12
The new england journal of medicine n engl j med 368;11 nejm.org march 14, 2013 1033 Mechanisms of Disease Review article The Pathogenesis of the Antiphospholipid Syndrome Bill Giannakopoulos, M.B., B.S., Ph.D., and Steven A. Krilis, M.B., B.S., Ph.D. From the Departments of Infectious Dis- eases, Immunology, and Sexual Health (B.G., S.A.K.) and Rheumatology (B.G.), St. George Hospital, Kogarah, and the Department of Medicine, St. George Clini- cal School, University of New South Wales (B.G., S.A.K.) — both in Sydney. Address reprint requests to Dr. Krilis at the De- partment of Infectious Diseases, Immu- nology, and Sexual Health, St. George Hospital, University of New South Wales, 2 South St., Kogarah, Sydney NSW 2217, Australia, or at [email protected], or to Dr. Giannakopoulos at the same mail- ing address, or at bill.giannakopoulos@ unsw.edu.au. N Engl J Med 2013;368:1033-44. DOI: 10.1056/NEJMra1112830 Copyright © 2013 Massachusetts Medical Society. T he antiphospholipid syndrome is a prothrombotic disorder that can affect both the venous and arterial circulations. 1,2 The deep veins of the lower limbs and the cerebral arterial circulation are the most common sites of venous and arterial thrombosis, respectively. 2 However, any tissue or organ vascular bed can be affected. Catastrophic antiphospholipid syndrome, which is characterized by clots in multiple small vascular beds and leads to multiorgan fail- ure with high mortality, develops in a small subgroup of patients. 2,3 In situations in which histopathological confirmation is sought, thrombosis should be present without evidence of inflammation in the vessel wall. 1 The other major clinical manifestations of the antiphospholipid syndrome are obstetrical. They include the unexplained death of one or more morphologically normal fetuses at or beyond the 10th week of gestation, the premature birth of one or more morphologically normal neonates before the 34th week of gestation because of either eclampsia or severe preeclampsia, and three or more unex- plained, consecutive spontaneous abortions before the 10th week of gestation. 1 The revised classification criteria for the antiphospholipid syndrome (2006) emphasize the presence of specific autoantibodies as an essential component of the diagnosis. 1 The persistence (for >12 weeks) of high titers of autoantibodies of the IgG or IgM isotype, detected by enzyme-linked immunosorbent assay (ELISA) for anti–β 2 -glycoprotein I or anticardiolipin antibodies or by lupus-anticoagulant assays, is required. 1 The lupus-anticoagulant assays detect autoantibodies that have the ability to prolong clotting time in vitro. Such antibodies target β 2 -glyco- protein I and prothrombin, both of which are plasma proteins that bind to an- ionic phospholipids. 4-6 The term “antiphospholipid antibodies” is often used to encompass any or all of the antibodies detected by ELISA and the lupus-anticoag- ulant assays. A diagnosis of the antiphospholipid syndrome is made if at least one of the above clinical criteria and one of the laboratory criteria are met. 1 Role of Autoantibodies with Lupus-Anticoagulant Activity A positive test for lupus anticoagulant is a stronger risk factor for thrombosis and adverse pregnancy outcomes after 12 weeks of gestation than positivity for either anti–β 2 -glycoprotein I or anticardiolipin antibodies. 7-9 A case–control study designed to estimate the contribution of genetic and acquired risk factors to a first episode of venous thrombosis in the general population of persons younger than 70 years of age (with no known cancer) showed that 3.1% of persons with venous thrombo- sis were positive for lupus anticoagulant, as compared with 0.9% of controls (odds The New England Journal of Medicine Downloaded from nejm.org at ANNE ARUNDEL MEDICAL CENTER on April 4, 2013. For personal use only. No other uses without permission. Copyright © 2013 Massachusetts Medical Society. All rights reserved.

The Pathogenesis of the Antiphospholipid Syndrome · The Pathogenesis of the Antiphospholipid Syndrome Bill ... have the ability to prolong clotting time in ... mation when infused

  • Upload
    dangthu

  • View
    215

  • Download
    2

Embed Size (px)

Citation preview

Page 1: The Pathogenesis of the Antiphospholipid Syndrome · The Pathogenesis of the Antiphospholipid Syndrome Bill ... have the ability to prolong clotting time in ... mation when infused

T h e n e w e ngl a nd j o u r na l o f m e dic i n e

n engl j med 368;11 nejm.org march 14, 2013 1033

Mechanisms of Disease

Review article

The Pathogenesis of the Anti phos pho lip id Syndrome

Bill Giannakopoulos, M.B., B.S., Ph.D., and Steven A. Krilis, M.B., B.S., Ph.D.

From the Departments of Infectious Dis-eases, Immunology, and Sexual Health (B.G., S.A.K.) and Rheumatology (B.G.), St. George Hospital, Kogarah, and the Department of Medicine, St. George Clini-cal School, University of New South Wales (B.G., S.A.K.) — both in Sydney. Address reprint requests to Dr. Krilis at the De-partment of Infectious Diseases, Immu-nology, and Sexual Health, St. George Hospital, University of New South Wales, 2 South St., Kogarah, Sydney NSW 2217, Australia, or at [email protected], or to Dr. Giannakopoulos at the same mail-ing address, or at [email protected].

N Engl J Med 2013;368:1033-44.DOI: 10.1056/NEJMra1112830Copyright © 2013 Massachusetts Medical Society.

The anti phos pho lip id syndrome is a prothrombotic disorder that can affect both the venous and arterial circulations.1,2 The deep veins of the lower limbs and the cerebral arterial circulation are the most common

sites of venous and arterial thrombosis, respectively.2 However, any tissue or organ vascular bed can be affected. Catastrophic anti phos pho lip id syndrome, which is characterized by clots in multiple small vascular beds and leads to multiorgan fail-ure with high mortality, develops in a small subgroup of patients.2,3 In situations in which histopathological confirmation is sought, thrombosis should be present without evidence of inflammation in the vessel wall.1

The other major clinical manifestations of the anti phos pho lip id syndrome are obstetrical. They include the unexplained death of one or more morphologically normal fetuses at or beyond the 10th week of gestation, the premature birth of one or more morphologically normal neonates before the 34th week of gestation because of either eclampsia or severe preeclampsia, and three or more unex-plained, consecutive spontaneous abortions before the 10th week of gestation.1

The revised classification criteria for the anti phos pho lip id syndrome (2006) emphasize the presence of specific autoantibodies as an essential component of the diagnosis.1 The persistence (for >12 weeks) of high titers of autoantibodies of the IgG or IgM isotype, detected by enzyme-linked immunosorbent assay (ELISA) for anti–β2-gly co pro tein I or anticardiolipin antibodies or by lupus-anticoagulant assays, is required.1 The lupus-anticoagulant assays detect autoantibodies that have the ability to prolong clotting time in vitro. Such antibodies target β2-gly co-pro tein I and prothrombin, both of which are plasma proteins that bind to an-ionic phospholipids.4-6 The term “anti phos pho lip id antibodies” is often used to encompass any or all of the antibodies detected by ELISA and the lupus-anticoag-ulant assays. A diagnosis of the anti phos pho lip id syndrome is made if at least one of the above clinical criteria and one of the laboratory criteria are met.1

Role of Au t oa n tibodies w i th Lupus-A n ticoagul a n t Ac ti v i t y

A positive test for lupus anticoagulant is a stronger risk factor for thrombosis and adverse pregnancy outcomes after 12 weeks of gestation than positivity for either anti–β2-gly co pro tein I or anticardiolipin antibodies.7-9 A case–control study designed to estimate the contribution of genetic and acquired risk factors to a first episode of venous thrombosis in the general population of persons younger than 70 years of age (with no known cancer) showed that 3.1% of persons with venous thrombo-sis were positive for lupus anticoagulant, as compared with 0.9% of controls (odds

The New England Journal of Medicine Downloaded from nejm.org at ANNE ARUNDEL MEDICAL CENTER on April 4, 2013. For personal use only. No other uses without permission.

Copyright © 2013 Massachusetts Medical Society. All rights reserved.

Page 2: The Pathogenesis of the Antiphospholipid Syndrome · The Pathogenesis of the Antiphospholipid Syndrome Bill ... have the ability to prolong clotting time in ... mation when infused

T h e n e w e ngl a nd j o u r na l o f m e dic i n e

n engl j med 368;11 nejm.org march 14, 20131034

Tabl

e 1.

