Radiation Therapy for Recurrent Esophageal Cancer after Surgery: Clinical Results and Prognostic Factors

Transcription

1 Radiation Therapy for Recurrent Esophageal Cancer after Surgery: Clinical Results and Prognostic Factors Yoshiyuki Shioyama 1, Katsumasa Nakamura 1, Saiji Ohga 1, Satoshi Nomoto 1, Tomonari Sasaki 1, Toshihiro Yamaguchi 1, Takashi Toba 1, Tadamasa Yoshitake 1, Hiromi Terashima 2 and Hiroshi Honda 1 1 Department of Clinical Radiology, Graduate School of Medical Sciences and 2 Department of Radiation Technology, School of Health Sciences, Faculty of Medicine, Kyushu University, Fukuoka , Japan Received September 19, 2006; accepted June 27, 2007 Jpn J Clin Oncol 2007;37(12) doi: /jjco/hym138 Objective: To evaluate the outcome of radiotherapy for recurrent esophageal cancer after surgery and to determine the prognostic factors. Methods: From 1987 through 2002, 82 patients treated with radiotherapy for loco-regional recurrences of esophageal cancer after surgery were retrospectively reviewed. The stage at initial surgery was I in 16, II in 41, III or higher in 24 and unknown in 1. The median size of recurrent tumors was 3.5 cm in diameter. Fifty-two patients were treated with radiotherapy alone, and 30 were treated with radiotherapy combined with chemotherapy. The median total dose of external radiotherapy given was 50.4 Gy in 28 fractions. Results: The median survival period after recurrence was 7.0 months. The 2- and 5-year overall survival rate for all patients was 22 and 11%, respectively. In univariate analysis, the patients with performance status (PS) ¼ 0 1, or tumor size,3.5 cm, and those treated with total dose 50 Gy showed a better survival outcome than each the other groups. The patients with a history of previous radiotherapy showed a poorer survival outcome in univariate analysis than each the other groups. In multivariate analysis, tumor size, PS and radiation dose were independent prognostic factors for overall survival. Conclusion: The prognosis of patients with post-operative loco-regional recurrence of esophageal cancer is poor. However, a long-term survival may be expected by definitive radiotherapy for the patients with small-size tumors and with a good PS. Key words: esophageal cancer recurrence after surgery radiation therapy INTRODUCTION A surgical resection has been the primary treatment choice for patients with esophageal cancer. However, the 5-year survival rate remains.40% (1). Loco-regional recurrence remains the major type of failure in the patients treated with surgery for esophageal cancer (2 4). Radiation therapy or a combination of radiation and chemotherapy has been used for the treatments for recurrent esophageal cancer after surgery. Radiotherapy with or without chemotherapy has been reported to have a beneficial symptomatic effect for a significant proportion of these patients and to have a possibility to obtain a long-term survival effect in some patients (5 7). In patients considered to have a favorable prognosis, radical treatment with radiotherapy or chemo-radiotherapy For reprints and all correspondence: Yoshiyuki Shioyama, Department of Clinical Radiology, Graduate School of Medical Sciences, Kyushu University, Maidashi, Higashi-ku, Fukuoka City, Fukuoka , Japan. yshioyama@yahoo.com seems to be beneficial for prolonging the survival time. For the patients with a poor prognosis, on the other hand, radiation therapy should be considered as one of palliation methods for symptom relief rather than as a radical treatment option. However, the role of radiotherapy and the prognostic factors for a postoperative recurrence of esophageal cancer patients have not yet been clearly elucidated. In this study, we evaluated the treatment outcome of radiation therapy for patients with post-operative recurrences of esophageal cancer treated from 1987 through 2002 in our hospital in order to determine the prognostic factors and optimal approach of radiotherapy for this disease. PATIENTS AND METHODS PATIENTS From 1987 through 2002, 96 patients with loco-regional recurrences of esophageal cancer after surgery were treated # The Author (2008). Published by Oxford University Press. All rights reserved.

