Abstract
Objective:
This study aimed to examine an association between intrauterine manipulator (IUM) use and frequency of lymphovascular space invasion (LVSI) in women with endometrial cancer undergoing minimally invasive hysterectomy.
Methods:
A retrospective case-control study was conducted among stage I–IV endometrial cancer patients who underwent hysterectomy between 2008 and 2015. Medical records were reviewed for patient demographics, surgical details, and tumor characteristics. Women who underwent total laparoscopic hysterectomy (TLH) with IUM use were compared with women who underwent total abdominal hysterectomy (TAH). Review of archived medical record for data collection and propensity score matching were performed to adjust for background differences between TLH-IUM and TAH groups. A systematic literature review with pooled analysis was performed to examine frequency of LVSI.
Results:
There were 687 women who underwent hysterectomy for endometrial cancer. Of those, 419 women underwent TLH with IUM use and 194 women underwent TAH. The most common type of IUM was VCare (89.5%). There was no statistically significant difference in the frequency of LVSI between the 2 groups: 15.1% for TLH-IUM vs 19.9% for TAH (P = 0.14). After propensity score matching, frequencies of LVSI were similar between the 2 groups: 21.2% for TLH-IUM vs 19.6% for TAH (P = 0.78). Systematic literature review identified 1371 cases of TLH-IUM and 1246 cases of TAH performed for endometrial cancer, and frequencies of LVSI were similar between the 2 groups (15.0% vs 13.6%, P = 0.31).
Conclusion:
Our study suggests that IUM use during TLH for endometrial cancer is not associated with increased frequency of LVSI.
Keywords: Endometrial cancer, Uterine manipulator, Minimally invasive surgery, Hysterectomy, LVSI, Lymphovascular space invasion
Endometrial cancer is the most common gynecologic malignancy in the United States, with an estimated 61,380 new cases identified in 2017.1 The incidence of endometrial cancer is steadily rising as risk factors, including obesity, diabetes mellitus, and single marital status, increase among the general population.2 Hysterectomy-based staging surgery is the standard treatment of endometrial cancer. Minimally invasive surgery, including total laparoscopic hysterectomy (TLH) and robot-assisted hysterectomy, offers many postoperative benefits such as less pain, faster recovery, and shorter hospital stays when compared with the conventional laparotomy approach.3–5 As a result, minimally invasive hysterectomy has gained popularity over the past few decades in gynecologic cancer surgery, largely replacing conventional total abdominal hysterectomy (TAH). The GOG-LAP2 trial showed less postoperative morbidity with minimally invasive approach and comparable recurrence risk between minimally invasive surgery and laparotomy for early-stage endometrial cancer.3,6 Therefore, minimally invasive surgery has been advocated by many as a feasible approach in the management of endometrial cancer.7
The effectiveness of using an intrauterine manipulator (IUM) during laparoscopic hysterectomy is well demonstrated.8 Nevertheless, some surgeons avoid using this device with endometrial cancer for fear of iatrogenic lymphovascular space invasion (LVSI) and retrograde tumor spillage into the peritoneal cavity. Theoretically, this instrument poses the potential risk of pushing tumor cells into the lymphovascular space with increased intrauterine cavity pressure created by the inflated balloon.9 Our previous study showed that timing of IUM insertion during TLH for endometrial cancer is not associated with frequency of malignant cells in the pelvic washing.10 However, the clinical significance of LVSI related to IUM use in endometrial cancer remains unclear with inconsistent findings across studies.10 The aim of our study was to evaluate the frequency of LVSI in women undergoing minimally invasive hysterectomy with IUM placement for the management of women with endometrial cancer.
MATERIALS AND METHODS
Study Design and Eligibility
After institutional review board approval at the University of Southern California, cases were identified for this study from an institutional surgical pathology database. Eligibility criteria included patients with a diagnosis of endometrial cancer who were treated with initial simple hysterectomy at Los Angeles County Medical Center and Keck Medical Center of the University of Southern California between January 1, 2008, and December 31, 2015. Exclusion criteria were as follows: (i) the patient did not undergo hysterectomy-based surgical staging; (ii) the type of hysterectomy was radical or modified radical, vaginal, or supracervical; and (iii) available information on the use of IUM and/or LVSI status was insufficient.
Eligible cases were divided into 2 groups: patients who underwent TLH using an IUM and patients who underwent TAH without IUM use. The decision to use an IUM and the type of IUM depends on surgeons. The Strengthening the Reporting of Observational Studies in Epidemiology guidelines were consulted for outlining the results of this retrospective cohort study.11 Some of the patients included were previously evaluated within the context of our prior studies.10,12
Clinical Information
Patient demographics at the time of surgery abstracted from the medical chart included age, ethnicity, body mass index (BMI; in kilograms per meter squared), presence of medical comorbidities (diabetes mellitus, hypertension, and hypercholesterolemia), and American Society of Anesthesiologists Physical Status (ASA-PS) at the time of surgery. We also abstracted the method of endometrial cancer diagnosis including a use of dilation and curettage, hysteroscopy, and Vabra aspirator®. Surgical demographics included estimated blood loss (EBL; in milliliters), operation time (in minutes), surgical staging details (type of hysterectomy and lymphadenectomy), intraoperative laparotomy conversion, and type of IUM. Tumor characteristics included cancer stage, histologic subtype, grade, uterine weight (in grams), tumor size (in centimeters), cervical involvement, LVSI, depth of myometrial invasion (in percent), and pelvic cytology. For survival outcome, disease-free survival (DFS) and cause-specific survival (CSS) were examined.
Study Definitions
Age and tumor size cutoff were based on prior studies.13,14 Cancer stage was reclassified based on the 2009 International Federation of Gynecology and Obstetrics system.15 Histologic subtypes were grouped as endometrioid, serous, clear cell, and others. Tumor grade was grouped as follows: grade 1 and 2 endometrioid tumors (low-grade), and grade 3 endometrioid, serous, and clear cell tumors (high-grade). Deep myometrial tumor invasion was defined as the presence of tumor in the outer half of the myometrium (>50%). Pelvic cytology was classified as negative, atypical, or malignant. Disease-free survival was defined as the time interval between the data of hysterectomy and the date of the first recurrence. Cause-specific survival was defined as the time interval between the date of hysterectomy and the date of death due to endometrial cancer. Patients were censored if there was no recurrence, alive at the last follow-up visit, or died of other causes.
