Essential surgical skills for a gynecologic oncologist

Gynecologic oncology has seen a tremendous growth as a surgical specialty over the past four decades. However, many regions of the world still lack structured training programs in this discipline. The aim of this article is to identify the essential skills for a gynecologic oncologist to be able to provide optimal care to women diagnosed with gynecologic cancer. Where the evidence exists in the literature we identify the learning curve necessary. Identifying essential skills required for the practice of gynecologic oncology should assist in standardizing care provision globally, and could be a starting point for health systems beginning structured training programs. Development of surgical skills requires adequate training, mentorship, and self‐evaluation as an ongoing process beyond the years spent in training programs.


| INTRODUCTION
Gynecologic oncology, introduced as a subspecialty of obstetrics and gynecology over four decades ago, has evolved considerably in recent years. The first step in diagnosis, staging, and treatment of gynecologic cancer is surgical intervention in almost all cases. The types of gynecologic cancers where surgical intervention may be needed are cervical, uterine or endometrial, ovarian, vulvar, and vaginal cancers; for these, the procedures may range from staging, debulking, total/ radical hysterectomy, unilateral/bilateral salpingo-oophorectomy, omentectomy, and lymph node removal, among others. Ovarian cancer is a unique disease where two-thirds of cases are advanced at presentation, with peritoneal carcinomatosis, rectosigmoid involvement, splenic hilum disease, diffuse miliary nodules on mesentery/ bowel, para-aortic nodal disease, and diaphragm disease. Surgical cytoreduction is an essential part of treatment for ovarian cancers.
Heterogeneity exits across various health systems globally regarding the level of surgical skill and team responsible for providing primary surgical care. Few regions of the world have structured training programs in gynecologic oncology, with an increasing number of countries/regions recognizing the need for structured training. In some countries, surgical management of gynecologic cancer is part of the wider specialty of surgical oncology. However, in the vast majority of health systems, gynecologists-with or without the help of general surgeons-undertake primary surgical management. It is important to note that owing to the varied presentation of gynecologic cancer, it is not unusual for the primary surgery to have been carried out by a general surgeon. There is increasing literature evidence that treatment outcomes, as judged by survival figures, are improved when specialists trained in gynecologic oncology undertake surgical management. 1,2 Given the wide disparities in available resources globally, and the burden of disease in economically less privileged parts of the world, a pragmatic approach is taken in the present article. Effort has been made to distinguish between what we consider mandatory skills and desirable skills. Additionally, we have identified a set of optional skills for procedures that are currently not part of mainstream practice.
Mandatory skills are the present standard of care, and which a gynecologic oncologist should be able to perform routinely in their individual practice.
Desirable skills are those that a gynecologic oncologist should be able to perform with the help of other specialty colleagues in a lowvolume center or with or without the assistance of specialty colleagues in a high-volume center. These include procedures that involve allied surgical specialties such as urological, colorectal, plastic, and upper gastrointestinal surgery. Some of these desirable skills currently lack level 1 evidence, but will impact on patient care and quality of care in years to come. A gynecologic oncologist should learn these skills during training and may require help from specialty colleagues during the initial learning curve.
Optional skills are those that currently do not have level 1 evidence and are performed in a research context but are likely to be accepted as standard practice. Optional skills are associated primarily with new technologies that may become part of standard care in the future and are therefore not discussed in detail in this article.
Depending on the resource setting of individual centers and region of the world, some of the procedures mentioned in the "desirable" category might be routine procedures in daily practice.
In our review, we have tried to identify the learning curve for individual procedures where possible. Given that the concept of a learning curve has only been used in the context of newer technologies, beginning with laparoscopy, this information is lacking for most of the "traditional" procedures in gynecologic oncology.

