Retrospective analysis of definitive radiotherapy for patients with superficial esophageal carcinoma: Consideration of the optimal treatment method with a focus on late morbidity

Radiotherapy and Oncology 95 (2010) 234–239

Contents lists available at ScienceDirect

Radiotherapy and Oncology journal homepage: www.thegreenjournal.com

Esophageal cancer radiotherapy

Retrospective analysis of definitive radiotherapy for patients with superficial esophageal carcinoma: Consideration of the optimal treatment method with a focus on late morbidity Takeshi Kodaira *, Nobukazu Fuwa, Hiroyuki Tachibana, Tatsuya Nakamura, Natsuo Tomita, Rie Nakahara, Haruo Inokuchi, Nobutaka Mizoguchi, Akinori Takada Department of Therapeutic Radiation Oncology, Aichi Cancer Center Hospital, Aichi, Japan

a r t i c l e

i n f o

Article history: Received 7 November 2008 Received in revised form 25 December 2009 Accepted 23 January 2010 Available online 12 February 2010 Keywords: Superficial esophageal cancer Second primary Definitive radiotherapy Iodine staining

a b s t r a c t Purpose: To evaluate the clinical efficacy of definitive radiotherapy for patients with superficial esophageal cancer. Material and methods: From 1990 through 2006, 97 patients with stage I disease were treated with radiotherapy with or without chemotherapy. All patients were diagnosed with panesophagoscopy and computed tomography. Chemotherapy was added in 61 patients, and intra-cavitary brachytherapy (ICBT) was used in 27 patients. Results: The patients were 90 men and seven women with a median age of 65.7 years (range; 41–89). At last follow-up with a median follow-up duration of 35.7 months, 3 year-overall and progression-free survival (PFS) rates were 81.5% (95% C.I. = 73.3–89.7%) and 55.8% (95% C.I. = 45.2–66.4%), respectively. Shorter tumor length was a significantly favorable factor for the PFS rate (P = 0.02) and local failure-free (LFF) rate (P = 0.007) on both univariate and multivariate analyses. Although the addition of ICBT had no apparent benefit for survival or tumor control, the rate of severe adverse effects including lethal esophageal ulcers, showed a higher tendency in patients receiving ICBT. Conclusions: Our results regarding efficacy from the viewpoint of organ preservation are promising. Special care would be taken for the use of ICBT for patients with superficial esophageal cancer, especially if they have received chemoradiotherapy. Ó 2010 Elsevier Ireland Ltd. All rights reserved. Radiotherapy and Oncology 95 (2010) 234–239

Esophageal cancer still remains disease with high mortality despite recent improvements in diagnosis and treatment. The mortality rate for Japanese patients with esophageal cancer is 9.0 per 100,000 (15.7 per 100,000 in males, 2.6 per 100,000 in females), representing 3.4% (4.9% in males, 1.3% in females) of all deaths by malignant neoplasms in 2006 [1]. Mortality rate of esophageal cancer has increased slightly over the past two decades. Although half of patients have adenocarcinoma in Western countries, the majority of patients in Japan have squamous cell carcinoma. In recent years, the number of patients with stage I disease, in which the primary lesion is limited to the mucosal layer of the esophagus, has increased. According to the third Comprehensive Registry of Esophageal Cancer in Japan, in 1999 clinical stage I disease (T1N0M0) accounted for 20.1% of all cases, 12.3% of cases in patients treated with radiotherapy or chemoradiotherapy, and 20.1% of cases in patients treated with esophagectomy [2]. The reported number of patients undergoing esophagectomy is more

than 3.6 times the number of patients receiving radiation or chemoradiation therapy (365 vs. 103). In Japan several retrospective studies [3–7] have shown promising clinical outcomes of definitive radiotherapy for early stage disease of the esophagus, and because the trend in Japan for a less-toxic treatment, definitive radiotherapy has become a standard treatment for patients with stage I esophageal cancer. The Japan Clinical Oncology Group is leading a randomized controlled trial (JCOG0502) to compare the clinical outcomes of surgery and chemoradiotherapy for patients with stage I esophageal cancer. Since 1990, more than 100 patients with stage I disease have received definitive radiotherapy at our hospital. To evaluate its clinical efficacy, we retrospectively analyzed clinical outcomes of this cohort according to disease characteristics and treatment contents.

