Curr Oncol Rep (2013) 15:559–565 DOI 10.1007/s11912-013-0345-1
GYNECOLOGIC CANCERS (NS REED, SECTION EDITOR)
Sentinel Lymph Node in Endometrial Cancer: A Review Cyril Touboul & Enrica Bentivegna & Catherine Uzan & Sebastien Gouy & Patricia Pautier & Catherine Lhommé & Pierre Duvillard & Christine Haie-Meder & Philippe Morice
Published online: 5 November 2013 # Springer Science+Business Media New York 2013
Abstract Lymph node status is a major prognostic factor in endometrial cancer (EC). Sentinel lymph node (SLN) biopsy has been reported in EC for more than 15 years but has not yet been incorporated as a standard-of-care procedure in EC. Complex uterine drainage, the various modalities of tracer injection, and the lack of large prospective series may explain this situation. In this review, we report an SLN detection rate of 81.7 %, a 10.9 % rate of metastatic SLN involvement, and a false-negative rate of 12.3 % in the main clinical trials. Thirtyfive percent of SLN metastases were low-volume disease (micrometastases or isolated tumor cells). These data raise the question of the clinical significance of low-volume disease in EC. SLN biopsy could allow upstaging in supposedly low- or C. Touboul : E. Bentivegna : C. Uzan : S. Gouy : P. Morice Department of Gynecologic Surgery, Gustave Roussy Cancer Campus, Villejuif, France P. Pautier : C. Lhommé Department of Oncology, Gustave Roussy Cancer Campus, Villejuif, France P. Duvillard Department of Pathology, Gustave Roussy Cancer Campus, Villejuif, France C. Haie-Meder Department of Radiotherapy, Gustave Roussy Cancer Campus, Villejuif, France P. Morice University Paris Sud, University de Kremlin Bicetre, Gustave Roussy Cancer Campus, Villejuif, France C. Uzan : P. Morice Unit INSERM 10-30, Villejuif, France P. Morice (*) Institut Gustave Roussy, 39 rue Camille Desmoulins, 94805 Villejuif, France e-mail: [email protected]
intermediate-risk patients in whom adjuvant therapy could be omitted. Further studies are required to precise the interest on the survival of this procedure in this subset of patients. Keywords Sentinel lymph node . Endometrial cancer . Lymphatic mapping . Nodal status . Micrometastasis
Introduction With nearly 50,000 new cases per year in the USA, endometrial cancer (EC) is the leading cause of gynecologic cancers in Western countries [1–3]. Most EC are limited to the uterus, with no evidence of extrauterine disease preoperatively [International Federation of Gynecology and Obstetrics (FIGO) stage I], thus explaining a generally favorable prognosis (95 % overall survival). Lymph node status is a crucial prognostic factor in this cancer, impairing survival when a lymph node metastasis is found [4, 5]. Recurrence-free survival (drops from 87 % without node involvement to 71 % and 36 % in women with pelvic and aortic node involvement, respectively [6]. However, the ASTEC randomized controlled trial showed no evidence of a benefit in terms of overall or recurrence-free survival for a systematic pelvic lymphadenectomy versus selective pelvic/aortic node dissection in women with early EC [2]. That study confirmed the data from an Italian randomized controlled trial in which a higher proportion of women underwent a para-aortic dissection (26 %) with no influence of lymphadenectomy on survival [7]. The indication for lymphadenectomy should therefore be limited to high-risk EC, and should be based on a preoperative biopsy and magnetic resonance imaging. Unfortunately, the relationship between the preoperative and the postoperative grade was shown to be discordant, with discrepancies in up to 46 % of cases [8–10]. Discordance in magnetic resonance imaging assessment is also considerable [11, 12], with better results since the
560
modification of the FIGO classification in 2009 [13, 14]. There is therefore a risk of surgical undertreatment or overtreatment with either the preoperative radiologic assessment or the histologic evaluation alone. Sentinel lymph node (SLN) biopsy could therefore solve this issue by acting as a “middle-of-the-road” procedure between an invasive complete pelvic and para-aortic lymphadenectomy and the absence of lymph node evaluation [15]. The concept of SLN biopsy was first developed in melanoma and breast cancer before its use in EC in 1996 [16]. SLN mapping is attractive in EC, which mainly arises in older menopausal women with comorbidities in whom a full lymphadenectomy may not be feasible or may cause further morbidity. We review here the clinical value of SLN biopsy, the injection technique used, the anatomopathological findings, and potential future indications in EC.
