Effect of Substantial Lymphovascular Space Invasion on Location of First Disease Recurrence in Surgical Stage I Endometrioid Endometrial Adenocarcinoma

Abstract

Objective

Lymphovascular invasion can predict nodal spread and recurrence in endometrioid endometrial cancer; however, the impact of lymphovascular invasion quantification on local versus distant recurrence in surgically staged patients has not been established.

Methods

This multicenter retrospective cohort study included surgically staged patients with International Federation of Obstetrics and Gynecology (FIGO) 2009 stage I node-negative endometrioid endometrial cancer. Patients were treated between 01/01/2012-12/31/2019 at two tertiary cancer centers. Staging included total hysterectomy and lymph node assessment. Lymphovascular invasion extent was defined using the World Health Organization criteria as focal (<5 vessels involved on at least 1 pathology slide) or substantial (≥5 vessels involved). Recurrence and death were considered events. Competing risk analysis was performed and controlled for multicenter clustering.

Results

Overall, 1555 patients met inclusion criteria; 65 (4.2%) had substantial, 119 (7.7%) focal, and 1371 (88.2%) no invasion. Median follow-up was 61.5 months (range, 0.8-133.9). There were 173 evaluable events among 1554 patients: 56 local recurrences, 43 distant recurrences, and 74 deaths without recurrence. Deep (≥50%) myoinvasion and grade 3 histology were more frequent in patients with substantial invasion. Overall, 323 patients (20.8%) received adjuvant therapy. The 5-year cumulative incidence failure rates for any recurrence were 6.0% for no, 19.5% for focal, and 19.0% for substantial invasion. Compared to no lymphovascular invasion, substantial invasion was associated with increased risk of distant recurrence (adjusted hazard ratio [aHR]: 2.29; 95% CI: 1.17-4.46).

Conclusion

In patients with surgical stage I endometrioid endometrial cancer, focal and substantial lymphovascular invasion were associated with a 3-fold increased risk of cumulative incidence failure versus no lymphovascular invasion. Patients with substantial invasion had more deeply invasive and grade 3 tumors and appear to experience more distant than local recurrences. These findings challenge the FIGO 2023 staging classification that combines no and focal lymphovascular invasion into one risk category.

Introduction

Endometrioid endometrial cancer has a relatively favorable prognosis, but approximately 2% to 10% of International Federation of Obstetrics and Gynecology (FIGO) 2009 stage I cases recur.[1] Lymphovascular space invasion is defined as the presence of tumor cells within blood or lymphatic vessels outside the primary tumor.[2] It is considered one of the strongest predictors of recurrence in patients with FIGO 2009 stage I endometrioid endometrial cancer,[3] warranting special consideration for adjuvant treatment.[4, 5]

In 2015, researchers identified an increased risk of pelvic nodal recurrence in non-surgically staged patients with apparent early-stage endometrioid endometrial cancer exhibiting “substantial” levels of lymphovascular invasion.[6] Over the last decade, increasing efforts have been made to determine a clinically meaningful cutoff that can identify patients with early-stage disease who would benefit from adjuvant treatment, with very heterogeneous findings.[7–9] A study by Pifer et al. of surgically staged patients with stage I endometrioid endometrial cancer and pathologically confirmed negative lymph nodes, using a cutoff of 4 or more vessels involved on at least one pathology slide to define substantial lymphovascular invasion, did not find any significant correlation between extent of invasion and pattern of recurrence.[8]

Given the inclusion of extent of lymphovascular invasion in the 2023 FIGO endometrial cancer staging system as a determinant of stage,[10] we sought to examine the rate and location of first disease recurrence associated with lymphovascular invasion extent using the current World Health Organization (WHO) definition,[11] while accounting for nodal status by including only patients with FIGO 2009 surgical stage I endometrioid endometrial cancer with confirmed negative lymph nodes at initial staging surgery.

