Reproductive outcomes in recurrent pregnancy loss associated with a parental carrier of a structural chromosome rearrangement

Abstract

BACKGROUND: Reproductive outcome studies of couples with a history of recurrent pregnancy loss (RPL) associated with a maternal or paternal carrier of a structural chromosome rearrangement are limited. Correlation of carrier status and cytogenetics of miscarriage specimens is critical to estimate subsequent pregnancy outcome. METHODS: Couples found to have a structural chromosome rearrangement were followed prospectively in a tertiary academic centre. Descriptive analysis and subsequent pregnancy outcomes were tabulated and compared to historic controls. RESULTS: In 1893 RPL couples, 51 carriers of a structural chromosome rearrangement were identified (2.7%). Overall, this cohort had a total of 273 documented pregnancies. Prior to evaluation, the mean maternal age at the time of delivery or miscarriage was 29.8 years and the live birth rate was 15%. Following evaluation and treatment of concomitant factors, there were 58 monitored pregnancies, with a live birth rate of 71%. Amniocentesis was performed on 22% of the ongoing pregnancies; all were diploid or balanced structural chromosome rearrangements. Thirty-six per cent of the miscarriages were found to have an unbalanced structural chromosome rearrangement. CONCLUSIONS: Following evaluation and management of RPL, the live birth rate for carriers of a structural chromosome rearrangement is highly encouraging at 71%, without the addition of assisted reproductive technology.

Introduction

Recurrent pregnancy loss (RPL), defined as two or more miscarriages of under 20 weeks of gestation, affects up to 5% of couples trying to establish a family (Roman, 1984). A parental carrier of a structural chromosome rearrangement, most commonly a reciprocal or Robertsonian translocation, is associated with a history of RPL in approximately 3.5% of couples (Stephenson, 1996). Despite this known association, the likelihood of having a miscarriage due to an unbalanced structural chromosome rearrangement is not known, primarily because miscarriages are not routinely sent for cytogenetic analyses. If such data existed, management and counselling of couples with a history of RPL associated with a maternal or paternal carrier of a balanced structural chromosome rearrangement could be evidence based, as opposed to hypothesis based, which, unfortunately, is the current practice.

The objectives of this prospective longitudinal study were (i) to report the distribution of structural chromosome rearrangements in carriers with a history of RPL; (ii) to describe subsequent monitored pregnancy outcomes following evaluation and management of concomitant factors associated with their history of RPL; and (iii) to compare cytogenetic results of miscarriages in this cohort to two sets of historical controls, specifically, a recurrent miscarriage cohort and a general reproductive population cohort (Jacobs and Hassold, 1987; Stephenson et al., 2002).

Materials and methods

University of British Columbia and British Columbia’s Women’s Hospital & Health Centre (BC Women’s) ethics approval was obtained. Inclusion criteria included couples with a history of at least two documented miscarriages before 20 weeks of gestation, and, one of the partners was a carrier of a structural chromosome rearrangement. All couples were evaluated by the author (MDS) in the Recurrent Pregnancy Loss Program at BC Women’s, according to the diagnostic screening protocol previously published (Stephenson, 1996). With a history of a fetal demise between 10 and 20 weeks of gestation, an inherited thrombophilia screen was performed. All patients were advised during the consultation to avoid conception while evaluation was underway. Once cytogenetic analysis of both partners was completed, subsequent pregnancies were considered to be post-evaluation. Monitoring of pregnancy consisted of a serum βHCG within 1 week of missed menses and repeated 1 week later. Endovaginal ultrasound was offered at approximately 6 and 10 weeks of gestation and more frequently if bleeding or cramping occurred. A nurse with expertise in RPL was available by phone during the week and an RPL physician by pager on a 24-h basis. Pregnancy outcomes were obtained from hospital records and/or a questionnaire completed by the physician who attended the delivery. The couples’ data were entered into an RPL database (ACCESS 2000, Microsoft, Redmond, WA, USA) created by the author (MDS) in 1992. The data set was transferred to EXCEL 2000 for tabulation. Mean values with standard deviations (SD) were reported for demographic results. Proportions were compared using two-sample tests of proportions and Chi-square where applicable. Calculations were done using the Stata/SE 9.0 Statistical data analysis package (© 1984–2005, Statacorp, College Station, Texas, USA). Live birth rates, prior and subsequent to evaluation and management of RPL, were compared using a generalized estimating equations approach with a logit link function (Zeger and Liang, 1986). This model accounts for the correlation among multiple pregnancies within the same woman, which ranged from 2 to 10. P < 0.001 was accepted as statistically significant.

