Fertility preserving management for postpartum haemorrhage secondary to subinvolution of the placental implantation site

Abstract

A 40-year-old woman presents with recurrent secondary postpartum haemorrhage (PPH) following her third normal vaginal delivery. Histology from subsequent evacuation of the uterus confirmed that she had subinvolution of the placental implantation site. Hysterectomy is the most common method of managing this condition and recurrent PPH, most often due to significant vaginal bleeding. We present a case of subinvolution of the placental implantation site with recurrent PPH managed with medical treatment alone, to offer a fertility-sparing treatment option.

Background

Postpartum haemorrhage (PPH) is estimated to affect 1%–5% of deliveries. It is an obstetric emergency and is associated with significant maternal morbidity and mortality. Primary PPH is bleeding of more than 500 mL, which occurs within the first 24 hours post delivery. Secondary PPH is abnormal or excessive vaginal bleeding, which occurs between 24 hours to 6 weeks post delivery. The estimated incidence of secondary PPH alone is 0.2%–3%.¹

The most common clinical diagnosis of secondary PPH are retained products of conception (RPOC, 60%) and endometritis (15%).¹ Studies looking at histological cause of secondary PPH shows that incidence of RPOC is 36%, endometritis is 10% and subinvolution of the placental implantation site is 13%.^1–3 Subinvolution of placental implantation site is a diagnosis based on histological evidence. However, it should be considered in any woman, who present with recurrent PPH in the absence of infection or RPOC. It is caused by abnormal persistence of dilated uteroplacental arteries that fail to obliterate after delivery of the placenta. On ultrasound, there is increased vascularity within the myometrium, and there is increased low resistance blood flow.

At present, women presenting with potential subinvolution of the placental implantation site are mostly managed by hysterectomy due to massive maternal haemorrhage. The diagnosis of subinvolution of the placental implantation site is only made retrospectively on histological examination of the uterus. In this rare case, we were able to reach the diagnosis based on histological diagnosis following a hysteroscopic removal of suspected RPOC. Despite hysteroscopic resection, the patient continued to have ongoing PPHs and we describe in detail the medical management and fertility sparing option offered to this woman.

Case presentation

A 40-year-old women had a pregnancy complicated by gestational diabetes that was well controlled and she had a normal vaginal delivery with an estimated blood loss (EBL) of 600 mL. She was discharged home on day 1 post delivery. This was her third normal vaginal delivery. She had no other known medical history of note.

Ten days post delivery, she presented to the Maternity Assessment Unit with her first episode of secondary PPH, with an EBL of 1000mLs. At admission, she was haemodynamically stable and her initial blood investigation revealed a haemoglobin (Hb) of 105 g/L, with normal platelet function, white cell count and C reactive protein. A bedside ultrasound scan showed thickened endometrium with evidence of RPOC. She was admitted to the obstetric high dependency ward for close observation. As her bleeding settled following admission, she received 24 hours of intravenous antibiotics, which was then followed by an evacuation of retained products of conception (ERPC) under ultrasound guidance. During the operation, blood clots of around 150 mL and small amounts of tissue were evacuated and sent for histopathological diagnosis. Histology reports later showed decidua only with no evidence of RPOC.

Two hours after the procedure, she had another episode of PPH of a further 2000 mL, and this dropped her Hb 66 g/L, requiring 3 units of blood transfusion. She underwent a CT angiogram due to concerns of ongoing bleeding. The CT angiogram highlighted a bulky postpartum uterus with no site of active bleeding and no evidence of arteriovenous malformations. She had no further bleeding episodes and was discharged to complete a course of oral antibiotics and advised to return for follow-up.

She returned on day 21 post delivery for a routine review. She had completed her full course of antibiotics and her bleeding had overall reduced with the ongoing vaginal spotting, with episodes of heavy period like bleeding in between. She also reported ongoing episodic lower abdominal pain. A repeat ultrasound scan was inconclusive on the persistence of RPOC. She was reassured to follow a watch and wait approach and discharged with further follow-up.

The patient returned the following day on day 22, with another episode of heavy vaginal bleeding and requiring admission to the postnatal ward.

Investigations

During her second readmission, she underwent a further pelvic ultrasound scan (figure 1). This showed that the endometrial cavity was distended by a hypoechoic and hyperechoic tissue structure. There was doppler enhancement suggestive of a vessel supplying this tissue. There was no clear break with the endometrial lining and no arteriovenous malformation seen to suggest an alternative pathology. This was suggestive of RPOC. Serum human chorionic gonadotrophin (hCG) at this point was performed and reported to be <1 IU/L.

Figure 1.

Ultrasound image showing the longitudinal cross section of the endometrium taken on day 22 post delivery, following readmission with secondary postpartum haemorrhage. The left figure (A) demonstrates appearances of retained product of conception within the endometrial cavity measuring 20×8 mm. The right figure (B) shows the endometrium with colour Dopplers.