Methods

We conducted a multi-institutional retrospective cohort study involving two tertiary care centers. Patients who received treatment at either centers between January 1, 2012, and December 31, 2019, were included. Our primary objective was to characterize the recurrence rate according to extent of lymphovascular invasion and to characterize the pattern of first disease recurrence in patients diagnosed with FIGO 2009 surgical stage I endometrioid endometrial cancer. This research was approved by the Institutional Review Boards of Memorial Sloan Kettering Cancer Center and Norwegian Radium Hospital/Oslo University Hospital, and the study was conducted in accordance with the Declaration of Helsinki. Informed consent was waived per each institution’s guidelines.

The study included patients diagnosed with FIGO 2009 surgical stage IA or IB endometrioid endometrial cancer, who underwent complete surgical staging including total hysterectomy with or without salingo-oophorectomy, and bilateral surgical pathologic evaluation of lymph nodes. Patients with positive nodes including isolated tumor cells, incomplete nodal assessment (empty nodal packets), positive peritoneal cytology, and synchronous ovarian tumors were excluded.

Patients were categorized into 3 groups based on the extent of lymphovascular invasion: no invasion, focal invasion, or substantial invasion. Lymphovascular invasion was defined as the presence of tumor cells within an endothelium-lined space beyond the invasive front of the tumor in the myometrium. The classification of lymphovascular invasion extent in this study adheres to the 2023 WHO definition,[11] which designates the involvement of 5 vessels on a single pathology slide as the threshold distinguishing focal from substantial invasion. Tumors were classified as having no invasion if no vessel involvement was identified on any pathology slide, focal invasion if 1 to 4 vessels were identified, and substantial invasion if 5 or more vessels were identified. To ensure accuracy in evaluation, a pathology review was undertaken by experienced gynecologic pathologists from both centers (L.H.E., R.A., B.D.). This review targeted tumors initially reported as having identified or suspected lymphovascular invasion for a detailed assessment of the extent of invasion. At our centers, a tumor is regarded as having a suspicion of lymphovascular invasion when the area suspected of invasion remains inconclusive, even with the use of immunohistochemical techniques to highlight endothelial cells. To mitigate potential discrepancies in interpretation among observers, the pathologists met regularly during the study’s planning stage to unify their definitions and approaches.

The method of nodal evaluation at surgery adhered to local institutional guidelines.[4, 5] The decision for postsurgical treatment was made individually, involved a shared decision between the patient and clinician, and varied by center according to local treatment guidelines.[4, 5] Following initial treatment, patients were followed every 3 to 6 months for the first 2 to 3 years, then every 6 months for the following 2 to 3 years. Imaging during follow up was at the physician’s discretion and based on occurrence of new symptoms or concerns for recurrence in keeping with local guidelines.[4, 5] Recurrences were broadly categorized as local or distant. Local recurrences were categorized as pure vaginal, pelvic non-vaginal, or isolated pelvic nodal. Pelvic non-vaginal recurrences included cases with non-nodal pelvic organ involvement, such as serosal or peritoneal pelvic disease. Additionally, recurrences affecting multiple pelvic locations, that included nodal recurrences (2 cases), were also classified under pelvic recurrences. Distant recurrences were categorized as isolated paraaortic nodal or distant. Patients with multifocal spread, defined as those satisfying criteria for both local and distant recurrences, were considered as having distant recurrence for the purpose of this study.

Data on demographics, disease characteristics, treatment modalities, and outcomes were extracted from electronic health records and compiled in a centralized database. Clinical, surgical, and pathological characteristics are reported with descriptive statistics using medians with ranges and percentages when appropriate. Comparisons between groups were performed using the Kruskal-Wallis test for continuous variables and Fisher exact test for categorical variables.

The primary outcome was 5-year cumulative incidence failure rate for local and distant recurrences. The secondary outcome was multivariate analysis for survival outcome using competing risk analysis, with the competing events being local recurrence, distant recurrence, or death. Survival time was calculated from the date of surgery to the date of local or distant recurrence or death of any other cause, whichever occurred first. Patients with no events were censored at their last follow-up visit. One patient was excluded from the competing risk analyses due to an unknown location of recurrence. For adjuvant therapy as a time-dependent variable, landmark analyses were applied using a landmark time of 3 months.

