Astract
Objective : The main objective was to assess the efficacy of ‘stepwise uterine devascularisation in management of uncontrollable postpartum haemorrhage.

Introduction

In spite of rapid advances in medicine, obstetric haemorrhage still continues to be important cause of maternal morbidity and mortality. Post partum haemorrhage is one of the most common and most serious haemorrhagic threats to the parturient.1 Uterine atony, couvalaire uterus, abruptio placentae and placenta previa are the common causes of non-traumatic post partum haemorrhage.

Usually such post partum haemorrhage is controlled by uterine massage, oxytocin, methyl ergometrine and/or prostaglandins. But sometimes these standard measures fail and obstetrician has to resort to hysterectomy with /without ligation of internal iliac artery.

Hysterectomy is a radical procedure that carries undesirable effects of reproductive sterility, secondary amenorrhoea along with physical and psychological trauma.2

During pregnancy the major blood supply to the uterus comes from uterine arteries (90%), rest 10% from cervical, vaginal, ovarian vessels.3 During pregnancy, uterine blood flow increases by many fold and ranges between 600-750 ml/min. In uncontrollable haemorrhage, uterine arteries fail toconstrict due to an unidentifiable mechanism, probably an effect of pregnancy.4 Thus, occlusion of uterine vessels can reduce most of the blood supply to the uterus causing uterine ischaemia resulting into substantial decrease in the bleeding. This occlusion of vessels is temporary and recanalisation and normal uterine circulation is generally established within few months.

In 1952, Waters reported ‘Bilateral uterine artery ligation’ as anatomically sound, physiologically rational and surgically possible method to control Post partum haemorrhage.1 Since then many authors have reported beneficial effects of bilateral uterine artery ligation in management of severe postpartum haemorrhage. Unfortunately, the failure of uterine artery ligation in these studies ranged from 8 to 20%.2,3 Abd Rabbo5 in 1994 reported technique of uterine devascularisation in stepwise manner by taking additional sutures for low uterine artery ligation and ovarian artery ligation in cases of failure with uterine artery ligation and achieved 100% success. In this paper, we are presenting series of 23 cases of ‘stepwise uterine devascularisation performed at our institute for management of severe postpartum haemorrhage.

The aim of study was to assess usefulness of ‘stepwise uterine devascularisation’ as an alternative to obstetric hysterectomy in control of non- traumatic postpartum hemorrhage.

Material and Methods

The study consists of 23 women in whom ‘stepwise uterine devascularisation’ was done for control of non-traumatic postpartum haemorrhage.

The study was done over a period of approximately four years in department of obstetrics and gynaecology at major tertiaryteaching hospital in Mumbai.

In all these patients, classic conservative methods to control bleeding i.e. - oxytocin drip, intra-muscular methyl ergometrine injection, rectal misoprostal or intramuscular/ intra myometrial injection 15-methyl prostaglandin F2a along with bimanual massage and compression of the uterus were tried. In cases of vaginal deliveries, exploration of uterus and removal of any retained bits of placenta and careful inspection of lower genital tract to exclude traumatic cause were done.

After failure of these standard measures, decision for ‘stepwise uterine devascularisation’ was taken. The technique includes ‘three’ procedures done bilaterally in step-by-step fashion. Thus, if step I failed to control bleeding, we proceeded to step II, and so on to step III. After each step, assessment of bleeding was done directly through the incision, in case of caesarean sections or by observation of blood coming from the vagina for 5-10 minutes in cases of vaginal deliveries.

Step I: Bilateral Uterine vessel ligation

Uterine artery was ligated at level where it runs along the uterine border besides upper end of lower uterine segment. In cases of caesarean sections, it was done just below the angle of incision on the lower segment.

In patients with vaginal delivery, utero-vesical peritoneum was incised and bladder was dissected down.

For left uterine artery ligation, uterus was grasped and elevated anteriorly and to the right by left hand. The palm was put over the left uterine border with thumb anterior over a lower uterine segment and four fingers posterior to broad ligament protecting the intestines. A large Mayo needle with No.1 chromic catgut suture was passed into and through the myometrium anterior to posterior and then, guided by the four fingers of the left hand, brought forward through the avascular area of the left broad ligament lateral to uterine vessels and suture was tied.

