Eur J Vasc Endovasc Surg (2010) 40, 777e782

Skin Iron Deposition CharacterisesLipodermatosclerosis and Leg Ulcer

A. Caggiati a,*, C. Rosi a, A. Casini a, M. Cirenza b, V. Petrozza b,M.C. Acconcia c, P. Zamboni d

aDepartment of Anatomy, Sapienza University of Rome, ItalybDepartment of Pathology, Sapienza University of Rome, ItalycDepartment of Cardiology, Sapienza University of Rome, ItalydDepartment of Surgery, University of Ferrara, Ferrara, Italy

Submitted 31 May 2010; accepted 1 August 2010Available online 28 September 2010

KEYWORDSChronic venousdisorders;Venous ulcer;Lipodermatosclerosis;Iron;Haemosiderin

* Corresponding author. A Caggiati, M06 4991 8114; fax: þ39 0 4991 8081.

E-mail address: alberto.caggiati@u

1078-5884/$36 ª 2010 European Sociedoi:10.1016/j.ejvs.2010.08.015

Abstract Background: It is commonly reported that chronic venous disease (CVD) increasesthe skin iron content in which the excess is stored as haemosiderin. Despite increasing interestin the role of haemosiderin in venous ulceration, no study has systematically evaluated theoccurrence of iron overload in the limbs of patients with CVD.Purpose: To evaluate skin haemosiderin deposition in relation to the presence and severity ofskin changes in CVD legs designated according to the clinical, etiologic, anatomic and patho-physiologic (CEAP) classification.Methods: A total of 85 skin biopsies were taken from the medial aspect of 49 limbs with CVD ofCEAP clinical stages C2, C3, C4 and C6. The content of ferric ions was assessed by Perl’s Prus-sian Blue (PPB) stain.Results: No haemosiderin deposition was found in normal skin of C2, C3 and C4A legs, in lesssevere regions of pigmentation and in some parts of more severely affected limbs. Haemo-siderin was always present in lipodermatosclerotic skin and ulcers. Occasionally, haemosi-derin was found in the apparently normal perilesional skin of C4b and C6 legs. Theregenerating dermis at the base of healing ulcers showed none or light haemosiderin depo-sition.Conclusion: Iron overload is not present in the less severe stages of skin damage due to CVDbut lipodermatosclerosis and leg ulcers are always accompanied by haemosiderin deposition.In fact, no severe skin changes occur in CVD legs until iron overload occurs. Our results are inagreement with previous reports suggesting that a genetic inability to counteract skin iron

.D., Department of Anatomy, University ‘La Sapienza’, Via Borelli 50, 00161, Rome, Italy. Tel.: þ39

niroma1.it (A. Caggiati).

ty for Vascular Surgery. Published by Elsevier Ltd. All rights reserved.

778 A. Caggiati et al.

Table 1 Exclusion criteria.

1) Arterial diseases, lymphaticvasculitis, collagen disease).2) Dermatological disorders.3) Haematological diseases, ca4) Any previous surgical treatm5) History or current administr

overload is present in these patients. A more detailed analysis of disordered iron metabolismshould be undertaken in CVD patients.ª 2010 European Society for Vascular Surgery. Published by Elsevier Ltd. All rights reserved.

Legs afflicted with chronic venous disease (CVD) mayundergo skin changes ranging from small spot-like pigmen-tation to large ulcers.1 According to the clinical, etiologic,anatomic and pathophysiologic (CEAP) classification, thepresence of skin lesions characterises the severity of CVD.1

Although a number of cellular and molecular mediators,biochemical mechanisms and structural changes have beendescribed in detail,2 the exact pathogenesis of stasisdermatitis is still not completely understood. However, allcurrent studies agree that skin changes are due to aninflammatory process triggered by venular hypertension.3

An increased content of iron in the skin of CVD legswas firstreported in 1965 by Myers4 and is currently explained byextravasation and disruption of erythrocytes, followed bydecomposition of haemoglobin. The resulting excessive tissueiron (Feþþþ, ferric ions) is stored as haemosiderin. Theabnormal presence of this compound in the skin of CVD legshas been demonstrated by Perl’s Prussian Blue stain (PPB).5e7