Five-year cumulative incidence failure rates were estimated using the Gray method.[12] Hazard ratios and P values were obtained using a method developed by Zhou et al. as an extension of the Fine-Gray proportional hazards model by considering clustering effect.[13] A multivariate model was built based on the univariate analysis results. P values were calculated as two-sided, and those <0.05 were considered statistically significant. Analyses were performed in R 4.3.1. In accordance with the journal’s guidelines, we will provide our data for independent analysis by a selected team by the Editorial Team for the purposes of additional data analysis or for the reproducibility of this study in other centers if such is requested.

Results

A total of 1555 patients met inclusion criteria (Supplemental Figure 1). Overall, 184 patients (11.8%) had tumors with lymphovascular invasion. After pathology re-review, 119 tumors (7.7%) were classified as having focal invasion, while 65 (4.2%) had substantial invasion (Table 1).

Table 1 –

Demographic, clinical, and pathologic characteristics of patients with stage I endometrioid endometrial cancer treated surgically with negative nodes, according to extent of lymphovascular space invasion (N=1555).

Characteristic	No LVSI n=1371 (88.2%)	Focal LVSI n=119 (7.7%)	Substantial LVSI n=65 (4.2%)	P value
Age at surgery, years (median, range)	60 (27-92)	67 (35-87)	64 (24-90)	<0.001
BMI at surgery, kg/m2 (median, range)	31 (15-70)	30 (19-59)	30 (22-52)	0.34
BMI ≥30 kg/m2 (n, %)	716 (52.3)	56 (47.9)	32 (49.2)	0.6
Missing	2	2	0
FIGO grade (n, %)				<0.001
G1	1011 (73.7)	51 (42.9)	21 (32.3)
G2	269 (19.6)	44 (37.0)	23 (35.4)
G3	91 (6.6)	24 (20.2)	21 (32.3)
Depth of myometrial invasion (n, %)				<0.001
None	839 (61.2)	2 (1.7)	1 (1.5)
<50%	417 (30.4)	73 (61.3)	29 (44.6)
≥50%	115 (8.4)	44 (37.0)	35 (53.8)
Method of nodal evaluation (n, %)				<0.001
SLN	998 (72.8)	81 (68.1)	32 (49.2)
LND	168 (12.2)	18 (15.1)	22 (33.8)
SLN+LND	205 (15.0)	20 (16.8)	11 (16.9)
Received adjuvant treatment (n, %)	207 (15.1)	80 (67.2)	36 (55.4)	<0.001
Type of adjuvant therapy (n, %)
IVRT	184 (13.4)	73 (61.3)	25 (38.5)	0.006
Chemotherapy	23 (1.7)	7 (5.9)	11 (16.9)

LVSI: lymphovascular space invasion, BMI: body mass index, FIGO: International Federation of Obstetrics and Gynecology, G: grade, SLN: sentinel lymph node biopsy, LND: lymphadenectomy, IVRT: intra-vaginal brachytherapy

Detailed clinico-demographic and pathological comparisons across the 3 groups are presented in Table 1. Differences between the groups are noted in age at surgery, at 60 years (range, 27-92) for the no lymphovascular invasion group versus 67 years (range, 35-87) for the focal and 64 years (range, 24-90) for the substantial groups (P<0.001). Deep myometrial invasion was found in 115 (8.4%), 44 (37.0%), and 35 patients (53.8%) with no, focal, and substantial invasion, respectively (P<0.001). Grade 3 tumors were found in 91 (6.6%), 24 (20.2%), and 21 patients (32.3%) with no, focal, and substantial invasion, respectively (P<0.001) (Figure 1). All patients were confirmed to have bilateral negative lymph nodes on final pathology. Overall, 323 patients (20.8%) received any form of adjuvant therapy, including 282 (18.1%) who received intravaginal brachytherapy and 41 (2.6%) who received chemotherapy. Intravaginal brachytherapy was given to 184 (13.4%), 73 (61.3%), and 25 patients (38.5%) with no, focal, and substantial lymphovascular invasion, respectively; chemotherapy was given to 23 (1.7%), 7 (5.9%), and 11 patients (16.9%) with no, focal, and substantial invasion, respectively (P=0.006).

