Lymphatic Mapping and Sentinel Node Biopsy in Vulvar, Vaginal, and Cervical Cancers
Lymphatic Mapping and Sentinel Node Biopsy in Vulvar, Vaginal, and Cervical Cancers
In 1977, Cabanas reported his initial experience with lymphatic mapping and sentinel lymph node detection in men with penile cancer. He hypothesized that tumors spread from the primary lesion to the draining nodal basins in an orderly, predictable pattern and that tumor emboli would not "skip" primary draining nodal basins for upper-echelon nodes. These initial draining basins were termed "sentinel nodes," as they were metaphorically the lookouts for all upper-echelon, or nonsentinel, nodes. A more modern and succinct definition of a sentinel lymph node is "any lymph node that receives lymphatic draining directly from a tumor site."
Although the early pioneering studies in lymphatic mapping and sentinel lymph node biopsy were in patients with penile cancer, it was breast and melanoma surgeons who led the majority of the research in the 1980s and 1990s.[3,4] Currently, lymphatic mapping and sentinel lymph node biopsy, in lieu of complete regional lymphadenectomy, are the standard of care for breast cancers and melanomas. This technique has also been explored in other malignancies, including cancers of the head and neck, oropharynx, thyroid, stomach, colon, and bladder. In the past 15 years, lymphatic mapping has been explored in almost all of the gynecologic malignancies, including those of the vulva, vagina, cervix, and uterus.
A variety of technologies exist to help surgeons locate the sentinel nodes. Preoperative lymphoscintigraphy uses a radiolabeled colloid injected peritumorally to identify the sentinel node. A scanning machine detects the gamma emissions from the colloid (typically technetium-99) after it has drained to the sentinel lymph node. Gamma emissions from a radiolabeled colloid can also be followed intraoperatively using a handheld gamma probe to pinpoint the sentinel node. Patent blue dyes may also be injected peritumorally for visual identification of sentinel nodes during surgery. Resected sentinel nodes may therefore be "hot" (positive for the radiolabeled colloid), "blue" (positive for patent blue dye), or both hot and blue. Most studies of lymphatic mapping for breast cancer and melanoma have found that a combination of radiolabeled colloid and blue dye leads to higher sentinel node detection rates. This appears to hold true for the gynecologic malignancies, as will be discussed in this review.
The objective of this article is to review the current literature on lymphatic mapping and sentinel node detection in women with gynecologic malignancies. We will pay particular attention to the rationale for mapping, the reliability of the techniques, the localization of sentinel nodes, and the utility of pretherapeutic lymphoscintigraphy. Among studies of cervical and vulvar cancers, there are multiple case reports, small series (< 20 patients), and large series (≥ 20 patients). We included only those series with more than 20 patients in our review, since a long-recognized limitation of lymphatic mapping is the higher rate of false-negative sentinel node detection in the first few cases a surgeon performs.[5,6] By restricting the eligible studies in this manner, we hoped to assure that the individual surgeons in each reported series would have obtained proficiency (ie, master the "learning curve") in sentinel lymph node mapping for gynecologic malignancies.
An estimated 3,490 new cases of vulvar cancer were diagnosed in the United States in 2007, and an estimated 880 deaths resulted from the disease. Most vulvar cancers have squamous cell histologies (more than 90%), with melanoma accounting for the majority of the remaining tumors.
For patients with clinical stage I squamous cell carcinoma of the vulva, the risk of lymph node metastasis is 10.7%. This risk increases to 26.2% for clinical stage II disease and to 64.2% for clinical stage III disease. The lymphatic drainage of the vulva is almost exclusively to the inguinofemoral triangle, which could be considered a sentinel lymph node basin. Although some anatomists have hypothesized that lymph drains directly from the vulva to the pelvis, this route has never been demonstrated clinically. Unilateral vulvar lesions, typically defined as tumors located more than 2 cm from the midline, primarily drain primarily to the ipsilateral groin nodes.
Unilateral and bilateral inguinofemoral lymphadenectomies have relatively high rates of postoperative complications, with as many as two-thirds of patients experiencing wound breakdown, lymphocyst formation, and/or lymphedema as a result of their groin dissection. For this reason, and because vulvar cancers have a highly predictable anatomic drainage pattern, lymphatic mapping and sentinel node dissection are seemingly ideal for this disease site.
