Three-dimensional conformal radiotherapy with concurrent chemotherapy for postoperative recurrence of esophageal squamous cell carcinoma: clinical efficacy and failure pattern

Esophageal cancer (EC) remains one of the most fatal malignancies in the world. In 2005, about 497,700 new cases occurred worldwide, and the prevalence is expected to increase by approximately 140% by 2025[1]. Unlike western countries, in China the predominant histological subtype of EC is squamous cell carcinoma (ESCC), and tumors are more likely to develop in the middle and upper thoracic esophagus[2, 3]. Surgical resection is the primary treatment for thoracic ESCC in many cancer institutes as it offers a chance of cure. Although the 5-year overall survival rates of patients who underwent curative tumor resection range from 31 to 55%[4, 5], postoperative recurrence remains the major type of failure. The recurrence rate of surgical patients ranges from 36 to 56% and the median time to recurrence ranges from 10 to 12 months; while anastomosis, regional (mediastinum and upper abdomen) lymph node and supraclavicular lymph node are the most common recurrence sites[6‐8]. Significant difficulty is often encountered in determining treatment options for recurrent disease after esophagectomy, and patient prognosis is generally poor[9‐11].

Although optimal treatment for patients with postoperative recurrence of ESCC has remained controversial, recent advances in anticancer drug and radiation techniques may help to improve treatment outcomes. Since these patients hadn’t received radiotherapy (RT) or chemotherapy before, RT combined with concurrent chemotherapy might have a beneficial symptomatic effect and a possibility to obtain long-term survival[12‐16]. Thus, the factors affecting this survival after postoperative recurrence in ESCC patients need to be fully explored. In our study, we evaluated the prognostic factors and treatment failure pattern of concurrent chemoradiotherapy (CCRT) for postoperative recurrence of ESCC.

CCRT has been established as a curative alternative for the treatment of ESCC, conferring substantial improvement in survival compared with RT alone. However, the optimal treatment of recurrent ESCC after surgery remained controversial. Some studies on the effectiveness of radiotherapy with or without chemotherapy for treatment of postoperative recurrent ESCC reported 2-year survival rates of only 15-31% with short-term observation. Lu et al. reported that the OS of all 73 recurrent ESCC patients was 46.7% and 4.7% at 1 and 3 years respectively, and patients receiving CCRT had better OS than those receiving RT alone with 1-year OS 62.5% in CCRT group vs. 33.8% in RT alone group, 3-year OS 10.5% in CCRT group vs. 0% in RT alone group[12]. A recent phase II study on radiotherapy combined with nedaplatin and 5-Fu for postoperative loco-regional recurrent ESCC showed improved 5-year OS of 27.0%. High objective response rate (76.7%) for recurrent tumors was obtained in the study, suggesting a potential benefit of concurrent chemotherapy[23].

In our study, the ESCC patients with localized recurrence treated by CCRT showed a promising improvement of prognosis compared to the previous reports in which the 3-year OS rate of all the patients was 51.8%, the median OS time was 43.0 months, and the overall tumor response rate was 75.9% (63/83). Similar to primary localized ESCC, prolonged survival is unlikely to be achieved in patients who had postoperative recurrent ESCC unless they have obtained a good local response to initial chemoradiotherapy. In our study, an improvement in survival was observed in patients who obtained CR after CCRT, with the 3-year OS of CR and non-CR patients being 75.7% and 35.6% (p = 0.000). Recent studies of preoperative CCRT followed by surgery showed that the excellent prognosis were associated with the patients who achieve a complete pathological response to preoperative CCRT[24‐26]. Although clinical CR after CCRT may not be significantly correlated to pathologic CR[27], maximizing the CR rate is likely to increase the proportion of patients with most favorable outcome and potentially affect survival of the group as a whole. In our study, it showed that in the CR group (37 patients), 7 patients developed in-field local recurrence and 6 patients had distant metastases; while in the non-CR group (46 patients) 22 had local recurrence and 14 had distant metastases. These results suggested that the survival benefit in CR group mainly came from the decrease of the rate of local recurrence. Moreover, local recurrence might have a relationship with the occurrence of distant metastasis.

