Feasibility of a Trial of Laparoscopic Hysterectomy Versus Laparoscopic Sub-Total Hysterectomy

Overview

Observational studies suggest faster recovery and quicker return to normal activities following laparoscopic supracervical hysterectomy (LSH which involves removal of the body of the womb but conservation of the neck of the womb or cervix) compared with laparoscopic hysterectomy (LH in which both the body and neck of the womb are removed) in women with benign uterine disease. Data from the only randomised controlled trial (RCT) on the topic does not support this observation. The investigators set out to investigate the feasibility of a double blind RCT comparing post-operative recovery following LH with that following LSH. The investigators set out to recruit 50 participants from a single gynaecological surgeon's caseload. Web based randomisation was carried out at the time of study laparoscopy. Surgery was performed using a standardised technique. Participants and the data handler were blinded to treatment allocation (double blind study where neither the patient nor the surgeon have knowledge of treatment group allocation at the time of questionnaire analysis). Primary outcome is feasibility of recruitment to the study. The null hypothesis for this study is that 'a double blind, randomised controlled comparison of recovery following laparoscopic hysterectomy with recovery following laparoscopic sub-total hysterectomy' is not feasible. Secondary outcomes included validated post-operative recovery and mood questionnaires at baseline, prior to discharge and at weekly intervals for 12 weeks. Validated questionnaires regarding pelvic floor function and sexual function will be assessed at baseline, 6 weeks and 6 months. The findings of the feasibility study will inform the power calculation for a planned definitive study ( the magnitude of the differences found between the 2 arms of the feasibility study will allow calculation of the total number of participants required in a definitive study to allow demonstration of statistically significant differences in outcomes).

Hysterectomy remains the last resort for a number of benign conditions which have failed to respond to conservative measures. Hysterectomy is also performed for operable gynaecological malignancy which is excluded from the present study. Conventional techniques for uterine extirpation involve open surgery through an abdominal or vaginal incision. In the laparoscopic approach small incisions are made in the abdomen to allow insertion of a laparoscope and operative instruments. In Laparoscopic Sub-total Hysterectomy (LSH) the entire procedure is performed laparoscopically, the cervix is left in situ and the specimen is removed through an abdominal port (See Appendix I for classification of laparoscopic hysterectomy). In Laparoscopic Hysterectomy (LH) part of the operation is performed laparoscopically, up to and including division of the uterine vessels, but the specimen is removed with the cervix through an open vaginal vault. (See Appendix II for detailed description of equipment and technique). Popularity of hysterectomy shows a geographical distribution from a high of 5.4 per thousand women in the USA (Farquhar 2002) to a nadir of 1.2/1000 in Norway (Backe et al 1993). Despite the high prevalence, how and in whom hysterectomy should be undertaken remains controversial. A Cochrane systematic review of methods of hysterectomy concluded that fewest complications occur with vaginal hysterectomy (VH) whereas abdominal hysterectomy (AH) carries the highest risk (Johnson et al 2006). The meta-analysis included 27 RCTs with a total of 3643 participants comparing one surgical approach of hysterectomy to another. The authors concluded that VH should be the approach of choice for benign indications as it is a fast, relatively cheap technique with rapid recovery and a low complication rate. LH should be offered when VH is not practicable. Despite this grade 1 evidence most hysterectomies worldwide are performed as total abdominal hysterectomies (TAH) when the whole uterus with attached cervix is removed through an abdominal incision (Garry 2005). Prior to the introduction of population cervical screening it remained routine practice to remove the non-diseased cervix at the time of hysterectomy to prevent cancer of the cervical stump (Munro 1997). Total hysterectomy (TH) was therefore standard and sub-total hysterectomy (STH) was usually reserved for those rare clinical circumstances when removal of the cervix was deemed unduly difficult or risky for the patient. A number of retrospective studies have shown reduced operating time, reduced ureteric complications and reduced blood loss with STH (Kilkku et al 1983, Kilkku 1985, Brown et al 2000, Lalos and Bjerle1986). It has been postulated that TH leads to more extensive disruption of pelvic innervation than STH with accompanied increased risk of poor bladder, bowel and sexual function. However, RCTs have failed to confirm this hypothesis (Thakar et al 2002, Gimbel et al 2003). Indeed a recent Cochrane systematic review has shown no difference in the rates of continence, constipation or measures of sexual function between STH and TH (Lethaby 2006). Why findings of the RCTs do not match observational data is unclear. Perhaps the tools used for assessment of the outcome measures studied are not sufficiently sensitive or have been applied too infrequently to detect a clinically important difference. Total laparoscopic hysterectomy (TLH) was introduced by Harry Reich (Reich et al 1989) and was modified into laparoscopic assisted