uva-dare (digital academic repository) familial
TRANSCRIPT
UvA-DARE is a service provided by the library of the University of Amsterdam (http://dare.uva.nl)
UvA-DARE (Digital Academic Repository)
Familial hypercholesterolemia: the Dutch approach
Huijgen, R.
Link to publication
Citation for published version (APA):Huijgen, R. (2012). Familial hypercholesterolemia: the Dutch approach
General rightsIt is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s),other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).
Disclaimer/Complaints regulationsIf you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, statingyour reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Askthe Library: http://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam,The Netherlands. You will be contacted as soon as possible.
Download date: 08 Feb 2018
Chapter 6
Cardiovascular Risk in Relation to
Functionality of Sequence Variants in
the Gene Coding for the Low-Density
Lipoprotein Receptor
A study among 29,365 individuals tested for 64 specific LDLR sequence variants
Roeland Huijgen, Iris Kindt, Joep C. Defesche and John J.P. Kastelein
European Heart Journal 2012;33(18):2325-2330.
R1
R2
R3
R4
R5
R6
R7
R8
R9
R10
R11
R12
R13
R14
R15
R16
R17
R18
R19
R20
R21
R22
R23
R24
R25
R26
R27
R28
R29
R30
R31
R32
R33
R34
Chapter 6
106
AbSTRACT
Background: A plethora of mutations in the LDL-receptor gene (LDLR) underlie the clinical phenotype of familial hypercholesterolemia (FH). For the diagnosis of FH, it is important, however, to discriminate between pathogenic and non-pathogenic mutations. The aim of the current study was to assess whether true pathogenic mutations were indeed associated with the occurrence of coronary artery disease (CAD) when compared to non-functional variants. The latter variants should not exhibit such an association with CAD.
Methods: We assessed 29,365 individuals tested for the 64 most prevalent LDLR variants. First, we determined pathogenicity for each of these sequence variants. Subsequently, a Cox-proportional hazard model was used to compare event-free survival, defined as the period from birth until the first coronary artery disease (CAD) event, between carriers and non-carriers of LDLR mutations.
Results: Fifty-four sequence variants in the LDLR gene were labeled as pathogenic and 10 as non-pathogenic. The 9,912 carriers of a pathogenic LDLR mutation had a shorter event-free survival than the 18,393 relatives who did not carry that mutation; hazard ratio 3.64 (95%CI: 3.24 to 4.08; p<0.001). In contrast, the 355 carriers of a non-pathogenic LDLR variant had similar event-free survival as the 705 non-carrying relatives; hazard ratio 1.00 (95%CI: 0.52 to 1.94; p=0.999).
Conclusions: These findings with respect to clinical outcomes substantiate our criteria for functionality of LDLR sequence variants. They also confirm the CAD risk associated with FH and underline that these criteria can be used to decide whether a specific sequence variant should be used in cascade screening
R1
R2
R3
R4
R5
R6
R7
R8
R9
R10
R11
R12
R13
R14
R15
R16
R17
R18
R19
R20
R21
R22
R23
R24
R25
R26
R27
R28
R29
R30
R31
R32
R33
R34
Cardiovascular risk and functionality criteria for the LDLR
107
6
InTRODuCTIOn
Familial hypercholesterolemia (FH) is a frequent autosomal co-dominant disorder of lipoprotein metabolism, with 1 in 500 persons affected with heterozygous FH in most Western countries.1 Patients with FH have elevated plasma low-density lipoprotein-cholesterol (LDL-C) levels and an increased risk of premature coronary artery disease (CAD).1, 2 Defects in genes that code for proteins involved in hepatic clearance of LDL-C underlie this hereditary disorder.1 In fact, more than a 1000 different mutations in the genes coding for the LDL-receptor (LDLR), apolipoprotein B (APOB) and proprotein convertase subtilisin/kexin type 9 (PCSK9) are now known to cause FH.3
The identification of a mutation that underlies FH in a particular family enables genetic testing of family members for the presence of the same mutation and makes it possible to initiate effective medical management before the cardiovascular consequences of FH become clinically manifest. This notion has led to the implementation of a nationwide genetic cascade screening program for FH in the Netherlands and approximately 27,000 subjects with FH have been found and treated since 1994.4
However, as widely appreciated, not every sequence variant in LDLR, APOB or PCSK9 results in an FH phenotype. If a novel sequence variant in one of those genes is identified in a patient with a clear clinical FH diagnosis, one has to rely on in silico or in vitro studies to ascertain pathogenicity. However, in vitro studies are cumbersome and therefore rarely performed. Also, the in silico results do not always correspond with the clinical observations.5 An alternative strategy is to perform cascade screening for such a novel variant and to identify new carriers of that specific variant to determine whether these carriers express an FH phenotype. Accordingly, we recently validated specific criteria that can discriminate pathogenic from non-pathogenic mutations, using lipoprotein and lipid levels as well as the use of lipid-lowering medication.5 Since then, collected a much larger cohort of carriers of non-pathogenic variants in order to assess possible consequences of these variants on occurrence of cardiovascular disease.
In the present study, we compared cardiovascular risk between individuals with established pathogenic FH mutations and those with sequence variants in LDLR that we consider non-pathogenic. Here, we present our results.