Systematic Literature Review
To evaluate the frequency of LVSI in hysterectomy specimens of patients with endometrial cancer who underwent minimally invasive hysterectomy with IUM, we conducted a comprehensive systematic literature search per the Meta-analysis of Observational Studies in Epidemiology guidelines for systematic review.16 A public search engine PubMed/ MEDLINE was searched, limited to the English language, with the keywords “lymphovascular space invasion” and “hysterectomy” (searched on December 5, 2016). Eligible publications included case series and cohort studies of patients with endometrial cancer who received primary treatment via minimally invasive hysterectomy. The references listed in each identified article were also reviewed, and eligible studies were enrolled in the analysis. Among these screened, publications lacking sufficient patient demographics, hysterectomy type, and documentation of LVSI status were excluded from the analysis. Calendar year of publication, surgical pathology results (stage, histology type, tumor grade, and LVSI), and treatment pattern (hysterectomy type and use of IUM during surgery) were abstracted from the identified publications.
Statistical Analyses
The primary interest of this analysis was to determine whether an association exists between LVSI and IUM use in hysterectomy specimens with endometrial cancer in our retrospective cohort. The secondary interest of analysis was to conduct a validation cohort for this association by using systematic literature review. In addition, we examined an association of IUM type and survival outcome among TLH-IUM cases, and survival outcome was examined based on the diagnostic modality in this group. In a post hoc sensitivity analysis, frequency of LVSI was compared between the TLH-IUM group and the TLH without IUM group.
Based on a prior prospective study that demonstrated no association of IUM use with increased frequency of LVSI,17 an assumption that the rate of LVSI is similar between TLH-IUM and TAH groups (noninferiority design) was made for sample size estimation for our study. We assumed that up to a 5% difference in frequency of LVSI can be clinically tolerated (noninferior limit margin). With an α level of 0.025 and power of 80%, 190 cases of TLH-IUM and 190 cases of TAH were estimated.18
Continuous variables were assessed by Student t test or Mann-Whitney U test, expressed as mean (SD) or median (range) as appropriate. Categorical or ordinal variables were expressed as number (%), and statistical significance was examined using χ2 or Fisher exact test as appropriate. A binary logistic regression test was used to determine independent risk factors for LVSI. All covariates with a significance level of P < 0.05 on univariate analysis were considered candidates in the initial model consisting of patient demographics, tumor factors, and treatment patterns. Then, a conditional backward method was used to determine independently significant covariates in the final model, expressed as odds ratio (OR), and 95% confident interval (CI). All statistical tests were 2-tailed, and a P value of less than 0.05 was considered to be statistically significant.
Because patient demographics, tumor factors, and treatment patterns were largely different between TLH-IUM and TAH groups, a propensity score matching was performed to adjust the background differences between the 2 groups. The rationale and methods underlying the use of a propensity score for a proposed causal exposure variable have been previously described.19,20 A propensity score for minimally invasive hysterectomy was computed for each case, determined by a multivariable logistic regression analysis (TLH-IUM vs TAH). Patient demographics, tumor characteristics, and treatment pattern were entered in the propensity score model. An automated algorithm was used for one-to-one propensity score matching between the TLH-IUM and TAH groups with a propensity score difference cutoff of 1%. All hypotheses were 2-tailed, and Statistical Package for the Social Sciences (version 24.0; IBM SPSS, Armonk, NY) was used for analysis.
RESULTS
Institutional Cohort Analyses
Selection schema is shown in Figure 1. There were 687 hysterectomies for endometrial cancer identified. Of those, 657 cases were either TLH or TAH. After exclusion of all TLH with inadequate IUM status information or nonuse of an IUM, 419 cases of TLH-IUM and 194 cases of TAH were eligible for analysis.
FIGURE 1.
Study schema. *No cases used intrauterine manipulator. EMCA, endometrial cancer; RH/mRH, radical hysterectomy/modified radical hysterectomy.
Clinical demographics are shown in Table 1. Patients in the TLH-IUM group were similar to those who underwent TAH with regard to age (median, 57.5 years), Hispanic ethnic majority (60.6%), and medical comorbidities (BMI, 32.8 kg/m2; diabetes mellitus, 32%; hypertension, 51.3%; and hypercholesterolemia, 30.3%) (all, P > 0.05). The ASA-PS in the TLH-IUM group was significantly better than that of the TAH group (ASA 1–2; 65.1% vs 41.1%, P < 0.05). The TLH-IUM group had lower EBL (100 mL vs 189 mL, P < 0.001) but longer operation time (253 minutes vs 225 minutes, P < 0.001) compared with the TAH groups. The most common type of minimally invasive hysterectomy was conventional laparoscopic approach (78.5%) followed by robotic-assisted approach (21.5%). The types of IUM used during minimally invasive surgery included VCare, RUMI/Koh, HUMI, and ENDOPATH. The most common type of IUM was VCare (89.5%) followed by RUMI/ Koh (8.8%). Rates of pelvic lymphadenectomy were significantly lower in the TLH-IUM group compared with the TAH group (36.5% vs 52.1%, P < 0.001). Twenty-six (6.2%) patients were converted to laparotomy among the TLH-IUM group, but IUM was inserted before the laparotomy conversion in all the cases.
TABLE 1.