Pro

pose

d Pa

thog

enet

ic M

echa

nism

s in

the

Ant

ipho

spho

lipid

Syn

drom

e (A

PS),

Supp

ortin

g Ev

iden

ce fr

om S

tudi

es in

Hum

ans

and

Ani

mal

s, a

nd Im

plic

atio

ns fo

r Tar

gete

d Th

erap

y.*

Prop

osed

Mec

hani

smSt

udie

s of

APS

Bio

mar

kers

in H

uman

sIn

Vitr

o M

echa

nist

ic S

tudi

es

Usi

ng H

uman

Tis

sue

Ani

mal

Mod

els

of A

PSA

nim

al M

odel

s of

Non

-APS

Th

rom

bosi

sTa

rget

ed T

hera

py

Incr

ease

d ox

idat

ive

stre

ssLe

vels

of o

xidi

zed

β2G

PI w

ere

incr

ease

d in

pat

ient

s w

ith

APS

25; p

arao

xona

se a

ctiv

ity

was

dec

reas

ed26

,27 ; l

ipid

pe

roxi

datio

n by

prod

ucts

w

ere

incr

ease

d28; m

ono-

cyte

s fr

om p

atie

nts

with

A

PS s

how

ed a

n in

crea

se

in in

trac

ellu

lar

RO

S29

Free

thio

l for

m o

f β2G

PI p

ro-

tect

s en

doth

eliu

m fr

om

RO

S30; a

ntip

hosp

holip

id

antib

odie

s pr

omot

e an

in

crea

se in

intr

acel

lula

r R

OS29

RO

S co

ntri

bute

to th

e pa

tho-

gene

sis

of m

urin

e th

rom

-bo

sis31

NA

C in

hibi

ts R

OS-

med

iate

d th

rom

bosi

s31; c

oenz

yme

Q10

inhi

bits

ant

ipho

spho

-lip

id a

ntib

ody–

med

iate

d RO

S ge

nera

tion29

Impa

ired

func

tion

of e

NO

SPa

tient

s w

ith A

PS h

ad im

paire

d en

doth

elia

l nitr

ic o

xide

–de

pend

ent v

ascu

lar

rela

x-at

ion27

and

dec

reas

ed

plas

ma

nitr

ite le

vels

32

HD

L ch

oles

tero

l fro

m w

omen

w

ith a

ntip

hosp

holip

id a

nti-

bodi

es in

hibi

ts e

ndot

helia

l ni

tric

oxi

de p

rodu

ctio

n27

Mic

e de

ficie

nt in

eN

OS

do

not h

ave

antip

hosp

holip

id

antib

ody–

med

iate

d po

ten-

tiatio

n of

thro

mbo

sis33

Stat

ins

up-r

egul

ate

eNO

S ac

-tiv

ity34

(w

hich

may

acc

ount

fo

r th

eir

prot

ectiv

e ef

fect

in

vitr

o35 a

nd in

an

in v

ivo

mur

ine

mod

el o

f APS

36)

Act

ivat

ion

of r

ecep

tors

by

anti-

β2G

PI a

ntib

odie

sA

nti-β

2GPI

ant

ibod

ies

(spe

cific

fo

r dom

ain

I) w

ith lu

pus-

antic

oagu

lant

act

ivity

st

rong

ly c

orre

late

with

th

rom

bosi

s as

soci

ated

with

A

PS13

,19

The

rele

vant

targ

et r

ecep

tors

on

pla

tele

ts a

re A

poE

rece

p-to

r 237

and

gly

copr

otei

n Ib

a37,3

8 ; on

mon

ocyt

es,

anne

xin

A2,

39 T

LR2,

40 a

nd

TLR

439,4

1 ; and

on

endo

the-

lial c

ells

, ann

exin

A2,

42,4

3 TL

R2,

40 a

nd T

LR443

,44

Apo

E re

cept

or 2

–kno

ckou

t m

ice,

33 a

nnex

in A

2–kn

ocko

ut m

ice,

45 an

d LP

S-in

sens

itive

mic

e46

are

prot

ecte

d fr

om a

n-tip

hosp

holip

id a

ntib

ody–

med

iate

d th

rom

bosi

s

A1

anal

ogue

s of

Apo

E re

cept

or

2 an

d sy

nthe

tic d

omai

n I

inhi

bit a

nti-β

2GPI

–med

i-at

ed e

ffect

s in

vitr

o47 a

nd

in v

ivo48

,49 ; N

AC

inhi

bits

th

rom

bosi

s as

soci

ated

w

ith T

TP in

a m

urin

e m

odel

50

Incr

ease

d ex

pres

sion

and

ac

tivat

ion

of ti

ssue

fa

ctor

Incr

ease

d ex

pres

sion

of t

issu

e fa

ctor

sho

wn

in p

atie

nts

with

APS

Up-

regu

latio

n of

tiss

ue fa

ctor

by

ant

ipho

spho

lipid

ant

i-bo

dies

has

bee

n sh

own

in

mon

ocyt

es39

,51 a

nd n

eutr

o-ph

ils52

and

on e

ndot

helia

l ce

lls53

Tiss

ue fa

ctor

pla

ys a

role

in

APS

-ass

ocia

ted

thro

mbo

tic

mic

roan

giop

athy

54

PDI i

nhib

itors

att

enua

te m

u-ri

ne th

rom

bosi

s55; s

tatin

s in

hibi

ted

thro

mbo

sis

in a

m

urin

e m

odel

of t

issu

e fa

ctor

–dep

ende

nt A

PS54

Incr

ease

in fr

ee th

iol f

orm

of

fact

or X

IPa

tient

s w

ith A

PS h

ave

elev

ated

le

vels

of t

he fr

ee th

iol f

orm

of

fact

or X

I56

PDI-

trea

ted

or th

iore

doxi

n-tr

eate

d fa

ctor

XI i

s m

ore

ra

pidl

y co

nver

ted

to

fact

or X

Ia56

Fact

or X

I pla

ys a

cri

tical

rol

e

in p

atho

logi

c th

rom

bus

form

atio

n57

PDI i

nhib

itors

and

fact

or X

I in

hibi

tors

att

enua

te in

vi

vo th

rom

bus

form

atio

n in

mic

e55 a

nd p

rim

ates

57

Dis

rupt

ion

of th

e an

nexi

n

A5

shie

ldA

nnex

in A

5 re

sist

ance

ant

ico-

agul

ant a

ctiv

ity c

orre

late

s w

ith c

linic

al A

PS

Dec

reas

e in

ann

exin

A5

show

n in

ant

ipho

spho

lipid

ant

i-bo

dy–t

reat

ed e

ndot

helia

l ce

lls58

Hyd

roxy

chlo

roqu

ine

inhi

bits

an

ti-β

2GPI

dis

rupt

ion

of

the

anne

xin

A5

shie

ld in

vi-

tro59

and

att

enua

tes

thro

m-

bosi

s as

soci

ated

with

APS

in

mic

e60

Ant

ibod

y-m

edia

ted

activ

a-tio

n of

com

plem

ent C

3 an

d C

5

Exce

ssiv

e co

mpl

emen

t act

iva-

tion

show

n in

pla

cent

as o

f pa

tient

s w

ho a

re p

ositi

ve

for

antip

hosp

holip

id a

nti-

bodi

es, a

s co

mpa

red

with

he

alth

y co

ntro

ls

C5a

bin

ds a

nd a

ctiv

ates

neu

-tr

ophi

ls, i

nduc

ing

up-

regu

latio

n of

tiss

ue

fact

or52

Com

plem

ent h

as a

med

iatin

g ro

le in

mod

els

of th

rom

bo-

sis23

,61 a

nd fe

tal l

oss62

,63

in m

urin

e A

PS

C5

inhi

bito

r ec

uliz

umab

am

e-lio

rate

s ca

tast

roph

ic

APS

64,6

5

The New England Journal of Medicine Downloaded from nejm.org at ANNE ARUNDEL MEDICAL CENTER on April 4, 2013. For personal use only. No other uses without permission.

Copyright © 2013 Massachusetts Medical Society. All rights reserved.

Page 3: The Pathogenesis of the Antiphospholipid Syndrome · The Pathogenesis of the Antiphospholipid Syndrome Bill ... have the ability to prolong clotting time in ... mation when infused

mechanisms of disease

n engl j med 368;11 nejm.org march 14, 2013 1035

ratio, 3.6).10 In a case–control study focusing on risk factors for stroke in women in the general population younger than 50 years of age, 17% of the patients with stroke were positive for lupus anticoagulant, as compared with 0.7% of con-trols (odds ratio, 43.1).8 The risk was further in-creased by taking oral contraceptive pills (odds ratio, 201.0) or smoking (odds ratio, 87.0). Ap-proximately 1% of women trying to get pregnant have recurrent miscarriages; of these women, ap-proximately 10 to 15% are estimated to have ob-stetrical anti phos pho lip id syndrome.11 Positivity for lupus anticoagulant is the strongest predictor of subsequent thrombosis in purely obstetrical anti phos pho lip id syndrome; the annual incidence of deep-vein thrombosis is 1.46%, and the annual incidence of stroke is 0.32%.12

Lupus anticoagulant due to anti–β2-gly co pro-tein I autoantibodies correlates more strongly with a risk of thrombosis than does lupus anticoagu-lant due to antiprothrombin autoantibodies.13,14 The risk of a first thrombotic event among asymp tomatic persons who are positive for lupus anticoagulant, anticardiolipin antibodies, and anti–β2-gly co pro tein I antibodies — so-called triple-positive patients — is 5.3% per year.15 These patients have high titers of autoantibodies that bind the major B-cell epitope on domain I of the β2-gly co pro tein I molecule.16-18 Domain I anti–β2-gly co pro tein I autoantibodies confer lupus-anticoagulant activity associated with the highest risk of thrombosis.19 Assays that detect autoan-tibodies to the phosphatidylserine–prothrombin complex (in contrast to prothrombin alone) may help establish the diagnosis of the anti phos pho-lip id syndrome and the associated level of risk, in conjunction with the lupus-anticoagulant as-says and the ELISA for anti–β2-gly co pro tein I antibodies.20,21 The usefulness of also performing the ELISA for anticardiolipin antibodies to diag-nose thrombotic anti phos pho lip id syndrome is being debated.22

Autoantibodies from patients with the anti-phos pho lip id syndrome potentiate thrombus for-mation when infused into mice in which the blood vessel has been injured.23,24 The thrombogenic properties are eliminated when the fraction of anti–β2-gly co pro tein I autoantibodies is re-moved.23,24 We review the mechanisms that may contribute to thrombosis in the syndrome, inte-grating clinical biomarker data, results of in vi-tro mechanistic studies, and relevant animal models (Table 1).25-71

