Article Text
Abstract
Introduction Several studies have reported optimizing ultrastaging protocols using immunohistochemistry for sentinel lymph node (SLN) biopsy in endometrial carcinoma; however, the clinical significance of isolated tumor cells (ITCs) detected by ultrastaging is unknown. This study aimed to: (1) determine the frequency of retrospective ITC detection in patients with endometrial carcinoma and reported negative SLNs determined by hematoxylin and eosin (H&E) examination only; and (2) determine the clinicopathological features and outcomes of patients with endometrial carcinoma and previously undetected ITCs.
Methods 474 SLNs from 155 patients with endometrial carcinoma and reported negative SLNs were subjected to an immunohistochemistry protocol which included staining slides with cytokeratin at 1, 10, 20, and 50 µm levels, to examine for ITCs. Clinicopathological data of patients with ITCs detected by this method were analyzed to determine patient outcomes.
Results Using immunohistochemistry, ITCs were detected in 5.7% (27/474) of SLNs and 13.5% (21/155) of patients with previously reported negative SLNs. In this patient cohort, 95.2% (20/21) had endometrioid histology, with the remaining case being carcinosarcoma. 38.1% (8/21) received adjuvant therapy (either brachytherapy alone (4/8) or chemotherapy and radiation (4/8)) based on other parameters, while 61.9% (13/21) had no adjuvant therapy. Of the patients who did not receive adjuvant therapy, all had endometrioid histology and 84.6% (11/13) were International Federation of Gynecology and Obstetrics (FIGO) stage IA. No patients (0/13) recurred after a median follow-up of 31.5 (range 2–84.4) months.
Discussion In this study, 38.1% of patients with previously undetected ITCs had adjuvant treatment based on other high risk factors; as such, reporting ITCs would not have altered patient management for those who received adjuvant chemotherapy. To date, no patients with previously undetected ITCs without adjuvant treatment had a recurrence, suggesting that ITC detection may not be clinically relevant.
- endometrial neoplasms
- surgical oncology
- sentinel lymph node
Statistics from Altmetric.com
HIGHLIGHTS
Immunohistochemistry for cytokeratin AE1/3 enhances isolated tumor cell detection.
In patients with endometrial carcinoma, adjuvant treatment is often determined by high-risk uterine factors.
Isolated tumor cells in endometrial cancer patients may not have clinical relevance in terms of recurrence risk.
Introduction
Pelvic and para-aortic lymphadenectomy became an important aspect of endometrial carcinoma management following a large prospective study (Gynecologic Oncology Group (GOG) study 33) which showed that 22% of stage I endometrial carcinoma patients harbored extrauterine disease1; however, it is not without morbidity.2 Lymph node dissection in women with endometrial carcinoma can result in increased operative time, as well as early and late complications including blood loss, vascular injury, lymphocyst formation, and lower extremity edema.3–9 Sentinel lymph node (SLN) mapping in endometrial carcinoma has been recognized by the National Comprehensive Cancer Network as an alternative to more comprehensive nodal assessment.10 It can negate the need for lymphadenectomy and is a good indicator of nodal status as it has high sensitivity and a low false negative rate.11–13
Pathological ultrastaging protocols are performed on SLNs, which improves isolated tumor cell (ITC) and micrometastasis detection; however, there is no consensus on the optimal evaluation protocol and studies have reported variable protocols that include multiple hematoxylin and eosin (H&E) levels at different micron separations and intervening cytokeratin immunohistochemistry.13–19 ITCs are considered negative in TNM staging and should be reported as pN0(i+).20 The clinical significance of detecting ITCs remains unclear; furthermore, there is no consensus regarding clinical management of patients with ITCs and only a few studies have assessed the clinical outcomes of patients with ITCs.21–23
The aims of this study were twofold: (1) to determine the frequency of retrospective ITC detection in patients with endometrial carcinoma and reported negative SLN biopsy determined by H&E examination only; and (2) to determine the clinicopathological features and outcomes of patients with endometrial carcinoma and previously undetected ITCs.
Methods
Case collection
Institutional Research Ethics Board approval was obtained for this study. Pathology reports of all patients with endometrial carcinoma and SLN biopsy at our institution from 2012 to 2016 were retrieved by performing a search of the pathology database. All reports were reviewed to determine patients with reported negative lymph node status.
