Journal of Human Hypertension (2007) 21, 393–400 & 2007 Nature Publishing Group All rights reserved 0950-9240/07 $30.00 www.nature.com/jhh ORIGINAL ARTICLE Inheritance of arterial lesions in renal fibromuscular dysplasia J Perdu1,2, P Boutouyrie3,4,5,10, C Bourgain6,7,10, N Stern1,8, B Laloux4, E Bozec4, M Azizi3,8, C Bonaiti-Pellie´6,7, P-F Plouin2,3,9, S Laurent3,4,5, A-P Gimenez-Roqueplo1,2,3 and X Jeunemaitre1,2,3 1AP-HP, Department of Genetics, Hoˆpital Europe´en Georges Pompidou, Paris, France; 2INSERM, Unit 772, Colle`ge de France, Paris, France; 3Faculte´ de Me´decine, Universite´ Paris Descartes, Paris, France; 4AP-HP, Department of Pharmacology, Hoˆpital Europe´en Georges Pompidou, Paris, France; 5INSERM, Unit 337, Paris, France; 6INSERM, Unit 535, Villejuif, France; 7Universite´ Paris-Sud, Villejuif, France; 8AP-HP, Hoˆpital Europe´en Georges Pompidou, Clinical Investigation Center CIC9201, Paris, France and 9AP-HP, Department of Hypertension, Hoˆpital Europe´en Georges Pompidou, Paris, France We have previously shown that patients with renal cases had also a high carotid score (4.17), very fibromuscular dysplasia (FMD) have asymptomatic significantly higher than that of controls (2.52, Po10À5) carotid lesions and that familial forms may occur. The even though lower than the corresponding index FMD objective of this study was to test whether carotid cases (4.81, P ¼ 0.01). Segregation analysis showed that lesions could be detected in relatives of familial cases. 52% of the descendants of subjects with a score 44 had High-resolution echotracking of the carotid artery was a score 44, a proportion consistent with autosomal- performed in 47 relatives of 13 cases from six families. dominant transmission of the trait. Altogether these This non-invasive investigation led to a semiquantitative results strengthen the hypothesis of renal FMD being arterial score that was compared with that obtained for a systemic arterial disease and argue for a familial 47 controls matched for age and sex and that for 125 resemblance that may be due to a major genetic effect. sporadic cases. Familial resemblance was tested by The carotid score obtained by high-resolution echo- using a generalized estimating equation approach tracking may provide a non-invasive surrogate marker taking into account the clustering of scores in families. for renal FMD of potential value for use in linkage As expected, FMD cases had a significantly higher score strategies on large pedigrees. than controls (4.02 vs 2.52, Po10À5). Familial cases were Journal of Human Hypertension (2007) 21, 393–400. not significantly different from sporadic cases. Of doi:10.1038/sj.jhh.1002156; published online 1 March 2007 interest, the 47 apparently healthy relatives of familial Keywords: genetics; fibromuscular dysplasia; renovascular; echotracking; arterial stiffness; generalized estimating equation approach Introduction and saccular aneurysms. In medial FMD, renal artery lesions usually present as multiple, multi- Renal artery fibromuscular dysplasia (FMD) is a focal stenoses with intervening aneurysms causing rare, non-atherosclerotic, non-inflammatory vascu- a ‘string of beads’ appearance, whereas unifocal lar disease of truncal or branch kidney arteries, of stenoses are usually associated with intimal FMD. If unknown pathophysiology. In medial FMD, the renal artery stenoses are severe, FMD may result in most prevalent form of FMD, pathologic examina- renovascular hypertension and, if dissection occurs, tion shows successive areas of parietal thickening in renal infarction.1,2 (corresponding to medial hyperplasia) and dilata- Medial FMD occurs predominantly in middle- tion (following disruption of internal elastic lami- aged Caucasian women, and a number of risk factors na). These arterial alterations may lead to stenoses have been identified in this group of women: smoking habits, oral contraceptive intake and genetic factors.1,3–5 Medial FMD may present as a Correspondence: Dr J Perdu, Department of Genetics, Hoˆpital familial disease, defined as the presence of angio- Europe´en Georges Pompidou, 20-40, rue Leblanc, 75908 Paris graphically documented FMD in at least two first- Cedex 15, France. degree relatives.6 It has also been suggested that E-mail: [email protected] 10These authors contributed equally to this work. FMD is transmitted as an autosomal-dominant Received 18 September 2006; revised 28 November 2006; disease with incomplete penetrance and variable accepted 24 December 2006; published online 1 March 2007 clinical symptoms.7,8 Evidence of arterial wall subclinical lesions in relatives of FMD pedigrees J Perdu et al 394 Identifying the genes responsible for FMD is Methods therefore an appropriate strategy for investigation of the pathophysiology of this disease. However, this Patients and subjects approach requires the identification of a reliable Fourteen (6.1%) of the 228 FMD