Nephrogenic systemic fibrosis is associated withtransforming growth factor b and Smad without

evidence of renin-angiotensin system involvement

Brent Kelly, MD, Matthew Petitt, DO, and Ramon Sanchez, MD

Galveston, Texas

Background: The mechanisms of fibrosis associated with nephrogenic systemic fibrosis (NSF) are largelyunknown. Transforming growth factor beta (TGF-b), a known profibrotic cytokine, is theorized to play acentral role. The renin-angiotensin system has been linked with both TGF-b expression and fibrosis inother organ systems.

Objective: We sought to investigate whether these mechanisms were involved with NSF.

Method: Eleven biopsy specimens from 8 patients with NSF were evaluated by immunohistochemistry forthe expression of TGF-b, Smad 2/3, angiotensin-converting enzyme (ACE), and angiotensin II receptor1 (AT1).

Results: TGF-b was detected in 8 of 11 samples of NSF. Smad 2/3 nuclear staining was seen in 8 of 11samples. Conversely, only faint staining for ACE was seen in 2 of the 11 specimens. No AT1 staining wasseen.

Limitations: We did not perform our studies on a cohort of comparable patients with renal dysfunctionwithout NSF. Our technique may not have been sufficiently sensitive to detect renin-angiotensin systeminvolvement.

Conclusions: TGF-b, as well as its second messengers, Smad 2/3, appears to be associated with the fibrosisseen in NSF. No definitive evidence of renin-angiotensin system involvement could be determined. ( J AmAcad Dermatol 2008;58:1025-30.)

INTRODUCTIONNephrogenic systemic fibrosis (NSF) was first de-

scribedbyCowper et al1 in2000as a scleromyxedema-like cutaneous disease at that time affecting 15patientsreceiving renal dialysis. Since this first description,more than 200 cases have been reported.2 The diseaseis a fibrosing disorder exclusively affecting patientswith some underlying degree of renal dysfunction,

From the Department of Dermatology, University of Texas Medical

Branch, Galveston.

Funding sources: None.

Conflicts of interest: None declared.

This work was presented at the Medical Dermatologic Society

meeting held during the 2008 American Academy of Derma-

tology meeting in San Antonio, Tex.

Reprint requests: Brent Kelly, MD, University of Texas Medical

Branch, 301 University Blvd, 4.112 McCollough Bldg, Galveston,

TX 77555-0783. E-mail: bckelly@utmb.edu.

0190-9622/$34.00

ª 2008 by the American Academy of Dermatology, Inc.

doi:10.1016/j.jaad.2008.02.038

many of whom receive dialysis. Common cutaneousmanifestations include symmetric skin tightening andinduration, particularly on lower extremities andforearms. Involvement of the trunk also occurs andthe disease may progress to contractures, immobility,and even death. Involvement of internal organs in-cluding fascia, myocardium, striated muscle, and duramater3 has led to the designation of this process as asystemic disease.

The histologic findings include increased colla-gen, a proliferation of spindle-shaped fibroblasts,mildly increased dermal mucin, and minimal to

Abbreviations used:

ACE: angiotensin-converting enzymeAng: angiotensinAT1: angiotensin II receptor 1Gd: gadoliniumNSF: nephrogenic systemic fibrosisTGF-b: transforming growth factor beta

1025

J AM ACAD DERMATOL

JUNE 2008

1026 Kelly, Petitt, and Sanchez

absent inflammation. These findings are most similarto those of scleromyxedema; however, the interest-ing finding of CD341/procollagen11 coexpressionon fibroblasts suggests that these spindle cells arederived from a circulating fibrocyte originating fromthe bone marrow.4

Since the original description by Cowper et al,many etiologies have been proposed for the devel-opment of the disease, including exogenous agents,components of dialysis equipment, erythropoietin,5

medications (including the absence of angiotensin-converting enzyme [ACE] inhibitors),6 recentvascular procedures, an unidentified virus, hyperco-agulability and autoimmune disorders. In 2006,Grobner7 observed an association with prior expo-sure to gadolinium in patients in his dialysis center.Five of 9 patients with end-stage renal disease, whoreceived magnetic resonance angiography, devel-oped symptoms of NSF. Similar findings were foundby Marckmann et al.8 These findings have led to apublic health advisory by the Food and DrugAdministration (FDA), suggesting that caution beused when considering gadolinium-containing con-trast agents in patients with a low glomerular filtra-tion rate; the FDA has requested boxed warningsfrom the manufacturer’s labels.9

