Sentinel lymph node mapping in the modern management of gynaecological malignancy

Sentinel lymph node (SLN) biopsy is an alternative to systematic lymphadenectomy in the surgical staging of gynaecological malignancy. It is recommended in the management of vulval cancer and is increasingly used in endometrial and cervical cancer. SLN failed mapping algorithms require that side‐specific lymphadenectomy should be performed in the case of failed mapping, and that suspicious lymph nodes are removed. Ultrastaging protocols improve detection of lymph‐node metastasis and should be used for the pathological processing of SLNs.

Sentinel lymph node (SLN) biopsy is an alternative to systematic lymphadenectomy in the surgical staging of gynaecological malignancy. It is recommended in the management of vulval cancer and is increasingly used in endometrial and cervical cancer. SLN failed mapping algorithms require that side-specific lymphadenectomy should be performed in the case of failed mapping, and that suspicious lymph nodes are removed. Ultrastaging protocols improve detection of lymph-node metastasis and should be used for the pathological processing of SLNs.

Learning objectives
To understand the rationale, use and evidence for SLN biopsy in women with vulval, cervical and endometrial malignancy.
To appreciate the concepts behind the technical performance of SLN biopsy. To understand the pathological processing of SLN samples and the implications this may have for the interpretation of results and adjuvant therapy.

Introduction
Sentinel lymph node (SLN) mapping is an alternative to complete lymphadenectomy in the surgical staging of gynaecological cancer. Its usage aims to reduce the extent of surgery and the associated complications of systematic inguinofemoral, pelvic and para-aortic lymphadenectomy. 1 SLN mapping was first reported in the 1970s, 2 and subsequent technical advances have led to the adoption of SLN biopsy in the surgical staging of various solid tumours, including melanoma and breast cancer. 3 In gynaecological malignancy, SLN biopsy was first widely adopted into treatment guidelines for vulval cancer and is used increasingly in the management of early-stage cervical and endometrial cancer. [4][5][6][7][8][9] Lymphatic drainage of the female pelvic organs The lymphatic drainage of the vulva is primarily to the ipsilateral inguinofemoral lymph nodes, but it can cross the midline if the primary tumour involves midline structures. 1 Superficial lymph nodes (typically [8][9][10] are identified medial to the saphenous vein above the cribiform fascia, with deep nodes found deep to the cribiform fascia. 1 The most cephalad deep node under the inguinal ligament is known as Cloquet's node, which drains to the external iliac lymph node. 1 The lymphatic drainage of the cervix is primarily to the paracervical, parametrial, obturator, hypogastric, external iliac and presacral lymph nodes and secondarily to the common iliac, para-aortic and inguinal lymph nodes, 1 with the sentinel node identified most frequently in the iliac, obturator and parametrial regions. 10 The uterus has a complex bilateral lymphatic drainage system, with the lower segment draining to the pelvic lymph nodes and the upper segment draining into the para-aortic nodes via the ovarian lymphatics. 1 Metastatic lymph node positivity can be sub-categorised as macrometastases (tumour deposit >2 mm), micrometastases (tumour deposit 0.2 mm to 2 mm), and isolated tumour cells (tumour deposit <0.2 mm). 11 The sentinel lymph node concept The SLN is defined as the first lymph node involved in the drainage of lymph from the primary tumour to the regional lymph node basin. 3 To perform SLN mapping, a tracer is injected into the area surrounding the tumour. The tracer travels through lymphatic channels to regional lymph nodes and the first draining node can be detected and removed. 3 If this node does not contain metastatic cells, non-sentinel nodes can be presumed to also be negative, according to the reported negative predictive value for the specific SLN procedure being undertaken. 3 To evaluate the success of SLN mapping, the main outcomes include detection rate (percentage of patients where the SLN is detected), sensitivity and false negative rate. 12,13 False negatives are important because they represent a failure to detect metastatic disease, resulting in downstaging and potential omission of adjuvant treatment, and these should represent less than 5% of cases. 13 SLN biopsy can replace either full lymphadenectomy, or alternatively no lymph node sampling, in patients without radiological evidence of lymphadenopathy.
Significant morbidity, particularly lymphoedema, is associated with systematic lymphadenectomy in gynae cological malignancy. 14 Lymphoedema is a progressive lifelong condition that can be associated with chronic pain, cellulitis, cosmetic disfigurement and reduced mobility. 14 It has been shown that in vulval cancer, SLN biopsy reduced the risk of lymphoedema to 1.9% from 25.2%, and further reductions were identified in rates of wound breakdown and cellulitis and the need for compression stockings. 15,16 A quality of life benefit from SLN biopsy has not yet been established, despite a reduction in surgical morbidity, 15 however several studies have reported poorer body image, sexual and cognitive function and increased fatigue in women following inguinofemoral lymphadenectomy compared with SLN biopsy. [15][16][17] Pelvic lymphadenectomy is associated with lymphocyst formation and a higher risk of intraoperative complications (nerve and vascular injury) as well as increased blood loss and prolonged operative duration. 18,19 Sentinel lymph node tracers and detection rates

