Comparison of salvage therapies for isolated para-aortic lymph node recurrence in patients with uterine cervical cancer after definitive treatment

Uterine cervical cancer is one of the most common gynecologic cancers in Japan and has been associated with an excellent tumor control rate and favorable prognosis after definitive radiotherapy (RT), radical hysterectomy with pelvic lymph node dissection for early stages, and chemo-radiation therapy (CCRT) for locally advanced stages. However, approximately one-third of patients who are treated for cervical cancer develop recurrence [1, 2]. The most common recurrence sites, other than the pelvis, are the para-aortic lymph nodes [1]. Recurrences of uterine cervical cancer in isolated para-aortic lymph nodes (PALNs) after definitive treatment occur in 1.7 to 12% of patients [3‐7]. Although PALN recurrences are defined as distant metastases, some patients with PALN recurrences have survived for long periods and were considered to be cured as nodal oligo-recurrence. Some reports [5‐7] have shown that concurrent chemo-radiotherapy is an effective salvage treatment for isolated PALN recurrence.

However, no studies have compared multi-treatment modalities, such as radiation therapy (RT), concurrent chemoradiotherapy (CCRT), surgery, chemotherapy, and best supportive care (BSC), across a sufficient number of patients with PALN recurrences. The aim of this study was to evaluate the outcome of patients with isolated PALN recurrence after definitive treatment for primary cervical cancer at our institution.

This study retrospectively examined the data of 50 patients who were first diagnosed with isolated PALN metastases after definitive cervical cancer treatment, from 2002 to 2016, at our institution. Initial definitive treatments were performed with RT alone, CCRT, or surgery with or without post-operative RT (PORT). The median follow-up time was 33 months, and the baseline patient characteristics are summarized in Table 1. The median age was 57 years (range, 26–84 years), the median time to PALN relapse was 10 months (range, 1–91 months), and the median maximum metastatic lesion size was 17 mm (range, 8–60 mm). Forty patients presented with squamous cell carcinoma and 10 with other histopathologies. Eleven patients had initial international federation of Gynecology and Obstetrics (FIGO) stage I disease, 24 had stage II, 10 had stage III, and five had stage IV disease. For PALN recurrence, 18 (36%) patients were treated with RT alone, 17 (34%) with chemotherapy, seven (14%) with CCRT, five (10%) with BSC, and three (6%) with surgery. The patient characteristics of each treatment group are summarized in Additional file 1: Table S1. There were no significant differences between the treatment groups in terms of the assessed variables except age at recurrence. PALN recurrence was defined as the appearance of clinical and radiologic evidence of disease. Radiologically, enlarged PALNs over 1 cm in size on a follow-up computed tomography (CT) scan or/and abnormal uptake on positron emission tomography with [18F]-fluorodeoxyglucose (FDG-PET) scans were defined as recurrences. Twenty-seven patients were diagnosed with FDG-PET scans. The maximum size of the metastatic lesion or the number of metastases were measured and evaluated by CT or FDG-PET CT. Based on the number of metastases, the cases were classified into two groups: 1–2 or ≥ 3. Two tumor markers, serum squamous cell carcinoma antigen (SCC-Ag) and carcinoembryonic antigen (CEA) were measured at the time of PALN recurrence. Elevated serum SCC-Ag (≥1.5 ng/mL) and CEA (≥5.0 ng/mL) levels were considered positive, whereas lower levels were considered negative.

Patients who had any evidence of metastatic disease in addition to PALN recurrence were excluded from this analysis.

The Institutional Review Board (IRB) of our institution approved this study (R-514).

The median follow-up time was 33 months, and the baseline patient characteristics are summarized in Table 1. The median age was 57 years (range, 26–84 years), the median time to PALN relapse was 10 months (range, 1–91 months), and the median maximum metastatic lesion size was 17 mm (range, 8–60 mm). For PALN recurrence, 18 (36%) patients were treated with RT alone, 17 (34%) with chemotherapy, seven (14%) with CCRT, five (10%) with BSC, and three (6%) with surgery.

