Histopathologic Factors

In the study by Sedlis et al,7 78.7% of patients had squamous cell carcinoma and 23% had non-squamous subtypes (adenocarcinoma and adenosquamous). It has been shown that non-squamous subtypes are considered more aggressive with different risk factors and prognoses.15 16 Different classifications have been developed for endocervical adenocarcinoma, such as the Silva classification,17 and the International Endocervical Adenocarcinoma Criteria and Classification modified the classification of endocervical adenocarcinomas, separating these into human papillomavirus associated and human papillomavirus independent categories.18 These classifications demonstrate different prognostic features, even in the same histological type. It could be argued that the study of Sedlis et al generates conclusions that primarily apply to squamous cell carcinomas

It is also interesting to mention that although the intermediate risk established by the combination of the factors (tumor depth of invasion, lymphovascular space invasion, and tumor size) predicted a 3 year recurrence of ≥30%,7 it seems that this risk differs according to the histological subtype, and in the presence of one or more factors. Levinson et al16 developed an ancillary analysis of GOG 49, 92, and 141, and included stage I patients who underwent surgery and received no neoadjuvant/adjuvant therapy; among them, depth of invasion was associated with the most significant recurrence risk for squamous carcinoma (16% of risk), while for adenocarcinoma it was tumor size (15% of risk with negative lymphovascular space invasion vs a 25% risk with positive lymphovascular space invasion). This study guides the creation of a nomogram to assess the risk of recurrence based on histology. With these findings, the classic combinations could be reconsidered as a criterion to define adjuvant therapy.

Other prognostic factors have been evaluated in the setting of intermediate risk. Tumor budding is defined as the presence of clusters of tumorous cells detaching from the invasive margin of the primary tumor. Park et al19 evaluated its prognostic value. High tumor budding was defined as ≥5 per high power field for disease free survival. The authors found on multivariate analysis that high tumor budding was an independent prognostic factor for predicting overall survival (HR, 4.96; 95% CI 1.06 to 23.29; p=0.0423). Addition of this factor to the conventional intermediate risk models improved the accuracy of recurrence prediction.

Cibula et al20 retrospectively analyzed different prognostic models, including 15 factors: age along with 14 other factors obtained from the preoperative imaging and pathological assessment (11 tumor related, and three lymph node status related: lymph node positivity, number of positive lymph nodes, and type of metastasis in lymph node). Tumor free distance was defined by measuring the image on a transversal plane as a minimum distance of uninvolved stroma between the tumor and pericervical ring. Adenosquamous tumor type, lymph node involvement, and tumor free distance ≤3.5 mm were the most significant independent factors for the risk of recurrence in the entire cohort. In the cohort of lymph node negative patients, the combination of adenosquamous tumor type and tumor free distance ≤3.5 defined a group of 3% of cases with a recurrence rate (RR) of 33% in comparison with a 6% RR in the rest of the cohort. Tumor free distance was the best marker in both predictive models for the entire population and for the lymph node negative subgroup. This factor, when evaluated preoperatively, could represent an advantage for planning the surgical approach.

Although the proposal to apply normograms to determine adjuvant therapy through an individualized approach according to risk of recurrence, and the addition of preoperative and postoperative prognostic factors is promising,16 19 20 these should be validated before their use in clinical practice.

Lastly, it is important to highlight that in the study by Sedlis et al,7 institutional pathologists measured the depth of stromal tumor invasion and made a determination of the presence or absence of tumor in the capillary lymphatic spaces. Subsequently, and as quality control, all histology slides were reviewed by two pathologists of the GOG pathology committee but tumor size was only clinically determined.

