{"paper_id":"7a606168-8dbb-4971-9f14-97eca5e5e1ab","body_text":"At present, the use of assisted reproductive\ntechnology (ART) is expanding worldwide. A related\nchallenge after ovarian hyperstimulation is\nmultiple gestation and the health of children born\nby ART. Recently, the rate of multiple pregnancies\nhas dramatically increased due to the widespread\nuse of ART ( 1 ). Multiple pregnancies are\nassociated with increased risk of maternal and\nfetal complications ( 2 ). The ideal of infertility\ntherapy is to achieve one healthy baby at a time\n( 3 ). Despite the attempts to limit the incidence of\nmultiple pregnancies after ART by elective single\nembryo transfer (SET), the average  in vitro  fertilization\n(IVF) treatment includes transfer of two,\nthree or sometimes more embryos into the uterus.\nSET, as a clinical practice, has not been executed\nin some countries.\nPrevious studies evaluated embryo and cyclespecific\nparameters associated with twin pregnancies\nafter double embryo transfer (DET)\n( 4 - 8 ). Niu et al. ( 4 ) found that four factors-the\nfirst treatment cycle, good ovarian response,\nhigher number of top-quality embryos, and development\nstage score of the second-best embryo\ntransferred had an independent association with\ntwin pregnancies after DET. Xu et al. ( 5 ) reported\nthat women’s age and the number of highquality\nembryos transferred were risk factors\nassociated with twin pregnancies after IVF with\nDET. Groeneveld et al. ( 6 ) demonstrated that the\nheight of the women and the number of oocytes\nretrieved were associated with an increased risk\nof twins after DET.\nIn a recent study, Kim et al. ( 8 ) concluded that\nyounger age, higher body weight, and better\nquality of transferred embryos showed an association\nwith increased chance for twin pregnancies\nafter DET at cleavage stage. Identification\nof risk factors for multiple pregnancies can enable\nmedical personnel to provide counseling and\ninformation to couples at high risk for multiple\npregnancies and suggest SET in their infertility\ntreatment program. The aim of present study is\nto evaluate the factors that affect the occurrence\nof multiple pregnancies after transfer of two or\nthree embryos in intracytoplasmic sperm injection\n(ICSI) by two different regression analyses\n(poisson and Logistic).\n\nThis retrospective cohort study reviewed the\nrecords of patients referred to Royan Institute between\nOctober 2011 and January 2012. The Institutional\nScientific Board of Royan Institute approved\nthis study. Admitted patients gave written\nconsent that stated which the treatment information\nwould be used for scientific purposes without\nmentioning names or personal details.\nThe data related to infertile couples who underwent\nICSI cycles and included: demographic\ninformation, medical records, and cycle characteristics.\nExcluded from the study were: women\nolder than 40 years, uterine factor infertility\n(myoma, polyps, and congenital malformations),\nrecurrent miscarriages, and stages II or III endometriosis.\nCycles with donor oocytes or embryos,\npreimplantation genetics diagnosis, and blastocyst\nembryo transfer were also excluded. In this\nstudy we reviewed all ICSI cycles that had complete\ndata. Embryo quality was recorded based on\nthe fragmentation degree and regularity of blastomeres\non day 3 after fertilization, as follows\n( 9 ): excellent: 6-8 equal sized blastomeres with\n≤10% frag mentation; good: 6-8 equal or unequal\nsized blastomeres with 10-20% fragmentation;\nand fair: uneven sized and few blastomeres with\n>20% fragmentation. We calculated the fertilization\nrate according to the number of fertilized oocytes\nper number of microinjected MII oocytes.\nMultiple pregnancies were considered as the observation\nof more than one gestational sac with\nheart beats by vaginal ultrasound evaluation six\nweeks after embryo transfer.\nThe findings were described as mean ± SD for\nquantitative variables and number (%) for qualitative\nvariables. We compared the study population\ncharacteristics according to the number of\nembryos transferred, either DET or triple embryo\ntransfer (TET) according to the student’s t test\nand chi-square test for quantitative and qualitative\nvariables, respectively. To investigate the\nrelationship between factors and gestational sac,\nwe used the statistical software STATA 11 program\nand the logistic and poisson regression (PR)\nmodels. On the basis of previous studies ( 5 ,  8 ,  10 ,\n 11 ), the possible variables that affected the number of gestational sacs included: women’s age,\nbody mass index (BMI), infertility type, infertility\ncause and duration, acquired uterine anomaly,\nendometrial thickness on the day of the human\nchorionic gonadotropin (hCG) injection, type of\nstimulation protocol, present cycle type, ovarian\nresponse type, total number of gonadotropin ampules,\ntotal number of retrieved oocytes, number\nof MII oocytes, fertilization rate, and quality of\ntransferred embryos were listed in the PR model\nfor both study populations (DET and TET). Multiple\nlogistic regression (LR) model was used to\ndetermine significant variables related to multiple\npregnancies. We considered P<0.05 to be statistically\nsignificant.