A double-blind randomised trial of leuprorelin acetate prior to hysterectomy for dysfunctional uterine bleeding
Correspondence: Mr S. R. G. Duffy, Department of Obstetrics and Gynaecology, Level 09, Gledhow Wing, St James's Hospital, Leeds LS9 7TF, UK.
Objective To evaluate the use of pre-operative leuprorelin acetate for reducing the morbidity from hysterectomy for nonfibroid menorrhagia.
Design A double-blind, randomised, placebo-controlled trial.
Setting Gynaecology department in a large university teaching hospital.
Sample Fifty-one women without uterine fibroids awaiting abdominal or vaginal hysterectomy for dysfunctional uterine bleeding.
Methods Participants received leuprorelin acetate or placebo for eight weeks prior to hysterectomy.
Main outcome measures Operative blood loss, operative difficulty, first day morphine use, speed of return to ‘normal health’.
Results The study and control groups were similar as regards prognostic factors. Two women in the study group withdrew because of side-effects. Although a 34% reduction in uterine volume was seen in those treated with leuprorelin, there were no significant differences in operative blood loss (183 mL in the study group vs 285 mL in controls, P= 0.27), operation time (39 vs 49 min, P= 0.64) or operative difficulty (visual analogue scale 3.0 vs 4.0, P= 0.09). Furthermore, there was no difference between the groups in post-operative morbidity or rate of recovery.
Conclusions Treating women with leuprorelin acetate for 8 weeks prior to surgery for nonfibroid menorrhagia has no significant operative or post-operative benefits.
Despite a rapid increase in use of minimally invasive techniques for the treatment of nonfibroid menorrhagia, nearly 90% of women undergoing surgical treatment have a vaginal or abdominal hysterectomy1. About 72,000 women are estimated to undergo hysterectomy every year in the UK making it one of the most common major operations performed in any specialty. The average recovery time of seven weeks2, means that over 3.5 million work days are lost annually at a huge social cost. Randomised trials have suggested a number of ways in which recovery rates and post-operative complication rates might be improved in order to decrease the individual and social burden of this operation. These include the increased use of vaginal hysterectomy (with or without laparoscopic assistance), intra-operative antibiotics, post-operative heparin, nonclosure of the peritoneum and use of pre-operative gonadotrophin releasing hormone agonists to reduce the vascularity of the uterus.
Gonadotrophin releasing hormone agonists such as leuprorelin acetate induce a reversible medical menopause by inhibiting the production of gonadotrophins from the pituitary gland. In 1988 Matta et al. demonstrated that gonadotrophin releasing hormone agonists decreased both the volume of uterine fibroids and the associated uterine artery blood flow3. This led to a number of trials investigating their use prior to hysterectomy as a means of decreasing operative morbidity in women with fibroids. They were found to be of benefit in three circumstances; when used in combination with iron therapy in anaemic women they decreased the need for peri-operative blood transfusion; they decreased the intra-operative blood loss and operating time; and they decreased the fibroid volume allowing more operations to be conducted vaginally4–6.
Women without uterine leiomyomata might also benefit from pre-operative gonadotrophin releasing hormone agonists. Gonadotrophin releasing hormone agonist therapy effects the myometrial and fibroid volume in similar ways, decreasing both by about 35%7, and studies conducted in women before and after the menopause suggest that uterine blood flow is decreased with oestrogen withdrawal even in the absence of fibroids8. Gonadotrophin releasing hormone agonists also reduce menstrual blood loss irrespective of the presence of fibroids9, and should therefore have the same beneficial effect on the pre-operative haemoglobin concentration. This randomised trial has therefore been designed to test the hypothesis that treatment with a gonadotrophin releasing hormone agonist prior to hysterectomy can reduce operative morbidity and improve the rate of recovery.