| TRAINING IN GYNECOLOGIC ONCOLOGY
Most regions offering structured training programs in gynecologic oncology define the scope of surgical procedures that trainees are expected to gain experience in during fellowship training. 3 These procedures can broadly be grouped into gynecologic procedures and extended scope surgical procedures that may include general surgical, plastic, colorectal, and urological procedures. We refer to extended scope surgical procedures as "general surgical procedures" in this article. Most structured training programs do not stipulate a minimum number of procedures required for accreditation as a gynecologic oncologist; instead the emphasis is on the overall volume of surgical exposure and some form of assessment process during training.
Regions with structured training programs have identified a need for exposure to extended scope surgical procedures. 3 This is largely related to extended cytoreductive procedures in surgical management of ovarian cancer. Trainee surveys have identified the inability of some training programs to provide adequate experience in general surgical procedures. 4 The standardized training curriculum recommended by the European Society of Gynaecological Oncology [ESGO] has set a minimum number of surgical procedures directly related to gynecologic cancer diagnosis. 5 The minimum training standards for the type and number of extended surgical procedures is difficult to standardize given the variation in surgical management of ovarian cancer across various centers, even in high-income countries. Some common themes can be identified based on a review of the curriculums from various structured training programs. Most of these programs have identified a need for additional training in extended surgical procedures, including small and large bowel surgery, urological procedures, plastic surgery, diaphragmatic surgery, and splenectomy.

| Radical hysterectomy
Any discourse on surgical skills in gynecological oncology must start with radical hysterectomy for cervical cancer (Fig. 1). Radical hysterectomy and pelvic lymph node dissection are the only surgical procedures in gynecologic oncology that have been described in terms of pelvic anatomic landmarks, with the aim to standardize and define the completeness of the surgical outcome. The Piver-Rutledge-Smith classification described in 1974 stood the test of time until 2008 when the Querleu and Morrow classification (Q-M classification, Table 1) based on more specific anatomic landmarks attempted to standardize the description of surgery. 6,7 Cibula et al. 8 attempted to further refine the description of the Q-M classification with emphasis on a three-dimensional description of the parametrium. We recommend that gynecologic oncology training programs provide adequate training in at least type B and type C1 radical hysterectomy, as described in the Q-M classification. The ability to perform a type C2 radical hysterectomy is a skill that is progressively acquired, preferably under the mentorship of a more skilled surgeon. Type D radical hysterectomy should be considered a variation of pelvic exenteration surgery. Surgical skills for management of cervical cancer are shown in Table 2.

| Nerve-sparing radical hysterectomy
Various authors have described nerve-sparing radical hysterectomy (NSRH). The essential component of the technique is to identify the inferior hypogastric nerves (in the mesoureter) that contain the autonomic nerves to the bladder, rectum, and vagina. The aim of NSRH is to prevent or contain the severity of bladder, bowel, and sexual dysfunction resulting from the damage to the autonomic nerve supply. 9 NSRH has been criticized for a lack of standardized surgical description. A recent systematic review of available evidence and meta-analysis of data from selected high-quality studies suggested that the procedure F I G U R E 1 Radical hysterectomy specimen.
is safe with oncologic outcomes comparable to conventional radical hysterectomy. 10 Time to micturition, as a surrogate marker of bladder function, was significantly shorter in the NSRH group. 10 The review identified significant heterogeneity in the studies included, and a significant proportion of the women in either group received adjuvant radiotherapy-an independent factor impacting on quality of life indices.

| Radical trachelectomy and fertility-sparing surgery for cervical cancer
Cervical cancer, with its bimodal peak in incidence, is often diagnosed in women of reproductive age. With increasing age of attempted child bearing, especially in high-resource countries, fertility-preserving surgery for gynecologic cancers is at the forefront of consideration, along with oncologic outcomes and quality of life issues. Dargent is credited with pioneering the concept of radical vaginal trachelectomy, followed by description of an abdominal approach by Smith et al. 11 Radical vaginal trachelectomy (RVT) had been the preferred approach for many years, especially in Europe, and the bulk of the data on safety comes from just over 1300 cases reviewed in the literature. 12 The oncologic outcomes were acceptable, with a reported recurrence rate of 4% and a 2% rate of death related to recurrent disease in women with node-negative disease. 12 Abdominal radical trachelectomy (ART) has gained popularity over the past decade owing to familiarity with the abdominal approach and its less resourceintensive nature for equipment and training. Data on oncologic safety are more heterogeneous for ART as the procedure has been attempted and results reported for tumors larger than those in VRT series. A recurrence rate of 4% has been reported from pooled data amounting to around 350 procedures. 13 Either VRT or ART is acceptable for carefully selected patients who are committed to fertility preservation, and who are diagnosed with squamous cell cancers, adenocarcinoma, or adenosquamous carcinomas less than 2 cm in size. Careful patient selection will minimize the rates of failed or abandoned procedures, with the exception of 10%-12% of patients who will still require adjuvant treatment following trachelectomy. 13

