Age and homocysteine plasma levels are risk factors ...

Hum. Reprod. Advance Access published June 3, 2004

Human Reproduction Page 1 of 4

DOI: 10.1093/humrep/deh346

Age and homocysteine plasma levels are risk factors for thrombotic complications after ovarian stimulation E.Grandone1,5, D.Colaizzo1, P.Vergura1, F.Cappucci1, G.Vecchione1, A.Lo Bue2, E.Cittadini3 and M.Margaglione4 1

Unita` di Aterosclerosi e Trombosi, IRCCS Casa Sollievo della Sofferenza, S. Giovanni R. (Foggia), Viale Padre Pio, 71013, Centro di Biologia della Riproduzione, Palermo, 3Cattedra di Ostetricia e Ginecologia, Istituto Materno-Infantile, Via V. Villareale, 54, 90141, Palermo and 4Genetica Medica, Universita` degli Studi di Foggia, Viale Pinto, Foggia 71100, Italy

2

5

To whom correspondence should be addressed at: IRCCS ‘Casa Sollievo della Sofferenza’, Poliambulatorio Giovanni Paolo II Viale Padre Pio, S. Giovanni R. (Foggia), 71013, Italy. E-mail: [email protected]

Key words: age/homocysteine/IVF/thrombophilia/thrombotic events

Introduction It is well established that, as in pregnancy and puerperium, a thrombotic risk is associated with hormonal therapies. IVF procedures involve the controlled pharmacological stimulation of ovarian follicles using GnRH and exogenous gonadotrophins, and women who need ovarian stimulation of ovarian follicles using GnRH analogues or exogenous gonadotrophins may as a consequence suffer ovarian hyperstimulation syndrome (OHSS) (Whelan and Vlahos, 2000). Some case reports of both arterial and venous thromboses have been published (Baumann and Diedrich, 2000), but no data are available about the prevalence of thrombotic complications in women approaching IVF. Thromboembolic disease, as a complication of ovarian stimulation, is generally considered to be a rare occurring mainly during OHSS (Stewart et al., 1997). In some cases the presence of acquired or inherited thrombophilia has been demonstrated in patients showing thrombosis occurring during IVF cycles (Aurousseau et al., 1995; Loret de Mola et al., 2000).

It is generally accepted that high estrogen levels and an increased number of small ovarian follicles at time of ovulation induction predispose to the development of OHSS (Belaen et al., 2001). However, high estrogen levels and the number of ovarian follicles do not have a high predictive value for the development of severe OHSS (Loret de Mola et al., 2000). It has been suggested that before commencement of IVF therapy, patients should be counselled and carefully evaluated for the presence of thrombophilic factors. It has also been proposed to screen for acquired and inherited thrombophilia all patients receiving assisted conception treatment, although the magnitude of venous and arterial risk associated with IVF techniques has not been established (Arya et al., 2001). We carried out a study aimed at evaluating the prevalence of thrombotic events in women approaching ovarian stimulation for IVF. Moreover, we investigated a possible relationship between inherited or acquired thrombophilias and the occurrence of thrombotic events in this setting.

Human Reproduction q European Society of Human Reproduction and Embryology 2004; all rights reserved