Mapping tracers
Several tracers have been developed to detect SLNs in gynaecological cancer. Tracers have been studied for use either alone or in combination.

Blue dye
Blue dyes (isosulfan blue, methylene blue and patent blue V) are approved for SLN mapping. Blue dye is injected around the tumour site and is detectable in SLNs within 10-20 minutes and visualised consistently for up to 30 minutes, after which the ability to identify blue dye decreases until it dissipates after approximately 50 minutes; therefore, dissection and identification of the sentinel node must be completed within this timeframe to avoid missing the SLN. 20,21 Blue dyes are considered safe, although administration can result in skin discolouration, reductions in non-invasively monitored peripheral oxygen saturation measurements and anaphylactic reactions in 1.1% of patients. 22,23 Radioactive colloid Radiolabelled colloid technetium 99 (Tc99) is injected around the tumour site, and high signal can be detected with a hand-held gamma probe with the option of preoperative SPECT-CT or lymphoscintigraphy to aid operative planning. 24 Radioactive tracers need to be injected prior to surgery as they can take longer to reach the sentinel node and are optimally detected between 1-6 hours following injection. 1,24 Indocyanine green and near infrared imaging Indocyanine green (ICG) is a water soluble tricarbocyanine that fluoresces in near infrared (NIR) light. This fluorescence is detected by NIR imaging systems available for use in open, laparoscopic or robotic surgery. 25,26 ICG contains 5% sodium iodide and undergoes hepatic metabolism, so it should not be used in those with iodine allergy or significant liver dysfunction. [25][26][27] High body mass index can reduce detection rates. [25][26][27] Detection rates are higher when a combination of tracers are used, and ICG mapping has higher overall and bilateral detection rates than blue dyes. 28 Detection rates for each cancer site and tracer are provided in Table 1.

Failed mapping
In cervical and endometrial cancer, use of SLN mapping algorithms ( Figure 1) reduced false negative rates to 2.1% and 1.9% with a sensitivity of 92.6% and 98.1% and negative predictive value of 98.9% and 99.8% in cervical and endometrial cancer, respectively. 29,30 Further troubleshooting recommendations in the context of failed mapping include performing the lymph node dissection on the contralateral side before reattempting on the original side, reinjection of the tracer and extending the retroperitoneal dissection to evaluate the common, pre-sacral and para-aortic areas. 31 If this is unsuccessful, complete lymphadenectomy can be performed on the side with failed mapping. 31 In vulval cancer, a side-specific complete inguinofemoral lymphadenectomy can be performed with preservation of the saphenous vein. 7 For tumours with a medial edge within 1 cm of a midline structure (clitoris, urethral meatus, vagina, perineal body or anus), if SLN are not detected bilaterally, this should be considered as failed mapping and a complete lymphadenectomy considered on the side where SLN are not detected. 7