The exact mechanism of fibrosis in NSF is cur-rently unknown. Leading theories postulate thatCD341 circulating fibrocytes, after being stimulatedfrom the bone marrow, are recruited into cutaneoustissues, resulting in the uncontrolled fibrosis.4

Indeed, CD341 spindle cells are one of the hallmarksof the histopathologic findings in NSF. How they arerecruited is unknown; however, speculation re-volves around gadolinium leaked into tissues.Although no causal evidence is present for thislink, gadolinium has been detected in skin samplesof NSF by means of electron-dispersive x-ray spec-troscopy.10-12

Transforming growth factor b (TGF-b) is a multi-functional cytokine known to have profibrotic prop-erties, particularly induction of synthesis ofextracellular matrix proteins.13 In other organs andfibrosing diseases, TGF-b has been associated withfibrosis.13 Additionally, in these other systems, par-ticularly cardiac14 and renal,15 the renin-angiotensinsystem has been intimately associated with fibrosisand TGF-b production both indirectly and directly.The relationship of a local renin-angiotensin systemand cutaneous fibrosing disorders is less clear, buthas been suggested in entities such as systemicsclerosis16 and keloids.17

The first aim of this study was to evaluate theexpression of TGF-b in skin specimens of NSF.Second, we theorized that if TGF-b played a role in

fibrosis, a cutaneous renin-angiotensin system mayup-regulate its production and therefore be detect-able in our samples. To investigate these aims, weevaluated the expression of TGF-b1, Smad 2/3(known second messengers of TGF-b), ACE, andangiotensin II receptor (AT1) by immunohistochem-istry on 11 paraffin-embedded biopsy specimensfrom 8 previously described12 patients with NSF.

MATERIAL AND METHODSEleven biopsy specimens from 8 patients with NSF

were obtained from our institution. Charts werereviewed for epidemiologic data. The hematoxylin-eosinestained slides, as well as any special stains,were reviewed by 3 dermatopathologists to confirmdiagnosis. Two additional specimens (#8 and #11)from uninvolved skin of patients with NSF wereincluded in the analysis (Table I). Seven of the 8patients have been described previously12 and havehad x-ray spectroscopy performed on the speci-mens. These data were also reviewed.

ImmunohistochemistryThe specimens, including control tissues of nor-

mal kidney, scar in regenerative phase, and normalskin, had been fixed in formalin and embedded inparaffin blocks according to standard procedures.Sections 4 to 6 �m in thickness were applied toslides, deparaffinized in xylene, hydrated throughgraded alcohols and washed. Antigen unmaskingwas performed via heat treatment for 20 minuteswith 10-mM sodium citrate buffer at 958C. Sectionswere then incubated at room temperature in 1.5%normal blocking serum in TRIS buffer for 20 minutes.After removal of blocking serum, primary antibodywas applied and incubated overnight at 48C. Theprimary antibodies included ACE [2E2] (sc-23908;Santa Cruz Biotechnology, Inc), TGF-b1 [TB21] (sc-52893; Santa Cruz Biotechnology, Inc), AT1 [N-10](sc-1173; Santa Cruz Biotechnology, Inc.) andp-Smad2/3 [Ser 433/435] (sc-11769; Santa CruzBiotechnology, Inc.) at dilutions of 1:25, 1:1000,1:500, and 1:500, respectively. Dilutions were basedon the manufacturer’s recommendations with ad-justments made by using control tissue of normalskin, scar, and kidney specimens. On day 2, primaryantibody was washed with 3 changes of phosphate-buffered saline (solution) (PBS) and then incubatedwith secondary antibody for 30 minutes. After againwashing with PBS, avidin-biotin enzyme reagent wasapplied for 30 minutes. Next, a Vulcan Fast RedBuffer was applied for 25 minutes after washing.Slides were then counterstained with hematoxylin,washed with deionized water, dehydrated throughgraded alcohols and xylene followed by coverslip

Table I. Review of 8 pa

Patient

Specimen

No.