vaginal hysterectomy (LAVH) with a sub-total variation, LSH, introduced by the minimal access pioneer Kurt Semm in 1991 (Mettler et al 1997). Only 1 randomised comparison on the subject has been published (Morreli et al 2007) with no significant differences found on comparison of LSH and TLH with respect to complications or post-operative recovery. Case series suggest recovery from LSH occurs much earlier with return to normal quoted from within 3-21 days (Lyons 1997, Jugnet et al 2001, Schwartz 1994, Zupi et al 2003 and Donnez \& Nisolle 1993). It follows that in order to detect a difference in recovery between the 2 treatment groups frequent assessments are necessary during the first 12 weeks following surgery. Post-operative recovery will be assessed in participants of the current study at 6 hours after surgery, at the time of discharge and at weekly intervals until 12 weeks from surgery and at study completion using a validated questionnaire (QoR-40 Myles 2000 Appendix VI). In a recent systematic review of 'Recovery Specific Quality of Life Instruments' (Kluivers et al 2008) the QoR-40 was the best validated tool with regards to short term post operative recovery. A second questionnaire will be completed at the same times to assess symptoms of depression (Center for Epidemiologic Studies Depression Scale CES-D Appendix VII). With regards to urinary, bowel and sexual symptoms a 3rd validated questionnaire, the e-PAQ version 10 (Radley 2007 Appendix VIII) will be completed at recruitment, at the 6 week follow-up visit and at 6 months. To our knowledge there has been no published RCT comparing LSH with LH. The approach of choice for hysterectomy for benign indications is vaginal. When VH is not practicable LH should be offered but there is paucity of evidence with regards to which type of laparoscopic hysterectomy should be offered. The feasibility study outlined is designed to assess whether it is realistic and practicable to recruit to an RCT aimed at addressing this enormous gap in the literature. Specific aims of the feasibility study are: 1 To determine the rate of recruitment to a randomised controlled trial comparing short term outcomes following laparoscopic hysterectomy versus laparoscopic sub-total hysterectomy 2 To assess the validated questionnaires used in discerning differences between the groups undergoing the two types of laparoscopic hysterectomy. 3 To provide pilot data to inform power calculations for the full RCT. 4 To determine the acceptability to randomisation. 5 To identify potential obstacles to recruitment to this study in our centre. Aims of the proposed definitive study are outlined in order justify the tools used. 1. To test the hypothesis that LSH is associated with an accelerated recovery and return to normal activity when compared with LH. 2. To test the hypothesis that LSH leads to a decreased incidence of postoperative troublesome urinary symptoms compared with LH. 3. To test the hypothesis that LSH leads to a decreased incidence of postoperative troublesome bowel symptoms compared with LH. 4. To test the hypothesis that LSH leads to a lesser deterioration of sexual function postoperatively compared with LH. 6\] To test the hypothesis that LSH is associated with fewer symptoms of depression when compared with LH. Women in whom hysterectomy is indicated for a benign condition and in whom a laparoscopic hysterectomy is feasible will be invited to take part in the study. The feasibility study sample will be invited from the current caseload at the pilot hospital (Author's main hospital of practice). All invitees will be provided with written information and informed consent obtained. Women who express their willingness will be assessed for their eligibility according to the criteria below. Eligible invitees will include all premenopausal women in whom hysterectomy is indicated for benign indications and in whom either technique is considered feasible after initial laparoscopic survey. Exclusion criteria were predefined as post menopausal status, suspected malignancy, abnormal cervical smears or a history of treatment of a high grade cervical intraepithelial neoplasia within 10 years of recruitment, uterine greater than 14 weeks size, raised BMI (Body Mass Index) \> 35Kg/M2, uterovaginal prolapse \> Stage 1 (Bump et al 1996, previous gynaecological malignancy, previous extensive pelvic surgery, current psychological/psychiatric disease, current anti-depressant/anti-psychotic therapy or inability to read and write English. Randomisation will be conducted by a web based computer programme to generate sealed opaque envelopes (www.sealedenvelope.com/freerandmiser/v1) using block randomization to allow periodic balancing of the 2 treatment groups. The opaque envelopes will be in a locked cabinet in the charge of an administrator with no connection to the trial. The envelope will be opened after initial laparoscopic survey confirms that either technique is feasible. The participants will not be informed of their treatment group allocation until 1 year after entry into the feasibility study. Treatment group allocation may be divulged early if required by the participants General Practitioner or another clinician to facilitate the participants care. Randomised women will receive laparoscopic hysterectomy (LH) or laparoscopic supracervical hysterectomy using a standardised technique There are numerous descriptions for LSH and LH in the literature. The procedures will be carried out according to the usual practise of the principal investigator. A large part of both procedures are the same and will be described