The procedure was repeated on other side. After bilateral uterine artery ligation, myometrium assumes pinkish blanched hue secondary to ischaemia. As mentioned earlier, the decrease in the bleeding was directly assessed through the incision in cases of caesarean section, while in cases of vaginal deliveries it was assessed by observing blood coming out of vagina. For this, uterus was compressed to remove clots then covered with warm pack, abdominal wound covered with sterile towel, thigh and legs flexed and abducted, vagina cleaned and bleeding observed for 5-10 min.

If haemostasis is not adequate, additional lower stitches on either side (Step II) were taken.

Step II: Bilateral Low uterine ligation

This step was particularly required in cases of bleeding from lower uterine segment as in cases of placenta previa.

In this step, bladder was further mobilized and pushed down. A stich was taken at lower part of the lower uterine segment, near uterine genu i.e. where it gives cervico-vaginal branch. It was important to include a good chunk of myometrium to avoid injury to vessels and to obliterate intramyometrial branches of cervico-vaginal branch.

Step III: Bilateral ovarian vessel ligation

This step was required only in those cases where significant bleeding continued in spite of performing step I and II.

In this step, the uterus was grasped to the contra lateral side by left hand, and a large atraumatic needle with No. 1 chromic catgut suture was passed through the avascular area in the infundibulopelvic ligament from posterior to anterior and suture was tied.

All patients were given proper supportive and post-operative care. Apart from this procedure, blood and blood product transfusions, antishock measures, fibrinogen and agents like tranaxemic acid and ethamsylate were used concomitantly, depending on patients’ need.

All women were asked to follow-up in the out patient department, to check the resumption of menstruation and possible conception.

Observations and Results

The age of the patients ranged from 18 to 36 years with the mean age being 27.6 years. There were 12 primigravidas and 11 multigravidas (Table 1).

Period of gestation at time of delivery ranged between 27 and 41 weeks. (Table 2).
The incidence of preterm delivery was 43.5%. The aetiologies of postpartum haemorrhage in these cases are listed in Table 3.

Associated antenatal risk factors or problems are shown in Table 4. There was one case of intra-uterine foetal death that later on developed disseminated intravascular coagulation. When these risk factors were analyzed with respect to cause of post partum haemorrhage, definite correlation was found in some cases (Table 4). Excessive enlargement of uterus due to factors like multiple pregnancy or hydramnios is known risk factors for uterine atony. In our series, there was one woman with hydramnios and one of twin pregnancy. Pregnancy induced hypertension, known risk factor for accidental haemorrhage; couvalaire uterus and coagulation disorder was present in six patients.

There were six vaginal deliveries, which included normal deliveries (3), forceps delivery (1), assisted breech delivery (1) and twin delivery (1). Seventeen women had caesarean sections. Mode of delivery in these 23 patients who had postpartum haemorrhage is shown

Placenta previa and previous caesarean section were common indications for caesarean sections. The indications for caesarean sections are summarized in Table 6.

Effectiveness of procedure

Stepwise uterine devascularisation was effective in all but one patient. Step I was sufficient to achieve haemostasis in 17 cases.

In remaining 6 women, additional step II was required to control bleeding. Out of these 6 patients, four had significant decrease in bleeding after step II. All four women in whom step II helped to achieve haemostasis were cases of placenta previa.

In other two patients even ovarian vessel ligation (Step III) was done. Though in one case bleeding decreased in one case, other patient continued to have substantial bleeding warranting additional intervention. Thus, technique was effective in curtaining the bleeding in all but one case.

This was a patient transferred from peripheral maternity home with intrauterine foetal death with impending afibrinogenaemia and disseminated intravascular coagulation. Patient had massive postpartum haemorrhage after vaginal delivery. In this case, significant bleeding continued even after step-III. Hence, hysterectomy with bilateral internal artery ligation was done. Patient was transfused with 11 unit of blood and 8 units of fresh frozen plasma. Patient was on ventilator for four days, later on came out and recovered completely.

Thus effectiveness of this procedure was 95.7%. All patients survived.

There were no complications noted except one case of broad ligament haematoma, which was self-limiting and required no additional treatment.

The average estimated blood loss was 1250 ml. All patients required blood transfusion. Sixteen patients required more than two units of blood. A woman with afibrinogenaemia also required transfusion of fresh frozen plasma.
Thus, technique was effective in curtailing the bleeding in all but one case.