* Corresponding author. Address: Department of Therapeutic Radiation Oncology, Aichi Cancer Center Hospital, 1-1 Kanokoden Chikusa-ku, 464-8681 Nagoya, Aichi, Japan. E-mail address: [email protected] (T. Kodaira).

Eligibility criteria

0167-8140/$ - see front matter Ó 2010 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.radonc.2010.01.005

Materials and methods

From 1990 through 2006, 97 patients with histologically confirmed stage I squamous cell carcinoma of the esophagus

T. Kodaira et al. / Radiotherapy and Oncology 95 (2010) 234–239

diagnosed according to the criteria of the International Union Against Cancer, sixth edition (2002), were eligible for this analysis. The laboratory examinations consisted of complete blood count (CBC), serum chemistry, 24 h creatinine clearance, arterial blood gas, electrocardiography, and pulmonary function tests. The staging work-up included chest radiography; computed tomography (CT) of the cervix, chest, and abdomen; esophagography; and panesophagoscopy with iodine staining. In a selected case diagnosed as having submucosal layer disease (n = 12), the endoscopic ultrasonography (EUS) was also used to help the evaluation of tumor depth. In weekly conference, disease stage including the depth of tumor invasion is usually determined by institutional tumor board consisted of surgeon, medical physician, radiologists, and radiation oncologists. Considering our sufficient diagnostic quality for tumor depth (>90%) by assessing surgical findings, we did not use EUS for all cases. After May 2002, positron emission tomography (PET) or PET/CT was used for staging work-ups. Lymph node swelling of 10 mm or more along the long axis was found in no patient on CT scan.

235

(HDR)–ICBT was performed with a remote-after-loading system using an iridium source. A double-balloon applicator system [10] was used in patients treated with HDR–ICBT. Five patients also received intra-cavitary hyperthermia [11] along with ICBT. ICBT was performed immediately after the completion of EBRT. The median dose of ICBT was 10 Gy (range, 4–20 Gy). The 2–5 fractions with 3– 5 Gy per fraction were usually used in their series. The estimated dose was calculated 5 mm beyond the surface of the double-balloon applicator [4,10]. Only four patients received relatively higher radiation dose of 80 Gy (60 Gy of EBRT and 20 Gy of ICBT), otherwise majority patients (n = 88, 91%) received total dose (EBRT + ICBT) with a range from 54 Gy to 66 Gy. To appropriately evaluate for dose impact for late mucosal reaction by both EBRT and ICBT, we have also calculated EQD2 which is the calculated biological dose equivalent, as if 2 Gy per fraction would have been applied. As for estimate of late adverse event, alpha/beta value of three was used for estimate of late mucosal and submucosal reactions. Follow-up

Chemotherapy Sixty-one of the patients had submucosal disease and, because laboratory studies indicated a sufficiently good general condition, received systemic chemotherapy. Thirty-nine patients received two cycles of fluorouracil (5-FU) and cisplatin administration via intravenous infusion every 4 weeks. Elderly patients and patients with medical illness (n = 8) received radiotherapy and protracted daily low-dose chemotherapy with 5-FU or cisplatin or both during external beam radiotherapy (EBRT). Seven patients with simultaneously diagnosed head and neck squamous cell carcinoma received intravenous 5-FU and nedaplatin with an alternating setting. The details of this protocol have been reported elsewhere [8]. Thirty-six patients received only radiation therapy without systemic chemotherapy. Patients with mucosal disease (n = 24) received radiotherapy alone, because lymphatic spread is thought to be rare for such patients. Radiotherapy Radiation therapy with a megavoltage photon beam (6 or 10 MV) was started on day one of systemic chemotherapy. The gross tumor volume (GTV) was defined as the total volume of the primary tumor. Before taking a simulator film or planning CT, the esophageal lesion was marked with a metallic clip to indicate the cranio-caudal extent. The clinical target volume (CTV) was defined as the GTV with a 2 cm margin in the longitudinal direction. No prophylactic lymph node region was included in the CTV. The planning treatment volume (PTV) of the primary tumor was defined as the CTV with a 1–1.5 cm cranio-caudal margin and a 0.5–1 cm lateral margin. Radiotherapy was given with a daily 2 Gy fraction to a total dose of 60 Gy. For patients who did not receive chemotherapy, the proposed dose was designed to be 66 Gy. Initially, a dose of 40 Gy to the PTV was planned for 1–4 weeks, using anteroposterior parallel opposed pairs of portals. Then, another 20 Gy was applied to the PTV with a proper margin in an oblique parallel opposed manner or with the dynamic conformal method during fifth and sixth weeks. The spinal cord never received more than 45 Gy. The radiation dose and the volume to the lung and heart were minimized in each case. Until 2004, patients with submucosal lesions shorter than 5 cm received intra-cavitary brachytherapy (ICBT) after the completion of EBRT with 46–56 Gy [9]. Twenty-seven patients received ICBT with a radium (n = 3) or iridium source (n = 24). High-dose-rate