Page 1 of 4

Downloaded from http://humrep.oxfordjournals.org/ by guest on June 3, 2013

BACKGROUND: The magnitude of thrombotic risk during ovarian stimulation cycles is not known. We calculated the magnitude of thrombotic risk in a cohort of women starting a new cycle of ovarian stimulation and investigated the role of inherited and acquired thrombophilia for these events. METHODS: This is an observational study involving outpatients of a clinical research centre. Consecutive women undergoing ovarian stimulation (n 5 305) were enrolled. Blood samples for studying inherited and acquired thrombophilia were obtained $2 months after the last cycle of treatment. Odds ratios (OR) and confidence intervals (CI) were determined for markers significantly associated with thrombotic events. Blood samples were analysed for inherited and acquired causes of thrombophilia (antithrombin, protein C, protein S, antiphospholipid antibodies, the Factor V Leiden and FIIA20210 mutations, the TT677 MTHFR genotype, and homocysteine plasma levels). RESULTS: Thrombotic events were observed in 4/747 cycles of ovarian stimulation, with a prevalence of 0.5%, corresponding to 1.6 per 100 000 cycles/woman. Age $39 years and homocysteine plasma levels above the 97.5 percentile were significantly associated with thrombotic events during IVF [OR 15.2 (95% CI 2.0– 115.0) and 14.4 (1.5 –141.3) respectively]. CONCLUSIONS: Age $39 years and mild hyperhomocysteinaemia are strongly associated with the occurrence of thrombotic events during IVF.

E.Grandone et al.

Materials and methods

DNA analysis Leukocyte DNA was obtained from frozen blood by the use of standard techniques. A 220 bp DNA fragment of the Factor V gene that includes nucleotide 1691, was amplified by PCR, as previously described (Grandone et al., 1997). To identify the G ! A mutation of the Factor II gene, a 345 bp fragment was obtained and then digested using the Hind III endonuclease, as described (Grandone et al., 1998). A fragment of 198 bp containing the C677T polymorphism of MTHFR gene was generated and digested using Hinf I endonuclease, as described elsewhere (Grandone et al., 1998). Statistical analysis All the analyses were performed according to the Statistical Package for Social Sciences (SPSS 6.1 for Macintosh). The significance of differences in means was evaluated by non-parametric test, whereas the significance of any difference in proportions was tested by x2-test. The significance of the difference of alleles and genotypes observed between the groups was tested using the x2-test, after grouping homozygous and heterozygous carriers of the FII

Page 2 of 4

Table I. Features of the whole sample Age years, median (range) Causes and type of infertility, n (%) Unexplained Male infertility Female infertility Cycles, median (range) Embryos transferred, median (range) Smoking habits, n (%) No smokers 1 –10 cigarettes/day 11 –20 cigarettes/day .20 cigarettes/day FV Leiden mutation, n (%) FIIA20210 mutation, n (%) MTHFR TT677 genotype, n (%) Confirmed antiphospholipid antibodies, n (%) Other Homocysteine (mmol/l), median (range) a

35 (20–46) 103 (33.8) 100 (32.8) 102 (33.4) 2 (1–10) 8 (1–38) 262 (85.9) 29 (9.5) 11 (3.6) 3 (1.0) 12 (3.9) 19 (6.2)a 54 (17.7) 15 (4.9) 1 Protein C deficiency 5.8 (1.2–32.0)

One homozygous.

mutation. Odds ratios (OR) and 95% confidence intervals (CI) were calculated.

Results Features of women are shown in Table I. Median age was 35 years (range 20 –46). Causes of infertility were equally distributed among male, female and unexplained causes (Table I). The majority (79.3%) of women did not smoke. The median number of cycles of ovarian stimulation was 2 (range 1– 10), with a median number of embryos transferred of 8 (range 1 – 38). Overall, these women were exposed to a total of 747 cycles of ovarian stimulation, in which 396 (53%) cycles used gonadotrophins and in 269 (36%) clomiphene citrate; for the remaining 82 (11%) cycles, this information was not available. Among them, 3.9% (n ¼ 12) carried the FV Leiden mutation (all heterozygotes) and 6.2% (n ¼ 19, one of them homozygote) the FII A20210 gene variant; finally, 17.7% (n ¼ 54) carried TT MTHFR genotype. The frequencies we observed were not significantly different from those found in a general population from the same ethnic background (Margaglione et al., 1998). Only one woman showed protein C deficiency, and she was also heterozygous for FV Leiden mutation. As far as acquired causes of thrombophilia are concerned, 4.9% (n ¼ 15) of women tested positive for lupus anticoagulant antibodies, two of them also showing anticardiolipin antibody IgG titre $ 40 GPL on at least two occasions at 6 –8 weeks apart, according to SSC SubCommettee (Exner et al., 1991). Median values of anticardiolipin antibody IgG were 2.9 GPL (range 0.1– 363): in four cases a moderate – high titre of anticardiolipin antibodies ($ 40 GPL) was recorded. Overall, two women suffered from a mild OHSS, defined as recommended by the Practice Commettee of the American Society for Reproductive Medicine (2003). One of them showed a thrombotic event. The other was a 32 year old