2 Jpn J Clin Oncol 2007;37(12) 919 with radiation therapy at Kyushu University Hospital. Of these, 82 patients had loco-regional recurrences alone, whereas 14 patients had distant metastases in addition to loco-regional recurrences. The 14 patients who had both loco-regional recurrence and distant metastases were excluded from this analysis. In this study, loco-regional recurrences were defined as anastomotic recurrences or lymph node metastases including supraclavicular and abdominal lymph node metastases. The patient characteristics are summarized in Table 1. Seventy patients were male and 12 were female. The median age of these patients was 61-year-old. For the initial stage of esophageal cancer, 16 patients were stage I, 41 were stage II and 24 were stage III or more (UICC, 1997) (8). The initial stages of the remaining one patient were unknown. The location of the primary tumors was cervical esophagus in 3, upper thoracic esophagus in 14, middle thoracic esophagus in 48, lower thoracic esophagus in 16, and abdominal esophagus in 1. Thirty-six patients received pre-operative radiotherapy of Gy in initial treatments. All patients had squamous cell carcinoma. The characteristics of recurrent tumors are summarized in Table 2. The site of loco-regional recurrence was the supraclavicular region in 24, the mediastinal region in 36, the upper abdominal region in 10 and multiple regions in 12. The maximum tumor size of recurrent tumors was 3.5 cm in diameter in median (range cm). Regarding the number of recurrent tumors, 50 patients had solitary tumors, whereas 32 patients had multiple tumors. Most of the patients (72/82, 88%) had lymph-node metastases, whereas the remaining 10 patients had only local recurrence near the site of anastomosis. The median interval of recurrence from initial surgery was 10 months (range Table 1. Patients characteristics No. of patients 82 Median age (range) 61 years (43 86 years) Gender (male:female) 70:12 Performance status Histologic type Squamous cell ca. 82 Initial stage I 16 II 41 III IV 24 Unknown 1 Median time to recurrence from surgery 10 months ( months) Previous radiotherapy Yes 36 No 46 Table 2. Characteristics of loco-regional recurrence Number of tumors Solitary 50 Multiple 32 Sites Supraclavicular region 24 Mediastinal region 36 Abdominal region 10 Multi-region 12 Pattern of recurrences Local 10 Lymph-node 72 Median maximum diameter (range) 3.5 cm ( cm) months). The diagnostic work-up for the detection of recurrence included a physical examination and a radiographic evaluation, including computed tomography (CT), esophagogram, and/or endoscopy. TREATMENT The treatments given to the patients are summarized in Table 3. External radiation therapy was given with 6 or 10 MV X-ray of a linear accelerator. The median fraction size and total doses of external radiation therapy given were 1.8 Gy (range: Gy) and 50.4 Gy (range: Gy), respectively. All cases were planned using an X-ray simulator. In most of patients (67/82, 82%), a local field covering recurrent tumors with a margin of 2 3 cm in both the oral and anal directions was used. In the remaining 15 patients, T-shaped field including the bilateral supraclavicular and mediastinal regions was used. After irradiation of Gy, the beam direction was changed to avoid the spinal cord. In the patients who had received pre-operative irradiation to the corresponding regions, before a cumulative dose of 50 Gy, the radiation field carefully avoided the spinal cord. The type of radiation field and the irradiation dose and the use of combined chemotherapy were all determined by a radiation oncologist based on various factors including the patient s age, performance status (PS) and tumor extent. Chemotherapy was combined with radiotherapy in the 30 patients younger than 75 years old and having PS 0 1 among the patients who were treated after Of these, 24 patients received a concurrent combination of chemotherapy and radiotherapy. Low-dose cisplatin (CDDP, 5 mg/m 2 )with protracted venous infusion of 5-fluorouracil (5-FU, mg/m 2 ) was combined with daily irradiation in 17 patients. Drip infusion of 30 mg carboplatin and bolus injection of 5-FU (250 mg/day) were administered before daily irradiation in five patients and two patients, respectively. In