R1
R2
R3
R4
R5
R6
R7
R8
R9
R10
R11
R12
R13
R14
R15
R16
R17
R18
R19
R20
R21
R22
R23
R24
R25
R26
R27
R28
R29
R30
R31
R32
R33
R34
Chapter 6
108
METHODS
In this observational study we used a stepwise approach. First, we applied our criteria to assess functionality of all variants we have identified in LDLR in order to expand the number of individuals with pathogenic and non-pathogenic variants. Second, we compared CAD risk between carriers and non-carriers of all non-pathogenic variants and, similarly, of all pathogenic mutations.
Study populationThe functionality criteria and event free survival were assessed in individuals who were screened by genetic testing in the period between January 1994 and December 2010. Of these subjects, lipoprotein profiles, an extensive history of cardiovascular disease, medication use, and specific carrier status were collected at the time of molecular diagnosis. We excluded index patients from the current analysis to avoid clinical sampling bias. The cascade screening program was launched by the Ministry of Health of the Dutch Government and approved by the National Ethics Committee. All participants gave written informed consent for genetic analysis.
Criteria to test functionalityThe three criteria to establish functionality of a specific DNA variant have been described before 5 and were: (1) a mean LDL-C level above the 75th percentile for age and gender in untreated individuals carrying such a mutation, (2) statistically significantly higher mean LDL-C levels in untreated carriers compared to untreated non-carriers, and (3) a statistically significant higher percentage of medication users among carriers compared to non-carriers at the time of molecular screening. We considered a mutation to be non-pathogenic when none of the three criteria were met.
All genetic variants had to be tested in at least 50 untreated subjects per variant. Additionally, the in silico analyses were applied to all prevalent variants by using the Alamut software (version 2, Interactive Biosoftware, Rouen, France: SIFT and PolyPhen prediction). Mutations were described according to the nomenclature as proposed by den Dunnen and Antonarakis,6 and we used the description based on the effect of the mutation on the protein.
Coronary artery diseaseCAD was defined as a history of one of the following non-fatal cardiac outcomes: a) sudden cardiac arrest; b) myocardial infarction; c) coronary artery bypass graft, or d)
R1
R2
R3
R4
R5
R6
R7
R8
R9
R10
R11
R12
R13
R14
R15
R16
R17
R18
R19
R20
R21
R22
R23
R24
R25
R26
R27
R28
R29
R30
R31
R32
R33
R34
Cardiovascular risk and functionality criteria for the LDLR
109
6
percutaneous transluminal coronary angioplasty. CAD event-free survival was defined as the period from year of birth until the year of the first CAD event or censoring at the moment of genetic screening.
Statistical analysisDifferences in LDL-C levels between subgroups were compared by means of the unpaired Student t-test. Proportions of subjects using lipid-lowering medication were compared with the chi-square test. We compared survival of mutation carriers with that of family members without mutations using Kaplan-Meier survival analysis. A Cox-proportional hazard model was used to compare CAD risk between the carriers and non-carriers. The non-carriers served as reference group. We adjusted for gender, hypertension, diabetes, smoking and body mass index. The Cox-proportional hazard analyses were repeated in R, taking also family ties into account. Sensitivity and specificity were calculated for the variants for which in silico analysis could be performed, using the described criteria as a gold standard for pathogenicity. 5 A two sided p-value < 0.05 was considered statistically significant. Data were analyzed with SPSS for Windows 16.0.2 (Chicago, IL).
Results
Test of functionality of new sequence variantsIn our study cohort, we extended previous findings with 12 sequence variants that had been tested in more than 50 untreated subjects just by 2010.5 All these variants reside in the LDLR and are presented in Table 1.3, 7-11 Three variants involve exonic deletions, whereas the other mutations result in amino acid substitutions.
Table 2 summarizes the in silico analysis and the assessment according to the criteria for all 12 variants. As observed, 11 mutations are pathogenic, whereas G701S (exon 14) is not pathogenic.
R1
R2
R3
R4
R5
R6
R7
R8
R9
R10
R11
R12
R13
R14
R15
R16
R17
R18
R19
R20
R21
R22
R23
R24
R25
R26
R27
R28
R29
R30
R31
R32
R33
R34
Chapter 6
110
Tab
le 1
: Nom
encl
atur
e of
mut
atio
ns in
LD
LR
Loc
atio
nn
ucl
eoti
de
chan
gen
ew n
omen
clat
ure
Old
nom
encl
atu
re b
Ref
eren
cec
Exo
n 1
c.1A
>G
p.M
et1V
alM
-21V
1, 3
Exo
n 3
c.26
8G>
Ap.
Asp
90A
snD
69N
2E
xon
4c.
622G
>A
p.G
lu20
8Lys
E18
7K1
Exo
n 4
c.66
2A>
Cp.
Asp
221G
lyD
200G
1E
xon
6c.
858C
>A
p.Se
r286
Arg
S265
R1
Exo
n 9
c.12
43G
>C
p.A
sp41
5His
D39
4H1
Exo
n 9
c.12
97G
>T
p.A
sp43
3Tyr
D41
2Y1
Exo
n 9
c.13
29G
>C
or
G>
Tp.
Trp4
43C
ysW
422C
1E
xons
12-
18c.
