- Top of page
Objective To analyse the fertility rates, complications and recurrences in a group of women who have undergone radical vaginal trachelectomy and pelvic lymphadenectomy for early-stage cervical cancer.
Design An observational series.
Setting A Gynaecological Oncology Centre.
Population One hundred and twenty-three consecutive women who underwent radical vaginal trachelectomy and pelvic lymphadenectomy for early-stage cervical cancer.
Methods Data were collected prospectively.
Main outcome measures Complications, recurrences, pregnancies and live births are presented as percentages of the total population. Fertility is presented as a 5-year cumulative rate, with women attempting to conceive as the denominator.
Results A total of 123 women were followed up for an average of 45 months. Eleven (8.9%) had completion treatment (two radical hysterectomies and nine chemoradiotherapy) at the time of initial treatment. There were three recurrences (2.7%) among the women who did not have completion treatment and two (18.2%) in those who did. There were 6 perioperative and 26 postoperative complications. Sixty-three women attempted pregnancy. There were 55 pregnancies in 26 women and 28 live births in 19. Three women had continuing pregnancies. The 5-year cumulative pregnancy rate among women trying to conceive was 52.8%. All but two women were delivered by classical caesarean section and seven (25.0%) babies were born at 31 + 6 weeks or less.
Conclusions For selected women with early-stage cervical cancer, radical vaginal trachelectomy and pelvic lymphadenectomy are fertility-sparing options, with a low incidence of recurrence and acceptable cumulative conception rates. Complications are few, although there is a high premature labour and miscarriage rate among pregnant women.
- Top of page
Early-stage cervical cancer often presents in women of fertile age.1 Traditional treatment is radical hysterectomy with pelvic lymphadenectomy, or chemoradiotherapy. The 5-year survival following radical hysterectomy for lymph node-negative early cervical cancer is 95% for stage Ib1 disease and 99% for stage Ia2 tumours.2 Early reports suggested that radical trachelectomy is as efficacious for the treatment of early cervical cancer and maintains the potential for preservation of fertility.3–11 Some authors have now published extended series that support the original data.12–16
Pregnancies in most series are expressed as total numbers and do not take into account the women using contraception or not trying to conceive. Many delay having children and others never attempt to become pregnant. We report the cumulative pregnancy rates based on women wishing to conceive as a denominator and include population, surgical and oncological data from the largest series of radical vaginal trachelectomies reported to date. Preliminary data on the first 30 women included in this series have been reported previously in a short communication.4
- Top of page
One hundred and twenty-three consecutive women who selected to have radical vaginal trachelectomy with pelvic lymphadenectomy between August 1994 and 2005 were included in the series. All women had early-stage cervical cancer and would have otherwise had a radical hysterectomy.
The protocol included careful preoperative counselling, formal staging by examination under anaesthetic (EUA), magnetic resonance imaging (MRI) of the pelvis and histopathological review of the tissue from the referring institution. The length of the cervix and the depth of the uterine cavity are measured carefully both at EUA and MRI to ensure that the correct amount of tissue is removed at surgery. The technique has been described previously3 and is only summarised here. The procedure commences with a laparoscopic pelvic lymphadenectomy. A 10-mm trocar is placed in the umbilicus for the laparoscope to pass. Secondary 5-mm ports are placed high in the right and left iliac fossae after identification of the epigastric vessels. A high suprapubic 10-mm port is also placed and used for retrieval of lymphatic tissue. Lymph node dissection commences with skeletalisation of the obturator nerve and retrieval of the obturator and internal iliac nodes. External iliac nodes are retrieved from the circumflex iliac vessels just above the bifurcation of the common iliac artery. Care is taken to ensure preservation of the genitofemoral nerve. Ligaclips and diathermy are used for haemostasis, and more recently, a harmonic scalpel has been introduced. After the laparoscopic part of the procedure is performed, the woman is placed in an extended lithotomy position and the radical trachelectomy is performed vaginally. A 1- to 2-cm cuff of vagina is mobilised so that it can cover the cervix in an anterior–posterior manner using Krobach clamps. The bladder is mobilised and the paravesical space entered. The ureter is palpated, bladder pillars are divided and the ureters are mobilised away. Further parametrial tissue is excised by ensuring lateral division of cardinal and uterosacral ligaments. The cervix is then amputated at the level of the internal os. The isthmus is identified by recognising the vesicocervical reflection of peritoneum, although care is taken to avoid opening this. A 1 nylon cerclage suture is then placed around the isthmus for support in future pregnancies. It is for this reason and because the lower segment has been excised during the operation that women require a classical caesarean section or a low vertical incision for delivery of subsequent children. The vagina is reanastomosed to the isthmus at the end of the procedure. Indications for completion treatment include lymph node involvement or incomplete resection margins after the final pathology has been reviewed.
