Outcomes and prognostic factors for patients with cervical esophageal cancer undergoing definitive radiotherapy or chemoradiotherapy

Cervical esophageal cancer (CEC) is uncommon, accounting for less than 5% of all esophageal cancers. The management of CEC is controversial. This study investigated treatment outcomes and prognostic factors of survival in CEC patients undergoing definitive radiotherapy or concurrent chemoradiotherapy (CCRT). Ninety-one CEC patients were treated by intensity-modulated radiation therapy (IMRT) and three-dimensional conformal radiation therapy (3DCRT) between July 2007 and September 2017. The mean prescription dose was 64 Gy (range 54−70 Gy) delivered as 1.8−2.2 Gy per fraction per day, 5 days a week. Out of 91 patients, 34 received concurrent cisplatin-based chemotherapy (CT) including 18 patients who also received neoadjuvant CT. Overall survival (OS), locoregional failure-free survival (LRFFS), and progression-free survival (PFS) were estimated by the Kaplan–Meier method. Prognostic factors of survival were determined in univariate (log-rank test) and multivariate (Cox proportional hazard model) analysis. Treatment-related toxicity was also assessed. Median follow-up time for all patients was 19 months. Two-year OS, LRFFS and PFS of all patients were 58.2%, 52.5% and 48.1%, respectively. Clinical stage was an independent prognostic factor for OS (HR = 2.35, 95% CI: 1.03−5.37, p = 0.042), LRFFS (HR = 3.84, 95% CI: 1.38−10.69, p = 0.011), and PFS (HR = 2.68, 95% CI: 1.11−6.45, p = 0.028). Hoarseness was an independent prognostic factor for OS (HR = 2.10, 95% CI: 1.05−4.19, p = 0.036). CCRT was independently associated with better LRFFS (HR = 0.33, 95% CI: 0.14−0.79, p = 0.012). 3DCRT and IMRT with concurrent CT is well-tolerated and may improve local tumor control in CEC patients. Advanced clinical stage and hoarseness are adverse prognostic factors for OS, LRFFS, and PFS in CEC.


INTRODUCTION
Cervical esophageal cancer (CEC) is relatively uncommon, accounting for less than 5% of all esophageal cancers [1].The management of CEC is controversial due to the low incidence and a lack of studies investigating specifically treatment strategies and outcomes in CEC.In most available studies, CEC is analyzed together with carcinomas of the hypopharynx and thoracic esophagus, even though it is anatomically distinct from both [2,3].Treatment modalities for CEC include surgical resection with or without neoadjuvant chemoradiotherapy (CRT) and definitive radiotherapy (RT) with or without concurrent chemotherapy (CT) [4][5][6].
Minimally invasive surgical approaches that preserve organ shape and function were the treatment of choice for CEC, however, the risk of complications and morbidity and mortality rates associated with surgical resection of CEC remain high.RT, on the other hand, has a major positive impact on the quality of life (QOL) of patients with CEC, since it allows the preservation of both the larynx and esophagus [7].Therefore, RT has become the preferred treatment for CEC in recent years.Moreover, randomized trials on esophageal cancer [8] and squamous cell head and neck cancer [9,10] showed that neoadjuvant chemotherapy (NAC) and concurrent CRT (CCRT) improve the locoregional tumor control and organ preservation in patients.
Technological advances led to the development of improved radiation delivery techniques such as intensity-modulated radiation therapy (IMRT) and its novel form volumetric-modulated arc therapy (VMAT), which have several advantages in cancer treatment over three-dimensional conformal radiation therapy (3DCRT).Studies comparing the efficacy of 3DCRT and IMRT in the treatment of esophageal cancer show that IMRT provides improved planning target volume coverage and dose conformity as well as a reduced dose to adjacent normal tissues [11][12][13][14][15].
In the current study, we investigated treatment outcomes and prognostic factors of survival in CEC patients undergoing definitive RT or CCRT, with the overall goal to help guide decision making for treatment of CEC.

Patients
A total of 91 patients met the following inclusion criteria: 1) pathological confirmation of cervical esophageal squamous cell carcinoma (CESCC); 2) permission with upper mediastinal lymph node metastasis (M1 lymph node/stage IV esophageal cancer), with no evidence of other distant metastases; 3) completed RT with/without CT; 4) Karnofsky Performance Status (KPS) score ≥70; and 5) age of 75 years or younger.All patients underwent definitive RT or CCRT at our institution from July 2008 to June 2015.Prior to RT, a detailed medical history was obtained from patients and physical examination, barium-swallow X-ray examination, a computed tomography (CT) scan of the neck, chest and abdomen, bronchoscopy, endoscopic ultrasound of the esophagus, and 18F-fluorodeoxyglucose-positron emission tomography (18F-FDG PET) were performed.Tumors were staged according to the 6 th edition of the American Joint Committee on Cancer (AJCC) staging system for esophageal cancer.

