Article Text
PDF
Abstract
Objective Patients with intermediate-risk cervical cancer receive external beam radiotherapy (EBRT) as adjuvant treatment. It is commonly administered with brachytherapy without proven benefits. Therefore, we evaluated the frequency of brachytherapy use, the doses for EBRT administered alone or with brachytherapy, and the overall survival impact of brachytherapy in patients with intermediate-risk, early-stage cervical cancer.
Methods This retrospective cohort study was performed using data collected from the National Cancer Database. Patients diagnosed with cervical cancer from 2004 to 2019 who underwent a radical hysterectomy and lymph node staging and had disease limited to the cervix but with tumors larger than 4 cm or ranging from 2 to 4 cm with lymphovascular space invasion (LVSI) were included. Patients with distant metastasis or parametrial involvement were excluded. Patients who underwent EBRT alone were compared with those who also received brachytherapy after 2:1 propensity score matching.
Results In total, 1174 patients met the inclusion criteria, and 26.7% of them received brachytherapy. After 2:1 propensity score matching, we included 620 patients in the EBRT group and 312 in the combination treatment group. Patients who received brachytherapy had higher equivalent doses than those only receiving EBRT. Overall survival did not differ between the two groups (hazard ratio (HR) 0.88 (95% confidence interval (CI), 0.62 to 1.23]; p=0.45). After stratification according to tumor histology, LVSI, and surgical approach, brachytherapy was not associated with improved overall survival. However, in patients who did not receive concomitant chemotherapy, the overall survival rate for those receiving EBRT and brachytherapy was significantly higher than that for those receiving EBRT alone (HR, 0.48 (95% CI, 0.27 to 0.86]; p=0.011).
Conclusion About one-fourth of the study patients received brachytherapy and EBRT. The variability in the doses and radiotherapy techniques used highlights treatment heterogeneity. Overall survival did not differ for EBRT with and without brachytherapy. However, overall survival was longer for patients who received brachytherapy but did not receive concomitant chemotherapy.
- Brachytherapy
- Radiation Oncology
- Surgical Oncology
Data availability statement
Data are available upon reasonable request.
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
Adjuvant external beam radiotherapy (EBRT) reduces recurrence rates in patients with intermediate-risk cervical cancer. However, the common use of combined brachytherapy and EBRT in clinical practice contrasts with a lack of consensus in guidelines regarding their use.
WHAT THIS STUDY ADDS
The study reflects the heterogeneity in radiation doses and adjuvant radiotherapy techniques administered to patients with intermediate-risk cervical cancer. Combining brachytherapy with EBRT improved overall survival in patients not receiving concomitant chemotherapy. However, this subanalysis was limited by sample size. We observed no impact of varying radiation doses and adjuvant radiotherapy techniques on overall survival in the general study population or in the other subgroup of patients.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
The indications and dosage for brachytherapy for intermediate-risk cervical cancer are not well defined as seen in this study’s broad population heterogeneity, highlighting the need for clearer guidelines for brachytherapy use. Given that combining brachytherapy with EBRT may benefit survival outcomes for patients with intermediate-risk cervical cancer who do not receive concomitant chemotherapy, this intervention requires further investigation.
Introduction
The standard management for patients with early-stage cervical cancer (based on 2018 International Federation of Gynecology and Obstetrics (FIGO) stage IA1 or IIA1) is surgical, consisting of a hysterectomy and lymph node assessment.1 2 Following a definitive pathology report, the risk factors for recurrence are assessed, and most guidelines recommend that patients with intermediate risk factors (a combination of lymphovascular space invasion (LVSI), stromal invasion, and tumor size) receive adjuvant external beam radiotherapy (EBRT).1–3 This recommendation is based on the results of the two available prospective randomized trials assessing this intervention, which were published more than 20 years ago.3 4 In particular, in the trial conducted by Sedlis et al.,3 patients with intermediate risk factors were randomized to two arms: 1) EBRT with no brachytherapy use with doses ranging from 46.0 to 50.4 Gy; and 2) observation. They reported a reduction in recurrence risk in the treatment arm (relative risk, 0.53; p=0.008). Nevertheless, a long-term follow-up study demonstrated no improvement in overall survival in the radiotherapy group.5 Since that time, it was proposed that higher pelvic radiation doses could further reduce the pelvic local failure rate,3 but unfortunately, this has yet to be assessed in prospective trials.