Incr

ease

d ex

pres

sion

of

TLR

7 an

d TL

R8

and

sen-

sitiz

atio

n to

TLR

7 an

d TL

R8

agon

ists

Ant

ipho

spho

lipid

ant

ibod

ies

in

duce

up-

regu

latio

n of

TL

R7 a

nd T

LR8

in p

lasm

a-cy

toid

den

driti

c ce

lls, s

en-

sitiz

ing

them

to th

e ef

fect

s of

TLR

7 an

d TL

R8

ago-

nist

s66

A s

pont

aneo

us m

odel

of m

u-ri

ne A

PS in

NZ

W x

BXS

B

F1 m

ice

is m

edia

ted

by

TLR

7 du

plic

atio

n16,6

7,68

Hyd

roxy

chlo

roqu

ine

inhi

bits

TL

R7

activ

atio

n in

vitr

o69;

hydr

oxyc

hlor

oqui

ne u

se in

SL

E is

ass

ocia

ted

with

lo

wer

odd

s of

per

sist

ently

po

sitiv

e an

tipho

spho

lipid

an

tibod

ies70

BA

FFB

AFF

inhi

bito

r be

limum

ab is

pr

otec

tive

agai

nst t

hrom

-bo

sis

in N

ZW

× B

XSB

F1

mic

e71

* A

poE

deno

tes

apol

ipop

rote

in E

, BA

FF B

-cel

l act

ivat

ing

fact

or, β

2GPI

β2-

glyc

opro

tein

I, e

NO

S en

doth

elia

l nitr

ic o

xide

syn

thas

e, H

DL

high

-den

sity

lipo

prot

ein,

LPS

lipo

poly

sacc

hari

de, N

AC

N

-ace

tylc

yste

ine,

PD

I pro

tein

dis

ulfid

e is

omer

ase,

RO

S re

activ

e ox

ygen

spe

cies

, SLE

sys

tem

ic lu

pus

eryt

hem

atos

us, T

LR2

toll-

like

rece

ptor

2, T

LR4

toll-

like

rece

ptor

4, T

LR7

toll-

like

rece

ptor

7,

TLR

8 to

ll-lik

e re

cept

or 8

, and

TTP

thr

ombo

tic t

hrom

bocy

tope

nic

purp

ura.

The New England Journal of Medicine Downloaded from nejm.org at ANNE ARUNDEL MEDICAL CENTER on April 4, 2013. For personal use only. No other uses without permission.

Copyright © 2013 Massachusetts Medical Society. All rights reserved.

Page 4: The Pathogenesis of the Antiphospholipid Syndrome · The Pathogenesis of the Antiphospholipid Syndrome Bill ... have the ability to prolong clotting time in ... mation when infused

T h e n e w e ngl a nd j o u r na l o f m e dic i n e

n engl j med 368;11 nejm.org march 14, 20131036

Thrombo tic Mech a nisms

Post-Translational Redox Modifications of β2-Gly co pro tein I

A number of findings suggest that the anti phos-pho lip id syndrome is characterized by increased oxidative stress. Paraoxonase activity, which ac-counts for the antioxidant properties of high-density lipoprotein cholesterol (preventing oxida-tion of low-density lipoprotein [LDL] cholesterol), is significantly decreased in persons with the syndrome,26,27 whereas levels of 8-epi-prosta-glandin F2α, a biomarker of lipid peroxidation, are elevated.28 Plasma levels of β2-gly co pro tein I–oxidized LDL complexes are elevated in persons with the anti phos pho lip id syndrome as compared with healthy controls.72

Oxidative stress plays a direct role in the structure and function of β2-gly co pro tein I in patients with the anti phos pho lip id syndrome. Purified β2-gly co pro tein I is composed of four domains (domains I through IV) that contain

two disulfide bridges each and a fifth domain (domain V) that contains an extra disulfide bridge linking cysteine (Cys) 288 with Cys32673

(Fig. 1). In healthy persons, the free thiol form of β2-gly co pro tein I predominates in the plasma, characterized by a broken disulfide bridge at Cys32 and Cys60 and another at Cys288 and Cys326.25,30,74,75 The former pair of free thiols are near the anti phos pho lip id syndrome B-cell epitope in domain I,16,17 and the latter are near the T-cell epitope in domain V.76 The disulfide bridges at these locations are broken by the oxi-doreductases thioredoxin-1 and protein disulfide isomerase (PDI).74,75 Under conditions of oxida-tive stress, disulfide bonds form at these sites.25

The relative proportion of plasma β2-gly co pro-tein I in the oxidized versus the free thiol form was significantly greater in patients with the anti phos pho lip id syndrome than in patients with autoimmune disease with or without per-sistent anti phos pho lip id antibodies but without the anti phos pho lip id syndrome, patients with vascular thrombosis without anti phos pho lip id antibodies, and healthy volunteers (P<0.001 for all comparisons).25 Patients with the syndrome who were positive for both lupus anticoagulant and anti–β2-gly co pro tein I antibodies had sig-nificantly higher levels of oxidized β2-gly co pro-tein I than patients who were positive for anti–β2-gly co pro tein I antibodies alone.25

Supporting the notion that oxidation unmasks the critical anti phos pho lip id syndrome B-cell epi-tope, anti–β2-gly co pro tein I antibodies purified from mice and rabbits that had been immunized with β2-gly co pro tein I displayed decreased binding to oxidoreductase-treated β2-gly co pro tein I,74 as did autoantibodies that were affinity-purified from patients with the anti phos pho lip id syn-drome (Fig. 2).

Conformations of β2-Gly co pro tein I

β2-gly co pro tein I can potentially exist in a circu-lar form, with domain I interacting with domain V.77 In this form, the critical B-cell epitope is hid-den from the immune system.77 On binding to an anionic phospholipid surface through domain V, the circular form of β2-gly co pro tein I opens up to a fishhook configuration, exposing the domain I epitope and allowing domain I anti–β2-gly co pro-tein I autoantibodies to bind.77 The presence of the circular form has yet to be directly shown in

Oxidoreductases

Oxidation

K19

DI

DII

DIII

DIVDV

R39

R43

T-cell epitope276–290

Oxidized Form (immunogenic)

Free Thiol Form (nonimmunogenic)

B-cell epitopeR39, R43

Major Species in Patients with the Antiphospholipid Syndrome

Major Species in Healthy Persons

K19

R39

R43

K19

DI

DII

DIII

DIVDV

R39

R43

T-cell epitope276–290

B-cell epitopeR39, R43

K19

R39

R43

SH326

SH32

B-cell epitope32

SH60

S60

S32

SH288

S281

S306

S326

S288

S281

S306

SS

S

S

S

S

S

S

SS

S

S

S

S

S

S

S SS SS SS S

S SS S

S SS S

S SS S

S SS S

S SS S

S SS S

Figure 1. Schematic Representation of the Crystal Structure (Fishhook Configuration) of β2-Gly co pro tein I (β2GPI).

The carboxy-terminal amino acid cysteine (Cys) 326 forms a disulfide (S–S) bridge with Cys288. D denotes domain, K lysine, R arginine, and SH free thiol. The numbers indicate the position of the amino acid starting from the amino-terminal end of the protein.

The New England Journal of Medicine Downloaded from nejm.org at ANNE ARUNDEL MEDICAL CENTER on April 4, 2013. For personal use only. No other uses without permission.

Copyright © 2013 Massachusetts Medical Society. All rights reserved.

Page 5: The Pathogenesis of the Antiphospholipid Syndrome · The Pathogenesis of the Antiphospholipid Syndrome Bill ... have the ability to prolong clotting time in ... mation when infused

mechanisms of disease

n engl j med 368;11 nejm.org march 14, 2013 1037

human plasma; however, circumstantial evidence points to its in vivo presence. Domain I anti–β2-gly co pro tein I autoantibodies were induced in mice in which protein H (derived from Streptococcus pyogenes) was administered.78 Protein H changes the conformation of β2-gly co pro tein I from the circular to the theoretically more immunogenic open form in vitro.78 The relationship of the cir-cular form of β2-gly co pro tein I to the free thiol form has not been determined.

Triggers of Thrombosis

The “two hit” model of thrombosis associated with the anti phos pho lip id syndrome proposes that an initiating “first hit” injury disrupts the endothe-lium, and a “second hit” potentiates thrombus formation.79 Autoantibodies from patients with the anti phos pho lip id syndrome that are infused into mice do not promote thrombus formation in the absence of vessel-wall injury.23,24 A key step in allowing β2-gly co pro tein I immune complexes to form on the cell surface is endothelial-cell priming.23 β2-gly co pro tein I does not bind un-stimulated endothelium in vivo.80 In catastrophic anti phos pho lip id syndrome, infection and recent surgery are recognized precipitants of endothe-lial injury.81 However, the initiating stimulus is not identified in most cases of thrombotic anti-phos pho lip id syndrome. We postulate that dis-turbance of the redox balance in the vascular milieu in patients with the anti phos pho lip id syn-drome may constitute a substantial first hit that primes the endothelium, allowing β2-gly co pro-tein I immune complexes to form on the cell sur-face and assert their pathogenicity.