| Pelvic lymph node dissection for cervical and endometrial cancers
Cervical cancer staging does not take into account pelvic nodal status.
However, nodal status has a significant impact on postsurgical treatment planning and prognostication in cervical cancer.
Nodal status is an essential part of the surgicopathologic staging for endometrial cancer. The pelvic component of the nodal dissection is essentially the same for the two pathologies.
The boundaries of pelvic nodal dissection are defined as follows: medial extent limited by the internal iliac artery and its continuation as the obliterated umbilical/hypogastric artery; lateral boundary of the genitofemoral nerve overlying the psoas muscle; distal limit defined by the circumflex iliac vein crossing the external iliac artery; deep T A B L E 1 Modified Querleu-Morrow classification of radical hysterectomy. dissection is usually limited to the obturator nerve and the proximal limit at the common iliac artery. The proximal limit has been variably described between the bifurcation of the common iliac to its origin at the bifurcation of the aorta.

| Minimally invasive surgery
Laparoscopic surgery has irreversibly shifted the paradigm of training in gynecological surgery. With the introduction of the levonorgestrel intrauterine system for the treatment of abnormal benign uterine bleeding, the incidence of hysterectomy has dropped sharply. This, coupled with a longer learning curve for laparoscopic surgery, has had an impact on the available opportunities for both learning and training in laparoscopic surgery for gynecologists and gynecologic oncolo-  16,17 Laparoscopic hysterectomy is resource intensive owing to the capital outlay for equipment costs and longer theatre times, but these costs are offset by a lower incidence of postoperative complications and shorter hospital stay. 18

| Laparoscopic hysterectomy for endometrial cancer
Endometrial cancer is the most common gynecologic cancer in highincome countries. It is certainly the most common gynecologic malignancy in the USA, Australia, and Europe-regions where the rates of obesity are high and from where the bulk of published literature on the safety and successful introduction of laparoscopic approach has been reported. 19 Given that the oncologic outcomes and survival compare favorably to open surgery, as evident from prospective randomized controlled trials, laparoscopic surgery for early stage endometrial cancer should be considered as the standard management. 20 The incidence of intraoperative complications is similar with laparo-  19,20 The difference in quality of life between the two groups levels off at 6 months after surgery. Surgical skills for management of endometrial cancer are shown in Table 3.
The success of laparoscopic surgery can be limited by both patient and surgeon limitations, as highlighted by the Gynecology Oncology Group (GOG) Lap 2 study. 21 While operative difficulty due to high BMI was cited as the leading reason for conversion to laparotomy, the authors do allude to level of surgeon experience as a possible factor in the relatively higher conversion rate, compared with the LACE Trial. 17,22 It should be noted that over 90% of patients underwent pelvic and para-aortic nodal staging in the Lap 2 study. 21

| Laparoscopic radical hysterectomy for cervical cancer
There is no definitive conclusion on the role of minimally invasive surgery for surgical management of cervical cancer. While laparoscopic/ robotic surgery brings with it the advantages of minimally invasive procedures to women needing to undergo radical hysterectomy and nodal dissection, the only data in the literature are from single institution or single surgeon series. Data from retrospective and single institution series suggest that laparoscopic and robotic radical hysterectomy are probably associated with acceptable oncologic outcomes, but the data are limited and long-term follow-up information is lacking. 23 We require data from adequately designed, randomized controlled trials to categorically state that a minimally invasive approach T A B L E 3 Surgical skills for management of endometrial cancer.