After the publication of the results obtained by Sedlis et al.,3 a debate emerged regarding the value of using adjuvant radiotherapy in this context, which included discussion about using additional brachytherapy and/or chemotherapy.6 Some guidelines7 recommend the use of a brachytherapy boost with EBRT, particularly for patients with extensive LVSI or deeply invasive tumors, whereas other guidelines1 2 8 do not mention the use of brachytherapy in this context.
In high-risk scenarios, brachytherapy may be appropriate for patients with positive margins or parametrial infiltration.9 However, extending this to intermediate-risk patients may be more complex, particularly if using brachytherapy as a boost. Nevertheless, some proponents use brachytherapy to reduce the dosage of EBRT in patients with high-risk factors or as primary treatment of locally advanced disease.10 11 Therefore, the purpose and optimal use of brachytherapy remain poorly defined. This approach and its use are not thoroughly described in the literature, and the lack of consensus in guidelines for it is reflected in clinical practice, with up to 27% of patients receiving brachytherapy.12 This highlights the heterogeneity in treatment received by patients with intermediate-risk cervical cancer.
The objective of our study was to assess the frequency of brachytherapy use, the doses administered in EBRT alone or in combination with brachytherapy, and the overall survival impact of brachytherapy as adjuvant treatment in patients with intermediate-risk, early-stage cervical cancer.
Methods
This was a retrospective cohort study using the most recent available information with complete survival data from the National Cancer Database (NCDB). All patients diagnosed with squamous cell carcinoma, adenosquamous carcinoma, or adenocarcinoma from 2004 to 2019 according to the International Classification of Diseases for Oncology, Third Edition, were identified in the database.13 Only patients who underwent a radical hysterectomy and lymph node staging (including sentinel lymph node procedures) as their primary treatment, irrespective of the surgical approach, were considered for study inclusion. Patients with clinically or pathologically reported positive nodes, metastatic disease, positive tumor margins, or parametrial invasion or with a previous history of another tumor were excluded from further analysis. Given that the NCDB lacks information regarding stromal invasion, the original criteria for radiation reported by Sedlis et al.,3 could not be completely used. Therefore, only tumor size and LVSI were considered for this analysis in two possible combinations: patients with limited cervical tumors larger than 4 cm irrespective of LVSI, and those with tumors ranging from 2 to 4 cm with LVSI. This categorization was described in previous reports of studies assessing patients with intermediate-risk cervical cancer and is currently under evaluation in a randomized controlled trial.6 14 For the overall survival analysis, only patients who received adjuvant radiotherapy with or without concurrent chemotherapy were included, and two treatment cohorts were analyzed: those who underwent EBRT alone and those who underwent EBRT plus brachytherapy.
Data imputation was performed for patients with missing LVSI information who potentially could be included in the analysis to compare adjuvant treatments. In multiple imputation analyses, five variables were used: age, race, and tumor histology, grade, and size. To validate our multiple imputations, the percentage of patients with LVSI and Kaplan–Meier survival curves created before and after the imputations were compared.
Demographic, pathological, and treatment characteristics were extracted from the deidentified data set. Propensity score matching was performed to create two comparable cohorts. In one cohort, patients received adjuvant EBRT without brachytherapy. In the other cohort, patients received adjuvant EBRT combined with brachytherapy. The variables applied to this matching based on its plausibility consisted of age, tumor size, race, LVSI, year of diagnosis, Charlson/Deyo comorbidity score, health insurance, histology, surgical approach, and chemotherapy administration.
