Apart from cervical and vaginal cancers that are staged by clinical examination, most gynaecological cancers are staged surgically. Not only can pelvic and para-aortic lymphadenectomy offer accurate staging information that helps determine patients’ prognosis and hence their treatment plan, but it may also provide a therapeutic effect under certain circumstances. In the past, such a procedure required a big laparotomy incision. With the advent of laparoscopic lighting and instrument, laparoscopic lymphadenectomy became popular since the late 1980s. Dargent et al. published the first report on laparoscopic staging in cervical cancers, and many studies then followed. To date, there are numerous case series and trials evaluating the efficacy and safety of laparoscopic surgery in managing gynaecological cancers. In general, compared with laparotomy, laparoscopic lymphadenectomy has less intraoperative blood loss and post-operative pain, fewer wound complications, shorter length of hospital stay and more speedy recovery. However, this is at the expense of longer operative time. The incidence of port-site metastasis is extremely low, although it may be higher in advanced ovarian cancer. Preliminary data showed that there was no significant effect on recurrence and survival, but long-term data are lacking. In this article, the roles of laparoscopy in staging of uterine, cervical and ovarian cancers, the three most common gynaecological cancers, will be reviewed. Novel technologies such as robot-assisted surgery, single-port surgery and sentinel node biopsy will also be discussed.
Highlights
- •
Laparoscopic staging is feasible and safe in cervical and endometrial cancers.
- •
Laparoscopic surgery is also feasible in early ovarian cancers.
- •
Laparoscopic-related port-site metastasis is rare.
- •
Robot-assisted surgery has a low conversion rate and a short length of hospital stay.
- •
The cost-effectiveness of laparoscopic staging is not worse than laparotomy.
Introduction
According to the International Federation of Obstetrics and Gynaecology (FIGO), most gynaecological malignancies, except cervical and vaginal cancers that are staged clinically with or without imaging examination, are staged surgically. In the past, staging was performed by laparotomy. Dargent first described the use of laparoscopic pelvic lymphadenectomy (PLND) together with radical vaginal hysterectomy in early cervical cancer in 1987 . After his pioneering work, Querleu et al. published a series of 39 stage Ib–IIb cervical cancer patients receiving laparoscopic PLND via a transperitoneal approach , where an average of 8.7 lymph nodes (range 3–22) was obtained and the mean operation time was 80 min. Nezhat et al. and Childers et al. subsequently reported the first few cases of laparoscopic para-aortic lymphadenectomy (PALND) in addition to hysterectomy in cervical and endometrial cancers, respectively . Laparoscopic surgery has a potential benefit compared to conventional laparotomy by reducing the amount of operative blood loss, the degree of post-operative pain, the rate of wound complications and the length of hospital stay (LOH). However, additional issues such as the yield of lymph nodes, risk of port-site metastasis, effect on patients’ survival and recurrence need to be addressed when evaluating the feasibility of laparoscopic surgery.
In this review article, the role and safety of laparoscopic staging in the three most common gynaecological cancers, cervical, endometrial and ovarian cancers, will be discussed. The indications and controversies of staging in these cancers will be covered in other chapters. Robot-assisted and single-port surgery, as well as sentinel node biopsy, will also be described.
Roles of laparoscopic staging
Carcinoma of cervix
About 85% of cervical cancer occurs in developing countries where resources are limited. As a result, its staging relies heavily on clinical examination, and modern imaging techniques such as computed tomography (CT), magnetic resonance imaging (MRI) and positron emission tomography-CT (PET-CT), as well as surgical staging, are not mandatory. Nevertheless, although lymph node status is not incorporated in the FIGO staging system, lymph node metastasis is an independent prognostic factor of cervical cancer and so lymphadenectomy forms an integral part in the treatment.
Early-stage diseases
The incidence of lymph node metastasis is about 0.1–2.6% in stage Ia1 diseases , and so lymphadenectomy can be safely obviated. However, the incidence increases after stage 1a2 8 and therefore lymph node evaluation is essential for the treatment planning.