Results

A total of 51 (2.7%) couples with a history of RPL associated with a structural chromosome rearrangement were identified in the RPL database between June 1992 and April 2005; 1893 couples had been entered. There were 273 documented pregnancies in this cohort; therefore, the mean number of pregnancies per couple was 5.4 (SD 1.8 pregnancies). There were 72 (26%) live births, of which 67 were term and 5 were pre-term. There were 192 (70%) miscarriages, defined as pregnancy demise at or under 20 weeks of gestation; 148 were miscarriages of less than 10 weeks and 44 were fetal demises of 10–20 weeks of gestation. Other outcomes included eight elective terminations, including a diploid (46,XY) fetus with major congenital anomalies, and one ectopic pregnancy. The distribution of the carrier status, past obstetrical histories and subsequent pregnancy outcomes have been summarized in Tables I–III.

The 51 couples had a history of 215 pregnancies prior to evaluation and management of RPL, with a mean maternal age at delivery or miscarriage of 29.8 years (SD 5.0 years). These pregnancies resulted in 31 live births, 175 miscarriages, 8 elective terminations and 1 ectopic pregnancy. Therefore, the live birth rate was 15% (31/206) prior to evaluation and management.

Following evaluation and management, there were 58 monitored pregnancies, with a mean maternal age at delivery or miscarriage of 34 years (SD 4.3 years). These pregnancies resulted in 41 (71%) live births and 17 (29%) miscarriages, which was a significant improvement.

Amniocentesis for cytogenetic analysis was performed on 16 of the ongoing pregnancies—2 prior and 14 subsequent to the RPL evaluation. The mean maternal age was 34.5 years (SD 6.4 years) and 34.5 years (SD 4.8 years), respectively. All 16 of the results were euploid; seven were either 46,XX or 46,XY, eight were balanced and one contained the same Y heterochromatin insertion as the maternal carrier.

Overall, 46 (24%) of the miscarriage specimens were sent for cytogenetic analysis; unfortunately 10/46 (22%) of the specimens failed to grow in culture. Therefore, cytogenetic results were obtained on 36 miscarriage specimens—26 prior and 10 subsequent to the RPL evaluation, with a mean maternal age at time of miscarriage of 33.5 years (SD 4.4 years) and 36.7 years (SD 3.5 years), respectively. The distribution of the cytogenetic results is shown in Table IV, along with two cohorts of historic controls.

Prior to evaluation and management of RPL, 11 (42%) of the 26 miscarriage specimens were euploid; six were diploid (46, XX or 46,XY) and five were balanced structural chromosome rearrangements involving the same chromosomes as the parental carrier. Eight (31%) were unbalanced structural chromosome rearrangements involving the same chromosomes as the parental carrier; one had an additional de-novo chromosome rearrangement. Five (19%) were trisomic; three had a concomitant balanced structural chromosome rearrangement and one had a concomitant unbalanced structural chromosome rearrangement, all involving the same chromosomes as the parental carrier. Two (8%) of the miscarriage specimens were polyploid; one had the identical Y heterochromatin insertion as the parental carrier.

Following the RPL evaluation, one of the 10 miscarriage specimens was diploid. Five were unbalanced structural chromosome rearrangements involving the same chromosomes as the parental carrier. Four were trisomic; two had a concomitant balanced structural chromosome rearrangement and one had a concomitant unbalanced structural chromosome rearrangement, all involving the same chromosomes as the parental carrier.

Parental carriers of a reciprocal translocation

There were 28 parental carriers of a balanced reciprocal translocation; 15 were females and 13 were males. In the female carriers, chromosome 6 was involved in the translocation most frequently (n = 5 couples), followed by chromosome 4 (n = 4 couples), chromosomes 2, 7 and 12 (n = 3 couples, each), chromosomes 3, 8, 10 and 11 (n = 2 couples, each) and chromosomes 1, 9, 18 and 19 (n = 1 couple, each). In the male carriers, chromosome 5 was involved in the translocation most frequently (n = 4 couples), followed by chromosomes 6, 8, 13 and 17 (n = 3 couples, each), chromosomes 1 and 7 (n = 2 couples, each) and chromosomes 9, 10, 11, 12, 14 and 18 (n = 1 couple, each).