Patient and tumor demographics (N = 613)
| Before Propensity Score Matching |
After Propensity Score Matching |
|||||
|---|---|---|---|---|---|---|
| Characteristics | TLH-IUM (n = 419) | TAH (n = 194) | P | TLH-IUM (n = 104) | TAH (n = 104) | P |
| Patient demographics | ||||||
| Age, y | 57.5 (23.8–86.8) | 56.4 (24.9–86.8) | 0.51 | 58.0 (29.1–79.8) | 56.1 (30.1–86.4) | 0.32 |
| Ethnicity | 0.18 | 0.93 | ||||
| White | 104 (24.8%) | 49 (25.3%) | 26 (24.0%) | 25 (24.0%) | ||
| African American | 11 (2.6%) | 12 (6.2%) | 4 (3.8%) | 3 (2.9%) | ||
| Hispanic | 254 (60.6%) | 110 (56.7%) | 58 (55.8%) | 62 (59.6%) | ||
| Asian | 50 (11.9%) | 23 (11.9%) | 16 (15.4%) | 14 (13.5%) | ||
| BMI, kg/m2 | 32.8 (17.5–74.4) | 33.9 (15.6–77.2) | 0.16 | 33.4 (19.4–61.9) | 33.1 (19.1–77.2) | 0.86 |
| Diabetes mellitus | 134 (32.0%) | 69 (35.6%) | 0.38 | 28 (26.9%) | 39 (37.5%) | 0.10 |
| Hypertension | 215 (51.3%) | 110 (56.7%) | 0.21 | 54 (51.9%) | 56 (53.8%) | 0.78 |
| Hypercholesterolemia | 127 (30.3%) | 59 (30.4%) | 0.98 | 36 (34.6%) | 35 (33.7%) | 0.88 |
| ASA-PS classification | <0.001 | 0.12 | ||||
| 1–2 | 273 (65.1%) | 79 (41.1%) | 56 (53.8%) | 44 (43.1%) | ||
| 3–4 | 146 (34.8%) | 113 (58.8%) | 48 (46.2%) | 58 (56.9%) | ||
| Surgical demographics | ||||||
| EBL, mL | 100 (10–2200) | 189 (10–3850) | <0.001 | 150 (15–2200) | 250 (50–1500) | <0.001 |
| Operation time, min | 253 (116–677) | 225 (65–735) | <0.001 | 252 (117–630) | 230 (65–735) | 0.47 |
| Hysterectomy | ||||||
| Laparoscopic | 329 (78.5%) | — | 77 (74.0%) | — | ||
| Robotic | 90 (21.5%) | — | 27 (26.0%) | — | ||
| Lymph node dissection | 153 (36.5%) | 101 (52.1%) | <0.001 | 44 (42.3%) | 44 (42.3%) | 0.80 |
| Laparotomy conversion | 26 (6.2%) | — | 10 (9.6%) | — | ||
| Type of uterine manipulator | ||||||
| VCare | 375 (89.5%) | — | 95 (91.3%) | — | ||
| RUMI/Koh | 37 (8.8%) | — | 9 (8.7%) | — | ||
| Other types | 7 (1.7%) | — | 0 | — | ||
| Tumor demographics | ||||||
| Uterine weight, g | 139 (18–1408) | 198 (12–4325) | <0.001 | 160 (39–740) | 145 (12–2520) | 0.84 |
| Tumor size, cm | 2.8 (1.0–21.6) | 4.0 (1.0–23.0) | <0.001 | 3.6 (1.0–11.0) | 3.5 (1.0–15.0) | 0.29 |
| Stage | <0.001 | 0.10 | ||||
| I | 368 (87.8%) | 122 (62.9%) | 82 (78.8%) | 68 (65.4%) | ||
| II | 12 (2.9%) | 12 (6.2%) | 5 (4.8%) | 9 (8.7%) | ||
| III | 31 (7.4%) | 31 (16.0%) | 12 (11.5%) | 14 (13.5%) | ||
| IV | 8 (1.9%) | 29 (14.9%) | 5 (4.8%) | 13 (12.5%) | ||
| Histology | 0.03 | 0.88 | ||||
| Endometrioid | 363 (86.6%) | 152 (78.4%) | 83 (79.8%) | 79 (76.0%) | ||
| Serous | 30 (7.2%) | 18 (9.3%) | 8 (7.7%) | 10 (9.6%) | ||
| Clear cell | 9 (2.1%) | 11 (5.7%) | 5 (4.8%) | 7 (6.7%) | ||
| Other | 17 (4.1%) | 13 (6.7%) | 8 (7.7%) | 8 (7.7%) | ||
| Grade | <0.001 | 0.81 | ||||
| 1 | 285 (68.0%) | 91 (46.9%) | 56 (53.8%) | 52 (50.0%) | ||
| 2 | 56 (13.4%) | 42 (21.6%) | 20 (19.2%) | 20 (19.2%) | ||
| 3 | 78 (18.6%) | 61 (31.4%) | 28 (26.9%) | 32 (30.8%) | ||
| Cervical involvement | 16 (3.8%) | 18 (9.4%) | <0.001 | 9 (8.7%) | 9 (9.0%) | 0.14 |
| LVSI | 63 (15.1%) | 38 (19.9%) | 0.14 | 22 (21.2%) | 20 (19.6%) | 0.78 |
| Deep myometrial invasion | 79 (18.9%) | 58 (30.2%) | 0.002 | 25 (24.0%) | 27 (26.5%) | 0.69 |
| Pelvic cytology | <0.001 | 0.40 | ||||
| Negative | 362 (90.7%) | 130 (83.9%) | 91 (89.2%) | 66 (84.6%) | ||
| Atypical | 1 (0.3%) | 1 (0.6%) | 11 (10.8%) | 11 (14.1%) | ||
| Malignant | 36 (9.0%) | 24 (15.5%) | 0 | 1 (1.3%) | ||
Median (range) or number (%) is shown. Kruskal-Wallis H test and χ2 test for P values. Significant variables are emboldened.
Most patients in the TLH-IUM group had lower uterine weight (139 g vs 198 g), small tumor size (2.8 cm vs 4.0 cm), stage I disease (87.8% vs 62.9%), endometrioid histology (86.6% vs 78.4%), and low tumor grade (81.4% vs 68.5%) compared with the TAH group (all, P<G 0.05). In addition, hysterectomy specimens in the TLH-IUM group showed significantly less cervical stromal involvement (3.8% vs 9.4%), less deep myometrial invasion (18.9% vs 30.2%), and lower incidence of malignant pelvic cytology (9.0% vs 15.5%) compared with the TAH group (all, P < 0.01).