Patients were followed up at 2–3 months intervals for the first 2 years and at 4–6 months intervals thereafter. Follow-up examinations included physical examinations, CBC, chemistry profiles including tumor markers, esophagoscopy, and CT of the neck, chest, and abdomen. Treatment response was evaluated 1 month after the completion of radiation therapy. Usually, patients underwent esophagoscopy and CT of the neck, chest, and abdomen 1 month after the completion of the treatment session. Statistical considerations The survival time was defined as the period from the start of treatment to death or the last follow-up evaluation, and progression-free survival (PFS) was defined as the period from the start of treatment to progression of disease (PD) or death for any reason. Development of second primary was not considered an event for PFS. New esophageal cancer out of radiation field was excluded from second primary. The local failure-free (LFF) time was defined as the period from the start of treatment to the failure of the esophagus within the radiation field. Newly esophageal cancer out of radiation field was counted as event for disease progression, but not for local failure. Statistical differences between groups were assessed with the x2 test. The overall survival (OAS) and PFS curves were calculated using the Kaplan–Meier method [12]. The log-rank test [13] was used to compare survival curves. Cox’s proportional hazard model [14] was used for multivariate analysis. Prognostic factors selected for univariate and multivariate analyses, included patient age, sex, tumor length, tumor location, mucosal or submucosal lesion, solitary or multicentric lesion, coexistence of second primary, use of ICBT, and use of chemotherapy. Results Patient’s characteristics and treatment contents Patients’ characteristics and treatment contents are summarized in Table.1. Seventeen patients had undergone endoscopic mucosal resection (EMR) for esophageal disease with mucosal lesions before consultation for definitive radiotherapy. Seventy-one (73.2%) patients underwent consultation for radiotherapy because radical surgery could not be performed owing to second primary or poor general condition or both. Twenty-six (26.8%) patients who were considered contradiction for radical surgery hoped to receive chemoradiotherapy.

236

Definitive radiotherapy for superficial esophageal cancer

rate

Table 1 Patient characteristics and treatment contents.

3/5 year-OAS rate 81.5/63%

Factor Sex Age (years) PS Depth of invasion

Length (cm) Location of primary lesion

Multicentric Second primary EBRT dose (Gy) ICBT Total dose (EBRT + ICBT; Gy) EQD2 Chemoradiotherapy

Male Female Median Range 0 1 Mucosal layer Submucosal layer NA Median Range Ce Ut Mt Lt Yes No Yes No Median Range Yes No Median Range Median Range Yes No

90 7 66 41–89 30 67 24 71 2 4 1–16 5 15 55 22 32 65 47 50 60 38–70.2 27 70 60 50–84 60 50–92 61 36

3/5 year-PFS rate 55.8/48.1% Patients at risk OAS PFS

93 71

79 55

39 24

29 19

years OAS = overall survival PFS = progression-free survival Fig. 1. Overall and progression-free survival curves of all patients.

rate 3year-PFS rate 63.3%

Length < 5 cm, N=75

PS = performance status, NA = not assessed, EBRT = external beam radiotherapy, ICBT = intra-cavitary brachytherapy, EQD2 = biological effective dose adjusted by 2 Gy per fraction.

