Genetic contribution to postpartum haemorrhage in Swedish population: cohort study of 466 686 births
BMJ 2014; 349 doi: https://doi.org/10.1136/bmj.g4984 (Published 13 August 2014) Cite this as: BMJ 2014;349:g4984
All rapid responses
Rapid responses are electronic comments to the editor. They enable our users to debate issues raised in articles published on bmj.com. A rapid response is first posted online. If you need the URL (web address) of an individual response, simply click on the response headline and copy the URL from the browser window. A proportion of responses will, after editing, be published online and in the print journal as letters, which are indexed in PubMed. Rapid responses are not indexed in PubMed and they are not journal articles. The BMJ reserves the right to remove responses which are being wilfully misrepresented as published articles or when it is brought to our attention that a response spreads misinformation.
From March 2022, the word limit for rapid responses will be 600 words not including references and author details. We will no longer post responses that exceed this limit.
The word limit for letters selected from posted responses remains 300 words.
Oberg et al1 is read with great interest. In this Swedish study, 18% of variation in postpartum haemorrhage (PPH) liability is put down to maternal genetics and a further 11% down to fetal genetics. It is suggested that paternal genetics may have little role in a mother’s liability to PPH.
As 50% of fetal DNA should be paternal, then it must be expected that at least 5.5% of the overall liability to PPH would be influenced by paternal genes. Oberg et al suggest that this may derive from the correlations seen in brother pairs and the reduced incidence in women who changed partner as opposed to those with the same partner.
The authors found that one of the most common types of PPH, retained placenta, clustered mostly within couples (Tetrachoric Correlation Coefficient 0.45) and mothers with new partners (0.37) but also clustered, albeit minimally, in fathers with new partners (0.04). However, when calculated for PPH of all types, there was no correlation found between PPH and fathers who changed partners in this study. This latter cluster level may suggest paternal genes may be involved in the occurrence of retained placenta but not other types of PPH.
Recent research has shown that genomic imprinting leads to a skew in maternal and paternal gene expression in the early trophoblast with the majority of genes being paternally expressed2. Hydatiform moles, a type of gestational trophoblastic disease, can be the result of dispermic fertilisation of an empty ovum or fertilisation of an empty ovum by a sperm which then duplicates. In other cases, the mole may be triploid with one maternal chromosome set and two paternal sets3. The hallmark feature of these conceptions is the excessive growth of placental tissue, which, in rare cases, may become invasive and, even, malignant. The depth of trophoblastic invasion has been suggested as a risk factor for PPH as mentioned by Oberg et al. Paternal genes are certainly involved in placental formation but their contribution to PPH liability has not been proven in this present study.
The authors also suggest that PPH may be so common because the genes involved may confer a selective advantage to offspring as long as, at the population level, the death of the mother will not confer a selective disadvantage. Another theory, although an over-simplification, could be that PPH is a recessive trait. If the genes involved are beneficial to the carrier’s fecundity when heterozygous, then they would be selected for. However, if present in a homozygous state, these genes may lead to PPH. For instance, a gene in heterozygous state may lead to deeper invasion of the trophoblast and so a stronger placental attachment, hence, greater fetal health. However, in homozygous state could lead to a placental attachment that is too strong and prone to retention after birth.
An interesting study would be to look at the number of offspring born to couples who have a personal or family history of PPH compared to those couples who have no personal or family history of PPH. If a personal or family history of PPM is associated with greater fecundity (i.e. greater number of offspring or fewer spontaneous miscarriages) then this could suggest a selective advantage for PPM predisposing genes. However, confounding factors such as social attitudes to number of children and contraception may make this study impractical.
References
1. Oberg AS, Hernandéz-Diaź S, Frisell T, Greene MF, Almqvist C, Bateman BT. Genetic contribution to postpartum haemorrhage in Swedish population: cohort study of 466 686 births. BMJ 2014;349:g4984. doi: 10.1136/bmj.g4984.
2. Wang X1, Miller DC, Harman R, Antczak DF, Clark AG. Paternally expressed genes predominate in the placenta. Proc Natl Acad Sci U S A. 2013 Jun 25;110(26):10705-10. doi: 10.1073/pnas.1308998110. Epub 2013 Jun 10.
3. Sunde L1, Niemann I, Hansen ES, Hindkjaer J, Degn B, Jensen UB, Bolund L. Mosaics and moles. Eur J Hum Genet. 2011 Oct;19(10):1026-31. doi: 10.1038/ejhg.2011.93. Epub 2011 Jun 8.
Competing interests: No competing interests
The BMJ has published an impressive research article1 on a relatively common condition of postpartum haemorrhage (PPH) which appears to be an excellent example of evidence based medicine (EBM) but the majority of clinicians may struggle to understand it. This article certainly contains an impressive cohort of 466,686 births from the Swedish population registry and extensive statistical modelling and analysis. However, it is difficult to understand if there was a 'control' group in this study. That is whether this study included similar number of births without PPH and whether it has been demonstrated that these non-PPH cases did not have the 'familial clustering of PPH'. Of course a belief about need for such a control group may be completely mistaken.
The other statistic difficult to find was the degree of increased relative risk of PPH because of different types of family history of PPH. One mention found was that the women with a sister with PPH had 70% increased odds of being affected.1 Is it likely that this is closer to the top of the range of increase in PPH risk due to family history? This would increase the risk of PPH from 4.6% (general prevalence) to 7.8%. This contrasts with multiple fold increase in risk of PPH associated with Obstetric and medical risk factors. What additional measures should be instituted for a woman with PPH risk of 7.8% that could be denied to women with PPH risk of 4.6% in developing or developed countries? In other words how does taking account of this family history of PPH would save lives given the constraints on human and financial resources?