CT, MRI and PET-CT have been used before the start of treatment. Nevertheless, their accuracy in detecting metastatic lymph nodes is variable and the need of lymphadenectomy could not be precluded even in the absence of positive findings. On the other hand, laparoscopic staging can allow a direct assessment of the lymph node status for those surgical candidates. It can be performed before definitive surgery at a separate setting, or more commonly, at the same operation with radical hysterectomy (RH) or trachelectomy for those who wish to preserve fertility. For the latter, lymphadenectomy is usually performed first and frozen section may be needed if any suspicious lymph nodes are detected. Frozen section of lymph nodes has 80% sensitivity, nearly 100% specificity and positive predictive value, 90–99.5% negative predictive value, 0% false-positive rate and 4.5–19% false-negative rate . In a study of 448 patients with stage Ia1–IIa undergoing laparoscopic lymphadenectomy before RH, 54 patients were found to have lymph node metastasis on frozen section and hence RH was abandoned . In the remaining patients who had completed hysterectomy, only 9.9% required adjuvant chemo-irradiation due to other histological risk factors. This approach can delineate a group of patients with lymph node metastasis so that they could be directly treated by chemo-irradiation, avoiding both RH and adjuvant radiotherapy and hence potentially over-treatment.
It is difficult to examine the performance of laparoscopic lymphadenectomy alone as it is usually accompanied by hysterectomy/trachelectomy. In the early era, most centres had to validate the reliability and safety of laparoscopic lymphadenectomy against an immediate laparotomy. In 1992, Childers et al. depicted that among their five patients who had an RH immediately following laparoscopic PLND and PALND, an average of 2.8 more lymph nodes were obtained in the laparotomy . The average number of lymph nodes yielded was 31.4 with operating time ranging from 75 to 175 min. Fowler et al. showed that the average number of pelvic nodes removed at laparoscopy was 23.5 (range 7–33) . In addition, among all the 377 lymph nodes being removed, 75% were obtained at laparoscopy and the rest were retrieved at laparotomy following laparoscopy. However, none of the patients with negative nodes at laparoscopic lymphadenectomy were found to have further positive nodes at laparotomy. In other words, it implied that all positive nodes could have been sampled by laparoscopy. A Gynecologic Oncology Group (GOG) prospective study included 73 patients with stage Ia–IIa cervical cancer planning to undergo laparoscopic PLND, para-aortic lymph node sampling and then immediate open RH . Seventeen women (23.3%) did not complete the laparoscopic procedures because of metastatic and unresectable lymph nodes or complications. Among the remaining 40 patients who were evaluable, the mean number of pelvic and para-aortic lymph nodes removed was 31.1 and 12.1, respectively. One patient (2.5%) experienced ureteric injury and seven (10.5%) had vascular injury, of which three required laparotomy. Six patients had residual lymph nodes as assessed on laparotomy, but none of them showed metastatic diseases.
With regard to the effect of laparoscopic lymphadenectomy on disease survival, one study compared 76 stage Ib1–2 patients receiving laparoscopic lymphadenectomy followed by either open RH at the same or separate session or chemo-irradiation in the case of positive lymph nodes (case), with 93 historical patients who had open RH and PLND (control) . In the whole cohort regardless of the lymph node status, both groups had similar intraoperative complication and survival rates, and the case group had a higher yield of lymph nodes (25.5 vs. 22, p = 0.01) and lower local recurrence rate (5.3% vs. 17.2%, p = 0.018). On the other hand, among those who required adjuvant radiotherapy or primary chemo-irradiation, the case group had better disease-free survival than the control. Therefore, laparoscopic lymphadenectomy did not have any deleterious effects on the patients’ survival.
Advanced-stage diseases
The incidence of pelvic lymph node metastasis is 10–27% in stage IIa diseases and it increases to 43% in stage IIIb, and that of para-aortic lymph node involvement is 7–25% . In a review including >700 patients with stage 1b-IV diseases undergoing systematic lymphadenectomy, eight (1.1%) had isolated para-aortic lymph node metastasis and two (25%) were above the inferior mesenteric artery (IMA) .