SLN Biopsy Protocol
As previously described,24 patients were staged following the National Comprehensive Cancer Network SLN mapping algorithm.6 13 The cervix was injected submucosally at the 3 o'clock and 9 o'clock positions with 1.5–2.0 mL of 1.25 mg/mL indocyanine green. A 10 mm near-infrared laparoscopic camera (Novadaq, Mississauga, Ontario, Canada) was used. The first green lymph node in each lymphatic chain and/or any grossly enlarged or suspicious lymph nodes were removed. When an SLN could not be identified, the uterus was sent for frozen section, and pelvic lymphadenectomy was performed on the side(s) where an SLN was not identified if indicated by the modified Mayo criteria.
Examination of SLNs by Immunohistochemistry
Available material from patients with reported negative SLNs were subjected to examination by immunohistochemistry. All SLNs had been previously sectioned at 2 mm intervals perpendicular to the long axis, formalin-fixed, paraffin embedded, and entirely submitted for pathological examination. Each had one to three H&E levels cut 100 µm apart that were reviewed at the time of issuance of the pathology report. For this study, an additional 12 sections 4 µm thick were cut from all SLN blocks for each case. Immunohistochemistry for cytokeratin AE1/3 (1:450 dilution, AE1 and AE3 clones; Dako, Carpinteria, CA, USA) was performed on four of these sections at 1, 10, 20, and 50 µm levels to examine for ITCs. ITCs are defined as clusters of malignant cells within a lymph node that are ≤0.2 mm in size and ≤200 cells, micrometastases measure >0.2 to 2 mm and are >200 cells, and macrometastases measure >2 mm.20 Additionally, following identification of ITCs by cytokeratin AE1/3, an adjacent section was stained with WT-1 (1:200 dilution; clone 6 F-H2; Dako, Carpinteria, CA, USA) to exclude the presence of mesothelial cells and confirm ITCs. The secondary antibody (The EnVision +System, HRP labeled polymer anti-mouse, DAKO, Carpinteria, CA, USA) and detection (Liquid DAB+, 2-component system, DAKO, Carpinteria, CA, USA) were used for both primary antibodies. ITCs were considered present when there was a cell or cells with strong cytoplasmic cytokeratin AE1/3 positivity and WT-1 staining was absent.
Collection and Analysis of Clinicopathological Data
Clinicopathological data of patients with ITCs detected by this method were collected from an institutional patient database and included patient age at diagnosis, histology, grade and stage of malignancy, surgical management, adjuvant therapy, recurrence, and length of follow-up.
Results
Overall, 246 patients with endometrial carcinoma had SLN biopsy ± concurrent lymph node dissection between 2012 and 2016. Eleven of the 246 (4.5%) had reported metastatic disease in SLNs. Nine were macrometastasis and the other two were micrometastasis. The remaining 95.5% (235/246) had reported negative SLNs. Available material from 155 patients with reported negative SLNs, which included a total of 474 SLNs, were examined using immunohistochemistry for ITC detection (Figure 1).
In this group, the most common histology was grade 1 endometrioid carcinoma (60.0%, 93/155) followed by 19.4% (30/155) grade 2 endometrioid carcinoma, 4.5% (7/155) grade 3 endometrioid, 11.6% (18/155) serous carcinoma, 3.2% (5/155) carcinosarcoma, and one (0.65%) each of clear cell carcinoma and mixed endometrioid and serous carcinoma. Patients had a median of three SLNs removed (range 1–16) at the time of surgery. Nineteen percent (29/155) had concurrent uni- or bilateral pelvic lymphadenectomy. Four patients with previously reported negative SLN biopsy and concurrent pelvic lymphadenectomy had micro- or macrometastases in non-SLNs; two of these patients had ITCs in SLN detected by immunohistochemistry.
