The proportion of cervical adenocarcinoma in Japan was only a few percent in the 1960s but has gradually been increasing, with adenocarcinoma accounting for 1,791 (22.4%) out of 7983 cervical cancer cases in 2019 (1). Adenocarcinoma has a longer cell cycle than squamous cell carcinoma, a lower percentage of radiosensitive G₂M and G₁ stage cells, and biological characteristics such as a lower oxygen effect ratio in radiotherapy, resulting in resistance to radiotherapy (2). Although there are scattered reports of superior outcomes associated with surgical treatment for cervical adenocarcinoma, there are no randomized controlled trials comparing surgery and radiotherapy, specifically for adenocarcinoma (3-5). Given the radioresistance of adenocarcinoma in stages I-II, radical hysterectomy (RH) is considered reasonable if complete tumor removal is possible. Pelvic lymph node (LN) metastasis is a poor prognostic factor in cervical adenocarcinoma, however, superior radiotherapy results have been reported in LN metastasis-positive cases (4,5). The National Comprehensive Cancer Network Guidelines broadly categorize indications for radical radiation therapy by the presence or absence of enlarged pelvic LNs, rather than by histological type. Concurrent chemoradiotherapy (CCRT) is recommended for patients with enlarged pelvic LNs (6).

In some cases, pelvic LN metastases are found on histopathological evaluation after RH in patients who had no enlarged pelvic LNs on preoperative imaging evaluation. The aim of this study was to clarify whether LN metastasis is also a prognostic factor in these patients, and whether any postoperative therapy is advisable.

From April 2004 to July 2019, 61 patients with stage I-II cervical adenocarcinoma by the 2009 International Federation of Obstetrics and Gynecology classification (7) with no enlarged pelvic LNs on preoperative imaging evaluation who underwent RH at our Institution were retrospectively examined for clinicopathological, treatment, and prognosis-related factors. For preoperative evaluation of pelvic LNs, significant enlargement was defined as a short diameter of 10 mm or more on pelvic magnetic resonance imaging.

For statistical analyses, Kaplan-Meier and log-rank tests were used for determination of overall (OS) and disease-free (DFS) survival, and Cox proportional hazards regression analysis was used for multivariate analysis of prognostic factors.

This retrospective study was conducted according to the principles stated in the 1964 Declaration of Helsinki and subsequent revisions and was approved by the Institutional Review Board of our university in September 2022 (#2000). We used the opt-out method to obtain consent from the patients.

The clinicopathological backgrounds of the 61 eligible patients with cervical adenocarcinoma are shown in Table I. The median age was 47 (range: 28-68) years. International Federation of Obstetrics and Gynecology stages were IB1 in 44 patients, IB2 in 10, IIA in 4, and IIB in 3. Most patients had usual-type endocervical adenocarcinoma (38 cases) or endometrioid carcinoma (12 cases). The median diameter of the cervical tumor was 30 (range=5-62) mm. Two patients (3.2%) had parametrial invasion, 26 patients (42.6%) had lymphovascular space involvement, 28 patients (45.9%) had more than 2/3 depth of cervical stromal invasion, and 17 patients (27.9%) had positive pelvic LNs. The median observation period was 62.8 (7.9-168.1) months.

OS and DFS according to clinicopathological factors are shown in Table II. OS and DFS were significantly poorer in patients with parametrial invasion, lymphovascular invasion, and those with pelvic LN metastasis. Table III shows the multivariate analysis of prognostic factors for OS and DFS; only pelvic LN metastasis was an independent prognostic factor for both of these endpoints.

Additional analysis of the influence of pelvic LN metastasis was performed for the number of metastases, extra-nodal invasion, and the maximum diameter of intra-nodal metastases. Of the 17 patients with positive pelvic LNs, 10 (16.4%) had multiple metastases, 9 (14.8%) had extra-nodal extension, and 8 (13.1%) had intra-nodal metastases greater than 5 mm. In a multivariate analysis of LN-positive cases, the presence of multiple metastases was an independent prognostic factor (p=0.048) for poorer DFS (Table IV), although it did not significantly affect OS (p=0.100).

Based on these results, we examined prognosis for three groups: Patients with negative pelvic LNs (n=44), single metastasis (n=7), and multiple metastases (n=10). The groups showed no significant differences in age, clinical stage, histology, tumor size, and pelvic LN enlargement. Sixteen out of the 17 patients with positive LNs received postoperative adjuvant therapy (chemotherapy in 15 and radiation therapy in one). OS rates were 100%, 83.3%, and 30.0% (p<0.001), and DFS rates were 85.5%, 83.3%, and 12.5% (p<0.001) for the three groups, respectively, showing a significant difference (Figure 1 and Figure 2). The only comparison that failed to reach significance was between the LN-negative and single-metastasis groups in DFS (p=0.390).