Radiotherapy
3DCRT and IMRT, optimized using the Pinnacle treatment planning system (Pinnacle3 version 9.6, Philips Medical Systems, Andover, MA), were applied to all patients.Treatments were delivered using a linear accelerator with dynamic multileaf collimator system (6 MV photon beams) and multiple field technique.
The gross tumor volume (GTV) included the primary tumor and involved regional lymph nodes, determined by multiple imaging examinations.The clinical target volume (CTV) was defined as GTV and additional 0.8-1.0-cmmargins in the radial direction and 3-cm margins in the cranial-caudal direction from the GTV.Elective nodal irradiation, including the area drained by adjacent involved lymph nodes, bilateral levels II-IV of the cervical lymph node area, supraclavicular fossa, and upper mediastinal of the lymph node area [16].The planning target volume (PTV) included the CTV plus a 0.5-cm margin.The organs at risk (OARs) were contoured and comprised the larynx, parotid gland, thyroid gland, trachea, spinal cord, lungs, and heart.

Patient follow-up
Patients were followed up one month after treatment completion, every three months during the first two years, every six months for three to five years, and annually after five years.Toxicity was graded according to the National Cancer Institute Common Terminology Criteria for Adverse Events (NCI CTCAE), version 3.0.Evaluation tools included imaging techniques and biopsy when tumor recurrence was suspected.

Statistics
Statistical analyses were performed using SPSS for Windows, Version 13.0.(SPSS Inc., Chicago, IL, USA).The endpoints were overall survival (OS), locoregional failure-free survival (LRFFS) and progression-free survival (PFS).Each endpoint was calculated from the date of initial diagnosis by biopsy.LRF was defined as local tumor persistence/recurrence, regional lymph node persistence/ recurrence, or death.OS, LRFFS and PFS was defined as the time from initial diagnosis to death from any cause or last follow-up, locoregional tumor persistence/recurrence and evidence of tumor progression, respectively.Locoregional recurrence was defined as recurrence at the primary site or regional lymph nodes.Survival data were analyzed using the Kaplan-Meier method and log-rank test.Univariate analysis of prognostic factors of OS, LRFFS and PFS was performed using the log-rank test and multivariate analysis was carried out using the Cox proportional hazards model.Two-sided tests were used and p <0.05 was considered statistically significant.

Patients
A total of 91 patients with CEC who were treated with RT were included in the study.Thirty-four patients (37.4%) received concurrent cisplatin-based CT, including 18 patients who also received NAC.Fifty-seven patients (62.6%) were treated with RT alone.Table 1 summarizes clinicopathological characteristics and treatment of CEC patients.

Acute and late treatment-related toxicities
Among 91 CEC patients, the most frequently observed acute toxicities were grade 1 and 2. Common grade 1 and 2 acute toxicities were mucositis and esophagitis, observed in 20 (22%) and 22 (24.2%)cases, respectively.Nine patients (10%) experienced grade 3 acute hematologic toxicity.Three patients (3.3%, all three treated with CCRT only) developed grade 3 acute gastrointestinal toxicity.Dysphagia was the most common late toxicity, and six patients (6.6%) experienced severe dysphagia requiring intervention.Other late toxicities were grade 1 radiation pneumonitis and radiation-induced brachial plexus injury.

Survival
The median follow-up time for all patients was 19 months.The two-year OS, LRFFS and PFS for all patients were 58.2%, 52.5% and 48.1%, respectively (Figure 1).

Prognostic factors
Table 2 shows prognostic factors of survival in CEC according to the univariate analysis.Hoarseness, advanced clinical stage (III-IV), tumor length (>5 cm), GTV volume (≥45 cc) and treatment time (>42 days) were significant prognostic factors for poor OS, LRFFS and PFS.GTV dose (<66 Gy), number of fractions (>30 fractions), and 3DCRT technique had an adverse effect on OS.Weight loss (≥10%), distant metastasis (nonregional lymph nodes) and non-CCRT were associated with a worse LRFFS.Lymph node metastasis (N1) was associated with a worse PFS (Figure 2).Table 3 summarizes the multivariate analysis results.