Haemosiderin is currently considered to be the cause ofthe brownish skin pigmentation typical of CVD.4,8 Anincreasing interest in the role of haemosiderin in thepathogenesis of venous disease followed the demonstrationof either the pro-inflammatory properties of ferric ions,9,10

and the link between venous ulcer and genetical disordersof iron metabolism.10,11 Despite these important data, nostudy has systematically evaluated haemosiderin accumu-lation in relation to the presence and severity of skinchanges.

To contribute to an understanding of the effective roleof ferric ions in the pathogenesis of CVD-related skinchanges and ulcer development, we have morphometricallyevaluated haemosiderin deposition in skin samples fromlegs afflicted with CVD at different clinical stages of theCEAP classification.

Patients, Materials and Methods

A total of 44 Caucasian patients in whom clinical investi-gation and duplex ultrasonography demonstrated thepresence of CVD8 were included in this study. The severityof CVD in each leg was assigned to the appropriate CEAPclinical class by a physician skilled in the management of

disorders, microangiopathies (d

rdiac, pulmonary, renal or hepaent, trauma or irradiation of tation of drugs known to affect

venous disease.1 The severity of pigmentation was scoredaccording to the Venous Clinical Severity Score (VCSS).12

History or evidence of morbidities or treatments possiblyinfluencing skin changes and metabolism were consideredto be exclusion criteria (Table 1). The local ethicscommittee approved this study and all patients gaveinformed written consent to their inclusion in this research.

As many as 85 skin biopsies were taken from the medialaspect of 49 legs with CVD (C2: 16 legs; C3: four legs; C4A:12 legs; C4B: 10 legs; C6: seven legs) in 44 patients (20 men,24 women; mean age 62 years). A total of 49 biopsies werefrom apparently normal skin, 12 were from pigmentedlesions (seven scored 1e2 by VCSS; five scored 3), 10 weretaken from skin areas with lipodermatosclerosis (LDS) and14 were from open ulcers. No samples were obtained fromC5 legs (Table 2).

In C2, C3 and C4 legs, punch biopsies (2e4 mm) wereobtained mostly during surgery for varicose veins (theresulting holes were then used to perform avulsion ofvaricosities). Specimens from C6 legs were also obtainedduring ulcer debridement or skin grafting. All biopsiesincluded the whole dermis and at least a thin layer ofsubcutaneous tissue.

All legs with skin changes (pigmentation, LDS or ulcer)underwent at least two biopsies, one from the affected skinand one from the adjacent apparently normal skin. In C6legs, ulcer tissues were taken from both the wound edgeand ulcer bed. Finally, four skin biopsies were taken fromthe medial aspect of the leg in four subjects (two male, twofemale; mean age 57.7 years) in whom there was no clinicalor duplex evidence of CVD. These patients were undergoingorthopaedic surgery. Skin biopsies were fixed in 10% buff-ered formalin solution and routinely processed for haema-toxylin/eosin (HE) and PPB stains. This technique is able todemonstrate the presence of haemosiderin in the dermis ashas been previously demonstrated.5e7,11

Quantitative measurements of haemosiderin depositionwereperformedondigital images capturedby theD-Sight Fluomicroscope (Menarini, Italy). PPB stain positive areas wereobjectively quantified using the D-Sight Viewer package(Menarini, Italy)where themarginof the tissuewas tracedandthe area containing haemosiderin was defined as the Area ofInterest (AoI). This resulted in colour segmentation expressed

iabetes, rheumatoid arthritis,

tic failure.he leg.iron metabolism and skin trophism

Table 2 Designation of samples according to C-Class.

C-Class C2 C3 C4A C4B C6 Total

Number of legs 16 4 12 10 7 49Number of biopsies 16 4 24 20 21 85

Skin Iron Deposition During CVD 779

as percentage of AoI. Data were finally transferred into anExcel (Microsoft) spreadsheet for further analysis.

