Introduction

Lymph node (LN) metastasis is an important adverse prognosticator for cervical cancer and detected in 15.3–25.5% of patients who have 2009 International Federation of Gynecology and Obstetrics (FIGO) stage IB-IIA disease [1, 2]. Nevertheless, the status of the LNs was not considered in the FIGO staging system till 2018. In 2018 the staging system for cervical cancer was revised by the FIGO Committee and LN metastasis was defined as stage IIIC disease [3].

Previous studies reported that women with node-positive cervical cancer had 5-year overall survival (OS) rates between 62.0 and 81.9% [4, 5]. It was argued that these patients were a heterogeneous group with certain clinic-pathologic risk factors for relapse and survival [5,6,7]. Tumor histology is among one of the most important risk factors considered. It is well-known that squamous cell carcinoma (SCC) and adenocarcinoma (AC) histologies exhibit similar prognosis in node-negative patients, whereas AC histology is related with worse prognosis in patients who are node-positive [8, 9]. Additionally, the number of metastatic lymph nodes (mLNs) has been proposed as a prominent prognostic factor in patients who have node-positive cervical cancer [5, 7, 10]. However, the prognostic significance of the number of mLNs has not been delineated in specific histologic types of cervical cancer.

This retrospective dual-institutional study presents the experience in patients who uniformly received radical hysterectomy plus systematic pelvic and para-aortic LN dissection followed by concurrent radiochemotherapy. The purpose of this study is to clarify the prognostic value of the number of mLNs in various histological types of node-positive cervical cancer. Therefore, the prognostic significance of the number of mLNs was investigated in stage IIIC patients with SCC histology and stage IIIC patients with non-SCC histology separately.

Methods

The cervical cancer patients who were subjected to radical hysterectomy with bilateral pelvic and para-aortic lymphadenectomy between January 1st, 2006 and May 6th, 2019 at two gynecologic oncology centers in Ankara, Turkey were retrospectively reviewed after the approval of Institutional Review Board (Başkent University Institutional Review Board Approval Number: KA 19/427).

We included women who met the following criteria: (1) FIGO stage IIIC (p) cervical cancer, (2) SCC, AC or adenosquamous carcinoma, (3) no neo-adjuvant therapy, (4) no synchronous malignancy. Women who had no complete medical records, those who had less than 10 pelvic LNs and less than five para-aortic LNs on their final pathology reports, and those who did not complete standard adjuvant radiochemotherapy were excluded.

We obtained the following data from medical records of patients; age at diagnosis, 2009 FIGO stage, 2018 FIGO stage, lymphovascular space invasion (LVSI) (absent/present), stromal invasion depth, vaginal involvement (absent/present), microscopic parametrial involvement (absent/present), primary tumor size (cm), positive surgical margin (absent/present), number of LNs removed, location of LNs involved, size of the mLNs > 1 cm (yes/no) and number of mLNs. The number of mLNs was categorized as mLNs ≤ 2 and mLNs > 2 according to previously published studies [5, 10, 11]. Also, the date of surgery, type of adjuvant therapy, the recurrence status, and the date of last visit or death, were abstracted.

Cervical biopsy and pelvic examination were the essential components of the pre-operative work-up. A pelvic ultrasound or MRI (Magnetic Resonance Imaging) was applied to all women in order to guide the treatment plan. Cystoscopy and rectoscopy were performed when clinically indicated. A radical hysterectomy with pelvic and para-aortic lymphadenectomy was planned for the women with 2009 FIGO stage IB-IIA disease [12]. After the emerging of the 2018 FIGO staging system, we routinely used MRI or PET-CT (positron-emission tomography) to assess LNs [3]. Thereafter, primary chemoradiation was planned for women with positive nodes on imaging.

All the patients underwent abdominal Piver type III hysterectomy with bilateral pelvic and para-aortic LN dissection. Pelvic and para-aortic lymphadenectomy was performed as we described in our previous study [13]. Intraoperative frozen section examination for detecting nodal disease was not routinely requested. The attending surgeon decided whether to request a frozen section examination for suspicious LNs or not. In the case of a frozen-section analysis indicating positive nodal disease, radical hysterectomy was abandoned. Additionally, the surgical procedure was terminated if intra-peritoneal disease was detected during surgery.