Univariate analysis was performed to determine risk factors for LVSI (Table 2). The use of IUM during surgery was not associated with frequency of LVSI (TLH-IUM vs TAH, 15.1% vs 19.9%; OR, 0.72; 95% CI, 0.46–1.12; P = 0.14). Type of IUM was not associated with frequency of LVSI (VCare vs other types; OR, 0.72; 95% CI, 0.27–1.90; P = 0.51). The presence of LVSI was significantly associated with age, ethnicity, BMI, stage, histology, tumor grade, tumor size, deep myometrial invasion, and pelvic cytology (all, P < 0.01). On multivariate analysis controlling for significant variables in the univariate analysis, age of at least 60 years (adjusted OR, 2.16; 95% CI, 1.25–3.71; P = 0.005), stage III–IV disease (adjusted OR, 2.25; 95% CI, 1.20–4.22; P = 0.011), high-grade tumor (adjusted OR, 3.83; 95% CI, 2.13–6.88; P < 0.001), tumor size of at least 2 cm (adjusted OR, 2.06; 95% CI, 1.07–3.99; P = 0.031), and deep myometrial invasion (adjusted OR, 4.61; 95% CI, 2.67–7.93; P < 0.001) remained independent risk factors associated with the presence of LVSI. In a post hoc analysis, we examined a frequency of LVSI between the TLH-IUM group and the TLH-without-IUM group. We found that the frequency of LVSI was similar between the 2 groups (15.1% vs 21.1%, P = 0.35).
TABLE 2.
Risk factors for LVSI
| Before Propensity Score Matching |
After Propensity Score Matching |
|||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Univariable |
Multivariable |
|
Univariable |
Multivariable |
||||||||
| Characteristics | No. | Positive LVSI, % | OR (95% CI) | P | OR (95% CI) | P | No. | Positive LVSI, % | OR (95% CI) | P | OR (95% CI) | P |
| Age, y | ||||||||||||
| <60 | 377 | 10.9 | 1 | 1 | 125 | 16.0 | 1 | 1 | ||||
| ≥60 | 236 | 25.6 | 2.81 (1.81–4.35) | <0.001 | 2.16 (1.25–3.71) | 0.005 | 83 | 27.2 | 1.96 (0.99–3.88) | 0.054 | 3.38 (1.10–10.4) | 0.03 |
| Ethnicity | ||||||||||||
| Hispanic | 364 | 12.7 | 1 | 125 | 17.6 | 1 | ||||||
| Non-Hispanic | 249 | 22.3 | 1.97 (1.28–3.03) | 0.002 | 88 | 24.1 | 0.67 (0.34–1.33) | 0.26 | ||||
| BMI, kg/m2 | ||||||||||||
| <40 | 449 | 19.3 | 1 | 146 | 22.9 | 1 | ||||||
| ≥40 | 164 | 9.1 | 2.38 (1.33–4.25) | 0.003 | 62 | 14.5 | 0.57 (0.26–1.28) | 0.17 | ||||
| Hysterectomy type | ||||||||||||
| TAH | 194 | 19.9 | 1 | 104 | 19.6 | 1 | ||||||
| TLH-IUM | 419 | 15.1 | 0.72 (0.46–1.12) | 0.14 | 104 | 21.2 | 1.10 (0.56–2.17) | 0.78 | ||||
| Type of IUM | ||||||||||||
| VCare | 375 | 15.5 | 1 | 95 | 20.0 | 1 | ||||||
| Other types* | 43 | 11.6 | 0.72 (0.27–1.90) | 0.51 | 9 | 33.3 | 2.00 (0.46–8.74) | 0.36 | ||||
| Stage | ||||||||||||
| I–II | 514 | 10.9 | 1 | 1 | 164 | 11.0 | 1 | 1 | ||||
| III–IV | 99 | 46.4 | 7.05 (4.33–11.5) | <0.001 | 2.25 (1.20–4.22) | 0.011 | 44 | 55.8 | 10.2 (4.68–22.1) | <0.001 | 5.60 (1.79–17.5) | 0.003 |
| Histology | ||||||||||||
| Endometrioid | 514 | 12.7 | 1 | 161 | 16.1 | 1 | 1 | |||||
| Nonendometrioid | 99 | 37.1 | 4.06 (2.49–6.61) | <0.001 | 47 | 35.6 | 2.87 (1.37–6.01) | 0.005 | 7.69 (1.39–42.5) | 0.02 | ||
| Tumor grade | ||||||||||||
| Low-grade | 474 | 8.7 | 1 | 1 | 148 | 9.5 | 1 | 1 | ||||
| High-grade | 139 | 43.8 | 8.19 (5.14–13.0) | <0.001 | 3.83 (2.13–6.88) | <0.001 | 60 | 48.3 | 8.93 (4.20–19.0) | <0.001 | 18.2 (3.57–92.5) | <0.001 |
| Tumor size, cm | ||||||||||||
| <2 | 211 | 7.6 | 1 | 1 | 56 | 5.4 | 1 | 1 | ||||
| ≥2 | 373 | 21.0 | 3.22 (1.82–5.68) | <0.001 | 2.06 (1.07–3.99) | 0.03 | 128 | 25.8 | 6.14 (1.80–21.0) | 0.004 | 7.00 (1.15–42.7) | 0.04 |
| Deep myometrial invasion | ||||||||||||
| No | 474 | 8.5 | 1 | 1 | 154 | 7.8 | 1 | 1 | ||||
| Yes | 137 | 44.9 | 8.80 (5.51–14.1) | <0.001 | 4.61 (2.67–7.93) | <0.001 | 52 | 57.7 | 16.1 (7.21–36.1) | 0.001 | 8.01 (2.73–23.5) | <0.001 |
| Pelvic cytology | ||||||||||||
| Negative | 492 | 12.7 | 1 | 157 | 15.4 | 1 | ||||||
| Positive | 60 | 36.7 | 3.99 (2.21–7.18) | <0.001 | 22 | 45.5 | 4.58 (1.78–11.8) | 0.002 | ||||
| Atypical | 2 | 50.0 | 6.89 (0.43–111.5) | 0.17 | 1 | 100 | n.a. | 0.99 | ||||
Binary logistic regression models for multivariate analysis entered all the listed variables. Significant variables are emboldened. In the original data, 29 cases were missing tumor size, 2 cases were missing myometrial invasion, and 58 cases were missing pelvic cytology results.