Patients with the anti phos pho lip id syndrome have significantly lower levels of the endothelium-protective free thiol form of β2-gly co pro tein I that provides a buffer against oxidative stress than do healthy persons.25,30 The odds ratio for reduced levels of the free thiol form of β2-gly co pro tein I in patients with the anti phos pho lip id syndrome, as compared with age-matched and sex-matched controls, was reported to be 4.1 (95% confidence interval [CI], 1.9 to 8.8).25 Oxidative stress from exogenous sources such as smoking82 may tip the vascular endothelial milieu toward a prothrom-botic phenotype. For example, oxidative stress can up-regulate the expression of annexin A2,83 an endothelial cell-surface receptor for β2-gly co pro-tein I that has an important role in the pathogen-

esis of the anti phos pho lip id syndrome.42 Among young women, the odds ratio for ischemic stroke in the presence of lupus anticoagulant is 43.1, and it increases to 87.0 with concurrent oxidative stress (and other pathophysiological disturbances) in-duced by smoking.8 In a murine model of throm-bosis, reactive oxygen species induced platelet aggregation, endothelial-cell stimulation, and ex-pression of von Willebrand factor.31 N-acetylcys-teine (NAC), a scavenger of reactive oxygen species, inhibited thrombus formation in this model.31 The therapeutic value of NAC in the anti phos pho lip id syndrome may be worth exploring (Fig. 3).

Endothelial Nitric Oxide Synthase

Patients with the anti phos pho lip id syndrome have decreased levels of plasma nitrite, as compared with controls.32 They also have impaired endo-thelium-dependent vascular responses,27 which suggests that the activity of endothelial nitric ox-ide synthase is abnormal.

Endothelium-derived nitric oxide plays an im-

Opt

ical

Den

sity

(405

nm

)

1.0

0.8

0.6

0.4

0.2

0.0Oxidized (S–S)Form of β2GPI

Free Thiol (SH)Form of β2GPI

Figure 2. Results of Direct Enzyme-Linked Immuno-sorbent Assays of Affinity-Purified Anti phos pho lip id Antibodies in Patients with the Anti phos pho lip id Syn-drome.

The oxidized (S–S) form or free thiol (SH) form of β2GPI was coated onto microtiter plates, and antibodies from individual patients were assessed for binding. Symbols denote individual antibody preparations. In-creased optical density denotes an increase in anti-β2GPI binding to oxidized β2GPI.

The New England Journal of Medicine Downloaded from nejm.org at ANNE ARUNDEL MEDICAL CENTER on April 4, 2013. For personal use only. No other uses without permission.

Copyright © 2013 Massachusetts Medical Society. All rights reserved.

Page 6: The Pathogenesis of the Antiphospholipid Syndrome · The Pathogenesis of the Antiphospholipid Syndrome Bill ... have the ability to prolong clotting time in ... mation when infused

T h e n e w e ngl a nd j o u r na l o f m e dic i n e

n engl j med 368;11 nejm.org march 14, 20131038

portant role in healthy endothelial function.84-86

It is produced by enzymatic conversion of L-argi-nine by endothelial nitric oxide synthase.87 Re-duced expression and activity of endothelial ni-tric oxide synthase can result in the generation of superoxide and peroxynitrite.88 Because nitric oxide has an exceptionally short half-life, the activity of endothelial nitric oxide synthase is estimated by measuring nitric oxide metabolites in plasma. Plasma nitrite most closely reflects changes in the activity of endothelial nitric oxide synthase in humans.89

In a murine model, domain I anti–β2-gly co-pro tein I autoantibodies decreased bioavailable nitric oxide by antagonizing the activity of endo-thelial nitric oxide synthase, which led to mono-cyte adhesion to the endothelium.33 The autoan-tibodies exerted their pathogenic effects in this model in a manner that was independent of complement and Fc receptor.33 Endothelial nitric

oxide–dependent arterial relaxation was inhibit-ed by domain I anti–β2-gly co pro tein I autoanti-bodies in these mice,33 reflecting vascular dis-turbances analogous to those in humans with anti phos pho lip id autoantibodies.27 Inhibition of the activity of endothelial nitric oxide synthase was mediated by the F(ab′)2 portion of domain I anti–β2-gly co pro tein I autoantibodies, which di-merized β2-gly co pro tein I molecules attached to apolipoprotein E (ApoE) receptor 2 (LDL receptor–related protein 8), cross-linking and activating ApoE receptor 2.33 Anti–β2-glycoprotein I auto-antibodies did not enhance leukocyte adhesion to the endothelium, nor did they potentiate in vivo thrombus formation in mice deficient in endothelial nitric oxide synthase or ApoE recep-tor 2, findings that indicate the critical role of these receptors in pathogenicity mediated by anti–β2-gly co pro tein I autoantibodies.33

Statins — inhibitors of 3-hydroxy-3-meth yl-

PDI

β2GPI β2GPI

C5a

C5

C3

Factor XI Factor XIa

Propagation

Neutrophil

B cell

Decrease in NO

eNOS

TLR7TLR8

1

7

6

2

3

4

11 5

8

9

10

S S

P

Increase in thrombus formation

Proinflammatory cytokine release

β2GPI immune complexes

β2GPI receptors

Plasmacytoid dendritic cells

and monocytes

Endothelial cell surface

S S

Factor XI

cTF deTF

SH SH SH SH

S S

ROS

XIncrease in PP2A

Smoking

Tissuefactor

5

Intracellular ROS

Figure 3. Pathophysiological Mechanisms of Thrombogenesis and Sites of Action of Possible Therapeutic Interventions.

Numbers indicate sites of action of the following possible interventions: (1) N-acetylcysteine (NAC), (2) statins and other molecules that up-regulate activity of endothelial nitric oxide synthase (eNOS), (3) A1–A1 dimer of apolipo-protein E receptor 2 or hydroxychloroquine, (4) synthetic domain I of β2GPI or NAC, (5) coenzyme Q10, (6) inhibitors of protein disulfide isomerase (PDI), (7) inhibitors of factor XIa, (8) inhibitors of complement 5 (C5) (e.g., eculizumab), (9) inhibitors of complement 3 (C3), (10) inhibitors of toll-like receptor 7 (TLR7) (e.g., hydroxychloroquine), and (11) inhibitors of B-cell activating factor (e.g., belimumab). β2GPI receptors are dimerized by β2GPI immune complexes, resulting in cell activation. C5a denotes complement fragment 5a, cTF cryptic tissue factor, deTF de-encrypted tissue factor, NO nitric oxide, P phosphorylation, PP2A protein phosphatase 2A, ROS reactive oxygen species, and TLR8 toll-like receptor 8.

The New England Journal of Medicine Downloaded from nejm.org at ANNE ARUNDEL MEDICAL CENTER on April 4, 2013. For personal use only. No other uses without permission.

Copyright © 2013 Massachusetts Medical Society. All rights reserved.

Page 7: The Pathogenesis of the Antiphospholipid Syndrome · The Pathogenesis of the Antiphospholipid Syndrome Bill ... have the ability to prolong clotting time in ... mation when infused

mechanisms of disease

n engl j med 368;11 nejm.org march 14, 2013 1039

glu taryl–coenzyme A (HMG-CoA) reductase — block the thrombogenic properties of anti phos-pho lip id autoantibodies in vitro35 and in vivo.36 Statins may be protective in the anti phos pho lip id syndrome owing in part to their up-regulation of endothelial nitric oxide synthase34 (Fig. 3).

A number of strategies are being pursued to disrupt the formation of β2-gly co pro tein I im-mune complexes on cell surfaces. Domain V of β2-gly co pro tein I binds the A1 ligand–binding type A module of ApoE receptor 2.90 A dimeric A1–A1 molecule blocks the formation of β2-gly-co pro tein I immune complexes on anionic phos-pholipid surfaces.47 Mice treated with soluble monomeric A1 are protected from the thrombo-genic effects of anti–β2-gly co pro tein I autoanti-bodies,48 providing proof of principle that A1–A1 dimers have therapeutic value (Fig. 3). Infusion of synthetic domain I of β2-gly co pro tein I was pro-tective in a murine model of thrombosis induced by anti–β2-gly co pro tein I autoantibodies49 (Fig. 3).

The binding of pathogenic domain I anti–β2-gly co pro tein I autoantibodies to β2-gly co pro tein I may be inhibited by breaking the disulfide bond and inducing free thiol formation at Cys32 and Cys60 within domain I of β2-gly co pro tein I74,75 (Fig. 2). NAC is able to break an analogous di-sulfide bond within von Willebrand factor, with implications for treating patients with throm-botic thrombocytopenic purpura,50 and explora-tion of its therapeutic potential in the anti phos-pho lip id syndrome may be of value (Fig. 3).

Endothelial Cells and Monocytes

Anti phos pho lip id autoantibodies may up-regu-late the cell-surface expression of proadhesive and procoagulant molecules such as tissue fac-tor.53 Anti–β2-gly co pro tein I autoantibodies may induce signaling by means of a multiprotein com-plex on the endothelial cell surface that includes annexin A2 (bound by β2-gly co pro tein I),42 toll-like receptor 4 (TLR4),44 calreticulin, and nucleo-lin.43 Intracellular activation downstream of TLR4 occurs through myeloid differentiation factor 88,44 culminating in activation of nuclear factor κB (NF-κB).43 The targeting of NF-κB is a thera-peutic option.91 The absence of annexin A2 was reported to protect mice against the prothrom-botic effects of infused autoantibodies from pa-tients with the anti phos pho lip id syndrome.45 Mice with resistance to lipopolysaccharide were also

protected, a finding that supports the relevance of TLR4 in the pathogenesis of the anti phos pho-lip id syndrome.46

β2-gly co pro tein I has been shown to colocal-ize with annexin A2 and TLR4 on the lipid rafts of monocytes.39 Anti–β2-gly co pro tein I autoanti-bodies stimulate monocytes to increase tissue factor expression and release tumor necrosis fac-tor α (TNF-α).39 Anti phos pho lip id autoantibody–induced tissue factor expression is mediated through a number of intracellular signaling pathways.51

In one study, 19 of 32 affinity-purified anti-bodies from patients with the anti phos pho lip id syndrome were shown to induce activation of hu-man monocytes and endothelial cells.40 However, in that series of experiments, the results showed that activation occurred through toll-like recep-tor 2 (TLR2) and CD14, not TLR4.40 Further work is needed to resolve these discrepancies.