Our results also demonstrated a benefit in survival for patients treated with a regimen of docetaxel-based concurrent chemotherapy. The 3 year OS rate was 59.2% for patients received concurrent TP chemotherapy and 43.3% for concurrent PF. The Swiss Group for Clinical Cancer Research (SAKK) established a preoperative induction schedule comprising docetaxel and cisplatin followed by CCRT[28]. This schedule proved to be effective and feasible, resulting in a 3-year survival rate of 53%. Indeed, taxanes promote tubulin conjugation and stabilize microtubule formation, thereby inhibiting mitosis. In addition to their cytotoxic effect, taxanes also act as excellent radiosensitizers, arresting the cell cycle in the G2/M phase[29]. Multiple studies have examined the use of docetaxel and platinum compounds as radiosensitizers for esophageal cancer and neoadjuvant chemotherapy, reporting high pathological CR rates[30, 31]. Zanoni et al. reported that the results of neoadjuvant concurrent radiotherapy with docetaxel, cisplatin, and 5-Fu was a promising regimen, with a 41.9% pathological CR rate and a good safety profile[32]. Encouraged by few recent studies suggesting the efficacy of taxanes in esophageal cancer, investigators performed a phase II trial to evaluate the feasibility and safety of docetaxel and cisplatin in 5-FU and cisplatin-pretreated esophageal cancer. A total of 38 patients were enrolled, and 35 patients were available for evaluation. The median and total numbers of cycles delivered were 3.5 (range, 1–9 cycles) and 162, respectively. One patient (2.6%) achieved complete response, 12 (31.6%) achieved partial response, 12 (31.6%) had stable disease, and 10 (26.3%) had progressive disease. The overall response rate was 34.2% (95% confidence interval, 19.6-51.3). The median progression-free survival and overall survival times were 4.5 ± 1.3 months (95% CI, 4.1-4.9) and 7.4 ± 0.4 months (95% CI, 7.3-7.5), respectively. The best results were observed in patients who had good performance status at baseline and who had longer treatment-free intervals after first-line chemotherapy[33]. In our study, 10 out of 83 patients (12.0%) had ECOG 2–3. Among them 6 patients received 2 cycles of cisplatin and 5-fluorouracil concurrently with RT, and 4 patients received concomitant weekly docetaxel and cisplatin for 4–6 weeks. This suggested that performance status at baseline is an important issue in choosing a concurrent chemotherapy regimen.

Another interesting finding in our study was that AR patients had a better OS compared with LR patients (3-year rate 74.4% vs. 43.6%, P = 0.028). Among the 20 AR patients, 11 patients had recurrent lesion only at anastomosis, 7 patients had anastomotic recurrence and one station lymph node involvement, and 2 patients had anastomotic recurrence and more than one stations lymph node involvement. After CCRT, 1 patient developed in-field local recurrence and 5 had distant metastases in the 20 AR patients; while 26 had local recurrence and 15 had distant metastases among the 63 LR patients. The recurrent lesion in AR patients seemed to be more localized and had less lymph node involvements. This might because the anastomotic recurrence is more likely to cause dysphagia and be diagnosed earlier. Other than that, anastomotic recurrence could be regarded as a second primary cancer; the prognosis would be different from extensive lymph node involvement.

The higher treatment response and local control rates in this study compared with other published data might be due to: 1) the different treatment target definition (the CTV comprised the anastomosis, supraclavicular, and station 1–5 and 7 lymph nodes in order to involve all the subclinical metastatic regions); 2) the use of 3D-CRT technique (3D-CRT had better GTV high dose coverage than two-dimensional conventional radiotherapy (2DCRT) which had been used by most previous studies); 3) the higher irradiation dose (median dose in present study was 60 Gy, range from 56 to 68 Gy. Zhang’s study suggested that higher dose of 60 Gy could achieve better local control)[34]. 4) 44 out of 83 (53.0%) patients receive docetaxel-based concurrent chemotherapy. This retrospective study has several limitations such as selection bias, various chemotherapy regimens, small study numbers and short follow-up time. Moreover, instrumental diagnostic procedures such as endoscopic ultrasound (EUS) and positron emission tomography (PET) were not used for a precise CR evaluation of the recurrent lesions after CCRT. However, our results revealed that ESCC patients with postoperative localized recurrent disease could achieve promising improvement in outcome when treated with CCRT and they have the potential to be cured. Patients could benefit from proper target definition, 3D-CRT technique and aggressive concurrent chemotherapy.

3D-CRT-based CCRT is effective and well-tolerated in patients with recurrent ESCC; patients obtaining complete remission after CCRT or receiving concurrent TP chemotherapy appeared to achieve long-term OS. Although local and distant failures remained high in the whole group, patients with localized disease could still benefit from aggressive CCRT. Prospective studies are needed to validate these predictive factors.