Background
Minimally invasive radical trachelectomy has emerged as an alternative to open radical hysterectomy for patients with early-stage cervical cancer desiring future fertility. Recent data suggest worse oncologic outcomes after minimally invasive radical hysterectomy than after open radical hysterectomy in stage I cervical cancer.
Objective
We aimed to compare 4.5-year disease-free survival after open vs minimally invasive radical trachelectomy.
Study Design
This was a collaborative, international retrospective study (International Radical Trachelectomy Assessment Study) of patients treated during 2005–2017 at 18 centers in 12 countries. Eligible patients had squamous carcinoma, adenocarcinoma, or adenosquamous carcinoma; had a preoperative tumor size of ≤2 cm; and underwent open or minimally invasive (robotic or laparoscopic) radical trachelectomy with nodal assessment (pelvic lymphadenectomy and/or sentinel lymph node biopsy). The exclusion criteria included neoadjuvant chemotherapy or preoperative pelvic radiotherapy, previous lymphadenectomy or pelvic retroperitoneal surgery, pregnancy, stage IA1 disease with lymphovascular space invasion, aborted trachelectomy (conversion to radical hysterectomy), or vaginal approach. Surgical approach, indication, and adjuvant therapy regimen were at the discretion of the treating institution. A total of 715 patients were entered into the study database. However, 69 patients were excluded, leaving 646 in the analysis. Endpoints were the 4.5-year disease-free survival rate (primary), 4.5-year overall survival rate (secondary), and recurrence rate (secondary). Kaplan-Meier methods were used to estimate disease-free survival and overall survival. A post hoc weighted analysis was performed, comparing the recurrence rates between surgical approaches, with open surgery being considered as standard and minimally invasive surgery as experimental.
Results
Of 646 patients, 358 underwent open surgery, and 288 underwent minimally invasive surgery. The median (range) patient age was 32 (20–42) years for open surgery vs 31 (18–45) years for minimally invasive surgery ( P =.11). Median (range) pathologic tumor size was 15 (0–31) mm for open surgery and 12 (0.8–40) mm for minimally invasive surgery ( P =.33). The rates of pelvic nodal involvement were 5.3% (19 of 358 patients) for open surgery and 4.9% (14 of 288 patients) for minimally invasive surgery ( P =.81). Median (range) follow-up time was 5.5 (0.20–16.70) years for open surgery and 3.1 years (0.02–11.10) years for minimally invasive surgery ( P <.001). At 4.5 years, 17 of 358 patients (4.7%) with open surgery and 18 of 288 patients (6.2%) with minimally invasive surgery had recurrence ( P =.40). The 4.5-year disease-free survival rates were 94.3% (95% confidence interval, 91.6–97.0) for open surgery and 91.5% (95% confidence interval, 87.6–95.6) for minimally invasive surgery (log-rank P =.37). Post hoc propensity score analysis of recurrence risk showed no difference between surgical approaches ( P =.42). At 4.5 years, there were 6 disease-related deaths (open surgery, 3; minimally invasive surgery, 3) (log-rank P =.49). The 4.5-year overall survival rates were 99.2% (95% confidence interval, 97.6–99.7) for open surgery and 99.0% (95% confidence interval, 79.0–99.8) for minimally invasive surgery.
Conclusion
The 4.5-year disease-free survival rates did not differ between open radical trachelectomy and minimally invasive radical trachelectomy. However, recurrence rates in each group were low. Ongoing prospective studies of conservative management of early-stage cervical cancer may help guide future management.
Introduction
Radical hysterectomy with pelvic lymphadenectomy is the standard treatment for early-stage cervical cancer. However, radical trachelectomy has emerged as an alternative to radical hysterectomy in patients with early-stage disease who wish to preserve fertility. In 2011, minimally invasive radical trachelectomy became more common than open trachelectomy.
Why was this study conducted?
A randomized prospective trial has demonstrated worse disease-free survival (DFS) in patients undergoing minimally invasive radical hysterectomy. Whether there is a difference in DFS or overall survival (OS) between open and minimally invasive radical trachelectomy for patients with early-stage cervical cancer (≤2 cm) has not been established.