3 920 Radiotherapy for recurrent esophageal cancer Table 3. Treatment characteristics Median total dose (range) 50.4 Gy ( Gy) Median fractions 30 (15 47) Median treatment time (range) 44 days ( days) Combination with chemotherapy No 52 Yes 30 Chemotherapy regimens Concurrent Low-dose CDDP þ 5-FU 17 CBDCA 5 5-FU 2 Sequential CDGP þ 5-FU 5 Tegafur 1 CDDP, cisplatin; CDGP, nedaplatin; CBDCA, carboplatin; 5-FU, 5-fluorouracil. five patients, two cycles of systemic chemotherapy with 5-FU 500 mg/m 2 on Days 1 5 with nedaplatin (CDGP) 80 mg/m 2 Day 6 was administered prior to radiotherapy. Oral administration of 600 mg tegafur was used in one patient after completion of radiotherapy. METHODS OF EVALUATION AND STATISTICAL ANALYSIS After the completion of therapy, the patients were followed at 1- or 3-month intervals. Follow-up evaluations included a history and physical examination, esophagograms and/or CT. Those who did not appear for a routine follow-up were contacted either by phone or by mail. The overall survival was defined as the interval between the start of treatment and the most recent follow-up or death. Cause-specific survival was calculated by taking a census of patients who died without any evidence of recurrence. The survival and local control rates were calculated using the Kaplan Meier methods (9) and the statistical significance of differences between survival curves by a log-rank test. The differences in the survival in univariate analysis was assessed by the log-rank test. In a multivariate analysis, Cox s regression analysis was used. the esophageal cancer, and three died of other diseases (cholangiocarcinoma in one, pneumonia in one and chronic heart failure in one). Twenty-eight patients were alive at the last follow-up, but two of these had a recurrence of cancer. The overall survival and cause-specific survival curves for all 82 patients are shown in Fig. 1. The 2- and 5-year overall survival rate for all patients was 22 and 11%, respectively. The 2- and 5-year cause-specific survival rate was 28 and 14%, respectively. Ten of 82 patients showed a long-term survival of longer than 2 years. All of these 10 patients were treated with 50 Gy or more doses and had PS 0-1 without distant metastases or history of previous radiotherapy. There was no patient with any serious complication directly related with radiotherapy in our present study. Several patients had radiation pneumonitis at 1 6 months after the completion of radiotherapy. However, there was no lung toxicity that required steroid administration. Although one patient died of chronic heart failure 20 months after the end of radiotherapy total dose 60 Gy, we considered that this chronic heart failure had not been related to the radiation therapy because the local field to the supraclavicular lymphnode metastases had been used in this patient. The results of a univariate analysis of the patient characteristics for overall survival are shown in Table 4. Patients with a PS 0 1 and those with a recurrent tumor measuring,3.5 cm in diameter showed a better overall survival than each other group; 2-year overall survival rate was 34% in PS 0 1 (versus 6% in PS 2 3, P ¼ ) and 50% in tumor size 3.5 cm (versus 0% in.3.5 cm, P, ). The patients who had received preoperative radiotherapy showed a worse prognosis in comparison to those without previous radiotherapy; the 2-year overall survival rates were 15 and 40%, respectively (P ¼ 0.005). However, there was no significant difference between the survival rates in the patient s age, gender, initial stage, the number of recurrent tumors, recurrent sites or interval periods from surgery to recurrence. In a univariate analysis of treatment variables for overall survival is shown in Table 5. The patients treated with 50 Gy or RESULTS The initial response to radiation therapy for the loco-regionally recurrent tumors was 15 patients with a complete response, 49 with a partial response, 17 with no change and 1 who showed progressive disease. The survival periods after start of radiotherapy for recurrences of all patients ranged from 1 to 120 months. The median survival period after recurrence was 7.0 months. Fifty-one patients died from Figure 1. The overall and cause-specific survival curves of all patients with recurrent esophageal cancer after surgery.