Figure 1 –

Side-by-side comparison of histological characteristics, cumulative failure rates, and recurrence patterns in patients with stage I endometrioid endometrial cancer according to the extent of lymphovascular space invasion (LVSI; N=1555).

*1 patient with no LVSI was missing recurrence location and was excluded from analysis.

With a median follow-up duration of 61.5 months (range, 0.8-133.9), 99 patients experienced disease recurrence with a known location of recurrence during the study period; 66 patients in the no lymphovascular invasion group, 21 in the focal group, and 12 in the substantial group. Recurrences by anatomical site are listed in Supplemental Table 1. Among patients with observed disease recurrence, local recurrences occurred in 40 (60.6%) of 66 patients, 12 (57.1%) of 21 patients, and 4 (33.3%) of 12 patients with no, focal, and substantial lymphovascular invasion, respectively; distant recurrences occurred in 26 (39.4%) of 66 patients, 9 (42.9%) of 21 patients, and 8 (66.7%) of 12 patients, respectively. Of the 43 observed distant recurrences, 14 were multifocal: 7 (26.9%) of 26, 3 (33.3%) of 9, and 4 (50.0%) of 8 recurrences in the no, focal, and substantial invasion groups, respectively (Figure 1).

The 5-year cumulative incidence failure rates for any recurrence were 6.0% (95% CI: 4.7%-7.6%) for no lymphovascular invasion, 19.5% (95% CI: 12.6%-27.6%) for focal invasion, and 19.0% (95% CI: 10.4%-29.6%) for substantial invasion. The 5-year cumulative incidence failure rates for local recurrences were 3.6% (95% CI: 2.6%-4.9%), 11.5% (95% CI: 6.2%-18.6%, P=0.009), and 6.3% (95% CI: 2.0%-14.1%, P=0.39) for cases with no, focal, and substantial lymphovascular invasion, respectively; for distant recurrences, rates were 2.4% (95% CI: 1.6%-3.4%), 8.0% (95% CI: 3.9%-14.0%, P=0.004), and 12.7% (95% CI: 5.9%-22.3%, P<0.001) for each group, respectively (Table 2, Figure 2).

Table 2 –

Univariate competing risk analysis showing the 5-year cumulative incidence failure rates (%) based on extent of lymphovascular space invasion and other risk factors (N=1554^*).

		Local recurrence		P value	Distant recurrence		P value
Variable	N	Event	5-y CIF (95% CI)		Event	5-y CIF (95% CI)
Overall	1554	56	4.4% (3.4%-7.7%)	-	43	3.3% (2.4%-4.5%)	-
Extent of LVSI
None	1370	40	3.6% (2.6%-4.9%)	-	26	2.4% (1.6%-3.4%)	-
Focal	119	12	11.5% (6.2%-18.6%)	0.009	9	8.0% (3.9%-14.0%)	0.004
Substantial	65	4	6.3% (2.0%-14.1%)	0.39	8	12.7% (5.9%-22.3%)	<0.001
BMI at surgery ^
<30 kg/m2	746	23	3.6% (2.3%-5.3%)	-	20	3.0% (1.9%-4.5%)	-
≥30 kg/m2	804	33	5.4% (3.7%-7.5%)	<0.001	23	3.7% (2.4%-5.5%)	0.71
FIGO 2009 stage
IA	1360	43	4.0% (3.0%-5.4%)	-	30	2.7% (1.8%-3.8%)	-
IB	194	13	6.6% (3.6%-10.8%)	0.044	13	7.0% (3.9%-11.2%)	<0.001
Grade
G1	1082	22	2.7% (1.7%-4.0%)	-	8	1.1% (0.5%-2.1%)	-
G2	336	25	8.5% (5.7%-12.1%)	<0.001	20	6.4% (4.0%-9.6%)	<0.001
G3	136	9	7.0% (3.4%-12.3%)	<0.001	15	11.5% (6.7%-17.7%)	<0.001
Myoinvasion
None	842	13	1.9% (1.1%-3.2%)	-	5	0.8% (0.3%-1.8%)	-
<50%	518	30	6.9% (4.7%-9.6%)	<0.001	25	5.3% (3.5%-7.7%)	<0.001
≥50%	194	13	6.6% (3.6%-10.8%)	<0.001	13	7.0% (3.9%-11.2%)	<0.001
Adjuvant treatment (landmark analysis)
None	1172	37	3.7% (2.6%-5.0%)	-	27	2.8% (1.9%-4.1%)	-
Any	320	19	6.9% (4.3%-10.4%)	0.22	16	5.8% (3.4%-9.1%)	0.39

