The MEMORY Study: MulticentEr study of Minimally invasive surgery versus Open Radical hYsterectomy in the management of early-stage cervical cancer: survival outcomes

Abstract

Objective:

The Laparoscopic Approach to Cervical Cancer (LACC) trial found that minimally invasive radical hysterectomy compared to open radical hysterectomy compromised oncologic outcomes and was associated with worse progression-free survival (PFS) and overall survival (OS) in early-stage cervical carcinoma. We sought to assess oncologic outcomes at multiple centers between minimally invasive (MIS) radical hysterectomy and OPEN radical hysterectomy.

Methods:

This is a multi-institutional, retrospective cohort study of patients with 2009 FIGO stage IA1 (with lymphovascular space invasion) to IB1 cervical carcinoma from 1/2007-12/2016. Patients who underwent preoperative therapy were excluded. Squamous cell carcinoma, adenocarcinoma, and adenosquamous carcinomas were included. Appropriate statistical tests were used.

Results:

We identified 1093 cases for analysis—715 MIS (558 robotic [78%]) and 378 OPEN procedures. The OPEN cohort had more patients with tumors >2 cm, residual disease in the hysterectomy specimen, and more likely to have had adjuvant therapy. Median follow-up for the MIS and OPEN cohorts were 38.5 months (range, 0.03-149.51) and 54.98 months (range, 0.03-145.20), respectively. Three-year PFS rates were 87.9% (95% CI: 84.9-90.4%) and 89% (95% CI: 84.9-92%), respectively (P=0.6). On multivariate analysis, the adjusted HR for recurrence/death was 0.70 (95% CI: 0.47-1.03; P=0.07). Three-year OS rates were 95.8% (95% CI: 93.6-97.2%) and 96.6% (95% CI: 93.8-98.2%), respectively (P=0.8). On multivariate analysis, the adjusted HR for death was 0.81 (95% CI: 0.43-1.52; P=0.5).

Conclusion:

This multi-institutional analysis showed that an MIS compared to OPEN radical hysterectomy for cervical cancer did not appear to compromise oncologic outcomes, with similar PFS and OS.

INTRODUCTION

Radical hysterectomy via exploratory laparotomy (“OPEN” surgery) has been the standard surgical management approach for early-stage cervical cancer for more than 100 years.^1-3 In 1992, Nezhat and colleagues published the first report of laparoscopic total radical hysterectomy.⁴ Multiple retrospective series since then have suggested that a laparoscopic, or minimally invasive surgical (MIS), approach with or without the use of a computer-enhanced (robotic) surgical platform is associated with improved perioperative outcomes, with similar oncologic outcomes compared to those of an OPEN approach.^5-20

In 2018, results from the first, and only, randomized controlled trial (RCT) to compare these approaches were published by Ramirez and colleagues.²¹ The results of the Laparoscopic Approach to Cervical Cancer (LACC) trial, a planned non-inferiority trial, demonstrated that an MIS approach was not non-inferior to an OPEN approach, with worse 3-year disease-free survival (DFS) and overall survival (OS).²¹ Subsequent reports of secondary outcomes from the LACC trial noted that perioperative adverse events as well as quality of life (QOL) also were not better with MIS.^22,23 Multiple retrospective single-institution, multicenter, and national database series since then have reported similar findings in terms of DFS, and some but not all in terms of OS.^24-33 Findings from other retrospective studies, however, have not shown compromised oncologic outcomes with MIS over OPEN surgery.^34-41

Although the results of the LACC trial have been met with some opposition, the National Comprehensive Cancer Network (NCCN) and many other societies have amended their guidelines in accordance with the study’s findings, stating that the preferred and standard approach for radical hysterectomy should be laparotomy ⁴² Regardless of the LACC trial’s findings, there are still many unanswered questions with regard to the role of MIS in relation to tumor size, stage, cone biopsy only lesions, tumor containment technique improvements, individual surgeon outcomes, and preoperative assessments, among others. In this study, we sought to evaluate the oncologic outcomes of MIS versus OPEN radical hysterectomy for early-stage cervical cancer among a group of skilled MIS gynecologic oncologists in the US and Canada.