Success of Lymphatic Mapping
Multiple investigators have published their experiences with lymphatic mapping and sentinel node biopsy in women with vulvar cancer.[10-19] Table 1 lists case series in the English-language literature that included more than 20 patients in whom sentinel node biopsy was followed by complete inguinofemoral lymphadenectomy and that provided enough data to calculate the sensitivity, false-negative rate, and negative predictive value of sentinel node biopsy.
At least one sentinel node was identified in 97% of the patients reported. However, a sentinel node was identified in only 84% of groins explored. This is likely a result of the standard practice of dissecting groins bilaterally for vulvar lesions less than 2 cm from the midline. Many of these lesions will have only ipsilateral lymphatic drainage, so the contralateral groin will not contain a sentinel node.
For the combined group—the 383 patients in the 10 series included in Table 1—the overall sensitivity of sentinel node biopsy was 97.6%, with a false-negative rate of 2.4%. Only three patients had a truly false-negative sentinel node result—ie, the sentinel node was negative but metastatic disease was present in other nodes.[10,14] Another five patients (1.3%) had metastatic disease in the groin, but no sentinel node was identified. Whether these five cases should be considered false-negatives or technique failures is debatable. One reason that lymphatic mapping might not be successful in identifying the sentinel node is that when the node is completely replaced with tumor, radiolabeled tracers and/or blue dye cannot enter the tissue as readily. Most would argue that in the absence of identification of a sentinel node in a draining groin, a complete inguinofemoral lymphadenectomy should be performed.
Early experiences using only blue dye resulted in a low identification rate, with sentinel nodes identified upon visual inspection in only 64% of groins dissected (Table 1). Since these blue-dye–only studies were among the earliest in vulvar cancer, the low detection rate may indicate that investigators were on the early portion of the learning curve and did not have the benefit of the knowledge accumulated by the research community since then. Another reason for the low detection rate may have been the lack of a radiolabeled tracer as part of the technique. Once radiolabeled colloids were added, sentinel node identification rates increased dramatically.
Overall, the detection rate per groin is 91% for intraoperative mapping using a radiolabeled colloid combined with blue dye. For preoperative lymphoscintigraphy and intraoperative mapping using a radiolabeled colloid only, the rate is 92%, and for preoperative lymphoscintigraphy and intraoperative mapping using a radiolabeled colloid combined with blue dye, the rate is 90% (Table 1).
As shown in Table 1, even the largest validation studies for this technique are relatively small. Currently in the United States, the Gynecologic Oncology Group (GOG) is prospectively evaluating the sensitivity, false-negative rate, and negative predictive value of lymphatic mapping and sentinel lymph node dissection in a multi-institutional study (GOG 173). The goal is to complete patient accrual in 2008.
Gynecologic oncologists in Europe recently reported their data from the GROningen INternational Study on Sentinel nodes in Vulvar cancer (GROINSS-V). In contrast to the validation study underway by the GOG, the GROINSS-V study is an observational study in which the investigators performed sentinel node–only biopsies instead of complete inguinofemoral lymphadenectomy in women with squamous cell carcinomas of the vulva ≤ 4 cm in size. At a median follow-up of 35 months, they observed 6 groin recurrences (2.3%) in 259 women with unifocal vulvar disease who had negative sentinel node biopsies. Interestingly, this reported recurrence rate of 2.3% is almost identical to the false-negative rate of 2.4% shown for the combined studies in Table 1.
Other smaller studies have also reported their experience with sentinel node–only biopsy in lieu of complete lymphadenectomy in women with vulvar cancer. Vidal-Sicart and colleagues reported a prospective study of 70 patients with vulvar cancer who underwent lymphatic mapping and sentinel node biopsy. The first 50 patients were considered the validation group, in which lymphatic mapping and sentinel node biopsy were followed by complete inguinofemoral lymphadenectomy. The next cohort of 20 women was the application group, in which only sentinel node biopsy was performed.
In 7 of these 20 patients, metastatic disease was found in the sentinel lymph node; complete lymphadenectomy followed by adjuvant chemotherapy and/or radiation was then performed in six women. In the seventh, only isolated tumor cells were found in the sentinel lymph nodes, so adjuvant radiation therapy was administered without reexploration of the groin. The remaining 13 patients without disease in their sentinel nodes were assigned to close observation.