Changes in FIGO Stage IB Classification

In 1958, the cervix was the first organ to be assigned to a clinical staging system for cancer by FIGO, and the pathologic (TNM) staging was used to report nodal and metastatic disease status. The classification has undergone modifications in recent decades.8 By 1994, 'bulky' stage IB was divided into IB1 (≤4 cm) and IB2 (>4 cm), to establish specific treatment strategies, according to their different prognostic impact.21

In the FIGO 2009 classification, staging was based mainly on physical examination; the use of diagnostic imaging techniques to assess the size of the primary tumor was encouraged but not mandatory, given the limited resources in some regions.22 In 2018, FIGO revised the staging classification to allow clinical, radiological, or pathological findings as valid tools to assign stage of disease.8 In addition, taking into account that recurrence rates are lower in stage I tumors <2 cm in size compared with tumors measuring 2–4 cm, stage IB was defined as an additional cut-off at 2 cm, establishing three stage categories (IB1, IB2 and IB3).23

The study by Sedlis et al7 included stage IB patients with tumors from <2 cm (28.1%) to >6 cm (10.4%). There were 74 (26.7%) patients with tumors >4 cm, a factor taken into account in two of the four combinations included in that trial. Based on this information, just over a quarter of the patients who participated in that study would not be candidates for surgery today, considering that they have stage IB3 by current guidelines and would, in turn, be treated with chemoradiation.

Tumor Size Measurement

Determining the actual size of the tumor has been the subject of debate in patients with cervical cancer. The evidence suggests that clinical examination, although a standard and necessary step in the initial evaluation of patients, is limited in its ability to reliably ascertain tumor size, with an accuracy of 47–70%.24 Among the limitations are that it is examiner dependent, has a high interobserver variability, its impacted by patient specific factors such as tolerance of examination or body mass index, and it becomes increasingly more challenging in patients with endocervical tumors.24 25

For measurement of tumor size by histopathology evaluation, it is important to highlight the discrepancy among institutions and even among pathologists on reporting.24 In the FIGO 2018 classification, the term 'diameter' is used interchangeably with 'dimension', when they represent two different definitions and measurements. In addition, inconsistencies have been found in the measurement according to the location of the tumor (eg, if it is in the anterior or posterior lip, or if the tumor is circumferential).26 Recent data have shown discrepancies in the estimation of preoperative tumor size compared with postoperative tumor size. Uppal et al27 analyzed preoperative tumor size during the clinic visit and tumor size as reported in the pathology report (final tumor size) in 744 patients with early cervical cancer undergoing radical hysterectomy. A total of 291 patients had no visible disease on preoperative assessment, but 58 (19.9%) patients had tumors >2 cm on final pathology, and of 257 patients with preoperative assessment of tumor size <2 cm, 89 (34.6%) had tumors >2 cm on final pathology.

In the study by Sedlis et al,7 it is mentioned that tumor diameter was estimated by palpation and there is no mention of who performed the physical examination or standardization of this measurement. In addition, there is no mention of tumor size by images. There was no documentation of evaluation of tumor size in the final surgical specimen. What the reproducibility of these selection criteria would be in current practice is difficult to determine when using magnetic resonance imaging, computed tomography scan, or ultrasound imaging, which allow a more precise estimation of tumor size.

Characteristics of Radical Surgery

Details regarding the radicality of surgery (type II or III radical hysterectomy) was not well established in the study by Sedlis et al. No details are provided on the surgical volume of radical surgeries performed by physician and/or center. In the study, all patients underwent bilateral pelvic lymphadenectomy but it is important to mention that the integration of sentinel lymph node biopsy and ultrastaging has been recently introduced into the routine management of cervical cancer,28 29 allowing detection of an additional 10–15% of cases with low volume disease.30 The technique also identifies sentinel lymph nodes in 'unexpected' nodal basins in up to 17% of patients.31 Detection of lymph node involvement using this technique identifies patients with high risk for recurrence who may benefit from adjuvant treatment with chemoradiation.30 With the evolution of radical surgery regarding the technique and indications in the last decade, the results of the Sedlis study could have less external validity in current clinical settings.