A

Et

E* AT1* TGF-b* Smad* Fibrosis Notes

A 1 3 � � 1 11

2 � 111 11 1111

B 3 6 � � 1 111

C 4 4 � 11 � 11 Deeper, morefibrotic area lostin processing

D 5 1 � 1 111 111

6 � 1 1111 1111

E 7 3 � � � 111

8 � � � � Uninvolved skinF 9 8 � 111 1111 1111

10 � 111 1 11

G 11 56/ � � � 1 Clinicallyuninvolved skin

12 1 � 11 11 1111

H 13 3 � 1 � 111

AA, African American; ACE, an Latina American; TGF-b, transforming growth factor beta; W, white.

*Scale: e, no staining; 1, \2 ing.

JA

MA

CA

DD

ER

MA

TO

L

VO

LU

ME

58,N

UM

BER

6

Kelly,

Petitt,

an

dSa

nch

ez1027

tients with nephrogenic fibrosing dermopathy

ge (y)/

hnicity/

Sex

Gd dose

(mmol/kg) Gd in tissue

Time (mo) from

exposure

to biopsy

Time from

symptom

onset to biopsy AC

3/AA/F 0.28 Yes 4 2 mo �No �

2/AA/F 0.16 Yes 7 7 mo �2/LA/M 0.26 Yes 1-2 1 mo �

4/LA/F 0.26 No 4-5 2 mo �No �

8/LA/F 0.19 Yes 4 1 mo 1

Yes �7/W/M 0.43 No 2 2 wk �

Yes �Asian/M 0.24 Yes 3 3 mo �

Yes 11

7/AA/F ? N/A ? ? �

giotensin-converting enzyme; AT1, angiotensin II receptor 1; Gd, gadolinium; LA, Latino/

5% staining; 11, 25%-50% staining; 111, 50%-75% staining; 1111, 75%-100% stain

J AM ACAD DERMATOL

JUNE 2008

1028 Kelly, Petitt, and Sanchez

application. Negative controls from each paraffinblock were prepared in the same manner, with theexception of PBS used in place of the primaryantibody. Interpretation was made by using a semi-quantitative scale with the following values: e (nostaining), 1 (\25% staining), 11 (25%-50% stain-ing), 111 (50%-75% staining), and 1111 (75%-100% staining).

RESULTSIn our normal skin sample, AT1 and ACE cyto-

plasmic staining was seen in the epidermis, particu-larly the basement membrane, sebaceous glands,eccrine glands, and in dermal vessel walls. Thesefindings were similar to those of other studies.17,18

No staining for AT1 or ACE was seen in the scarsample. The kidney revealed cytoplasmic staining inthe kidney tubule cells.

TGF-b staining in our controls, including normalskin, scar in remodeling phase, and kidney, showedstaining of extravasated and intravascular red bloodcells with only mild, focal staining of the scar. Smad2/3 staining showed nuclear staining in the spindlecells of the scar and approximately 50% of the kidneytubules.

In 8 of 11 samples of NSF tested, TGF-b wasdetected in the spindle cells (Fig 1). Smad 2/3 nuclearstaining was seen in 8 of 11 samples (Fig 2). Six of the11 stained for both TGF-b and Smad 2/3 (Table I).

Conversely, only faint staining for ACE was seenin the spindle cells in 2 of the 11 specimens. Thesetwo specimens also showed more intense staining ofthe endothelial cells. No AT1 staining was seen in anyof the specimens.

There was no correlation with any of the stains tothe intensity of CD34 staining (data not shown).There did appear to be a positive trend with thedose of gadolinium received and TGF-b expression,with all patients receiving over 0.2 mmol/kg ofgadolinium, showing TGF-b expression in 7 of 8

Fig 1. TGF-b staining pattern (with red chromagen) inNSF. (Original magnification: 3200.)

biopsy specimens. Neither of the two biopsyspecimens of patients receiving doses below0.2 mmol/kg of gadolinium showed TGF-b expres-sion. One patient’s gadolinium dose was unknown(see Table I).

DISCUSSIONTGF-b has previously been found in skin and

fascia samples of patients with NSF by in-situ hy-bridization.19 Our results further suggest the associ-ation with TGF-b and fibrosis in NSF, with anadditional suggestion of Smad involvement as asecond messenger. TGF-b is believed to be a centralmediator in fibrosis as it induces fibroblasts tosynthesize and contract the extracellular matrix. Ifindeed TGF-b is an important participator in theprocess of fibrosis, the mechanisms that lead to itsproduction are unclear.

