Lymph node enlargement after definitive chemoradiotherapy for clinical stage I esophageal squamous cell carcinoma

This retrospective study used the database at the Department of Gastroenterology and Hepatology of Osaka University Hospital. From February 2006 to August 2011, 55 consecutive patients with stage I (T1 [submucosal cancer] N0 M0) ESCC who were treated with CRT were analysed. TNM staging was determined according to the Union for International Cancer Control criteria. None of the patients chose the surgical treatment. The median patient age was 66 years (range, 49–82 years). The clinical stage was diagnosed by endoscopy, endoscopic ultrasonography, cervical and thoraco-abdominal CT. Tumor localization was identified by combining chromoendoscopy with Lugol staining. Tumor invasion depth was evaluated using magnification endoscopy with narrow band imaging (NBI) and endoscopic ultrasonography in addition to conventional CT. The CRT was performed as following criteria: 1) Eastern Cooperative Oncology Group performance status of 0–2; 2) leucocyte count >3,000/m³, platelet count >100,000/m³; 3) aspartate aminotransferase or alanine transferase level within 3 times the normal upper limit; 4) creatinine level <1.5 mg/dl and creatinine clearance >50 ml/min; and 5) no other serious complications. All patients had adequate hepatic and renal functions and performance status scores of 0. The tumor histological type was diagnosed as squamous cell carcinoma based on an endoscopic biopsy. This study was approved by the Institutional Review Board of Osaka University.

All patients were treated with cisplatin and 5-FU chemotherapy. Cisplatin was administered at a dose of 70 mg/m² body surface area on Day 1 and Day 29, and 5-FU was administered at a dose of 700 mg/m² per day by continuous infusion for 24 hours on Days 1–5 and Days 29–33. Nedaplatin was administered instead of cisplatin to 3 patients on Day 29 because of renal dysfunction induced by cisplatin. The concurrent radiotherapy consisted of external administrations of 2 Gy daily to a total dose of 60 Gy without a planned break. The gross tumor volume was limited to the primary tumor. The planning target volume was defined by adding 2- to 3-cm margins above and below the tumor without prophylactic lymph node coverage according to the clips marked during the endoscopic procedure. The lateral, anterior, and posterior margins were limited to 1–2 cm. The radiation therapy field comprised the planning target volume for up to 40 Gy with anterior/posterior opposed portals and exposed to an additional 20 Gy with bilateral oblique portals excluding the spinal cord.

After CRT, the clinical response was assessed with endoscopic observation accompanied by biopsy specimens and a thoraco-abdominal multidetector CT scan. CR was defined when a tumor was not detected by endoscopic observation and CT scan for >4 weeks. First, the initial evaluation was performed 1 month after CRT and subsequently followed up at 3-month intervals up to 1 year and subsequently at 6 months and up to 5 years by endoscopy and thoraco-abdominal CT scan or until recurrence was diagnosed. Metachronous esophageal recurrence out of the radiation field was not included in the present study. A lymph node enlargement was defined when the size of a lymph node was larger than the initial size before CRT or newly detected by CT scan. The early cases were diagnosed as a recurrence or a benign enlargement pathologically by surgical resection. In the later cases, the lymph node enlargement was followed by CT scan with closed interval. During the closed follow-up, the enlarged lymph nodes were diagnosed as a recurrence when they increased in size and number in the CT scan. When the exacerbation was not detected for more than 6 months, lymph node enlargement was defined as “benign”.

Adverse events were retrospectively evaluated using the Common Terminology Criteria version 4.0. We evaluated therapeutic late toxicity according to the Radiation Therapy Oncology Group/European Organization for Research and Treatment of Cancer late radiation morbidity scoring schema (available at http://​www.​rtog.​org/​).

If there were missing cases, the date of the last observation was defined as the censor date. The overall survival was calculated using the Kaplan-Meier method from the date of CRT initiation to death from any cause. Cumulative progression-free survival was calculated using the Kaplan-Meier method from the date of CRT initiation to the recurrence or death from any cause. A Wilcoxon test was performed to compare diameters and intervals until lymph node enlargement between benign enlargement and recurrence. All analyses were performed using JMP software (V.10.0.2, SAS Institute Inc., Cary, NC, USA).