Gynecological pathologists examined the surgical specimens and used conventional pathology for diagnosis. The data regarding the histopathological features were extracted from the original pathology reports and central pathology review was not performed. LVSI was described as the existence of tumor cells inside the lumen of the lymphatics or capillaries [14]. The histologic subtype of cervical cancer was determined according to the World Health Organization (WHO) classification system [15]. The patients were allocated to the FIGO 2018 stages based on their final pathology reports.

Adjuvant treatment of the patients was decided by the multidisciplinary tumor board at each institution. Whole pelvic external beam radiation therapy (EBRT) was administered at a dose of 45–50 Gy to the patients who had pelvic LN metastasis. During course of EBRT patients received systemic cisplatin (40 mg/m2) weekly. Chemoradiation was administered to all cases those who had para-aortic LN metastasis on final pathology. By employing a four field arrangement in the pelvis and, anterior and posterior fields in the para-aortic area, extended-field radiotherapy delivered 45–50 Gy to the pelvis and para-aortic area over 5 weeks at 1.8–2 Gy per fraction. During the period of extended-field radiotherapy, cisplatin (40 mg/m2) infusion was applied once-weekly. If there was vaginal involvement, vaginal brachytherapy was administered to the patients. Cylinders were employed to deliver vaginal brachytherapy to the upper 4 cm of the residual vagina. Three fractions of high dose rate brachytherapy of 7 Gy every other day, was prescribed to 0.5 cm from the surface of the applicator.

Recurrence was defined as pathological or radiological confirmation of tumor metastasis after three months disease-free period following primary surgery. Disease-free survival (DFS) was defined as the period between the date of initial surgery and first recurrence, death from any cause, or the date of last follow-up for women without any event. The duration between the date of initial surgery and death, or the last visit was regarded as OS. The women were categorized into two cohorts according to tumor histology as patients with SCC or patients with non-SCC. The non-SCC cohort consisted of adenosquamous carcinoma and AC histologies. The prognostic value of mLN count was determined in each of the cohorts separately.

We carried out the statistical analysis by employing the SPSS version 23.0 statistical software (IBM Corp., Armonk, NY, USA). Continuous variables were given as medians and ranges and categorical variables were reported in terms of counts and percentages. The continuous variables were compared with student’s t test whereas Chi-square test was utilized to compare the categorical variables. Furthermore, we employed Kaplan Meier method to form survival curves and the long-rank test to document the difference between survival curves. Multivariate analysis was applied to each variable which exhibited p value of less than 0.05 in the univariate analysis. Multivariate analysis was performed by using the Cox proportional hazard model. For all analysis p < 0.05 was regarded as statistically significant.

Results

Patient characteristics

This study includes a total of 191 women with FIGO stage IIIC (p) cervical cancer. Among these women 148 (77.5%) had SCC and 43 (22.5%) had non-SCC histology. The SCC and non-SCC cohorts were well balanced with regard to median age at diagnosis, 2018 FIGO stages, 2009 FIGO stages, LVSI, depth of cervical stromal invasion, vaginal involvement, microscopic parametrial involvement, tumor size, positive surgical margin, and the duration of median follow-up. The median number of removed pelvic, para-aortic and total lymph node numbers was similar between the groups. The two cohorts were also comparable in terms of parametrial, obturator, common iliac, presacral and external iliac LN metastasis. Sixty-five patients (43.9%) in the SCC cohort had mLNs > 2, and 24 (55.8%) patients in the non-SCC cohort had mLNs > 2 (p = 0.22). Table 1 demonstrates the clinical and pathological characteristics of the patients. We analyzed the prognostic value of the number of mLNs for patients with SCC histology and non-SCC histology separately.

Table 1 Baseline characteristic of 191 women with node-positive cervical cancer
Full size table

SCC cohort

In the SCC cohort, we compared the baseline characteristics of women with mLNs ≤ 2 and mLNs > 2. The women with mLNs > 2 were more likely to have FIGO stage IIIC2 disease (27/65 vs .7/83, respectively; p < 0.01), external iliac LN metastasis (48/65 vs . 30/83, respectively; p < 0.01), common iliac LN metastasis (13/65 vs .1/83, respectively; p < 0.01) and presacral LN metastasis (6/65 vs .0/83, respectively; p < 0.01) than those with mLNs ≤ 2.

