Sentinel Lymph Node Mapping for Endometrial Cancer

Importance of Lymph Node Assessment

Endometrial cancer is the most common gynecologic malignancy; approximately 10% to 15% of patients will have metastatic nodal disease, and nearly 15% of patients with grade 1 tumors at presentation preoperatively at office biopsy or dilation and curettage will actually have higher grade disease at final pathologic review after hysterectomy ; therefore it is of utmost importance to stage and treat patients properly and limit missing undetected metastatic disease that could upstage the patient’s condition and change adjuvant therapy. For most gynecologic oncology practices, a low-risk endometrial cancer is a retrospective diagnosis, meaning that one does not know that the patient is at low risk until total hysterectomy has been done and the permanent pathologic assessment has been completed.

The majority of patients with newly diagnosed endometrial cancer will undergo initial surgical treatment that will include a total hysterectomy, bilateral salpingo-oophorectomy, and pelvic washings. Proper surgical staging, one of the most important prognostic factors, provides information on the actual extent of disease rather than on perceived risks based on uterine factors such as grade, histologic type, and depth of myometrial invasion, which helps tailor adjuvant therapy. Endometrial cancer frequently develops after menopause, and comprehensive lymphadenectomy in elderly women may be associated with side effects, such as lower extremity lymphedema and lymphocyst formation, which can negatively affect quality of life. The more pelvic lymph nodes removed, the greater the likelihood a patient will develop these side effects. The importance of lymph node assessment for proper surgical staging in this patient population cannot be stressed enough. In a study of 1289 patients with uterine corpus malignancies, 16 (3.4%) of 469 patients who had 10 or more lymph nodes removed at operation developed new postoperative symptomatic leg lymphedema, and this is likely an underestimation of the true incidence. Complete lymphadenectomy is also likely associated with greater operating time, prolonged anesthesia, and other potential untoward effects such as blood loss, vascular and nerve injury, and increased conversion rate from laparoscopy to laparotomy to complete the operation successfully. However, because accurate surgical staging is one of the most important prognostic factors, staging with the SLN algorithm will provide the necessary pathologic information in the majority of women with apparent uterine-confined disease while limiting morbidity. The SLN algorithm, when applied for staging of all patients with newly diagnosed endometrial cancer, will at a minimum permit bilateral pelvic nodal assessment as part of the surgical staging. The SLN algorithm also increases surgical precision and avoids the “circumflex iliac” lymph nodes, which were often removed during routine bilateral pelvic lymphadenectomy. These nodes are usually benign, especially when other nodal areas are also negative. Removing these nodes frequently causes lymphatic obstruction in the lower extremity, increasing the risk of leg lymphedema.

Unfortunately, many patients with early-stage endometrial cancer will undergo surgical procedures with insufficient nodal evaluation, with their nodes being palpated (with biopsy performed if they are enlarged), sampled, or completely ignored. Studies have shown rates of nodal assessment as low as 30%, although that percentage has increased as the importance of nodal assessment has been realized. Rates are higher when a gynecologic oncologist, as opposed to a general gynecologist, performs the operation (83% vs. 26% in one study). When surgical staging is inadequately performed, patients can be subjected to unnecessary adjuvant pelvic radiation therapy and its associated side effects. An MSKCC study showed that with the increase in lymph node assessment over a 12-year time period, there was an inverse decrease in the use of adjuvant whole-pelvis radiation therapy.

In an attempt to clarify clinical variables that affect overall survival in women with endometrial cancer, the classification and regression tree (CART) method, a form of recursive partitioning, was used in a study of 1035 International Federation of Gynecology and Obstetrics (FIGO) stage I to IV endometrioid-type endometrial cancer patients. The study showed that stage, age, adjuvant therapy, and the removal of 10 or more lymph nodes were predictors of overall survival in patients with stage I to IIIA disease. Only stage was a predictor of overall survival in patients with stage IIIC to IV disease. The CART method is a tree-building technique in which “predictor” variables are analyzed to determine how they affect an “outcome” variable (overall survival). The study confirmed the importance of age and surgical staging in influencing overall survival, but the increasing number of nodes removed was not a factor, particularly when a patient was assigned stage IIIC (node-positive disease), confirming the observation that a minimum number of nodes is needed to assign stage properly, but removal of more normal-appearing nodes, particularly in stage IIIC disease, did not improve overall survival.

The use of an SLN mapping algorithm in endometrial cancer patients is an acceptable staging strategy, providing a middle ground between the polarized schools of thought: complete lymphadenectomy and no nodal evaluation. Modern studies of SLN mapping have used the cervical injection technique in the majority of cases. In a study of 42 patients with grade 1 endometrioid endometrial cancer, the most common anatomic sites at which SLNs were identified were the internal iliac (52 [36%]), external iliac (43 [30%]), obturator (34 [23%]), and common iliac (11 [8%]) regions. Only five patients (3%) had paraaortic SLN involvement. Figs. 10.2 and 10.3 demonstrate the most common and less common drainage patterns following a cervical injection of dye.

Barlin and colleagues sought to evaluate clinical and pathologic factors that influenced overall survival and to determine if a paraaortic nodal assessment at the initial staging operation in patients with endometrial cancer affected overall survival. The study of 1920 patients who had at least one lymph node removed for staging, which also used CART analysis, showed no association between the removal of paraaortic nodes and overall survival ( P = .450). The CART method did show that stage I versus stages II to IV and grades 1 or 2 versus grade 3 (a binary grading system of low vs. high grade) were predictors of overall survival. In other words, what appear to be important with regard to staging are the proper determination of uterine fundus–contained disease versus disease outside the uterine fundus and the pathologist’s determination of the grade of the tumor—low grade versus high grade (grade 1 or 2 endometrioid disease is considered low grade, and grade 3 endometrioid or serous, clear cell carcinoma, or carcinosarcoma is considered high grade).

Sentinel Lymph Node Mapping Techniques

Historically, a radioactive tracer and blue dye were used to locate “hot” nodes or colored nodes. There are three different types of SLN mapping techniques based on site of injection: (1) uterine subserosal, (2) cervical ( Fig. 10.4 ), and (3) endometrial, by means of hysteroscopy. The majority of current SLN users prefer a cervical injection. A rationale for use of a cervical injection includes the following: (1) the main lymphatic drainage to the uterus is from the parametria, and therefore a combined superficial (1–3 mm) and deep (1–2 cm) cervical injection is adequate; (2) the cervix is easily accessible; (3) the cervix in women with endometrial cancer is rarely distorted by anatomic variations, such as myomas, which sometimes make uterine serosal mapping impossible; (4) the cervix in women with endometrial cancer is rarely scarred from prior procedures such as conization or bulky tumor infiltration; and (5) uterine fundal serosal mapping does not reflect the parametrial lymphatic drainage of the uterus (the main route of drainage), and the majority of early-stage endometrial cancers do not have disease infiltrating and ulcerating the uterine fundal serosa. The main argument against cervical injection is that it has a lower paraaortic detection rate, as opposed to the hysteroscopic approach, but as is well documented, when the pelvic lymph nodes are negative for metastasis, disease is unlikely to be found in the paraaortic nodes (<5% isolated aortic nodal metastasis with negative pelvic nodes), and to date there has been no definitive, well-documented association between paraaortic nodal assessment and improved overall survival. In a large meta-analysis, Kang and colleagues reported a decrease in detection rates when the cervical method was not used and also recommended that the “subserosal injection only” technique be avoided because of decreased sensitivity. The importance of the cervical injection technique has also been recently supported by large review studies.