The mean value of AoI and standard deviation (SD) werecalculated in PPB þ specimens grouped according to theCEAP-Class of the donor leg. Due to the small sample size,these values were compared by using ManneWhitney test.A p-value < 0.05 was regarded as significant. Bonferroni’scorrection was applied to control the possibility of a type Ierror.

Results

Haemosiderin was visible as granules (siderosomes), whichappeared brownish or black in HE sections and blue in PPBsections (Fig. 1). Siderosomes were found within dermalmacrophages (siderophages) (Fig. 1(A)) as well as in theinterstitium between the collagen bundles of the dermis(Fig. 1(B)). In case of more intense haemosiderin deposi-tion, PPB stained diffusely the dermis (Fig. 1(C)) and eventhe stroma of the subcutaneous layer (Fig. 2(D)). Theprevalence of PPB þ specimens and the mean values of AoIare reported in Table 3.

Control legs: All samples from control legs were nega-tive for haemosiderin deposition (AoI Z 0%).

Apparently normal skin from CVD legs: No haemosiderinwas found in most (46/49) biopsies taken from apparently

Figure 1 Skin biopsies from medial face of CVD legs. A) Granules(HE, original magnification 20�). B) Siderosomes may be found(HE, original magnification 20�). C) Border of an active ulcer. Ininterstitial spaces. (PPB, original magnification 20�).

normal skin. The only exceptions were two samples takenfrom C4B legs in proximity of a lipodermatosclerotic plaque(AoI Z 1.6 and 9.6, mean 5.6, SD 5.7) and one from a C6 legtaken close to the border of an active ulcer (AoI Z 2.8%).

Pigmented skin: Haemosiderin was absent in all biopsiesinvolving mild-to-moderate pigmentation (scored 1e2 byVCSS). It was found in only three out of five biopsies takenfrom severe pigmentations that scored 3 by VCSS (meanAoI Z 10.3%, SD 11.4).

Lipodermatosclerotic skin: All these biopsies showedsevere haemosiderin deposition in the dermis (Fig. 2(C))and even in the subcutaneous layer (Fig. 2(D)) (meanAoI Z 45%, SD 9.0).

Specimens from ulcerated skin: Specimens from ulcertissues showed haemosiderin deposition (mean AoI of 28, SD9.5) (Fig. 1(C)). No significant differences were seenbetween specimens from wound edge and bed (mean AoI:28, SD 9.3 and 25 � 17, respectively). In two ulcers, whichshowed reparative phenomena (Fig. 3), the regeneratingdermis at the wound bed was nearly completely devoid ofhaemosiderin (AoI: 0% and 7.4%, respectively).

Statistical evaluation

A ManneWhitney U test with Bonferroni’s correctiondemonstrated that haemosiderin deposition was

of haemosiderin (siderosomes) are visible within macrophagesalso in the interstitial spaces between dermal fibroblasts.presence of excessive iron overload, PPB diffusely stains the

Figure 2 Differences in the intensity of haemosiderin deposition in skin biopsies from medial face of CVI-legs. A) Absence ofhaemosiderin deposition (PPB, original magnification 20�). B) Light haemosiderin deposition. This specimen belonged to a biopsy ofapparently normal skin taken at the border of an LDS area (PPB, original magnification 20�). C) Biopsy of LDS skin (PPB, originalmagnification 20�). D) Haemosiderin deposition extending to the subcutaneous layer (PPB, original magnification 20�).

780 A. Caggiati et al.

significantly higher in lipodermatosclerotic skin if comparedwith samples from heavily pigmented skin (p < 0.01) andulcer tissues (p < 0.01).