1705
+?_
2580
+?d
elD
elet
ion
exon
s 12
-18
16 k
b de
leti
on o
f exo
n 12
-18
(Cat
ania
-1)
2, 4
Exo
n 14
c.21
01G
>A
p.G
ly70
1Ser
G68
0S2
Exo
n 16
c.23
11+
?_23
90+
?del
Del
etio
n ex
on 1
62.
0 kb
del
etio
n ex
on 1
6 (P
adov
a-2)
2, 5
Exo
ns 1
6-18
c.23
11+
?_25
80+
?del
Del
etio
n ex
ons
16-1
88.
0 kb
del
etio
n ex
on 1
6-18
2
Num
beri
ng o
f the
nuc
leot
ides
of t
he L
DL
R g
ene
was
bas
ed o
n th
e cD
NA
, wit
h +
1 be
ing
the
A o
f the
AT
G t
rans
lati
on in
itia
tion
cod
on. a N
ew n
ame
repr
esen
t nu
mbe
ring
of
the
codo
ns w
ith
the
init
iati
on c
odon
is
1. b O
ld n
ame
repr
esen
t nu
mbe
ring
of
the
codo
ns w
ith
init
iati
on c
odon
is
-21
for
LD
LR
. Ref
eren
ce s
eque
nce
LD
LR
: Gen
Ban
k A
cces
sion
NM
_000
527.
3. c R
efer
ence
s of
the
mut
atio
ns i
dent
ified
in
the
Net
herl
ands
: (1)
Fou
chie
r et
al
., H
um G
enet
109
:602
–615
, 20
01;
(2)
Fouc
hier
et
al.,
Hum
Mut
at 2
6:55
0–55
6, 2
005;
(3)
Lom
bard
i et
al.,
Cli
n G
enet
51:
430
-431
, 19
97;
(4)
Ber
toli
ni e
t al
., A
TV
B 1
9:40
8-41
8, 1
999;
(5) B
erto
lini
et
al.,
AT
VB
15:
81-8
8, 1
995.
R1
R2
R3
R4
R5
R6
R7
R8
R9
R10
R11
R12
R13
R14
R15
R16
R17
R18
R19
R20
R21
R22
R23
R24
R25
R26
R27
R28
R29
R30
R31
R32
R33
R34
Cardiovascular risk and functionality criteria for the LDLR
111
6
Tab
le 2
: Mut
atio
ns w
ith
findi
ngs
of in
sili
co r
esul
ts a
nd o
f the
app
lica
tion
of c
rite
ria
to a
sses
s pa
thog
enic
ity
In si
lico
ana
lysi
sIn
viv
o ob
serv
atio
nW
itho
ut c
hole
ster
ol lo
wer
ing
med
icat
ion
All
All
FH+
FH-
FH+
FH-
Sequ
ence
var
iant
SIFT
Pol
yPhe
nN
pLD
LLD
L-C
aLD
L-C
apb
%M
ed%
Med
p
M1V
not
tol.
prob
ably
5097
5.2
± 1
.13.
2 ±
1.2
<0.
001
576
<0.
001
D90
Nno
t to
l.pr
obab
ly51
793.
8 ±
1.0
3.0
± 0
.90.
021
5317
0.01
6E
208K
tole
rate
dbe
nign
5093
4.6
± 1
.03.
2 ±
1.1
0.00
149
14<
0.00
1D
221G
not
tol.
prob
ably
7289
4.9
± 1
.52.
9 ±
1.0
<0.
001
5221
0.00
1S2
86R
not
tol.
prob
ably
8484
4.5
± 1
.23.
0 ±
1.0
<0.
001
6313
<0.
001
D41
5Hno
t to
l.be
nign
5081
4.3
± 1
.33.
0 ±
0.8
<0.
001
3615
0.04
1D
433Y
not
tol.
prob
ably
5096
5.9
± 1
.43.
3 ±
0.9
<0.
001
6021
0.00
4W
443C
not
tol.
prob
ably
9095
5.4
± 1
.43.
0 ±
0.9
<0.
001
6212
<0.
001
Del
etio
n ex
ons
12-1
8-
-60
976.
4 ±
0.9
3.1
± 1
.0<
0.00
163
9<
0.00
1G
701S
not
tol.
poss
ibly
5036
2.9
± 0
.73.
1 ±
0.7
0.67
3117
0.17
Del
etio
n ex
on 1
6-
-80
965.
5 ±
0.8
3.0
± 1
.1<
0.00
185
9<
0.00
1
Del
etio
n ex
ons
16-1
8-
-87
925.
3 ±
1.6
2.9
± 1
.0<
0.00
167
23<
0.00
1
a LD
L-C
leve
ls (m
mol
/l) a
re e
xpre
ssed
as
mea
n ±
sta
ndar
d de
viat
ion.
b p-v
alue
s ba
sed
on c
ompa
riso
n of
LD
L-C
leve
ls b
etw
een
untr
eate
d ca
rrie
rs a
nd n
onca
rrie
rs.
The
in
sili
co a
naly
ses
coul
d on
ly b
e pe
rfor
med
for
mut
atio
ns t
hat
resu
lt i
n an
am
ino
acid
sub
stit
utio
n an
d no
t fo
r de
leti
ons.