The mainstay treatment in advanced diseases is chemo-irradiation, and imaging is commonly performed prior to this for staging and planning purposes. However, a systematic review showed that 4–35% of histologically proven para-aortic lymph node metastasis was missed by CT, MRI or PET-CT . By contrast, not only can pretreatment lymphadenectomy or debulking of enlarged lymph node ascertain the lymph node status and provide a prognostic value but it may also reduce the risk of pelvic recurrence or improve the survival by modulating the extent of the radiotherapy field and determining the need of systematic chemotherapy. Besides, the knowledge of the extent of lymph node involvement can also avoid unnecessary extended-field radiotherapy, which is associated with 15% risk of grade 3–4 toxicity and even death .
Nonetheless, Lai et al. conducted a randomized trial comparing clinical staging with the aid of CT or MRI and surgical staging using either a laparoscopic or extraperitoneal approach in advanced cervical carcinoma . The study accrual was prematurely terminated because these surgically staged patients had a significantly worse progression-free survival than the clinically staged patients. At follow-up of 2 more years, this difference remained significant ( p = 0.003) and the difference in the overall survival (OS) also became significant ( p = 0.024). However, this study had a certain bias as the surgical arm had more poor prognostic factors. After this trial, another prospective trial examined 237 stage Ib2–IVa patients who had negative para-aortic lymph nodes on PET imaging . Twenty-nine patients (12.2%) were found to have nodal metastasis on laparoscopic PALND. The 3-year event (i.e., recurrence, disease progression or death) free survival rate for those with no para-aortic lymph node involvement and those with ≤5-mm and >5-mm para-aortic lymph node metastasis was 74%, 69% and 17%, respectively. The good prognosis of those with ≤5-mm para-aortic node metastasis might be due to the presence of a single node only, the small size of the metastasis or the use of extended-field chemo-irradiation. The Lymphadenectomy in Locally Advanced Cervical Cancer Study (LiLACS) is a multicentre phase III trial trying to recruit those patients with stage Ib2–IVa cervical cancers whose PET shows hypermetabolic pelvic but not para-aortic lymph nodes and to evaluate whether pretreatment laparoscopic extraperitoneal staging followed by tailored chemo-irradiation could improve their survival compared with those who have PET alone followed by chemo-irradiation . It is still in the recruitment phase and the result will probably be available 8 years later.
Roles of laparoscopic staging
Carcinoma of cervix
About 85% of cervical cancer occurs in developing countries where resources are limited. As a result, its staging relies heavily on clinical examination, and modern imaging techniques such as computed tomography (CT), magnetic resonance imaging (MRI) and positron emission tomography-CT (PET-CT), as well as surgical staging, are not mandatory. Nevertheless, although lymph node status is not incorporated in the FIGO staging system, lymph node metastasis is an independent prognostic factor of cervical cancer and so lymphadenectomy forms an integral part in the treatment.
Early-stage diseases
The incidence of lymph node metastasis is about 0.1–2.6% in stage Ia1 diseases , and so lymphadenectomy can be safely obviated. However, the incidence increases after stage 1a2 8 and therefore lymph node evaluation is essential for the treatment planning.
CT, MRI and PET-CT have been used before the start of treatment. Nevertheless, their accuracy in detecting metastatic lymph nodes is variable and the need of lymphadenectomy could not be precluded even in the absence of positive findings. On the other hand, laparoscopic staging can allow a direct assessment of the lymph node status for those surgical candidates. It can be performed before definitive surgery at a separate setting, or more commonly, at the same operation with radical hysterectomy (RH) or trachelectomy for those who wish to preserve fertility. For the latter, lymphadenectomy is usually performed first and frozen section may be needed if any suspicious lymph nodes are detected. Frozen section of lymph nodes has 80% sensitivity, nearly 100% specificity and positive predictive value, 90–99.5% negative predictive value, 0% false-positive rate and 4.5–19% false-negative rate . In a study of 448 patients with stage Ia1–IIa undergoing laparoscopic lymphadenectomy before RH, 54 patients were found to have lymph node metastasis on frozen section and hence RH was abandoned . In the remaining patients who had completed hysterectomy, only 9.9% required adjuvant chemo-irradiation due to other histological risk factors. This approach can delineate a group of patients with lymph node metastasis so that they could be directly treated by chemo-irradiation, avoiding both RH and adjuvant radiotherapy and hence potentially over-treatment.