Four variables were explored in subgroup analyses given the influence on oncologic outcome in patients with cervical cancer in previous studies. Minimally invasive surgery vs open surgery in the Laparoscopic Approach to Cervical Cancer Trial,15demonstrated worse outcomes for patients with early-stage cervical cancer undergoing minimally invasive surgery than for those undergoing open surgery, a finding reproduced in other retrospective studies.16 17 Also, histological comparison of squamous cell carcinoma with adenocarcinoma was considered, as retrospective evidence demonstrated worse outcomes for patients with nonsquamous carcinomas.18 19 Finally, LVSI was analyzed because it has a negative impact on recurrence and overall survival20 as well as whether to use chemotherapy given that concomitant chemotherapy has produced improved outcomes when given as adjuvant treatment in patients with high-risk21 and locally advanced cervical cancer22 but also because it is an option for patients with intermediate-risk cervical cancer under the National Comprehensive Cancer Network guidelines.1 The hazard ratios (HRs) for overall survival were estimated after refitting separate propensity score–weighted survival models for each subgroup.
The primary outcome of our study was overall survival. Categorical variables were compared using a Chi-squared test or Fisher’s exact test for the cells containing small values, and continuous variables were compared using independent sample t-tests in univariate analysis of patient characteristics. Kaplan–Meier survival analysis was performed to assess the survival trends in the groups. A Cox regression model was used to estimate the HR for death for the two cohorts.
The secondary outcome was the use and dose distribution of EBRT alone or with brachytherapy as adjuvant treatment. To analyze the total doses of ERBT and brachytherapy, the equivalent dose was calculated for patients with available data on dose and number of fractions.
Quantitative bias analyses were performed to assess the robustness of the estimates of results with regard to potential unmeasured confounding using a sensitivity analysis with the E-value, which quantifies the strength of an unmeasured confounder’s association with both the exposure and outcome needed to explain the observed relationship. The E-value was used to calculate the magnitude of the association an unmeasured confounder would be required to have with both the exposure (combined treatment) and outcome to fully explain the derived estimate (HR). Statistical uncertainty was addressed by calculating the E-value required to explain away the upper CI limit in our estimates. All statistical analyses were performed using SAS Enterprise Guide software (version 7.1; SAS Institute, Cary, NC, USA).
Results
We initially identified 1 57 882 patients with the International Classification of Diseases for Oncology codes for cervical cancer from 2004 to 2019 in the NCDB. After excluding patients who did not undergo surgery and those with other histologies, lymph node metastases, any other metastases, or parametrial invasion, we considered 17 194 patients with tumors larger than 4 cm and patients with tumors 2–4 cm in size with LVSI for our analysis. We addressed missing data on LVSI in our study, with LVSI information absent for 8008 of the 17 194 patients through multiple imputation techniques. Before imputation, 28.6% of the patients had LVSI. Following multiple imputations, the percentage of patients with LVSI among the whole cohort was 29.4%. We then compared the trends in Kaplan–Meier survival curves before and after the imputations, with the curves exhibiting similar trends at both time points. Additionally, we calculated the HRs for LVSI before and after imputation, which were 0.592 and 0.588, respectively. After excluding the patients who did not undergo adjuvant radiotherapy, we included 1174 patients in the definitive analysis. In total, 861 patients (73.3%) received EBRT alone, whereas 313 patients (26.7%) received EBRT and brachytherapy (Figure 1).
Cohort selection from the National Cancer Database.
Univariate analysis of patient characteristics demonstrated significant differences between the brachytherapy group and the EBRT-only group. Patients with tumors larger than 4 cm and without LVSI were more likely to receive the combination with brachytherapy than were those with tumors 2–4 cm in size and LVSI (Table 1). After 2:1 propensity score matching, we matched 620 (66.5%) patients who received EBRT alone with 312 (33.5%) patients who received EBRT combined with brachytherapy. We found no standardized differences in the matched cohorts greater than 5% for any of the characteristics (Table 1). With this sample size (n=932), we had 79% power to detect an 8% survival probability difference between the two groups if such an effect truly existed.