Mandatory
Desirable Optional • Total abdominal hysterectomy and bilateral salpingo-oophorectomy • Pelvic lymph node dissection • Para-aortic lymph node dissection up to renal veins • Other standard staging procedures such as peritoneal biopsy, infracolic omentectomy, peritoneal cytology collection • Laparoscopic approach for skills mentioned in mandatory procedures • Robotic surgery skills • Pelvic sentinel lymph node detection does not compromise oncologic outcomes. One such trial, the Laparoscopic Approach to Cervical Cancer (LACC trial), is currently underway and the mature data will not be available until 2022. The aim is to compare the disease-related outcomes as well as quality of life indices between abdominal radical hysterectomy and laparoscopic or robotic radical hysterectomy.
The data from a single institution study, where all surgeries were performed by a single surgeon, suggest that the learning curve for laparoscopic radical hysterectomy with pelvic and/or para-aortic lymph node dissection can be as high as the first 50 cases before there was a significant improvement in surgical outcomes and reduction in complication rates. 24

| Laparoscopic pelvic lymph node dissection
The ability to carry out pelvic lymph node dissection laparoscopically should be an essential skill, together with laparoscopic hysterectomy, given that a significant proportion of these patients would require comprehensive staging surgery to include pelvic with or without para-aortic lymph node assessment (Fig. 2). This further adds to the learning curve with increased surgical risks of injury to the ureter, pelvic vessels, and nerves. Teaching advanced laparoscopic procedures for pelvic malignancies adds to the complications resulting from the procedure. 15 When compared with open surgery, minimally invasive surgery did not compromise on nodal count or detection of nodal metastatic disease, but was associated with significantly lower blood loss and hospital stay. 25 A lack of training in laparoscopic pelvic lymph node dissection is likely to negate the benefits of training in laparoscopic hysterectomy in two ways: either the operating team is likely to convert the surgery from laparoscopic to open to complete the procedure, or there may be a temptation to omit lymph node dissection in cases where it would have been reasonably indicated. Either scenario has the potential to compromise patient outcomes. Hence, any training program with emphasis on minimally invasive procedures should include laparoscopic pelvic lymph node dissection along with hysterectomy in the curriculum.

| Para-aortic lymph node dissection
Lymphatic drainage of the ovaries and uterine body along the ovarian lymphatics follows their embryonic origin and blood supply from the para-aortic region. Para-aortic lymph node dissection is an essential part of staging for apparent early stage ovarian cancer and for highrisk endometrial cancer. Para-aortic lymph nodes may be microscopically involved in 18%-24% of cases of apparent Stage I and II ovarian cancer. 26,27 Additionally, debulking of para-aortic nodal disease in a subset of advanced ovarian cancer with predominantly nodal spread, and in some cases of endometrial cancer, presents unique anatomic challenges owing to the proximity of the nodal disease to important vessels-namely the inferior mesenteric artery, renal vessels, and proximal ovarian vessels; and vital structures such as the ureters and duodenum. Debulking of para-aortic nodal disease requires dexterity and an intimate knowledge of the anatomy of the region (Figs 3a and 3b).
F I G U R E 2 Open para-aortic node dissection and robotic pelvic node dissection. learning curve for laparoscopic para-aortic lymph node dissection in a porcine model in the initial days of laparoscopy. 28 They noted a learning curve of 14 cases to achieve adequate lymph node retrieval and the ability to complete the procedure laparoscopically without need for conversion to laparotomy for complications. 28 It is expected that the learning curve would be even longer for surgical training with patients.
The robotic-assisted minimally invasive approach has a shorter learning curve, an increased ability to perform complex procedures, such as high para-aortic lymph node dissection in women with high BMI, and low conversion rates. Single institution prospective data and multicenter retrospective data have indicated the feasibility of roboticassisted para-aortic lymph node dissection, with low conversion rates and acceptable morbidity. 29,30 Currently there is no evidence on the clinical superiority of the robotic-assisted approach compared with traditional laparoscopic surgery in the management of endometrial cancer.