Fifty-one women awaiting hysterectomy for dysfunctional uterine bleeding were enrolled from a gynaecological outpatient department. All had socially disabling menorrhagia unresponsive to medical therapies and had been investigated to exclude uterine malignancy. Women were excluded if they had been treated with a gonadotrophin releasing hormone agonist within the preceding six months, if the ultrasound examination revealed uterine leiomyomas greater than 2.5 cm in diameter, or if the women were on any hormonal treatments. Ethical approval was obtained from the local ethics committee. All women participating in the study did so voluntarily and provided written informed consent prior to randomisation. The study size was calculated using data on operative blood loss from Stovall et al.6 It was calculated that with 24 women in each group the study would have 95% power to detect a 20% decrease in operative blood loss that would be statistically significant at the 5% level.
Identical syringes, pre-filled with vehicle solution, were placed in identical boxes with or without vials of leuprorelin acetate depot 3.75 mg powder (Prostap SR, Wyeth-Lederle, Maidenhead, UK) and numbered consecutively according to a computer generated randomisation programme (in blocks of four). Departmental staff not involved in the trial carried out the numbering of boxes and administration of the two injections (given monthly) of placebo (sterile vehicle alone) or reconstituted leuprorelin acetate. The boxes were opened and the syringes prepared out of sight of the trial participants to ensure effective blinding. All other staff were blind to the treatment allocation.
Ultrasound scans were conducted at the time of entry to the trial and after eight weeks of treatment. All the scans were performed by A.D.W. on a Toshiba SSA-270 A using a 7.5 MHz transvaginal probe. The uterine volume was calculated using the formula for a prolate ellipsoid (volume = 0.52 × length × antero-prosterior diameter × transverse diameter).
Serum samples were obtained from the participants at the time of entry of the trial and again at the end of eight weeks therapy. Analysis of serum oestrogen levels was performed in the hospital's laboratories using the standard Delfia fluoroimmuno assay (coefficient of variation for intra-assay and inter-assay variability 3.8% and 9.7% respectively). Haemoglobin levels were measured at entry to the trial, pre-operatively and on the second day post-operatively.
The intended route of hysterectomy was chosen by the clinical team responsible for the woman's care prior to entry to the trial. Women were booked for a total abdominal hysterectomy, a vaginal hysterectomy, or for assessment on admission. On the day prior to surgery the intended route of operation was re-evaluated by the clinical team and a final decision made. The clinicians were blinded to the treatment allocation and were asked not to enquire about the effects of the medication. All operations were conducted under general anaesthesia by surgeons with at least seven years gynaecological training. The excised uteri underwent routine histological examination. The women received a single dose of antibiotic during the operation and heparin prophylaxis post-operatively until mobile. At the time of surgery all blood-stained swabs were collected and weighed. The excess weight of the swabs (above that of identical dry swabs) was calculated and added to the volume of any aspirated blood to calculate the total blood loss (the density of blood was assumed to be 1 mg/mL). The operations were timed and the surgeons asked to record their assessment of the ease of the operation on a 10 cm visual analogue scale marked as ‘easiest operation imaginable’ at one end and ‘most difficult operation imaginable’ at the other.
Post-operative care was carried out according to standard protocols by the clinical firms in charge of the woman's care. The number of doses of analgesia required were recorded daily, and the women completed daily visual analogue scores of the degree of pain until discharge. Post-operative infection was defined as a temperature of over °C for greater than 12 hours, a laboratory proven urinary tract infection, or a wound or chest infection requiring treatment with antibiotics. The need for blood transfusion was determined by the ward clinicians based on the woman's clinical condition. When calculating the peri-operative change in haemoglobin concentration, the women receiving transfusion were excluded from the statistics because of their artificially raised haemoglobin concentrations. The decision regarding discharge from hospital was taken clinically based on the rate of recovery of the women and their social circumstances.
Four months following surgery a postal questionnaire was sent to all participants. They were asked about their speed of recovery, post-operative complications, and their rate of return to normal activities. Nonresponders were sent two further reminders and finally contacted by telephone.
Computerised statistical analysis was conducted using the Astute statistics add-in (DDU Software, University of Leeds, Leeds, UK) for Microsoft Excel for Windows 3.1 (Microsoft, USA). Paired and unpaired t tests were used for parametric data, the Mann-Whitney test for nonparametric data. Relative risk calculations were used to compare categorical data. Results were considered statistically significant if P < 0.05 or if the 95% confidence limits of the relative risk lay outside unity.