The patient underwent a hysteroscopy and transcervical removal of RPOC under general anaesthesia. At hysteroscopy, a white and blueish tissue was seen on the posterior wall, along with blood clots. This tissue sample was sent for histology. The endometrial cavity appeared empty at the end of the hysteroscopic procedure with no evidence of retained tissue. The histology report showed degenerate decidua and fragments of smooth muscle with dilated blood vessels highly suggestive of subinvolution of the placental implantation site.

Treatment

She remained in hospital for 7 days after the hysteroscopic removal of tissue and continued to have episodes of heavy bleeding that settled spontaneously. She was extremely keen on preserving her fertility and given the complexity of the case, a multidisciplinary discussion between the gynaecologist, interventional radiologist and gynae oncologist was conducted to ascertain further management. All options of treatment were discussed with the patient. This included medical management in the form of hormonal suppression, interventional radiological in the form of selective vessel embolisation and uterine artery embolisation (UAE),^{4 5} and surgical options. The definitive surgical option was a hysterectomy, which was also how these cases appeared to have been diagnosed retrospectively in the current literature. This was the common management option in patients with recurrent major PPH. However, this would take away her option of any future pregnancies and carries with it associated surgical risks. The patient expressed her strong desire to preserve her fertility and declined surgical management in the form of a hysterectomy.

In order to preserve her fertility, medical option included a trial of progesterone only pill (POP) and Gonadotrophin releasing hormone (GnRH) analogue, but its efficacy is not well documented in the literature and carried a risk of further haemorrhage. Selective vessel emoblisation was an alternative option if specific vessels could be identified to be causative of the bleeding. If no vessel could be identified, UAE could be performed; however, there are limited data at present on its effect on fertility and there were no data to date on the use of this method to treat subinvolution.

Following a detailed discussion, the patient decided to have a trial of medical management and selective vessel embolisation. The option of a further CT angiogram to identify any specific bleeding vessel was offered to the patient for selective embolisation. The risk of repeat radiation from CT scan was discussed with the patient in detail, and on balance the need for a diagnosis, she was in agreement to proceed with the investigation. A CT angiogram was performed, which did not show evidence of arteriovenous malformation and there were no vessels for selective embolisation. The patient declined UAE. She was discharged on day 28 post delivery with desogestrel 150 μg once a day, GnRH analogue and tranexamic acid. The patient was counselled on the impact of GnRH analogue on breast milk production; as she was bottle feeding, this was not of ongoing concern for her.

Outcome and follow-up

The patient was subsequently followed up on a 2 weekly basis. On day 35 post delivery, a pelvic ultrasound scan showed a hypoechoic mass, likely to be a haematoma (figure 2A). It was otherwise normal. The patient remained well and tranexamic acid was stopped.

Figure 2.

Ultrasound images showing the longitudinal cross section of the endometrium taken on. (A) Day 35 2 weeks after hysteroscopy, showing a hypoechoic mass likely representing a haematoma. (B) Day 50 on high-dose medical therapy of GnRH analogue and progesterone only pill, showing resolution of the haematoma. (C) Day 90 on progesterone only pill alone with thinned endometrium of 1.4 mm. (D) Day 118 showing thickening endometrium off medical treatment.

On day 50, a further ultrasound scan showed resolution of the previously seen haematoma (figure 2B). Similarly, at 3 months post delivery (day 90), her endometrium had normal appearances on ultrasound scan, and the endometrial thickness was 1.4 mm (figure 2C). The patient continued to have on going spotting and lower abdominal cramping pain. Decision was made to stop desogestrel with the aim to allow her endometrium to regenerate.

Her follow-up at 4 months post delivery (day 118) showed improvement of her endometrial thickness to 5 mm (figure 2D). A further scan 2 weeks later on day 132 showed endometrial thickness of 3.7 mm.

The patient had her final check-up at 7 months post delivery (day 209). The pelvic ultrasound showed normal endometrium, with normal regularity with no evidence of intrauterine adhesion on saline sonogram (figure 3). Her menstrual cycles had returned to normal, and she was discharged from follow-up.

Figure 3.

Ultrasound image showing the longitudinal cross section of the endometrium taken 7 months post delivery and 4 months off medical treatment. The left figure (A) demonstrates a heathy endometrium with returning endometrial thickness. The right figure (B) was produced with saline sonography to demonstrate a patent endometrium with no scarring.

Discussion

Subinvolution of the placental implantation site should be considered in women who present with recurrent PPH, as it can be associated with life-threatening haemorrhage. PPH secondary to subinvolution is typically more common in the second week post delivery.⁶ The exact cause of subinvolution is unknown, but it is associated with risk factors such as age and multiparty.