| Sentinel lymph node detection for pelvic cancers
Evidence for sentinel lymph node (SLN) detection for endometrial and cervical cancers has largely been described in the context of lap-

| Surgery for vulvar cancer
Vulvar cancer is a rare condition, accounting for approximately 5% of all gynecologic cancers and less than 1% of all cancers diagnosed in women. Typically, vulvar cancer has been considered a disease of old age, but is increasingly being diagnosed in younger women.
Until the pioneering work of Taussig and Way in the middle of  Table 4.
The technique for inguinofemoral lymph node dissection has also undergone a major shift, from the days when removal of skin overlying the femoral triangle and the skin bridge between the vulva and the groin was advocated, to an increasingly conservative surgical approach that can be carried out through much smaller incisions. The dissection should aim to remove all lymph nodes medial to the femoral artery, both superficial and deep to the cribriform fascia. 35

| Groin sentinel lymph node detection
Detection of SNL in the groin or inguinofemoral group of lymph nodes has been around for several years. Although there are no randomized controlled trials on its efficacy and safety, data are available from two multicenter prospective observational studies on its feasibility and safety in unifocal vulvar cancers under 4 cm in maximal dimension.
Concerns related to false-negative rates persist, but the procedure is an option available to women who cannot [owing to medical issues] or would not undergo full lymph node dissection (Fig. 5). Full lymph node dissection should still be considered as the gold standard of care in women presenting with vulvar cancer.
Long-term, disease-specific, and survival data following performance of the sentinel node procedure from the GROningen

INternational Study on Sentinel Nodes in Vulvar cancer (GROINSS-V)
I was recently published. 36 This is one of the two largest prospective observational studies on the role of sentinel nodes in vulvar cancer.
The long-term data suggest an isolated groin recurrence rate of 2.5% for SLN-negative patients with vulvar tumors less than 4 cm who only underwent SLN detection without inguinofemoral lymph node dissection. All recurrences were diagnosed within 16 months of first surgery.
All patients diagnosed with isolated groin recurrences succumbed to the disease.
The main reason for selecting the SLN procedure over complete inguinofemoral lymph node dissection is to avoid a 50%-70% risk of lower limb lymphedema following the removal of all groin lymph nodes. 37

| Plastic surgery for vulvar cancer
Long-term data from the GROINSS-V study would suggest that 40% of women experience a recurrence of vulvar cancer over a 10-year follow-up. 36 This means that nearly one in two women would require repeat surgery for vulvar cancer. A significant proportion of women present with a large-volume primary tumor that may require plastic reconstructive procedures to close the defect with good blood supply and avoid undue tension on the wound. Re-excision for recurrent vulvar cancer after previous radiotherapy is associated with a high risk of wound complications, and plastic repair with nonirradiated musculocutaneous grafts offers the best chance to close the defect and promote healing as they bring fresh blood supply to the affected area.
Plastic surgical procedures can offer some semblance of normal appearance after a surgical procedure that may leave a patient with significant disfigurement of the vulvar region.
Plastic surgical procedures may range from simple advancement flaps or V-Y plasty to allow the closure of the defect, to more complex plastic procedures such as lotus flaps or Singapore flap reconstructions. Gynecologic oncologists should be able to perform simple advancement flaps to allow adequate wound closure without tension, but may require the assistance of plastic surgeons for more complex vascular pedicle or musculocutaneous flap repairs.

| Surgery for advanced stage ovarian cancer
Ovarian cancer presents with peritoneal and/or nodal metastatic disease in 75% of women at diagnosis. Gynecologic oncology has witnessed an increasingly aggressive approach to cytoreductive surgery for ovarian cancer since the time Meigs described the benefit of reduction in tumor volume, to Griffiths' landmark paper that provided the first scientific evidence of an inverse relationship between tumor volume and patient survival. The definition of "optimal cytoreduction" has changed over the years to less than 2 cm maximal residual disease to the current definition of less than 1 cm residual disease as defined by the GOG. 40 The consensus from the Gynecological Cancer InterGroup (GCIG) Fourth Ovarian Cancer Consensus Conference in 2010 was to define "optimal debulking" as the absence of any residual disease or no gross residual disease. 41 The scope of surgical resections undertaken by a gynecologic oncologist has progressively expanded to include small and large bowel resections, diaphragmatic surgery (Figs 6a and 6b), splenectomy, distal pancreatectomy, subsegmental liver resection, and mesenteric peritoneal resection. In some regions the scope has gone even further to include partial/sleeve gastrectomy, cholecystectomy, and resection of disease from the porta hepatis. Surgical skills for management of ovarian cancer are shown in Table 5. The other issue that arises from the retrospective review of surgical outcomes from the GOG 182 and SCOTROC-1 trials is whether ability to achieve no gross residual disease is a function of better disease biology rather than maximal surgical effort. 43,44 The authors argue that maximal surgical effort to include increasingly complex procedures cannot overcome aggressive disease biology. ing complete debulking. 47 It is important to note that while this information has influenced our practice, this is post hoc analysis of data that was not originally powered or intended to answer the question of impact of surgical debulking on survival outcomes. In the absence of any relevant data from well-designed randomized controlled trials, this is the best information currently available to be able to counsel women on the benefits of extended cytoreduction for ovarian cancer.
Despite the information from these studies with a large volume of patient data, a Cochrane review on the role of extended cytoreductive surgery or "ultra-radical surgery" involving complex upper abdominal procedures failed to reveal good quality evidence in favor of these procedures when evaluated against survival as the outcome measure. 48 There was also a lack of data on the impact of these procedures on quality of life. This is a reflection of lack of prospective evidence from well-designed randomized trials and a difficulty in standardizing the most important factors impacting on outcome, i.e. disease burden, surgical effort, and surgeon performance.