Pathological assessment
Ultrastaging Histological assessment of lymph nodes was traditionally performed by bisecting the node longitudinally and evaluating the haematoxylin and eosin (H&E)-stained appearances of each side of the node. 32 Ultrastaging protocols have been shown to improve detection of lymph node metastasis by 10-15% in cervical cancer [33][34][35] and 37-43% in endometrial cancer 32,36 and are recommended for use in SLN assessment. 4,5,7,8 Ultrastaging involves serial sectioning of the SLN into 50-250 lMsections, and assessing each slice wth H&E. 33 Additional immunohistochemistry (pancytokeratin antibody AE1-AE3) is performed on a single slice. 32 Slicing can be performed either longitudinally or in a 'bread-loaf' fashion by slicing perpendicularly to the longitudinal axis. 32 Low volume metastases (micrometastases and isolated tumour cells [ITCs]), which would not have been identified using traditional techniques, are identified through ultrastaging, resulting in upstaging of disease. However, the clinical significance of these and the impact on adjuvant treatment is yet to be determined. ICG = indocyanine green; NR = not reported (pooled false negative rates were not reported in these meta-analyses in cervical cancer).

Vulval cancer
SLN biopsy is now recommended as standard-of-care in patients with unifocal tumours <4 cm in size in the absence of suspicious groin nodes. For tumours within 1 cm of the midline, bilateral detection is mandatory, otherwise a sidespecific lymphadenectomy should be performed. 4,7 Surgical technique International guidelines recommend the use of radioactive tracers alone or in combination with blue dye. 4,7 A typical example is 4 ml of blue dye and 2-4 ml of 1-2.5 mCi Technetium-99 in four 1-ml aliquots injected intradermally in four orthogonal peritumoral injections. 24 Deeper injection should be avoided as this may result in the tracer draining along the deeper lymphatics to the pelvic nodal basin, rather than to the inguinofemoral nodes. 24 Technetium-99 should be injected pre-operatively to allow sufficient time to reach the nodes, with blue dye injected immediately prior to groin dissection at the time of surgery under anaesthesia. 24 Topical local anaesthetic creams can be applied 30-45 minutes prior to the administration of Technetium-99 and Entonox â can be offered, as vulval intra-dermal injections can be painful. An incision is made over the area displaying the highest signal using handheld gamma probes and careful dissection is then performed to identify lymphatic channels and sentinel node. 24 The gamma probe should be used to confirm radioactivity of the sentinel node post-dissection and to confirm there is no significant residual tracer in the lymphatic basin. Sentinel nodes are typically defined as having a ratio of at least 10:1 (SLN count:background count). 3 If multiple nodes are identified, all of these should be removed. 24 Residual radioactivity should usually be <10% of the counting rate in the sentinel node. 3 Frozen section and identification of positive sentinel nodes could avoid a second operation for completion lymphadenectomy, and this approach is accepted in current European guidelines, although caution is advised owing to the potential loss of tissue in sample processing, which may increase the risk of missing micrometastasis. 4 The GROINSS-V II trial identified that patients with positive sentinel nodes should have further groin treatment irrespective of node size, as patients with micrometastases who did not receive further adjuvant treatment had higher groin recurrence rates. 37 Macrometastases should be treated with inguinofemoral lymphadenectomy, as radiotherapy was associated with more groin recurrences. 38 However, for micrometastases, inguinofemoral radiotherapy appeared to be a safe alternative to inguinofemoral lymphadenectomy, with reduced toxicity and morbidity. 38,39 Oncological safety Two key studies provided the safety evidence for the establishment of SLN biopsy in vulval cancer: the GOG-173 and the GROINSS-V studies (Table 2). 40,41 In the GROINSS-V multicentre observational study, 41 inguinofemoral lymphadenectomy was performed in the case of metastatic disease, and in the absence of nodal involvement, patients (n = 259) were followed up for 2 years at 2-monthly intervals. 41 The recurrence rate was 2.7% with a 3-year survival rate of 97% and a reduction in short-and long-term morbidity. 41 The long-term survival data from this study showed disease-specific survival of 91% for SLN-negative women compared with 65% for SLN-positive women. 42 Despite this, the vulval recurrence rate was 36.4% for SLNnegative women and 46.4% for SLN-positive women, and disease-specific survival is far lower in women with vulval recurrence (69%) compared with those without (90%). 42 A meta-analysis evaluating the accuracy of SLN biopsy in vulval cancer reported detection rates between 94-97.7% and sensitivity of 91% with a negative predictive value (NPV) of 95.6% in those studies with clinical follow-up data available for SLN-negative patients, concluding that SLN biopsy with Tc-99 and blue dye with ultrastaging is safe in carefully selected patients with close follow-up. 43 Cervical cancer