Discussion

Iron is normally present in the skin because about 20e25%of absorbed iron is daily eliminated by exfoliation ofepidermal cells.13 As a protection against its toxicity, iron issequestered in iron-binding or haeme-containing proteins(ferritin and haemosiderin, respectively). Haemosiderin iscurrently considered the histological marker of tissue ironexcess.7,13

In the past, skin iron overload was demonstrated in CVDlegs by PPB,6,7 X-ray spectrometry14 and atomic absorptionspectrum analysis.9 However, most of these investigationsconsidered ulcerated legs. The present study is thus thefirst to have systematically evaluated skin haemosiderindeposition in legs at any stage of severity of CVD. The mainfindings can be summarised as follows:

No haemosiderin deposition was found in the normal skinof C2, C3 and C4A legs, in less severe pigmentation and inpart of more severe ones. By contrast, haemosiderin wasalways present in lipodermatosclerotic skin and activeulcers. Occasionally, haemosiderin was found in theapparently normal perilesional skin of C4b and C6 legs.

Table 3 Prevalence of PPB-positive samples and intensity of haboth C-class of the donor leg and clinical appearance of the skin

C Class C2 C3 C4

Apparently normal skin 0/16(0%) 0/4 (0%) 0/1Pigmentations þ/þþ e e 0/7Pigmentations þþþ e e 3/5Lipodermatosclerotic e e e

Ulcer edge e e e

Ulcer bed e e e

These results indicate that haemosiderin deposition doesnot intervene in the initial phases of stasis dermatitis. Inturn, PPB-positive ferric overload accompanies the wors-ening of skin damage and, from the histological point ofview, it characterises LDS and ulcers. Iron loss due to ulcersecretion6 could explain why iron content in ulcer samplesis lower than in LDS skin.

The regenerating dermis at the base of two healingulcers showed none or very light haemosiderin deposition.This agrees with previous studies, which demonstratedthat iron levels decrease when venous ulcer heals.5e7

Unfortunately, biopsies of healing or healed ulcers aredifficult to obtain due to ethical implications. We couldonly hypothesise that reduction of PPB staining mightcorrelate with ulcer healing.

Iron is a well-known mediator of skin toxicity in a varietyof pathological conditions including sunburn,15 porphyriacutanea tarda,16 inflammation17 and even skin cancer.18 InCVD legs, it has been demonstrated that ferric ions con-tained in haemosiderin intervene in the pathogenic chainleading to matrix disruption and ulcer development bygeneration of free radicals,6,9 activation of metal-loproteinase19 or, finally, down-regulation of tissue inhibi-tors of metalloproteinase.9,20,21 Our findings do not explainthe exact iron-derived mechanism of skin damage butclearly demonstrate that severe skin changes and tissueiron overload are highly correlated.

emosiderin deposition. Samples are designated according towhere biopsies were taken. Mean AoI between brackets.

A C4B C6 Total

2 (0%) 2/10 (5.6%) 1/7 (2.8%) 49(0%) e e 7(10.3%) e e 5

10/10 (45.0%) e 10e 6/7 (28.0%) 7e 7/7 (25.4%) 7

Figure 3 A) 38 years old male, left leg. B) Note the great deposition of haemosiderin in the ulcer border (PPB, originalmagnification 20�). C) Haemosiderin is nearly completely absent in the regenerating epithelium of the ulcer base (PPB, originalmagnification 20�).

Skin Iron Deposition During CVD 781

In 1868, John Gay noted that “.varicose veins can bepresent for many years without any ulcer or bronzing ofthe skin.”.22 In medical practice, it is a common finding tosee legs with severe and long-lasting CVD and normal skinor, by contrast, young subjects with heavy pigmentationand even LDS or ulcer (Fig. 3). Similarly, skin iron overloadcorrelates with the entity of skin changes but not with theduration and severity of CVD. It is evident that impairmentof venous haemodynamic is “.an essential but insufficientfactor in explaining subsequent skin changes.”.11

Our findings suggest that patients with LDS or ulcer areunable to counteract CVD-induced skin iron overload. It hasbeen hypothesised that such inability could be geneticallydetermined.23 This has been demonstrated in other cuta-neous pigmentation diseases such as porphyria cutaneatarda24 and explains why cutaneous siderosis occurs in onlyabout one-half of patients with haemochromatosis.25,26

Recently, a strong genetic component in the pathogenesisof ulcer has been reported.19,27 In particular, a positivecorrelation has been demonstrated between susceptibility,healing time and size of venous ulceration with mutations ofgenes involved in iron metabolism.19 If these mutations arecurrently considered asymptomatic for systemic diseases, incombination with an acquired condition (venular hyperten-sion) they could be responsible of LDS and ulcer.23

The weaknesses of this study include the fact that fewbiopsies were taken from healing ulcers and none from C5

legs due to ethics committee restrictions. These limited theextent of this research. We have so far not performedgenetic investigation to assess iron metabolisms disordersin these patients.