It is difficult to examine the performance of laparoscopic lymphadenectomy alone as it is usually accompanied by hysterectomy/trachelectomy. In the early era, most centres had to validate the reliability and safety of laparoscopic lymphadenectomy against an immediate laparotomy. In 1992, Childers et al. depicted that among their five patients who had an RH immediately following laparoscopic PLND and PALND, an average of 2.8 more lymph nodes were obtained in the laparotomy . The average number of lymph nodes yielded was 31.4 with operating time ranging from 75 to 175 min. Fowler et al. showed that the average number of pelvic nodes removed at laparoscopy was 23.5 (range 7–33) . In addition, among all the 377 lymph nodes being removed, 75% were obtained at laparoscopy and the rest were retrieved at laparotomy following laparoscopy. However, none of the patients with negative nodes at laparoscopic lymphadenectomy were found to have further positive nodes at laparotomy. In other words, it implied that all positive nodes could have been sampled by laparoscopy. A Gynecologic Oncology Group (GOG) prospective study included 73 patients with stage Ia–IIa cervical cancer planning to undergo laparoscopic PLND, para-aortic lymph node sampling and then immediate open RH . Seventeen women (23.3%) did not complete the laparoscopic procedures because of metastatic and unresectable lymph nodes or complications. Among the remaining 40 patients who were evaluable, the mean number of pelvic and para-aortic lymph nodes removed was 31.1 and 12.1, respectively. One patient (2.5%) experienced ureteric injury and seven (10.5%) had vascular injury, of which three required laparotomy. Six patients had residual lymph nodes as assessed on laparotomy, but none of them showed metastatic diseases.
With regard to the effect of laparoscopic lymphadenectomy on disease survival, one study compared 76 stage Ib1–2 patients receiving laparoscopic lymphadenectomy followed by either open RH at the same or separate session or chemo-irradiation in the case of positive lymph nodes (case), with 93 historical patients who had open RH and PLND (control) . In the whole cohort regardless of the lymph node status, both groups had similar intraoperative complication and survival rates, and the case group had a higher yield of lymph nodes (25.5 vs. 22, p = 0.01) and lower local recurrence rate (5.3% vs. 17.2%, p = 0.018). On the other hand, among those who required adjuvant radiotherapy or primary chemo-irradiation, the case group had better disease-free survival than the control. Therefore, laparoscopic lymphadenectomy did not have any deleterious effects on the patients’ survival.
Advanced-stage diseases
The incidence of pelvic lymph node metastasis is 10–27% in stage IIa diseases and it increases to 43% in stage IIIb, and that of para-aortic lymph node involvement is 7–25% . In a review including >700 patients with stage 1b-IV diseases undergoing systematic lymphadenectomy, eight (1.1%) had isolated para-aortic lymph node metastasis and two (25%) were above the inferior mesenteric artery (IMA) .
The mainstay treatment in advanced diseases is chemo-irradiation, and imaging is commonly performed prior to this for staging and planning purposes. However, a systematic review showed that 4–35% of histologically proven para-aortic lymph node metastasis was missed by CT, MRI or PET-CT . By contrast, not only can pretreatment lymphadenectomy or debulking of enlarged lymph node ascertain the lymph node status and provide a prognostic value but it may also reduce the risk of pelvic recurrence or improve the survival by modulating the extent of the radiotherapy field and determining the need of systematic chemotherapy. Besides, the knowledge of the extent of lymph node involvement can also avoid unnecessary extended-field radiotherapy, which is associated with 15% risk of grade 3–4 toxicity and even death .