From February 2000 to March 2002, 305 women were consecutively referred to our Unit for a thrombophilic screening before starting a new cycle of ovarian stimulation. At that time a questionnaire was administered to all women. Detailed information about personal or family history of thromboembolic events, in addition to personal obstetric history and causes of infertility, was collected. No woman suffered from arterial hypertension, diabetes mellitus, dyslipidaemia, or had a history of arterial or venous thrombosis. Diagnosis of venous thromboembolism or superficial vein thrombosis was objectively confirmed by Doppler ultrasonography, ventilation – perfusion lung scanning, or pulmonary angiography. Cerebral ischaemic events were confirmed by magnetic resonance imaging. The arterial thromboses were documented by means of arteriography. Sixteen women had a diagnosis of polycystic ovary syndrome; none of them developed an ischaemic event. The study was carried out after informed consent was obtained as approved by the local Ethics Committee. Blood was drawn for determination of markers of thrombophilia, $ 2 months after the last cycle of ovarian stimulation. Citrated blood was centrifuged within 30 min at 3000 g for 10 min, and plasma aliquots were stored at 2808C. Antiphospholipid antibodies lupus anticoagulant (LA) and IgG anticardiolipin antibodies (aCL) (enzyme-linked immunosorbent assay; Byk Gulden, Italy) were determined in all patients, as reported elsewhere (Grandone et al., 1998). Activity of protein C, antithrombin and free protein S were measured by using chromogenic assays (Berichrom antithrombin III, Berichrom Protein C, DADE Behring; Asserachrom Free Protein S, Diagnostica Stago respectively). Clotting assays were performed on a BCT (DADE Behring). Inter- and intra-assay coefficients of all the variables were , 8.0% and , 5.0% respectively. Total plasma homocysteine (tHcy) was measured using LCMS/MS tandem spectrometry, as reported elsewhere (Magera et al., 1999). Briefly, after sample reduction and deproteinization, the analysis was performed in the multiple reaction monitoring mode in which tHcy and Hcy-d4 were detected through the transition from the precursor to the product ion. The retention time of tHcy and Hcy-d4 was 1.5 min in a 2.5 min analysis.

Thrombotic risk during ovarian stimulation

Table II. Features of women experiencing thrombotic events after IVF Patient

Age at event (years)

Body mass index (kg/m2)

Smoke

Cycle

Type of event

Type of thrombophilia

LE

28

29.6

0

1

Ischaemic stroke

DCA ST CN

41 39 41

20.9 18.7 27.3

1–10 0 0

2 2 5

SVT VTE and ischaemic strokea Radial arterial thrombosis and mesenteric vein thrombosisa

Antiphospholipid antibodies (87GPL) and mild hyperhomocysteinaemia 13.1 mmol/l Mild hyperhomocysteinaemia 12.7 mmol/l – OHSS

a After the same cycle. SVT ¼ superficial vein thrombosis; VTE ¼ venous thromboembolism.

Discussion This study shows that the thrombotic risk associated with IVF procedures is rather low, but age . 39 years and mild hyperhomocysteinaemia represent important risk factors for the occurrence of these events. Many women undergo repeated transfers without success, so they may need to be stimulated many times before achieving pregnancy. For this reason they may also undergo administration many times in a brief period. Severe forms of thromboses following ovarian stimulation for IVF have been reported in women with concomitant signs of OHSS