\n\nA total of 1000 patients referred to Royan Institute\nfor ART during the study period. We included 606\npatients in this study. There were 336 patients with\nDET and 270 patients with TET. Table 1 lists the\ndemographic and medical characteristics of patients\naccording to group (DET or TET). The women’s\nmean age, duration of infertility, and number of\nprevious ART cycles in the TET group were greater\nthan those in the DET group (P<0.001). The two\ngroups showed no significant differences in terms\nof BMI, infertility type, endometrial thickness, and\nquality of transferred embryos. The DET group had\na higher mean number of retrieved and MII oocytes.\nDespite a higher fertilization rate in the TET group,\nwe observed clinical pregnancy rate between the\ntwo groups (P=0.1).\nThe distribution of the frequency of the numberof\nsacs in the DET and TET groups were shown in\nTable 2. The results showed that the mean of the\nobserved gestational sac in the TET group (0.64\n± 0.88) was greater than the DET group (0.48 ±\n0.72, P=0.01). We calculated incidence rate ratios\n(IRRs) for explanatory variables by the PR models\nto identify the related variables to the number of\nobserved gestational sacs after the ICSI cycles. If\nan IRR is greater than 1, the incidence rate (IR,\nsac no.) increases as x (explanatory variable) increases.\nIf an IRR is less than 1, the IR decreases\nas x increases.\nStudy population characteristics in patients with double embryo transfer (DET) and three embryo transfer (TET)\nResults are presented as mean ± SD, n (%). ART; Assisted reproductive technology, hCG; Human chorionic gonadotropin, and *; Significant at the 0.05 level.\nDistribution of the frequency of observed gestational\nsacs in the double embryo transfer (DET) and triple embryo\ntransfer (TET) groups\n*; Significant at the 0.05 level.\nThe quality of the transferred embryos and woman's\nage significantly impacted the number of observed\nsacs in the DET group ( Table 3 ). It means that\nthe IR of multiple sac in excellent grade embroyos\ngroup was 9 times and meanwhile in good grade\nembroyos group was 6 times respect to fair grade\nembryos as refrence group. The results showed that\nthe IRR for women’s age was 0.96 (95% CI: 0.93-\n0.99). In other words, one year increases in female\nage showed an IR of the observed gestational sac\nthat decreased 4%. The other explanatory variables\ndid not impact the number of sacs.\nIn the TET group, women's age, type of infertility,\nand fertilization rate significantly impacted the\nnumber of sacs ( Table 4 ). The IRR for maternal\nage was 0.94 (95% CI: 0.90-0.97) and IRR for the\nfertilization rate was 6.02 (95% CI: 2.96-12.22).\nThe one year increase in female age showed an IR\nof the observed gestational sac that decreased 6%.\nThe IR of the observed gestational sac increases 6\ntimes when one unit increases in the fertilization\nrate. The results demonstrated that the IR of the\nobserved gestational sac in patients with secondary\ninfertility decreased 42% with respect to those\nwith primary infertility ( Table 4 ).\nAll the possible affecting variables that included\nwomen’s age, BMI, infertility type, infertility cause\nand duration, acquired uterine anomaly, endometrial\nthickness on the day of the hCG injection, type\nof stimulation protocol, present cycle type, ovarian\nresponse type, total number of gonadotropin ampules,\ntotal number of retrieved oocytes, number of\nMII oocytes, fertilization rate, and quality of transferred\nembryos were entered in the multiple logestic\nregression (MLR) model. The result showed that\nwomen's age, duration of infertility, and number of\ntransferred embryos were the most important variables\nrelated to multiple pregnancies in this population\n( Table 5 ). Therefore, we repeated the MLR\nanalysis in the two separate populations (DET and\nTET). The results showed that none of the variables\nin the model had a significant association with multiple\npregnancies in the DET group. Multiple LR\nshowed that in the TET group, women's age and\nfertilization rate were significantly related variables\n( Table 6 ). Women with increased age had a 20% reduction\nin the odds for multiple pregnancies. When\nthe fertilization rate increased one unit, the odds for\na multiple pregnancy increased 19.7 times.\nResults of poisson regression for assessing the effect of explanatory variables on number of sac in patients with double embryos transfer\nSE; Standard error, IRR; Incidence rate ratio; CI; Confidence interval, and * ; Significant at the 0.05 level.\nResults of poisson regression for assessing the effect of explanatory variables on number of sac in patients with three embryos transfer\nSE; Standard Error; IRR; Incidence rate ratio; CI; Confidence interval, and * ; Significant at the 0.05 level.\nMultiple logistic regression analysis by backward manner for\npredicting multiple pregnancy in all study population\nOR; Odds ratio, CI; Confidence interval, and *; Significant at the 0.05 level.\nMultiple logistic regression analysis by backward manner for\npredicting multiple pregnancy in women with three embryos transfer in ICSI cycles\nOR; Odds ratio, CI; Confidence interval, ICSI; Intracytopasmic sperm injection.\nand *; Significant at the 0.05 level.\n\nIn present study, on the basis of the PR model, we\nfound that the quality of transferred embryos and\nwomen’s age had a significant effect on the number\nof observed sacs in patients who underwent ICSI\ncycles with DET. However, based on the multiple\nLR model, we found no significant predictive\nvariable for multiple pregnancy in these patients.