Key findings
There was no difference in the rates of disease recurrence, 4.5-year DFS, or OS between open radical trachelectomy and minimally invasive radical trachelectomy.
What does this add to what is known?
Based on this large multicenter retrospective study, surgical approach (minimally invasive vs open) for radical trachelectomy in patients with early-stage cervical cancer (≤2 cm) may not affect oncologic outcomes.
A randomized noninferiority trial comparing open vs minimally invasive radical hysterectomy showed that the minimally invasive approach was associated with lower rates of disease-free survival (DFS) and overall survival (OS). Subsequent studies confirmed the findings from that trial. The unanticipated results of the Laparoscopic Approach to Cervical Carcinoma (LACC) trial raised concern regarding the oncologic safety of minimally invasive radical trachelectomy. Given the limited number of patients who were candidates for radical trachelectomy and the low recurrence rate, a randomized controlled trial comparing open surgery with minimally invasive surgery (MIS) was unlikely. Therefore, we performed an international retrospective study comparing 4.5-year DFS rates in patients with preoperative early-stage cervical cancer (≤2 cm) who underwent open vs minimally invasive radical trachelectomy.
Materials and Methods
After obtaining approval from each center’s institutional review board, we obtained data from 18 centers in 12 countries on all patients with early-stage cervical cancer who underwent open or minimally invasive (robotic or laparoscopic) radical trachelectomy. Eligible patients had squamous carcinoma, adenocarcinoma, or adenosquamous carcinoma; had a preoperative tumor size of ≤2 cm (via physical examination, imaging, or pathology assessment); and underwent radical trachelectomy with nodal assessment (pelvic lymphadenectomy and/or sentinel lymph node biopsy) during 2005–2017. The exclusion criteria included neoadjuvant chemotherapy or preoperative pelvic radiotherapy, previous lymphadenectomy or pelvic retroperitoneal surgery, pregnancy, stage IA1 disease with lymphovascular space invasion, aborted trachelectomy (conversion to radical hysterectomy), or vaginal approach. Surgical approach, indication, and adjuvant therapy regimen were at the discretion of the treating institution. Written informed consent was waived because the data were deidentified. The study design was previously published.
Patient characteristics were summarized using descriptive statistics. Categorical variables were compared using the chi-square or the Fisher exact test; continuous variables were compared using the Wilcoxon rank-sum test. Preoperative tumor size was categorized as <1 cm and 1 to 2 cm using tumor size reported at conization, tumor size on imaging if conization was not performed or tumor size was not reported at conization, or tumor size on physical examination if tumor size from conization or imaging was not available. DFS was measured from the radical trachelectomy date until the date of first recurrence or death from any cause. Patients were censored at the date of the last clinic visit when they were known to be disease-free. An evaluation of DFS and recurrence rates at 4.5 years was performed as defined in the LACC trial. OS was measured from the diagnosis date until death. Patients were censored at their last contact date. Kaplan-Meier methods were used to estimate DFS and OS. Survival distributions were compared using the log-rank test. The proportion of patients with recurrence was calculated for each group (open surgery vs MIS), and associations were tested using the chi-square or the Fisher exact test. The methods of Gooley et al were used to estimate the cumulative incidence of disease recurrence (with 95% confidence interval [CI]) as a function of the surgical method with noncervical cancer death as a competing risk. The methods of Fine and Gray were used to compare the 2 groups concerning a cumulative incidence of recurrence. Exploratory analyses in subsets of patients investigated possible associations among variables using proportions and logistic regression modeling when possible and recurrence and survival endpoints using Kaplan-Meier methods. Given the low number of events, a multivariable proportional-hazards model for DFS and overall survival (OS) could not be performed. Separate Cochran-Mantel-Haenszel tests were used to test the association of surgical approach and recurrence between the groups while adjusting for variables. Except for the primary endpoint, DFS, all testing methods were 2-sided using α =0.05. A 2-sided test with α =0.10 was used for DFS. Missing data were ignored when completing the analyses, but the impact on the primary analyses was minimal because of the low number of missing data. All patients had information regarding survival endpoints. Furthermore, when examining recurrence adjusting for propensity scores, only 33 of 646 patients (5%) had missing data when calculating the propensity scores. Therefore, bias owing to casewise deletion should be minimal.