The median follow-up period after initiating CRT was 46 months (range, 2–83 months). The main clinical characteristics of the patients are presented in Table 1. Before the treatment, 5 patients had been treated for malignancies in other organs, which were inactive when CRT was performed for ESCC. During the follow-up period, the second primary cancers in the other organs were observed in 10 patients: pancreas, skin, and prostate, 1; biliary duct, 1; urinary tract, 1; pharynx, 1; colon, 1; lung and stomach, 1; vocal cord and stomach, 1; valve, 1; thyroid, 1; and stomach, 1. Esophageal intraepithelial neoplasia, which occurred metachronously out of the radiation therapy field, was detected in 3 patients after CRT; all of whom were treated with ER.

All but 2 patients successfully underwent total treatment, and the treatment completion rate was 96%. One of the 2 patients received only 52 Gy irradiation because he withdrew further treatment, and the other patient received an 80% dose of chemotherapy on Days 29–33 because of digestive toxicity. One patient died 2 weeks after CRT completion. The causal correlation with the treatment was not confirmed; however, treatment-related death was suspected.

With regard to hematologic adverse events, grade 3 or 4 leukocytopenia was observed in 8 patients, and grade 3 anaemia was observed in 1 patient. Grade 3 appetite loss and nausea were seen in 3 patients, and grade 3 mucositis was observed in 3 patients. Late toxicity events were observed 3 to 24 months after the treatment. Grade 3 interstitial lung disease was observed in 1 patient (1.8%). Grade 3 pericardial effusions were observed in 3 patients (5.5%). Two of these patients had been treated for cardiac disease before CRT.

The 3-year overall survival rate in all patients was 90.7% (Figure 1a). The CR rate was 87.2% (48/55). Although 6 non-CR patients, except for 1 patient who was suspected of treatment-related death, underwent salvage treatment, including surgery and chemotherapy, 2 died because of lymph node recurrence and liver metastasis (Table 2). Of the 48 CR patients, 5 died from lymph node recurrences, and 4 patients died without ESCC recurrences during the follow-up period: urinary tract cancer, 1; biliary tract cancer, 1; sudden death of unknown cause, 1; and acute heart failure, 1.The 3-year progression-free survival rate was 71.2% (Figure 1b). Of the 48 CR patients, recurrences were detected in 9 patients from 8 to 41 months after CRT initiation. The local recurrences of 2 patients were treated completely with ER. Lymph node recurrences were found in 7 patients who underwent salvage surgery or chemotherapy. Of the limited 48 CR patients, the 3-year progression-free survival rate was 81.5%.

We performed a scheduled thoraco-abdominal CT before and after the treatment according to the protocol for detecting metastatic recurrence. Although no enlarged lymph node was detected by CT before CRT, the enlarged lymph nodes were detected in 20 (36.3%) of the 55 patients during the scheduled follow-up (Table 3). Initially, the salvage surgery including lymphadenectomy for lymph node enlargements were performed in 4 patients. Two patients who underwent surgery were pathologically diagnosed with benign enlarged lymph nodes. Then, we monitored patients with newly enlarged lymph nodes using CT to determine whether the nodes were metastatic recurrences. Among the 20 patients in whom enlarged lymph nodes were detected after CRT, 7 patients were diagnosed with lymph node recurrences because of the additional growth of lymph nodes within 6 months or surgical pathological examinations. The size of the lymph nodes of 13 patients did not change within 6 months, and they satisfied the “benign” definition. However, in 2 patients, the lymph nodes (patients 3 and 6) grew slowly and were finally diagnosed as metastatic recurrences more than 1 year after their initial detections. Consequently, 9 of these 20 patients (45%) were diagnosed with metastasis, and 11 patients (55%) were considered to have reactive enlargement of lymph nodes. The average interval from initiating CRT to the point of lymph node enlargement was 14.1 months in the patients with recurrent metastasis and 12.0 months in the patients with benign lymph node enlargement. The average lymph node diameters were 9.3 mm in the patients with recurrent metastasis and 8.3 mm in the patients with benign lymph node enlargement. We did not find significant differences between the recurrence and the benign enlargements in the interval after CRT or lymph node size. The benign lymph node enlargements that were diagnosed were followed for >19 months, except for 1 patient who was followed for 8 months, which suggests that these patients had benign lymph node enlargements. Regarding the benign lymph node enlargements, FDG accumulation on PET-CT was positive in one patient and negative in 5 patients, and 5 patients did not undergo PET-CT. All of enlarged lymph nodes diagnosed as metastatic recurrence finally showed FDG accumulation. Regarding the alteration of lymph node diameter after detecting lymph node enlargement, we observed that several nodes shrank promptly, whereas others maintained the same size during the follow-up period. Half of the lymph node enlargements that occurred after CRT were eventually diagnosed as false-positive, which means that they could avoid unnecessary treatment.