Patient characteristics
In total, 337 patients (36.2%) had tumors that were 2–4 cm in size, and 583 (62.6%) had tumors larger than 4 cm. Tumor size information was missing for 12 patients (1.3%), consequently we based study inclusion on the reported disease stage. According to the histology data, 594 patients (63.7%) had squamous cell carcinoma, 257 patients (27.6%) had adenocarcinoma, and 81 patients (8.7%) had adenosquamous carcinoma. Additionally, 621 patients (66.6%) had LVSI, whereas 311 patients (33.4%) did not. Furthermore, 349 patients (37.4%) underwent open surgery, and 310 patients (33.3%) underwent minimally invasive surgery; and data regarding surgical approach were missing for 273 patients (29.3%). In terms of adjuvant chemotherapy, 535 patients (57.4%) received it, 392 patients (42.1%) did not, and five patients (0.5%) had no data regarding chemotherapy administration.
Regarding the radiotherapy doses, information was available for 740 patients (79.4%). The median EBRT dose and number of fractions were 45 Gy (range, 45–50 Gy) and 28 (range, 25–28), respectively. In the cohort study, 79.4% of the patients had four distinct treatment patterns: group 1 (258 patients (34.9%)) received EBRT only with a median dose of 45 Gy; group 2 (209 patients (28.2%)) received EBRT with a median dose of 45 Gy combined with brachytherapy; group 3 (218 patients (29.5%)) received EBRT only with a median dose of 50.4 Gy; and group 4 (55 patients (7.4%]) received EBRT with a median dose of 50.4 Gy combined with brachytherapy. We found no significant differences in overall survival according to facility type (32.6% for an academic facility vs 35.7% for another facility; p=0.54) or year of treatment administration (p=0.16) (Online Supplemental Figure 1).
Supplemental material
Inclusion of doses and fractions in brachytherapy reports in the NCDB has been mandatory only since 2018. Therefore, complete dose and fraction information was available for only 86 patients since then. The total median equivalent dose in group 2 was 63.75 Gy, whereas that in group 4 was 67.57 Gy.
The median follow-up durations for patients who did and did not receive brachytherapy were 74.6 months and 68.8 months, respectively (p=0.1). We did not observe a difference in overall survival between patients who received EBRT only and those who also received brachytherapy (HR, 0.88 (95% CI, 0.63 to 1.23)); p=0.45. At 5 years after treatment, the overall survival rates were 86% for patients given EBRT only and 88% for those given EBRT with brachytherapy, with no significant difference observed (p=0.48). At 10 years, the overall survival rates were 77% for patients given EBRT only and 80% for those given EBRT with brachytherapy, with no significant difference observed (p=0.32). Figure 2 shows Kaplan–Meier estimates of overall survival according to adjuvant radiotherapy modality.
Kaplan–Meier (K–M) estimates of overall survival for patients with intermediate-risk cervical cancer according to adjuvant radiotherapy mode at the end of follow-up. EBRT, External beam radiotherapy.
In our exploratory subgroup analyses, brachytherapy was not associated with improved overall survival following stratification according to histology (HR, 0.80 (95% CI, 0.50 to 1.27) for squamous cell carcinoma), LVSI (HR, 1.18 (95% CI, 0.63 to 2.19)), or surgical approach (HR, 0.82 (95% CI, 0.40 to 1.65) for open surgery). However, among patients not receiving concomitant chemotherapy, we observed a markedly longer overall survival rate for patients receiving brachytherapy than for those not receiving it (HR, 0.48 (95% CI, 0.27 to 0.86); p=0.011) (Figure 3, Online Supplemental Figures 2–9).