P=0.02

3year-PFS rate 30.8%

Survival and failure site The median duration of follow-up for all patients was 35.7 months (range, 4.7–132 months). At last follow-up in June 2008, 46 patients were alive without recurrent disease and 16 patients were alive with recurrent disease. Sixteen patients had died of progression of esophageal cancer, and 19 patients had died for other reasons. Esophageal cancer recurred in 25 patients a median duration of 10.6 months (range, 3.5–44.4 months) after the completion of radiotherapy. Twenty-one failures in the esophagus were located within a previously irradiated field. Four new esophageal cancers developed outside the radiation field. Mediastinal lymph node metastasis developed in six patients, of whom three received additional radiation therapy with or without chemotherapy. In our analysis distant metastasis to the liver occurred in only one case. The 3- and 5-year OAS rates were 81.5% (95% C.I. = 73.3–89.7%), and 63% (95% C.I. = 51.0–75.0%), respectively (Fig. 1). On univariate analysis, no factor adversely affecting OAS was identified. The 3and 5-year PFS rates were 55.8% (95% C.I. = 45.2–66.4%), and 48.1% (95% C.I. = 36.5–59.7%), respectively (Fig. 1). Patients with a lesion longer than 5 cm had a significantly worse PFS rate than did patients with lesions 5 cm or shorter (P = 0.02; Fig. 2). Multivariate analysis showed that a tumor length greater than 5 cm was a significant adverse factor for PFS (hazard ratio [HR]) = 2.134, 95% C.I. = 1.075–4.234; P = 0.03; Table 2). The greater tumor length was also a significant adverse prognostic factor for the LFF rate on univariate analysis (P = 0.007; Table 2). On multivariate analysis, shorter tumor length (HR = 3.776, 95% C.I. = 1.285–11.091; P = 0.02) and the addition of chemotherapy (HR = 5.208, 95% C.I. = 1.367–20.00; P = 0.02) were significant favorable prognostic factors for the LFF rate (Table 2). Our cohort had a relatively different time, so we divided into two groups regarding start of radiotherapy (1990–1999; n = 25, and 2000– 2006; n = 72). We found no significant difference of OAS, PFS, and LFF between the groups divided at year started (P >0.05).

49 32

Patients at risk <5cm 63 5 cm 12

Length 47 9

28 5

20 5

5 cm, N=21

16 4

years PFS = progression-free survival Fig. 2. Progression-free survival curves of groups divided by tumor length.

Chemotherapy Survival rates (OAS, PFS) of patients who received chemoradiotherapy did not differ significantly from those of patients who received radiotherapy alone in univariate analysis (P >0.05). Patients treated with chemoradiotherapy had a slightly, but not significantly, higher LFF rate than did patients treated with radiation alone (P = 0.07). Patients who received chemoradiotherapy had significantly greater rates of submucosal lesions (P < 0.0001), second primary (P = 0.011), and the addition of ICBT (P = 0.0002) than did patients treated with radiotherapy alone. ICBT The OAS and PFS rates of both groups undergone ICBT and that without ICBT did not differ significantly (P >0.05). The addition of ICBT had no apparent effect on the LFF rate (P >0.05). The average estimate of EQD2 of group received ICBT proved to be significantly higher compared to that of group without ICBT (69.5 vs. 60.9; P < 0.0001). Second primary’ Sites of second primary are shown in Table 3. The most common site of second primary was the stomach (n = 13), followed by the

0.5895 0.6441 0.965 0.0302*

0.1297 0.65

0.0652

0.9598 0.3065

0.8222 0.7526

0.2932 0.5534

0.3339 0.5094

0.6056 0.6562

0.0978 0.0764

0.741 0.9352

0.9064 0.7491

0.0165*

0.8637

0.1734

0.3064

0.4034

1.327 (0.68–2.58) Base Base 1.811(0.58–5.66) Base 1.503 (0.4–6.6) 1.352 (0.38–4.86) 1.033 (0.24–4.46) 2.134 (1.08–4.23) Base Base 1.048 (0.48–2.29) Base 1.126 (0.56–2.28) 1.757 (0.9–3.42) Base 1.235 (0.55–2.75) Base Base 1.446 (0.56–2.28) 0.5826

0.3084

0.44

0.51

0.9091

0.4972

0.4508 0.9639 0.2947

0.5406

0.7058

0.531 0.582 0.577 Not reached 0.4 0.629 0.551 0.595 0.308 0.633 0.5 0.572 0.578 0.551 0.452 0.661 0.577 0.532 0.63 0.515 0.2701

0.443

0.5893

0.9011

0.8142

0.4721

0.811

0.4317

2.355 (1.03–5.39) Base Base 1.35 (0.28–6.41) 1.794 (0.28–11.67) Base 1.651 (0.45–6.08) 1.035 (0.23–4.67) 1.586 (0.67–3.76) Base 1.396 (0.53–3.66) Base 1.006 (0.43–2.34) Base 1.321 (0.58–3.03) Base 1.49 (0.54–4.1) Base 1.583 (0.65–3.83) Base 0.1666