Univariate analysis revealed that the FIGO stage IIIC2 disease (p = 0.04), tumor size ≥ 4 cm (p = 0.04) and mLNs > 2 (p = 0.005) were significantly related with decreased 5-year DFS. On multivariate analysis, mLNs > 2 remained as the only factor that was significantly associated with decreased 5-year DFS (hazard ratio [HR] = 2.06; 95% confidence interval [CI] 1.03–4.09; p = 0.03) (Table 2).

Table 2 Univariate and multivariate analyses of DFS in patients with FIGO stage IIIC SCC
Full size table

Univariate analyses revealed that there was no factor other than mLNs > 2 that significantly related to decreased 5-year OS (HR = 2.35, 95% CI 1.11–4.99; p = 0.02). As there was only one factor related to decreased 5-year OS, we did not perform multivariate analysis (Table 3).

Table 3 Univariate and multivariate analyses of OS in patients with FIGO stage IIIC SCC
Full size table

Non-SCC cohort

In the non-SCC cohort, we compared the baseline characteristics of women with mLNs ≤ 2 and mLNs > 2. The women with mLNs > 2 were more likely to have FIGO stage IIIC2 disease (14/24 vs.1/19, respectively; p < 0.01), LVSI (24/24 vs.15/19, respectively; p = 0.03), external iliac LN metastasis (17/24 vs.11/19, respectively; p < 0.01), obturator LN metastasis (21/24 vs.10/19, respectively; p = 0.01) and size of the mLNs > 1 cm (24/24 vs.9/19, p < 0.01) than those patients with mLNs ≤ 2.

Univariate analysis revealed that vaginal involvement (p = 0.02), microscopic parametrial involvement (p = 0.001), primary tumor size ≥ 4 cm (p = 0.02) and 2009 FIGO stage II disease (p = 0.03) were significantly associated with decreased 5-year DFS. However, on multivariate analysis, no independent prognostic factor was defined (Table 4).

Table 4 Univariate and multivariate analyses of DFS in patients with FIGO stage IIIC non-SCC
Full size table

Univariate analysis revealed that vaginal involvement (p = 0.006), presacral LN metastasis (p = 0.01) and FIGO stage II disease (p = 0.01) were significantly related with decreased 5-year OS. However, on multivariate analysis, no independent prognostic factor was defined (Table 5).

Table 5 Univariate and multivariate analyses of OS in patients with FIGO stage IIIC non-SCC
Full size table

Effect of tumor histology and number of mLNs on survival for the whole study population

We grouped the patients with SCC histology based on the number of mLNs as mLNs ≤ 2, and mLNs > 2. However, we did not stratify the patients with non-SCC histology because the number of mLNs was not a prognostic factor for survival outcome in non-SCC cohort according to the results of our study. Thereafter we analyzed the baseline characteristics and survival outcomes of SCC with mLNs ≤ 2, SCC with mLNs > 2, and non-SCC groups.

There was no statistically significant difference between the groups in terms of age at diagnosis, LVSI, depth of cervical stromal invasion, vaginal involvement, microscopic parametrial involvement, tumor size, positive surgical margin, parametrial LN metastasis, obturator LN metastasis and the duration of median follow-up. However, the SCC with mLNs ≤ 2, SCC with mLNs > 2, and non-SCC groups were significantly different with regard to FIGO stage IIIC2 disease (27/65, 7/83, 15/43, respectively; p < 0.01), external iliac LN metastasis (48/65, 30/83, 28/43, respectively; p < 0.01) common iliac LN metastasis (13/65, 1/83, 8/43, respectively; p < 0.01), presacral LN metastasis (6/65, 0/83, 3/43, respectively; p = 0.02) and size of the mLNs > 1 cm (48/65, 48/83, 33/43, respectively; p = 0.04).

During the study period, 41 (21.4%) women had recurrent disease. In the SCC group with mLNs ≤ 2, there were three (3.6%) vaginal vault, two (2.4%) pelvic, three (3.6%) lymphatic, two (2.4%) distant relapses. In the SCC group with mLNs > 2, there were four (6.2%) vaginal vault, two (3.1%) pelvic, six (9.2%) lymphatic, three (4.6%) distant and one (1.5%) vaginal vault plus pelvic relapses. The recurrence locations in the non-SCC group as follows; three (7.0%) vaginal vault, four (9.3%) pelvic, three (7%) lymphatic, four (9.3%) distant, one (2.3%) lymphatic plus distant. At the time of reporting 142 women were alive and 49 were dead.