Advances in Radiotherapy Techniques

Treatment modalities with radiotherapy have been modified over time.32 Initially, teletherapy was administered using a two-dimensional technique. Subsequently, three-dimensional conformal radiation therapy was introduced and with the advancement of radiotherapy, the standard treatment modality is intensity modulated radiation therapy, which makes it possible to deliver doses to target volumes with less toxicity to nearby critical structures33 without differences in disease outcomes.34

The clinical practice guideline of the American Society for Radiation Oncology35 currently suggests that for patients meeting intermediate risk criteria, whole pelvic radiotherapy may be delivered to a total dose of 4500–5040 cGy, in 180 cGy per fraction, or 4000–4400 cGy in 200 cGy per fraction. In the Sedlis study,7 pelvic irradiation was given with a fourfield technique with a megavoltage beam, although cobalt-6 was allowed if the source–skin distance was >80 cm. The radiation dose was 46–50.4 Gy with daily fractions of 1.8–2 Gy over 4.5–6 weeks. Currently, radiation techniques such as three-dimensional conformal radiation therapy, intensity modulated radiation therapy, and volumetric modulated arc therapy are the standard of care, with potentially lower long term morbidity. For future studies, implementation of these techniques should be considered, seeking to improve pelvic control and thereby disease free survival, with a lower threshold (8–10% risk of pelvic relapse).36

Further Points of Consideration

The study of Sedlis et al demonstrated a reduction in risk of recurrence after radiotherapy, but not in overall survival. One potential reason is that the sample size might have been too small as this study was planned to have 80% power to detect risk reductions of 46% or greater in overall survival; however, patients with radiotherapy had only a 30% reduction in the risk of mortality and significance was not reached.37 Other aspects related to the design of the study include the evidence of a higher percentage of worse prognostic factors in the radiotherapy group, more adenocarcinomas, and tumors >3 cm.11 A potential risk of bias is to be highlighted because the method of randomization, allocation concealment, or blinding of the investigators was not reported.7 It has been described that inadequate concealment of allocation and lack of blinding are often associated with overestimation of the effects of treatment.38 All of these aspects should be considered in the design of a new study.

Impact of Observation Versus Adjuvant Therapy

When adjuvant therapy is administered in the presence of intermediate risk criteria, the oncological benefit must be balanced against morbidity. Regarding the oncological results, there are inconsistent results in the literature and this is likely because patient populations are heterogeneous, most studies are retrospective, and different treatment strategies have been used. Treatment recommendations for patients with intermediate risk disease include radiotherapy with or without brachytherapy, chemoradiotherapy, or observation.2 12

Rogers et al39 evaluated the effectiveness and safety of adjuvant therapies after radical hysterectomy for early stage cervical cancer. Two prospective randomized trials were included.11 40 No significant difference was found in risk of death within 5 years between women who received radiation and those who received no further treatment (RR 0.8; 95% CI 0.3 to 2.4). However, women who received radiation had a significantly lower risk of disease progression at 5 years (RR 0.6; 95% CI 0.4 to 0.9). The authors concluded that the available evidence is not of high quality.

Regarding the role of concomitant chemotherapy, Kim et al41 evaluated 316 patients with stage IB–IIA cervical cancer who underwent radical hysterectomy with pelvic lymphadenectomy and had one or more intermediate risk factors and no high risk factors. Based on adjuvant treatment (adjuvant radiotherapy alone vs adjuvant chemoradiotherapy), the 5 year recurrence free survival rates (90.8% vs 88.9%, p=0.63) and 5 year overall survival rates (95.9% vs 91.0%, p=0.29) were not different.