first. General anaesthesia is induced and maintained by a single anaesthetist. A 5mm primary bladeless, dilating tip trocar (Excel™, Ethicon, Cincinnati, USA) is inserted into the umbilicus under direct vision. Once intraperitoneal placement is confirmed a carbon dioxide pneumoperitoneum is established at a pressure of 15mmHg. The patient is placed in 45° Trendelenberg and 3 accessory 5mm ports are inserted under direct vision. The position of the ports are lateral at the level of the umbilicus on either side and the third 2 fingers breadth above the symphysis pubis. In the case of LH all 3 accessory ports are 5 mm diameter but in LSH the suprapubic port is 12mm diameter. Vapor pulse coagulation via a single use Plasma Kinetic forceps (PK™ Gyrus ACMI PLC) is used to coagulate the utero-ovarian ligament or the infundibulopelvic ligament if removal of the ovary is desired, after identification of the position of the ureter at the pelvic brim. The uterovesical fold of peritoneum is dissected with scissors and monopolar diathermy at 50W coagulation. The bladder flap is reflected caudad. A posterior peritoneal flap is similarly dissected and reflected to mobilise the ureters caudad. The ascending branch(es) of the uterine arteries are coagulated and transected with the PK forceps. The remainder of the procedure differs according to treatment allocation. For LH the main trunk of the uterine artery is coagulated superomedially to the course of the ureter. The uterosacral ligaments are coagulated and transected close to the arch of the uterosacrals. The remainder of the procedure is carried out via the vagina. Forty mls of a 1:200,000 adrenaline in normal saline solution is infiltrated into the subcuticular space around the cervix taking care to avoid intravascular injection The cervix is circumcised with a monopolar angled needle using blended diathermy of 100W cutting and 60W coagulation. The supra-vaginal septum is divided to allow cranial reflection of the bladder. The Pouch of Douglas is entered with a single cut into the peritoneum underlying the vaginal skin in the posterior vaginal fornix. The Cardinal ligaments are clamped with curved Zeppelin clamps, cut,transfixed and ligated with 1 Vicryl® (Ethicon division of Johnson \& Johnson, New Jersey, USA) prior to removal of the uterus and cervix. The vaginal skin and peritoneum are sutured en-masse with locking 1 Vicryl®. In the case of LSH a 100W monopolar diathermy loop (Lap Loop™, Roberts Surgical, Worcester, UK) is used to amputate the body of the uterus from the cervix. The uterus is morcellated and removed piecemeal with a 15mm disposable morcellator (X-Tract™ Gynecare division of Johnson \& Johnson, New Jersey, USA). The endocervical canal and cervical stump are coagulated with bipolar diathermy. The accessory ports are removed under direct vision and in the case of LSH the suprapubic port is closed with 1 Vicryl® on Endoclose® (Tyco, Hampshire UK). The remainder of the procedure is the same for both techniques. Haemostasis is confirmed laparoscopically.The pneumoperitoneum is deflated and 20 mls 2.5% chirocaine is instilled intraperitoneally. The laparoscope and primary trocar are removed under direct vision. In LSH the 15mm suprapubic port site is closed with 1 Vicryl® on an Endoclose™ (Tyco Healthcare, Hampshire UK) needle. The skin is closed with subcuticular 3-0 Vicryl Rapide® (Ethicon). 5mls 2.5% chirocaine is injected subcuticularly at the port sites. The bladder catheter is removed in all patients when the patient is mobile. Analgaesia is commenced per-operatively with intravenous diclofenac 75 mg and paracetamol 1g together with 8 mg dexamethasone and 4mg ondansetron as antiemetic. Questionnaires will be administered without prior knowledge of type of hysterectomy. The operating surgeon will not be involved with administering questionnaires for the duration of the study. Analysis of the results will be performed without prior knowledge of treatment group allocation. Participants and their GPs will be informed of treatment group at the completion of the feasibility study (or sooner if requested by GP or participant who wishes to drop out of the study) in order to continue on the cervical screening programme if required (LSH group). The feasibility study will describe recruitment to the study and acceptability to randomisation. Statistical analysis of the feasibility study results will inform power calculations for the subsequent definitive study. All statistical analyses will be performed using SPSS version 17 for Windows.. The QoR-40 questionnaire domains are scored on a scale from 0 (worst recovery) to 200 (best recovery). The scores are compared preoperatively, postoperatively and at weekly intervals up to and including 6 months using the t-test where data satisfies the test of normality. Normality will be tested using the Kolmogorov-Smirnov test (P value set at 0.050) and where data does not satisfy normality the Wilcoxon Signed Rank Test will be used. The CES-D questionnaire elements are scored as 0-3 with a maximum score of 60. A score of over 16 indicates depression. The CES-D scores will be converted to presence or absence of depression and compared between the 2 groups using the Wilcoxon Signed Rank test. The e-PAQ has 4 domains. 3 domains have 5 items and 1 domain has 4 items. Each item is scored from a minimum of 0 (best possible health) to 100 (worst possible health. The preoperative item scores will be compared with postoperative scores and at six months using the Wilcoxon Signed Rank test. The primary outcome measures scores will also be analysed as above.