The 3-year OS rates of all patients were 47.0 and 36.2%, respectively (Fig. 1). With respect to the 3-year OS according to the PALN recurrence treatment method, the 3-year OS of patients who underwent CCRT was 85.7%, surgery, 66.7%; chemotherapy, 48.8%; RT, 41.3%; and BSC, 0% (p = 0.014) (Fig. 2). A clear significant correlation was observed between the PALN recurrence treatment method and overall survival. Of all the potential prognostic factors that we investigated, only the recurrence treatment method was significantly associated with OS in the univariate analysis (Table 2).

With respect to LCR according to the recurrence treatment method, the 3-year LCR was 100% for CCRT and surgery, 33.6% for chemotherapy, and 55.5% for RT (p = 0.025) (Fig. 3). The univariate analysis showed that the number of metastatic lesions (p = 0.001), serum SCC-Ag level (p = 0.012), and recurrence treatment method were significantly associated with LCR (Table 2).

With respect to PFS according to the recurrence treatment method, the 3-year PFS was 71.4% for CCRT, 66.7% for surgery, 13.7% for chemotherapy, and 14.1% for RT (p = 0.059) (Fig. 4). The univariate analysis showed that age at recurrence (p = 0.031), serum SCC-Ag (p = 0.002), and number of metastatic lesions (p = 0.009) were significantly associated with PFS (Table 2).

Salvage treatment modalities for isolated PALN recurrences after primary definitive cervical cancer treatment are equivocal. In the current study, we retrospectively examined a population of 50 patients with isolated PALN recurrence of cervical cancer. Our patient sample size was substantially larger than that of prior studies, and this is the first report to show the superiority of CCRT over other treatment modalities. Concurrent para-aortic irradiation and cisplatin-based chemotherapy generates an excellent prognosis, with a 3-year OS rate of 85.7%, 3-year LCR of 100%, and 3-year PFS of 71.4%. Previously, it has been suggested that RT confers long survival. Niibe et al. [5] reported on 84 patients who were treated with RT for isolated PALN recurrence after curative treatment, from 13 Japanese hospitals. Although the number of patients in the study seemed larger, the chemoradiotherapy group in the the study consisted mostly of sequential, non-CCRT. The 3- and 5-year overall survival rates of all patients were 49.5 and 31.3%, respectively. The authors indicated that irradiation of 51 Gy or greater was associated with better prognoses. Although 32 patients received chemo-radiation therapy in this study, a survival benefit of chemo-radiation therapy over RT alone was not identified. However, the chemo-radiation therapy in this study consisted mostly of sequential, nonconcurrent chemo-radiation therapy.

The reports from Singh et al. [7], Chou et al. [6], and Jeon et al. [8] support our findings. In a series of 14 patients, Singh et al. [7] demonstrated a 5-year OS of 100% in asymptomatic patients with isolated PALN who were treated with CCRT (n = 7), in which radiation doses ranged from 45 to 50.4 Gy. Chou et al. [6] examined 26 patients with isolated PALN recurrence after definitive RT, 14 of whom received CCRT, one received RT, and four received chemotherapy. They found that the 5-year survival rate of patients who received CCRT was 51.2%. Jeon et al. [8] examined 22 patients with PALN (n = 9), supraclavicular lymph node (n = 10), and inguinal lymph node (n = 3) recurrences. Patients who had both a long no-evidence-of-disease period and CCRT showed a 72.9% 5-year PFS rate and a 60% 5-year OS rate. In this study, concurrent chemotherapy was identified as an important factor for achieving a good prognosis. Radiation was delivered to PALN regions, up to 41.4–50.4 Gy, and then a radiation boost was administered to PALN recurrences, up to a total dose of 60 Gy. In our study, we found that patients who were treated with CCRT had better survival outcomes compared to those treated with other modalities. In patients treated with RT, RT was performed for the regional PALN area, up to a median dose of 44 Gy, followed by for the PALN metastases, up to a total median dose of 55 Gy. Regarding the total dose that should be administered to patients, considering results from previous studies [5‐8] and the current study, we recommend administration of more than 50 Gy irradiation to PALN metastases.