At a median follow-up of 10 months, no groin recurrences had been observed in these women. After publication of the manuscript, however, and 12 months after surgery, one patient with a negative sentinel lymph node biopsy had a recurrence in the same inguinofemoral basin in which the sentinel node was identified. She underwent complete lymphadenectomy and, at the time of this writing, has no evidence of disease (personal communication, Vidal-Sicart).
A similar experience with sentinel node biopsy for T1 squamous cell carcinomas of the vulva was reported by Terada et al. Of the 21 patients who underwent sentinel lymph node biopsy only, 3 had metastatic disease in the sentinel nodes. At a median follow-up of 4.6 years, none of the 18 patients with a negative sentinel node biopsy had had a groin recurrence, although 2 had local recurrence on the vulva.
Anatomic Location of Sentinel Nodes
As previously stated, no study has reported direct drainage of lymph from the vulva to the pelvis. The location of the sentinel node within the inguinofemoral triangle, however, varies greatly. Our group previously suggested that groin recurrence in women with no sentinel node identified and a negative lymphadenectomy specimen may be associated with insufficient dissection of the groin, most likely with nodal tissue left in the medial part of the inguinofemoral triangle. Locating the sentinel node (the most likely site of metastatic disease) is of the utmost importance.
Rob et al published the most detailed study about sentinel node location in vulvar cancer. They divided the lymph nodes of the groin into four regions: the superficial medial group (Sf-M), located above and medial to the femoral vein and medial to the saphenous vein; the superficial intermediate group (Sf-IM), located in the vicinity of but superior and lateral to the saphenous and femoral veins; the superficial lateral group (Sf-L), located in the outer third of the groin; and the profundum nodes group (Pf), located deep, medially along the femoral vein. Of the 118 sentinel lymph nodes found in 82 groins in this study, 58 (49%) were in the Sf-M region, 41 (35%) were in the Sf-IM region, and 19 (16%) were located in the Pf region. No sentinel lymph nodes were found in the Sf-L region. These data are important for anyone who performs groin dissections, even those who do not perform lymphatic mapping.
Utility of Pretherapeutic Lymphoscintigraphy
It is not certain whether preoperative lymphoscintigraphy for the detection of sentinel nodes in vulvar cancer is necessary. Multiple arguments exist for not performing the procedure. First, consistent drainage of primary vulvar lesions to the inguinofemoral triangle suggests that the scan might add little to surgical planning. This is in contrast to, for example, cutaneous melanoma of the trunk, for which lymphatic drainage is unpredictable and the surgical approach changes according to the lymphoscintigraphy findings. Another argument against routine use of lymphoscintigraphy is the added cost, time, and patient discomfort involved. Patients are required to visit the nuclear medicine clinic hours or days before the surgical procedure and undergo injection with the radiolabeled colloid around the vulvar lesion, without an analgesic.
For well-lateralized lesions (for which only unilateral groin dissection is planned) and for lesions crossing the midline (for which bilateral groin dissections are indicated), preoperative lymphoscintigraphy probably is unnecessary. However, it may be useful for lesions near the midline for which bilateral dissection is being considered. Hauspy et al reported identification of bilateral sentinel lymph nodes in only 6 (46%) of 13 patients with lesions "close to midline," as opposed to in 13 (93%) of 14 patients with lesions abutting or crossing the midline. For 12 (92%) of those 13 women with lesions "close to midline"—defined as a location less than 1 cm from but not touching the midline—sentinel lymph nodes were found only in the ipsilateral groin.
Perhaps preoperative lymphoscintigraphy could have identified some women with lesions "close to midline" at low risk for bilateral drainage who could have safely undergone ipsilateral groin dissection only. However, one must be cautious in abandoning standard-of-care bilateral groin dissections for lesions near the midline based on lymphoscintigraphy results only. Louis-Sylvestre and colleagues found that in 13 patients who had lesions less than 1 cm from the midline and in whom lymphoscintigraphy identified unilateral drainage only, 3 had metastatic disease in nodes located in the contralateral, lymphoscintigraphy-negative groin.