The SLN immunohistochemistry protocol detected ITCs in 5.7% (27/474) of SLNs and 13.5% (21/155) of patients that had previously reported negative SLN on final pathology. No additional micro- or macrometastases were detected. The clinicopathological data for these patients is summarized in Table 1. Briefly, the average age of patients with ITCs was 59 years, all had laparoscopic hysterectomy and either SLN biopsy only (76.2%, 16/21), SLN and pelvic lymph node dissection (19.0%, 4/21), or SLN and pelvic and para-aortic lymph node dissection (4.8%, 1/21). Of the patients with ITCs detected by immunohistochemistry, 95.2% (20/21) had endometrioid histology with the remaining case being carcinosarcoma. The majority were International Federation of Gynecology and Obstetrics (FIGO) stage IA (57.1%, 12/21) (reflects stage before the study and ITC detection). Sixty-two (13/21) of patients with endometrial carcinoma and ITCs did not receive adjuvant therapy, while the remaining 38.1% (8/21) received adjuvant therapy (either vaginal brachytherapy alone (4/8) or chemotherapy and radiation (4/8) (either brachytherapy (1/4), external beam radiation and brachytherapy (1/4), or whole pelvic radiation (2/4)) based on other parameters. The clinicopathological features of the eight patients who received adjuvant therapy are shown in Table 2. In all cases, adjuvant treatment was initiated due to high risk uterine factors or advanced stage disease. To summarize, all but one had endometrioid carcinoma (87.5%) in which 25.0% (2/8) were grade 1, 37.5% (3/8) were grade 2, and 25.0% (2/8) were grade 3; the remaining was carcinosarcoma. The majority had >50% myometrial invasion (62.5%, 5/8), one had <50%, and one had no myometrial invasion. Lymphovascular invasion was present in 37.5% (3/8), one patient (12.5%) had a positive non-SLN, and 42.6% (3/7) had positive peritoneal cytology. FIGO stage was 25.0% (2/8) stage IA, 25.0% (2/8) stage IB, 25.0% (2/8) stage IIIA, 12.5% (1/8) stage IIIC, and 12.5% (1/8) stage IV. Metastases were present in 50.0% (4/8) at the time of surgery and involved pelvic lymph nodes, ovaries, round ligament, and omentum. After a median follow-up of 36 (range 14.4–89.4) months, the patient with stage IV carcinosarcoma was the only one to recur at 18 months.
In order to determine if ITCs had any prognostic significance in patients with endometrial carcinoma, we focused on the group of patients with previously undetected ITCs that had not received adjuvant therapy. Of the 13 patients in this group, 84.6% (11/13) were FIGO stage IA and the remaining 15.4% (2/13) were either FIGO stage II or IIIC1, the latter of which had a concurrent positive non-sentinel pelvic lymph node metastasis. The two patients with higher stage disease did not undergo adjuvant therapy due to personal and other medical factors; however, they would have received adjuvant therapy based on treatment protocols and were therefore not included in the patient group with ITCs and no adjuvant treatment. The clinical and pathological characteristics of patients with ITCs and without adjuvant therapy are shown in Table 3. All (11/11) had laparoscopic hysterectomy (either standard or robotic assisted) and all (11/11) had SLN biopsy only. One hundred percent (11/11) had endometrioid carcinoma, were grade 1, and all but one were FIGO stage IA. The depth of myometrial invasion was <50% in 72.7% (8/11), >50% in 9.1% (1/11), and in 18.1% (2/11) there was no myometrial invasion. Two (patients 5 and 6) had multiple SLNs that were positive for ITCs, 3/11 (27.3%) had lymphovascular invasion, and 2/11 (18.1%) had positive peritoneal cytology. Two (patients 4 and 9) also had a concurrent primary stage I ovarian tumor. No patients (0/11) recurred after a median follow-up interval of 32.6 (range 2–84.4) months, with four patients having >36 months follow-up with no evidence of disease recurrence.
Discussion
The clinical significance of ITCs in patients with endometrial carcinoma, in terms of prognosis and recurrence, remains uncertain. We examined a population of patients with endometrial carcinoma and ITCs and then separated out patients with ITCs and no adjuvant therapy to better determine the natural history of ITCs. In our cohort, 38.1% of patients with previously undetected ITCs had adjuvant treatment based on other parameters, and reporting of ITCs may not have altered patient management for those who received chemotherapy. Although our interpretation of the data is limited by the relatively short follow-up interval, no patients with previously undetected ITCs and no adjuvant treatment had a recurrence. Four of these patients were disease-free after >3 years of follow-up, suggesting that these patients may not require additional treatment and that ITC detection may not have an impact on clinical outcome.
This study, similar to others, demonstrates that immunohistochemistry enhances ITC detection.19 Immunohistochemistry for cytokeratins detected ITCs in 13.5% of patients with previously reported negative SLNs. In order to detect low volume disease, Plante et al utilized an extensive ultrastaging protocol that examined six H&E sections at 50 μm apart and one immunohistochemical slide (cytokeratin AE1/3), and found that ITCs were detected in 6.0% of cases.21 In our study, we examined four cytokeratin immunohistochemical slides to detect ITCs and found the rate of detection to be over twice as high. A study comparing two ultrastaging protocols, one with five H&E levels and a pankeratin stain and the other with one H&E slide and a pankeratin stain, found that there was no significant advantage of more extensive sectioning when comparing number of positive SLNs detected, size of metastasis, and false-negative rate with the more conservative approach.19 As our study involved multiple immunohistochemical levels (and not H&Es), this suggests that while additional H&E levels do not increase ITC detection, additional immunohistochemistry does. Thus SLNs appear to more frequently harbor unrecognized ITCs than previously recognized by published methods, which again raises the possibility that ITC detection may not impact clinical outcome. In addition, reported 5 year recurrence rate for patients with low grade, low stage endometrioid carcinoma treated with surgery alone and no adjuvant therapy is 2.6–3.1%,25 which is much less than the ITC detection rate, further supporting the proposal that ITCs may not be clinically relevant. For this reason, and the fact that in our study immunohistochemistry did not detect any additional micro- or macrometastases, this suggests that immunohistochemical evaluation of SLNs in endometrial carcinoma may not be necessary.