Article Selection Process We reviewed all articles from the first report of the technique in 1996 by Burke et al. [16], using “endometrial cancer” (EC) and “sentinel node” as search keywords in the PubMed database. We selected all clinical trials written in English reporting SLN biopsy results, and analyzed only the last publication when it was published by the same team with shared patients. Twenty-three studies fulfilled our criteria. We analyzed the type of study and year of publication, the size of the population, the disease stage, the detection method, the detection rate, the number and location of biopsied SLN, the histologic results, and the number of false negatives.
EC Drainage into SLN Unlike the case in cervical cancer where the cervix is the only injection site, several injection sites are reportedly used in EC [17–19]. Cervical, endometrial, or myometrial injections have been used in EC with different approaches: hysteroscopic, laparoscopic, and more recently transvaginal myometrial injections. These differences regarding the injection sites are relative to differences in the mode of drainage between these two cancers. Whereas cervix lymphatic drainage is exclusively via the pelvic lymph nodes, uterine corpus drainage may flow directly into the aortic lymph nodes via the ovarian vein, thus explaining the skip metastasis phenomenon. The location of the endometrial tumor is therefore determinant in disease dissemination and the site of lymph node metastasis. This review focuses on the performance of SLN biopsy and SLN mapping according to the injection site. The first series on SLN in EC used a myometrial injection into the anterior and posterior walls of the uterine fundus [16, 20]. Burke et al. [16] described, as early as this first report, blue dye
Curr Oncol Rep (2013) 15:559–565
diffusion in the uterine cornua and lateral parametria within 2–3 min of injection and along the ovarian vessels after 5–10 min. The distribution of the 31 biopsied SLN was around 40 % in the para-aortic nodes, 20 % in common iliac nodes, and 40 % in pelvic nodes [16]. Furthermore, the para-aortic SLN were above the inferior mesenteric artery, around ovarian vessel implantation, confirming drainage via this route. This is the strength of the myometrial injection, which is in line with potential lymphatic drainage of uterine corpus tumors. These results were confirmed by more recent larger series, reporting 40–50 % of patients with biopsied SLN in the para-aortic basin [21–23]. However, other reports showed discrepancies in the distribution of SLN [24–26]. For small series, this may be due to the impact of the learning curve on the overall results. Nevertheless, in the 152 patients with successfully detected SLN, 37.5 % (0–58 %) had at least one SLN in the para-aortic area (Tables 1 and 2). The study by Robova et al. [25], which compared myometrial injection with hysteroscopic injection, found 8.3 % of patients with SLN biopsied in the para-aortic area. This myometrial/subserosal injection was performed intraoperatively with blue dye during a laparotomy in most of these series. However, Torné et al. [23] performed a transvaginal ultrasound-guided injection of a radioactive tracer into the anterior and posterior uterine wall at the level of the tumor. This technique has the advantage of being consistent with tumor drainage, 45.4 % of which was para-aortic, and the detection rate exceeded 90 %. However, the technique is more complex than cervical injection, as it requires expertise in transvaginal ultrasonography. Moreover, specific equipment must be available in the operating room when a sequential combined technique is used in order to perform a second injection of blue dye. The endometrial or subendometrial injection is performed by hysteroscopy before surgery or just after general anesthesia. This injection site also enables the surgeon to orient the injections in relation to the tumor location. Thus, the distribution of biopsied SLN is comparable with that achieved with myometrial injection. The para-aortic basin was indeed the location of the SLN biopsy site in 47.8 % of the 193 patients with successfully detected SLN using hysteroscopic injections (Table 2). This injection site is associated with a higher rate