| Extended cytoreductive procedures for management of advanced stage ovarian cancer
Ovarian cancer is a locoregional disease with locoregional effects leading to symptoms and presentation typical of advanced ovarian cancer.
Typically hysterectomy to facilitate dissection off the rectum (Fig. 7).
The pelvic disease process may invade into the pelvic colon or its mesentery or present with impending obstruction. Surgery may require primary en-bloc resection of the rectosigmoid colon with low or ultra-low anastomosis to relieve the obstruction or to achieve optimal debulking (Fig. 8). The technique and results of en-bloc rectosigmoid resection have been well detailed in the literature and the procedure is not without its risks. 51  Surgical skills for colorectal and urologic procedures are shown in Table 6.

| FACTORS ASSOCIATED WITH SURGICAL OUTCOMES
It is a well-established fact that surgical, especially oncologic, outcomes vary between surgeons and institutions. 56 This is true for most surgical oncology specialties. Ovarian cancer is one of the conditions for which the impact of surgeon specialty and place of surgery has been addressed extensively in the literature from different parts of the world. In a retrospective review of cases, gynecologic oncologists were more likely to achieve complete or optimal debulking for ovarian cancer and were more likely to adhere to the guidelines for postoperative treatment with platinum-based chemotherapy. 1,2 The overall survival was significantly better for the cohort of women treated by a gynecologic oncologist. A systematic review of impact of surgeon specialty and institutional specialty on survival following ovarian cancer treatment found a similar trend. The survival was better when a gynecologic oncologist treated women diagnosed with ovarian cancer; this difference was more profound for advanced stage disease. The impact of the institutional specialty showed a similar trend although did not reach statistical significance. 1 Another review into the impact of hospital and physician volume of work on the outcome of women diagnosed with advanced stage ovarian cancer found that women treated by high-volume physicians (those with a case load of more than 10 cases per year of advanced F I G U R E 7 Resection of pelvic peritoneal disease. F I G U R E 8 En-bloc resection of rectosigmoid colon with pelvic disease. ovarian cancer) in high-volume hospitals (those with a case load of more than 20 cases per year) had a much better outcome than any other combination of physician and hospital case load.

| CONCLUSION
We have alluded to ongoing training and mentorship in this article. As surgical techniques evolve and newer technologies are introduced, gynecologic oncologists will need to seek additional training from their peers as ongoing process in not only the immediate post-fellowship phase of a young consultant, but as an ongoing process that addresses not only their surgical skills but also the decision making associated with complex presentations and myriad management options. There is a lesson to be learnt in ongoing learning from the mentorship program introduced at the MD Anderson Cancer Centre. 59 The program highlights the need to support a young gynecologic oncologist even after a grueling 3-4-year fellowship program.
The key message from this article is that surgical skill is a continuous spectrum and achieving expertise in complex pelvic surgery is a continuous process that develops over several years. It requires ongoing mentorship, self-reflection, and seeking training opportunities with peers even when the surgical role has transitioned from a trainee to a trainer.

AUTHOR CONTRIBUTIONS
VA and SSP contributed equally to the literature search, concept and design, and writing of the paper.

ACKNOWLEDGMENTS
The clinical photographs presented are from the authors' personal library and were taken with the consent of the patients.