Several studies were identified that have evaluated the impact of medical and family history on the prognosis in RPL couples. The chance of a live birth, time to live birth and the risk of a pregnancy loss in the next pregnancy are considered relevant outcomes for prognosis in RPL.

In the absence of any interventions proven to ameliorate the chances of a live birth in couples with unexplained RPL, investigators have attempted to develop prognostic tools, based on the identified factors affecting prognosis. Although not an intervention as such, informing couples confronted with RPL about their individual prognosis in a next pregnancy and in the long term is an essential part of the management of couples and allows the couples to decide for or against further pregnancy attempts (Lund et al., 2012).

ESHRE Guideline group on RPL (2022)

Summary of literature

Factors affecting prognosis

Reproductive history

The impact of the number of prior pregnancy losses for the chance of live birth has been investigated in a number of cohort studies. The authors consistently find that the number of prior pregnancy losses is an important prognostic factor for chance of live birth in both the first pregnancy after referral and in the long term (Parazzini et al., 1988, Knudsen et al., 1991, Quenby and Farquharson, 1993, Brigham et al., 1999, Bhattacharya et al., 2010, Lund et al., 2012, Kolte et al., 2014, Greenberg et al., 2015, Kling et al., 2016).

In a nested cohort study of 251 women with two or more miscarriages from the ALIFE trial, it was demonstrated that the number of prior miscarriages was a determinant both for time to live birth and cumulative incidence of live birth. Follow-up was limited to 24 months after enrolment in the trial (Kaandorp et al., 2014).

One retrospective cohort study of 587 women with unexplained RPL (≥3 PLs) following spontaneous conception showed that among the 499 women who subsequently became pregnant, the relative risk of live birth in the first pregnancy after referral was the same for miscarriages and non-visualized pregnancy losses (Kolte et al., 2014). This suggests that the type of pregnancy loss is less important for chance of live birth, but needs corroboration in independent cohorts.

For secondary unexplained RPL, a recent cohort study suggested that only consecutive pregnancy losses after the birth influenced the subsequent prognosis, while the number of losses prior to the birth did not affect the prognosis in the next pregnancy (Egerup et al., 2016).

In a multicenter study on 777 patients, subsequent pregnancy success rate was found to be significantly associated with pregnancy loss history (i.e. time (in years) between first and last miscarriage prior to assessment) and subfertility index (i.e. the product of the number of PLs and the pregnancy loss history), suggesting an effect of the time needed to conceive (Cauchi et al., 1995). In this study, the maternal age was only borderline significant associated with the subsequent pregnancy success rate, but only if treated as a dichotomous variable (< 30 years or ≥ 30 years). The number of spontaneous pregnancy losses was significantly associated with the subsequent pregnancy success rate.

Sex of firstborn

In secondary RPL, the sex of the firstborn may be important for prognosis. In a study of 358 Danish women with unexplained secondary RPL compared to the Danish general population, sex ratios were shown to be significantly skewed in the RPL population: sex ratio (boy/girl) of the children born prior to secondary RPL was 1.49 compared to 1.05 in the general population. The sex-ratio of live born children in the first pregnancy after referral was 0.76, and thus the sex ratio significantly changed from firstborn (more boys) to the first pregnancy after referral (more girls) in couples with secondary RPL (Nielsen et al., 2010). In an Irish cohort study of 85 women with secondary RPL, sex-ratios prior to secondary RPL was 1.66, but there were no significant differences in chances of live birth according to sex of the firstborn (Ooi et al., 2011). In a study of 170 women with secondary RPL, another observational study reported a skewed sex ratio for first stillborn children, but not live born children (Li et al., 2014).

Family history

The results from a recent register-based study of 2138 women with RPL (Kolte et al. 2021) indicate that pregnancy losses among first-degree family members is not an important risk factor for outcome of the first pregnancy after referral among women with RPL (Kolte et al. 2021).

A number of studies have reported that sporadic or recurrent (≥2) pregnancy loss is more common among RPL patients’ first-degree relatives than controls, approximately a doubled incidence or per pregnancy loss rate (Alexander et al., 1988, Johnson et al., 1988, Christiansen et al., 1990, Ho et al., 1991, Zhang et al., 2010, Kolte et al., 2011). While this may suggest a familial or hereditary component to RPL, none of the abovementioned studies investigated whether affected family members are important for the prognosis of an individual patient. Furthermore, it should be remembered that studies evaluating risk of pregnancy loss among patients’ relatives may be subject to information bias, especially if information on relatives’ pregnancy losses is derived from the patients. In families where one person suffers from RPL, there may be more openness about reproductive history than in other families and women are referred earlier than women without family members with pregnancy losses (Kolte et al. 2021).