Guo et al42 performed a systematic review with meta-analysis that evaluated the effect of post-surgical adjuvant radiotherapy versus adjuvant chemoradiotherapy in patients with intermediate risk. Eleven retrospective cohort studies and two randomized clinical trials were included. Adjuvant chemoradiotherapy was found to have better recurrence free survival in patients with multiple intermediate risk factors (odds ratio (OR) 3.11; 95% CI 1.04 to 4.99; p<0.0001; i²=6%.) However, similar benefit was observed between both regimens in the presence of a single intermediate risk factor (OR 1.80; 95% CI 0.96 to 3.36; p=0.07; i²=0%). To confirm the effect of chemotherapy in the intermediate risk group, the results of a prospective study from the GOG 263 (NCT01101451) are awaited43

Brachytherapy has been included in some institutions as adjuvant therapy in the setting of intermediate risk. This treatment technique is recommended in the presence of positive surgical margins as a high risk factor.2 13 Rotman et al11 noted a high locoregional failure rate (14% in the radiation arm), suggesting the need to increase radiotherapy doses, either by external beam or in conjunction with brachytherapy. Further control in the pelvis, specifically in the vagina, may be achieved by using additional vaginal brachytherapy.36 Bronsart et al44 examined the role of adjuvant brachytherapy alone as an alternative to external beam radiation therapy in 40 patients. This treatment was indicated when tumor size was ≥2 cm and/or in the presence of lymphovascular space invasion. Ninety per cent of patients were in complete remission and four patients (10%) experienced tumor relapse, all in the peritoneal cavity. At 5 years, overall survival was 83.6% (95% CI 67.8% to 100%) and disease free survival was 85.1% (95% CI 72.6% to 99.9%). On the other hand, Yu et al45 evaluated the role of postoperative adjuvant chemotherapy combined with intracavitary brachytherapy. One hundred and two patients were subjected to chemotherapy combined with intracavitary brachytherapy (chemotherapy+intracavitary brachytherapy) and the remaining 70 patients were treated with concurrent chemoradiation. No significant difference was observed in 3 year disease free survival (93.2% vs 95%, p=0.862), and 3 year overall survival (96.0% vs 95.4%, p=0.599).

Observation in the presence of intermediate risk factors is becoming increasingly more common. Cibula et al46 published a retrospective study comparing outcomes in patients with intermediate risk factors that underwent observation or radiotherapy according to the protocol of three different centers. In all, 231 patients were evaluated in the study; 127 patients were treated with radical surgery without adjuvant treatment (surgery only group) and 104 patients with radical surgery, followed by adjuvant radiotherapy (control group). No differences were found for total recurrence rate between the groups (surgery only group 6.3% vs control groups 11.5%; p=0.168), 5 year overall survival (95% vs 91%; p=0.317), and disease specific survival (96% vs 95%; p=0.745). It is interesting to note that this study demonstrated improved oncologic outcomes in the group without adjuvant treatment, compared with the original GOG 92 (recurrence rate 6.3% and local recurrence rate 1.6% vs Sedlis study which were 27.9% and 19.3% respectively).

Van der Velden et al47 retrospectively analyzed 161 patients with intermediate risk factors who did not receive adjuvant treatment within a cohort of 937 patients with cervical cancer stage IB (FIGO 1995 classification) who underwent open radical surgery. Twenty-five (15.5%) patients had a recurrence, and nine (5.6%) patients had an isolated locoregional recurrence. The 5 year recurrence free survival was 86.6% and the 5 year overall survival was 90%.

Cao et al48 retrospectively compared observation with adjuvant radiotherapy or chemoradiation in 861 patients with stage IB and IIA disease (FIGO 2009 classification) and intermediate risk. In total, 75.2% (n=646) of patients had tumors >4 cm and 46.5% (n=400) of patients had lymphovascular space invasion. Eighty-five (9.9%) underwent no adjuvant treatment, 283 (32.9%) underwent adjuvant radiotherapy, and 493 (57.2%) underwent concurrent chemoradiation. The 5 year recurrence free survival was 87.1%, 84.2%, and 89.6% (p=0.27) for the no adjuvant treatment, radiotherapy alone, and concurrent chemoradiotherapy groups, respectively. Adjuvant treatment was not a significant independent prognostic factor for recurrence free survival. On multivariate analysis, lymphovascular space invasion was an independent negative prognostic factor for recurrence free survival (HR 2.26; 95% CI 1.44 to 3.55; p<0.001).