Abb
revi
atio
ns f
or S
IFT:
no
t tol
., no
t tol
erat
ed. F
or P
olyP
hen:
pos
sibl
y, p
ossi
bly
dam
agin
g; p
roba
bly,
pro
babl
y da
mag
ing.
Oth
er a
bbre
viat
ions
: FH
+, L
DL
R m
utat
ion
carr
iers
; FH
-, n
onca
rrie
rs o
f an
LD
LR
var
iant
; LD
L-C
, lo
w-d
ensi
ty l
ipop
rote
in c
hole
ster
ol;
p, p
-val
ue;
pLD
L, m
ean
perc
enti
le f
or a
ge a
nd g
ende
r; %
Med
, pe
rcen
tage
cho
lest
erol
-low
erin
g m
edic
atio
n us
ers.
R1
R2
R3
R4
R5
R6
R7
R8
R9
R10
R11
R12
R13
R14
R15
R16
R17
R18
R19
R20
R21
R22
R23
R24
R25
R26
R27
R28
R29
R30
R31
R32
R33
R34
Chapter 6
112
Risk of coronary artery disease based on functionalityAll tested mutations combined amounted to a total of 64 sequence variants in the LDLR gene that could be labeled as either pathogenic or non-pathogenic.5 The characteristics of these variants including the in silico findings are shown in supplemental table 1. The total cohort consisted of 29,365 individuals, who were tested for the presence of one particular LDLR mutation (Table 3). This constitutes 83.3% of the overall population tested in the cascade screening program. Of these 29,365, 1,060 (3.6%) individuals were tested for one of the 10 non-pathogenic sequence variants, of whom 355 subjects carried that variant.
As can be expected, carriers of pathogenic mutations were more often treated with cholesterol lowering medication compared to carriers of non-pathogenic variants in LDLR (4,193/9,912 (42%) vs. 38/355 (11%); p<0.001). Also, the mean levels (± SD) of the highest total cholesterol levels measured over lifetime were significantly higher in the carriers of a pathogenic mutation than in those identified with a non-pathogenic variants (9.5 ± 3.1 vs 6.2 ± 4.3 mmol/L; p<0.001).
Figure 1 and 2 provide Kaplan Meier curves for CAD event free survival for individuals tested for pathogenic and non-pathogenic LDLR variants. The 9,912 carriers of a pathogenic LDLR mutation had on average a shorter event free survival than the 18,393 relatives who did not carry that mutation. The hazard ratio associated with carriership was 3.64 (95% confidence interval: 3.24 to 4.08; p<0.001) (Figure 1). In contrast, the Kaplan Meier curves overlapped between carriers of the non-pathogenic sequence variants and the relatives that did not carry the variants (Figure 2). In particular, the risk of CAD was similar for both carriers and their unaffected relatives: the hazard ratio was exactly 1.00 (95% confidence interval: 0.52 to 1.94; p=0.999). The analyses described above were also performed with R statistics - taking family ties into account - and these analyses yielded results that were in essence similar (data not shown).
The in silico analyses could be performed for 34 pathogenic mutations and 9 non-pathogenic variants according to the criteria,5 as shown in supplemental table 1. Of these 43 variants, 33 were amino acid substitutions that were not tolerated according to SIFT, indicating a non-pathogenic nature. The associated specificity was 0.22 (2 out of 9) and the sensitivity was 0.97 (33 out of 34). Polyphen identified 13 variants as benign, i.e. non-pathogenic. Consequently the specificity was 0.44 (4 out of 9) and the sensitivity 0.74 (25 out of 34).
R1
R2
R3
R4
R5
R6
R7
R8
R9
R10
R11
R12
R13
R14
R15
R16
R17
R18
R19
R20
R21
R22
R23
R24
R25
R26
R27
R28
R29
R30
R31
R32
R33
R34
Cardiovascular risk and functionality criteria for the LDLR
113
6
Tab
le 3
: Cha
ract
eris
tics
of t
he s
tudy
pop
ulat
ion L
DL
R m
utat
ion
yes
noP
atho
geni
c by
cri
teri
aal
lye
sno
all
p#
Num
ber
10,2
679,
912
355
19,0
97A
ge y
ears
**37
(22-
51)
37 (2
2-51
)42
(23-
55)
45 (3
1-57
)<
0.00
1M
ale
sex
n (%
)4,
941
(48%
)4,
773
(48%
)16
8 (4
7%)
9,06
1 (4
7%)
0.27
Bod
y m
ass
inde
x kg
/m2 *
23.7
± 4
.823
.6 ±
4.8
24.4
± 4
.724
.5 ±
4.5
<0.
001
Hig
hest
tot
al c
hole
ster
ol e
ver
mm
ol/L
* 9.
5 ±
3.1
9.5
± 3
.06.
2 ±
4.3
5.9
± 3
.7<
0.00
1St
atin
use
at
gene
tic
scre
enin
g n
(%)
4,23
1 (4
1%)
4,19
3 (4
2%)
38 (1
1%)
1,77
4 (9
%)
<0.
001
Cor
onar
y ar
tery
dis
ease
n (%
)69
8 (7
%)
683
(7%
)15
(4%
)67
6 (4
%)
<0.
001
**m
edia
n (b
orde
rs o
f qua
rtil
es),
*mea
n ±
sta
ndar
d de
viat
ion,
# p-v
alue
for
com
pari
son
of a
ll L
DL
R m
utat
ion
carr
iers
ver
sus
all n
on c
arri
ers.