Subinvolution of the placental implantation site can be a challenging diagnosis as it typically requires histological specimen to confirm the diagnosis. This is most commonly obtained following surgical evacuation of or hysterectomy following a major obstetric haemorrhage. Histology specimen characteristically have large, tortuous and dilated vessels close by to each other. These are spiral arteries, which are often filled with red blood cells or organised thrombi. It is common to have normally involuted vessels nearby.

The use of sonography can be crucial in the diagnosis of subinvolution of the placental implantation site. Ultrasound images demonstrate low resistance vessels within the myometrium; these are hypoechoic tortuous vessels visualised within the myometrium. These are different to arteriovenous malformations, which are typically high-flow uterine vascular malformation.⁷ RPOC can often be confused with the above findings.

Management of subinvolution is often challenging, because the case commonly present as an obstetric emergency due to severe maternal haemorrhage. A literature search of case reports between 2000 and 2020 revealed that the majority of cases are managed surgically (table 1) The first approach is uterine curettage, followed by hysterectomy due to ongoing bleeding. All the cases reported in the medical literature are diagnosed retrospectively based on histological findings either from biopsy or hysterectomy specimen.

Table 1.

Case report summary from literature search showing the type of patients and their management outcome

Case	Parity	Age	Delivery	Postpartum interval time (days)	Management
1,13	G2P3 (twins)	23	Caesarean Section for Pre Eclampsia, 35 weeks	21	Curettage, then hysterectomy 1 week later
2,13	G2P2	25	Single vaginal delivery, term	5	Uterine ablation and ergotin
3,17	G1P1	24	Caesarean Section for breech, term	27, 43, 63	Day 43—estrophem/2 mg/+duphaston/dydrogesterone/10 mg daily Day 63—abdominal hysterectomy
4,20	G5P1	35	Caesarean Section for twins	60	Curettage, then hysterectomy due to ongoing bleeding
5,21	G1P1	19	Caesarean Section	8	Death
6,22	G1P1	21	Caesarean Section for Failure to progress, term	21	Hysterectomy
7,23	G1P1	36	CS for twins	10	Curettage followed by hysterectomy

In a PPH scenario, it is vital to consider all conventional treatment options that have been proven to be effective in treating this condition. Treatments such as the use of uterotonic medications, bimanual compression, an intrauterine balloon device (eg, Bakri balloon, Ebb balloon or Rusch balloon),⁸ B-Lynch suture,⁹ UAE,^{4 5} and a hysterectomy are all recognised methods with adequate safety data to support their usage in a life threatening haemorrhage.¹⁰ In an unstable patient, conservative treatment has been shown to be associated with increased mortality and thus early recourse to optimal surgical treatment should be sought.¹¹

While hysterectomy is considered the definitive curative treatment in managing subinvolution in particular, it carries significant surgical challenges and increased maternal morbidity and mortality. Other methods have been documented such as the use of interventional radiology for embolisation of either selective vessels or the uterine artery.^{4 5} A review by McLucas et al details several case series of successful pregnancies following UAE; however, complications such as miscarriage, preterm delivery, malposition and PPH were reported by other studies.¹² Uterine ablation has been demonstrated in one case to have been effective in its management for subinvolution.¹³ There has been one further case report on the use of uterine ablation as a method for management of PPH.¹⁴ In the context of future fertility, there are well documented consequences on future pregnancy outcomes following uterine ablation, such as increased caesarean births, preterm delivery and stillbirth.¹⁵

Progesterone and oestrogen play an important role in decidual angiogenesis during pregnancy. Reduction of these hormones in the postpartum period leads to low levels of Vascular Endothelial Growth Factor-A (VEGF-A), which causes vascular regression.¹⁶ Aggressive suppression of the hypothalamus–pituitary–ovarian (HPO) axis reduces progesterone and oestrogen levels further, which was thought to be the main driving factor for the successful management of this case. While previous medical managements have not been shown to be successful, this may be due to insufficient dosage.¹⁷ A combination of both POP and GnRH analogue ensures that the HPO signal pathway is turned off, leading to medically induced menopause. This reduces the dilated spiral arteries to prevent further bleeding. Incidentally, there has been documented successful medical managements of subinvolution in animal case series using progestogen alone.^{18 19}

While fertility was very important for this particular patient, she was clearly counselled that this would not be at the cost of increasing morbidity or mortality. The ability to manage her successfully via medical management was done with careful counselling and establishing clear threshold for intervention.

In summary, aggressive medical therapy can be a consideration for management of subinvolution of the placental implantation site. Not only can this can reduce maternal mortality and morbidity associated with major surgery, it also offers a fertility sparing option. Expertise in sonography can provide invaluable information in its diagnosis as well as follow-up to ensure resolution.

Patient’s perspective.