All women are followed up once in 3 months for the first year, once in 4 months for the second year, once in 6 months for the third, fourth and fifth years and annually thereafter, with discharge to the GP at the tenth year. At each follow-up appointment, reported complications are noted and vaginal vault and isthmic smears are taken using a spatula and endocervical brush. Colposcopy is performed at 6 and 18 months and a follow-up pelvic MRI scan is performed at 6, 12 and 24 months. Women are asked to use contraception until 6 months postsurgery.
Demographic, pathological, surgical and follow-up data are recorded prospectively onto a database at each follow-up visit and the women are asked whether or not they were trying to conceive. When a woman reported that she was not trying to conceive, this is noted on the database. Cumulative pregnancy rates are reported in an actuarial manner using the Kaplan–Meier method.
- Top of page
A total of 123 women underwent the procedure. The age, stage, parity and histological characteristics are shown in Table 1. The mean follow-up period was 45 months (SD 32 months, range 1–120 months).
Table 1. Demographic details of 123 women selected for radical trachelectomy
|Age, mean years (SD; range)||30.6 (4.3; 21–45)|
|Parity, n (%)|
|Stage, n (%)|
|Lymph-vascular space invasion, n (%)||39 (31.7)|
|Histological type, n (%)|
|Grade, n (%)|
|No residual cancer on specimen, n (%)||77 (62.6)|
|No residual dysplasia on specimen, n (%)||62 (50.4)|
|Lymph node count, mean (SD; range)||19.4 (9.1; 3–51)|
|Positive pelvic lymph nodes, n (%)||7 (5.7)|
Eleven women (8.9%) had completion treatment. Two had completion surgery and nine had chemoradiotherapy. Three of the women who went on to have chemoradiotherapy had their surgery abandoned perioperatively due to the presence of positive lymph nodes on frozen section. The other six women who had completion chemoradiotherapy had positive lymph nodes on paraffin section (three women), had close parametrial margins (two women) or had an involved isthmic margin (one women). Both the women who had completion surgery had radical hysterectomies for close endocervical margins. An additional woman with a single microscopic, intracapsular lymph node metastases declined completion treatment and is disease free 10 years later.
Of the 11 women who had completion treatment, 2 (18.2%) developed recurrent disease and both have died. One woman was 41 years old and nulliparous. She had completion chemoradiotherapy following positive nodes on paraffin section. The disease recurred on the pelvic sidewall at 15 months and the woman died at 26 months. The other woman was 28 years old and previously had one child, which was given for adoption. Her surgery was abandoned perioperatively due to positive lymph nodes and she received chemoradiotherapy. She had recurrent disease at 21 months and died of the disease at 32 months.
Of the 112 women who did not have completion treatment, 3 (3.3%) have had recurrent disease. One woman was nulliparous and 34 years old at the time of radical trachelectomy. She had a pelvic sidewall recurrence at 7 years following treatment. In the intervening time, she had three children. She had radical chemoradiotherapy for her recurrence and is currently disease free 45 months following her second-line treatment. Another woman was 32 years old at the time of trachelectomy and already had one child. She had a central recurrence at 19 months and died of disease at 32 months. The third woman was 36 years old and had two previous children. She had a trachelectomy for a poorly differentiated squamous cell carcinoma with lymph-vascular space invasion. She developed a 1.5-cm recurrence in the pelvis just lateral to the isthmus at 31 months. She received chemoradiotherapy and died of disease at 26 months after the diagnosis of recurrence.