Supplemental material
Supplemental material
Supplemental material
Supplemental material
Supplemental material
Supplemental material
Supplemental material
Supplemental material
Subgroup analysis of the associations between brachytherapy boost with EBRT and all-cause mortality according to histology, LVSI, surgical approach, and concomitant chemotherapy. MIS, minimally invasive surgery; EBRT, External Beam Radiotherapy; LVSI, Lymph Vascular Space Invasion; MIS, Minimally Invasive Surgery.
We performed the E-value calculation for the HR of overall survival which was 0.88 (95% CI, 0.62 to 1.23) for EBRT combined with brachytherapy compared with EBRT alone. The E-value needed to shift the upper confidence limit to 0.99 and make it statistically significant was 1.50. This would have reversed the finding of overall survival in this study and demonstrated a clinically meaningful improvement in the HR for overall survival for EBRT and brachytherapy compared with EBRT alone. However, a weaker confounder would not have reversed this finding.
Discussion
Summary of Main Results
In the present study, the use of brachytherapy with EBRT was not associated with an overall survival benefit in patients with intermediate-risk cervical cancer after radical surgery. However, patients who did not undergo concomitant chemotherapy but did receive brachytherapy had a statistically significantly longer overall survival rate. Additionally, we found heterogeneity regarding the radiotherapy modalities and their doses, highlighting an absence of consensus in clinical practice regarding adjuvant treatment of intermediate-risk cervical cancer.
Results in the Context of the Published Literature
Since publication of the results of the Sedlis et al., trial3 5 authors have raised concerns about the role of adjuvant radiotherapy for intermediate-risk cervical cancer, particularly when considering the limitations identified in the trial.6 12 Furthermore, whereas most cervical cancer guidelines1 2 do not address the routine use of vaginal brachytherapy combined with adjuvant EBRT in patients with intermediate-risk cervical cancer, we found that 26% of the original cohort in our database search received brachytherapy. Some advocates of brachytherapy argue that in certain cases, patients could receive brachytherapy to increase the total radiotherapy dose, whereas in other cases, it is given to reduce the pelvic EBRT dose with the aim of improving morbidity based on other clinical settings.23 The percentage of patients receiving brachytherapy reported in the present study is consistent with previous reports.12 24
Additionally, we identified four distinct cohorts with different treatment patterns based on the initial EBRT dose (45 Gy vs 50 Gy) and use of brachytherapy. Data available for calculating the equivalent dose before 2018 were limited, resulting in only a small proportion of patients displaying the total radiation doses received. However, for 2018 and2019, most of these data were complete, revealing that patients receiving brachytherapy tended to receive higher total doses (>60 Gy) than did patients not receiving it.
Published research on the role of additional brachytherapy in the treatment of cervical cancer is limited and often characterized by varied inclusion criteria or small samples. In a retrospective cohort study, researchers assessed the use of brachytherapy boost with EBRT in 480 patients with intermediate- or high-risk cervical cancer at five tertiary cancer centers in Turkey.24 According to institutional protocols, the criteria used for vaginal brachytherapy administration consisted of close (≤5 mm) or positive surgical margins, grade 3 histology, and extensive LVSI. Three hundred of the patients (62.5%) received a brachytherapy boost. The authors reported no significant differences in both cohorts in terms of overall survival (p=0.11), recurrence-free survival (p=0.49), local recurrence-free survival (p=0.16), and distant metastasis-free survival (p=0.08) rates, which was mostly consistent with our findings.
In other retrospective cohort studies including 94 patients with intermediate-risk cervical cancer,25 authors reported a benefit for patients receiving a brachytherapy boost with EBRT. The brachytherapy group had a significantly superior 5 year locoregional recurrence-free survival rate (87.7% vs 72.5%; p=0.004), whereas the difference in overall survival rate between the groups was not significant (78.4% vs 75.3%; p=0.055). However, that study was limited by a small sample.