\nPrevious studies reported that maternal age ( 5 ,  8 ),\nbody composition ( 6 ,  8 ), good ovarian response,\ncycle number, and the number of retrieved oocytes\n( 4 ) had a relationship to multiple pregnancies after\nDET. The current study results showed no impact\nby body composition, cycle number, number of retrieved\noocytes, and type of ovarian response. In\nagreement with previous studies ( 5 ,  8 ), we found\nthat the quality of transferred embryos significantly\ninfluenced the IR of sac numbers. Kaser et al.\n( 7 ) stated that there were six risk factor for twin\nlive births after cryopreserved cleavage stage DET\ncycles: age <35 years, resumption of mitosis, 7-8\nviable cells in the non-lead embryo, transfer of a\nlead embryo with ≥7 cells and a total of ≥14 viable\ncells.\nGroeneveld et al. ( 6 ), in a large nationwide Dutch\ncohort, demonstrated that tall stature and increased\nnumber of retrieved oocytes independently increased\nthe chance for dizygotic twins after IVF\nwith DET. In contrast to this study, we did not find\nany relationship between multiple pregnancies and\nBMI or number of retrieved oocytes in ICSI cycles\nwith DET.\nIn agreement with Niu et al. ( 4 ), we found that\nexcellent and good quality transferred embryos\nindependently increased the chances of multiple\nimplantation after ICSI with DET. They suggested\nthat it was advisable to perform SET when patients\nhad high risk factors for twin pregnancies\nthat included initial IVF-ET treatment, good or\nhigh ovarian response, more number of top-quality\nembryos, and development stage score of the\nsecond best embryo transferred. Fauque et al. ( 12 )\nconcluded that not only the implantation and pregnancy\nrates, but also the live birth rate depended\non embryo quality.\nOn the basis of our knowledge, the present study\nwas the first to evaluate risk factors for multiple\npregnancies in patients who underwent ICSI with\nTET. Our findings demonstrated that both regression\nmodels (poisson and logistic) had the same\noutputs. In both models, women’s age and fertilization\nrate showed significant relationships with\nmultiple pregnancies in these patients. It could be\ntaken as a measure of oocyte quality, which has\nbeen shown to be associated with increased implantation\npotential. A higher rate of fertilization\nwould likely result in a higher number of MII oocytes\nand, consequently, a higher chance of good\nquality embryos would be associated with higher\nchances for multiple pregnancies. In agreement\nwith previous studies we found a negative correlation\nbetween female age and multiple pregnancies\nin ICSI cycles with TET. Younger women had increased\nimplantation potential and multiple pregnancies\n( 10 ,  11 ).\nIn the present study, we have sought to evaluate\nthe risk factors of multiple pregnancies after transfer\nof two or three embryos in ICSI cycles. If the\nrisk factors of multiple pregnancies could be identified,\nwe can give proper counseling to couples at\nrisk for multiple pregnancies and suggest SET in\ntheir infertility treatment programs.\nCurrently an ideal target is considered in ART;\na healthy baby is the success, not a positive pregnancy.\nThe majority of twin or high order multiple\npregnancies (HOMP) are not successful or associated\nwith poor maternal and neonatal outcomes\n( 13 ). De Neubourg and Gerris ( 14 ) have reported\nthat the twin rate after IVF/ICSI dropped by at\nleast 50% simply by transferring only one goodquality\nembryo in the first and second fresh IVF/\nICSI cycles in young women, without a reduction\nin the overall pregnancy rate. They believed that\npreventing ‘the second half‘ of IVF/ICSI twins\nconstituted another, probably tougher challenge\nbecause the target group was a heterogeneous\nmix consisting of patients in very different clinical\nsituations. However, they suggested expanding the SET policy to women <38 years of age until\nthe third cycle and to cryoperservation cycles. In\nmany European countries the SET policy is the\nprimary prevention method used to prevent multiple\npregnancies. However, in Iran, the transfer of\njust one embryo, even with excellent morphology\nand quality, is taboo because it is feared that the\npregnancy rate will decline ( 15 ). Despite warnings\nfrom physicians about the risks associated\nwith multiple pregnancies, patients that have longterm\ninfertility problems perceive twin or multiple\npregnancies as blessings from God. In some cases\nthey are not satisfied with multifetal pregnancy\nreduction. However, physicians are ethically obligated\nto inform the patients regarding the risks of\nmultiple pregnancies.\nA limitation of the present study is its retrospective\nnature; therefore, we could not evaluate the\neffect of some possible factors reported in previous\nstudies.\n\nWe suggest the SET policy for prevention of\nmultiple pregnancies in women younger than 35\nyears of age who undergo ICSI cycles with high\nfertilization rates and excellent or good quality\nembryos. The remaining embryos should be cryopreserved.\nHowever, further prospective studies\nare necessary to evaluate whether SET in women\nwith these risk factors can significantly decrease\nthe multiple pregnancy rate and improve cycle\noutcomes.","source_license":"CC-BY-4.0","license_restricted":false}