The complications during and following surgery are summarised in Table 2. There were six perioperative complications in six (4.9%) women. These included a uterine perforation in a woman who had a concurrent hysteroscopic resection for a uterine septum. The woman has since succeeded in achieving a pregnancy. Other intraoperative complications included a woman who required a laparotomy and primary repair of an external iliac artery injury and two women who required blood transfusion following pelvic sidewall bleeding and bleeding from a lacerated vaginal fornix, respectively (Table 2). There were 26 postoperative complications in 24 (19.5%) women (Table 2).
Table 2. Complications
|External iliac artery rupture||1|
|Vaginal fornix laceration||1|
|Uterine perforation during uterine septum resection||1|
|Failure of pneumoperitoneum||1|
|Laparotomy for pelvic sidewall bleeding||1|
|Temporary thigh numbness||5|
|Mild leg oedema||1|
|Infected epidural site||1|
|Urinary tract infection||1|
There were 55 pregnancies in 26 of 63 (41.3%) women who attempted to conceive. The cumulative pregnancy rate by month of trying to conceive in women who did not have completion treatment is shown in Figure 1. The 5-year percent probability of conception is 52.8%. Of the 55 pregnancies, 28 (50.9%) were live births and these occurred in 19 women. When discounting two pregnancies that were terminated and three that were continuing, then 56.0% of pregnancies resulted in the birth of a child (Table 3).
Figure 1. Cumulative actuarial probability of conception adjusted for reported contraceptive usage among women who are trying to conceive following successful radical trachelectomy for early cervical cancer.
Download figure to PowerPoint
Table 3. Pregnancies
|Live births||28 in 19 women|
|Stillbirths||0 in 0 women|
|Neonatal deaths||0 in 0 women|
|Miscarriages <14/40||14 in 9 women|
|Miscarriage ≥14/40||7 in 3 women|
|Terminations||2 in 2 women|
|Ectopic pregnancies||1 in 1 woman|
|Continuing pregnancies||3 in 3 women|
|Surrogate live births||1 in 1 woman|
A total of 18 women had some form of fertility treatment. Of these women, three have had a total of four live births and a further three have conceived and not carried to term. In addition, one woman achieved a genetic offspring, with her mother as a surrogate. The cause of infertility for all these women is unknown in most cases. However, three women are known to have partners with oligospermia.
All children were born by caesarean section. A low classical incision was used in all cases except two, wherein transverse incisions were made in the lower uterine body. Fifteen (53.6%) were born at 36 weeks of gestation or more (Table 4). Six (21.4%) were born between 32 and 35 + 6 weeks of gestation (Table 4). Seven (25.0%) were born at 31 + 6 weeks of gestation or less (Table 4).
Table 4. Gestation of births following radical trachelectomy
- Top of page
This data demonstrate that radical trachelectomy can be performed safely. The complication rates are low compared with those for radical hysterectomy17 as are the recurrence rates.17 The data are comparable with other published series, with five (4.1%) recurrences and four (3.3%) deaths in 123 women who were selected to have trachelectomy (Table 5). Excluding those who had abandoned procedures or completion treatment, there were three (2.7%) recurrences and two (1.8%) deaths in 112 women. Covens et al.11 also compared a group of women who had trachelectomy with another group of women who had radical hysterectomy and found no difference in disease-free survival. On that evidence,11 we do not advocate hysterectomy following completion of a family. In the current series, one woman had a recurrence 7 years following surgery and after completion of her family. However, this recurrence occurred on the pelvic sidewall and would not have been prevented by completion hysterectomy.