DISCUSSION
The management of CEC remains controversial.Prospective randomized data for CEC are lacking, due to its low incidence and a small number of clinical studies specifically investigating treatment outcomes and prognostic factors of survival in CEC patients.RT has become the primary treatment option for CEC, due to the fact that it allows preservation of the esophagus and adjacent organs.Nevertheless, studies investigating outcomes of CEC patients treated with  RT, CCRT, or surgery showed controversial results.For example, the two-and five-year survival rates of patients treated with CRT are 24% to 60% and 20% to 55%, respectively [5,[17][18][19][20], while the five-year survival rates of CEC patients undergoing surgical resection with larynx preservation are higher, i.e. between 40.6% and 57.8% (Table 4) [6,[21][22][23].Moreover, 23.8% to 40.7% of CEC patients treated with CRT undergo salvage surgery [17][18][19][20].On the other hand, other studies showed comparable disease control and survival between surgery with/ without larynx preservation and RT with/without concurrent CT in CEC patients [7,20,24,25].Chou et al. [7] retrospectively analyzed 15 patients who underwent radical resection (total laryngopharyngectomy with neck dissection, total esophagectomy, and reconstruction with stomach) and 14 patients who received CCRT.They showed no significant difference in the QOL and survival between the two treatment groups (mean survival time was 36.2 months for surgical resection vs. 34.4months for CCRT [p = 0.97]) [7].In a matched-case analysis of 58 patients with CEC, Cao et al. [24] did not show any significant differences in two-year survival rates between surgery group (in most cases pharyngolaryngoesophagectomy [ and OS (17.1 vs. 14.4 months, p = 0.943) between CCRT and PLE groups.Among the above-described studies three performed PLE, thus increasing the possibility of a better therapeutic effect due to radical surgical approach.Furthermore, some studies reported that PLE is associated with high mortality rate and risk of complications in CEC patients [20,[26][27][28].Generally, a direct comparison between RT and surgery in CEC is difficult due to the retrospective design and inherent selection bias of these studies.In our study, the two-year OS, LRFFS and PFS of CEC patients treated with RT were 58.2%, 52.5% and 48.1%, respectively and only mild complications were recorded for these patients.Similar findings were reported by other studies [4,5].Thus, CCRT appears to be optimal treatment strategy for CEC and should be considered individually.Randomized studies on esophageal cancer [8] and squamous cell head and neck cancer [9,10] showed that CCRT results in organ preservation and improves patient OS compared with RT alone.A rationale for CCRT is that CT can sensitize tumors to RT by preferentially killing hypoxic cells, inhibiting tumor repopulation, inhibiting the sublethal radiation damage repair, and by improving blood supply and reoxygenation of organ [29].However, in CEC, local and regional failure rates after CCRT remain high [1,29,30].Studies with adequate follow-up reported a local relapse rate between 34% and 85% in CEC patients treated with CCRT [4,5,17,18,31,32] suggesting that a more aggressive local approach, such as the use of NAC or higher radiation doses, may be helpful.Our study showed that CCRT improved LRFFS in CEC patients.Several other studies on CEC patients reported a weak positive trend between OS or LRFFS and CCRT, although without significant difference compared to RT alone.Overall, CCRT may improve local tumor control in CEC patients [17,31,33].In addition, Huang et al. [17] showed no significant difference in OS (p = 0.94) and LRFFS (p = 0.19) between patients treated with a lower dose, hypofractionated, 2D RT with 5-FU-based CT protocol and those treated with high-dose cisplatin-based, conventionally fractionated, conformal CCRT with prophylactic nodal RT.
Similar to previous studies [4,16,34], we showed that hoarseness was a significant prognostic factor for OS in CEC patients, i.e., it was associated with advanced clinical stage and disease progression.Hoarseness results from recurrent laryngeal nerve involvement due to direct tumor invasion or lymph node metastasis.Therefore, hoarseness can be considered as a late symptom of CEC and should be taken into account when planning the treatment.
Due to the low incidence of CEC, relatively short follow-up and patient heterogeneity, only a few studies have investigated the effect of advanced stage on survival in CEC.In the studies of Ludmir et al. [35] and Cao et al. [4] advanced stage had a significant impact on LRFFS and PFS in patients treated with definite RT or CCRT.In our study advanced clinical stage, together with hoarseness, was an adverse prognostic factor for OS, LRFFS and PFS.I.e., CEC patients with clinical stage I-II had a better OS (71.4% vs. 40.0%,p = 0.000), LRFFS (76.5% vs. 36.6%,p = 0.000) and PFS (72.0% vs. 32.9%,p = 0.000) compared to patients with stage III-IV.Moreover, CEC patients with advanced clinical stage with N1 and M1 (nonregional lymph node metastasis) tended to have a worse OS, LRFFS and PFS in our study.
The major limitations of our study are retrospective design, small number of included patients, selection bias, and heterogeneity among patients.Larger prospective randomized studies are necessary for better insight into the effect of CCRT on CEC.

CONCLUSION
In summary, 3DCRT and IMRT with concurrent CT is well-tolerated and may improve local tumor control in CEC patients.Advanced clinical stage and hoarseness are adverse prognostic factors for OS, LRFFS, and PFS in CEC.

FIGURE 2 .
FIGURE 2. Hoarseness (A) and clinical stage (B) were independent prognostic factors of OS in CEC.Clinical stage (C) and CCRT (D) were independent prognostic factors of LRFFS.CEC: Cervical esophageal cancer; OS: Overall survival; LRFFS: Locoregional failure-free survival.

TABLE 1 .
Demographic and clinicopathological characteristics of patients

TABLE 2 .
Univariate analysis of two-year OS, LRFFS and PFS in CEC patients