Other types of IUM include RUM/Koh, HUMI, and Endopath.
n.a., not available.
Propensity score matching was performed (Tables 1 and 2). After matching, all variables except EBL were statistically similar between the TLH-IUM group and TAH group (all, P > 0.05). On univariate analysis, frequency of LVSI was similar between the TLH-IUM group and the TAH group (21.2% vs 19.6%, P = 0.78). On multivariate analysis, age of at least 60 years (adjusted OR, 3.38; 95% CI, 1.10–10.4; P = 0.03), stage III–IV disease (adjusted OR, 5.60; 95% CI, 1.79–17.5; P = 0.003), nonendometrioid histology (adjusted OR, 7.69; 95% CI, 1.39–42.5; P = 0.02), high-grade tumor (adjusted OR, 18.2; 95% CI, 3.57–92.5; P < 0.001), tumor size of at least 2 cm (adjusted OR, 7.00; 95% CI, 1.15–42.7; P = 0.04), and deep myometrial invasion (adjusted OR, 8.01; 95% CI, 2.73–32.5; P < 0.001) remained independent risk factors for increased risk of LVSI in endometrial cancer.
Survival outcome of women in the TLH-IUM group was examined. The median follow-up time was 25.6 months. There were 35 women who developed recurrence and 14 women who died of endometrial cancer. Type of IUM was not associated with DFS (5-year rates for VCare vs others, 85% vs 90%; P = 0.82) and CSS (5-year rates for VCare vs others, 94% vs 94%; P = 0.52). Diagnostic modality including endometrial biopsy (n = 293), Vabra aspirator® (n = 70), dilation and curettage (n = 42), and hysteroscopy (n = 14) was not associated with survival (DFS, P = 0.65; CSS, P = 0.50).
Systematic Review
Selection schema for systematic review is shown in Figure 2. Our search identified 203 relevant articles, for which titles and abstracts were screened. Of these, 12 articles meeting our criteriawere reviewed and 2617 cases of endometrial cancer treated with initial staging hysterectomy by TLH-IUM or TAH were examined.9,10,17,21–29 Details of individual study are shown in Table 4. Pooled demographic results of a systematic review for the frequency of LVSI in endometrial cancer are summarized in Table 3. There were 1371 TLH-IUM and 1246 TAH cases that met the eligibility criteria for inclusion. In a pooled analysis of our systematic literature review, the frequency of LVSI was 15.0% for TLH-IUM cases and 13.6% for TAH cases. Similar to our institutional cohort, there was no statistically significant difference in LVSI between the 2 groups (P = 0.31). Histologic subtype, tumor grade, and cancer stage in the TLH-IUM group were significantly different when compared with patients undergoing TAH (all, P < 0.05).
FIGURE 2.
Selection criteria for systematic literature review.
TABLE 4.
Details of individual study for systematic literature
| LVSI Results |
|||||||
|---|---|---|---|---|---|---|---|
| Author | Year | No. (TLH vs TAH) | Study Design | Population of TLH/IUM* (+), No. (%) | TLH/IUM (+), No. (%) | TAH, No. (%) | P |
| Zhang et al24 | 2014 | 214 vs 242 | Hospital-based control study | Histology: endometrioid, 191 (89); eonendometrioid, 23 (11) | 39 (18) | 44 (18) | 0.99 |
| Retrospective | Tumor grade: G1, 101 (49); G2–3, 105 (51) | ||||||
| Stage: I, 173 (81); II–IV 41 (19) | |||||||
| Hopkins et al23 | 2014 | 45 vs 54 | Hospital-based control study | Tumor grade: G1, 29 (69); G2–3, 13 (31) | 12 (27) | 26 (48) | 0.03 |
| Retrospective | Stage: I, 41 (91); II–IV, 4 (9) | ||||||
| Momeni et al25 | 2013 | 270 vs 287 | Hospital-based control study | Histology: endometrioid, 270 (100%) | 36 (13) | 59 (21) | 0.02 |
| Retrospective | Tumor grade: G1, 113 (42); G2–3, 156 (58) | 17 (8)† | 20 (10)† | 0.40 | |||
| Stage: I, 245 (91); II–IV, 25 (9) | |||||||
| Lee et al17 | 2013 | 110 | Hospital-based control study | Histology: endometrioid, 98 (89); nonendometrioid, 12 (11) | 13% vs 9% with or without IUM | 0.76 | |
| Prospective | Tumor grade: G1, 65 (59); G2–3, 44 (40) | ||||||
| Stage: I, 96 (87); II–IV, 14 (13) | |||||||
| Krizova et al21 | 2011 | 57 vs 103 | Histopathologic review | Hysterectomy with or without IUM | 16 (28) | 4 (4) | <0.001 |
| Retrospective | Histology, grade, stage: n.a. | ||||||
| Fanfani et al26 | 2011 | 114 vs 204 | Hospital-based control study | Histology: endometrioid, 87 (76.3); nonendometrioid, 27 (23.7) | 4.9% | 3.7% | 0.75 |
| Retrospective | Tumor grade: G1, 54 (47.4); G2–3, 60 (52.6) | ||||||
| Stage: I, 106 (93.0); II–III, 8 (7.0) | |||||||
| Nevadunsky et al27 | 2010 | 66 vs 43 | Hospital-based control study | Histology: endometrioid, 56 (85); nonendometrioid, 10 (15) | 10 (15) | 4 (9) | 0.37 |
| Retrospective | Tumor grade: G1, 35 (55); G2–3, 24 (36) | ||||||
| Stage: I, 53 (87); II–IV, 8 (13) | |||||||
| Folkins et al28 | 2010 | 58 vs 39 | Hospital-based control study | Histology: endometrioid, 58 (100) | 9 (16) | 3 (7) | 0.08 |
| Retrospective | Tumor grade: G1–2, 58 (100) | ||||||
| Stage: undetected | |||||||
| Kitahara et al22 | 2009 | 21 vs 28 | Hospital-based control study | Histology: undetected | 7 (33) | 0 (0) | 0.001 |
| Retrospective | Tumor grade: G1, 13 (62); G2, 8 (38) | ||||||
| Stage I: 21 (100) | |||||||
| Sonoda et al29 | 2001 | 131 vs 246 | Hospital-based control study | Histology: endometrioid, 131 (100) | 9 (7) | 22 (9) | 0.62 |
| Retrospective | Tumor grade: G1, 104 (81); G2–3, 25 (19) | ||||||
| Stage I: 131 (100) | |||||||
| Machida et al10 | 2016 | 333 | Hospital-based control study | Histology: endometrioid, 287 (86.2); nonendometrioid, 46 (13.8) | 50 (15.0) | ||
| Retrospective | Tumor grade: G1, 227 (68.2); G2–3, 106 (11.8) | ||||||
| Stage: I, 268 (80.5); II–IV, 39 (19.5) | |||||||
| Logani et al9 | 2008 | 7 | Histopathologic review | All cases are endometrioid, stage I | 5 (71) | ||
| Retrospective | |||||||
Number (%) is shown.