MATERIALS AND METHODS

This is a retrospective, multicenter cohort analysis of all consecutive and concurrent patients who presented with 2009 International Federation of Gynecology and Obstetrics (FIGO) stage IA1 (with lymphovascular space invasion [(LVSI]) to IB1 cervical carcinoma between 1/2007 and 12/2016 at one of 15 centers in the United States and Canada. All surgeons were fellowship-trained gynecologic oncologists with experience in MIS approaches, both with and without a robotic platform, in the management of appropriate patients diagnosed with gynecologic malignancies. We included cases with squamous cell carcinoma, adenocarcinoma, and adenosquamous carcinoma. Neuroendocrine tumors and gastric-type adenocarcinomas were excluded. All patients had undergone a radical hysterectomy, as described per surgeon, as their initial therapy. We excluded patients who had undergone any form of preoperative radiation therapy and/or chemotherapy, a vaginal approach, or fertility-sparing procedure. MIS cases converted to an OPEN procedure were analyzed within the MIS cohort. Patients who underwent adjuvant postoperative therapies based on established pathologic risk criteria and NCCN guidelines were included.³

Institutional Review Board approvals were obtained at each institution, and data use agreements were implemented. Multiple clinicopathologic and survival data were collated from each institution. Intraoperative complications were those that occurred during surgery, and postoperative complications were those reported ≥30 days after surgery. Pathologic tumor size was defined as the greatest diameter of the cervical lesion measured by the pathologist either micro- or macroscopically at the time of radical hysterectomy. Patients had undergone surgical nodal assessment with either comprehensive lymphadenectomy and/or sentinel lymph node mapping per institutional protocols. Disease recurrence was determined clinically, radiologically, and/or histologically per standard institutional protocols.

Association of clinical factors with surgery approach were tested with the Wilcoxon rank sum test or Fisher exact test as appropriate. Progression-free survival (PFS) was defined from date of surgery to disease progression post-surgery or death or last follow-up, whichever came first. Both progression and death were considered as an event. Overall survival (OS) was defined from date of surgery to death from any cause or last follow-up. Median survival and survival rate were estimated using the Kaplan-Meier method. The hazard ratios and P values for PFS and OS were obtained using the Cox proportional hazards (CoxPH) model with Wei, Lin, Weissfeld (WLW)⁴³ estimate while considering different centers’ clustering effect. Landmark analysis was applied in order to assess the effect of adjuvant treatment on OS/PFS; associated tests and models used landmark time as 6 weeks. Significant variables (P<0.05) in the univariate analysis as well as the surgical approach (MIS/OPEN) were used to create a multivariate model.

RESULTS

Of 1093 patients from the 15 participating centers, 715 (65%) were performed via an MIS approach and 378 (35%) via an OPEN approach. A robotic platform was used in 558 (78%) of the MIS patients. Conversion from an MIS to OPEN procedure occurred in 16 (2.2%). There was imbalance between the cohorts in terms of pathologic tumor size, presence of residual tumor in the hysterectomy specimen, FIGO stage, node positivity, presence of LVSI, and use of adjuvant therapy (Table 1). Median estimated blood loss (EBL) was 100 mL (range, 0-2500 mL) in the MIS group compared to 300 mL (range, 50-3000 mL) in the OPEN group (P<0.001). Median length of stay (LOS) was 1 day (range, 0-29 days) and 4 days (range, 1-17 days), respectively (P<0.001). Median operative times were 222 min (range, 65-562 min) and 205 min (range, 69-381 min), respectively (P=0.05). Overall complication rates (none, intra-, post-operative, or both) were similar (P= 0.39).

Table 1.

Baseline clinicopathologic characteristics and perioperative outcomes of the MIS and OPEN cohorts