Prognostic tools
In a descriptive cohort study, prognosis was evaluated in 987 women with primary or secondary RPL referred to a tertiary center in Denmark (Lund et al., 2012). Five years after the first consultation, 66.7% (95% CI 63.7-69.7) had achieved a live birth, increasing to 71.1% (95% CI 68.0-74.2) after 15 years. There was a significantly decreased chance of at least one subsequent live birth with increasing maternal age; of women aged 40 years or older, 41.7% (95% CI 29.8-56.1) achieved a live birth within 5 years compared to 81.3% (95% CI 69.2-90.7) of women aged 20–24 years. There was also a significant decrease in chance of a live birth by increasing number of miscarriages before first consultation ranging from 71.9% (95% CI 67.5-76.1) in women with 3 miscarriages to 50.2% (95% CI 40.5-60.8) in women with 6 or more previous miscarriages. There was no evidence of an interaction between maternal age and the number of previous miscarriages. The Lund model was not designed for individual risk assessment, given the descriptive scope of the study. Furthermore, the study does not discriminate between unexplained and explained RPL

Another longitudinal study prospectively collected data of 716 RPL patients (325 idiopathic) attending a referral clinic in Liverpool over a 10-year period (Brigham et al., 1999). Of the patients achieving a further pregnancy, 167/222 (75%) had a successful outcome with survival beyond 24 weeks. There was no statistically significant difference in outcome between primary (77%) and secondary losers (74%). From a survival curve, it was shown that the most perilous time for women with idiopathic RPL was between 6 and 8 weeks’ gestation. By 8 weeks’ gestation, if a fetal heartbeat had been identified, the chances of a successful outcome in a subsequent pregnancy were 98%, climbing to 99.4% at 10 weeks’ gestation. Previous miscarriage history and age of the patient significantly affected the chances of a successful outcome, age being slightly more significant than previous number of miscarriages.

Fosse et al.  aimed to explore whether predicting the chance of a subsequent ongoing pregnancy in couples with unexplained RPL could be improved by taking, besides maternal age and the number of previous pregnancy losses, additional candidate predictors into account. Indeed, they showed that predicting the chance of a subsequent ongoing pregnancy beyond 24 weeks of gestation in couples with RPL becomes more accurate when, besides the conventional predictors maternal age and number of previous pregnancy losses, more variables are incorporated into the model. The additional predicting variables include both male and female characteristics, advocating a couple-focused rather than a female-focused approach in RPL. However, the predictive ability of the current model remains limited, and we emphasize that more research is needed to develop a model that can be used in clinical practice.

Although the Brigham model and the Lund model were both reviewed with high methodological quality and both studies have consistent results, these models did not follow the nowadays recommended TRIPOD guideline in the development and reporting of a prediction model. Both models were never internally nor externally validated, which leaves their predictive performance unknown. Youssef et al externally validated the Brigham model in a Dutch population of 739 couples and showed that there is poor predictive performance of this model on a new population (Youssef et al 2022). The model overestimates, has too extreme predictions and has a poor discriminative ability .

Recommendations

Justifications

The number of pregnancy losses before referral for RPL is of prognostic importance for future chance of a live birth. Although the studies are of high quality and consistent, evidence on the prognostic potential of reproductive history can only be obtained by observational studies, which is reflected in the low evidence level. The guideline development group concludes that a thorough reproductive history should be taken in couples presenting with RPL and stresses that number of preceding pregnancy losses and female age provide the best available prognostic information.

The studies of Lund and Brigham show that RPL couples have a good prognosis for a next live birth, especially if female age and the number of previous miscarriages are low. However, one should be aware that the model might overestimate (Youssef et al. 2022). There is a need for revising the prediction model in order to estimate the chance of a successful pregnancy in couples with unexplained RPL more accurately. Until new studies have been performed, the information from the studies (summarized in the graphs and tables below) can be used by clinicians to estimate the chance of a live birth in a next pregnancy in couples with RPL. The new prognostic tool Kolte & Westergaard could also be used to predict a live birth in the next pregnancy (Kolte et al. 2021). However, it should be noticed that this tool had not been externally validated.

Unproven therapeutic interventions for unexplained RPL should only be applied in research setting with relying on reasonable pathophysiological hypotheses.

Additional information:

Effect of female age and number of previous pregnancy losses on live birth in RPL

Table Predicted percentage success rate of subsequent pregnancy according to age and previous miscarriage history (Brigham et al., 1999) (reproduced with permission) (international guideline group on RPL, 2022)

The results of the summary of literature were adapted from the international ESHRE guideline Recurrent Pregnancy Loss (2022). More details about the methodology of the ESHRE guideline on RPL (2022) can be found in “Annex 5: Methodology”of the ESHRE guideline and in the Manual for ESHRE guideline development, version 2013, which can be consulted at the ESHRE website. The evidence tables and literature studies can be found at the ESHRE website in Supplementary data files.