At the population level, Nasioudis et al49 reported a study using the National Cancer Database. The authors investigated the outcomes of observation alone versus adjuvant radiotherapy for 765 patients with squamous, adenosquamous carcinoma, or adenocarcinoma of the cervix, pathological stage IB disease (FIGO 2018 classification), who underwent radical hysterectomy with at least 10 lymph nodes removed (all negative), negative tumor margins, and presence of intermediate risk factors, defined as tumor size 2–4 cm with lymph-vascular invasion or tumor size >4 cm. As the National Cancer Database does not collect data on the depth of stromal invasion, the traditional criteria of Sedlis et al7 were not applied. The 4 year overall survival rates were 88.4% and 87.1% (p=0.44) among patients who received or did not receive adjuvant radiation therapy, respectively. When the prognostic variables age, histology, and surgical approach were controlled, there were also no significant differences in overall survival between the groups. The authors found no benefit of administering concurrent chemotherapy with 4 year overall survival rates of 89.8% and 86.3% (p=0.36). Of note, in this study, most patients (57.9%) received, in addition to radiotherapy, concomitant chemotherapy thus reflecting the heterogeneity in the treatment approach. On the other hand, it should be highlighted that the National Cancer Database does not collect data on tumor recurrence, an outcome that demonstrated a benefit from the addition of adjuvant radiotherapy in the Sedlis study.7

Similar results have been published in other retrospective studies comparing the use of adjuvant radiotherapy (with or without chemotherapy) versus observation for intermediate risk patients.50–53 Gómez-Hidalgo et al54 recently conducted a systematic review with meta-analysis including articles with ≥10 patients with two comparative arms. The relative risk of recurrence (RR 1.49; 95% CI 0.81,2.75) and the relative risk of mortality (RR 1.34; 95% CI 0.71,2.54) were similar in both groups independently whether they did or did not receive adjuvant therapy.

Treatment Related Morbidity

Another aspect to consider when deciding whether to provide adjuvant therapy in intermediate risk patients is treatment related morbidity. This is important because approximately 28% of patients undergoing radical surgery for early cervical cancer receive adjuvant therapy.55 The reported incidence of toxicity associated with adjuvant treatment varies in different studies due to the use of many classification systems or incomplete records, especially regarding mild toxicity. Late severe toxicity was observed in about 5% of patients treated with postoperative radiotherapy and in up to 15% of those treated with chemoradiation.56 Potential morbidity of pelvic irradiation includes diarrhea, urinary urgency or frequency, hematuria, as well as potential urinary or bowel incontinence57 The morbidity of radical surgery followed by radiation therapy is sufficiently great that clinicians should consider primary chemoradiation for patients with a need for subsequent radiation after surgery.58

Sedlis et al7 reported 6% grade 3–4 adverse events in the radiotherapy arm versus 2.1% in the no further therapy group. Kim et al41 reported that combined treatment with chemoradiotherapy was associated with a significantly higher risk of grade 3 toxicity compared with radiation therapy alone. Sandadi et al59 compared the morbidity of radical surgery with or without adjuvant radiation therapy in stages IB1–IB2 cervical carcinoma (FIGO 1995 classification). The rate of grade 3 or higher complications was similar (5% vs 4%, respectively; p=0.999). Although it is assumed that the addition of adjuvant therapy to surgical management increases severe morbidity, modern surgical and radiotherapy techniques could limit its impact. Lastly, several studies suggest that the quality of life and sexual functioning of cervical cancer survivors treated with radiotherapy are worse than surgery alone or surgery followed by chemotherapy without radiotherapy.60 61 This aspect is crucial when deciding to provide adjuvant therapy.

Considering the technological advances in radiotherapy described above, modifications in the surgical technique, and growing evidence of similar oncological outcomes in providing adjuvant versus no additional treatment, a reduction in treatment related morbidity is expected for this population of patients with early cervical cancer in future studies.