Surgical technique
Technetium-99 with blue dye, or ICG, should be injected into the cervix either in two or four positions ( Figure 2). 7,24 Following inspection of the pelvic cavity and opening of the  peritoneum, 'hot' nodes can be detected using a gamma probe; if ICG is used, the surgeon will be able to identify green fluorescence in the lymphatic tracts leading to the fluoresced sentinel node. 24 Radioactivity or fluorescence should be confirmed ex-vivo following node removal. 24 A key priority in cervical cancer is avoiding dual-treatment with radical surgery and chemoradiation; therefore, the value of frozen section of SLNs has been explored. If positive lymph nodes are identified intraoperatively, radical hysterectomy can be abandoned and the patient referred for primary chemoradiation. However, the initial results of the prospective SENTIX trial reported poor sensitivity of frozen section for the detection of macrometastasis (75.9%), and macrometastasis and micrometastasis combined (45.8%). 44 Frozen section failed to detect 50% of nodepositive disease and should not be recommended in cervical cancer. 44

Oncological safety
In cervical cancer, lymph node status is an important prognostic indicator, and SLN mapping can be performed in early cervical cancer (1A1 with lymphovascular space invasion or 1A2-1B1 disease with tumours <2 cm) in patients recruited into registered clinical trials. 7,8 Positive lymph nodes are identified in 0-22% of women with stage 1A-1B cervical cancer, 45 and therefore up to 80% of women are undergoing pelvic lymphadenectomy (with the associated morbidity) for negative nodes. The prospective SENTICOL study ( Table 2) demonstrated high detection rates, sensitivity and NPV for SLN biopsy in early-stage cervical cancer, with a bilateral detection rate of 76.5%. No false negatives were identified when SLNs were identified bilaterally, highlighting the role of SLN algorithms for failed mapping. 46 A retrospective cohort study also reported high sensitivity (96.4%) and NPV (99.3%) with no difference in detection rates between mapping agents, surgical approach, previous conisation and size of tumour (up to 4 cm). 47 The SENTIX trial ( Table 2) demonstrated that all sentinel nodes were detected in the pelvic external iliac, interiliac, common iliac and presacral regions. 44 The SENTICOL III trial 49 is a prospective multicentre randomised study comparing disease-free survival and health-related quality of life in patients with negative SLNalone versus patients with negative SLN and pelvic lymphadenectomy. The trial aims to validate the use of SLN biopsy in early-stage cervical cancer, assuming noninferiority of SLN biopsy in terms of disease-free survival and superiority of SLN biopsy in health-related quality of life. 48 Survival data from the SENTIX and SENTICOL III trials is anticipated to provide an evidence-base for which to prospectively validate the safety of SLN biopsy in earlystage cervical cancer. 11,44,48 A recent meta-analysis reported survival outcomes comparing women with micrometastases or ITCs and women with micrometastases-only to those with negative nodes. 49 In terms of disease-free survival, the hazard ratio (HR) was 2.6 (95% CI 1.6-4.3) and 4.1 (95% CI 2.7-6.2), respectively 49 and for overall survival, the HR was 5.7 (95% CI 2.8-11.4) and 6.9 (95% CI 2.6-18.8), respectively. 49 This suggests that low-volume metastatic nodal disease appears to have a negative impact on survival in cervical cancer. However, interpretation of these findings is limited by a lack of information on the adjuvant therapy in the included studies, and the limited prospective data available do not appear to show a clear benefit to adjuvant therapy in terms of recurrence-free survival in this setting. 50,51 A small prospective study of 19 patients with micrometastases or ITCs found no differences in recurrence-free survival, irrespective of adjuvant therapy. 51 Prospective data from the SENTIX and SENTICOL III trials will provide further information on patient outcomes in the setting of lowvolume metastatic disease. 44,48