(Kodama et al., 1996; Stewart et al., 1997; Baumann and Diedrich, 2000). This is a condition known to be associated with thrombotic events, and usually treated also by means of antithrombotic therapy (Stewart et al., 1997; Baumann and Diedrich, 2000). So far, jugular vein thromboses have been mostly reported to be associated with hormonal ovarian stimulation (Arya et al., 2001). Many authors agree that these events (arterial and/or venous) are quite rare, but to date the magnitude of this risk is not known. The present is, to our knowledge, the first report investigating the thrombotic risk during cycles of ovulation induction. We observe an absolute prevalence of thrombotic events of 0.5%, corresponding to a rate of 1.6 per 100 000 cycles/woman, which is very low according to suggestions from case reports and anecdotal reports. In agreement with other thrombotic series (Abbott et al., 2003; Anderson and Spencer, 2003), we found that age is significantly associated with a higher risk of arterial and venous diseases in women treated with hormones for ovarian stimulation and induction of ovulation. Only in one case did we observe an adverse event at young age (LE), occurring in a woman presenting hyperhomocysteinaemia and antiphosholipid antibodies, both abnormalities known to be associated with arterial thrombotic events. Moreover, homocysteine plasma levels were significantly associated with the occurrence of both arterial and venous events in this setting, as documented in the literature (D’Angelo and Selhub, 1997). The role of inherited causes of thrombophilia that are known to predispose to thromboembolic events is not evident here, probably because we recorded a prevalence of arterial events. However, in our sample we had 31 women with FV Leiden or FII A20210 mutation, and none of them developed a thrombotic event during ovulation induction. Although thrombotic events after ovarian stimulation are quite rare, they can markedly affect the quality of life of woman. Our data show that age and tHcy plasma levels are strong risk factors for these events. In our setting, age $ 39 years and tHcy . 97.5th percentile confer a risk of thrombotic events 14- or 15-fold higher than that of women who are younger or who have lower tHcy plasma levels. Some authors recommend that before commencement of ovarian stimulation women should be counselled appropriately. Screening for thrombophilia in all women receiving Page 3 of 4


healthy woman with a body mass index of 21 kg/m2, who showed the OHSS at the first attempt of ovarian stimulation. tHcy median values were 5.8 mmol (range 1.22– 32); in all, 18 (6.8%) women showed a value . 97.5th percentile (10.14 mmol/l) calculated in a sample (n ¼ 100) of women within the same age range and from the same ethnic background. In this cohort, we recorded four thrombotic events: one superficial vein thrombosis, one deep vein thrombosis with pulmonary embolism, one ischaemic stroke and, finally, one contemporary presence of a radial arterial thrombosis and a mesenteric vein thrombosis requiring amputation of a large section of ileum. The latter event was recorded after a mild OHSS in woman who underwent five cycles of IVF. Her last ovulation induction was obtained by administration of purified FSH (75 IU 10 times per day for 12 days) and chorionic gonadotrophin (5000 IU). Nine oocytes were obtained. Among women who suffered from thrombotic events, two showed tHcy values . 97.5th percentile (Table II), one of them also tested positive for antiphospholipid antibodies and suffered from juvenile arterial thrombosis (Table II). One woman showed protein C deficiency in association with FV Leiden heterozygosis and was asymptomatic. Overall, we observed in our setting adverse events in 4/747 cycles of ovulation induction, with a prevalence of 0.5%, corresponding to 1.6 per 100 000 cycles/woman. We calculated that age $ 39 years was significantly associated with a thrombotic event after a cycle of ovarian stimulation (OR 14.4, 95% CI 1.5 – 141.3, P ¼ 0.01). Moreover, in our setting, homocysteine plasma levels . 97.5th percentile were significantly associated with these events (OR 15.2, 95% CI 2.0 – 115.0, P ¼ 0.03).