Nonetheless, Lai et al. conducted a randomized trial comparing clinical staging with the aid of CT or MRI and surgical staging using either a laparoscopic or extraperitoneal approach in advanced cervical carcinoma . The study accrual was prematurely terminated because these surgically staged patients had a significantly worse progression-free survival than the clinically staged patients. At follow-up of 2 more years, this difference remained significant ( p = 0.003) and the difference in the overall survival (OS) also became significant ( p = 0.024). However, this study had a certain bias as the surgical arm had more poor prognostic factors. After this trial, another prospective trial examined 237 stage Ib2–IVa patients who had negative para-aortic lymph nodes on PET imaging . Twenty-nine patients (12.2%) were found to have nodal metastasis on laparoscopic PALND. The 3-year event (i.e., recurrence, disease progression or death) free survival rate for those with no para-aortic lymph node involvement and those with ≤5-mm and >5-mm para-aortic lymph node metastasis was 74%, 69% and 17%, respectively. The good prognosis of those with ≤5-mm para-aortic node metastasis might be due to the presence of a single node only, the small size of the metastasis or the use of extended-field chemo-irradiation. The Lymphadenectomy in Locally Advanced Cervical Cancer Study (LiLACS) is a multicentre phase III trial trying to recruit those patients with stage Ib2–IVa cervical cancers whose PET shows hypermetabolic pelvic but not para-aortic lymph nodes and to evaluate whether pretreatment laparoscopic extraperitoneal staging followed by tailored chemo-irradiation could improve their survival compared with those who have PET alone followed by chemo-irradiation . It is still in the recruitment phase and the result will probably be available 8 years later.
Carcinoma of corpus
The gold-standard treatment of endometrial cancer is total hysterectomy, bilateral salpingo-oophorectomy with or without PLND/PALND. Positive peritoneal cytology is no longer incorporated in the FIGO staging system, but peritoneal washing is still needed as it may carry prognostic value. Omentectomy is indicated in high-risk diseases such as clear cell and serous subtypes.
The role of lymphadenectomy in endometrial cancer has been a long debate in the past decades, and this will be discussed in another chapter. Systematic lymphadenectomy is advised if high-risk features are present, such as grade 3 or type II carcinomas, deep myometrial invasion, size of >2 cm and cervical or extrauterine involvement . There are robust data supporting laparoscopic surgery in endometrial cancer. The earlier GOG LAP2 trial found that the conversion was up to 25.8% mainly due to poor visibility (14.6%), metastatic cancer (4.1%) and bleeding (2.9%) . Ten laparoscopic patients and eight laparotomy patients had perioperative deaths (0.59 vs. 0.88%, p = 0.404) mainly attributed to thromboembolism. Nevertheless, a systematic review published in 2013 encompassing nine randomized controlled trials showed that laparoscopic surgery outperformed laparotomy by having shorter LOH (mean difference (MD) −3.42 days, 95% confidence interval (CI) −3.81 to −3.03 days; p < 0.01), and fewer post-operative complications (15.8 vs. 23.4%, odds ratio (OR) 0.62, 95% CI 0.52–0.73; p < 0.01), despite having higher rates of intraoperative complications (9.4 vs. 7.0%, OR 1.35, 95% CI 1.05–1.74; p = 0.02) and longer duration of surgical procedures (MD 32.73 min, 95% CI 16.34–49.13 min; p < 0.01) .
On the other hand, the 3-year OS (87.0 vs. 88.0%, OR 0.91, 95% CI 0.49–1.71; p = 0.77), disease-free survival (87.5 vs. 88.0%, OR 0.95, 95% CI 0.29–1.80; p = 0.89) and recurrence rate (12.8 vs. 11.6%, OR 1.11; 95% CI 0.60–2.06; p = 0.74) were similar between those undergoing laparoscopic and open staging . The non-inferiority of the laparoscopic approach in terms of survival was also confirmed by the latest survival analysis of the LAP2 trial, where the estimated 3-year cumulative recurrence rate was 10.2% and 11.4% for laparoscopic and open routes, respectively, and the estimated 5-year OS of both was 89.8% .
Obesity is associated with type I endometrial cancer, and it can pose difficulty in surgery due to limited exposure, anaesthetic risks and potential post-operative complications such as thromboembolism and wound complications . Although morbid obesity may hinder complete laparoscopic staging and those being converted tend to have a higher body mass index (30.8 ± 9.7 vs. 26.7 ± 11.4; p = 0.01) , >88% of obese patients could successfully undergo laparoscopic surgery with an acceptable perioperative complication rate . Age is another risk factor for endometrial cancer and it is also a potential limiting factor for minimally invasive surgery. Based on the reports albeit scarce in the literature, it appeared that laparoscopic surgery had lesser blood loss and shorter LOH than vaginal or open surgery in the elderly and their outcomes resembled those of younger patients .