LD
LR
=
low
-den
sity
lipo
prot
ein
rece
ptor
.
R1
R2
R3
R4
R5
R6
R7
R8
R9
R10
R11
R12
R13
R14
R15
R16
R17
R18
R19
R20
R21
R22
R23
R24
R25
R26
R27
R28
R29
R30
R31
R32
R33
R34
Chapter 6
114
Figure 1: Kaplan Meier curves for individuals tested for pathogenic LDLR mutations
Carrier of pathogenic LDLR mutation in red and non-carrier in blue
Figure 2: Kaplan Meier curves for individuals tested for non-pathogenic LDLR sequence variants
Carrier of non-pathogenic LDLR mutation in red and non-carrier in blue
R1
R2
R3
R4
R5
R6
R7
R8
R9
R10
R11
R12
R13
R14
R15
R16
R17
R18
R19
R20
R21
R22
R23
R24
R25
R26
R27
R28
R29
R30
R31
R32
R33
R34
Cardiovascular risk and functionality criteria for the LDLR
115
6
DISCuSSIOn
In this study, we first applied a set of specific criteria to determine functionality of the 64 most prevalent LDLR sequence variants used in the screening program in our country.5 Almost 30,000 individuals were tested for one of these variants and, as expected, carriers of pathogenic FH mutations had consistently elevated LDL-C levels. Moreover, FH patients with a pathogenic mutation had a more than 3 fold higher hazard ratio for CAD compared to unaffected relatives.
In contrast, lipid levels and the risk of CAD in individuals carrying a non-pathogenic LDLR variant were similar to those of their healthy relatives. As such, this provides further evidence that the proposed criteria to determine non-pathogenicity did label sequence variants correctly.
Only patients with a strong clinical suspicion of genetic FH are referred for analysis of LDLR mutations in the Netherlands. Once a novel sequence variant is identified in such an index patient, the clinical characteristics of the index patient will evidently not reveal whether that variant is functional or not. In fact, we showed previously that the clinical FH diagnostic scores were in essence not different for the index patients in whom an established pathogenic mutation was discovered from the index patients with a novel non-pathogenic variant in the LDLR.5 Thus, when in vitro tests are lacking and no segregation data is available yet, one often has to rely on the in silico analyses to decide whether further cascade screening should be performed for the novel LDLR variant. Table 1 and supplemental table 1 show that the findings in silico not correspond well with the co-segregation data, with a specificity below 50%. For example, the p.Gly701Ser mutation (c.2101 G>A) was identified as non-pathogenic rather convincingly by co-segregation while the SIFT and Polyphen indicated the amino acid substitution to be not tolerated and possibly damaging, respectively. Conversely, clinical data on the p.Glu208Lys mutation (c.662 G>A) showed that this mutation is clearly pathogenic, while the in silico analyses suggested that this was a non-pathogenic sequence variant. Thus, the clinical data from the index patients referred for genetic analysis combined with the in silico results are often not conclusive for determining functionality for novel LDLR sequence variants.
Therefore, in case the newly discovered sequence variant in LDLR was identified in a patient with a strong clinical suspicion of FH and the in silico analyses also hinted at pathogenic nature, the further approach was to include it in the genetic cascade screening. After a robust number of individuals was been tested for a variant that was assumed to cause FH, that variant was re-evaluated based on the criteria we proposed.5
R1
R2
R3
R4
R5
R6
R7
R8
R9
R10
R11
R12
R13
R14
R15
R16
R17
R18
R19
R20
R21
R22
R23
R24
R25
R26
R27
R28
R29
R30
R31
R32
R33
R34
Chapter 6
116
We advocate not to discard a mutation from inclusion in a screening program based on a pedigrees of less than 50 persons.5 In addition, three recent publications show that certainly not all carriers of a confirmed pathogenic LDLR mutation show an FH phenotype.12-14
Previously, most other reports on non-pathogenic variants in LDLR focussed on single mutations. Based on in silico, in vitro or co-segregation analyses or a combination of these, authors then concluded on the functionality of that specific variant.15-17 The sample size of the pedigrees that harboured such variants was often modest. In contrast, in our pooled analyses, we were able to assess a large dataset of LDLR variants that all were initially thought to be pathogenic and were therefore used for genetic screening in relatives. In fact, at the current time more than 500 different variants in LDLR have been identified in the Netherlands.18 Of the 29,365 individuals tested for the 64 most prevalent LDLR sequence variants, 355 (1.2%) were identified to carry a variant that was initially thought to cause FH, but was later shown not to be pathogenic. Thus, re-evaluation of functionality of sequence variants seems a reasonable approach to select only those FH mutations for genetic cascade screening that are associated with increased LDL-C levels, and consequently increased CAD risk.
Not surprisingly, we found that sequence variants in LDLR that did not lead to hypercholesterolemia did not affect CAD risk. In fact, we recently measured carotid intima-media thickness (IMT) in subjects with pathogenic LDLR mutations who had untreated LDL-C levels below the 75th percentile. Their IMT values were similar to those in unaffected relatives and less than those in genetic FH patients with hypercholesterolemia.12
The strength of the present study is that we could directly compare CAD risk between carriers of an LDLR variant and unaffected relatives. Several factors other than non-pathogenicity could potentially explain the lack of phenotypic expression of a LDLR mutation. These include healthy life style or co-incidence of neutralizing mutations.19-21 However, non pathogenicity of the sequence variants is by far the most likely explanation, because risk of CAD was not increased in individuals that carried these variants.