These 28 couples, in whom one was a carrier of a reciprocal translocation, had 118 pregnancies prior to the RPL evaluation. Of these, 16 (14%) were live births, 98 (83%) were miscarriages and 4 were elective terminations.

Evaluation revealed a concomitant diagnosis of antiphospholipid syndrome (APS), according to the International APS criteria (Wilson et al., 1999), in 8 of 28 (29%) couples (patient numbers 2, 24, 41, 13, 17, 35, 40 and 47); acetylsalicylic acid and heparin were prescribed. A concomitant diagnosis of a luteal phase deficiency, defined as two mid-luteal phase endometrial biopsies with maturation delay of ≥3 days based on the first day of the next menstrual period (Noyes et al., 1975) or glandular-stromal dyssynchrony, was made in 2 of 28 (7%) couples (patient numbers 30 and 44); vaginal progesterone or clomiphene citrate was prescribed. A concomitant diagnosis of heterozygous Factor V Leiden was made in one of the couples (patient number 44); heparin was prescribed at eight weeks of gestation. One of the female patients (patient number 27) had Crohn’s disease requiring prednisone and gestational diabetes requiring insulin in pregnancy. Empiric acetylsalicylic acid and vaginal progesterone were prescribed preconceptually in five of the patients (patient numbers 2, 31, 43, 53 and 44).

There were 35 monitored pregnancies subsequent to the RPL evaluation, including 22 (63%) live births. There was no difference in the live birth rates between the female and male carriers of a reciprocal translocation. Seven of the pregnancies underwent amniocentesis; three were diploid and four were balanced reciprocal translocations. There were 13 miscarriages <20 weeks of gestation, of which seven (54%) could be explained; two were trisomic and five were unbalanced reciprocal translocations. The other six miscarriages were unexplained; one was diploid (46,XX) and five did not have cytogenetic results.

Parental carriers of a Robertsonian translocation

There were a total of 12 parental carriers of a balanced Robertsonian translocation; nine were females and three were males. Eleven of the Robertsonian translocations were 13;14 and one was 13;22.

Overall, there were 70 pregnancies in the Robertsonian translocation carriers. There were 19 (27%) live births; no difference in the live birth rate was found between the female and male carriers. Amniocentesis was performed in four pregnancies; one was diploid and three were balanced Robertsonian translocations involving the same chromosomes as the parental carrier. There were 46 (66%) miscarriages; 37 in <10 weeks and 9 between 10 and 20 weeks of gestation. There were four elective terminations and one ectopic pregnancy.

A concomitant diagnosis of APS was made in 2 of the 12 (17%) couples (patient numbers 36 and 15); acetylsalicylic acid and heparin were prescribed. A concomitant diagnosis of a luteal phase deficiency was made in 2 of 12 (16%) couples (patient numbers 18 and 51); vaginal progesterone was prescribed.

There were 13 monitored pregnancies subsequent to the RPL evaluation, including 9 (69%) live births. Two of the four miscarriages could be explained; both were trisomic. The other two miscarriages were unexplained; they did not have cytogenetic results.

Parental carriers of inversions

There were a total of seven parental carriers of a balanced inversion; five were females and two were males. Two of the inversions were of chromosome 2 and others were of chromosomes 3, 6, 7, 10 and 17.

Overall, there were 35 pregnancies in the inversion carriers. There were 11 (31%) live births; no difference in the live birth rate was found between the female and male carriers. Amniocentesis was performed in two pregnancies; one was diploid and the other had a balanced inversion involving the same chromosome as the parental carrier. There were 24 miscarriages; 18 in <10 weeks and 6 between 10 and 20 weeks of gestation.

A concomitant diagnosis of APS was made in 2 of the 7 couples (patient numbers 4 and 45); acetylsalicylic acid and heparin was prescribed. A concomitant diagnosis of a luteal phase deficiency was made in 1 of 7 couples (patient number 6); vaginal progesterone was prescribed.

There were seven monitored pregnancies subsequent to the RPL evaluation; all were live births.