Previous studies have shown that recurrence rates in the open surgery group range from 3.8% to 7.6%. If the 4.5-year disease-free survival rate for patients who underwent open surgery was 92.4%, we had 80% power to detect a 0.53 hazard ratio using a 2-sided test with α =0.10. This corresponds to an 86.1% DFS rate at 4.5 years in the MIS group. An estimated 845 patients would need to be included in this study: 456 open surgery and 389 MIS. This power calculation was performed with PASS 13 (Power Analysis and Sample Size Software [2014]; NCSS, LLC, Kaysville, UT; ncss.com/software/pass).
A post hoc weighted analysis was performed to compare the recurrence rates between the surgical approaches. Propensity score methods were used to assign weights to each observation with adjustment for body mass index, International Federation of Gynecology and Obstetrics (FIGO)-defined stage, tumor size, preoperative cone biopsy, and preoperative histologic type. Open surgery was considered standard, and MIS was considered experimental. The risk of recurrence and 95% CI were estimated. This report summarized the data collected through July 12, 2020. Statistical significance was defined as P <.05. Statistical analyses were completed using SAS (version 9.4; SAS Institute Inc, Cary, NC) and R (version 3.6.1; R Core Team [2019]; R Foundation for Statistical Computing, Vienna Austria).
Results
The study included 646 patients, 358 with open surgery and 288 with MIS (121 laparoscopic; 167 robotic) ( Figure 1 ). The median age was 32 years in the open surgery group and 31 years in the MIS group. Moreover, 549 of 646 patients (85.0%) had FIGO 2009 stage IB1 disease ( Table 1 ). Residual disease was present in the final specimen in 204 patients (57%) with open surgery and 118 patients (41%) with MIS ( P <.001) ( Table 2 ).
| Characteristic | Number of patients | P value | |
|---|---|---|---|
| Open surgery (n=358) | Minimally invasive surgery (n=288) | ||
| Age (y) | 32.0 (20.0–42.0) | 31.0 (18.0–45.0) | .11 |
| BMI (kg/m 2 ) | 21.8 (16.0–36.6) | 23.5 (16.1–48.4) | <.001 |
| FIGO 2009 stage | .54 | ||
| IA2 | 51 (14.2) | 46 (16.0) | |
| IB1 | 307 (85.8) | 242 (84.0) | |
| Preoperative tumor size (cm) | .54 | ||
| <1 | 137 (38.3) | 117 (40.6) | |
| 1–2 | 221 (61.7) | 171 (59.4) | |
| Cone biopsy | 217 (60.6) | 229 (79.5) | <.001 |
| Preoperative histologic type | .01 | ||
| Squamous carcinoma | 234 (65.4) | 168 (58.3) | |
| Adenocarcinoma | 108 (30.2) | 115 (39.9) | |
| Adenosquamous carcinoma | 16 (4.5) | 5 (1.7) | |
| Preoperative grade | .02 | ||
| I | 36 (10.1) | 32 (11.1) | |
| II | 80 (22.3) | 99 (34.4) | |
| III | 58 (16.2) | 35 (12.2) | |
| Unknown or not reported | 184 (51.4) | 122 (42.4) | |
| Surgical time (min) | 171 (50–425) | 262 (120–472) | <.001 |
| Estimated blood loss (mL) | 200 (50–4500) | 50 (0–3000) | <.001 |
| Nodal assessment | <.001 | ||
| Pelvic lymphadenectomy only | 267 (74.6) | 163 (56.6) | |
| SLN only | 17 (4.7) | 28 (9.7) | |