Follow - up

Discussion

Postpartum haemorrhage is one of the most commonly encountered and most dangerous obstetric emergencies, requiring prompt and effective treatment.1 Though in some cases, it is because of injury to genital tract, in majority of cases, atony of uterus is the cause for the bleeding. In most patients, judicious use of oxytocin or methyl ergometrine and uterine massage is effective in controlling the atonic postpartum haemorrhage. In last few years, prostaglandins have revolutionized the management of postpartum haemorrhage with its strong oxytocic property, which could cause effective uterine contractions, and control of intractable postpartum haemorrhage. Recently, misoprostal, in oral or rectal route has been successfully used in control of postpartum haemorrhage. Unfortunately, in certain cases, these standard methods fail to act, for some unknown reasons, where more aggressive intervention is required.

Hysterectomy is the usual technique for management of such uncontrollable, non-traumatic postpartum haemorrhage when the classic conservative measures fail to control bleeding.^2,6,7 This surgical procedure is often done in emergency when patient herself is in a state of shock and hence unable to take decision. When the woman recovers, she is confronted with the fact that her uterus has been removed and that she will be permanently sterile and ammenorrhic. This can lead to considerable psychological trauma, particularly in primiparous or women with low parity. Very often, there is foetal death in conditions like abruptio placenta or placenta previa. Moreover, hysterectomy is a major procedure associated with blood loss.

Hypogastric artery ligation, although at times life saving, doesn’t result in complete cessation of uterine blood flow, as does direct ligation of uterine vessels. Also, the procedure is technically more difficult, takes much longer time, requires more dissection and may be associated with troublesome complications.⁸

Angiographic embolisation of bleeding vessel with gelatin sponge has been reported for treatment of postpartum haemorrhage.9 Materials used for embolisation include gelfoam, subcutaneous fatty tissue, autologous clot or small sialastic spherols. Gelfoam is commonly preferred because it is sterile, nonantigenic, remains in vessel for 20-50 days forming fibrin meshwork on which the blood clot is formed. After injecting the material, repeat angiography is done to confirm the haemostasis and then catheter is removed. However, most of these patients who are in state of hypovolaemia or even shock and difficult to shift them to the place of angiographic embolisation. Also, facility, special instrumentation and trained personnel may not be available in all centres.

Christopher B-Lynch described technique of ‘Brace sutures’ for control of intractable postpartum haemorrhage.¹⁰ B-Lynch described using this technique successfully in five patients. Though, there was no mortality, all patients required massive blood transfusions and needed intensive care for few days. The technique requires more detailed evaluation.

The stepwise uterine devas-cularisation is simple and easy conservative procedure. Any obstetrician with moderate experience can perform this after initial training. It does not require much dissection, as in cases of hysterectomy or internal iliac ligation. The time required is much shorter, which is important in such critical patients. This technique is not associated with blood loss, which is important in such serious patients who had already lost a considerable amount of blood.

In this technique, absorbable suture was used and vessels are not divided, so recanalisation will occur. Meanwhile collateral circulation appears to be sufficient to prevent complications. In our study, in one patient who became pregnant after stepwise uterine devascularisation technique and who was delivered by caesarean section, the uterine vessels appeared to have been recanalised.

As main blood supply of uterus comes from uterine arteries, it is logical that ligation of these vessels could help in controlling bleeding in majority of cases. Our study shows that procedure was effective in 95.7% of cases. A studies by O’Leary,^2,3 Sabzposh,¹¹ bilateral uterine artery ligation was successful in 80 to 92% cases. In their studies, cases that did not respond to bilateral uterine artery ligation were cases of placenta previa and afibrinogenaemia.

The lower segment and cervix are supplied by descending cervico-vaginal branch of uterine artery, which originates near uterine genu or where artery crosses ureter and then runs downwards along the lateral border. Thus, in placenta previa a significant portion of the placental site blood supply comes from these cervical and vaginal arteries. Since in bilateral uterine artery ligation, the sutures are placed high above these arteries, it generally doesnot help to control bleeding from the placental bed of lower uterine segment. In such cases, it is necessary to take additional lower sutures to occlude these branches (Step-II).