patients referred to phenotype for separating unaffected and affected the Hypertension Unit of our institution since 1986 subjects, even in the absence of symptoms. Renal were considered to correspond to familial cases artery duplex sonography may be not sensitive because at least one of the first-degree relatives of enough, especially in cases of non-stenotic lesions1 these patients (the sister in 13 cases) also had and computed tomographic or magnetic resonance angiographically proven renal FMD. Three of the 14 angiography have a poor specificity in patients with families concerned were impossible to trace, one FMD.9 Renal angiography is an invasive test and could not be studied for geographical reasons and would be unethical to use for the characterization two consisted of two FMD cases with no living of normotensive members of a pedigree. We recently relatives. The relatives of the FMD cases in the eight showed, with carotid artery high-resolution echo- remaining families were asked to participate in a tracking techniques, that FMD is actually a systemic clinical research protocol based on a routine clinical disease and that subclinical lesions of the common examination, together with arterial echotracking and carotid artery (CCA) wall are commonly found in blood sampling for DNA extraction and analysis. patients with renal artery medial FMD. An arterial These eight families contained 113 eligible relatives, phenotypic score of 2 (normal) to 7 (highly abnor- of whom 26 refused to participate, 23 could not be mal) has been shown to be strongly associated with contacted and 17 were willing to participate but FMD.10 could not because of geographical constraints. The In this study, we used this score as a surrogate final analysis included 47 Caucasian relatives from marker for renal FMD phenotype, in the character- six pedigrees studied in detail (Figure 1). These ization of first-degree relatives of affected subjects. subjects were compared with 47 normotensive We selected pedigrees containing at least two Caucasian control subjects matched for age and patients with angiographically proven renal FMD. sex, who were recruited during the same period We showed, using a generalized estimating equation among the medical staff members and their families. (GEE) approach, that relatives of renal FMD familial Control subjects known to be hypertensive were cases have significantly higher arterial score values excluded. They underwent the same procedures, than matched controls, suggestive of the presence of carried out by the same trained technicians (EB, BL), subclinical arterial lesions. We performed a segrega- as FMD patients and relatives. The study was tion analysis in six families, treating arterial score approved by our institutional ethics committee as a dichotomous trait. We found that ‘score 44’ was (Approval CCPPRB Paris-Cochin # RBM 00–028) a heritable trait with an autosomal-dominant mode and all subjects gave written informed consent. As of inheritance. previously described, for a part of this cohort FMD 004 FMD 008 FMD 143 ? ? ? ND ND ND ND ND ND ? ? ND ND 4.90 ND 6.75 4.16 4.50 6.00 4.08 5.17 3.70 3.08 3.75 4.00 ND 5.50 3.67 3.71 3.94 5.80 2.00 4.37 2.66 4.00 4.00 3.50 5.00 ? FMD 125 ? FMD 074 ?? 4.50 4.25 2.34 4.04 ND ND ND ND ND ND FMD 018 ? 5.75 4.21 3.66 4.83 ND ND 3.00 4.00 ND 7.00 4.75 5.00 5.16 ND 4.90 4.83 4.80 4.91 5.55 5.91 5.00 2.00 4.83 4.33 4.50 4.58 3.25 4.66 4.33 3.50 4.16 2.08 5.16 2.16 6.83 2.33 2.00 FMD-YES HT-YES HT-NO? HT-UNK Figure 1 Structure of the six renal FMD families. Arterial phenotypic score is presented beneath each subject icon. HT: hypertension; UNK: unknown; FMD: renal fibromuscular dysplasia. Journal of Human Hypertension Evidence of arterial wall subclinical lesions in relatives of FMD pedigrees J Perdu et al 395 (N ¼ 104), each FMD patient underwent a standar- included in the regression are independent (con- dized medical evaluation.6 The procedures followed trols, sporadic cases), this is strictly equivalent to a were in accordance with institutional guidelines. Student’s t-test for comparing mean values of the score in the two groups, adjusted for the covariates. Echotracking measurements When several related individuals are included This investigation was performed in a controlled (multiple cases or multiple relatives from a parti- environment kept at þ 22711C, after the subject had cular family), tests should account for it. Indeed, remained in a recumbent position for 15 min. Right as we suspect a familial correlation of the score, CCA was studied 2 cm beneath the carotid bifurca- neglecting the relatedness of the individuals would tion, with a 7.5 MHz linear array ultrasound probe misestimate the variance of the regression para- (Pie Medical 350, Maastricht, The Netherlands) meters and artificially either increase or decrease coupled with an echotracking system (Walltrack the significance of the tests.