A potential limitation of our event-free survival analysis is that statin use at the time of genetic diagnosis differs between the groups. Particularly, statin use was more prevalent in the group of FH patients with a pathogenic mutation, compared to unaffected relatives and carriers of a non-pathogenic variant. As a consequence, the natural course of FH has likely been altered in this group and the estimate of CAD risk will be an underestimation of the true risk. However, medication use was similar
R1
R2
R3
R4
R5
R6
R7
R8
R9
R10
R11
R12
R13
R14
R15
R16
R17
R18
R19
R20
R21
R22
R23
R24
R25
R26
R27
R28
R29
R30
R31
R32
R33
R34
Cardiovascular risk and functionality criteria for the LDLR
117
6
between individuals carrying a non-pathogenic variant and their unaffected relatives. As a consequence, the assessment of CAD risk in the group of interest, i.e. carriers of a non-pathogenic LDLR sequence variant, was not biased by medication use.
In conclusion, 10 out of 64 sequence variants in the LDLR gene are proven here to be non-pathogenic both in terms of lipids as well as with regard to event free CAD survival. We therefore have demonstrated that our clinical outcome findings have substantiated the criteria for functionality of LDLR sequence variants. They also confirm the CAD risk associated with FH and underline that these criteria can be used to decide whether a specific sequence variant should be used in cascade screening.
R1
R2
R3
R4
R5
R6
R7
R8
R9
R10
R11
R12
R13
R14
R15
R16
R17
R18
R19
R20
R21
R22
R23
R24
R25
R26
R27
R28
R29
R30
R31
R32
R33
R34
Chapter 6
118
Sup
ple
men
tal t
able
1: C
hara
cter
isti
cs o
f the
LD
LR s
eque
nce
vari
ants
Loc
atio
nM
uta
tion
Seq
uen
ce v
aria
nt
Pat
hog
enic
SIF
TP
olyP
hen
Cod
ing
Syn
onym
ous
P-e
xon
1c.
1-?_
67-?
del
-Y
ES
--
YE
SN
Oex
on 1
c.1A
>G
M1V
*Y
ES
not
tol.
prob
ably
YE
SN
Oex
on 2
c.10
8C>
AD
36E
NO
not
tol.
beni
gnY
ES
NO
exon
2c.
131G
>A
W44
XY
ES
--
YE
SN
Oex
on 2
c.14
8G>
TA
50S
NO
tole
rate
dbe
nign
YE
SN
Oin
tron
2c.
191-
2A>
G-
YE
S-
-N
ON
Oex
on 3
c.24
1C>
AR
81C
YE
Sno
t to
l.be
nign
YE
SN
Oex
on 3
c.26
8G>
AD
90N
*Y
ES
not
tol.
prob
ably
YE
SN
Oin
tron
3c.
313+
1G>
A-
YE
S-
-N
ON
Oin
tron
3c.
313+
1G>
C-
YE
S-
-N
ON
Oin
tron
3c.
313+
2T>
C-
YE
S-
-N
ON
Oin
tron
3c.
314-
1G>
A-
YE
S-
-N
ON
Oex
on 4
c.42
9C>
AC
143X
YE
S-
-Y
ES
NO
exon
4c.
518d
elG
-Y
ES
--
YE
SN
Oex
on 4
c.55
0T>
CC
184R
YE
Sno
t to
l.pr
obab
lyY
ES
NO
exon
4c.
621C
>T
G20
7GY
ES
--
YE
SY
ES
exon
4c.
622G
>A
E20
8K*
YE
Sto
lera
ted
beni
gnY
ES
NO
exon
4c.
662A
>C
D22
1G*
YE
Sno
t to
l.pr
obab
lyY
ES
NO
exon
4c.
682G
>A
E22
8KY
ES
not
tol.
prob
ably
YE
SN
Oex
on 5
c.74
2T>
GC
248G
YE
Sno
t to
l.pr
obab
lyY
ES
NO
exon
5c.
757C
>T
R25
3WN
Ono
t to
l.be
nign
YE
SN
Oex
on 5
c.76
3T>
CC
255R
YE
Sno
t to
l.pr
obab
lyY
ES
NO
exon
6c.
858C
>A
S286
R*
YE
Sno
t to
l.pr
obab
lyY
ES
NO
exon
6c.
877d
elG
-Y
ES
--
YE
SN
Oex
on 6
c.91
7C>
TS3
06L
YE
Sno
t to
l.pr
obab
lyY
ES
NO
exon
6c.
[932
A>
G;9
38G
>A
]K
311R
/C31
3WY
ES
not
tol.;
beni
gn;
YE
SN
Oex
on 7
-8c.