Autoantibodies from patients with the anti-phos pho lip id syndrome can disrupt the mito-chondrial function of monocytes and neutrophils, leading to the generation of various intracellular reactive oxygen species and the subsequent ex-pression of tissue factor.29 Antibodies from pa-tients with the syndrome do not colocalize with mitochondria, suggesting that mitochondrial dysfunction is induced through undefined indi-rect pathways.29 In one study, the inhibition of intracellular reactive oxygen species in monocytes with the use of NAC, vitamin C, or mitochon-drial cofactor coenzyme Q10 prevented the up-regulation of tissue factor induced by anti phos-pho lip id autoantibodies29 (Fig. 3).

Human monocytes are activated in a distinct manner by polyclonal autoantibodies derived from patients with purely thrombotic anti phos-pho lip id syndrome as compared with monocyte activation by autoantibodies from patients with obstetrical anti phos pho lip id syndrome.41 Auto-antibodies from patients with thrombotic anti-phos pho lip id syndrome induce tissue factor ex-pression, which is caused by the autoantibody fraction that binds β2-gly co pro tein I.41

Tissue Factor

Tissue factor is the key initiator of the extrinsic coagulation pathway.92 It is located on cell sur-faces and microparticles in an encrypted, inac-tive form.92 On vessel injury leading to exposure

The New England Journal of Medicine Downloaded from nejm.org at ANNE ARUNDEL MEDICAL CENTER on April 4, 2013. For personal use only. No other uses without permission.

Copyright © 2013 Massachusetts Medical Society. All rights reserved.

Page 8: The Pathogenesis of the Antiphospholipid Syndrome · The Pathogenesis of the Antiphospholipid Syndrome Bill ... have the ability to prolong clotting time in ... mation when infused

T h e n e w e ngl a nd j o u r na l o f m e dic i n e

n engl j med 368;11 nejm.org march 14, 20131040

of phosphatidylserine, tissue factor becomes de-encrypted and activated, enabling it to bind fac-tor VIIa, which leads to activation of factor IX and factor X.92 The relevance of tissue factor in the pathogenesis of the anti phos pho lip id syn-drome is supported by the results of in vitro and in vivo murine studies.39,54 Thiol exchange reac-tions play an important role in the regulation of tissue factor from the encrypted to the de- encrypted form.92 PDI, an extracellular regulator of thiol exchange, is associated with cell-surface tissue factor and is required for tissue factor–dependent thrombosis in vivo.92 Hence, PDI in-hibitors55 may therapeutically target tissue factor in the pathogenesis of the anti phos pho lip id syn-drome (Fig. 3).

Factor XI

Elevated levels of coagulation factor XI confer a predisposition to venous thrombosis93 and stroke94 in the general population, mirroring the distribution of thrombosis in patients with the anti phos pho lip id syndrome. A link has been dis-covered between the anti phos pho lip id syndrome and dysregulated activation of factor XI.56 Factor XI is a proenzyme that is cleaved to its active form (factor XIa) by factor XIIa or thrombin.95 Factor XIa is responsible for factor IX activation, ultimately leading to a burst of thrombin genera-tion.95 Factor XI can be a substrate of the oxido-reductases thioredoxin 1 and PDI,56 which target the factor XI intrachain disulfide bonds at Cys118–Cys147 and Cys362–Cys482 and the in-terchain disulfide bond Cys321–Cys321, generat-ing free thiols at these positions.56 Both the free thiol form and the intact disulfide-bridge form of factor XI are found in human plasma.56 In one study, patients with the anti phos pho lip id syn-drome had significantly higher levels of the free thiol form of factor XI than age-matched and sex-matched controls.56 Oxidoreductase-treated fac-tor XI was more rapidly converted to factor XIa by factor XIIa and thrombin than was the untreated form. The interaction between factor XI and PDI in the context of thrombosis associated with the anti phos pho lip id syndrome warrants further ex-ploration. Inhibitors of PDI and factor XIa are ef-fective in treating thrombosis in animal mod-els55,96 (Fig. 3). Inhibition of factor XI provides protection against thrombosis but is not associ-ated with an increased risk of bleeding in these models, making factor XI an attractive therapeu-tic target.57

The Cys362–Cys482 free thiols may not be critical for the potentiation of factor XI activa-tion. Mutagenesis of Cys362 and Cys482 to ala-nine, which eliminates the disulfide bridge be-tween the heavy chain and light chain of factor XI, leads to decreased ligation of factor IX by the factor XIa mutants.97

Platelets

β2-gly co pro tein I can interact with the von Wille-brand factor receptor gly co pro tein Ibα37,38 and ApoE receptor 2.37 This enables anti–β2-gly co-pro tein I autoantibodies to cross-link these re-ceptors, leading to potentiation of platelet activa-tion, the release of thromboxane A2, and an increase in platelet adhesiveness.37,38 Platelet fac-tor 4, a cationic protein released by activated plate-lets, can facilitate the dimerization of β2-gly co-pro tein I, enhancing the formation of pathogenic immune complexes on the platelet surface.98

Annexin A5 Anticoagulant Shield and Hydroxychloroquine

In one model of the pathogenesis of the anti-phos pho lip id syndrome, annexin A5 binds to phosphatidylserine surfaces, forming a shield that inhibits the formation of procoagulant complex-es.58 An in vitro study has shown that domain I anti–β2-gly co pro tein I autoantibodies in complex with β2-gly co pro tein I can disrupt the shield, ex-posing procoagulant phosphatidylserine and hence predisposing to thrombosis.99 Hydroxychloro-quine inhibits the ability of β2-gly co pro tein I im-mune complexes to disrupt the annexin A5 matrix on the endothelial-cell surface59 (Fig. 3). Hydroxy-chloroquine diminished anti phos pho lip id auto-antibody–mediated thrombosis in vivo in a mu-rine model.60

Complement and Neutrophils

Case reports document the use of the C5 inhibi-tor eculizumab to prevent anti phos pho lip id syn-drome–associated thrombotic microangiopathy that complicates renal transplantation,64 and to treat patients with acute catastrophic anti phos-pho lip id syndrome65 (Fig. 3). In vivo murine studies implicating the activation of the classical complement pathway in thrombosis associated with the anti phos pho lip id syndrome23,61 were the basis for the use of eculizumab in these case reports. Activation of complement by anti phos-pho lip id autoantibodies generates C5a, which binds and activates neutrophils, leading to tissue

The New England Journal of Medicine Downloaded from nejm.org at ANNE ARUNDEL MEDICAL CENTER on April 4, 2013. For personal use only. No other uses without permission.

Copyright © 2013 Massachusetts Medical Society. All rights reserved.

Page 9: The Pathogenesis of the Antiphospholipid Syndrome · The Pathogenesis of the Antiphospholipid Syndrome Bill ... have the ability to prolong clotting time in ... mation when infused

mechanisms of disease

n engl j med 368;11 nejm.org march 14, 2013 1041

factor expression.52 On the basis of murine stud-ies, C3 and C5 have been proposed as possible therapeutic targets for treating obstetrical anti-phos pho lip id syndrome62,63 (Fig. 3).

Disturbance of Innate Immunity

The prevalence of lupus-anticoagulant positivity among patients with systemic lupus erythema-tosus (SLE) is 30%,100 and the presence of lupus-anticoagulant positivity in such patients is associated with an increased risk of thrombo-sis (odds ratio, 5.6).101 Forty percent of patients with the anti phos pho lip id syndrome also have SLE,2 and 37% of patients with SLE have anti–β2-gly co pro tein I autoantibodies.102 These find-ings suggest that there is overlap in the patho-genesis of SLE and that of the anti phos pho lip id syndrome.

A relationship between the two disorders is supported by the spontaneous development of domain I anti–β2-gly co pro tein I autoantibodies16 and the development of a syndrome analogous to human anti phos pho lip id syndrome in a murine model of lupus, NZW x BXSB F1 male mice.67 In this model, a major contributor to pathogenesis is TLR7 duplication resulting from translocation of TLR7 from the X to the Y chromosome in the BXSB male mice.68 Dysregulated activation of TLR7 in plasmacytoid dendritic cells by RNA containing immune complexes and the genera-tion of autoantibodies (e.g., anti-Sm and anti-RNP antibodies) create a positive-feedback loop

for further autoantibody generation.103,104 Anti-bodies from patients with the anti phos pho lip id syndrome are able to up-regulate the expression of TLR7 and TLR8 in plasmacytoid dendritic cells and monocytes, respectively, as well as their translocation from the endoplasmic reticulum to the endosome, sensitizing the cells to TLR7 and TLR8 ligands.66 These effects depend on the uptake of anti phos pho lip id autoantibodies into the endosome, the activation of NADPH oxidase, and the generation of superoxide.66 Inhibition of TLR7 and TLR8 may be an attractive therapeutic target in patients with SLE and anti phos pho lip id autoantibodies. In keeping with this idea, hydroxy chloroquine has been shown to inhibit TLR769 (Fig. 3), and it is associated with a re-duced odds ratio for the persistence of anti phos-pho lip id autoantibodies in patients with SLE.70

B-cell activating factor (BAFF) is a cytokine that is crucial for B-cell survival.105 The BAFF-inhibiting antibody belimumab has recently been approved for the treatment of SLE.106 BAFF inhi-bition prevented thrombosis in NZW x BXSB F1 male mice,71 a finding that suggests it may have a role in the prevention of thrombosis associated with the anti phos pho lip id syndrome in high-risk patients with SLE (Fig. 3).