Includes laparoscopically assisted vaginal hysterectomy and robot-assisted hysterectomy.
Proportion of LVSI in stage IA.
G, histologic tumor grade; n.a., not available.
TABLE 3.
Pooled analysis for systematic literature review
| Characteristics | TLH-IUM* (n = 1371) | TAH (n = 1246) | P |
|---|---|---|---|
| Age, y† | 60.0 (31–92) | 61.9 (32–91) | n.a. |
| Histology, n (%) | <0.001 | ||
| Endometrioid | 1126 (82.1) | 1002 (80.4) | |
| Nonendometrioid | 115 (8.4) | 61 (4.9) | |
| Unknown | 130 (9.5) | 183 (14.7) | |
| Tumor grade, n (%) | <0.001 | ||
| 1 | 782 (57.0) | 437 (35.1) | |
| 2–3 | 517 (37.7) | 670 (53.8) | |
| Unknown | 72 (5.3) | 139 (11.1) | |
| Stage% | 0.01 | ||
| I | 1141 (83.2) | 976 (78.3) | |
| II–IV | 139 (10.1) | 162 (13.0) | |
| Unknown | 91 (6.7) | 108 (8.7) | |
| LVSI% | 206 (15.0) | 170 (13.6) | 0.31 |
| IUM% | |||
| VCare | 582 (42.5) | ||
| RUMI | 81 (5.9) | ||
| Unknown | 708 (51.6) |
Median (range) or number (%) is shown. χ2 Test for P values. Significant variables are emboldened.
Includes laparoscopically assisted vaginal hysterectomy and robot-assisted hysterectomy.
Median of 12 articles reporting median age.
n.a., not available.
DISCUSSION
Our study suggests that IUM insertion during minimally invasive hysterectomy for endometrial cancer is not associated with increased risk of LVSI as shown in our institutional data set. Subsequently, a propensity score matching analysis in our study cohort and a systematic literature review of relevant published articles were performed and observed similar results, further supporting our conclusion that IUM does not seem to pose an increased risk of iatrogenic LVSI at the time of minimally invasive hysterectomy for endometrial cancer.
The National Comprehensive Cancer Network guidelines recommend that patients with endometrial cancer undergo thorough surgical staging and be stratified by adverse risk factors (age, positive LVSI, tumor size, depth of invasion, and lower uterine segment involvement); the decision regarding adjuvant therapy should be based on the stratification of these risk factors.30,31 As such, the presence of LVSI in endometrial cancer is a well-known predictive factor for relapse and decreased DFS.32,33 However, the potential risk of LVSI with IUM use during minimally invasive hysterectomy for endometrial cancer raises an important clinical question with controversial implications for endometrial cancer survival.
Prior studies have suggested that TLH-IUM, including laparoscopic-assisted vaginal hysterectomy and robot-assisted hysterectomy for endometrial cancer, is associated with increased frequency of LVSI compared with TAH (13%–33% vs 0–48%; Table 4).20–22 Supported by observations that artifacts of pseudoinvasion are more common in TLH-IUM specimens compared with TAH, some studies indicate that vascular pseudoinvasion may be caused by the pressure created with intrauterine balloon inflation.8,20 In contrast, other recent studies have reported that IUM use during minimally invasive hysterectomy is not associated with increased frequency of LVSI in endometrial cancer (TLH-IUM vs TAH, 4.9%–18% vs 3.7%–21%; Table 4).23–28 However, weaknesses of these previous studies include differing clinicopathologic backgrounds between the groups analyzed, relatively small sample size (median, 62 cases in the TLH-IUM group), and lack of description of the IUM type for each candidate. There is insufficient evidence evaluating LVSI either caused by IUM insertion or originally present in the endometrium. These weaknesses make the results difficult to interpret for proper analysis and integration into clinical practice. Therefore, concerns regarding the potential increased risk of disease spread with IUM use have remained unsettled.
Our study revealed a 15.1% prevalence of LVSI among 419 cases of TLH-IUM for endometrial cancer, and the rates of LVSI were not significantly different between TLH-IUM and TAH even after adjusting for clinicopathologic backgrounds using propensity analysis. In a pooled analysis of prior published studies within our systematic literature review, the frequency of LVSI was quite similar between the TLH-IUM group and the TAH group (15.0% vs 13.6%). With an α level of 0.025 and a noninferiority margin of 5% for tolerable clinical difference in the frequency of LVSI between the 2 groups, we obtained adequate sample sizes to power this analysis for systematic review cohort (>80%). Both our study cohort and systematic literature review cohort show similar results regarding the frequency of LVSI for minimally invasive surgery using IUM. Altogether, this supports our conclusion that there is no apparent association between IUM use during TLH and increased risk of LVSI compared with conventional abdominal hysterectomy.