940+
?_11
87-?
del
-Y
ES
--
YE
SN
Oex
on 7
c.10
04G
>T
G33
5VY
ES
not
tol.
prob
ably
YE
SN
Oex
on 7
c.10
27G
>A
G34
3SY
ES
not
tol.
prob
ably
YE
SN
Oex
on 7
c.10
48C
>T
R35
0XY
ES
--
YE
SN
Oex
on 7
c.10
57G
>A
E35
3KY
ES
not
tol.
beni
gnY
ES
NO
R1
R2
R3
R4
R5
R6
R7
R8
R9
R10
R11
R12
R13
R14
R15
R16
R17
R18
R19
R20
R21
R22
R23
R24
R25
R26
R27
R28
R29
R30
R31
R32
R33
R34
Cardiovascular risk and functionality criteria for the LDLR
119
6
exon
8c.
1069
G>
AE
357K
YE
Sno
t to
l.pr
obab
lyY
ES
NO
exon
9-1
2c.
1186
+?_
1846
-?du
pY
ES
--
YE
SN
Oex
on 9
c.12
43G
>C
D41
5H*
YE
Sno
t to
l.be
nign
YE
SN
Oex
on 9
c.12
47G
>A
R41
6QY
ES
not
tol.
beni
gnY
ES
NO
exon
9c.
1265
T>
CL4
22P
YE
Sno
t to
l.be
nign
YE
SN
Oex
on 9
c.12
85G
>A
V42
9MY
ES
not
tol.
poss
ibly
YE
SN
Oex
on 9
c.12
91G
>A
A43
1TY
ES
not
tol.
beni
gnY
ES
NO
exon
9c.
1297
G>
TD
433Y
*Y
ES
not
tol.
prob
ably
YE
SN
Oex
on 9
c.13
29G
>C
or
G>
TW
443C
*Y
ES
not
tol.
prob
ably
YE
SN
Oin
tron
9c.
1358
+1G
>A
-Y
ES
--
NO
NO
intr
on 9
c.13
59-1
G>
A-
YE
S-
-N
ON
Oex
on 1
0c.
1549
T>
CS5
17P
YE
Sno
t to
l.po
ssib
lyY
ES
NO
exon
11+
17c.
[169
0A>
C;2
393_
2401
del]
N56
4H/2
393d
el9b
pY
ES
not
tol./
-pr
obab
ly/-
YE
SN
Oex
ons
12-1
8c.
1705
+?_
2580
+?d
elD
elet
ion
exon
s 12
-18
*Y
ES
--
YE
SN
Oex
on 1
2c.
1729
T>
CS5
87R
NO
not
tol.
prob
ably
YE
SN
Oex
on 1
2c.
1775
G>
AG
592E
YE
Sno
t to
l.pr
obab
lyY
ES
NO
exon
12
c.18
20A
>G
H60
7RY
ES
not
tol.
prob
ably
YE
SN
Oex
on 1
2c.
1835
G>
CL6
11F
YE
Sno
t to
l.po
ssib
lyY
ES
NO
exon
12
c.18
43G
>A
E61
5KY
ES
not
tol.
prob
ably
YE
SN
Oex
on 1
3c.
1898
G>
AR
633H
YE
Sno
t to
l.pr
obab
lyY
ES
NO
exon
13
c.18
98G
>T
R63
3LY
ES
not
tol.
prob
ably
YE
SN
Oex
on 1
4c.
2054
C>
TP
685L
YE
Sno
t to
l.pr
obab
lyY
ES
NO
exon
14
c.20
96C
>T
P69
9LY
ES
not
tol.
prob
ably
YE
SN
Oex
on 1
4c.
2101
G>
AG
701S
*N
Ono
t to
l.po
ssib
lyY
ES
NO
exon
14
c.21
13G
>C
A70
5PY
ES
not
tol.
beni
gnY
ES
NO
intr
on 1
4c.
2140
+5G
>A
-N
O-
-Y
ES
NO
exon
15
c.21
77C
>T
T72
6IN
Ono
t to
l.pr
obab
lyY
ES
NO
exon
15
c.22
31G
>A
R74
4QN
Oto
lera
ted
beni
gnY
ES
NO
exon
15
c.22
82C
>T
T76
1MN
Ono
t to
l.pr
obab
lyY
ES
NO
exon
s 16
-18
c.23
11+
?_25
80+
?del
Del
etio
n ex
ons
16-1
8*
YE
S-
-Y
ES
NO
exon
s 16
-18
c.23
11+
?_23
90+
?del
Del
etio
n ex
ons
16-1
8Y
ES
--
YE
SN
Oin
tron
16
c.23
90-2
A>
G-
YE
S-
-N
ON
Oex
on 1
7c.
2479
G>
AV
827I
NO
not
tol.
prob
ably
YE
SN
O
*Ind
icat
es a
var
iant
new
ly t
este
d by
the
cri
teri
a, s
o no
t de
scri
bed
in t
he H
uman
Mut
atio
n (2
010,
31
(6):7
52-6
0) p
ubli
cati
on (r
efer
ence
5).
R1
R2
R3
R4
R5
R6
R7
R8
R9
R10
R11
R12
R13
R14
R15
R16
R17
R18
R19
R20
R21
R22
R23
R24
R25
R26
R27
R28
R29
R30
R31
R32
R33
R34
Chapter 6
120
REFEREnCES
1. Goldstein JL, Hobbs HH, Brown MS. In: The metabolic and molecular bases of inherited disease. 8th ed. New York: McGraw-Hill; 2001: 2863-2913.