Characteristic	Overall (N=1093)	MIS (N=715)	OPEN (N=378)	P value (MIS v OPEN)
Age, years	44 (22-87)	44 (22-87)	43 (22-5)	0.1
Missing data (n)	1	1	-
BMI, kg/m2	27 (16-58.6)	27 (17-58.6)	26.7 (16.6-51.9)	0.8
Missing data (n)	52	21	31
2009 FIGO stage				<0.001
IA1 (LVSI)/IA2	187 (17)	147 (21)	40 (11)
IB1	906 (83)	568 (79)	338 (89)
Histology				0.4
Squamous cell carcinoma	567 (52)	360 (51)	207 (55)
Adenocarcinoma	473 (43)	318 (45)	155 (41)
Adenosquamous carcinoma	50 (4.6)	34 (4.8)	16 (4.2)
Missing data (n)	3	3	-
Pathologic tumor size				<0.001
<2 cm	592 (56)	416 (61)	176 (49)
2-4 cm	354 (34)	218 (32)	136 (38)
≥4 cm	101 (9.6)	51 (7.4)	50 (14)
Missing data (n)	46	30	16
Residual tumor in hysterectomy	608 (62)	360 (56)	248 (72)	<0.001
Missing data (n)	108	76	32
Presence of nodal metastasis	119 (12)	63 (9.7)	56 (16)	0.008
Missing data (n)	84	67	17
Positive margins	44 (4.4)	28 (4.3)	16 (4.4)	>0.99
Missing data (n)	83	68	15
Presence of LVSI				0.007
Yes	352 (35)	209 (31)	143 (41)
Suspicious	28 (2.8)	19 (2.9)	9 (2.6)
Missing data (n)	82	50	32
Adjuvant therapy given	329 (30)	182 (26)	147 (39)	<0.001
Missing data	4	2	2
MIS approach				-
Robotic assisted		558 (78)	-
Non-robotic assisted		141 (20)
Converted to laparotomy		16 (2.2)
Operative time, min	221 (65-562)	222 (65-562)	205 (69-381)	0.05
Missing data (n)	573	336	237
Estimated blood loss, mL	150 (0-3000)	100 (0-2500)	300 (50-3000)	<0.001
Missing data (n)	53	30	23
Transfused	49 (4.5)	12 (1.7)	37 (9.8)	<0.001
Missing data (n)	2	2	-
LOS, days	2 (0-2()	1 (0-29)	4 (1-17)	<0.001
Missing data (n)	132	55	77
Complications				0.4
None	893 (82)	589 (83)	304 (81)
Intraoperative	32 (2.9)	24 (3.4)	8(2.1)
Postoperative	157 (14)	96 (13)	61 (16)
Both	7 (0.6)	4 (0.6)	3 (0.8)
Missing data (n)	4	2	2

MIS, minimally invasive surgical radical hysterectomy; OPEN, open radical hysterectomy; BMI, body mass index; FIGO, International Federation of Gynecology and Obstetrics; LVSI, lymphovascular space invasion; LOS, length of stay

Continuous variables are presented as medians (range). Categorical variables are presented as n (%) and tested with the Wilcoxon rank sum test. For yes/no variables, only “YES” counts shown.

Among the 1093 patients, 18 had either a missing surgery date (n=16) or information on recurrence (n=2). This left 1075 patients for the PFS analyses (Table 2). Progression of disease occurred in 120 patients, and 14 patients died prior to documented progression. The median follow-up for the 941 progression-free survivors was 42.8 months (range, 0-149.5 months). Figure 1 depicts the Kaplan-Meier estimates for PFS between the MIS and OPEN cohorts. The 3-year PFS rate for MIS was 87.9% (95% CI: 84.9-90.4%) compared to 89% (95%CI: 84.9-92%) for OPEN surgery (HR, 0.92; 95% CI: 0.66-1.27; P=0.6). On multivariate analysis, surgical approach was not associated with PFS (HR, 0.70; 95% CI: 0.47-1.03; P=0.07). FIGO stage (IA/IB), margin positivity, and pathologic tumor size were independently associated with PFS.

Table 2.