5-y CIF: 5-year cumulative incidence failure rate reported in %, CI: confidence interval, LVSI: lymphovascular space invasion, BMI: body mass index, FIGO: International Federation of Obstetrics and Gynecology, G: grade.

^*1 patient with no LVSI was missing recurrence location and was excluded from analysis.

^{^}BMI information was missing for 4 patients.

Figure 2 –

Five-year cumulative incidence failure rates for any recurrence according to the extent of lymphovascular space invasion (N=1554*).

CIF: cumulative incidence failure rates, LVSI: lymphovascular space invasion

*1 patient with no LVSI was missing recurrence location and was excluded from analysis.

On univariate analysis, extent of lymphovascular invasion, body mass index at time of surgery, histological grade, and depth of myoinvasion were significantly associated with disease recurrence (Supplemental Table 2). These factors were subsequently included in a multivariate competing risk model (Table 3). A pyramid diagram (Figure 1) visually represents results of the multivariate competing risk analysis focused on extent of lymphovascular invasion. Focal invasion was associated with an increased rate of both local and distant recurrences (Table 3), which is reflected by the triangle growth being even in all directions, indicating that no pattern of recurrence predominates in this group. In contrast, substantial invasion demonstrated a non-uniform increase in risk across the 3 axes (Figure 1), with a significant tendency toward distant recurrence (adjusted hazard ratio [aHR]: 2.29; 95% CI: 1.17-4.46; P=0.015) (Table 3).

Table 3 –

Multivariate competing risk analysis showing the adjusted hazard ratio of developing disease progression based on extent of lymphovascular space invasion and other risk factors (N=1550^*).

Variable	aHR for local recurrence (95% CI); Pv	aHR for distant recurrence (95% CI); Pv	aHR for death without recurrence (95% CI); Pv
Extent of LVSI
None	Reference	Reference	Reference
Focal	1.78 (0.92-3.43); 0.086	1.52 (0.37-6.21); 0.56	1.30 (1.10-1.54); 0.002
Substantial	1.04 (0.33-3.30); 0.94	2.29 (1.17-4.46); 0.015	1.98 (1.12-3.50); 0.019
Age at surgery
1-year increment	1.04 (1.03-1.06); <0.001	1.04 (1.01-1.06); 0.003	1.12 (1.10-1.14); <0.001
BMI at surgery
<30 kg/m2	Reference	Reference	Reference
≥30 kg/m2	1.71 (1.56-1.88); <0.001	1.48 (0.70-3.15); 0.31	1.18 (0.79-1.76); 0.41
Grade
G1	Reference	Reference	Reference
G2	2.32 (1.41-3.82); <0.001	4.60 (3.40-6.21); <0.001	0.47 (0.46-0.47); <0.001
G3	1.81 (1.52-2.16); <0.001	7.20 (2.16-24.02); 0.001	0.61 (0.37-1.01); 0.052
Depth of invasion
None	Reference	Reference	Reference
<50%	1.99 (1.59-2.50); <0.001	3.16 (3.12-3.20); <0.001	1.27 (0.93-1.73); 0.13
≥50%	1.54 (1.17-2.03); 0.002	2.48 (1.98-3.09); <0.001	1.36 (1.24-1.50); <0.001

aHR: adjusted hazard ratio, CI: confidence interval, Pv: P value, LVSI: lymphovascular space invasion, BMI: body mass index, G: grade.

^*4 patients missing BMI information, and 1 patient missing recurrence location were excluded from the analysis.