I was initially admitted into hospital to have an induction of labour, because I was told that I had gestational diabetes. I had quite a tough labour. I became hypertensive during the end of the labour and baby’s heart rate kept dropping and I suffered a small placental abruption. I gave birth 2 days after the initial induction. I was kept in hospital for a few days then discharged home.

When I got home, I started experiencing painful cramping and I would suddenly have gushes of very dark blood. I saw my midwives a couple of days after I was discharged, and they diagnosed me with an infection in the womb and gave me antibiotics. The midwife saw me every few days after being on the antibiotics, but the bleeding never really stopped and the cramping continued. They advised me to go to the Maternity Assessment Unit to be assessed.

In the assessment unit, I was scanned and they saw that my womb was full of blood. They diagnosed me with retained products. I was admitted straight away and given antibiotics. I had an operation the next day to remove the products. After the operation, I didn’t feel too bad. However, when I got up to use the bathroom with the help of the nurses, I passed a massive clot the size of my hand and I started bleeding heavily. I was in a lot of pain and I kept gushing. As this was happening, my heart rate and blood pressure appeared fine. Then I suddenly started feeling really poorly and my blood pressure dropped. I had a blood test to check my haemoglobin and it was OK to start with, and then it dropped. They tested it again and it dropped again to 72. My blood pressure was also low. I had three blood transfusions. I was on the observation ward for a few days till I was transferred the postnatal ward. I still had some blood in the womb and my blood pressure and haemoglobin were still slightly low, so I had an iron infusion. I was allowed to keep my baby with me the whole time as I was breastfeeding and when I went for my scans, the nurses would look after the baby.

I was discharged from hospital after a few days stay, but kept under the care of my local midwifes. I saw them a couple of days after I had been discharged, but I was still bleeding on and off. It would stop for a while, then I would have a gush and fill a pad up in minutes. The midwife told me to go back into hospital as she still wasn’t happy.

After an assessment in the hospital, the doctors felt that a lot of the pain and bleeding I had was from the birth and the operation. I was asked to return the next day for an ultrasound scan. This showed that I had a black area in the womb, which represented something in the womb. The consultant wanted me to have another scan on a better machine a few days later. I was told to come straight back if the bleeding or pain got worse.

I went to bed that night and had a lot of cramping. I woke in the morning feeling very wet, when I stood up, clots the size of my hand started dropping out of me and the pain was awful. I was having another haemorrhage, and my husband rushed me straight back to the hospital. I was admitted into the hospital, where a lovely nurse helped to clean me up. I saw another consultant, who was lovely and put me at ease. I had an internal scan done and my womb still had clots inside. As it looked like there was still something there, I had a hysteroscopy to remove them.

After the operation and more monitoring, I was transferred back to the postnatal ward. It was nice that I could have my baby with me the whole time. Unfortunately I was examined regularly, because I was still bleeding. My uterus was still high and my stomach was “boggy”. My haemoglobin was also a little low. The doctors informed me that they think a vessel hadn’t shut properly from the birth, so even though I had some scans, but unfortunately as it was a intermittent leaking it never showed up on the scans.

As I was still bleeding, I was told I needed to have another operation, where they would clamp the uterine artery. However that could cause problems and there was a chance it could damage the womb, which I didn’t want as I still wanted more children.

The consultant who performed the hysteroscopy on me was brilliant and he later came to explain the results of the biopsy from the hysteroscopy. The consultant talked to me on what the next steps were. I explained that I wanted to have more children, and he listened to me talked to me through a process where I could be treated with medical intervention. I was given an injection in the tummy of a GnRH analogue, tablets of progesterone and tranexamic acid to prevent any more haemorrhages in the hope that given time, the vessel would heal.

I was discharged from the hospital and seen regularly in clinic. I eventually came off the anti-haemorrhage medication and I stayed on the progesterone tablets for a little longer. Eventually I came off those as well. During the follow up clinics, I had regular scans and check ups. My womb lining was slowing returning to normal.

At the final follow up clinic, I had another scan as well as filling my womb with saline to see if I had any scar tissue. I received good news that there was no scar tissue was found and my womb lining looks good, which means if I wanted to there would be no reason why I couldn’t carry more children. My periods are still erratic and fairly light, and I do still have tummy pain, which the doctor thinks are due to my pelvic floor muscles rather than my uterus. Hopefully with a bit more time my periods will settle down and the discomfort I experience will get better.

Learning points.

• Subinvolution of the placental bed should be considered in women, who presents with recurrent presentation of postpartum haemorrhage in the absence of infection or retained tissue

• A shared decision-making model should be employed particularly in these complex cases. Women should be fully informed of the long-term and short-term implications, particularly as it may affect their future fertility.

• Medical management of subinvolution of the placental bed can be considered in a woman, who wish to preserve their fertility and avoid a high-risk operation.

Ethics statements

Patient consent for publication

Obtained.