of para-aortic SLN drainage, which is consistent with data showing 67 % of EC patients with metastasis in para-aortic nodes and 16 % of EC patients with isolated para-aortic nodal involvement [27]. This detection rate is not statistically different from the myometrial injection results (p =0.2). However, after a cervical injection, most of the series reported a maximum of 5 % of patients with para-aortic drainage (Table 1). This is the injection site with the greatest number of biopsied SLN (more than 700; Table 1). Only the series by How et al. [28] found 23 % of patients with paraaortic drainage, even though they used the same cervical injection as the team with the largest series of 498 patients,
FN rate (%)
14.89 11.11 15.79 0.00 0.00 0.00
0.00 0.00 0.00 10.28
7 1 3 0 0 0
0 0 0 11 3 NA 3 40 5 2 3 96 Pelvic Pelvic Pelvic 2.5 2.5 1.6 2.4 5.26 NA 0.00 6.49 82.61 84.00 93.75 82.87 Isotopic Blue Dual
Retrospective analysis of a prospective database
Aortic and common iliac SLN location included b
a
FN false negative, LN lymph node, NA not available
I–II I Ib 23 25 16 899 2007 2004 2003 Delpech et al. [46] Holub et al. [47] Pelosi et al. [48] Total
Prospective NA NA
9 4 8 4 2 7 40 8 16 4 8 10 Pelvic +/- aortic Pelvic +/- aortic Pelvic +/- aortic Pelvic Pelvic +/- aortic Pelvic +/- aortic 3.2 2.5 2.2 1.7 2.9 2.6 4.24 23.91 4.50 0.00 3.33 2.50 80.52 92.00 88.80 69.57 69.77 86.96 Dual Dual Dual Isotopic Dual Dual 498 100 125 23 43 46 2012 2012 2011 2008 2008 2008 Barlin et al. [29] How et al. [28] Ballester er al. [10] Perrone et al. [32] Bats et al. [45] Ballester et al. [11]
Retrospectivea Prospective Prospective Prospective Prospective Retrospective
I–IV I–III I–II I–II I–III I–II
Low-volume disease SLN involved LN surgery Mean no. of SLN Aortic SLNb (%) Detection rate (%) Method No. of patients Year
Design
Stage
with only 4.25 % of patients with para-aortic drainage [29]. Overall, the location of the SLN was statistically different from that after a myometrial (p
GYNECOLOGIC CANCERS (NS REED, SECTION EDITOR)
Sentinel Lymph Node in Endometrial Cancer: A Review Cyril Touboul & Enrica Bentivegna & Catherine Uzan & Sebastien Gouy & Patricia Pautier & Catherine Lhommé & Pierre Duvillard & Christine Haie-Meder & Philippe Morice
Published online: 5 November 2013 # Springer Science+Business Media New York 2013
Abstract Lymph node status is a major prognostic factor in endometrial cancer (EC). Sentinel lymph node (SLN) biopsy has been reported in EC for more than 15 years but has not yet been incorporated as a standard-of-care procedure in EC. Complex uterine drainage, the various modalities of tracer injection, and the lack of large prospective series may explain this situation. In this review, we report an SLN detection rate of 81.7 %, a 10.9 % rate of metastatic SLN involvement, and a false-negative rate of 12.3 % in the main clinical trials. Thirtyfive percent of SLN metastases were low-volume disease (micrometastases or isolated tumor cells). These data raise the question of the clinical significance of low-volume disease in EC. SLN biopsy could allow upstaging in supposedly low- or C. Touboul : E. Bentivegna : C. Uzan : S. Gouy : P. Morice Department of Gynecologic Surgery, Gustave Roussy Cancer Campus, Villejuif, France P. Pautier : C. Lhommé Department of Oncology, Gustave Roussy Cancer Campus, Villejuif, France P. Duvillard Department of Pathology, Gustave Roussy Cancer Campus, Villejuif, France C. Haie-Meder Department of Radiotherapy, Gustave Roussy Cancer Campus, Villejuif, France P. Morice University Paris Sud, University de Kremlin Bicetre, Gustave Roussy Cancer Campus, Villejuif, France C. Uzan : P. Morice Unit INSERM 10-30, Villejuif, France P. Morice (*) Institut Gustave Roussy, 39 rue Camille Desmoulins, 94805 Villejuif, France e-mail: [email protected]
intermediate-risk patients in whom adjuvant therapy could be omitted. Further studies are required to precise the interest on the survival of this procedure in this subset of patients. Keywords Sentinel lymph node . Endometrial cancer . Lymphatic mapping . Nodal status . Micrometastasis