Disclosure forms provided by the authors are available with the full text of this article at NEJM.org.

We thank Michael Lockshin, Chris Weatherall, and Bob Graham for reading an earlier version of the manuscript and making sug-gestions to improve it; and Qi Miao for assistance with earlier versions of the figures.

References

1. Miyakis S, Lockshin MD, Atsumi T, et al. International consensus statement on an update of the classification criteria for definite anti phos pho lip id syndrome (APS). J Thromb Haemost 2006;4:295-306.2. Cervera R, Piette JC, Font J, et al. Anti-phos pho lip id syndrome: clinical and im-munologic manifestations and patterns of disease expression in a cohort of 1,000 pa-tients. Arthritis Rheum 2002;46:1019-27.3. Cervera R, Bucciarelli S, Plasin MA, et al. Catastrophic anti phos pho lip id syn-drome (CAPS): descriptive analysis of a series of 280 patients from the “CAPS Registry.” J Autoimmun 2009;32:240-5.4. McNeil HP, Simpson RJ, Chesterman CN, Krilis SA. Anti-phospholipid antibod-ies are directed against a complex antigen that includes a lipid-binding inhibitor of coagulation: beta 2-gly co pro tein I (apoli-poprotein H). Proc Natl Acad Sci U S A 1990;87:4120-4.5. Roubey RA, Pratt CW, Buyon JP, Win-field JB. Lupus anticoagulant activity of

autoimmune anti phos pho lip id antibodies is dependent upon beta 2-gly co pro tein I. J Clin Invest 1992;90:1100-4.6. Bevers EM, Galli M, Barbui T, Com-furius P, Zwaal RF. Lupus anticoagulant IgG’s (LA) are not directed to phospholip-ids only, but to a complex of lipid-bound human prothrombin. Thromb Haemost 1991;66:629-32.7. Galli M, Luciani D, Bertolini G, Barbui T. Lupus anticoagulants are stronger risk factors for thrombosis than anticardio-lipin antibodies in the anti phos pho lip id syndrome: a systematic review of the lit-erature. Blood 2003;101:1827-32.8. Urbanus RT, Siegerink B, Roest M, Rosendaal FR, de Groot PG, Algra A. Anti phos pho lip id antibodies and risk of myocardial infarction and ischaemic stroke in young women in the RATIO study: a case-control study. Lancet Neurol 2009;8:998-1005.9. Lockshin MD, Kim M, Laskin CA, et al. Prediction of adverse pregnancy out-

come by the presence of lupus anticoagu-lant, but not anticardiolipin antibody, in patients with anti phos pho lip id antibod-ies. Arthritis Rheum 2012;64:2311-8.10. de Groot PG, Lutters B, Derksen RH, Lisman T, Meijers JC, Rosendaal FR. Lupus anticoagulants and the risk of a first epi-sode of deep venous thrombosis. J Thromb Haemost 2005;3:1993-7.11. Ruiz-Irastorza G, Crowther M, Branch W, Khamashta MA. Anti phos pho lip id syndrome. Lancet 2010;376:1498-509.12. Gris JC, Bouvier S, Molinari N, et al. Comparative incidence of a first throm-botic event in purely obstetric anti phos-pho lip id syndrome with pregnancy loss: the NOH-APS observational study. Blood 2012;119:2624-32.13. de Laat HB, Derksen RH, Urbanus RT, Roest M, de Groot PG. Beta2-gly co pro tein I-dependent lupus anticoagulant highly correlates with thrombosis in the anti-phos pho lip id syndrome. Blood 2004;104: 3598-602.

The New England Journal of Medicine Downloaded from nejm.org at ANNE ARUNDEL MEDICAL CENTER on April 4, 2013. For personal use only. No other uses without permission.

Copyright © 2013 Massachusetts Medical Society. All rights reserved.

Page 10: The Pathogenesis of the Antiphospholipid Syndrome · The Pathogenesis of the Antiphospholipid Syndrome Bill ... have the ability to prolong clotting time in ... mation when infused

T h e n e w e ngl a nd j o u r na l o f m e dic i n e

n engl j med 368;11 nejm.org march 14, 20131042

14. Devreese K, Peerlinck K, Hoylaerts MF. Thrombotic risk assessment in the anti-phos pho lip id syndrome requires more than the quantification of lupus antico-agulants. Blood 2010;115:870-8.15. Pengo V, Ruffatti A, Legnani C, et al. Incidence of a first thromboembolic event in asymptomatic carriers of high-risk anti phos pho lip id antibody profile: a mul-ticenter prospective study. Blood 2011; 118:4714-8.16. Reddel SW, Wang YX, Sheng YH, Krilis SA. Epitope studies with anti-beta 2-gly co pro tein I antibodies from autoan-tibody and immunized sources. J Autoim-mun 2000;15:91-6.17. Iverson GM, Reddel S, Victoria EJ, et al. Use of single point mutations in do-main I of beta 2-gly co pro tein I to deter-mine fine antigenic specificity of anti-phos pho lip id autoantibodies. J Immunol 2002;169:7097-103.18. Banzato A, Pozzi N, Frasson R, et al. Antibodies to domain I of beta(2)gly co-pro tein I are in close relation to patients risk categories in anti phos pho lip id syn-drome (APS). Thromb Res 2011;128:583-6.19. de Laat B, Derksen RH, Urbanus RT, de Groot PG. IgG antibodies that recog-nize epitope Gly40-Arg43 in domain I of beta 2-gly co pro tein I cause LAC, and their presence correlates strongly with throm-bosis. Blood 2005;105:1540-5.20. Otomo K, Atsumi T, Amengual O, et al. Efficacy of the anti phos pho lip id score for the diagnosis of anti phos pho lip id syn-drome and its predictive value for throm-botic events. Arthritis Rheum 2012;64: 504-12.21. Sciascia S, Murru V, Sanna G, Roc-catello D, Khamashta MA, Bertolaccini ML. Clinical accuracy for diagnosis of anti phos pho lip id syndrome in SLE: evalu-ation of 23 possible combinations of anti-phos pho lip id antibody specificities. J Thromb Haemost 2012 October 1 (Epub ahead of print).22. Galli M, Reber G, de Moerloose P, de Groot PG. Invitation to a debate on the serological criteria that define the anti-phos pho lip id syndrome. J Thromb Hae-most 2008;6:399-401.23. Fischetti F, Durigutto P, Pellis V, et al. Thrombus formation induced by antibod-ies to beta2-gly co pro tein I is complement dependent and requires a priming factor. Blood 2005;106:2340-6.24. Arad A, Proulle V, Furie RA, Furie BC, Furie B. β(2)-gly co pro tein-1 autoantibod-ies from patients with anti phos pho lip id syndrome are sufficient to potentiate ar-terial thrombus formation in a mouse model. Blood 2011;117:3453-9.25. Ioannou Y, Zhang JY, Qi M, et al. Nov-el assays of thrombogenic pathogenicity for the anti phos pho lip id syndrome based on the detection of molecular oxidative modification of the major autoantigen β2-gly co pro tein I. Arthritis Rheum 2011; 63:2774-82.

26. Delgado Alves J, Ames PR, Donohue S, et al. Antibodies to high-density lipo-protein and beta2-gly co pro tein I are in-versely correlated with paraoxonase activ-ity in systemic lupus erythematosus and primary anti phos pho lip id syndrome. Ar-thritis Rheum 2002;46:2686-94.27. Charakida M, Besler C, Batuca JR, et al. Vascular abnormalities, paraoxonase activity, and dysfunctional HDL in pri-mary anti phos pho lip id syndrome. JAMA 2009;302:1210-7.28. Ames PR, Nourooz-Zadeh J, Tomma-sino C, Alves J, Brancaccio V, Anggard EE. Oxidative stress in primary anti phos pho-lip id syndrome. Thromb Haemost 1998; 79:447-9.29. Perez-Sanchez C, Ruiz-Limon P, Aguirre MA, et al. Mitochondrial dysfunction in anti phos pho lip id syndrome: implications in the pathogenesis of the disease and ef-fects of coenzyme Q(10) treatment. Blood 2012;119:5859-70.30. Ioannou Y, Zhang JY, Passam FH, et al. Naturally occurring free thiols within {beta}2-gly co pro tein I in vivo: nitrosyla-tion, redox modification by endothelial cells, and regulation of oxidative stress-induced cell injury. Blood 2010;116: 1961-70.31. Nishimura S, Manabe I, Nagasaki M, et al. In vivo imaging visualizes discoid platelet aggregations without endotheli-um disruption and implicates contribu-tion of inflammatory cytokine and integ-rin signaling. Blood 2012;119(8):e45-e56.32. Ames PR, Batuca JR, Ciampa A, Iannaccone L, Delgado Alves J. Clinical relevance of nitric oxide metabolites and nitrative stress in thrombotic primary anti phos pho lip id syndrome. J Rheumatol 2010;37:2523-30.33. Ramesh S, Morrell CN, Tarango C, et al. Anti phos pho lip id antibodies promote leukocyte-endothelial cell adhesion and thrombosis in mice by antagonizing eNOS via β2GPI and apoER2. J Clin Invest 2011;121:120-31.34. Laufs U, La Fata V, Plutzky J, Liao JK. Upregulation of endothelial nitric oxide synthase by HMG CoA reductase inhibi-tors. Circulation 1998;97:1129-35.35. Meroni PL, Raschi E, Testoni C, et al. Statins prevent endothelial cell activation induced by anti phos pho lip id (anti-beta2-gly co pro tein I) antibodies: effect on the proadhesive and proinflammatory pheno-type. Arthritis Rheum 2001;44:2870-8.36. Ferrara DE, Liu X, Espinola RG, et al. Inhibition of the thrombogenic and in-flammatory properties of anti phos pho-lip id antibodies by fluvastatin in an in vivo animal model. Arthritis Rheum 2003;48:3272-9.37. Urbanus RT, Pennings MT, Derksen RH, de Groot PG. Platelet activation by dimeric beta2-gly co pro tein I requires sig-naling via both gly co pro tein Ibalpha and apolipoprotein E receptor 2′. J Thromb Haemost 2008;6:1405-12.