Fifty-one women were recruited to the trial. Full results were available on 49 women as two women in the treatment group withdrew because of side-effects. Two other women (one in the study group and one in the control group) were discovered to have fibroids of 3.5 cm on histological examination which had not been detected on ultrasound. Data for these two women are included in the final results. Four women in the trial did not have a second, pre-operative scan. Two women having vaginal hysterectomies in the control group also had colporrhaphies performed. Thirty-eight percent of those having an abdominal hysterectomy also had an oophorectomy (29% in the treatment group, 45% in the study group, RR 0.64, 95% CI 0.24 to 1.66). There was no difference between the groups in terms of age, parity, or eventual histological diagnosis (Table 1).
Table 1. Characteristics of trial participants. Values are given as n or mean (SD).
|No. of participants||26||25|
|No. of withdrawals with side-effects||2||0|
|Age (years)||39 (4.4)||40 (5.7)|
|Weight (kg)||72 (13.8)||66 (12.3)|
|No. of nulliparous women||1||3|
|No. of women with previous caesarean section||2||5|
|Serum oestradiol concentration (pmol/L)|| || |
| At trial entry||296 (245)||250 (148)|
| he-operatively||70 (90)||260 (200)|
|Haemoglobin concentration (g/dL)|| || |
| At trial entry||13.0 (1.2)||12.7 (1.1)|
| Pre-operatively||13.2 (1.3)||12.6 (1.5)|
|Uterine volume (cmD)|| || |
| At trial entry||81.4 (31.7)||87.0 (32.5)|
| he-operative1y||53.5 (21.7)||79.6 (32.4)|
|Hysterectomy route|| || |
|Histological diagnosis|| || |
| Endometrial polyp*||3||0|
| Fibroids (< 2.5 cm)||1||1|
In the women receiving leuprorelin acetate there was a marked reduction in oestradiol level (P= 0.0002), while there was no significant change in the placebo group. In the leuprorelin acetate group the mean haemoglobin concentration rose by 0.20 g/dL (P= 0.37; 95% CI −0.09 to +0.49) and in the control group it fell by 0.12 g/dL (P= 0.93; CI −0.50 to +0.26). There were only three women with a haemoglobin concertration of < 11.5 g/dL at trial entry (two in the treatment group and one in the placebo group).
Baseline scans performed prior to treatment revealed wide variation in initial uterine volumes with a range of 19 cm3 to 174 cm3. However, both treatment and placebo groups had comparable initial mean volumes (81.4 cm3vs 87.0 cm3, difference 5.6 cm3, 95% CI −14.2 to 25.4). In the treatment group there was a statistically significant decrease in uterine volume of 34% while there was no change in uterine volume in the placebo group. Full details of the ultrasonographic changes are published elsewhere10.
Two women in the leuprorelin acetate group withdrew because of side-effects. Overall 58% of women in the treatment group reported side-effects compared with 36% in the placebo group. The most common side-effect was hot sweats that were reported by 42% of women randomised to leuprorelin acetate and 24% of those randomised to placebo.
At the time of recruitment to the trial 63% of the women were booked to have an abdominal hysterectomy. All of the women booked for a vaginal hysterectomy underwent a vaginal hysterectomy, and all women except one who were booked for an abdominal hysterectomy had an abdominal hysterectomy. The one exception was a woman in the leuprorelin group who was planned for an abdominal hysterectomy but was changed to a vaginal hysterectomy on reassessment prior to surgery. There were three women in the leuprorelin acetate group who were booked for assessment on admission. Two of these underwent a vaginal hysterectomy and one underwent an abdominal hysterectomy. Of the five women in the placebo group who were assessed for route on admission two had a vaginal hysterectomy and three an abdominal hysterectomy. No woman in whom vaginal hysterectomy was intended was converted to an abdominal hysterectomy intraoperatively. The difference between the two groups was not statistically significant (for those booked for assessment on arrival, relative risk of vaginal hysterectomy in the treatment group = 1.67, 95% CI 0.44 to 6.36).
Intra-operative blood loss varied widely between 20 and 1939 mL. Although the median blood loss in the leuprorelin acetate group was only 64% of that in the placebo group, this was not statistically significant (Table 2). There was no difference between the groups in the median change in haemoglobin level (Table 3). There was also no difference between the groups with regard to operating time or operative difficulty.