Carcinoma of ovary
Traditionally, laparotomy is the gold standard in managing ovarian cancer. On entering the abdomen through a midline skin incision, ascitic fluid or peritoneal washing is collected for cytological examination, and the whole abdominal cavity has to be explored. Other than total hysterectomy and bilateral salpingo-oophorectomy, or unilateral salpingo-oophorectomy in those who wish to preserve fertility, omentectomy, peritoneal biopsy, PLND and PALND should be performed, and all suspicious lesions should be debulked and sent for histological examination. Staging is important as 16–39.5% of apparently early-stage ovarian cancers are upstaged after a staging procedure .
Unlike cervical and endometrial cancers, there is a relative paucity of publications on laparoscopy in ovarian cancer, and the recent Cochrane review failed to identify any randomized controlled or prospective case–control studies . The lack of good evidence on minimally invasive surgery in ovarian cancer can be explained by the relatively high frequency of advanced-stage diseases, which would require more extensive surgery and hence surgical skills, making laparoscopic operation less popular and/or feasible . The other deterrents may be the concern of dissemination by exfoliation of cells and possibility of tumour rupture during laparoscopic surgery.
Bogani et al. performed a meta-analysis including five comparative studies as well as their own series . Compared to patients having laparotomy, those undergoing laparoscopy had lesser blood loss (MD −175.7 ml; 95% CI −219 to −132.3 ml), longer operative time (MD 16.8 min; 95% CI 8.8–24.8 min), shorter LOH (MD −3.3 days; 95% CI −3.9 to −2.7 days) and earlier commencement of adjuvant chemotherapy (MD −4.9 days 95% CI −6.7 to −3.2 days). While there was no difference in the intraoperative complication rate (2.0 vs. 0.8%, OR 2.47; 95% CI 0.5–12.2), the laparoscopy group had fewer post-operative complications than the laparotomy group (5.3 vs. 20.3%, OR 0.25; 95% CI 0.12–0.53). There was no difference in the rates of spillage (7.2 vs. 9.5%, OR 0.78; 95% CI 0.35–1.73), upstaging (17.1 vs. 16.6%, OR 0.7; 95% CI 0.38–1.27) as well as the recurrence (5.3 vs. 8.3%, OR 0.5; 95% CI, 0.21–1.21) between the two approaches. In another meta-analysis including those single-arm studies, the intraoperative tumour rupture rate was 25.4%, the upstaging rate was 22.6% and the recurrence rate within a median follow of ≥19 months was 9.9% . Although these rates appeared to be higher than those reported in Bogani’s meta-analysis, direct comparison is difficult as the natures of the studies included were different. The overall conversion rate to laparotomy was 3.7% after adjustment, showing that laparoscopic staging is a feasible option in early-stage ovarian cancer.
The role of laparoscopic staging in stage III–IV ovarian carcinoma is uncertain. Laparoscopic scoring using parameters such as omental cake, peritoneal and diaphragmatic carcinomatosis, mesenteric retraction, bowel and stomach infiltration and spleen and/or liver superficial metastasis has been proven to be accurate in predicting resectability in advanced ovarian cancers and it could avoid unnecessary upfront laparotomy, which might otherwise result in suboptimal debulking . The LapOvCa trial is a multicentre randomized controlled trial evaluating the role of diagnostic laparoscopy in triaging patients to upfront surgery and neoadjuvant chemotherapy and is nearly finishing the patient recruitment . Under certain circumstances, optimal primary debulking might also be attempted using laparoscopy although the evidence is scarce .
Other safety concerns
At the beginning, the complications related to laparoscopic staging were not uncommon. Spirtos et al. reported a 12.5% failure rate among 40 patients scheduled for laparoscopic lymphadenectomy . Two were due to bleeding, two had unsuspected intra-abdominal diseases and one was due to equipment failure. Other complications included rehospitalization, hernia, bowel obstruction and deep vein thrombosis. In 1998, Possover et al. reported their series of 150 patients with laparoscopic PLND and PALND . Although the yield of pelvic (mean 26.8, range 1–56) and para-aortic lymph nodes (mean 7.3, range 0–19) was comparable to the literature, major vessel injury occurred in 4.7% of their patients. On the other hand, Scribner et al. showed that the top three reasons for conversion to laparotomy were obesity (29.1%), adhesions (16.7%) and intraperitoneal disease (16.7%) . In their series, two patients died of vascular injury and pulmonary embolism, respectively.