Surgical technique
Moloney et al. 32 performed an international Delphi exercise to achieve consensus on the mandatory, optional and prohibited steps when performing SLN biopsy for endometrial cancer. They recommended superficial injection of ICG in two or four positions on the ectocervix (Figure 1) using a 20-25 gauge needle, noting that the surgeon should feel resistance at the time of injection. 31 Although different injection sites (uterine subserosal and hysteroscopic endometrial) have been evaluated, the rationale for cervical injection is that drainage of the lymphatics occurs mainly through the parametria and there is increased likelihood of preserved normal anatomy and improved access for the surgical team at the cervix. 24 If a uterine manipulator is used, this should be inserted after injection of the tracer. 31 Dissection should be commenced at the level of the uterine artery and continued in a lateral direction, away from the uterus, using a blunt or electrosurgical technique. 31 The sentinel node is defined as the most proximal node to the uterus, irrespective of the nodal station, and is a single mapped node or single node plus its next station echelon node. 31 Nodes should be examined exvivo to confirm green fluorescence. 31 Frozen section is not recommended in endometrial cancer; the SENTOR prospective cohort study and the SENTI-ENDO trial reported poor sensitivity for frozen section detection of lymph node metastasis. 52,53 One potential benefit is the intraoperative detection of 'empty nodes' where no lymphatic tissue is present in the specimen. 54 Oncological safety SLN mapping is increasingly the standard-of-care in endometrial cancer, and in early-stage endometrial cancer SLN biopsy is now listed as a European quality-standard indicator for endometrial cancer surgery, with a target of 90%. 55 The British Gynaecological Cancer Society guidelines and SLN consensus statement advise that SLN mapping can be used as an alternative to lymphadenectomy in the management of endometrial cancer provided that ultrastaging and complete lymphadenectomy for unmapped cases are performed. 7,9 The SENTI-ENDO and FIRES studies demonstrated high sensitivity for the performance of SLN biopsy in early-stage endometrial cancer with low false-negative rates (Table 2), 56,57 and a Cochrane review, which pooled 33 studies, reported a mean detection rate of 86.9% with a sensitivity of 91.8%. 12 However, data are limited to diagnostic accuracy or prognostic studies and two randomised controlled trials (RCTs), the Sentinel Node Biopsy in Endometrial Cancer (ENDO 3) trial and the SELECT (Sentinel Lymph Node Endometrial Cancer) trial examining the role of SLN in the treatment pathway, are ongoing. 58,59 A recent systematic review described three studies where fewer patients received adjuvant therapy following SLN biopsy compared with systematic lymphadenectomy and two studies where no differences in the rates of adjuvant treatment were observed. 18 A further two studies compared no nodal dissection with SLN biopsy; in one study there was no difference between adjuvant therapy rates, in the second study radiation treatment rates increased in the SLN group compared with no nodal dissection. 18,19,60 The authors noted that high-risk tumour factors were a greater determinant of adjuvant therapy than lymph node dissection, although this data was from heterogeneous groups of patients. 18 In early-stage endometrial cancer, a meta-analysis of two RCTs found that pelvic lymphadenectomy does not improve either overall or relapse-free survival and is associated with an increase in surgical morbidity. 61 Therefore, the prognostic value and therapeutic benefit of performing a systematic pelvic lymphadenectomy in the context of failed mapping, particularly for low-and intermediate-risk disease, is yet to be determined, and local protocols may reflect this to allow clinicians to take an individualised approach based on uterine factors and patient choice. 62 Furthermore, with the recent adoption of the molecular classification of endometrial cancer into national and international guidelines, 5,9 the role of lymphadenectomy or lymph-node staging in determining adjuvant therapy in addition to uterine and molecular factors needs further exploration.
The prognostic value of systematic pelvic and para-aortic lymphadenectomy compared to pelvic lymphadenectomyalone or no lymphadenectomy has not been evaluated in prospective RCTs. 9 A systematic review of 13 retrospective observational studies (at critical risk of bias) comprising 7349 patients reported a 46% reduction in risk of death (HR 0.54, 95% CI 0.35-0.83) and 49% reduced risk of recurrence (HR 0.51, 95% CI 0.28-0.93) in women with intermediate or highrisk endometrial cancer who underwent combined pelvic and para-aortic lymphadenectomy compared with pelviclymphadenectomy alone. 63 Isolated para-aortic nodal metastases in the absence of pelvic nodal metastases are reported to occur in less than 3% of cases, [64][65][66] and SLNs are only identified in the para-aortic region in around 5% of cases following cervical injection. 67 Broader understanding of the role of para-aortic lymphadenectomy in the surgical staging of endometrial cancer and the impact of this on patient outcomes is needed to address whether a para-aortic nodal assessment (either by SLN biopsy or by systematic lymphadenectomy) is required and for which groups of patients. The prospective Endometrial Cancer Lymphadenectomy Trial (ECLAT) (ClinicalTrials.gov Identifier: NCT03438474 trial) is open to help address this question.
Endometrial cancer patients with micrometastases or ITCs have improved progression-free survival compared with those with macrometastases; 68 however, a meta-analysis identified a higher relative risk of recurrence (1.34, 95% CI 1.07-1.67) in women with low-volume metastatic disease (micrometastases and ITCs) compared with node-negative women, even if adjuvant therapy was given, although prognostic uterine factors were not controlled for in the included studies. 69 The largest retrospective multicentre cohort study of 175 women with SLN ITCs with a median follow-up time of 31 months reported chemotherapy and EBRT were not associated with improved recurrence-free survival (HR = 0.63, 95%CI 0.11-3.52, and HR = 0.90, 95%CI 0.22-3.61, respectively) after controlling for stage, grade and lymphovascular space invasion, although the overall recurrence rate in this population was low (5.1%). 70 Plante et al. 71 conducted a prospective series, which included 31 patients with ITCs; at a median follow up of 29 months, the recurrence-free survival was similar (95.5%) to the node-negative (87.6%) and micrometastases (85.5%) groups. In women with ITCs, 35% received adjuvant chemoradiation, 32% received EBRT and 32% received either vaginal brachytherapy or no adjuvant treatment, and only one recurrence was detected in a patient with other high-risk uterine factors (stage IB, carcinosarcoma). 71 Taken together, this suggests that women with SLN ITCs appear to have good prognosis, irrespective of adjuvant therapy, and therefore decisions regarding adjuvant therapy should be based on uterine and pathological factors, rather than lymph node status. Prospective multinational registry database studies would provide further data to determine the impact of low-volume metastatic disease on patient-outcomes in endometrial cancer.