4 Jpn J Clin Oncol 2007;37(12) 921 Table 4. Univariate analysis of the pretreatment variables for overall survival Variables Survival rates. (%) P-Value 2 years 5 years Age, 61 (n ¼ 46) (n ¼ 36) Gender Male (n ¼ 70) Female (n ¼ 12) Initial stage I(n¼16) II (n ¼ 41) III IV (n ¼ 24) Performance status 0 1 (n ¼ 64) (n ¼ 18) 6 0 Number of tumors Solitary (n ¼ 50) Multiple (n ¼ 32) 13 0 Sites of recurrence Supraclavicule (n ¼ 24) Mediastinum (n ¼ 36) Abdominal (n ¼ 10) Multi-region (n ¼ 12) 11 0 Pattern of recurrence Local (n ¼ 10) Lymphnode (n ¼ 72) Tumor size, 3.5 cm (n ¼ 45) 50 25, cm (n ¼ 37) 0 0 Time to recurrence, 10 M (n ¼ 42) M (n ¼ 40) Previous radiotherapy Yes (n ¼ 36) No (n ¼ 46) more doses in total had a better prognosis than those treated with,50 Gy; the 2-year overall survival rates were 26 and 10%, respectively (P ¼ 0.003). However, no significant survival difference was observed between the patients treated with 60 Gy or more doses and those treated with,60 Gy; the 2-year overall survival rates were 26 and 20%, respectively (P ¼ 0.10). Also, there was no significant difference between the patients receiving the combination chemotherapy and radiotherapy and radiotherapy alone. Table 5. Univariate analysis of the treatment variables for overall survival Variables Survival rates P-Value 2 years 5 years Total dose, 50 Gy (n ¼ 16) Gy (n ¼ 66) 26 13, 60 Gy (n ¼ 48) Gy (n ¼ 34) Total treatment time, 44 days (n ¼ 40) days (n ¼ 42) Chemotherapy Yes (n ¼ 30) No (n ¼ 52) In order to determine the independent prognostic factors, the initial stage, PS, the maximum tumor size, the total radiation dose, a history of previous radiation therapy were entered into a multivariate analysis. The results for all 82 patients are summarized in Table 6. In a multivariate analysis, a recurrent tumor size 3.5 cm (P ¼ ), a PS 0 1 (P ¼ 0.04) and treatment with 50 Gy or more doses (P ¼ 0.04) were independent positive predictive factors for the overall survival. DISCUSSION Through a retrospective study comprising a large group of patients treated with radiotherapy for loco-regional recurrences of esophageal cancer after surgery, we evaluated the treatment outcome and the prognostic factors of these patients. From our results, the prognosis of these patients was found to be still poor. However, our results also suggest that a long-term survival could be expected with definitive radiotherapy for the patients with small-size tumors and with a good PS. In our study, the survival rates at 2 and 5 years were 22 and 11%, respectively. Recently, similar survival results Table 6. Multivariate analysis of the pre- and treatment variables for overall survival Variables P-Value HR 95% CI Initial stage I versus II III Performance status 0 1 versus Tumor size,3.5 cm versus 3.5 cm Total dose,50 Gy versus 50 Gy Previous radiotherapy No versus yes HR, hazard ratio; CI, confidence interval.