Introduction With nearly 50,000 new cases per year in the USA, endometrial cancer (EC) is the leading cause of gynecologic cancers in Western countries [1–3]. Most EC are limited to the uterus, with no evidence of extrauterine disease preoperatively [International Federation of Gynecology and Obstetrics (FIGO) stage I], thus explaining a generally favorable prognosis (95 % overall survival). Lymph node status is a crucial prognostic factor in this cancer, impairing survival when a lymph node metastasis is found [4, 5]. Recurrence-free survival (drops from 87 % without node involvement to 71 % and 36 % in women with pelvic and aortic node involvement, respectively [6]. However, the ASTEC randomized controlled trial showed no evidence of a benefit in terms of overall or recurrence-free survival for a systematic pelvic lymphadenectomy versus selective pelvic/aortic node dissection in women with early EC [2]. That study confirmed the data from an Italian randomized controlled trial in which a higher proportion of women underwent a para-aortic dissection (26 %) with no influence of lymphadenectomy on survival [7]. The indication for lymphadenectomy should therefore be limited to high-risk EC, and should be based on a preoperative biopsy and magnetic resonance imaging. Unfortunately, the relationship between the preoperative and the postoperative grade was shown to be discordant, with discrepancies in up to 46 % of cases [8–10]. Discordance in magnetic resonance imaging assessment is also considerable [11, 12], with better results since the
560
modification of the FIGO classification in 2009 [13, 14]. There is therefore a risk of surgical undertreatment or overtreatment with either the preoperative radiologic assessment or the histologic evaluation alone. Sentinel lymph node (SLN) biopsy could therefore solve this issue by acting as a “middle-of-the-road” procedure between an invasive complete pelvic and para-aortic lymphadenectomy and the absence of lymph node evaluation [15]. The concept of SLN biopsy was first developed in melanoma and breast cancer before its use in EC in 1996 [16]. SLN mapping is attractive in EC, which mainly arises in older menopausal women with comorbidities in whom a full lymphadenectomy may not be feasible or may cause further morbidity. We review here the clinical value of SLN biopsy, the injection technique used, the anatomopathological findings, and potential future indications in EC.
Article Selection Process We reviewed all articles from the first report of the technique in 1996 by Burke et al. [16], using “endometrial cancer” (EC) and “sentinel node” as search keywords in the PubMed database. We selected all clinical trials written in English reporting SLN biopsy results, and analyzed only the last publication when it was published by the same team with shared patients. Twenty-three studies fulfilled our criteria. We analyzed the type of study and year of publication, the size of the population, the disease stage, the detection method, the detection rate, the number and location of biopsied SLN, the histologic results, and the number of false negatives.
EC Drainage into SLN Unlike the case in cervical cancer where the cervix is the only injection site, several injection sites are reportedly used in EC [17–19]. Cervical, endometrial, or myometrial injections have been used in EC with different approaches: hysteroscopic, laparoscopic, and more recently transvaginal myometrial injections. These differences regarding the injection sites are relative to differences in the mode of drainage between these two cancers. Whereas cervix lymphatic drainage is exclusively via the pelvic lymph nodes, uterine corpus drainage may flow directly into the aortic lymph nodes via the ovarian vein, thus explaining the skip metastasis phenomenon. The location of the endometrial tumor is therefore determinant in disease dissemination and the site of lymph node metastasis. This review focuses on the performance of SLN biopsy and SLN mapping according to the injection site. The first series on SLN in EC used a myometrial injection into the anterior and posterior walls of the uterine fundus [16, 20]. Burke et al. [16] described, as early as this first report, blue dye
Curr Oncol Rep (2013) 15:559–565