Intra-cavitary brachytherapy

Chemoradiotherapy

Second primary

Multicentric

Depth

Length

Location

Sex

Age

=66 years <66 years Male Female Ce Ut Mt Lt =5 cm <5 cm M Sm Yes No Yes No Yes No Yes No

77.2 85.8 81.6 75 60 92.9 78.2 90.9 70.6 85.2 76.7 83.2 81.1 82.8 74.5 88.8 81.3 81.9 85.2 79.2

P-value P-value

OAS = overall survival, PFS = progression-free survival, LFF = local failure-free, HR = hazard ratio, C.I. = confidential intervals, CE = cervical esophagus, Ut = upper thoracic, Mt = middle thoracic, Lt = lower thoracic.

0.0157*

0.7028 0.8532 0.0156* 0.005*

0.9782

0.8543

0.8222 0.5752

0.9861

1.01(0.35–2.95) Base Base 1.144(0.16–12.05) Base 2.237(0.13–37.4) 1.379(0.29–33.2) 1.19(0.2–36.2) 3.776(1.29–11.09) Base Base 2.012(0.55–7.41) Base 1.144(0.35–3.69) Base 1.031(0.32–3.36) Base 5.208(1.367–20) Base 2.728(0.75–9.99) 0.789 0.814 0.798 0.857 0.8 0.929 0.79 0.791 0.604 0.854 0.735 0.819 0.806 0.799 0.76 0.837 0.856 0.712 0.843 0.781

0.8523

P-value HR(95% C.I.) HR(95% C.I.) Rate

Factors

Table 2 Univariate and multivariate analyses of survival and tumor control.

0.0427*

Univariate

Rate P-value HR(95% C.I.)

Univariate Univariate

Rate

3-Year PFS

Multivariate 3-Year OAS

P-value

Multivariate

3-Year LFF

P-value

Multivariate

T. Kodaira et al. / Radiotherapy and Oncology 95 (2010) 234–239

237

larynx (n = 8), hypopharynx (n = 7), and oral cavity and tongue (n = 6). There included both synchronous and metachronous diseases in our cohort. The 3-year OAS and PFS rates in patients with second primary did not differ significantly from those in patients without second primary (P >0.05; Table 2). Failure sites and salvage treatment Disease recurred locally within the radiation field in 21 cases: four cases of new esophageal lesions, six cases of nodal recurrence, and one case of distant failure. Eight patients underwent EMR for salvage intent, and three patients underwent argon plasma coagulation treatment for in-field esophageal recurrence. Additional radiotherapy with or without chemotherapy was performed for primary recurrence or mediastinal lymph node recurrence or both in 10 patients. Only one patient underwent salvage ICBT for primary tumor recurrence. Fourteen of 25 patients (56%) with local recurrence, including new esophageal cancers, could undergo salvage treatment. Two patients with local recurrence underwent salvage surgery. Two patients underwent palliative chemotherapy for PD. Toxicity Hematologic toxicity was not examined in this study, because considerable variations existed in the content of chemotherapy. In our analysis, late adverse events were focused on treatment content. There were 14 late adverse events of grade 3 or higher (14.3%; Table 4): three cardiac events, two pulmonary events, and nine esophageal events. Among these patients, only one patient received a relatively higher total radiation dose of 80 Gy (EQD2 92). Patients treated with ICBT had a slightly, but not significantly, higher incidence of esophageal ulcers or stenosis or both than did patients not treated with ICBT (18.5% vs. 5.7%; P = 0.07). In addition patients treated with ICBT also showed a tendency for a higher rate of late sequelae of grade 3 or higher (25.9% vs. 10%; P = 0.057). We did not find any difference in morbidity rate between the group with chemotherapy and that without chemotherapy (P >0.05). The estimates of EQD2 for group developed late sequelae of grade 3 or higher showed no apparent difference with those of group without late event (65.6 vs. 62.9; P >0.05). Four patients had treatment-related deaths. All four patients had received ICBT, and three of the four patients had received chemoradiotherapy. Esophageal ulcers developed in three patients and did not heal with conservative treatment. In one of the three patients a mediastinal abscess developed because of the esophageal ulcer and led to a fatal systemic bacterial infection (case six in Table 4). In another patient (case seven in Table 4), a severe esophageal ulcer became resistant to conservative treatment and salvage operation was planned, however, a fatal arrhythmia suddenly developed before surgery. An esophageal ulcer accompanied by pericarditis developed in another patient. In two patients (cases nine and ten in Table 4), including the former patient, cardiac tamponade led to death from heart failure. Discussion Clinical outcomes are more favorable in patients with superficial esophageal cancer than in patients with more advanced disease. In Japan, the majority of patients with esophageal cancer have squamous cell cancer and the high incidence of superficial lesions is higher than in Western country [2]. In this retrospective analysis the 3-year OAS and PFS rates were 81.5% (95% C.I. = 73.3–89.7%) and 55.8% (95% C.I. = 45.2–66.4%), respectively. We believe that these results are comparable to the outcomes of surgery at other Japanese institutes.