Relationship between the renin-angiotensin sys-tem and fibrosis in renal and cardiovascular tissuehas been well described.20 Up-regulated ACE levelshave been observed in many fibrous disorders, suchas cardiac fibrosis, renal fibrosis, idiopathic pulmo-nary fibrosis, asbestosis, and silicosis.

There is a suggestion that angiotensin (Ang) II andTGF-b do not act independently of each other, atleast in the process of cardiac remodeling.14

Numerous studies have shown that the renin-angio-tensin system is activated in response to hemody-namic overload and that it contributes to myocardialhypertrophy, fibrosis, and dysfunction. Clinical trialshave also shown that inhibition of Ang II by ACE-inhibitors prevents or reverses ventricular remodel-ing and improves survival in patients with heartfailure.21 There is indirect evidence for a functionallink between Ang II and TGF-b1 in the heart. In vitrostudies show that TGF-b messenger RNA and proteinare up-regulated by Ang II in cardiac fibroblasts,myofibroblasts and myocytes. Ang II blockade by anAT1 receptor antagonist reverses both cardiac TGF-b

Fig 2. Smad 2/3 nuclear staining pattern (with redchromagen) in NSF. (Original magnification: 3400.)

J AM ACAD DERMATOL

VOLUME 58, NUMBER 6

Kelly, Petitt, and Sanchez 1029

expression and myocardial hypertrophy/fibrosis inrat models in vivo. More direct evidence of a causalrelationship of Ang II and TGF-b with respect tofibrosis came in a hallmark paper by Schultz et al22

showing that knockout mice deficient in TGF-b(TGF-b�/�)could not produce cardiac hypertrophyin the presence of Ang II compared to wild-type(TGF-b1/1) mice who did develop cardiac hyper-trophy. Additionally, treatment of scleroderma renalcrisis, as well as other fibrosing diseases of thekidney, with ACE inhibitors has been shown todramatically improve prognosis.23

This relationship among fibrosis, the renin-angi-otensin system, and TGF-b in cutaneous diseases isless clear, and it is unknown whether drugs shown toprevent fibrosis in other organs (eg, ACE inhibitors)would apply to the skin. A locally functioning tissuerenin-angiotensin system, however, has been de-scribed in the skin.18 Immunohistochemistry hasshown increased angiotensinogen in skin from pa-tients with systemic sclerosis, but not in normalskin.16 Cutaneous tissue ACE activity as well asangiotensin receptors have been found to be in-creased in both hypertrophic scars and keloids.17

Interestingly, they were most concentrated in myo-fibroblasts. Some studies have shown that NSFlesions do not have increased myofibroblasts, per-haps suggesting that the fibrocytes are not differen-tiating as they do in normal scars.24

Our study did not find evidence of an AT1 byimmunohistochemistry. It is possible that the renin-angiotensin system still plays a role in NSF, perhapsby another receptor. AT2, for example, has beenshown to be involved in mechanisms other thanregulation of blood pressure and body fluids, in-cluding proliferation control, cell differentiation,apoptosis, tissue repair, and tissue remodeling.18

Controversy exists, however, about the presence ofAT2 in the skin. Some studies have found AT2 in theskin,18 whereas others have not.25 Alternatively, ourtechnique may not have been sufficiently sensitive todetect AT1 activity.

Interestingly, Fazeli, Lio, and Liu6 suggested thatthe lack of ACE inhibitors in dialysis patients maycontribute to the development of NSF. None of the7 patients in their series with NSF were receiving ACEinhibitors. In the series of Grobner,7 only 1 of 5 wasreceiving an ACE-inhibitor. In our study, only 1 of the8 patients was receiving an ACE inhibitor.

