We applaud Naik et al.1 for undertaking an RCT in the surgical treatment of early-stage cervical cancer. Clinical recruitment for such an infrequent disease is always problematic and this is reflected in the small numbers of women in this study.
We agree that adequate vaginal access is essential for a radical vaginal approach, but question why cervical defects caused by previous large loop excision of the transformation zone (LLETZ) were an exclusion criterion to the study. Almost all women with invasive cervical cancer considered for definitive surgery will have had at least one loop excision, and often two may have been performed. In the Coelio–Schuata laparoscopic assisted radical vaginal hysterectomy (LARVH) technique the cervix is not grasped directly during the operation, instead a vaginal envelope is created, which is used to manipulate the cervix indirectly. Adequate vaginal access is therefore more relevant than number of previous LLETZ.
We agree with the authors that LARVH is associated with shorter hospital stay and faster recovery with probable reduced morbidity overall. However, we are surprised that duration of catheterisation was used as primary outcome measure because their protocol directs that the Foley catheter used in each arm of the study is removed on a different day.
One of the advantages of learning to perform LARVH is that it enables one to perform radical vaginal trachelectomy (RVT) for women to whom radical hysterectomy (RH) would otherwise be offered, but where the woman wishes to maintain fertility choices. In our centre RVT is performed much less frequently than RH, and we have found performing LARVH to be important in maintaining the necessary surgical skills for this operation. Vaginal steps in both procedures are essentially the same, but the uterine corpus is conserved in RVT whereas in LARVH, the uterine artery is divided at source laparoscopically. In our experience, confidence in vaginal palpation and visualisation of bladder pillars and the lower ureter is key to LARVH and RVT, enabling the surgeon to decide how wide a paracervical resection is required.
Our case–control study compared 14 women undergoing LARVH with 12 women undergoing RAH and suggested greater postoperative urinary tract dysfunction following RAH than after LARVH, with shorter hospital stay after LARVH (4.4 days [LARVH] versus 7.9 days [RAH]).2 Intraoperative injuries were higher in the LARVH group early in the learning curve but this is not reflected in our larger series (manuscript in preparation).
We have found LARVH to be a very acceptable treatment for women with lower risk tumours suitable for a single-stage radical hysterectomy with concomitant pelvic lymphadenectomy where fertility conservation is not desired. In the majority of women this is a Type II RH, but it is feasible to perform a Type III procedure transvaginally.3
We adopt Dargent’s two-stage approach4 for women with adverse histological features, large tumours >2 cm diameter, or enlarged nodes on pelvic magnetic resonance imaging, i.e. an initial pelvic lymphadenectomy followed by an interval Type III RH if lymph nodes are negative.
The authors concur with previous studies that a total laparoscopic approach to RH may be the way to avoid difficulties. This may be true, but may merely present a different set of problems and complications, particularly with regard to ureteric injury and risk of urinary fistula.
Recently published studies in Europe and USA suggest that robotic radical hysterectomy is a safe technique with decreased blood loss, operating time and minimal complications intraoperatively and may be applicable to a wider population than may be achieved with a conventional laparoscopy.5
We are one of the first centres in the UK to actively explore the potential of robotics in radical pelvic gynaecological surgery, performing robotic surgery for endometrial and early cervical cancer since late 2009. Prospective data are required to evaluate this but our initial experience is very promising. Comparative studies with a laparoscopic/laparo-vaginal approach are required.