Other structural rearrangements

There were four parental carriers of rare and/or complex structural rearrangements that require separate documentation. One of the maternal carriers (patient number 20) had three cell lines including a ring rearrangement of chromosome 13. Unfortunately, cytogenetic analysis was not performed on any of the prior three miscarriages. The couple decided not to attempt another pregnancy.

Another of the maternal carriers (patient number 46) had a Y heterochromatin insertion on the distal end of chromosome 15. This couple had an uncomplicated term delivery, followed by three explained miscarriages; two were due to an unbalanced translocation involving the same chromosome as the parental carrier, one was due to polyploidy. Following evaluation for RPL, the couple had one pregnancy that was a healthy term live birth with the Y heterochromatin insertion. This case has been reported previously (Rajcan-Separovic et al., 2001).

There was a maternal carrier (patient number 50) with a complex insertion involving two chromosomes. A concomitant diagnosis of a luteal phase deficiency was made; clomiphene citrate was prescribed. Subsequent pregnancy outcome consisted of a term live birth.

There was a paternal carrier (patient number 7) with a complex rearrangement consisting of three break points on a single chromosome. Following evaluation for RPL, this couple had a term live birth with the same complex rearrangement.

Comparison with historic controls

The cytogenetic results of the 36 miscarriage specimens from this cohort of couples with a history of RPL associated with a parental carrier of a structural chromosome rearrangement were compared to historic controls (see Table IV). The recurrent miscarriage cohort was defined as couples having at least three or more unexplained consecutive miscarriages of less than 20 weeks of gestation, excluding any aneuploid or polyploid miscarriages (Stephenson et al., 2002). The general reproductive population cohort was a tabulated data set of seven miscarriage surveys (Jacobs and Hassold, 1987). Compared to both cohorts of historic controls, our present cohort had a higher likelihood of having an unbalanced structural chromosome rearrangement in a miscarriage specimen (P < 0.001 for both comparisons).

Discussion

This reproductive outcomes study suggests that a thorough evaluation for concomitant factors associated with a history of RPL, followed by close monitoring in pregnancy, can result in approximately two-thirds of subsequent pregnancies being successful in carriers of a structural chromosome rearrangement. This and other recently published studies provide the foundation for an evidence-based approach to the management of RPL associated with a structural chromosome rearrangement.

In 2004, Sugiura-Ogasawara et al. (Sugiura-Ogasawara et al., 2004) published a prospective cohort, nested case–control study of 1284 Japanese couples with RPL, defined as two or more consecutive first-trimester miscarriages, in whom 100 (7.8%) carriers of a structural chromosome rearrangement were identified. The distribution of carrier status was statistically different (P = 0.002) between this Japanese cohort and the present study; there were fewer carriers of a Robertsonian translocation (11 versus 24%) and more carriers of an inversion, mostly involving chromosome 9 (31 versus 14%). Carriers of a reciprocal translocation had a live birth rate following a RPL evaluation, which was significantly lower than that found in the present study, 36 versus 63% (P = 0.005). Evaluation of RPL differed between the Japanese and the present study, including endocrine and immunological testing. Management differed in that Japanese patients were hospitalized for 1 month from 4 to 8 weeks of gestation. In the present study, acetylsalicylic acid and heparin were only used if the patient met strict criteria for APS.

Also in 2004, Goddijn et al. (Goddijn et al., 2004) published a retrospective cohort, nested case–control study of 1324 Dutch couples with a history of RPL, defined as two or more miscarriages <20 weeks of gestation, in whom 51 (2.5%) carriers of a structural chromosome rearrangement were identified. Of the 41 carriers with complete miscarriage data, there were 26 (63%) carriers of a reciprocal translocation (18 females and 8 males), 3 (7%) carriers of a Robertsonian translocation, 9 (22%) carriers of an inversion and 3 others. Following ascertainment of carrier status, 43 pregnancies in 25 carrier couples were documented, of which 30 (70%) were live births, identical to the present study. Amniocentesis had been performed in 26 of these pregnancies; 58% were diploid and 42% were balanced structural chromosome rearrangements. These amniocentesis results parallel the present study. This Goddijn et al. study questioned the utility of parental cytogenetic analyses as part of the diagnostic screening protocol of couples with RPL. Although it appears that carriers of a structural chromosome rearrangement, ascertained on the basis of RPL, have a low likelihood of a live birth with an unbalanced rearrangement, the numbers of cases reported are limited.