2. Huijgen R, Vissers MN, Defesche JC, Lansberg PJ, Kastelein JJ, Hutten BA. Familial hypercholesterolemia: current treatment and advances in management. Expert Rev Cardiovasc Ther 2008;6(4):567-581.
3. Leigh SE, Foster AH, Whittall RA, Hubbart CS, Humphries SE. Update and Analysis of the University College London Low Density Lipoprotein Receptor Familial Hypercholesterolemia Database. Ann Hum Genet 2008;72:485-498.
4. Umans-Eckenhausen MA, Defesche JC, Sijbrands EJ, Scheerder RL, Kastelein JJ. Review of first 5 years of screening for familial hypercholesterolaemia in the Netherlands. Lancet 2001;357(9251):165-168.
5. Huijgen R, Kindt I, Fouchier SW et al. Functionality of sequence variants in the genes coding for the low-density lipoprotein receptor and apolipoprotein B in individuals with inherited hypercholesterolemia. Hum Mutat 2010;31(6):752-760.
6. den Dunnen JT, Antonarakis SE. Mutation nomenclature extensions and suggestions to describe complex mutations: a discussion. Hum Mutat 2000;15(1):7-12.
7. Bertolini S, Cassanelli S, Garuti R et al. Analysis of LDL receptor gene mutations in Italian patients with homozygous familial hypercholesterolemia. Arterioscler Thromb Vasc Biol 1999;19(2):408-418.
8. Fouchier SW, Defesche JC, Umans-Eckenhausen MW, Kastelein JP. The molecular basis of familial hypercholesterolemia in The Netherlands. Hum Genet 2001;109(6):602-615.
9. Fouchier SW, Kastelein JJ, Defesche JC. Update of the molecular basis of familial hypercholesterolemia in The Netherlands. Hum Mutat 2005;26(6):550-556.
10. Lombardi P, Defesche JC, Kamerling SW, Kastelein JJ, Havekes LM. A novel mutation M-21V in exon 1 of the low density lipoprotein receptor gene causing familial hypercholesterolemia. Clin Genet 1997;51(6):430-431.
11. Bertolini S, Garuti R, Lelli W et al. Four novel partial deletions of LDL-receptor gene in Italian patients with familial hypercholesterolemia. Arterioscler Thromb Vasc Biol 1995;15(1):81-88.
12. Huijgen R, Vissers MN, Kindt I et al. Assessment of carotid atherosclerosis in normocholesterolemic individuals with proven mutations in the low-density lipoprotein receptor or apolipoprotein B genes. Circ Cardiovasc Genet 2011;4(4):413-417.
13. Garcia-Garcia AB, Ivorra C, Martinez-Hervas S et al. Reduced penetrance of autosomal dominant hypercholesterolemia in a high percentage of families: Importance of genetic testing in the entire family. Atherosclerosis 2011;218 (2):423-430.
14. Huijgen R, Hutten BA, Kindt I, Vissers MN, Kastelein JJ. Discriminative Ability of LDL-Cholesterol to Identify Patients with Familial Hypercholesterolemia: A Cross-Sectional Study in 26,406 Individuals Tested for Genetic FH. Circ Cardiovasc Genet 2012;5(3):354-359.
15. Lombardi P, Sijbrands EJ, Kamerling S, Leuven JA, Havekes LM. The T705I mutation of the low density lipoprotein receptor gene (FH Paris-9) does not cause familial hypercholesterolemia. Hum Genet 1997;99(1):106-107.
R1
R2
R3
R4
R5
R6
R7
R8
R9
R10
R11
R12
R13
R14
R15
R16
R17
R18
R19
R20
R21
R22
R23
R24
R25
R26
R27
R28
R29
R30
R31
R32
R33
R34
Cardiovascular risk and functionality criteria for the LDLR
121
6
16. Ekstrom U, Abrahamson M, Sveger T, Sun XM, Soutar AK, Nilsson-Ehle P. Expression of an LDL receptor allele with two different mutations (E256K and I402T). Mol Pathol 2000;53(1):31-36.
17. Gudnason V, Patel D, Sun XM, Humphries S, Soutar AK, Knight BL. Effect of the StuI polymorphism in the LDL receptor gene (Ala 370 to Thr) on lipid levels in healthy individuals. Clin Genet 1995;47(2):68-74.
18. Defesche JC, Kastelein JJ. www.jojogenetics.nl. Website DNA diagnostic for Familial Hypercholesterolemia, last accessed May 2012.
19. Pimstone SN, Sun XM, du Souich C, Frohlich JJ, Hayden MR, Soutar AK. Phenotypic variation in heterozygous familial hypercholesterolemia: a comparison of Chinese patients with the same or similar mutations in the LDL receptor gene in China or Canada. Arterioscler Thromb Vasc Biol 1998;18(2):309-315.
20. van der Graaf A, Fouchier SW, Vissers MN et al. Familial defective apolipoprotein B and familial hypobetalipoproteinemia in one family: two neutralizing mutations. Ann Intern Med 2008;148(9):712-714.
21. Huijgen R, Sjouke B, Vis K et al. Genetic variation in APOB, PCSK9, and ANGPTL3 in carriers of pathogenic autosomal dominant hypercholesterolemic mutations with unexpected low LDL-C Levels. Hum Mutat 2012;33(2):448-455.