Univariate and multivariate analysis of progression-free survival

Variable	N	Event#	3-yr PFS rate (95% CI)	Univariate HR (95% CI)	P value1	Multivariate HR (95% CI)	P value
All	1075	134	88.3 (85.9-90.3)	-	-	-	-
Age, years
10-year increments			-	1.13 (0.92-1.4)	0.2	-	-
BMI, kg/m2
5-point increments			-	1.02 (0.85-1.22)	0.8	-	-
Surgical approach
MIS	698	85	87.9 (84.9-90.4)	ref	0.6	ref	0.073
OPEN	377	49	89 (84.9-92)	0.92 (0.66-1.27)		0.70 (0.47-1.03)
2009 FIGO stage
IA1 (LVSI)/IA2	183	4	98 (93.9-99.4)	ref	<0.001	ref	0.036
IB1	892	130	86.4 (83.6-88.7)	6.91 (2.67-17.84)		3.77 (1.09-13.0)
Histology
Squamous cell carcinoma	557	67	87.4 (83.8-90.2)	ref		ref
Adenocarcinoma	465	52	90.9 (87.5-93.4)	0.81 (0.58-1.13)	0.004	1.04 (0.71-1.53)	0.85
Adenosquamous carcinoma	50	13	74.9 (59.2-85.3)	2.07 (1.08-3.97)		2.22 (0.96-5.12)	0.063
Pathologic tumor size
<2 cm	579	34	94.4 (91.7-96.2)	ref	<0.001	ref
2-4 cm	350	65	82.6 (77.6-86.5)	3.24 (2.27-4.65)		2.55 (1.66-3.9)	<0.001
≥4 cm	101	31	73 (62.7-81)	5.67 (3.71-8.67)		3.75 (1.98-7.1)	<0.001
Residual tumor in hysterectomy
No	372	25	93.5 (89.8-95.9)	ref	0.054	-	-
Yes	595	99	84.8 (81.4-87.7)	2.33 (0.98-5.5)		-	-
Presence of nodal metastasis
No	872	91	90.1 (87.6-92.2)	ref	<0.001	ref	0.3
Yes	119	36	73.2 (63.5-80.7)	2.95 (1.61-5.42)		1.68 (0.65-4.37)
Positive margins
No	948	105	88.9 (86.5-91)	ref	<0.001	ref	0.3
Yes	44	21	69.7 (53.5-81.1)	4.3 (3.01-6.13)		2.07 (1.08-3.99)
Presence of LVSI
No	619	57	91.4 (88.6-93.6)	ref	<0.001	ref	0.06
Yes/Suspicious	375	70	82.4 (77.6-86.3)	2.06 (1.55-2.74)		1.32 (0.99-1.77)
Adjuvant therapy given 2
No	706	63	91.6 (89-93.6)	ref	<0.001	ref	0.6
yes	322	69	81.6 (76.5-85.7)	2.43 (1.86-3.18)		0.85 (0.47-1.53)

PFS, progression-free survival; BMI, body mass index; MIS, minimally invasive surgical radical hysterectomy; OPEN, open radical hysterectomy; FIGO, International Federation of Gynecology and Obstetrics; LVSI, lymphovascular space invasion

¹Obtained by applying Cox proportional hazards model considering the cluster effect of each center using the Wei, Lin, Weissfeld (WLW) estimate

²Time-dependent variable analyzed using landmark analyses with landmark time=6 weeks; 44 patients who were lost to follow-up or progressed before 6 weeks were removed

Figure 1.

Progression-free survival (PFS) for the minimally invasive surgery (MIS) and OPEN cohorts

Among the 1093 patients, 19 had either a missing surgery date (n=16) or vital status at last follow-up (n=3). This left 1074 patients for OS analyses (Table 3). Sixty-two patients among the 1074 analyzed died. The median follow-up time for the 1012 survivors was 43.1 months (range, 0-149.5 months). The median follow-up time was 38.5 months (range, 0.03-149.5 months) in the MIS cohort and 54.98 months (range, 0.03-145.2 months) in the OPEN cohort. Figure 2 depicts the Kaplan-Meier estimates for OS between the MIS and OPEN cohorts. The 3-year OS rate for MIS was 95.8% (95% CI: 93.6-97.2%) compared to 96.6% (95%CI: 93.8-98.2%) for OPEN surgery (HR, 1.05; 95% CI: 0.66-1.67; P=0.8). On multivariate analysis, surgical approach was not associated with OS (HR, 0.81; 95% CI: 0.43-1.52; P=0.5). Margin positivity and pathologic tumor size were independently associated with OS.

Table 3.

