Subtotal abdominal hysterectomy: a surgical advance or a backward step?
Correspondence: Mr. A. A. A. Ewies, Obstetrics and Gynaecology Department, Birmingham Heartlands Hospital, Bordesely Green East, Birmingham B9 5SS, UK.
Objective To review the short and medium term outcomes of subtotal abdominal hysterectomy. We also describe the management of cervical stump complications by vaginal trachelectomy or large loop excision of the transformation zone.
Design Retrospective analysis.
Setting Warwick General Hospital, Warwickshire, UK.
Sample One hundred and fifty women underwent subtotal abdominal hysterectomy between 1993 and 1999. Five women had vaginal trachelectomy and another five had large loop excision of the transformation zone for complications relating to the cervical stump.
Results The prevalence of intra-operative and early post-operative complications was 4% and 7.3%, respectively. Twenty women (13.3%) had late complications, of whom 17 (11%) presented with symptoms directly related to the stump (two had also genuine stress incontinence). Three presented with genuine stress incontinence alone. The commonest problem was regular menstruation, which occurred in 12 women (8%). Ten of these women underwent vaginal trachelectomy or large loop excision of the transformation zone. None had intra-operative or post-operative complications.
Conclusions The high prevalences of cervical stump problems should be taken into account before a change in surgical procedure from total to subtotal hysterectomy is recommended. Further prospective studies with prolonged follow up are needed to evaluate the risks and benefits of retaining the cervix at hysterectomy. Total hysterectomy, preferably by the vaginal route, remains the procedure of choice for most women. Should a problem develop, vaginal trachelectomy or large loop excision of the transformation zone by an experienced surgeon are the best options for these women.
Hysterectomy performed in the UK usually involves removal of the uterine cervix. Owing to suggested benefits during and after surgery, it recently has been argued that total hysterectomy for benign conditions could be substituted by subtotal hysterectomy, particularly since cancers of the cervix are declining as a consequence of screening1,2.
Subtotal hysterectomy is a safer and technically easier operation than total hysterectomy. It is said to be associated with a lower incidence of ureteric damage and vesico-urethral dysfunction. Also, the utero-sacral and cardinal ligaments remain intact, thus preserving the pelvic floor support. The cervix appears to play an important role in the arousal phase of intercourse and quality of orgasm in some women3–5. Additional benefits of the subtotal hysterectomy include less peri-operative blood loss and less post-operative infection and haematoma. Also, vault granulations do not complicate subtotal hysterectomy1. The contemporary lifetime risk of cervical cancer in a monogamous woman with at least three normal Papanicolaou smears is 0.05%6.
Some women with cervices may require cervical stump removal because of prolapse, abnormal cervical cytology, cancer, pelvic pain, or recurrent bleeding or discharge. Cervical stump removal can be accomplished laparoscopically, without major complications, by an experienced surgeon7.
We report the details of 150 women treated by subtotal abdominal hysterectomy at Warwick Hospital between 1993 and 1999. The aim of the study was to review the short and medium term outcomes of subtotal abdominal hysterectomy and to present our experience with vaginal trachelectomy (five cases) and large loop excision of the transformation zone (five cases) in management of stump problems.
This study was a retrospective review of 150 women who underwent subtotal abdominal hysterectomy at Warwick General Hospital, Warwickshire, UK between 1993 and 1999. Information was derived from the Hospital Information Support System, using the Office of Population Censuses and Surveys code Q07.5 - Subtotal Hysterectomy. A further case of subtotal hysterectomy, which took place during the actual collection of information, was also included to make the final total of 150 cases for this review. A proforma was devised and case notes were retrieved. The following data were collected: women's age at the time of hysterectomy, date of operation, indication, intra-operative difficulties, intra-and post-operative complications, short or medium term sequelae, and the need for further surgery to treat stump problems. Data from the proforma were transferred to a database (D. Base III) and analysed. Early and late complications were defined as complications occurring during the first post-operative week and after the first week, respectively. Post-operative infection was defined as any serious operative site or wound infection needing treatment. Bleeding was intra-operative or post-operative blood loss > 1000 mL or that required blood transfusion.
All vaginal trachelectomies were performed by the same gynaecologist (K.S.J.O.). The procedure involved insertion of a Foley's catheter in the bladder. The cervix was grasped with two volsellum forceps and circumcised from the vagina. The bladder was then dissected from the cervix using blunt and sharp dissection. Clamping and transfixing the cardinal-uterosacral ligament complex was performed on both sides followed by anterior and posterior colpotomy, and removal of the stump. Meticulous haemostasis during these steps was essential to maintain a clear operative field. The vaginal cuff was closed with continuous locked Vicryl stitch. To achieve good suspension and prevent prolapse of the vaginal vault, the cuff was anchored to the cardinal-uterosacral ligament complex on each side. The average time for this procedure was 30 minutes. Large loop excision of the transformation zone was performed by another gynaecologist, using a standard technique as described for the colposcopic treatment of cervical intraepithelial neoplasia8. All procedures are conducted under general anaesthesia. The statistical analyses were performed using GraphPAD InStat Software (Version 1.11).