38. Shi T, Giannakopoulos B, Yan X, et al. Anti-beta2-gly co pro tein I antibodies in complex with beta2-gly co pro tein I can ac-tivate platelets in a dysregulated manner via gly co pro tein Ib-IX-V. Arthritis Rheum 2006;54:2558-67.39. Sorice M, Longo A, Capozzi A, et al. Anti-beta2-gly co pro tein I antibodies in-duce monocyte release of tumor necrosis factor alpha and tissue factor by signal transduction pathways involving lipid rafts. Arthritis Rheum 2007;56:2687-97.40. Satta N, Kruithof EK, Fickentscher C, et al. Toll-like receptor 2 mediates the ac-tivation of human monocytes and endo-thelial cells by anti phos pho lip id antibod-ies. Blood 2011;117:5523-31.41. Lambrianides A, Carroll CJ, Pierangeli SS, et al. Effects of polyclonal IgG derived from patients with different clinical types of the anti phos pho lip id syndrome on monocyte signaling pathways. J Immunol 2010;184:6622-8.42. Ma K, Simantov R, Zhang JC, Silver-stein R, Hajjar KA, McCrae KR. High af-finity binding of beta 2-gly co pro tein I to human endothelial cells is mediated by annexin II. J Biol Chem 2000;275:15541-8.43. Allen KL, Fonseca FV, Betapudi V, Willard B, Zhang J, McCrae KR. A novel pathway for human endothelial cell acti-vation by anti phos pho lip id/anti-β2 gly co-pro tein I antibodies. Blood 2012;119: 884-93.44. Raschi E, Testoni C, Bosisio D, et al. Role of the MyD88 transduction signaling pathway in endothelial activation by anti-phos pho lip id antibodies. Blood 2003;101: 3495-500.45. Romay-Penabad Z, Montiel-Manzano MG, Shilagard T, et al. Annexin A2 is in-volved in anti phos pho lip id antibody- mediated pathogenic effects in vitro and in vivo. Blood 2009;114:3074-83.46. Pierangeli SS, Vega-Ostertag ME, Ras-chi E, et al. Toll-like receptor and anti-phos pho lip id mediated thrombosis: in vivo studies. Ann Rheum Dis 2007;66: 1327-33.47. Kolyada A, Lee CJ, De Biasio A, Beg-lova N. A novel dimeric inhibitor target-ing beta2GPI in beta2GPI/antibody com-plexes implicated in anti phos pho lip id syndrome. PLoS One 2010;5(12):e15345.48. Romay-Penabad Z, Aguilar-Valenzue-la R, Urbanus RT, et al. Apolipoprotein E receptor 2 is involved in the thrombotic complications in a murine model of the anti phos pho lip id syndrome. Blood 2011; 117:1408-14.49. Ioannou Y, Romay-Penabad Z, Pericle-ous C, et al. In vivo inhibition of anti phos-pho lip id antibody-induced pathogenicity utilizing the antigenic target peptide do-main I of beta2-gly co pro tein I: proof of concept. J Thromb Haemost 2009;7: 833-42.50. Chen J, Reheman A, Gushiken FC, et al. N-acetylcysteine reduces the size and activity of von Willebrand factor in hu-

The New England Journal of Medicine Downloaded from nejm.org at ANNE ARUNDEL MEDICAL CENTER on April 4, 2013. For personal use only. No other uses without permission.

Copyright © 2013 Massachusetts Medical Society. All rights reserved.

Page 11: The Pathogenesis of the Antiphospholipid Syndrome · The Pathogenesis of the Antiphospholipid Syndrome Bill ... have the ability to prolong clotting time in ... mation when infused

mechanisms of disease

n engl j med 368;11 nejm.org march 14, 2013 1043

man plasma and mice. J Clin Invest 2011; 121:593-603.51. López-Pedrera C, Buendia P, Cuadra-do MJ, et al. Anti phos pho lip id antibodies from patients with the anti phos pho lip id syndrome induce monocyte tissue factor expression through the simultaneous ac-tivation of NF-kappaB/Rel proteins via the p38 mitogen-activated protein kinase pathway, and of the MEK-1/ERK pathway. Arthritis Rheum 2006;54:301-11.52. Ritis K, Doumas M, Mastellos D, et al. A novel C5a receptor-tissue factor cross-talk in neutrophils links innate immunity to coagulation pathways. J Immunol 2006;177:4794-802.53. Vega-Ostertag M, Casper K, Swerlick R, Ferrara D, Harris EN, Pierangeli SS. Involvement of p38 MAPK in the up-regu-lation of tissue factor on endothelial cells by anti phos pho lip id antibodies. Arthritis Rheum 2005;52:1545-54.54. Seshan SV, Franzke CW, Redecha P, Monestier M, Mackman N, Girardi G. Role of tissue factor in a mouse model of thrombotic microangiopathy induced by anti phos pho lip id antibodies. Blood 2009; 114:1675-83.55. Jasuja R, Passam FH, Kennedy DR, et al. Protein disulfide isomerase inhibitors constitute a new class of antithrombotic agents. J Clin Invest 2012;122:2104-13.56. Giannakopoulos B, Gao L, Qi M, et al. Factor XI is a substrate for oxidoreduc-tases: enhanced activation of reduced FXI and its role in anti phos pho lip id syndrome thrombosis. J Autoimmun 2012;39:121-9.57. Tucker EI, Marzec UM, White TC, et al. Prevention of vascular graft occlusion and thrombus-associated thrombin gen-eration by inhibition of factor XI. Blood 2009;113:936-44.58. Rand JH, Wu XX, Andree HA, et al. Pregnancy loss in the anti phos pho lip id-antibody syndrome — a possible throm-bogenic mechanism. N Engl J Med 1997; 337:154-60. [Erratum, N Engl J Med 1997; 337:1327.]59. Rand JH, Wu XX, Quinn AS, et al. Hy-droxychloroquine protects the annexin A5 anticoagulant shield from disruption by anti phos pho lip id antibodies: evidence for a novel effect for an old antimalarial drug. Blood 2010;115:2292-9.60. Edwards MH, Pierangeli S, Liu X, Barker JH, Anderson G, Harris EN. Hy-droxychloroquine reverses thrombogenic properties of anti phos pho lip id antibodies in mice. Circulation 1997;96:4380-4.61. Pierangeli SS, Girardi G, Vega-Oster-tag M, Liu X, Espinola RG, Salmon J. Re-quirement of activation of complement C3 and C5 for anti phos pho lip id antibody-mediated thrombophilia. Arthritis Rheum 2005;52:2120-4.62. Holers VM, Girardi G, Mo L, et al. Complement C3 activation is required for anti phos pho lip id antibody-induced fetal loss. J Exp Med 2002;195:211-20.63. Girardi G, Berman J, Redecha P, et al.