The 5-year DFS rates were 82.2% for the SCC group with mLNs ≤ 2, 57.5% for the SCC group with mLNs > 2 and 31.6% for non-SCC group (Fig. 1). The results of pairwise comparisons revealed that the SCC group with mLNs ≤ 2 had better 5-year DFS rate when compared to the SCC group with mLNs > 2 (82.2% vs. 57.5%, respectively; p = 0.005), and the non-SCC group (82.2% vs. 31.6%, respectively; p < 0.001). Nevertheless, the SCC group with mLNs > 2 and the non-SCC group were comparable in terms of 5-year DFS rates (57.5% vs. 31.6%, respectively; p = 0.06).

Fig. 1
figure 1

Kaplan–Meier curves for DFS according SCC with mLNs ≤ 2, SCC with mLNs > 2, and non-SCC groups (SCC with mLNs ≤ 2 vs. SCC with mLNs > 2, p = 0.005; SCC with mLNs ≤ 2 vs. non-SCC, p < 0.001; SCC with mLNs > 2 vs. non-SCC, p = 0.06) DFS disease-free survival, mLNs metastatic lymph nodes, SCC squamous cell carcinoma

Full size image

The 5-year OS rates were 85.3% for the SCC group with mLNs ≤ 2, 65.5% for the SCC group with mLNs > 2 and 46.8% for the non-SCC group (Fig.2). The results of pairwise comparisons revealed that the SCC group with mLNs ≤ 2 had a better 5-year OS rate when compared to the SCC group with mLNs > 2 (85.3% vs. 65.5%, respectively; p = 0.02), and the non-SCC group (85.3% vs. 46.8%, respectively; p < 0.001). Nevertheless, the SCC group with mLNs > 2 and the non-SCC group were similar in terms of 5-year OS rates (65.5% vs. 46.8%, respectively; p = 0.16).

Fig. 2
figure 2

Kaplan–Meier curves for OS according SCC with mLNs ≤ 2, SCC with mLNs > 2, and non-SCC groups (SCC with mLNs ≤ 2 vs. SCC with mLNs > 2, p = 0.02; SCC with mLNs ≤ 2 vs. non-SCC, p < 0.001; SCC with mLNs > 2 vs. non-SCC, p = 0.16) OS overall survival, mLNs metastatic lymph nodes, SCC squamous cell carcinoma.

Full size image

Discussion

Our study revealed that mLNs > 2 was not a prognostic factor in the non-SCC cohort, whereas it was independently associated with decreased 5-year DFS and 5-year OS rates in the SCC cohort. Furthermore, when the study population was classified as SCC with mLNs ≤ 2, SCC with mLNs > 2 and non-SCC, the SCC group with mLNs ≤ 2 had better survival outcomes when compared to the other two groups. To our knowledge, this study is the first to investigate the prognostic value of metastatic LN count with regard to histologic subtype in a node-positive cervical cancer population who underwent standard surgical and adjuvant treatment.

However, there are some shortcomings in the current study. The retrospective design makes the study susceptible to selection and referral bias and the absence of central pathologic review limits the data integrity between the participating centers. Small sample size of the non-SCC group is another limitation of our study.

Lymphadenectomy is a component of the surgical management of cervical cancer. Systematic lymphadenectomy is useful for tailoring adjuvant treatment and predicting prognosis [3]. Therefore, the thoroughness of lymphadenectomy is crucial for decreasing the risk of occult lymphatic metastasis. The total LN count was proposed as an indicator of quality control for pelvic LN dissection procedure in the surgical management of cervical cancer [16]. However, the minimum number of LNs for considering lymphadenectomy as adequate is still debatable. European Organization for Research and Treatment of Cancer–Gynecological Cancer Group has indicated that resection of more than 11 pelvic LNs as one of the quality indicators of surgical treatment for cervical cancer [17]. Very recently, European Society of Gynaecological Oncology quality indicators for surgical management of cervical cancer have been published [18]; however, there was no information regarding the number of LNs to be harvested for describing a lymphadenectomy as systematic. In this study, we included only patients with more than 10 pelvic and five para-aortic LNs on their final pathology report in order to maintain a minimum standard for surgical quality.