The characteristics of 150 women who had a subtotal abdominal hysterectomy are summarised in Table 1. The mean age (SD; range) was 42.8 years (8; 18–73). The commonest indications for subtotal abdominal hysterectomy were dysfunctional uterine bleeding (43%) and a uterus with fibroids (38%). In 75 of 141 elective procedures (53%), the decision to perform a subtotal abdominal hysterectomy was made in the outpatient clinic after giving the women an informed choice. Sixty women agreed to have subtotal hysterectomy in view of the suggested benefits. Conservation of the cervix was offered to 15 women because of expected surgical difficulties. The procedure was unplanned in another 40 women (28.5%), when it was difficult to remove the cervix because of severe endometriosis, pelvic adhesions, obesity, deep pelvis and/or the presence of a long cervix. Twenty-six (19.5%) women signed consent forms for ‘hysterectomy’, but it was not clear in their case notes whether subtotal abdominal hysterectomy was planned or not. Nine women (6%) in our series had an emergency subtotal abdominal hysterectomy; eight of these were due to pregnancy-related complications, and one woman presenting with an acute abdomen was found to have a large fibroid, equal to 40 weeks of gestation in size.
Table 1. The characteristics of 150 women who had a subtotal abdominal hysterectomy. Values are given as n or n (%). DUB = dysfunctional uterine bleeding.
|Age group (years)|| |
| < 30||5 (3)|
| 31–40||52 (35)|
| 41–50||65 (43)|
| > 50||28 (17)|
|Type of operation|| |
| Elective||141 (94)|
| Emergency||9 (6)|
|Decision (for elective cases)|| |
| Planned||75 (53)|
| Unplanned||40 (29)|
| Unsure*||26 (20)|
| DUB||65 (43)|
| Fibroid uterus||57 (38)|
| Endometriosis||5 (3)|
| Adnexal mass/cyst||5 (3)|
| Pelvic pain/dysmenorrhoea||5 (3)|
| Miscellaneous||4 (3)|
| Atonic haemorrhage||7 (5)|
| Cornual ectopic||1 (1)|
| Fibroid (= 40 weeks)/acute abdomen||1 (1)|
Intra-operative, short and medium term complications of subtotal abdominal hysterectomy are summarised in Table 2. The intra- and early post-operative complication rates were 4% and 7.3%, respectively. Twenty women (13.3%) had late complications, 17 of whom (11%) presented with symptoms directly related to the stump (two had also genuine stress incontinence); three presented with genuine stress incontinence alone. The commonest problem was regular menstruation, which was experienced by 12 (8%) women. Nine of these 17 women had a planned subtotal hysterectomy: seven because of suggested benefits and two because of expected surgical difficulties (one with history of three caesarean sections and the other was a known case of severe diverticular disease).
Table 2. Complications of subtotal abdominal hysterectomy. Values are given as n or n (%).
| Haemorrhage||5 (3)|
| Bladder injury||1|
|Early complications||11 (7)|
| Wound infection||4 (3)|
| Haemorrhage||3 (2)|
| Urine infection||1|
| Stump haematoma||1|
| Wound haematoma||1|
| Deep vein thrombosis||1|
|Late complications||20 (13)|
| Cyclical/ acyclical bleeding||12 (8)|
| Genuine stress incontinence||5 (3)|
| Pelvic pain||4 (3)|
| Vaginal discharge||3 (2)|
| Deep dyspareunia & reduced libido||1|
| Post-coital bleeding||1|
| Abnormal cytology||1|
The details of 10 patients (plus one on the waiting list), who had vaginal trachelectomy or large loop excision of the transformation zone for stump problems are summarised in Table 3. The mean length of time (SD; range) between the hysterectomies and vaginal trachelectomies/large loop excision of the transformation zone procedures was 16 months (15; 6–57). The mean age at trachelectomy/large loop excision of the transformation zone was 38.3 years (7.9; 28–54). Bleeding was the main indication for further surgery which required a cervical stump operation in 9/11 cases. Six of these women had a planned subtotal hysterectomy, five in view of the suggested benefits of the procedure and one because of history of three caesarean sections. None of these women had intra-operative complications. Post-operatively, one woman experienced spotting due to the formation of granulation tissue in the vaginal cuff after trachelectomy.