With respect to the prognostic factors, serum SCC-Ag levels were significantly associated with LCR (p = 0.012) and PFS (p = 0.002). Jeon et al. [8] also reported that SCC-Ag levels above 8 ng/mL, at the time of recurrence, were associated with good prognoses. Contradictory to findings from other studies [7‐9], our results demonstrated that the time to recurrence after initial therapy completion was not significantly associated with OS, LC, or PFS. Times to recurrence after initial therapy completion of more than 24 months, as reported by Singh et al. [7], and greater than 18 months, as reported by Jeon et al. [8], have been identified as good prognostic factors. Although we found that a median time to recurrence of more than 10 months tended to result in good prognosis, no significant differences were noted for even other cut-off durations, including 18 and 24 months. Kim et al. [9] reported on 12 patients with PALN recurrence after definitive or postoperative radiotherapy who were treated with concurrent chemotherapy and hyperfractionated RT with a total dose of 60 Gy in 50 fractions (1.2 Gy per fraction). They reported a 3-year survival rate of 19% and found more favorable outcomes in patients with a time to recurrence duration greater than 24 months after completion of initial therapy. Due to their inferior results, low dose per fraction may not be recommended.

Stereotactic body radiation therapy (SBRT) may provide a much higher radiation dose to the tumor with a steep dose gradient and reduce the radiation dose to normal tissues. Choi et al. [10] reported on 30 patients with isolated PALN metastases from uterine cervical and corpus cancer who were treated with SBRT (CyberKnife). The 4-year OS rate was 50.1%, and the 4-year LCR was 67.4% in their study, and they concluded that the SBRT results were promising. However, the LCR of CCRT observed in our study was not inferior to that of SBRT. Additionally, we recommend prophylactic regional PALN irradiation, especially for squamous cell carcinoma, because of difficulty in re-irradiation to re-recurrence in PALN. Only patients who are not candidates for long-course RT or have recurrences in previously irradiated fields may be candidates for SBRT.

Although the number of the patients was small in the current study, surgery also showed promise as a treatment, with an OS rate of 66.7% and LCR of 100%. Although surgical treatment is not widely accepted for patients such as ours, surgery has been reported to be feasible for PALN recurrences in colorectal cancer [11, 12]. For young patients and those with recurrences on the edge of the previous initial RT field, surgery might be a feasible treatment for PALN recurrences of cervical cancer. Debulking large nodes followed by PORT may be a good practice if the nodes are bulky and amenable to surgery.

On the basis of the reported results, multidisciplinary management might be a promising treatment for isolated PALN recurrences. If we consider the outcomes reported for the patients undergoing surgery as well as another treatment (most of them underwent adjuvant chemotherapy) and the outcomes for CCRT, there was a higher 3-year survival rate and better local control in the former group, while RT alone and chemotherapy alone was associated with a lower rate of survival (p = 0.036), local control (p = 0.006) and progression free survival (p = 0.007).

The study has some limitations. First, because this study was retrospective and performed at a single institution, careful consideration for possible selection bias is important. Clinical decisions for recurrent PALN treatment were frequently made based on time to recurrence after initial therapy completion, the location, number, and size of the metastatic lesions. For patients with a time to recurrence of less than 6 months and those with recurrence at the margin of the field of initial therapy, we often chose chemotherapy, rather than RT alone or CCRT. For patients with multiple and relatively larger metastatic lesions, we often chose CCRT, rather than RT alone. However, there were no significant differences between the treatment groups in terms of the assessed variables, except age at recurrence. Second, this study involved a relatively small number of patients. The clinical experience of fifty patients is insufficient to conclude that CCRT is superior to the other treatment modalities. However, isolated PALN recurrence is an infrequent occurrence after pelvic definitive therapy, and past studies have included a smaller number of patients. In the management of a relatively rare but curable oligo-recurrence, it is important to suggest this promising treatment strategy. Finally, the diagnoses of recurrent PALN were not histopathologically proven.

In conclusion, although the patient cohort was small and there was selection bias, our results suggest that CCRT may effectively prevent local progression of disease recurrence in association with PALN and improve PFS and OS. Negative SCC-Ag and a small number of metastases may be one of the indicators of curable disease, and such patients may be candidates for salvage CCRT.