Table 5. Recurrences and deaths for women in published series of trachelectomy
|Authors||n||Mean follow up (months)||Recurrences, n (%)||Deaths, n (%)|
|Current series||123||45||5 (4.1)||4 (3.3)|
|Dargent et al.9,10||95||76||4 (4.2)||3 (3.1)|
|Steed and Covens,15 Covens16||93||30||7 (7.3)||4 (4.2)|
|Plante et al.12||72||60||2 (2.8)||1 (1.4)|
|Burnett et al.7||19||31||0 (0.0)||0 (0.0)|
|Schlaerth et al.8||10||47||0 (0.0)||0 (0.0)|
|Total||412||51||18 (4.4)||12 (2.9)|
The protocols used for the selection and follow up of women were largely empirical. MRI was used as it is the most effective modality at determining parametrial disease18 and also help identify endocervical involvement and suitability for trachelectomy.18 Isthmic surveillance using cytology can produce confusing results,19,20 which is why a colposcopic assessment was also used for follow up. The reliability of cytology is greater if assessed by an experienced cytopathologist.19,20 The isthmus can be difficult to locate on clinical examination following trachelectomy. This has led to one group to modify the technique of vaginal reconstruction to leave the raw area of the uterus from the excised cervix exposed.12 It may be possible with some squamous lesions to try and conserve a portion of the upper cervix at the internal os if margins of excision can be guaranteed. This would probably help reduce the risk of prematurity. However, careful assessment by MRI and intraoperative frozen section of the endocervical margin is necessary and this would only be suitable for exophytic lesions.
The problem of complete isthmic stenosis has already been reported21 and has resulted in either a haematometra or further surgery in four women in this series. In an attempt to reduce this complication, the authors now place a urinary catheter through the isthmus as a stent and leave this in place for 3 days postoperatively. The complication rates are favourable when compared with those for radical vaginal or abdominal hysterectomy 17,22 and there have not been any long-term urinary or bowel problems as a result of trachelectomy in this series.
The complication of greatest concern is that of premature labour (Table 4). It is because of this that women are counselled to book their pregnancy in a unit with level 3 neonatal intensive care available on site. Furthermore, as part of preoperative counselling, women are informed of both the risk of extreme prematurity during pregnancy and the possible neurological consequences to a child if this were to happen. Caesarean sections must be performed through a classical incision because of the lack of a lower segment following isthmic excision. A transverse incision risks the extension, tearing the uterine vessels.
Pregnancy rates, when taking into account contraceptive usage, are good (cumulative probability of conception at 5 years is 52.8%) and comparable with fertility rates following surgery for an ectopic pregnancy.23–25 However, it should be noted that second trimester miscarriage occurred in 7 of 55 pregnancies (12.7%) in addition to the seven babies born prematurely before 32 weeks of gestation. Because of this, it may be of value to have ultrasound assessments of isthmic competence in early pregnancy. Only 19 (30%) of the 63 women who were not known to be using contraception had a live birth.
The success of radical vaginal trachelectomy has led some authors to hypothesise about the possibility of an even more conservative approach, with some women selected for conisation alone.26 It is now our policy to offer a large conisation with pelvic lymphadenectomy to selected women with stage Ia2 disease. In this series, 62.6% of women had no residual disease on the final specimen. This is consistent with data from Plante et al.12 Covens et al.27 reported that the incidence of parametrial involvement in women with a tumour size of less than 20 mm and with a depth of invasion of less than 10 mm was 0.6%. However, in this series, there were two women who had completion treatment for parametrial involvement and a further woman who developed a parametrial recurrence. A further six women had lymph node involvement. Although some women have been followed up for 10 years, the median follow up is 45 months and some women have only been followed up for a few months at the time of publication. One woman had a recurrence at 7 years. Most women in this series have not reached 7-year follow up, and until a longer follow-up period has been reached, we will not know the true effectiveness of the procedure. More conservative surgery may be associated with a higher live birth rate, a lower prematurity rate and less complications. However, there is insufficient evidence to support a less radical approach for low-volume stage Ibi disease, and in the opinion of the authors, more conservative surgery should be performed only as part of a study.
Further studies and long-term follow up of this and other published series may provide additional information on long-term morbidity to children born following radical vaginal trachelectomy. In addition, more detailed information may become available from urodynamic studies and in-depth assessments of menstruation and sexual function. However, this study demonstrates that radical vaginal trachelectomy is a fertility-sparing option in women with early-stage cervical cancer. The operation has an acceptable complication profile but is associated with a significant incidence of miscarriage and premature labour.