The management of patients with ITCs is controversial and, depending on the center, some patients receive adjuvant treatment on the basis of the presence of ITCs, while others are observed. Studies examining patients with endometrial carcinoma and ITCs have shown that patients with ITCs often receive adjuvant therapy with the percentage of patients receiving adjuvant treatment ranging from 57–87%,18 21 22 so the assessment of the clinical significance of ITCs is confounded.23 Due to the retrospective nature of detecting ITCs in our study, patients were not known to have ITCs at the time of diagnosis and their presence was not factored into treatment decisions. As such, the majority of patients (61.9%) with ITCs did not receive treatment; the 38.1% that did, received treatment based on other factors and not the presence of ITCs.
Furthermore, some studies considered patients with ITCs to be of FIGO stage IIIC and therefore recommended adjuvant therapy.18 A study by St Clair et al sought to determine treatment patterns and oncologic outcomes of patients with endometrial carcinoma and low volume nodal metastasis and found that of 844 patients, 23 patients had ITCs, of which 20 (87%) received adjuvant therapy. In this group, 2/23 (8.7 %) patients recurred, one locally and one distantly. When compared with patients with macrometastasis, 3 year recurrence-free survival was improved for patients with ITCs that received adjuvant therapy (86% for patients with ITCs vs 71% for patients with macrometastasis).18 While the authors concluded that patients with ITCs who received adjuvant therapy had improved recurrence-free survival when compared with patients with macrometastasis, they acknowledged that they were unable to comment on the outcomes of untreated low volume disease because the majority had received adjuvant treatment—therefore the question of optimal post-surgical treatment of these patients still remains. A study by Plante et al that evaluated the role of adjuvant treatment in the management of patients with endometrial carcinoma and ITCs draws similarities in terms of progression-free survival to the St Clair et al study.21 They showed that in a prospective study of 519 patients with endometrial carcinoma and SLN mapping, 85 patients (16.4%) had SLN metastases21; of these, 31 (36%) were ITCs. None of the 28 patients with endometrioid histology recurred. Of the 10 patients with no adjuvant therapy or brachytherapy alone, none recurred. When comparing patients with ITCs to those with micro- and macrometastases, 3 year progression-free survival was similar to the micrometastases group (95.5% for ITCs and 85.5% for micrometastases), but was statistically better than patients with macrometastases (58.6%).21 Despite a larger proportion of patients with ITCs receiving no adjuvant treatment or vaginal brachytherapy alone, progression-free survival in this cohort was similar to the cohort of patients with ITCs in the St Clair et al study in which most received adjuvant therapy.21 The authors concluded that patients with endometrial carcinoma and ITCs in SLNs have an excellent prognosis and probably derive little benefit from adjuvant therapy.21
The true clinical significance of ITCs is yet to be determined and data looking specifically at patients with ITCs in SLNs who did not undergo adjuvant treatment are sparse. Our study is one of the largest examining this population. Yet in other studies, micrometastases and ITCs are grouped together, making it difficult to determine the true clinical relevance of ITCs. Todo et al examined 61 patients with FIGO stage I to II endometrial carcinoma with at least one risk factor for recurrence, including high grade histology, deep myometrial invasion, cervical involvement, lymphovascular invasion or positive peritoneal cytology, and reported that low volume disease (ITC or micrometastasis) in regional pelvic lymph nodes was an independent risk factor for recurrence. They further reported that an 8 year overall and recurrence-free survival was >20% lower in the patients with ITCs or micrometastases compared with the node-negative patients; however, this did not reach statistical significance.26 Caveats that prevent direct comparison to our current study include: the ultrastaging protocol examined all regional pelvic lymph nodes, not SLNs only; and patients with ITCs and micrometastases were grouped to draw conclusions. Of the six patients with ITCs in the Todo et al study, three had a recurrence, one to a para-aortic lymph node, one to a para-aortic lymph node and mediastinum, and one to the liver and diaphragm.26 Only one of these patients did not receive adjuvant therapy, again making it difficult to determine the natural history of patients with ITCs due to such low numbers.