The multivariate analysis also revealed that grades 2 and 3 disease had higher risk of local recurrence compared to grade 1 disease, with aHRs of 2.32 (95% CI: 1.41-3.82, P<0.001) and 1.81 (95% CI: 1.52-2.16, P<0.001), respectively. Grades 2 and 3 disease also had higher risk of distant recurrence compared to grade 1 disease, with aHRs of 4.60 (95% CI: 3.40-6.21, P<0.001) and 7.20 (95% CI: 2.16-24.02, P=0.001), respectively. Patients with either superficial or deep myoinvasion also displayed increased risks of local recurrence compared to those with no myoinvasion, with aHRs of 1.99 (95% CI: 1.59-2.50, P<0.001) and 1.54 (95% CI: 1.17-2.03, P=0.002), respectively. An increased risk of distant recurrence was observed in patients with superficial or deep myoinvasion compared to those without myoinvasion, with aHRs of 3.16 (95% CI: 3.12-3.20, P<0.001) and 2.48 (95% CI: 1.98-3.09, P<0.001), respectively. Body mass index ≥30 kg/m² at time of surgery was associated with an increased risk of local recurrence (aHR: 1.71, 95% CI: 1.56-1.88, P<0.001) (Table 3).

The maximum number of vessels involved on at least one pathology slide per case was available for the 1279 patients in the New York cohort; a detailed analysis is listed in Supplemental Table 3. The most frequent number of involved vessels identified per slide in patients with lymphovascular invasion were 1, 2, and 4 vessels (Supplemental Table 3).

Discussion

Summary of Main Results

Our group recently reported that focal lymphovascular invasion (with ≤4 vessels involved in at least one pathology slide) and no lymphovascular invasion do not share the same prognostic outcome.[14] In this follow-up retrospective cohort study, we assessed how extent of lymphovascular invasion, as defined by WHO, affects recurrence rate and location of first recurrence in FIGO 2009 surgical stage I endometrioid endometrial cancer with negative nodes. Our analysis revealed the detection of lymphovascular invasion on a slide by a gynecologic pathologist, no matter the extent, was associated with a 3-fold increase in 5-year cumulative incidence failure for recurrence compared to no detected lymphovascular invasion. Patients with substantial lymphovascular invasion (≥5 vessels involved) also had more grade 3 tumors and deeper myoinvasion, portending distant recurrence.

Results in the Context of Published Literature

Limited research has explored to what degree the extent of lymphovascular invasion detection influences recurrence patterns in surgical stage I endometrioid endometrial cancer.[15, 16] A study by Bosse et al. utilized data from the Post-Operative Radiation Therapy in Endometrial Carcinoma (PORTEC) 1 and 2 trials.[6] In their analysis, substantial lymphovascular invasion (defined as diffuse or multifocal invasion without a specific vessel cutoff) emerged as a significant prognostic factor for both local and distant recurrence, with aHRs of 6.19 (95% CI: 2.35-16.3; P<0.001) for local recurrence and 3.61 (95% CI: 1.90-6.84; P<0.001) for distant recurrence. In our analysis, substantial lymphovascular invasion was a significant prognostic factor for only distant recurrence after adjusting for other covariates, with an aHR of 2.29 (95% CI: 1.17-4.46; P=0.015). This difference in results could be attributed to the lack of uniform nodal evaluation in the patients included in the PORTEC trials, suggesting the association of lymphovascular invasion with pelvic nodal recurrence could be due to unrecognized nodal involvement in incompletely surgically staged patients.