The American Brachytherapy Society recommends that cervical cancer patients with large or deeply invasive tumors or extensive LVSI should receive a brachytherapy boost with EBRT.7 However, besides our findings, evidence supporting the benefit of using brachytherapy in terms of oncological outcomes for patients with intermediate-risk cervical cancer as well as the potential for reduction in toxicity for the combined treatment group is lacking. For patients with intermediate-risk cervical cancer, the National Comprehensive Cancer Network cervical cancer guidelines1 also allow the use of concomitant chemotherapy, even when it is not part of any randomized controlled trial with available results.3 Our study demonstrated that combining EBRT with brachytherapy resulted in a significant improvement in overall survival for patients not receiving concomitant chemotherapy. Concomitant chemotherapy was common in our original cohort, as 52.8% of the patients received it as part of their adjuvant treatments. The use of chemotherapy may reduce the brachytherapy benefit in this population. However, precise reasons for the different outcomes associated with the use of chemotherapy have yet to be established.
Strengths and Weaknesses
The main strengths of our cohort study are the rigorous process used for the intermediate-risk patient selection and the large cohort. However, our study had certain limitations that must be considered, such as the requirement of multiple imputation analysis for LVSI description, including for patients with missing data after 2010 as well as patients with missing data before 2010, as the LVSI variable was not reported in the NCDB before that time. Another limitation was the absence of stromal invasion information in the database and the resulting study inclusion based only on tumor size and LVSI. Given that the database had no data regarding recurrence-free survival, information about the possible impact of brachytherapy on recurrence rates in this population is lacking. Most patients before 2018 had no information regarding brachytherapy reasons for administration, dose and fractioning, making calculation of equivalent doses for our analysis impossible.
Implications for Practice and Future Research
To assess the benefits of adjuvant radiotherapy with radical surgery in patients with intermediate-risk cervical cancer in a contemporary context, researchers launched CERVANTES, a phase 3 randomized controlled trial for which they expect to recruit 514 patients.14 Unfortunately, a confounding factor in that trial is that the intervention arm includes the option of using brachytherapy and/or chemotherapy as part of adjuvant schemes. Chemotherapy and brachytherapy use are expected to be analyzed as predefined subgroups. Evidence from prospective assessment of the use of brachytherapy boost and chemotherapy in this context is lacking thus far and this intervention requires further investigation.
Conclusions
The use of various radiotherapy doses and adjuvant radiotherapy techniques highlights the treatment heterogeneity in patients with intermediate-risk cervical cancer. We found no difference in overall survival for EBRT alone vs EBRT and brachytherapy. However, the patients who did not receive concomitant chemotherapy had an increased overall survival rate when they received brachytherapy.
Data availability statement
Data are available upon reasonable request.
Ethics statements
Patient consent for publication
Ethics approval
Not applicable.
Acknowledgments
NA acknowledges the support of Fundación Alfonso Martin Escudero in achieving her international research training at MD Anderson. Editorial support was provided by Donald Norwood in Editing Services, Research Medical Library at MD Anderson.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
X @RParejaGineOnco, @colbertle
Contributors The authors affirm that they have each met criteria for authorship as defined by the International Committee of Medical Journal Editors. NA: conceptualization, data curation, methodology, writing of the original draft, manuscript review and editing, guarantor. AK: writing of the original manuscript draft, manuscript review and editing. AM: methodology, writing of the original manuscript draft, manuscript review and editing. RP: writing of the original manuscript draft, manuscript review and editing. C-FW: data curation, formal analysis, methodology, validation, manuscript review and editing. RN: methodology, writing of the original manuscript draft, manuscript review and editing. LC: methodology, writing of the original manuscript draft, manuscript review and editing. JAR-H: conceptualization, methodology, supervision, writing of the original manuscript draft, manuscript review and editing. DV-C: conceptualization, data curation, methodology, writing of the original manuscript draft, manuscript review and editing.
Funding This work was supported by grants from the NIH (JAR-H: K08CA234333, R01MD017999; RN, AK, NA, and JAR-H: P30CA016672; AK: 5T32 CA101642).
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.