Table 3. Details of 10 cases who had vaginal trachelectomy or large loop excision of the transformation zone (LLETZ) after subtotal hysterectomy for stump problems. NAD = nothing abnormal was detected; HPV = human papillomavirus; GSI = genuine stress incontinence; W/L = waiting list; CIN 2 = cervical intraepithelial neoplasia; PCB = post-coital bleeding.
| || || || || ||Colposuspension||GSI|
|4||33||Planned||7||Trachelectomy||Menstruation/pelvic pain||Chronic inflammation|
|5||45||Planned||19||Trachelectomy||Menstruation/discharge/back pain||Chronic inflammation|
|6||40||Unplanned||W/L||Trachelectomy||Deep dyspareunia/reduced libido/pelvic pain|| |
|7||54||Unplanned||57||LLETZ||Moderate dyskaryosis||CIN 2|
Increased interest in subtotal hysterectomy by doctors and the general public reflects the recent trend towards more conservative pelvic surgical procedures such as endometrial ablation. However, it is the responsibility of those who promote any new surgical procedure to prove that it is better than the one it is to replace. The primary reason advanced for subtotal hysterectomy is to retain normal sexual function, but there are few convincing data to show that total hysterectomy is related to long term sexual dysfunction. In studies claiming an advantage for subtotal hysterectomy, criteria used to evaluate sexual satisfaction are inconsistent and outcome values are not well defined, and one component sometimes worsens while another improves. The preponderance of evidence suggests that detrimental effects on sexual function are rare with both total and subtotal hysterectomy and pre-operative sexual activity appears to be the most important factor in predicting post-operative sexual satisfaction. The data are equally uncertain regarding post-operative bladder symptoms, and they are inadequate to evaluate all other proposed advantages of subtotal hysterectomy6.
There are few published data on adverse effects of subtotal hysterectomy, but subsequent trachelectomy has been necessary in some patients because of continued bleeding, intraepithelial neoplasia, or prolapse of the cervical stump. Van der Stege and van Beek4 reported that 10% of women who had laparoscopic supracervical hysterectomy suffered from discharge and 25% continued to menstruate (two were treated by vaginal cervical stumpectomy). In previous reports, 10% of women who had laparoscopic supracervical hysterectomy experienced menstrual bleeding9. Women who have undergone subtotal hysterectomy require regular screening conforming to the established guidelines for cervical cancer prevention. This would mean additional expense for abnormal smears6. Ensuring that women who undergo subtotal abdominal hysterectomy have had no previous cytological abnormalities and appropriate screening subsequently should effectively reduce the possibilities of the development of cervical carcinoma to almost zero in these women. Routine cauterisation of the endocervical canal and/or excision of the endocervix by ‘reverse conisation’, should minimise the problems of vaginal bleeding, which was the main disadvantage of subtotal hysterectomy in this series. However, where residual endometrium exists there is the potential for continued menses, and it is possible that endometrial carcinoma may develop, particularly if unopposed oestrogen is used for hormone replacement.
The vaginal approach to trachelectomy is usually preferable to an abdominal or laparoscopic one because the tissue planes approached vaginally are devoid of scar tissue, and the initial dissection is much easier. However, this technique does not allow treatment of endometriosis or removal of the ovaries. The major concern related to the resection of a severely adherent cervix is the high risk of injury to the ureter, rectum, or bladder. In a report of six cases of laparoscopic trachelectomy, the mean duration of the procedure was 225 minutes (range 155–320 minutes). This time was largely dependent on the treatment of associated adhesions and endometriosis. The actual removal of the stump took ≤ 25 minutes. In our series the mean time of vaginal trachelectomy was 30 minutes and that of large loop excision of the transformation zone was 10 minutes. None of our vaginal trach-electomy/large loop excision of the transformation zone patients had intra-operative complications. Post-operatively, one woman experienced spotting due to the formation of granulation tissue in the vaginal cuff after trachelectomy. In view of the available data, there is little to choose between vaginal trachelectomy and large loop excision of the transformation zone in treating stump symptoms. Nevertheless, trachelectomy is probably a better option for women with pelvic pain.
In summary, the alleged sexual and genitourinary benefits of subtotal hysterectomy over total hysterectomy are not proven, and well-designed prospective trials are needed to address these issues. Women should understand that a second operation may be required to remove the cervix mainly because of continued menstruation and pelvic pain. The high rate of stump problems (11%) should be made clear to women before making the decision on whether to carry out total abdominal hysterectomy or subtotal abdominal hysterectomy. Whichever option the woman chooses, discussing her preferences is likely to increase her satisfaction with the operation. Further prospective studies with prolonged follow up are needed to evaluate the risks and benefits of retaining the cervix at hysterectomy. Total hysterectomy, preferably by the vaginal route, is better studied with longer follow up, is less costly in the long term, and therefore remains the procedure of choice for most women6,7. Should a problem develop, vaginal trachelectomy or large loop excision of the transformation zone by an experienced surgeon are probably the best options for these women.
The authors would like to thank Mrs. R. I. Walczak-Doughty, Audit Analyst, Clinical Effectiveness Department, Warwick Hospital for her help in collecting and analysing the data.