Table V shows the recurrence frequency and site in cases with negative LN s, single LN metastasis, and multiple LN metastases. Six (13.6%) out of 44 node-negative patients and two (28.6%) out of seven single node-positive patients experienced recurrence. However, eight (80.0%) recurrences in 10 multiple node-positive patients were observed, with six (75%) being an intrapelvic recurrence (all six patients received postoperative adjuvant chemotherapy). Regarding treatment after recurrence, five cases were treated with radiation alone, after four were dead of disease and only one had no evidence of disease (NED). Another five cases were treated with CCRT using paclitaxel and cisplatin (TP-CCRT) and achieved NED. The remaining two cases were treated with CCRT using cisplatin but died of their disease. Four cases were treated with systemic chemotherapy, resulting in one with NED, one alive with disease, and two died of disease. In the remaining two cases, one underwent surgical removal of the LN recurrence, and the other received best supportive care.

We examined 17 patients with pelvic LN metastases and no enlarged pelvic LNs (i.e., with short diameters of <10 mm on preoperative pelvic magnetic resonance imaging). Positron-emission tomography/CT detects LN metastases of 5 mm or larger (8). Tumors that occupy a high percentage of the excised LNs are more likely to be detected (8,9). Even with the increased accuracy of positron-emission tomography/CT in detecting LN metastases, there are still false-positive and false-negative cases. For these patients, histopathological diagnosis is the most accurate. Our study provides important information about managing these borderline cases.

We examined risk factors for cervical cancer recurrence. We evaluated pelvic LN metastases for the number of metastases, the presence of extra-nodal invasion, and the maximum diameter of metastases. Importantly, only multiple LN metastases emerged as an independent prognostic factor. Regarding the number of LN metastases, Ishikawa et al. reported that 5-year OS rates for 193 cases of cervical adenocarcinoma with 0, 1-2, and ≥3 LN metastases were 89.2%, 61.3%, and 13%, respectively; the 5-year DFS rates were 85%, 51%, and 13%, respectively (10). Multiple LN metastasis was previously identified as a strong risk factor in 243 patients undergoing RH and pelvic lymphadenectomy with positive pelvic lymph nodes, parametrial involvement, or positive surgical margins (11). Although we only examined patients with adenocarcinoma of the cervix, five out of the seven patients with no postoperative recurrence had a single pelvic LN metastasis. This finding suggests that the number of pelvic LNs is important in identifying high-risk groups for postoperative recurrence. OS and DFS rates were significantly different among the three groups. The only comparison that failed to reach significance was between the LN-negative and single metastasis groups in DFS.

Our Institute currently uses chemotherapy (paclitaxel plus carboplatin) as postoperative adjuvant treatment for patients at high risk of recurrence. This approach considers the possibility that radiotherapy after RH may impair long-term quality of life due to adverse effects such as radiation cystitis and colitis (12). In the present study, there was no significant difference in prognosis between LN-negative cases and cases with a single metastasis. Therefore, postoperative chemotherapy improves DFS in cases with a single metastasis. The prognosis was clearly poor in patients with multiple LN metastases, with 70% (7/10) of patients with pelvic recurrence. These results suggest that chemotherapy alone is insufficient for local control, and CCRT should be considered, as Matsuo et al. previously reported (13). TP-CCRT for cervical adenocarcinoma is superior to both distant control by multi-agent chemotherapy and local control by radiation therapy and may improve prognosis in patients with multiple metastases (14,15).

Regarding radiotherapy outcomes after recurrence, Haasbeek et al. reported OS rates of 45% and 28% for pelvic wall recurrence and 93% and 63% for central recurrence at 2 and 5 years, respectively (16). Other reports have shown the usefulness of intensity-modulated radiation therapy and high-dose-rate interstitial brachytherapy; however, these studies were not limited to adenocarcinoma (17,18). Refinement of radiotherapy methods has improved treatment outcomes. Whether radiotherapy should be combined with chemotherapy in patients with pelvic recurrence is debatable (19). CCRT is considered when the intensity of radiation-only treatment is insufficient, such as when recurrent lesions are widely distributed throughout the pelvis, or in cases with a large tumor volume. Our Department does not have fixed criteria and cases are considered individually. In this study, five patients received radiation therapy alone for recurrent disease. Outcomes were four deaths and only one case with NED. Another five cases were treated with CCRT; three cases treated with TP-CCRT achieved NED, and the remaining two cases treated with cisplatin-CCRT died of their disease. TP-CCRT for recurrent cervical adenocarcinoma might be promising and should be considered (14,15).

Among patients with no preoperative enlarged LNs who underwent RH for cervical adenocarcinoma and had positive pelvic LNs confirmed postoperatively, multiple positive pelvic LNs are an independent poor prognostic factor. Because pelvic recurrence is common, CCRT, possibly TP-CCRT, should be considered as adjuvant therapy for such patients.

The Authors have no conflicts of interest to declare regarding this study.

The work presented herein was carried out in collaboration among all Authors. TW and YA made substantial contribution to the conception, designed methods, interpreted the results, and wrote the article. TW, NT, YS, YA, YT, TN, WK, and YA engaged in data acquisition and analyzed the data. TW and YA substantively revised the article. All Authors treated patients, read the article, and approved this submission.

The Authors would like to thank Enago (www.enago.jp) for the English language review of this article.