The other major concern in laparoscopic oncology operations is port-site metastasis. The overall incidence is low, being 0.1–0.43% in cervical cancer and 0.06–0.33% in endometrial cancer . In a systematic review including 397 articles, only 12 cases of port-site recurrence were identified and no preventive strategies could be concluded . The incidence in ovarian cancer was reported to be 0.89%. Park et al. found only one case of port-site metastasis among the 11 studies on the use of laparoscopy in ovarian cancer . However, Vergote et al. found that 17% of their cohort with stage III–IV ovarian cancer developed port-site metastasis after diagnostic open laparoscopy . Heitz et al. showed an even higher rate (47%) of abdominal wall metastasis in their 66 patients who underwent cytoreductive surgery after laparoscopy at a median of 31 days . FIGO stage IV as compared to stage III (OR 17.0, 95% CI 1.7 – 168; p = 0.015) as well as the presence of ascites of >500 ml (OR 7.2, 95 CI 1.5–35.8; p = 0.016) were independent risk factors for abdominal wall metastasis. On the other hand, Ramirez et al. showed that the median time to the development of port-site metastasis varied from 17 days (4–730 days) in ovarian cancer to 13.5 months (6–21 months) in uterine cancer . Interestingly, the prognosis of those advanced ovarian cancer patients with port-site metastasis might not be worsened , probably because their prognosis was poor even without port-site metastasis. On the other hand, Zivanovic et al. demonstrated that those patients who developed port-site metastases within 7 months from the laparoscopic procedure had a median survival of 12 months, whereas those who developed after 7 months had a median survival of 37 months ( p = 0.004) .
The state of the art
Robot-assisted surgery
The US Food and Drug Administration (FDA) first approved the use of Da Vinci Surgical Systems in gynaecology in 2005. Although the surgeon loses the tactile feedback with the robotic system, it can allow access to difficult areas with its dexterity and hence a greater variety of radical operations can be performed. The chief surgeon can also operate with a three-dimensional view enhancing the precision of the operation. As the camera is held by the robotic arm, muscle fatigue and tremor of the assistant can be avoided.
A meta-analysis comparing the performance of open, laparoscopic and robot-assisted RH and lymphadenectomy in cervical cancer showed that the mean operation time, number of lymph nodes and nodal metastasis were similar among the three methods . However, the risk of blood loss, transfusion rate and post-operative infection were higher in the laparotomy group, and LOH was significantly shorter in the robot-assisted group compared to the other two groups. Another meta-analysis on endometrial cancer analysed 22 studies and concluded that robot-assisted surgery was superior to laparoscopy by having lower estimated blood loss, conversion rate and complication rate while other parameters such as length of operation, LOH, transfusion rate and yield of lymph nodes were similar . It also surpassed laparotomy in terms of LOH, complication and transfusion rates. In addition, robot-assisted surgery does not adversely affect the survival and recurrence rates compared with conventional laparoscopy . The data in ovarian cancer are rather limited. However, recent retrospective studies showed that the outcomes of ovarian cancer patients undergoing robot-assisted surgery were comparable to those having laparotomy or laparoscopy, where the 1-year survival and recurrence rates between the robot-assisted and open approaches were all similar .
Single-port surgery
With the development of flexible laparoscope and instruments, single-port surgery, also known as laparoendoscopic single-site surgery (LESS), has emerged into the oncological surgery field. Since the first series published by Fader et al., in 2009 , there have been <10 series up to date, and only one case–control retrospective cohort and a prospective study on endometrial cancer are identified in the literature . Preliminary data showed that LESS was not inferior to traditional laparoscopic surgery and it was associated with better pain control post-operatively. Together with another prospective study on two-port access surgery , a summary of those comparative studies is illustrated in Table 1 . The technology of robot-assisted single-port surgery has just been developed and further studies are needed.