In fact, studies examining patients with endometrial carcinoma and ITCs in SLN with no treatment, even when combined with patients who received vaginal brachytherapy alone, are sparse and total numbers are generally low, and require a longer follow-up period (Table 4).18 21 22 If we group patients with no treatment and vaginal brachytherapy alone, there are a total of 15 patients, none of whom recurred after a median follow-up of 32.8 months, making our study one of the largest published group of these patients with the longest follow-up to date.
More recently, a larger study of 175 patients with grade I or II endometrioid carcinoma and ITCs found that 43% received vaginal brachytherapy alone or no adjuvant treatment, with the remainder receiving either external beam radiation alone or chemotherapy ± radiation.22 Overall, the authors found that extravaginal recurrences were similar in patients with and without chemotherapy and were generally low (5.2% vs 3.8%, respectively). They suggest that adjuvant therapy does not appear to have a significant effect on recurrence-free survival in this population.22
In the Plante et al study, the only patient to recur was a stage IB carcinosarcoma.21 In our study, the single recurrence was a stage IV carcinosarcoma with metastases to round ligament, bilateral ovaries, and omentum, and clearly this outcome was driven by factors other than ITCs. Recurrences in other studies have occurred in patients who also received adjuvant treatment.18 26 In the St Clair et al study, of the two patients with ITCs that recurred, both received adjuvant therapy; however, the other tumor characteristics were not reported.18 In another study of 508 patients with endometrial cancer, of which 6.9% had nodal metastasis detected by H&E stain alone, ultrastaging of SLNs detected an additional 19 cases of ITCs.15 Of these cases, two patients recurred, one with grade 2 endometrioid carcinoma and one with carcinosarcoma; both received adjuvant treatment, pelvic radiation and chemotherapy, and vaginal brachytherapy, based on other parameters.15
Although some studies suggest that deep myometrial invasion and lymphovascular invasion may be predictors of ITCs,26 our study showed that 14.3% (3/21) had ITCs and no myometrial invasion. This is contrary to Kim et al who showed that ITC detection in patients with endometrial carcinoma and no myometrial invasion was only 0.8% compared with 7–8% in patients with any degree of myometrial invasion. The authors argued that ultrastaging is required for endometrial carcinoma with any degree of myometrial invasion and that in cases with no myometrial invasion, ultrastaging of SLN can be eliminated.15 We have shown that ITCs can occur even when there is no myometrial invasion. Lymphovascular invasion was only present in 33% (7/21) of cases with ITCs.
There are a few limitations to this study. First, although this is one of the largest studies examining outcomes of endometrial carcinoma patients with ITCs without adjuvant therapy, the absolute number of untreated patients remains small. However, our study adds to previously published data suggesting that ITCs may not have prognostic relevance.21 Second, which also contributes to the first limitation, material was only available to perform the immunohistochemistry protocol on a subset of patients with endometrial carcinoma at our institution during the study period (155/235) and may further account for the low number of patients in our study. Last, the mean follow-up interval in our study was 32.8 months; a longer follow-up period may be required to make firm conclusions about the prognostic significance of ITCs.
Older studies suggest that cytokeratin staining in lymph nodes in patients with endometrial carcinoma and otherwise no known metastasis is a predictor of recurrence27; however this is not supported by ours and other studies.21–23 We have shown that no patients with endometrial carcinoma and ITCs and no adjuvant treatment recurred in a median follow-up of 32.6 months. If we combine this with patients who received vaginal brachytherapy only, there was still no recurrence with a median follow-up of 32.8 months. Although we have shown that ITC detection is increased with the use of immunohistochemistry, we suggest that the presence of ITCs, in the absence of other high-risk uterine factors or advanced stage disease, may not be clinically relevant.
Acknowledgments
The authors thank Liping Yuan for performing the technical duties required for the immunohistochemistry protocol.
References
Footnotes
Editor's note This paper will feature in a special issue on sentinel lymph node mapping in 2020.
MRN and CMF contributed equally.
Contributors All authors contributed to the study.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Patient consent for publication Not required.
Provenance and peer review Not commissioned; externally peer reviewed.
Data availability statement All data relevant to the study are included in the article or uploaded as supplementary information.