Table 2. Intra-operative differences between the experimental and control groups. Values are given as median (interquartile range).
|Operative||183 (89–398)||285 (130–402)||0.27|
| blood loss (mL)|| || || |
|Operative difficulty†||2.4 (1.8–3.8)||3.2 (2.3–6.6)||0.09|
|Length of||39 (36–50)||49 (34–56)||0.64|
| operation (min)|| || || |
Table 3. Post-operative outcomes of the experimental and control groups. Values are given as n, median (interquartile range) or mean [SD]. Hb = haemoglobin concentration.
|Morphine use in first 24 h (mg)||40 ||37 ||0.50 (t= 0.68)*|
|Analgesia doses on day 2 and 3||6 (4–8)||6 (3–7)||0.94 (Z =−0,077)†|
|Day of discharge||5 (4–5)||5 (5–6)||0.40 (Z =−0.95)†|
|Pen-operative change in Hb (g/dL)||−1.3 (−0.7 to −1.75)||−1.2 (−0.45 to −2.08)||0.80 (Z =−0.26)†|
|Wound haematomas||0||2|| |
|Infections|| || || |
| Urinary||1||1|| |
| Wound||1||2|| |
| Chest||0||1|| |
| Pyrexia of unknown cause||3||3|| |
| Total||5||7|| |
|Other complications||1||1|| |
Overall blood loss (in the combined experimental and control groups) was lower in those having a vaginal hysterectomy than those having an abdominal hysterectomy (87 vs 287 mL, P < 0.01). These women also used less morphine in the first 24 hours (31 vs 42 mg, P < 0.05) and had an earlier median day of discharge day than those having an abdominal hysterectomy (4 vs 5 days, P < 0.01). There were, however, no significant differences seen in complication rate, speed of recovery, or in any of the other outcome measures.
Five women required transfusion (two in the experimental group, three in the control group). One woman in the leuprorelin acetate group suffered hypovolaemic shock post-operatively due to a bleeding ovarian pedicle. She returned to theatre for re-suturing and required four units of blood. She made an uneventful recovery following this.
There were no differences seen between the groups in the amount of morphine used in the first 24 hours or in the number of analgesic doses used on days two and three post-operatively (Table 3). The median day of discharge was day five in both groups. Two cases of wound haematoma were seen; both of these were in the control group. There was no significant difference in the incidence of febrile morbidity between the two groups. Of the seven infections in the control group, two originated from the wound and one from the chest. Of the five in the leuprorelin acetate group, one was a chest infection. There was one urinary tract infection in each group. All the other pyrexias were of unknown origin but responded to empirical antibiotic treatment.
There was a 98% response rate to the four month questionnaire. Women varied widely in the time required to return to their normal activities (Table 4). The median time taken to return to the usual state of health was less in the leuprorelin acetate group than in the control group, but this was not statistically significant (P= 0.06). There was no difference between the groups in the length of time taken to return to work or the number of days that analgesia was required. Those treated with placebo returned to sexual intercourse significantly more quickly than those treated with leuprorelin acetate (P= 0.04).
Table 4. Long term follow up of the experimental and control groups from a questionnaire survey four months after surgery. Those continuing to be affected at the time of the questionnaire were recorded as 120 days. Values are given as median no. of days (interquartile range).
|Return to ‘usual health’ after operation||37 (28–63)||64 (42–82)||0.06*|
|Use of analgesia after operation||8 (5–17)||13 (8–23)||0.28*|
|Return to work after operation||90 (74–112)||100 (90–110)||0.47*|
|Return to sexual intercourse after operation†||56 (49–90)||42 (34–51)||0.04*|
In this randomised trial, women awaiting abdominal or vaginal hysterectomy were treated for eight weeks with leuprorelin acetate or placebo prior to surgery. The uterine volume of those in the treatment group decreased by 34%. However, there was no difference between the groups with regard to intra-operative blood loss, operation time, operative difficulty or recovery time.