5 922 Radiotherapy for recurrent esophageal cancer have been shown by several retrospective studies (5 7). Raoul et al. (5) reported the survival rates of patients with postoperative recurrence treated with chemoradiotherapy to be 47.1% at 1 year, 17.1% at 2 years and 4.3% at 3 years. Nemoto et al. (6) also previously showed the results of radiation therapy for loco-regionally recurrent esophageal cancer after surgery. In their series, the survival rates at 1, 2 and 3 years were reported to be 33, 15 and 12%, respectively. Nishimura et al. (10) combined chemotherapy of protracted infusion of 5-FU and cisplatin with radiotherapy of 60 Gy concurrently for 18 patients with postoperative recurrence of esophageal cancer. For patients without distant metastases in their series, the 2-year survival rate was reported to be 19%. In our series, there was no patient with serious late toxicities related to radiotherapy despite that 36 patients who had a history of preoperative radiotherapy were included. The reason of this result was unclear but considered as follow. First, a local field was used in most of the patients. Secondly, total dose used for our series was modest (50.4 Gy in median). Thirdly, there is the possibility that the survival period after treatment (7 months in median) was too short to evaluate late toxicities adequately. The survival outcome of radiotherapy with or without chemotherapy for postoperative recurrences of esophageal cancer was considered to be poor in the previous retrospective series and also in our present study. However, high objective response rates for recurrent tumors were shown in these previous studies: 65% by Raoul et al. (5), 91% by Nemoto et al. (6) and 72% by Nishimura et al. (10). In addition, in our presented study, an objective response for recurrent tumors was obtained in 78% of the patients. The objective response rates are similar between these studies including ours. For patients with postoperative recurrences of esophageal cancer, therefore, radiotherapy with or without combined chemotherapy seems to be an effective method, at least as one of the palliative therapy for the symptoms related to recurrent tumors. In fact, a symptomatic improvement was shown in 70% or more of patients treated with radiotherapy (5,6). Chemotherapy combined with radiotherapy has been shown to be effective as a primary management for esophageal cancer (11,12). However, it has not yet been determined whether the combination of chemotherapy is superior to radiation alone for recurrent esophageal cancer patients. Nemoto et al. (6) reported a better local control in the patients treated with a combination of chemotherapy and radiotherapy in comparison to radiation alone. However, no survival difference was shown between the two groups in their report. Although the chemotherapy regimes varied in our study, there was no significant survival difference between the patients who received radiotherapy combined with chemotherapy and those who received radiotherapy alone. Recently, Nemoto et al. (13) reported the clinical results of radiotherapy combined with chemotherapy using CDGP and 5-FU for recurrent or inoperable esophageal cancer. CDGP is a platinum derivative with less renal toxicity that is considered to have an effectiveness comparable to cisplatin (14). The 2-year survival rate of seven patients with postoperative recurrent esophageal cancer treatedwithradiotherapycombinedwithcdgpand5-fu was reported to be 69% (13). Furthermore, Jingu et al. (15) recently reported excellent results of a phase II study using radiotherapy combined with CDGP and 5-FU; 3-year local control and overall survival were 72 and 56%, respectively (15). Recently, docetaxel has been shown to be an effective agent for primary and recurrent esophageal cancer, and several promising results have been reported (16,17). However, the optimal schedule and dosage of these chemotherapy agents when combined with radiotherapy have not been determined yet. In order to improve the prognosis of recurrent esophageal cancer patients, it will thus be necessary to further investigate the effectiveness of radiotherapy combined with these chemotherapy regimens. There have been few studies evaluating the prognostic factors of patients with postoperative recurrences who were treated with radiotherapy. Nemoto et al. (6) reported the time interval between surgery and recurrence and the patient age to be significant prognostic factors. In our present series, the PS, tumor size and radiation dose were selected as independent prognostic factors by a multivariate analysis. The PS was reported to be a prognostic factor of recurrent esophageal cancer also by Jingu et al. (15). The time interval between surgery and recurrence or age was not selected as prognostic factors in a univariate or multivariate analysis in our study. The reason for these differences of the results between the studies is unclear. A variation of patient s background and/or a difference