In conclusion, our findings indicate that haemosiderindoes not intervene in the initial phases of skin changes invenous disease and that its deposition appears to beobligatory when severe skin changes and ulcerationdevelop. Our results are in agreement with previous reportssuggesting a genetic inability to counteract skin iron over-load. Considering that at least 10% of Caucasians arecarriers of iron metabolism disorders,28,29 it is desirablethat genetic investigation should be included in the clinicalevaluation of patients with CVD.19,27 This could help clini-cians to identify leg ulcers that require advanced wound-healing treatment or surgery, in conjunction withcompression therapy.

Conflict of Interest

None.

Acknowledgements

This study was supported by funds from Italian MURST andby a grant from Servier Company.

782 A. Caggiati et al.

References

1 Eklof B, Bergan JJ, Gloviczki P, Kistner RL, Meissner MH,Smith PC, et al. American venous forum international Ad Hoccommittee for revision of the CEAP classification revision of theCEAP classification for chronic venous disorders: consensusstatement. J Vasc Surg 2004;40:1248e52.

2 Meissner MH, Moneta G, Burnand K, Gloviczki P, Lohr JM,Lurie F, et al. The hemodynamics and diagnosis of venousdisease. J Vasc Surg 2007;46:4Se24S.

3 Bergan JJ, Schmid-Schombein GW, Smith PD, Nicolaides AN,Bopisseau MR, Eklof B. Chronic venous disease.N Engl J Med 2006;355:488e98.

4 Myers HL. Topical chelation therapy for varicose pigmentation.Angiology 1965;17(1):66e68. American College of Angiology11th Annual Meeting.

5 Herrick SE, Sloan P, McGurk M, Freak L, McCollum CN,Ferguson MW. Sequential changes in histologic pattern andextracellular matrix deposition during the healing of chronicvenous ulcers. Am J Pathol 1992;141:1085e95.

6 Yeoh-Ellerton S, Stacey MC. Iron and 8-isoprostane levels inacute and chronic wounds. J Invest Dermatol 2003;121:918e25.

7 Tan J, Smith A, Abisi S, Eastham D, Burnand KG. Tissue andurinary haemosiderin in chronic leg ulcers. Eur J Vasc EndovascSurg 2007;34:355e60.

8 Eberhardt RT, Raffetto JD. Chronic venous insufficiency.Circulation 2005;111:2398e409.

9 Wenk J, Foitzik A, Achterberg V, Sabiwalsky A, Dissemond J,Meewes C, et al. Selective pick-up of increased iron bydeferoxamine-coupled cellulose abrogates the iron-driveninduction of matrix degrading metalloproteinase 1 and lipidperoxidation in human dermal fibroblasts in vitro: a newdressing concept. J Invest Dermatol 2001;116:833e9.

10 Zamboni P, Izzo M, Fogato L, Carandina S, Lanzara V. Urinehaemosiderin: a novel marker to assess the severity of chronicvenous disease. J Vasc Surg 2003;37:132e6.

11 Zamboni P, Izzo M, Tognazzo S, Carandina S, De Palma M,Catozzi L, et al. The overlapping of local iron overload and HFEmutation in venous leg ulcer pathogenesis. Free Radic Biol Med2006;40:1869e73.

12 Ricci MA, Emmerich J, Callas PW, Rosendaal FR, Stanley AC,Naud S, et al. Evaluating chronic venous disease with a newvenous severity scoring system. J Vasc Surg 2003;38:909e15.