The primary outcome measure of this trial was operative blood loss, with the power calculations being calculated to detect a reduction of 20%. As no previous data on measured operative blood loss in women without fibroids were available, data from a trial containing women with fibroid uteri were used. In our trial we found that the median blood loss in the leuprorelin acetate group was 36% lower than in the placebo group. The range of measured blood loss, however, was greater than in the trial data used for the power calculations. This resulted in the trial being insufficiently powerful to detect significant differences. A reduction in operative blood loss could be expected as ultrasound assessment of the trial participants showed that treatment with leuprorelin acetate resulted in vasoconstriction of the uterine vessels10. The amount of operative blood loss does not, however, depend exclusively on uterine blood flow, especially in abdominal hysterectomies where blood loss also occurs in areas which are not so sensitive to oestrogen, like the abdominal wall. Furthermore, the amount of blood lost from larger bleeding vessels is dependent primarily on the rapidity of suturing (or resuturing), with the uterine blood flow being of secondary importance. The results of this trial may in fact overestimate the effect of leuprorelin acetate on operative blood loss as there were more women having vaginal hysterectomies in the leuprorelin acetate group. This will have contributed to a reduction in overall blood loss in this group as the overall blood loss in women undergoing vaginal hysterectomy was less than half those having an abdominal hysterectomy, irrespective of whether they were pre-treated with leuprorelin acetate or placebo (115 vs 285 mL, P < 0.05).
One of the advantages of using pre-operative gonadotrophin releasing hormone agonists in women with uterine fibroids is the ability to conduct the hysterectomy vaginally rather than abominally4. For hysterectomy without oophorectomy the choice between abdominal hysterectomy or vaginal hysterectomy is usually flexible, the intention being to conduct the operation vaginally if possible so as to minimise postoperative morbidity. In trials of pre-operative gonadotrophin releasing hormone agonists it is therefore important to consider vaginal and abdominal hysterectomies together in the same group. Surgeons considering suitability for vaginal surgery take into account both cervical descent and uterine size. The uteri studied, despite their lack of fibroids, were still up to 174 cm3 in volume (equivalent to over seven weeks of gestational size)11. While treatment with gonadotrophin releasing hormone agonists does not influence uterine descent, it does decrease the uterine volume by a third, an amount that can be critical when attempting a vaginal hysterectomy on a bulky uterus. In this study most women underwent the operation that had been initially planned. Of those booked for assessment on admission, however, the majority in the treatment group had a vaginal hysterectomy. The opposite was true in the placebo group where the majority had an abdominal hysterectomy, suggesting that pre-operative gonadotrophin releasing hormone agonists may favour vaginal hysterectomy in some women.
The recovery rates are similar to those reported in other randomised trials2. Surprisingly, in this trial the women having vaginal surgery recovered no faster than those having an abdominal hysterectomy. This is in contrast to other published work that suggests a more rapid recovery in those having a vaginal hysterectomy, although this has never been confirmed by a randomised trial2,12,13. The large number of women who returned to work after exactly three months and to intercourse after exactly six weeks suggests that the timing of these events is governed more by instruction than by rate of recovery. Use of analgesia and ‘time to return to usual state of health’ are more objective measures. Both were numerically less in the leuprorelin acetate group although neither achieved statistical significance. The time to return to intercourse was significantly longer in those treated with leuprorelin acetate. The reasons for this are not clear but could relate to the pre-operative vaginal dryness experienced by some women taking gonadotrophin releasing hormone agonists.
In this placebo-controlled randomised trial of women with dysfunctional uterine bleeding undergoing hysterectomy there were no significant benefits in operative morbidity with pre-operative leuprorelin acetate compared with placebo. The trial was, however, underpowered, and a larger trial in women undergoing abdominal and vaginal hysterectomy may be useful to ascertain the true place of leuprorelin acetate in the treatment of women undergoing hysterectomy.
This trial was supported by a grant from Wyeth-Lederle, UK. We wish to thank Janice Porter and Liz Nelson in the Fetal Assessment Unit for assistance with the vaginal ultrasonography. Thanks also to all the medical, nursing, and theatre staff at St James's Hospital who assisted with the running of the trial. Finally, we are grateful to the participants for their cooperation.