Adjuvant chemoradiation following radical hysterectomy and systematic lymphadenectomy is currently recommended in women with positive LN, microscopic parametrial involvement, or a positive surgical margin [19]. However, combined treatment with radical hysterectomy and radiotherapy was associated with increased morbidity [20]. The European Society of Oncology/European Society for Radiotherapy and Oncology/European Society of Pathology Guidelines for the Management of Patients with Cervical Cancer recommended intraoperative assessment of LN status with a frozen section [21]. If mLN is detected, they have recommended abandoning radical hysterectomy to avoid morbidity [21]. However, there are only a few retrospective studies that directly compared the outcomes of women in whom radical hysterectomy was aborted or completed in early-stage node-positive cervical cancer [22,23,24,25]. In a Surveillance Epidemiology and End Results (SEER) analysis, Richard et al. [26] included 218 stage IB cervical cancer patients who received pelvic and para-aortic lymphadenectomy and reported positive LN. Among these patients, 163 had undergone radical hysterectomy while radical hysterectomy was abandoned in 55 women. The results of this study revealed that the radical hysterectomy completed and abandoned groups had similar five-year survival (69% vs. 71%, respectively, p = 0.46). Phanedra et al. [27] assessed the value of a routine intraoperative frozen section examination of pelvic LNs in early-stage cervical cancer and reported 86.7% sensitivity and 100% specificity. Despite the success rates, the authors did not suggest a routine frozen section examination of LNs as it was time-consuming, expensive, and not available in all centers.

On the other hand, clinical management of LNs in early-stage cervical cancer varies widely among gynecologic oncologists. A survey study evaluated the practice patterns of Society of Gynecologic Oncology (SGO) members concerning intraoperative management of LNs in early-stage cervical cancer [28]. It has been reported that only 38.6% of the responders preferred routine intraoperative frozen section examination for LNs. Of those, only 21% preferred to cancel radical hysterectomy when an isolated metastatic microscopic LN detected. In another survey study, Gubbala et al. [29] addressed the approach of the British Gynaecological Cancer Society (BGCS) members to intraoperative frozen section examination of LNs in patients with cervical cancer. Only 12.5% of the responders stated that they routinely perform frozen examination. Similarly, in our study, the decision to perform a frozen section examination in an attempt to identify positive nodes was at the discretion of the attending surgeon. The radical hysterectomy was abandoned in the event of metastatic LN. In order to standardize the study population, we did not include women whose radical hysterectomy was canceled as a result of the frozen section analysis.

The LN status was incorporated into the recently revised FIGO staging system in order to emphasize the importance of the nodal metastasis [3]. Patients with metastatic LN either detected by imaging (r) or histopathology (p) were classified as stage IIIC [3]. The 2018 FIGO staging system permits to perform any of CT (computed tomography), MRI, or PET-CT to detect nodal disease [3]. However, both CT and MRI suffer from low sensitivity. In a meta-analysis [30] that evaluated the diagnostic performances of imaging modalities, it was reported that CT had a sensitivity of 57% and a specificity of 91%. MRI showed a sensitivity of %52 and a specificity of 94% [30]. The PET-CT is more accurate than CT or MRI in the detection of mLNs with a sensitivity of 66% and specificity of 97% [30]. Although the diagnostic performance of the imaging modalities is sufficient to spare many patients from an unnecessary lymphadenectomy, histology-proven LN metastasis still seems to be the gold standard in assessing nodal status [31]. Avoiding the double morbidity of radiation and surgery in the case of positive LNs should be the main goal of gynecologist oncologists while managing locally-early stage cervical cancer. It is clear that a Piver III operation followed by chemoradiation carries high morbidity without any survival benefit. However, it should be emphasized that most of the patients in the current study had undergone surgery before the 2018 FIGO staging system was introduced. The low sensitivity of the imaging techniques [30] as well as the lack of stage IIIC (r) in the previous FIGO staging system resulted in inappropriately staged lymph nodes in the pre-operative setting which seems to be the major limitation of our study.

The prognostic significance of mLN count, in different histologic subtypes of cervical cancer has been questioned in a limited number of studies. Liu et al. [10] analyzed the prognostic value of the number of mLNs in 60 pelvic LN positive SCC cases. They found out that the women with positive pelvic nodes ≤ 2 had better 5-year OS rates when compared to women with positive pelvic nodes > 2 (76% and 35%, respectively). However, it must be noted that the cases in this study underwent only pelvic lymphadenectomy. The omission of para-aortic lymphadenectomy might have resulted in underestimation of metastatic nodal disease. Furthermore, the Liu study represented the results of a heterogeneous population because neoadjuvant chemotherapy was administered to 50% of the study population.