diffusion in the uterine cornua and lateral parametria within 2–3 min of injection and along the ovarian vessels after 5–10 min. The distribution of the 31 biopsied SLN was around 40 % in the para-aortic nodes, 20 % in common iliac nodes, and 40 % in pelvic nodes [16]. Furthermore, the para-aortic SLN were above the inferior mesenteric artery, around ovarian vessel implantation, confirming drainage via this route. This is the strength of the myometrial injection, which is in line with potential lymphatic drainage of uterine corpus tumors. These results were confirmed by more recent larger series, reporting 40–50 % of patients with biopsied SLN in the para-aortic basin [21–23]. However, other reports showed discrepancies in the distribution of SLN [24–26]. For small series, this may be due to the impact of the learning curve on the overall results. Nevertheless, in the 152 patients with successfully detected SLN, 37.5 % (0–58 %) had at least one SLN in the para-aortic area (Tables 1 and 2). The study by Robova et al. [25], which compared myometrial injection with hysteroscopic injection, found 8.3 % of patients with SLN biopsied in the para-aortic area. This myometrial/subserosal injection was performed intraoperatively with blue dye during a laparotomy in most of these series. However, Torné et al. [23] performed a transvaginal ultrasound-guided injection of a radioactive tracer into the anterior and posterior uterine wall at the level of the tumor. This technique has the advantage of being consistent with tumor drainage, 45.4 % of which was para-aortic, and the detection rate exceeded 90 %. However, the technique is more complex than cervical injection, as it requires expertise in transvaginal ultrasonography. Moreover, specific equipment must be available in the operating room when a sequential combined technique is used in order to perform a second injection of blue dye. The endometrial or subendometrial injection is performed by hysteroscopy before surgery or just after general anesthesia. This injection site also enables the surgeon to orient the injections in relation to the tumor location. Thus, the distribution of biopsied SLN is comparable with that achieved with myometrial injection. The para-aortic basin was indeed the location of the SLN biopsy site in 47.8 % of the 193 patients with successfully detected SLN using hysteroscopic injections (Table 2). This injection site is associated with a higher rate of para-aortic SLN drainage, which is consistent with data showing 67 % of EC patients with metastasis in para-aortic nodes and 16 % of EC patients with isolated para-aortic nodal involvement [27]. This detection rate is not statistically different from the myometrial injection results (p =0.2). However, after a cervical injection, most of the series reported a maximum of 5 % of patients with para-aortic drainage (Table 1). This is the injection site with the greatest number of biopsied SLN (more than 700; Table 1). Only the series by How et al. [28] found 23 % of patients with paraaortic drainage, even though they used the same cervical injection as the team with the largest series of 498 patients,
FN rate (%)
14.89 11.11 15.79 0.00 0.00 0.00
0.00 0.00 0.00 10.28
7 1 3 0 0 0
0 0 0 11 3 NA 3 40 5 2 3 96 Pelvic Pelvic Pelvic 2.5 2.5 1.6 2.4 5.26 NA 0.00 6.49 82.61 84.00 93.75 82.87 Isotopic Blue Dual
Retrospective analysis of a prospective database
Aortic and common iliac SLN location included b
a
FN false negative, LN lymph node, NA not available
I–II I Ib 23 25 16 899 2007 2004 2003 Delpech et al. [46] Holub et al. [47] Pelosi et al. [48] Total
Prospective NA NA
9 4 8 4 2 7 40 8 16 4 8 10 Pelvic +/- aortic Pelvic +/- aortic Pelvic +/- aortic Pelvic Pelvic +/- aortic Pelvic +/- aortic 3.2 2.5 2.2 1.7 2.9 2.6 4.24 23.91 4.50 0.00 3.33 2.50 80.52 92.00 88.80 69.57 69.77 86.96 Dual Dual Dual Isotopic Dual Dual 498 100 125 23 43 46 2012 2012 2011 2008 2008 2008 Barlin et al. [29] How et al. [28] Ballester er al. [10] Perrone et al. [32] Bats et al. [45] Ballester et al. [11]
Retrospectivea Prospective Prospective Prospective Prospective Retrospective
I–IV I–III I–II I–II I–III I–II
Low-volume disease SLN involved LN surgery Mean no. of SLN Aortic SLNb (%) Detection rate (%) Method No. of patients Year
Design
Stage
with only 4.25 % of patients with para-aortic drainage [29]. Overall, the location of the SLN was statistically different from that after a myometrial (p