Progress continues to be made in the under-standing of the pathogenesis of TGF-beinducedfibrosis. TGF-b activators include matrix metallo-proteinases MMP-2 and MMP-9, thrombospondin-1,plasmin, and integrin avb6. Activated TGF-b canthen bind to a receptor complex, which will lead to

phosphorylation of Smad2/3. These phosphory-lated mediators can bind with Smad4 and translo-cate to the nucleus. Smad 6 and Smad 7 act asinhibitory proteins to the above mechanism. TGF-bcan inhibit its own actions by the induction ofSmad7.13,26 A recent article suggested that TGF-bactivation may occur by transglutaminase 2, whichshowed increased expression in NSF samples.27

Targeting some of these cytokines may be impor-tant for the discovery of future treatment options.TGF-b has many complex physiologic roles inhumans in addition to its profibrotic properties,including suppression of the immune response andepithelial proliferation. However, caution should betaken in any attempt to block this cytokine. Dentonet al,28 for example, showed that CAT-192, a recom-binant human antibody against TGF-b caused in-creased adverse events and mortality when used inpatients with early-stage systemic sclerosis.Targeting Smad3 may also be problematic as Smad3-mutant mice can develop degenerative joint dis-ease and showed chronic inflammation and colorec-tal adenocarcinomas when exposed to TGF-b.29

In addition to the small number of cases pre-sented, a shortcoming of our findings is that we didnot perform our studies on a cohort of comparablepatients with renal dysfunction who did not haveNSF.

In conclusion, TGF-b, as well as its second mes-sengers Smad 2/3, appear to be associated with thefibrosis seen in NSF. No definitive evidence of renin-angiotensin system involvement could be deter-mined in our study.

REFERENCES

1. Cowper SE, Robin HS, Steinberg SM, Su LD, Gupta S, LeBoit PE.

Scleromyxoedema-like cutaneous diseases in renal-dialysis

patients. Lancet 2000;356:1000-1.

2. Cowper SE. Nephrogenic fibrosing dermopathy [NFD/NSF

Website]. 2001-2007. Available at: http://www.icnfdr.org.

Accessed October 7, 2007.

3. Saenz A, Mandal R, Kradin R, Hedley-Whyte ET. Nephrogenic

fibrosing dermopathy with involvement of the dura mater.

Virchows Archiv 2006;449:389-91.

4. Quan TE, Cowper S, Wu SP, Bockenstedt LK, Bucala R.

Circulating fibrocytes: collagen-secreting cells of the periphe-

ral blood. Int J Biochem Cell Biol 2004;36:598-606.

5. Swaminathan S, Ahmed I, McCarthy JT, Albright RC, Pittelkow MR,

Caplice NM, et al. Nephrogenic fibrosing dermopathy and high-

dose erythropoietin therapy. Ann Intern Med 2006;145:234-5.

6. Fazeli A, Lio PA, Liu V. Nephrogenic fibrosing dermopathy: are

ACE inhibitors the missing link? [letter] Arch Dermatol 2004;

140:1401.

7. Grobner T. Gadolinium—a specific trigger for the develop-

ment of nephrogenic fibrosing dermopathy and nephrogenic

systemic fibrosis? Nephrol Dial Transplant 2006;21:1104-8.

8. Marckmann P, Skov L, Rossen K, Dupont A, Damholt MB, Heaf

JG, et al. Nephrogenic systemic fibrosis: suspected causative

J AM ACAD DERMATOL

JUNE 2008

1030 Kelly, Petitt, and Sanchez

role of gadodiamide used for contrast-enhanced magnetic

resonance imaging. J Am Soc Nephrol 2006;17:2359-62.

9. U.S. Food and Drug Administration. Public health advisory:

Gadolinium-containing contrast agents for magnetic reso-

nance imaging (MRI): Omniscan, OptiMark, Magnevist, Pro-

Hance, and MultiHance. June 8, 2006. Available at:

http://www.fda.gov/cder/drug/advisory/gadolinium_agents.

htm. Accessed October 8, 2007.

10. High WA, Ayers RA, Chandler J, Zito G, Cowper SE. Gadolinium

is detectable within the tissue of patients with nephrogenic

systemic fibrosis. J Am Acad Dermatol 2007;56:21-6.

11. Boyd AS, Zic JA, Abraham JL. Gadolinum deposition in

nephrogenic fibrosing dermopathy. J Am Acad Dermatol

2007;56:27-30.

12. Wiginton CD, Kelly B, Oto A, Jesse M, Aristimuno P, Ernst R,

Chaljub G. Gadolinium-based contrast exposure, nephrogenic

systemic fibrosis and gadolinium detection in tissue. AJR

2008;190:1060-8.