of analysed factors may be one of the reasons. The univariate analysis in our series showed the patients treated with 50 Gy or more of radiotherapy to have a better prognosis in comparison to those with less than 50 Gy. Furthermore, a total dose of 50 Gy or more tended to be positively predictive for the overall survival in a multivariate analysis. Our result suggests that 50 Gy or more doses should be considered in the radiotherapy for postoperative loco-regional recurrence of esophageal cancer. Nemoto et al. (13) and also Jingu et al. (15) reported excellent survival results without serious complications in the patients treated with radiotherapy with total dose of 60 Gy for post-operative loco-regional recurrences. In our series, however, there was no significant difference in overall survival between the patients treated with 60 Gy or more doses and those treated with less than 60 Gy. The optimal dose of definitive radiotherapy for postoperative recurrent esophageal cancer remains to be determined in a future prospective randomized trial. In conclusion, our study suggests that PS, size of tumors and total irradiation dose are significant prognostic factors in the patients treated with radiotherapy for the loco-regional recurrences of esophageal cancer after surgery. However, there were numerous variations in the treatment methods in this series because of the retrospective study including many patients treated through long periods. These significant

6 Jpn J Clin Oncol 2007;37(12) 923 biases are thought to make it difficult to interpret our results. To confirm the prognostic significance of these factors, we are planning to analyse the clinical outcome of the patients treated with our current protocol using a concurrent combination of radiotherapy and CDDP plus 5-FU. Finally, further evidence should be accumulated based on prospective trials to determine the prognostic factors and to individualize the treatment strategy for patients with postoperative recurrent esophageal cancer. Conflict of interest statement None declared. References 1. Ando N, Oazawa S, Kitagawa Y, et al. Improvement in the results of treatment of advanced squamous esophageal carcinoma over fifteen consecutive years. Ann Surg 2000;232: Bhansali MS, Fujita H, Kakegawa T, et al. Pattern of recurrence after extended radical esophagectomy with three-field lymph node dissection for squamous cell carcinoma in the thoracic esophagus. World J Surg 1997;21: Collard JM, Otte JB, Reynaert MS, et al. Extensive lymph node clearance for cancer of the esophagus or cardia: merits and limits in reference to 5-year absolute survival. Hepatogastroenterology 1995;42: Tam PC, Siu KF, Cheung HC, Ma L, Wong J. Local recurrence after subtotal esophagectomy for squamous cell carcinoma. Ann Surg 1987;205: Raoul JL, Le Prise E, Meunier B, et al. Combined radiochemotherapy for postoperative recurrence of oesophageal cancer. Gut 1995;37: Nemoto K, Ariga H, Kakuto Y, et al. Radiation therapy for loco-regionally recurrent esophageal cancer after surgery. Radiother Oncol 2001;61: Yamashita H, Nagasawa K, Tago M, Nakamura N, Shiraishi K, Ohtomo K. Salvage radiotherapy for postoperative loco-regional recurrence of esophageal cancer. Dis Esophagus 2005;18: Sobin LH, Wittekind CH. International Union Against Cancer (UICC). TNM Classification of Malignant Tumors. 5th ed. New York: Wiley Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;53: Nishimura Y, Koike R, Nakamatsu K, et al. Concurrent chemoradiotherapy with protracted infusion of 5-FU and cisplatin for postoperative or residual esophageal cancer. Jpn J Clin Oncol 2003;33: Coia LR, Engstrom PF, Paul AR, Stafford PM, Hanks GE. Long-term results of infusional 5-FU, mitomycin-c and radiation as primary management of esophageal carcinoma. IntJRadiatOncolBiolPhys 1991;20: Herskovic A, Martz K, al-sarraf M, et al. Combined chemotherapy and radiotherapy compared with radiotherapy alone in patients with cancer of the esophagus. N Engl J Med 1992;326: Nemoto K, Matsushita H, Ogawa Y, et al. Radiation therapy combined with cis-diammine-glycolatoplatinum (Nedaplatin) and 5-fluorouracil for untreated and recurrent esophageal cancer. Am J Clin Oncol 2003;26: Sasaki Y, Amano T, Morita M, et al. Phase I study and pharmacological analysis of cis-diammine (glycolato) platinum (254-S; NSC D) administered by 5-day continuous intravenous infusion. Cancer Res 1991;51: Jingu K, Nemoto K, Matsushita H, et al. Results of radiation therapy combined with nedaplatin (cis-diammine-glycoplatinum) and 5-fluorouracil for postoperative locoregional recurrent esophageal cancer. BMC Cancer 2006;6: Muro K, Hamaguchi T, Ohtsu A, Boku N, Chin K, Hyodo, et al. A phase II study of single-agent docetaxel in patients with metastatic esophageal cancer. Ann Oncol 2004;15: Airoldi M, Cartesina G, Giordano C, Pedani F, Bumma C, Gabuiele P. Docetaxel and vinorelbine: an effective regimen in recurrent squamous cell esophageal carcinoma. Med Oncol 2003;20:19 24.