| SLN and pelvic lymphadenectomy | 74 (20.7) | 97 (33.7) | |
| Number of nodes removed | 17 (2–63) | 18 (2–65) | .79 |
| Length of stay (d) | 6 (1–23) | 2 (0–24) | <.001 |
| Readmission | 7 (2.0) | 28 (9.7) | <.001 |
| Reoperation | 6 (1.7) | 10 (3.5) | .14 |
| Characteristic | Number of patients | P value | |
|---|---|---|---|
| Open surgery (n=358) | Minimally invasive surgery (n=288) | ||
| Histologic type | <.001 | ||
| No residual disease a | 150 (42.0) | 165 (57.3) | |
| Squamous carcinoma | 138 (38.5) | 61 (21.2) | |
| Adenocarcinoma | 58 (16.2) | 53 (18.4) | |
| Adenosquamous carcinoma | 8 (2.2) | 4 (1.4) | |
| Not reported | 4 (1.1) | 5 (1.7) | |
| Tumor size (mm) b | 15 (10.00–18.00) | 12 (0.89–17.00) | .33 |
| Tumor size (cm) b | .02 | ||
| <1 | 55 (27.00) | 42 (35.6) | |
| 1–2 | 124 (60.8) | 49 (41.5) | |
| Not reported | 25 (12.2) | 27 (22.9) | |
| Grade b | <.001 | ||
| I | 31 (15.2) | 14 (11.8) | |
| II | 58 (28.4) | 54 (45.8) | |
| III | 65 (31.9) | 19 (16.1) | |
| Unknown or not reported | 50 (24.5) | 31 (26.3) | |
| LVSI b | .40 | ||
| Yes | 43 (21.1) | 28 (23.7) | |
| No | 144 (70.6) | 74 (62.7) | |
| Not reported | 17 (8.3) | 16 (13.6) | |
| Depth of invasion (mm) b | .06 | ||
| <10 | 85 (41.7) | 91 (77.1) | |
| ≥10 | 16 (7.8) | 7 (5.9) | |
| Not reported or unknown | 103 (50.5) | 20 (17.0) | |
| Pelvic nodal involvement | 19 (5.3) | 14 (4.9) | .81 |
| Adjuvant treatment c | <.001 | ||
| Yes | 46 (12.9) | 15 (5.2) | |
| No | 310 (86.6) | 273 (94.8) | |
| Not reported or unknown | 2 (0.6) | 0 (0.0) | |
| Follow-up time (y) | 5.5 (3.40–8.30) | 3.1 (2.0–4.80) | <.001 |
| Recurrence | 17 (4.7) | 18 (6.3) | .40 |
| Time to recurrence (range) d | 0.3–3.0 | 0.3–3.6 | .34 |
| Location of recurrence | .27 | ||
| Local | 15 | 13 | |
| Regional | 1 | 0 | |
| Distant | 0 | 1 | |
| Local and regional | 1 | 4 | |
a No residual disease after conization or initial biopsy
b Only patients with residual disease in the final pathology (204 with open surgery and 118 with minimally invasive surgery)
c Radiation therapy, chemotherapy, chemoradiation, and/or brachytherapy
d For current number of events, the median time to recurrence could not be estimated.
Median follow-up time was 5.5 (0.20–16.70) years for open surgery and 3.1 (0.02–11.10) years for MIS ( P <.001). Rates of adjuvant treatment in the open surgery and MIS groups were 13% and 5%, respectively ( P <.001). The odds of adjuvant treatment did not differ by tumor size (Breslow-Day, P =.88). At 4.5 years, 4.7% of patients with open surgery and 6.2% with MIS had recurrence ( P =.40). ( Supplemental Table 1 ). DFS did not differ between surgical approaches (log-rank P =.31). Recurrence rates did not differ between robotic surgery (9 of 167 patients [5.4%]) and laparoscopy (9 of 121 patients [7.4%]) ( P =.48). The 4.5-year DFS rates were 94.3% (95% CI, 91.6–97.0) for open surgery and 91.5% (95% CI, 87.6–95.6) for MIS (log-rank P =.37) ( Figure 2 , A).