The possible biological plausibility for no association between IUM use during TLH and frequency of LVSI merits an in-depth speculation. The characteristics of LVSI in endometrial cancer is that LVSI is generally located in the deep layer of the uterine myometrium far away from the intrauterine cavity, and LVSI located in the superficial layer of the uterine myometrium is rare and not associated with survival.34 This implies that the mechanical pressure effects of the balloon via IUM insertion would be unlikely to transmit into the deep layer of the myometrium, resulting in no increase in the frequency of LVSI with IUM use.
One of salient outcomes related to IUM use in the surgical management of women with endometrial cancer is survival as mentioned previously. A hypothetical risk of IUM use would be an increased risk of recurrence due to iatrogenic LVSI resulting in nodal tumors spread related to IUM insertion. However, our study not only found that IUM use does not increase a risk of LVSI but also found that type of IUM was not associated with endometrial cancer survival. These findings partly support what has shown in a recent large-scale study that demonstrated no association between IUM use and survival of women with endometrial cancer.35 Therefore, our results further support the safety of IUM use during minimally invasive hysterectomy for endometrial cancer.
Strengths of this study include its relatively large sample size and adjustment for patient clinical demographic characteristics by propensity score matching analysis. The quality of our analysis in this study was greatly enhanced by propensity score matching, which has been argued to be a powerful means of accounting for baseline confounding and selection biases.18,19 In addition, systematic review of LVSI in TLH-IUM for endometrial cancer with a pooled analysis validated the quality and consistency of our results.
Limitations of this study include its retrospective nature with potential missing confounding factors, lack of information regarding the indications for IUM use and type, duration of IUM insertion during the surgery, and nonrandomized patient selection for hysterectomy procedure type. In addition, selection bias may exist because only published cases were used for the systematic literature review. Lack of central pathology review is also a weakness of the study because the diagnosis criteria for LVSI vs pseudo-invasion may differ among pathologists. The optimal investigation to examine this association would be a prospective study of patients with similar demographics and endometrial cancer surgically staged using hysterectomy with IUM.
In summary, IUM use during minimally invasive surgical staging hysterectomy is not associated with an increase in the frequency of LVSI. Intrauterine manipulator use seems acceptable in the treatment for women with early-stage endometrial cancer.
Acknowledgments
Funding support: This study was supported by the Ensign Endowment for Gynecologic Cancer Research (K.M.).
Footnotes
The authors declare no conflict of interest in this study.
The abstract of the study was presented at 18th APAGE Annual Congress 2017 (Okayama, Japan; September 6–9, 2017) and 2017 Western Association of Gynecologic Oncology (Rancho Mirage, CA; June 14–17, 2017).
REFERENCES
- 1.Siegel RL, Miller KD, Jemal A. Cancer statistics, 2017. CA Cancer J Clin. 2017;67:7–30. [DOI] [PubMed] [Google Scholar]
- 2.Wright JD, Ananth CV, Lewin SN, et al. Robotically assisted vs laparoscopic hysterectomy among women with benign gynecologic disease. JAMA. 2013;309:689–698. [DOI] [PubMed] [Google Scholar]
- 3.Walker JL, Piedmonte MR, Spirtos NM, et al. Laparoscopy compared with laparotomy for comprehensive surgical staging of uterine cancer: Gynecologic Oncology Group Study LAP2. J Clin Oncol. 2009;27:5331–5336. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Leitao MM Jr, Malhotra V, Briscoe G, et al. Postoperative pain medication requirements in patients undergoing computer-assisted (“Robotic”) and standard laparoscopic procedures for newly diagnosed endometrial cancer. Ann Surg Oncol. 2013;20:3561–3567. [DOI] [PubMed] [Google Scholar]
- 5.Cardenas-Goicoechea J, Adams S, Bhat SB, et al. Surgical outcomes of robotic-assisted surgical staging for endometrial cancer are equivalent to traditional laparoscopic staging at a minimally invasive surgical center. Gynecol Oncol. 2010;117:224–228. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Walker JL, Piedmonte MR, Spirtos NM, et al. Recurrence and survival after random assignment to laparoscopy versus laparotomy for comprehensive surgical staging of uterine cancer: Gynecologic Oncology Group LAP2 Study. J Clin Oncol. 2012;30:695–700. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Wright JD, Burke WM, Tergas AI, et al. Comparative effectiveness of minimally invasive hysterectomy for endometrial cancer. J Clin Oncol. 2016;34:1087–1096. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Tinelli R, Cicinelli E, Tinelli A, et al. Laparoscopic treatment of early-stage endometrial cancer with and without uterine manipulator: our experience and review of literature. Surg Oncol. 2016;25:98–103. [DOI] [PubMed] [Google Scholar]
- 9.Logani S, Herdman AV, Little JV, et al. Vascular “pseudo invasion” in laparoscopic hysterectomy specimens: a diagnostic pitfall. Am J Surg Pathol. 2008;32:560–565. [DOI] [PubMed] [Google Scholar]