E.Grandone et al.

assisted conception treatment has been advocated, but it would be difficult to justify for such an infrequent complication. We think that every woman should be made properly aware before treatment of the risks she is going to undertake and that this risk is probably much higher if she is aged . 39 years and with elevated tHcy levels. In conclusion, our data suggest that, in addition to women with OHSS, an antithrombotic regimen and probably a supplementation with folic acid and/or B vitamins, should be recommended to this subgroup of women. However, controlled trials aimed at choosing correct thromboprophylaxis are warranted. References

Page 4 of 4

Submitted on January 29, 2004; accepted on May 7, 2004


Abbott RD, Curb JD, Rodriguez BL, Masaki KH, Popper JS, Ross GW and Petrovitch H (2003) Age-related changes in risk factor effects on the incidence of thromboembolic and hemorrhagic stroke. J Clin Epidemiol 56,479–486. Anderson FA, Jr and Spencer FA (2003) Risk factors for venous thromboembolism. Circulation 107(23 Suppl 1),I 9 –16. Arya R, Shehata HA, Patel RK, Sahu S, Rajasingam D, Harrington K, Nelson-Piercy C and Parsons JH (2001) Internal jugular vein thrombosis after assisted conception therapy. Br J Haematol 115,153– 155. Aurousseau MH, Samama MM, Belhassen A, Herve F and Hugues JN (1995) Risk of thromboembolism in relation to an in-vitro fertilization programme: three case reports. Hum Reprod 10,94–97. Baumann P and Diedrich K (2000) Thromboembolic complications associated with reproductive endocrinologic procedures. Haematol/Oncol Clin North Amer 14,431–443. Belaen B, Geerinckx K, Vergauwe P and Thys J (2001) Internal jugular vein thrombosis after ovarian stimulation: case report. Hum Reprod 16,510–512.

D’Angelo A and Selhub J (1997) Homocysteine and thrombotic disease. Blood 90,1– 11. Exner T, Triplett DA, Taberner D and Machin SJ (1991) Guidelines for testing and revised criteria for lupus anticoagulants. SSC Subcommittee for the Standardization of Lupus Anticoagulants. Thromb Haemost 65,320–322. Grandone E, Margaglione M, Colaizzo D, d’Addedda M, Cappucci G, Vecchione G, Scianname N, Pavone G and Di Minno G (1997) Factor V Leiden is associated with repeated and recurrent fetal losses. Thromb Haemost 77,822–824. Grandone E, Margaglione M, Colaizzo D, D’Andrea G, Cappucci G, Brancaccio V and Di Minno G (1998) Genetic susceptibility to pregnancyrelated venous thromboembolism: role of Factor V Leiden, prothrombin G20210A and C677T methylenetetrahydrofolate reductase mutations. Am J Obstet Gynecol 179,1324–1328. Kodama H, Fukuda J, Karube H, Matsui T, Shimuzu Y and Tanaka T (1996) Status of the coagulation and fibrinolytic systems in ovarian hyperstimulation syndrome. Fertil Steril 3,417–424. Loret de Mola JR, Kiwi R, Austin C and Goldfarb JM (2000) Subclavian deep vein thrombosis associated with the use of recombinant follicle-stimulating hormone (Gonal-F) complicating mild ovarian hyperstimulation syndrome. Fertil Steril 73,1253–1256. Magera MJ, Lacey JM, Casetta B and Rinaldo P (1999) Method for the determination of total homocysteine in plasma and urine by stable isotope dilution and electrospray tandem mass spectrometry. Clin Chem 45,1517– 1522. Margaglione M, Brancaccio V, Giuliani N, D’Andrea G, Cappucci G, Iannaccone L, Vecchione G, Grandone E and Di Minno G (1998) Increased risk for venous thrombosis in carriers of the prothrombin G ! A 20210 gene variant. Ann Intern Med 129,89–93. Practice Commettee of the American Society for Reproductive Medicine (2003) Ovarian hyperstimulation syndrome. Fertil Steril 80,1309–1314. Stewart JA, Hamilton PJ and Murdoch AP (1997) Thromboembolic disease associated with ovarian stimulation and assisted conception techniques. Hum Reprod 12,2167–2173. Whelan JG III and Vlahos NF (2000) The ovarian hyperstimulation syndrome. Fertil Steril 73,883–896.