The reasons for maternal mortality in developed countries are massive PPH compounded sometimes by errors in management like delay in seeking expert help or definitive measures. The morbidity and mortality of PPH in developing countries is due to lack of health care facilities and/or reluctance of patients to seek available help even in the presence of strong clinical risk factors. Will 'recording of family history of PPH' have any impact in either of these situations? It may be premature and mistaken to claim that taking family history of PPH from all pregnant women could save lives.2 It has been suggested that detailed family history should be taken to stratify women at special risk for PPH.2 But there seems no evidence that reliable history will be available and indeed conclusions of the Swedish study1 mention that such risk stratification is not possible at present and awaits future research. It is difficult to understand how the way has been paved for future "genetic screening" for one of the world's leading maternal killers (PPH)2, when genetic screening for even BRCA mutations is restricted in a rich country like UK. The case for addition of family history of PPH to the current maternity care has not been made as it may not be easy/reliable or economical or productive and could be unnecessary distracting additional task.
Declaration of interest: The author has no conflict of interest or funding to declare.
Author: Mr. Shashikant L Sholapurkar, MD, DNB, MRCOG
Obstetrician and Gynaecologist, Royal United Hospital, Bath, UK
References:
1. Oberg AS, Hernandéz-Diaź S, Frisell T, Greene MF, Almqvist C, Bateman BT. Genetic contribution to postpartum haemorrhage in Swedish population: cohort study of 466 686 births. BMJ 2014;349:g4984. doi: 10.1136/bmj.g4984.
2. Chong YS, Chi C, Holbrook JD. Genetic contribution to postpartum haemorrhage. BMJ 2014;349:g5285. doi: 10.1136/bmj.g5285.
Competing interests: No competing interests
Audit of three, independent, UK units delivering over 4000 women per annum in the past twelve months shows increases in postpartum haemorrhage rates (>1000 mls) of over 500% in the past decade (2003-2013). In 2003, such a unit might experience a massive PPH (>2000 mls) twice per year though is now having one of these events once a fortnight. Not only have PPH rates increased dramatically but they appear to be less predictable and unrelated to traditional risk factors.
Injured uteri, inexperienced clinical staff, and, increasing rates of induction of labour may all contribute to this increasing incidence of serious morbidity and mortality (1). Injuries to uterine nerves result from prolonged straining during defaecation (endemic in Western countries) and excessive traction to the cervix and uterosacral ligaments during surgical evacuation of the uterus in the second trimester (endemic in some developing countries). Whether “injured” uteri respond appropriately to ergometrine, oxytocin and prostaglandins is not known.
It is unlikely that these dramatic increases in the incidence of PPH in the UK result from genetic factors. Clustering of cases in Swedish families may result from similar diets and bowel habits though Oberg et al. do not report similar changes in incidence in Sweden (2) ? Whatever the reasons, obstetricians and haematologists need to be aware of any such changes in their clinical services and respond appropriately.
(1) Quinn MJ.
Postpartum hemorrhage and “injured” uteri.
AJOG 2014 Jun; 210(6):588.
(2) Oberg AS, Hernandéz-Diaź S, Frisell T, Greene MF, Almqvist C, Bateman BT,
Genetic contribution to postpartum haemorrhage in Swedish population: cohort study of 466 686 births.
BMJ 2014;349:g4984
Competing interests: No competing interests
Re: Will recording family history save lives or make a difference? Authors respond.
We appreciate Dr. Sholapurkar's interest in our study and welcome the opportunity to clarify our methods and what we see as the implications of our findings.
As our study concerns the extent of PPH clustering in differentially related births (=within family cluster correlation) there is no "control group" in the traditional sense. Irrespective of outcome, all births (i.e. affected and unaffected) that belong to a certain type of relation will contribute to the assessment of correlation in that family cluster.
The pair-wise odds ratios (POR) that we report quantify the within-cluster correlation by contrasting the odds of PPH in persons who had a related birth affected by PPH to those who did not. For each POR, the denominator thus constitutes the reference (no family history). Our main objective for performing this particular analysis was to explore the degree of familial clustering while taking potentially shared risk factors into account, but the PORs could also be taken as an approximation of the change in PPH risk by having a related birth affected by PPH. For the sister cluster e.g., this will correspond to the mean excess risk of women who had a sister affected by PPH (compared to those whose sister’s delivery was not affected by PPH).
While our findings show that PPH clusters in families, we agree with Dr. Sholapurkar that it is premature to conclude that assessing family history will add value in stratifying patients’ individual risk for PPH. Future work is needed to determine if incorporating such a history, in addition to traditional risk factors, will be clinically useful. We also agree that speculation about genetic testing for PPH risk is not yet warranted, but do believe our work motivates research aimed at identifying the mechanisms for genetic influence, as a means of better understanding the pathways that mediate PPH risk.
Anna Sara Oberg, MD MPH PhD
Research Fellow, Harvard School of Public Health and Karolinska Institutet
Brian T Bateman, MD MSc
Anesthesiologist and Assistant Professor, Massachusetts General Hospital and Brigham and Women's Hospital, Harvard Medical School
Competing interests: No competing interests