Based on a limited number of amniocenteses, none of the ongoing pregnancies were affected by an unbalanced translocation. The identification of a translocation carrier, ascertained on the basis of RPL, may be associated with a lower risk of having an ongoing pregnancy with an unbalanced translocation, compared to ascertainment on the basis of a personal or family history of an ongoing pregnancy with an unbalanced translocation.

Overall, approximately one-third of the miscarriages were found to have an unbalanced structural chromosome rearrangement in this study. This is certainly higher than seen in a general recurrent miscarriage cohort and in the general reproductive population, but does not account for the majority of the miscarriages since two-thirds of the miscarriages were euploid, polyploid or aneuploid. Analysis of miscarriage specimens gives us vital information in regard to whether the miscarriage is cytogenetically abnormal, either numeric or unbalanced, or normal, including diploid or balanced structural chromosome rearrangements. With the use of DNA technology such as comparative genomic hybridization, the rate of no results due to culture failure or erroneous results due to maternal contamination could be markedly reduced (Lomax et al., 2000). Therefore, we must not assume that miscarriages in couples with a history of RPL, associated with a maternal or paternal carrier of a structural chromosome rearrangement, are unbalanced, since the majority of them are not.

PGD has been offered to carriers of a structural chromosome rearrangement. A multi-centre retrospective analysis of 469 IVF/PGD cycles, in carriers of a translocation, reported a pregnancy rate per cycle started of 34.5% and a take-home baby rate of 23.0% per embryo transfer (Verlinsky et al., 2004). There have been no prospective studies evaluating IVF/PGD in couples with a history of RPL associated with a maternal or paternal carrier of a structural chromosome rearrangement. To date, there has only been one prospective study addressing the effectiveness of IVF/PGD in patients with advanced maternal age (Staessen et al., 2004), and another in patients with idiopathic recurrent miscarriage (Platteau et al., 2005). The results from both of these studies showed little advantage of IVF/PGD for fertile couples with a history of RPL. A prospective randomized trial comparing IVF/PGD to non-IVF/PGD intervention, as described in the present study, for couples with a history of RPL associated with a maternal or paternal carrier of a structural chromosome rearrangement, would address the utility of assisted reproductive technology in this cohort.

The present study suggests that evaluation and management of other factors associated with RPL improves the subsequent live birth rate in carriers of a structural chromosome rearrangement, ascertained on the basis of RPL. In addition, the close monitoring and supportive care likely contributed to the improved live birth rate, as reported by Stray-Pedersen et al. (Stray-Pedersen and Stray-Pedersen, 1984). The reporting of prior and subsequent live birth rates is biased towards accentuating the effect of evaluation and management of RPL for two reasons. First, referrals were prompted by the last pregnancy, or string of pregnancies, being a miscarriage(s), as documented in Tables I, II and III. Consequently, the estimated live birth rate in the prior obstetrical history is biased downwards. Second, it is likely that following evaluation and management, if the first monitored pregnancy results in a live birth, the patient will not attempt another pregnancy for some time. Consequently, the live birth rate is likely biased upwards to some degree. While it is difficult to assess the degree of bias resulting from these two sources, we would argue that it is not likely to account for the very large difference in live birth rates prior to and subsequent to evaluation and management of their history of RPL.

In conclusion, a thorough evaluation, evidence-based treatment, close monitoring and supportive care, in couples with a history of RPL in whom one of the partners is a carrier of a structural chromosome rearrangement, is associated with a marked improvement in subsequent live birth rate.

Acknowledgements

The authors thank T.Pacheco, CCHRA(C), BC Women’s, for data entry and processing, and, P.Schultz, RN, MHA, University of Chicago, for preparation of the Tables. We appreciate the guidance of S.Langlois, MD, University of British Columbia, and S. Schwartz, PhD, University of Chicago, in regard to cytogenetic data, and T.Karrison, PhD, University of Chicago, for assistance with statistical analyses.

References

Author notes

1Department of Obstetrics and Gynecology, University of Chicago, Chicago, IL, USA, 2Department of Obstetrics and Gynecology, University of British Columbia, Vancouver, Canada and 3Recurrent Pregnancy Loss Program, British Columbia’s Women’s Hospital & Health Centre, Vancouver, Canada