Complement C5a receptors and neutro-phils mediate fetal injury in the anti phos-pho lip id syndrome. J Clin Invest 2003; 112:1644-54. [Erratum, J Clin Invest 2004;113:646.]64. Lonze BE, Singer AL, Montgomery RA. Eculizumab and renal transplanta-tion in a patient with CAPS. N Engl J Med 2010;362:1744-5.65. Shapira I, Andrade D, Allen SL, Salm-on JE. Brief report: induction of sustained remission in recurrent catastrophic anti-phos pho lip id syndrome via inhibition of terminal complement with eculizumab. Arthritis Rheum 2012;64:2719-23.66. Prinz N, Clemens N, Strand D, et al. Anti phos pho lip id antibodies induce translocation of TLR7 and TLR8 to the endosome in human monocytes and plas-macytoid dendritic cells. Blood 2011;118: 2322-32.67. Hashimoto Y, Kawamura M, Ichikawa K, et al. Anticardiolipin antibodies in NZW x BXSB F1 mice: a model of anti-phos pho lip id syndrome. J Immunol 1992; 149:1063-8.68. Pisitkun P, Deane JA, Difilippantonio MJ, Tarasenko T, Satterthwaite AB, Bol-land S. Autoreactive B cell responses to RNA-related antigens due to TLR7 gene duplication. Science 2006;312:1669-72.69. Kuznik A, Bencina M, Svajger U, Jeras M, Rozman B, Jerala R. Mechanism of en-dosomal TLR inhibition by antimalarial drugs and imidazoquinolines. J Immunol 2011;186:4794-804.70. Broder A, Putterman C. Hydroxychlo-roquine use is associated with lower odds of persistently positive anti phos pho lip id antibodies and/or lupus anticoagulant in systemic lupus erythematosus. J Rheuma-tol 2012 August 1 (Epub ahead of print).71. Kahn P, Ramanujam M, Bethunaickan R, et al. Prevention of murine anti phos-pho lip id syndrome by BAFF blockade. Ar-thritis Rheum 2008;58:2824-34.72. Kobayashi K, Kishi M, Atsumi T, et al. Circulating oxidized LDL forms complex-es with beta2-gly co pro tein I: implication as an atherogenic autoantigen. J Lipid Res 2003;44:716-26.73. Schwarzenbacher R, Zeth K, Diederichs K, et al. Crystal structure of human beta2-gly co pro tein I: implications for phos pho-lipid binding and the anti phos pho lip id syndrome. EMBO J 1999;18:6228-39.74. Passam FH, Rahgozar S, Qi M, et al. Beta 2 gly co pro tein I is a substrate of thiol oxidoreductases. Blood 2010;116:1995-7.75. Passam FH, Rahgozar S, Qi M, et al. Redox control of β2-gly co pro tein I–von Willebrand factor interaction by thiore-doxin-1. J Thromb Haemost 2010;8: 1754-62.76. Arai T, Yoshida K, Kaburaki J, et al. Autoreactive CD4(+) T-cell clones to beta2-gly co pro tein I in patients with anti-phos pho lip id syndrome: preferential rec-ognition of the major phospholipid-bind-ing site. Blood 2001;98:1889-96.

77. Agar C, van Os GM, Morgelin M, et al. Beta2-gly co pro tein I can exist in 2 confor-mations: implications for our under-standing of the anti phos pho lip id syn-drome. Blood 2010;116:1336-43.78. van Os GM, Meijers JC, Agar C, et al. Induction of anti-β2-gly co pro tein I auto-antibodies in mice by protein H of Strep-tococcus pyogenes. J Thromb Haemost 2011;9:2447-56.79. Meroni PL, Borghi MO, Raschi E, Te-desco F. Pathogenesis of anti phos pho lip id syndrome: understanding the antibodies. Nat Rev Rheumatol 2011;7:330-9.80. Agostinis C, Biffi S, Garrovo C, et al. In vivo distribution of β2 gly co pro tein I under various pathophysiologic condi-tions. Blood 2011;118:4231-8.81. Asherson RA. The catastrophic anti-phos pho lip id syndrome, 1998: a review of the clinical features, possible pathogene-sis and treatment. Lupus 1998;7:Suppl 2: S55-S62.82. Basu S, Helmersson J, Jarosinska D, Sällsten G, Mazzolai B, Barregård L. Reg-ulatory factors of basal F(2)-isoprostane formation: population, age, gender and smoking habits in humans. Free Radic Res 2009;43:85-91.83. Madureira PA, Hill R, Miller VA, Gia-comantonio C, Lee PW, Waisman DM. Annexin A2 is a novel cellular redox regu-latory protein involved in tumorigenesis. Oncotarget 2011;2:1075-93.84. Palmer RM, Ferrige AG, Moncada S. Nitric oxide release accounts for the bio-logical activity of endothelium-derived relaxing factor. Nature 1987;327:524-6.85. De Caterina R, Libby P, Peng HB, et al. Nitric oxide decreases cytokine-induced endothelial activation: nitric oxide selec-tively reduces endothelial expression of adhesion molecules and proinflammato-ry cytokines. J Clin Invest 1995;96:60-8.86. Radomski MW, Palmer RM, Moncada S. Endogenous nitric oxide inhibits hu-man platelet adhesion to vascular endo-thelium. Lancet 1987;2:1057-8.87. Palmer RM, Ashton DS, Moncada S. Vascular endothelial cells synthesize ni-tric oxide from L-arginine. Nature 1988; 333:664-6.88. Beckman JS, Beckman TW, Chen J, Marshall PA, Freeman BA. Apparent hy-droxyl radical production by peroxyni-trite: implications for endothelial injury from nitric oxide and superoxide. Proc Natl Acad Sci U S A 1990;87:1620-4.89. Lauer T, Preik M, Rassaf T, et al. Plas-ma nitrite rather than nitrate reflects re-gional endothelial nitric oxide synthase activity but lacks intrinsic vasodilator ac-tion. Proc Natl Acad Sci U S A 2001;98: 12814-9.90. Lee CJ, De Biasio A, Beglova N. Mode of interaction between beta2GPI and lipo-protein receptors suggests mutually ex-clusive binding of beta2GPI to the recep-tors and anionic phospholipids. Structure 2010;18:366-76.

The New England Journal of Medicine Downloaded from nejm.org at ANNE ARUNDEL MEDICAL CENTER on April 4, 2013. For personal use only. No other uses without permission.

Copyright © 2013 Massachusetts Medical Society. All rights reserved.

Page 12: The Pathogenesis of the Antiphospholipid Syndrome · The Pathogenesis of the Antiphospholipid Syndrome Bill ... have the ability to prolong clotting time in ... mation when infused

n engl j med 368;11 nejm.org march 14, 20131044

mechanisms of disease

91. Montiel-Manzano G, Romay-Penabad Z, Papalardo de Martinez E, et al. In vivo effects of an inhibitor of nuclear factor-kappa B on thrombogenic properties of anti phos pho lip id antibodies. Ann N Y Acad Sci 2007;1108:540-53.92. Versteeg HH, Ruf W. Thiol pathways in the regulation of tissue factor pro-thrombotic activity. Curr Opin Hematol 2011;18:343-8.93. Meijers JC, Tekelenburg WL, Bouma BN, Bertina RM, Rosendaal FR. High lev-els of coagulation factor XI as a risk factor for venous thrombosis. N Engl J Med 2000;342:696-701.94. Yang DT, Flanders MM, Kim H, Rodg-ers GM. Elevated factor XI activity levels are associated with an increased odds ra-tio for cerebrovascular events. Am J Clin Pathol 2006;126:411-5.95. Emsley J, McEwan PA, Gailani D. Structure and function of factor XI. Blood 2010;115:2569-77.96. Müller F, Gailani D, Renne T. Factor XI and XII as antithrombotic targets. Curr Opin Hematol 2011;18:349-55.97. Marcinkiewicz MM, Sinha D, Walsh PN. Productive recognition of factor IX by factor XIa exosites requires disulfide link-

age between heavy and light chains of fac-tor XIa. J Biol Chem 2012;287:6187-95.98. Sikara MP, Routsias JG, Samiotaki M, Panayotou G, Moutsopoulos HM, Vla-choyiannopoulos PG. β2 Gly co pro tein I (β2GPI) binds platelet factor 4 (PF4): im-plications for the pathogenesis of anti-phos pho lip id syndrome. Blood 2010;115: 713-23.99. de Laat B, Wu XX, van Lummel M, Derksen RH, de Groot PG, Rand JH. Cor-relation between anti phos pho lip id anti-bodies that recognize domain I of beta2-gly co pro tein I and a reduction in the anticoagulant activity of annexin A5. Blood 2007;109:1490-4.100. Love PE, Santoro SA. Anti phos pho-lip id antibodies: anticardiolipin and the lupus anticoagulant in systemic lupus er-ythematosus (SLE) and in non-SLE disor-ders: prevalence and clinical significance. Ann Intern Med 1990;112:682-98.101. Wahl DG, Guillemin F, de Maistre E, Perret C, Lecompte T, Thibaut G. Risk for venous thrombosis related to anti phos-pho lip id antibodies in systemic lupus ery-thematosus — a meta-analysis. Lupus 1997;6:467-73.102. Bruce IN, Clark-Soloninka CA,

Spitzer KA, Gladman DD, Urowitz MB, Laskin CA. Prevalence of antibodies to beta2-gly co pro tein I in systemic lupus erythematosus and their association with anti phos pho lip id antibody syndrome cri-teria: a single center study and literature review. J Rheumatol 2000;27:2833-7.103. Lau CM, Broughton C, Tabor AS, et al. RNA-associated autoantigens activate B cells by combined B cell antigen recep-tor/Toll-like receptor 7 engagement. J Exp Med 2005;202:1171-7.104. Christensen SR, Shupe J, Nickerson K, Kashgarian M, Flavell RA, Shlomchik MJ. Toll-like receptor 7 and TLR9 dictate autoantibody specificity and have oppos-ing inflammatory and regulatory roles in a murine model of lupus. Immunity 2006;25:417-28.105. Mackay F, Schneider P. Cracking the BAFF code. Nat Rev Immunol 2009;9:491-502.106. Fairfax K, Mackay IR, Mackay F. BAFF/BLyS inhibitors: a new prospect for treatment of systemic lupus erythemato-sus. IUBMB Life 2012;64:595-602.Copyright © 2013 Massachusetts Medical Society.

specialties and topics at nejm.org

Specialty pages at the Journal’s website (NEJM.org) feature articles in cardiology, endocrinology, genetics, infectious disease, nephrology, pediatrics, and many other medical specialties. These pages, along with collections of articles on clinical and nonclinical topics, offer links to interactive and multimedia content and feature

recently published articles as well as material from the NEJM archive (1812–1989).

The New England Journal of Medicine Downloaded from nejm.org at ANNE ARUNDEL MEDICAL CENTER on April 4, 2013. For personal use only. No other uses without permission.

Copyright © 2013 Massachusetts Medical Society. All rights reserved.