More recently, Barnes et al. analyzed 325 patients with FIGO 2009 stage I/II endometrioid endometrial cancer using a cutoff of 2 or more vessels involved on at least one pathology slide to define substantial lymphovascular invasion. The authors found that substantial invasion was associated with both local and distant recurrence, but only 32% of patients included in the study underwent nodal evaluation.[7] A study by Tortorella et al. of patients with FIGO 2009 stage IA grade 1 or 2 endometrioid endometrial cancer, using a cutoff of 3 or more vessels to define substantial lymphovascular invasion, found an association between substantial lymphovascular invasion and distant recurrence.[9] In 2023, Pifer et al. analyzed data from 319 patients with node-negative stage I endometrioid endometrial cancer and found no association between extent of lymphovascular invasion and local or distant recurrence;[8] however, the median follow-up was only 25.8 months. In 2024, Bhatnagar and colleagues [17] analyzed 1052 patients with node-negative FIGO 2009 stage I endometrioid endometrial cancer, reporting similar distributions of focal and substantial lymphovascular invasion. They found slightly fewer tumors with focal invasion and more with substantial invasion, likely due to their use of a lower cutoff (three or more foci) for substantial invasion. Both studies found that oncologic outcomes were worse for patients with lymphovascular invasion, regardless of its extent; however, our findings suggest a difference in recurrence patterns, with substantial lymphovascular invasion associated with more distant spread. This discrepancy could be explained by our use of the WHO cutoff for substantial lymphovascular invasion, which selected for a more invasive, higher-grade cohort. When compared to patients with no lymphovascular invasion, we found that patients with focal invasion had a 3-fold increase in 5-year cumulative incidence failure rate for both local (11.5% vs 3.6%) and distant failure rates (8.0% vs 2.4%), challenging the current FIGO 2023 staging classification that combines no lymphovascular invasion and focal invasion as a single risk category.

Strengths and Weaknesses

Strengths of our study include the large sample size, multi-institutional collaboration, and extended follow-up period. We also provide an accurate vessel count per slide on a very large number of cases to demonstrate the distribution of lymphovascular invasion noted per slide in this disease. Limitations include the inherent biases of retrospective study designs and the limited number of patients with grade 3 tumors. In addition, the absence of tumor molecular information precludes further analysis of these findings within each molecular subtype. Finally, variations in adjuvant therapy practices between the two institutions, along with evolving practice guidelines over the years, while reflecting real-world scenario, may have introduced potential bias into the results, by affecting who received adjuvant treatment at various time points, the choice of adjuvant treatment received, and the potential implications this could have had on recurrence rates and patterns.

Implications for Practice and Future Research

Our study underscores other independent risk factors for local and distant recurrence, some of which are well known like histological grade and depth of myometrial invasion.[18, 19] Within the substantial lymphovascular invasion group, we noted a higher rate of tumors with deep (≥50%) myometrial invasion and FIGO grade 3 histology compared to the other 2 groups. The co-occurrence of these high-risk uterine factors likely further contributed to the elevated risk of distant recurrence seen in patients with substantial lymphovascular invasion.[15, 20]

The inclusion of lymphovascular invasion quantification in the FIGO 2023 classification should be reconsidered. In addition, deep myometrial invasion and grade 3 disease were prevalent in patients with substantial lymphovascular invasion and emerged as independent prognostic factors of distant recurrence.

Conclusions

In this study of patients with stage I endometrioid endometrial cancer and negative nodes, detection of lymphovascular invasion, no matter the extent, was associated with a 3-fold increase in 5-year cumulative incidence of disease recurrence compared to no detection of lymphovascular invasion. This finding calls into question the grouping of endometrioid endometrial cancers with no or focal lymphovascular invasion into the same risk category, as well as upstaging of patients with endometrioid endometrial cancer and substantial lymphovascular invasion, in the FIGO 2023 classification.

Key Messages.

What is already known on this topic:

Research on the impact of lymphovascular invasion extent on disease recurrence patterns in stage I endometrioid endometrial cancer is limited and inconsistent. Some findings suggest an association between substantial (and not focal) lymphovascular invasion and locoregional disease recurrence, while others suggest that any lymphovascular invasion, regardless of extent, is associated with worse oncological outcomes.

What this study adds:

Our study found that lymphovascular invasion, regardless of extent, was associated with a 3-fold increase in recurrence rates compared to no lymphovascular invasion. While patients with substantial invasion had more deeply invasive and grade 3 tumors and experienced more distant than local recurrences, patients with focal invasion showed no predominant recurrence pattern but still demonstrated a 3-fold rise in the 5-year cumulative incidence of both local and distant failure.

How this study might affect research, practice or policy:

Our findings challenge the 2023 International Federation of Obstetrics and Gynecology (FIGO) staging system, specifically the grouping of endometrioid endometrial cancers with no or focal lymphovascular invasion into the same risk category, and the upstaging of patients with substantial lymphovascular invasion.