Implementation into clinical practice
There is a well-recognised learning curve associated with performing SLN biopsy. SLN detection rates increase with surgical experience, 72,73 and although the precise number of procedures required to exceed the learning curve is not yet established, efforts are being made to standardise surgical techniques in SLN. A competency assessment tool has been validated for use in endometrial cancer, which outlines the mandatory, optional and prohibited steps of SLN biopsy in this setting. 7,31 Surgeons involved in SLN biopsy should therefore audit their performance, surgical outcome, morbidity and survival data and be encouraged to recruit patients into prospective trials. National quality-assurance standards and full open reporting of outcomes are desirable. SLN biopsy is associated with lower costs compared with systematic lymph node dissection. 18 Conclusions and future directions SLN biopsy is expanding in gynaecology oncology and is used increasingly as an alternative to systematic lymphadenectomy, particularly in vulval cancer where SLN biopsy is now considered to be the standard-of-care. Although the diagnostic accuracy of SLN biopsy is high for cervical and endometrial cancer, key trials are awaited on the oncological outcomes following SLN biopsy to provide an evidence base for the safety of SLN mapping. Data on patient outcomes, including adverse events, adjuvant therapy, quality of life and patient-reported outcome measures are required to inform decision-making on the role of SLN in the treatment pathway of these women. Future research into the prognostic significance of micrometastases and ITCs and the optimal management of this cohort of women would be desirable.

Disclosure of interests
AC is a trainee member of the RCOG e-Learning Editorial Board (gynaecology oncology module). AP is treasurer of the British Gynaecological Cancer Society.

Contribution to authorship
AC and AP instigated the article. AC researched and wrote the article. AP researched and edited the article. Both authors approved the final version.