Univariate and multivariate analysis of overall survival

Variable	N	Event#	3-yr OS rate (95% CI)	Univariate HR (95% CI)	P value1	Multivariate HR (95% CI)	P value
All	1074	62	96.1 (94.5-97.2)	-	-	-	-
Age, years
10-year increments			-	1.29 (0.96-1.74)	0.09	-	-
BMI, kg/m2
5-point increments			-	1.08 (0.85-1.37)	0.5	-	-
Surgical approach
MIS	696	36	95.8 (93.6-97.2)	ref	0.8	ref	0.2
OPEN	378	26	96.6 (93.8-98.2)	1.05 (0.66-1.67)		0.81 (0.43-1.52)
2009 FIGO stage
IA1 (LVSI)/IA2	183	2	98.6 (94.6-99.7)	ref	0.01	ref	0.2
IB1	891	60	95.6 (93.7-96.9)	6.15 (1.47-25.74)		4.83 (0.56-41.6)
Histology
Squamous cell carcinoma	558	36	94.6 (91.8-96.4)	ref	0.02	ref
Adenocarcinoma	464	22	98.1 (96-99.1)	0.63 (0.42-0.95)		0.71 (0.35-1.43)	0.33
Adenosquamous carcinoma	49	4	92.1 (77-97.4)	1.0 (0.35-2.88)		1.02 (0.34-3.09)	0.98
Pathologic tumor size
<2 cm	577	15	98.5 (96.8-99.3)	ref	<0.001	ref
2-4 cm	351	29	95.2 (91.8-97.2)	3.07 (1.72-5.49)	2.27 (1.38-3.72)		0.001
≥4 cm	101	16	87.1 (77.8-92.6)	6.04 (3.46-10.54)		2.71 (1.57-4.65)	<0.001
Residual tumor in hysterectomy
No	371	14	98.7 (96.6-99.5)	ref	0.5	-	-
Yes	595	41	95 (92.6-96.7)	1.57 (0.47-5.28)
Presence of nodal metastasis
No	871	41	97 (95.4-98.1)	ref	0.04	ref	0.3
Yes	119	16	90.7 (82.8-95.1)	2.69 (1.06-6.88)		1.74 (0.63-4.82)
Positive margins
No	947	44	97 (95.5-98.1)	ref	<0.001	ref	0.002
Yes	44	13	82.3 (66.4-91.2)	6.23 (3.85-10.09)		3.43 (1.59-7.40)
Presence of LVSI
No	620	24	97.3 (95.2-98.4)	ref	0.01	ref	0.2
Yes/Suspicious	373	33	93.9 (90.5-96.1)	2.2 (1.19-4.08)		1.63 (0.75-3.57)
Adjuvant therapy given 2
No	706	30	97.2 (95.4-98.3)	ref	<0.001	ref	0.3
Yes	323	32	92.6 (88.6-95.2)	2.22 (1.45-3.42)		0.67 (0.31-1.41)

OS, overall survival; BMI, body mass index; MIS, minimally invasive surgical radical hysterectomy; OPEN, open radical hysterectomy; FIGO, International Federation of Gynecology and Obstetrics; LVSI, lymphovascular space invasion

¹Obtained by applying Cox proportional hazards model considering the cluster effect of each center using the Wei, Lin, Weissfeld (WLW) estimate

²Time-dependent variable analyzed using landmark analyses with landmark time=6 weeks; 44 patients who were lost to follow-up or progressed before 6 weeks were removed

Figure 2.

Overall survival (OS) for the minimally invasive surgery (MIS) and OPEN cohorts

DISCUSSION

The findings of our retrospective multi-institutional analysis demonstrated that an MIS compared to OPEN radical hysterectomy for the management of patients with stage IA1 to IB1 cervical carcinoma did not appear to be associated with a compromise in oncologic outcomes, showing similar PFS and OS between the two approaches. We recognize the retrospective nature of this analysis, as well as the higher level of evidence with an RCT. The median follow-up time in the MIS group was also shorter than the OPEN group (38.5 months vs. 54.98 months, respectively). This is partially related to the inherent nature of a retrospective study, as more patients choose to pursue an MIS approach in recent years. Nonetheless, the range of the follow-up times are close between these two groups (0.03-149 months vs. 0.03-145 months, respectively), and although this analysis did not account for potential time trends, future follow-up studies could focus on patients with longer follow-up.

After the success of the first reported laparoscopic (MIS) radical hysterectomy in 1992 and the multiple subsequent retrospective series that have confirmed similar oncologic outcomes with improved perioperative outcomes,^5-20 enthusiasm for MIS radical hysterectomy was further strengthened by RCTs that showed non-inferior oncologic outcomes with improved perioperative outcomes and QOL in patients with endometrial cancer.^44,45 Given these positive findings, MIS radical hysterectomy was accepted as a viable approach by many guidelines and societies.

The 2018 published findings of the LACC trial, on the other hand, demonstrated that MIS compared to an open approach was not non-inferior in terms of DFS, did not result in lower adverse events, and did not improve QOL;^21-23 however, the results may be considered inconclusive since the 95% CI includes the pre-specified non-inferiority boundary.⁴⁶ Furthermore, secondary endpoint analyses suggested that an MIS approach may be associated with higher rates of local failure and reduced OS.²¹ As a result, the NCCN and many other societies changed their guidelines to reflect that the preferred and standard approach for radical hysterectomy should be laparotomy.⁴²

There are, however, reasonable debates over the findings of the LACC trial.⁴⁷ It is important to note the LACC trial is susceptible to a natural error rate, as are other RCTs, which is related to the power of any RCT leading to accepting a null hypothesis incorrectly. The open arm of the LACC trial had the best ever reported outcomes for such a cohort. Preoperative magnetic resonance imaging was not required, which may have affected the inclusion of tumors larger than 4 cm. Possibly the greatest potential contributing factor to the findings of the LACC trial was the lack of adherence to the basic oncologic principle of tumor containment and non-fragmentation in women undergoing an MIS approach. Furthermore, while the LACC trial is truly an internally valid trial, the external validity in applying these results to each and every surgeon and to all patients with any stage of cervical cancer must be considered carefully.

