, including hepcidin, in a patient with an aberrantHFE-protein and inflammation, before and after anti-inflammatorytreatment. The patient presented with a combination of Schnitzler’ssyndrome and C282Y homozygosity of the HFE-gene under a regimeof maintenance phlebotomies, and was subsequently treated withanti-inflammatory anti-IL-1ra.

Schnitzler’s syndrome is a rare disabling auto-inflammatorydisease. It is characterized by varying degrees of recurrentbouts of unexplained fever, arthralgia or arthritis, bone pain,lymphadenopathy, hepato- or splenomegaly, leukocytosis, andelevated erythrocyte sedimentation rate accompanied by an urticarialrash in the presence of a monoclononal IgG or IgM gammopathy.The complaints respond well to treatment with IL-1ra.¹²

A biochemically penetrating C282Y homozygous patient is typicallycharacterized by elevations of both serum ferritin and serumiron levels.¹⁵ In the patient presented herein the ongoing andperiodically exacerbating inflammation increased serum ferritinlevels and lowered TS, at times leading to a pattern of serumiron indices less pathognomonic for HH. Thus, the iron statusin the patient might be understood from a combination of HHtreated with phlebotomies, and both sequestering of iron inthe reticulo-endothelial (RE) macrophages and reduced intestinaliron uptake due to inflammation induced elevated hepcidin levels.

The bouts of fever and IL-6 peaks in the patient were followedby peaks of serum hepcidin. Hepcidin serum concentrations beforetreatment in our patient ranged from 2.9 to 11.2 nmol/L, highfor an C282Y homozygote in the maintenance phase of phlebotomy.⁷Normalization of IL-6, triggered by IL-1ra treatment, led toundetectable hepcidin and higher serum iron levels. These findingsconfirm the (temporal) relations, observed by us in an in vivomodel of LPS induced inflammation,⁶ and by Kawabata et al. inan IL-6 overproducing patient with multicentric Castleman’sdisease treated with IL-6 receptor antibodies.¹⁶ In additionto this, IL-1ra may directly inhibit IL-1 dependent hepcidinproduction through the blockade of the IL-1 receptor.⁵

At first glance, our data seem to contrast with the observationthat in C282Y homozygotes the total body iron does not correlatewith the momentary IL-6 and CRP levels.¹⁷ The latter reporthowever can be criticized because of the different kineticsof IL-6, CRP and body iron. On the other hand, our observationsof a patient with a defect of the HFE protein that can stillrespond to inflammatory stimuli by increasing hepcidin levelsare in agreement to those described for patients with deficiencyof the TfR2 protein.¹⁸ Interestingly, HFE and TfR2 proteinshave been found to interact in a common signal transductionpathway to hepcidin.¹⁹

From our observation it is clear that C282Y homozygosity and/oriron deficient anemia do not abrogate cytokine induced hepcidinproduction. However, in the absence of appropriate controlsfor this natural biological model as provided by the describedpatient, no conclusions can be made as to whether the levelachieved in the patient is of the same level seen in humanswithout HFE-related HH. Nevertheless, the retained ability ofC282Y homozygotes to increase hepcidin production upon inflammatorystimuli, and thus, according to the hepcidin mode of action,²⁰to redistribute iron out of the bloodstream into RE macrophages,might be advantageous in the setting of infection by extracellularpathogens, but on the other hand might also increase hepaticdamage through iron-induced mesenchymal activation.²¹

In conclusion, our in vivo human model of a C282Y homozygouspatient with an auto-inflammatory disease treated with IL-1rasupports the importance of an HFE-independent cytokine-hepcidinaxis in the development of hypoferremia and anemia of inflammation.These findings on the cross-talk between pathways regulatinghepcidin provides indirect information on the iron availabilityto certain pathogens relevant in critical host defense strategies.^2,22–24It furthermore suggests that chronic inflammation may reduceiron stores in C282Y homozygotes, contributing to the variablepenetrance of the C282Y mutation.

Acknowledgments

we thank the patient for his consent, J. van der Ven-Jongekrijgfor the cytokine assays and the team of the www.hepcidinanalysis.cominitiative for fruitful discussions.

Footnotes

Authorship and Disclosures

MvD and DWS designed research, collected and interpreted dataand wrote the paper. MvD contributed patient samples. DWS andJJCK contributed laboratory tools. JJCK performed hepcidin measurements,collected and interpreted data.

The authors reported no potential conflicts of interest.

Received for publication January 14, 2009. Revision received March 6, 2009. Accepted for publication March 24, 2009.

References

TOP