13 Adams BD, Lazova R, Andrews NC, Milstone LM. Iron in skin ofmice with three etiologies of systemic iron overload. J InvestDermatol 2005;125:1200e5.

14 Ackerman Z, Seidenbaum M, Loewenthal E, Rubinow A. Over-load of iron in the skin of patients with varicose ulcers: possiblecontributing role of iron accumulation in progression of thedisease. Arch Dermatol 1988;124:1376e8.

15 Bissett DL, ChatterjeeR,HannonDP. Chronic ultraviolet radiation-induced increase in skin iron and the photoprotective effect oftopically applied iron chelators. Photochem Photobiol 1991;54:215e23.

16 Takeshita K, Takajo T, Hirata H, Ono M, Utsumi H. In vivo oxygenradical generation in the skin of the protoporphyria modelmouse with visible light exposure: an L-band ESR study. J InvestDermatol 2004;122:1463e70.

17 Gira AK, Casper KA, Otto KB, Naik SM, Caughman SW, Swerlick RA.Induction of interferon regulatory factor 1 expression in humandermal endothelial cells by interferon-gamma and tumornecrosis factor-alpha is transcriptionally regulated and requiresiron. J Invest Dermatol 2003;121:1191e6.

18 Bhasin G, Kausar H, Sarwar Alam M, Athar M. Progressive ironoverload enhances chemically mediated tumor promotion inmurine skin. Arch Biochem Biophys 2003;409:262e73.

19 Gemmati D, Federici F, Catozzi L, Gianesini S, Tacconi G,Scapoli GL, et al. DNA-array of gene variants in venous legulcers: detection of prognostic indicators. J Vasc Surg 2009;50:1444e51.

20 Herouy Y, May AE, Pornschelegel G, Stetter CH, Grenz H,Preissner KT, et al. Lipodermatosclerosis is characterized byelevated expression and activation of matrix metal-loproteinases: implications for venous ulcer formation. J InvestDermatol 1998;111:822e7.

21 Abd-El-Aleem SA, Ferguson MW, Appleton I, Kairsingh S,Jude EB, Jones K, et al. Expression of nitric oxide synthaseisoforms and arginase in normal human skin and chronic venousleg ulcers. J Pathol 2000;191:434e42.

22 Loudon IS. Leg ulcers in the eighteenth and early nine-teenth centuries, II: treatment. J R Coll Gen Pract 1982;32:301e9.

23 Zamboni P, Scapoli G, Lanzara V, Izzo M, Fortini P, Legnaro A,et al. Serum iron and MMP-9 variations in limbs affected bychronic venous disease and venous leg ulcers. Dermatol Surg2005;31:19e22.

24 Skowron F, Berard F, Grezard P, Wolf F, Morel Y, Perrot H. Roleof the hemochromatosis gene in prophyria cutanea tarda.Prospective study of 56 cases. Ann Dermatol Venereol 2001;128:600e4.

25 Conte D, Piperno A, Mandelli C, Fargion S, Cesana M,Brunelli L, et al. Clinical, biochemical and histologicalfeatures of primary haemochromatosis: a report of 67 cases.Liver 1986;6:310e5.

26 Milman N, Agger AO, Nielsen OJ. Iron supplementation duringpregnancy. Effect on iron status markers, serum erythropoietinand human placental lactogen. A placebo controlled study in207 Danish women. Dan Med Bull 1991;38:471e6.

27 Charles CA, Tonic-Canic M, Vincek V, Nassiri M, Stojadinovic O,Eaglestein WH, et al. A gene signature of nonhealing venousulcers: potential diagnostic markers. J Am Acad Dermatol 2008;59:758e71.

28 McCune CA, Ravine D, Worwood M, Jackson HA, Evans HM,Hutton D. Screening for hereditary haemochromatosis withinfamilies and beyond. Lancet 2003;362:1897e8.

29 Beutler E, Felitti VJ, Koziol JA, Ho NJ, Gelbart T. Penetrance of845G. A (C282Y) HFE hereditary haemochromatosis mutation inUSA. Lancet 2002;359:211e8.