Professional Documents
Culture Documents
Background: Recurrent pregnancy loss (RPL) is a sig- firmed by means of fixed- and random-effects meta-
nificant clinical problem. Recently, thrombophilias have analyses models.
been implicated as a possible cause. Factor V Leiden (FVL)
and prothrombin gene (G20210A) mutations are the most Results: The combined odds ratios for the association
common types of hereditary thrombophilias, but are usu- between RPL and FVL and between RPL and G20210A
ally undiagnosed because most carriers are asymptom- were 2.0 (95% confidence interval, 1.5-2.7; P⬍.001) and
atic. The relationship between FVL, G20210A, and RPL 2.0 (95% confidence interval, 1.0-4.0; P = .03), respec-
has been investigated with conflicting results. This study tively. Similar results were produced by the logistic re-
analyzed existing data to determine whether an associa- gression and both fixed- and random-effects meta-
tion exists. analysis models.
Methods: A systematic review of the literature was per- Conclusions: Carriers of FVL or prothrombin gene mu-
formed. Only case-control studies that defined RPL as 2 tations have double the risk of experiencing 2 or more
or more pregnancy losses in the first or second trimes- miscarriages compared with women without thrombo-
ter and that confirmed mutations by DNA analysis were philias. Hereditary thrombophilias may be an unrecog-
included. Sixteen studies were selected for the FVL meta- nized cause of RPL. We recommend testing for these mu-
analysis and 7 for the G20210A analysis. Stratified and tations in women with RPL.
multivariate logistic regression analyses were per-
formed with the use of aggregate data. Results were con- Arch Intern Med. 2004;164:558-563
R
ECURRENT PREGNANCY LOSS cant thrombosis. However, when carriers
(RPL), defined as 2 or more are exposed to additional risk factors, such
spontaneous abortions, af- as pregnancy or possibly oral contracep-
fects approximately 5% of tives, the risk of life-threatening throm-
women of reproductive botic events is significantly increased and
age.1 Although several causes of RPL have may become clinically evident.16-17
been established, more than 50% of cases The 2 most common causes of heredi-
remain unexplained. This is a challeng- tary thrombophilias are factor V Leiden
ing dilemma for both patients and physi- (FVL) and prothrombin gene (G20210A)
From the Division of cians.1 Recently, thrombophilias have been mutations.18 Factor V Leiden is a point mu-
Reproductive Endocrinology
suggested as a possible cause of RPL.2-8 He- tation (G1691A) that results in an altered
and Infertility, Department of
Obstetrics and Gynecology reditary thrombophilias are a group of ge- factor V, which is resistant to inactivation
(Drs Kovalevsky, Gracia, and netic disorders of blood coagulation re- by protein C. This results in a hypercoagu-
Barnhart), and Center for sulting in a hypercoagulable state, which lable state with a 5- to 10-fold risk of throm-
Clinical Epidemiology and in turn can result in abnormal placenta- bosis in heterozygotes and an 80-fold risk
Biostatistics (Drs Kovalevsky, tion. Early in pregnancy this may mani- in homozygotes.17 Factor V Leiden is re-
Berlin, Sammel, and Barnhart), fest as spontaneous loss.9-10 In later preg- sponsible for 20% to 40% of isolated throm-
University of Pennsylvania nancy, thrombophilias have been associated botic events and 40% to 45% of familial
Medical Center, Philadelphia. with complications such as preeclampsia, thrombophilias. Its prevalence in the United
Dr Kovalevsky is now with
intrauterine growth restriction, placental States is estimated to be between 3% and
CONRAD, Department of
Obstetrics and Gynecology, abruption, and stillbirth.11-15 7%, with the highest frequency in whites.18-20
Eastern Virginia Medical Most carriers of the mutation will not Many studies have investigated the rela-
School, Norfolk. The authors develop any clinical signs and remain un- tionship between FVL and RPL, and the ma-
have no relevant financial diagnosed because these conditions result jority found an association, with odds ra-
interest in this article. in a small absolute risk of clinically signifi- tios ranging from 0.5 to 18.21-43
Age, y
No. of Minimum White Unexplained
Source Cases/Controls Cases Controls No. of Losses Trimester Only Only OR (95% CI)
Factor V Leiden
Balasch et al,31 1997 55/50 31.8 34.4 2 First No Yes 0.9 (0.1-14.9)
Dizon-Townson et al,42 1997 40/25 33.0 31.0 3 First and second Yes Yes ND*
Grandone et al,30 1997 27/118 31.0 35.0 2 First Yes Yes 1.8 (0.3-9.9)
Metz et al,29 1997 100/85 NA NA 3 First and second No No 1.7 (0.4-7.2)
Ridker et al,36 1998 206/437 NA NA 3 First and second Yes Yes 2.2 (1.1-4.5)
Pauer and Neesen,43 1998 84/87 32.9 NA 2 First and second No Yes 1.2 (0.4-3.2)
Souza et al,41 1999 56/384 29.6 24.3 3 First and second No No 4.8 (1.3-17.7)
Kutteh et al,32 1999 50/50 33.6 34.3 3 First and second Yes No 0.5 (0.04-5.6)
Tal et al,37 1999 78/125 31.4 30.7 2 First and second No Yes 4.4 (1.7-11.1)
Hashimoto et al,28 1999 52/55 NA NA 3 First No Yes ND*
Foka et al,33 2000 80/100 33.0 35.0 2 First and second Yes No 5.5 (1.8-17.4)
Wramsby et al,38 2000 84/69 NA NA 3 First and second No No 6.1 (1.3-28.2)
Younis et al,40 2000 78/139 30.0 30.7 2 First and second No Yes 3.9 (1.6-9.7)
Rai et al,34 2001 904/150 34.0 33.0 3 First Yes No 0.8 (0.4-1.5)
Raziel et al,35 2001 36/40 34.0 33.5 2 First and second Yes Yes 3.8 (0.7-20.2)
Pihusch et al,39 2001 102/128 35.0 32.0 2 First and second Yes Yes 0.9 (0.3-2.3)
Total or mean 2032/2042 32.4 32.2 2.5 8 Yes, 8 no 10 Yes, 6 no
Prothrombin Gene (G20210A)
Pickering et al,49 1998 122/66 35.0 31.0 3 First and second No Yes 0.7 (0.2-3.3)
Kutteh et al,32 1999 50/50 33.6 34.3 3 First and second Yes Yes 1.0 (0.06-16.4)
Souza et al,41 1999 56/384 29.6 24.3 3 First and second No No 3.5 (0.6-19.7)
Foka et al,33 2000 80/100 33.0 35.0 2 First and second Yes No 4.7 (0.9-23.3)
Wramsby et al,38 2000 84/69 NA NA 3 First and second No No 0.8 (0.2-4.2)
Raziel et al,35 2001 36/40 34.0 33.5 2 First and second Yes Yes 2.3 (0.2-26.4)
Pihusch et al,39 2001 102/128 35.0 32.0 2 First and second Yes Yes 6.5 (0.8-56.9)
Total or mean 530/837 33.4 31.7 2.6 4 Yes, 3 no 4 Yes, 3 no
Abbreviations: CI, confidence interval; NA, not available; ND, not determined; OR, odds ratio.
*Could not be calculated because none of the cases or controls had the mutation.