The 4.5-year DFS rate in the open arm of the LACC trial was 96.5%, compared to 86% in the MIS arm.²¹ This outcome in the open arm was much higher than the pre-specified expected outcome of 90% during trial design.²¹ It is unclear why the open arm performed much better than expected, and it is plausible this could have been a chance occurrence within a cohort sampled from a larger population. The many retrospective studies published since the LACC trial, both supportive and not, have nearly all reported 4.5-5–year DFS/PFS rates of approximately 90%.^{28,32,33,35,37,39,41} The 3-year PFS rate for our OPEN cohort was 89%, which is similar to that of these other trials and the upfront expected outcome for the OPEN arm when the LACC trial was designed.

There are many retrospective series, which we have not referenced specifically here, with similar findings to those of the LACC trial and are often considered “confirmatory”. However, there have also been large population-based studies that have not confirmed the LACC trial findings. An analysis of the Netherlands Cancer Registry, which captures all patients with cervical cancer undergoing radical hysterectomy in their country, did not show worse outcomes with an MIS compared to open procedure.³⁷ The 5-year PFS rate for the open cohort of the study was 89%, compared to 90% for the MIS cohort.³⁷ An analysis of the Swedish Quality Register of Gynecologic Cancer also noted similar oncologic outcomes for MIS and open approaches, with a 5-year DFS rate of only 85% for the open cohort.³⁵ Our current analysis, with a relatively large cohort of cases from multiple centers of expert MIS surgeons/centers, is in line with these other large national analyses from the Netherlands and Sweden, where patients are centralized to highly expert institutions.

One of the most likely explanations, if not a statistical chance error, for the findings of the LACC trial as well as other studies, is that many who had adopted an MIS approach for radical hysterectomy may have overlooked the importance of the basic surgical oncologic principle of tumor containment and non-fragmentation. The value of tumor containment with avoidance of tumor exposure intraperitoneally has been reported by multiple authors and makes sense when one considers that in an open approach, the vagina is clamped prior to colpotomy.^32,48 In our series, nearly all patients in the MIS cohort underwent an intraperitoneal colpotomy, so we cannot specifically address this factor; and it is unlikely that a future RCT will specifically randomize patients to tumor containment versus no containment.

One of the strengths of our study is that it encompasses all patients who were seen at multiple centers and not merely a sampling as is seen in an RCT. This study is obviously limited by its retrospective nature, which is associated with potential biases of selection and information ascertainment. This, however, also applies to the many retrospective studies that have confirmed the findings of the LACC trial in high-impact journals. As with any retrospective analysis, the adjudication of recurrence and timing of recurrence is also prone to the surveillance patterns of each institution, which varied. We also cannot comment on the use of manipulators or method of colpotomy, as the vast majority of the cases were performed with a manipulator and an intracorporeal colpotomy. We also did not have information on how many cases had undergone a prior cone biopsy.

In summary, we did not find obvious worse oncologic outcomes in patients who had undergone an MIS radical hysterectomy by an experienced gynecologic oncologist. We recognize that the findings of a retrospective analysis are not superior to those of an RCT. However, RCTs are also prone to error, and external validity is always a significant issue, especially with surgical interventions. Our data are not intended to represent a superior analysis or finding compared to an RCT. However, we feel that our real-world study and findings, along with those of other series, prompt some doubt about the findings of the LACC trial, highlight the variability in outcomes between cohorts of surgeons, and further strengthen the equipoise needed to support the currently enrolling robot-assisted approach to cervical cancer (RACC) trial⁴⁹ and GOG 3043, a randomized controlled trial of robotic versus open radical hysterectomy for cervical cancer (ROCC trial [(ClinicalTrials.gov identifier: NCT04831580]).

Highlights.

• Minimally invasive (MIS) compared to open radical hysterectomy did not compromise outcomes in early-stage cervical cancer

• Our findings are contrary to those of the Laparoscopic Approach to Cervical Cancer (LACC) trial favoring laparotomy

• As the NCCN has changed guidelines in accordance with LACC, women may miss out on the benefits of MIS radical hysterectomy