In step-I and Step-II, sometimes even unilateral sutures can substantially decreases bleeding. But, unilateral ligation is practical only in cases of caesarean section where assessment of decrease in bleeding can be done immediately through the incision on the lower segment. In cases of vaginal deliveries, pervaginal assessment of bleeding after each step is too time consuming and not practical. In our study, we performed bilateral ligations in each step before assessing effect on bleeding.

As part of blood supply to uterus comes from ovarian arteries. In cases where step-I and II fail to stop bleeding, it is logical to occlude ovarian vessels in infundibulo-pelvic ligament. This is particularly useful in cases of afibrinogenaemia and DIC.

In a study by AbdRabbo,5 the technique was effective in 100% cases as he performed stepwise uterine devascularisation rather than only bilateral uterine artery ligation. Unfortunately, in our study, success rate of stepwise uterine devascularisation was only 95.7%. In one woman, bleeding continued even after step-III warranting hysterectomy.

Broad ligament haematoma is possible complication in some cases. O’Leary et al reported 2 cases of broad ligament haematoma.2 We also had one case of broad ligament haematoma , which was self limiting and required no additional treatment. Such haematomas can be avoided by going through the myometrium on medial side and choosing avascular area in broad ligament lateral to uterine vessels.

The general concern during this procedure is ureteral injury. But in step-I, the suture is much higher and far away from ureter. In step-II, such injury can be well avoided by proper dissection of bladder down, thus pushing ureters away from uterine genu. In our study, we did not come across any urinary tract complications.

Howard (1968) in his study of 6 cases reported formation of arteriovenous fistula in one case.12 This complication has not been reported by any other author and can be easily avoided by using absorbable suture material like chromic catgut, including substantial amount of the myometrium at the ligation site and by avoiding the figure of eight suture.

There is no long term adverse effect reported by any of the author. The resumption of menstruation was normal and 5 patients conceived within one year of procedure. Similar findings are reported by AbdRabbo6 (1994) and O’Leary et al2,3 (1994).

Conclusion

From the results of our study and review of literature, it can be concluded that ‘stepwise uterine devascularisation’ is simple, safe, effective and life saving alternative in management of postpartum haemorrhage and should be offered and done in all cases of non-traumatic intractable postpartum haemorrhage, before resorting to hysterectomy.

References

1. Waters EG. Surgical management of postpartum hemorrhage with particular reference to ligation of uterine arteries. Am J Obstet Gynecol 1952; 64 : 1143-8.
2. O’Leary J, O’Leary JA. Uterine artery ligation in the control of of intractable postpartum hemorrhage. 1966; 94 : 920-4.
3. O’Leary J, O’Leary JA. Uterine artery ligation for control of post-cesarean section hemorrhage. Obstet Gynecol 1974; 43 : 849-53.
4. Nelson SH, Suresh NS. Lack of reactivity of uterine arteries from patients withobstetric hemorrhage. Am 1992; 166 : 1436-43.
5. AbdRabbo SA. Stepwise uterine devascularisation: A novel technique for management of uncontrollable postpartum hemorrhage with preservation of the uterus. Am J Obstet Gynecol 1994; 171 : 694-700.
6. Clark SL, Phelan JP, Bruce SR, Paul RH. Emergency hysterectomy for obstetric hemorrhage. 1984; 64 : 376-80.
7. Strudee DW, Rushton DI. Cesarean and postpartum hysterectomy. Br J Obstet Gynecol 1986; 93 : 270-4.
8. Burchell R. Internal iliac artery ligation hemodynamics. Obstet Gynecol 1964; 24 : 737-9.
9. Brown BJ, Heaston DK, Poulson AM, et al. Uncontrollable postpartum bleeding: a new approach to hemostasis through angiographic arterial embolisation. Obstet Gynecol 1979; 54 : 361-5.
10. B-Lynch C, Coker A, Laval AH, et al. Brit J Obstet Gynecol 1997; 104 : 372-75.
11. Sabzposh NA, Hisamuddin S, Sultana K. Bilateral uterine artery ligation: An alternative to hysterectomy for control of non-traumatic post-partum hemorrhage. J Obstet Gynecol India .1997; 47 : 780-84.
12. Howard V. Iatrogenic arteriovenous sinus of uterine artery and vein. Obstet Gynecol 1968; 31 : 255-7.