Zhou et al. [7] utilized the database of the SEER and analyzed 2222 cervical cancer patients who had pathologically-confirmed positive LNs. The Zhou study revealed that positive-LNs > 2 was significantly associated with decreased OS and cause-specific survival (CSS) in cases who had SCC histology (CSS, p < 0.001; OS, p = 0.001) or adeno-squamous histology (CSS, p = 0.017; OS, p = 0.045). However, the number of positive-LNs was not a prognostic factor among patients with AC subtype (CSS, p = 0.215; OS, p = 0.184). Nevertheless, it should be emphasized that the SEER database has some limitations such as lack of information about the type of hysterectomy, the extent of lymphadenectomy, the use of neo-adjuvant chemotherapy, adjuvant treatment modalities, and the recurrence of disease.

Previous studies on node-positive cervical cancer have indicated poor prognosis for non-SCC histology when compared to SCC histology. Nakanishi et al. [8] reported that non-SCC histology was significantly related with decreased OS (HR = 2.86, 95% CI 1.60–5.09; p < 0.001) and DFS (HR = 2.33, 95% CI 1.34–4.05; p = 0.003) in the presence of LN metastasis. Similarly, Hosaka et al. [32] evaluated 108 patients with node-positive cervical cancer and showed that pure AC histology was an independent prognostic factor for decreased OS (HR = 4.0, 95% CI 0.11–0.55, p = 0.0005). Also, Zhou et al. [1] analyzed the SEER database and reported that AC histology exhibited a worse OS (HR = 1.474, 95% CI 1.114–1.951; p = 0.001) and CSS (HR = 1.614, 95% CI 1.114–1.951; p < 0.001) when compared to SCC histology in node-positive cervical cancer patients. Our study represents the survival outcomes of a uniform study population who underwent standard surgical and adjuvant treatment. According to our findings, regardless of the number of mLNs, the non-SCC group had worse 5-year OS (46.8% vs. 85.3%, respectively; p < 0.001) and 5-year DFS rates (31.6% vs. 82.2%, respectively; p < 0.001) when compared to those of the SCC group with mLNs ≤ 2. However, the non-SCC group and the SCC group with mLNs > 2 had similar 5-year OS (46.8% vs. 65.5%, respectively; p = 0.16) and 5-year DFS rates (31.6% vs. 57.5%, respectively; p = 0.06).

In the present study, we administered concurrent radiochemotherapy to all patients as the adjuvant therapy. However, our results imply that radiochemotherapy seemed to be insufficient for controlling the disease for the SCC group with mLNs > 2 and the non-SCC group. Therefore, new adjuvant treatment strategies other than concurrent radiochemotherapy may be adopted for the management of those patients. Radiotherapy with concurrent doublet chemotherapy based on cisplatin is one of the alternatives. In a meta-analysis, Petrelli et al. [32] demonstrated that concurrent cisplatin-based doublet chemotherapy was associated with significantly enhanced OS (OR 0.65; 95% CI 0.51–0.81; p = 0.0002) and progression-free survival (PFS) (OR 0.71; 95% CI 0.55–0.91; p = 0.006) when compared to weekly cisplatin plus radiotherapy. Concurrent consolidation chemotherapy after chemo-radiation may be another adjuvant treatment option for those patients. In a phase II study, Mabuchi et al. [33] assessed the prognostic value of the consolidation chemotherapy following concurrent carboplatin and paclitaxel plus pelvic radiotherapy in node-positive cervical cancer patients which were surgically treated. The results of the Mabuchi study revealed that this kind of adjuvant treatment was related with better PFS (p = 0.026) when compared to concurrent radiochemotherapy. Additionally, the authors [33] reported that consolidation chemotherapy following concurrent carboplatin and paclitaxel plus pelvic radiotherapy has resulted in better PFS (p = 0.0004) and OS (p = 0.034) when compared to radiotherapy alone.

In conclusion, we demonstrated that patients node-positive cervical cancer patients who had non-SCC histology, as well as those having SCC histology with mLNs > 2, had worse survival outcomes when compared to women who have SCC histology with mLNs ≤ 2.This new stratification might be helpful in predicting the prognosis and determining the choice of adjuvant treatment in patients with stage IIIC (p) cervical cancer. However, our results need to be validated in a prospective fashion.