13. Verrecchia F, Mauviel A. Tranforming growth factor-b and

fibrosis. World J Gastroenterol 2007;13:3056-62.

14. Rosenkranz S. TGF-b1 and angiotensin networking in cardiac

remodeling. Cardiovasc Res 2004;63:423-32.

15. Remuzzi G, Ruggenenti P, Perico N. Chronic renal diseases:

renoprotective benefits of renineangiotensin system inhibi-

tion. Ann Intern Med 2002;136:604-15.

16. Kawaguchi Y, Takagi K, Hara M, Fukasawa C, Sugiura T,

Nishimagi E, et al. Angiotensin II in the lesional skin of systemic

sclerosis patients contributes to tissue fibrosis via angiotensin II

Type 1 receptors. Arthritis Rheum 2004;50:216-26.

17. Morihara K, Takai S, Takenaka H, Sakaguchi M, Okamoto Y,

Morihara T, et al. Cutaneous tissue angiotensin-converting

enzyme may participate in pathologic scar formation in

human skin. J Am Acad Dermatol 2006;54:251-7.

18. Steckelings UM, Wollschlager T, Peters J, Henz BM, Hermes B,

Artuc M. Human skin: source of and target organ for angio-

tensin II. Exp Dermatol 2004;13:148-54.

19. Jimenez SA, Artlett CM, Sandorfi N, Derk C, Latinis K, Sawaya H,

et al. Dialysis-associated systemic fibrosis (nephrogenic fibrosing

dermopathy): study of inflammatory cells and transforming

growth factor b1 expression in affected skin. Arthritis Rheum

2004;50:2660-6.

20. Pfeffer MA, Braunwald E, Moye LA, Basta L, Brown EJ Jr, Cuddy

TE, et al. Effect of captopril on mortality and morbidity in

patients with left ventricular dysfunction after myocardial

infarction: results of the survival and ventricular enlargement

trial. The SAVE Investigators. N Engl J Med 1992;327:669-77.

21. Pfeffer MA, McMurray JJV, Velazquez EJ, Rouleau JL, Køber L,

Maggioni AP, et al. Valsartan, captopril, or both in myocardial

infarction complicated by heart failure, left ventricular dys-

function, or both. N Engl J Med 2003;349:1893-906.

22. Schultz JEJ, Witt SA, Glascock BJ, Nieman ML, Reiser PJ, Nix SL,

et al. TGF-b1 mediates the hypertrophic cardiomyocyte growth

induced by angiotensin II. J Clin Invest 2002;109:787-96.

23. Steen VD, Costantino JP, Shapiro AP, Medsger TA Jr. Outcome

of renal crisis in systemic sclerosis: relation to availability of

angiotensin converting enzyme (ACE) inhibitors. Ann Intern

Med 1990;113:352-7.

24. Pryor JG, Poggioli G, Galaria N, Gust A, Robison J, Samie F,

et al. Nephrogenic systemic fibrosis: a clinicopathologic study

of six cases. J Am Acad Dermatol 2007;57:105-11.

25. Tekeda H, Katagata Y, Kondo S. Immunohistochemical study

of angiotensin receptors in human anagen hair follicles and

basal cell carcinoma. Br J Dermatol 2002;147:276-80.

26. Leask A. Scar wars: is TGFb the phantom menace in sclero-

derma? Arthritis Res Ther 2006;8:213.

27. Parsons AC, Yosipovitch G, Sheehan DJ, Sangueza OP, Green-

berg CS, Sane DC. Transglutaminases: the missing link in

nephrogenic systemic fibrosis. Am J Dermatopathol 2007;29:

433-6.

28. Denton CP, Merkel PA, Furst DE, Khanna D, Emery P, Hsu VM,

et al. Recombinant human anti-transforming growth factor

beta 1 antibody therapy in systemic sclerosis: a multicenter,

randomized, placebo-controlled phase I/II trial of CAT-192.

Arthritis Rheum 2007;56:323-33.

29. Yang X, Letterio JJ, Lechleider RJ, Chen L, Hayman R, Gu H,

et al. Targeted disruption of SMAD3 results in impaired

mucosal immunity and diminished T cell responsiveness to

TGF-beta. EMBO J 1999;18:1280-91.