- 10.Machida H, Casey JP, Garcia-Sayre J, et al. Timing of intrauterine manipulator insertion during minimally invasive surgical staging and results of pelvic cytology in endometrial cancer. J Minim Invasive Gynecol. 2016;23:234–241. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.von Elm E, Altman DG, Egger M, et al. Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. BMJ. 2007;335:806–808. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Matsuo K, Moeini A, Cahoon SS, et al. Weight change pattern and survival outcome of women with endometrial cancer. Ann Surg Oncol. 2016;23:2988–2997. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Matsuo K, Cahoon SS, Yoshihara K, et al. Association of low-dose aspirin and survival of women with endometrial cancer. Obstet Gynecol. 2016;128:127–137. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Mariani A, Webb MJ, Keeney GL, et al. Low-risk corpus cancer: is lymphadenectomy or radiotherapy necessary? Am J Obstet Gynecol. 2000;182:1506–1519. [DOI] [PubMed] [Google Scholar]
- 15.FIGO Committee on Gynecologic Oncology. FIGO staging for carcinoma of the vulva, cervix, and corpus uteri. Int J Gynaecol Obstet. 2014;125:97–98. [DOI] [PubMed] [Google Scholar]
- 16.Stroup DF, Berlin JA, Morton SC, et al. Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis Of Observational Studies in Epidemiology (MOOSE) Group. JAMA. 2000;283:2008–2012. [DOI] [PubMed] [Google Scholar]
- 17.Lee M, Kim YT, Kim SW, et al. Effects of uterine manipulation on surgical outcomes in laparoscopic management of endometrial cancer: a prospective randomized clinical trial. Int J Gynecol Cancer. 2013;23:372–379. [DOI] [PubMed] [Google Scholar]
- 18.Blackwelder WC. “Proving the null hypothesis” in clinical trials. Control Clin Trials. 1982;3:345–353. [DOI] [PubMed] [Google Scholar]
- 19.Joffe MM, Rosenbaum PR. Invited commentary: propensity scores. Am J Epidemiol. 1999;150:327–333. [DOI] [PubMed] [Google Scholar]
- 20.Austin PC. An introduction to propensity score methods for reducing the effects of confounding in observational studies. Multivariate Behav Res. 2011;46:399–424. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Krizova A, Clarke BA, Bernardini MQ, et al. Histologic artifacts in abdominal, vaginal, laparoscopic, and robotic hysterectomy specimens: a blinded, retrospective review. Am J Surg Pathol. 2011;35:115–126. [DOI] [PubMed] [Google Scholar]
- 22.Kitahara S, Walsh C, Frumovitz M, et al. Vascular pseudoinvasion in laparoscopic hysterectomy specimens for endometrial carcinoma: a grossing artifact? Am J Surg Pathol. 2009;33:298–303. [DOI] [PubMed] [Google Scholar]
- 23.Hopkins MR, Richmond AM, Cheng G, et al. Lymphovascular space invasion in robotic surgery for endometrial cancer. JSLS. 2014;18:pii: e2014.00021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Zhang C, Havrilesky LJ, Broadwater G, et al. Relationship between minimally invasive hysterectomy, pelvic cytology, and lymph vascular space invasion: a single institution study of 458 patients. Gynecol Oncol. 2014;133:211–215. [DOI] [PubMed] [Google Scholar]
- 25.Momeni M, Kolev V, Cardenas-Goicoechea J, et al. Does the type of surgery for early-stage endometrial cancer affect the rate of reported lymphovascular space invasion in final pathology specimens? Am J Obstet Gynecol. 2013;208:71.e1–71.e6. [DOI] [PubMed] [Google Scholar]
- 26.Fanfani F, Gagliardi ML, Zannoni GF, et al. Total laparoscopic hysterectomy in early-stage endometrial cancer using an intrauterine manipulator: is it a bias for frozen section analysis? Case-control study. J Minim Invasive Gynecol. 2011;18:184–188. [DOI] [PubMed] [Google Scholar]
- 27.Nevadunsky N, Clark R, Ghosh S, et al. Comparison of robot-assisted total laparoscopic hysterectomy and total abdominal hysterectomy for treatment of endometrial cancer in obese and morbidly obese patients. J Robot Surg. 2010;4:247–252. [DOI] [PubMed] [Google Scholar]
- 28.Folkins AK, Nevadunsky NS, Saleemuddin A, et al. Evaluation of vascular space involvement in endometrial adenocarcinomas: laparoscopic vs abdominal hysterectomies. Mod Pathol. 2010;23:1073–1079. [DOI] [PubMed] [Google Scholar]
- 29.Sonoda Y, Zerbe M, Smith A, et al. High incidence of positive peritoneal cytology in low-risk endometrial cancer treated by laparoscopically assisted vaginal hysterectomy. Gynecol Oncol. 2001;80:378–382. [DOI] [PubMed] [Google Scholar]
- 30.Keys HM, Roberts JA, Brunetto VL, et al. A phase III trial of surgery with or without adjunctive external pelvic radiation therapy in intermediate risk endometrial adenocarcinoma: a Gynecologic Oncology Group study. Gynecol Oncol. 2004;92:744–751. [DOI] [PubMed] [Google Scholar]
- 31.Creutzberg CL, van Putten WL, Koper PC, et al. Surgery and postoperative radiotherapy versus surgery alone for patients with stage-1 endometrial carcinoma: multicentre randomised trial. PORTEC Study Group. Post Operative Radiation Therapy in Endometrial Carcinoma. Lancet. 2000;355: 1404–1411. [DOI] [PubMed] [Google Scholar]
- 32.Briet JM, Hollema H, Reesink N, et al. Lymphvascular space involvement: an independent prognostic factor in endometrial cancer. Gynecol Oncol. 2005;96:799–804. [DOI] [PubMed] [Google Scholar]
- 33.Gemer O, Arie AB, Levy T, et al. Lymphvascular space involvement compromises the survival of patients with stage I endometrial cancer: results of a multicenter study. Eur J Surg Oncol. 2007;33:644–647. [DOI] [PubMed] [Google Scholar]
- 34.Matsuo K, Garcia-Sayre J, Medeiros F, et al. Impact of depth and extent of lymphovascular space invasion on lymph node metastasis and recurrence patterns in endometrial cancer. J Surg Oncol. 2015;112:669–676. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Uccella S, Bonzini M, Malzoni M, et al. The effect of a uterine manipulator on the recurrence and mortality of endometrial cancer: a multi-centric study by the Italian Society of Gynecological Endoscopy. Am J Obstet Gynecol. 2017;216:592.e1–592.e11. [DOI] [PubMed] [Google Scholar]