238

Definitive radiotherapy for superficial esophageal cancer

Table 3 Incidence of second primary. Site

Number

Stomach Larynx Hypopharynx Tongue/oral cavity Lung Colon Oropharynx Nasopharynx Liver Skin Kidney Pancreas

13 8 7 6 4 2 2 1 1 1 1 1 47

Patients with mucosal lesions limited to the m2 layer show a low incidence of lymph node metastasis, and in our institute patient with disease limited to m2 layer are treated with EMR; thus, definitive radiotherapy is suggested for patients with extensive disease (>5 cm in length) or circumferential extension for EMR is considered technically inadequate. Patients with submucosal disease are thought to be at greater risk (10–50%) for invasive disease with lymphatic spread. Standard treatment for patients with submucosal disease is surgery, consisting of total esophagectomy and three-region lymph node dissection. In Japan surgical outcomes are reportedly acceptable and are associated with lower morbidity rates [15], but patients are forced to accept considerable organ and functional impairment. Several Japanese investigators have reported the results of definitive radiotherapy for patients with submucosal disease without lymph node metastasis [3,5,7,16] (Table 5). In these reports, clinical outcomes in cases of stage I disease were comparable to those of surgery and had an acceptably low morbidity. In our cohort, radiotherapy was performed because the patients refused to undergo surgery or their medical condition, such as second primary, did not allow surgery. In the majority of patients in our cohort, medical conditions did not allow radical surgery. Although survival was apparently correlated with these prognostic factors in our analysis, several biases could have affected clinical outcomes. Thus, we believe our results indicate this treatment is quite acceptable as a standard treatment for stage I disease. Nemoto et al. [7] have reported Japanese clinical outcomes in multicenter of Japanese patients with stage I esophageal cancer treated with radiotherapy. They reported on 147 patients, largest reported number, with 5-year OAS rates of 43–62%. The majority of these

patients underwent ICBT without chemotherapy [7] (Table 5). Our study involved the largest number of patients at a single institute; thus, our data have shown a meaningful effect on clinical practice for superficial esophageal cancer. Regarding late morbidity, we found slightly high rates of morbidity, including death from esophageal ulceration and pericardial effusion. In our analysis, the addition of ICBT might affect the severe late adverse events, but ICBT had no significant beneficial effect on clinical outcome. We added further analysis for dose relationship of late adverse event using EQD2, however, apparent difference could not be seen between two groups. In addition, several treatment-related deaths occurred from late esophageal adverse effects among group treated with ICBT. RTOG 92–07 [17] examined the effectiveness of concurrent chemoradiotherapy with 50 Gy of EBRT and four courses of cisplatin and 5-FU followed by a 10-Gy boost via ICBT. In this report, severe morbidity or mortality occurred in 13 and four of 50 eligible patients, respectively. Therefore, the authors concluded that extreme caution should be taken when adding ICBT to the standard chemoradiotherapy. Indeed, increasing the risk of esophageal ulcer or perforation or both has also been reported [10]. In Japan, an ICBT applicator with a centering function through a double-balloon structure is usually used [10,4,3]. Several reports have shown its clinical effectiveness with low toxicity rates for superficial esophageal cancer [16]. In most of these reports a cobalt source was used. In our study the same type of balloon was used using iridium source, and doses were calculated with the same method. We have changed the fraction size of ICBT because of significant ulceration of the esophagus; however, ulceration caused the death of one patient treated with a small fraction size of 3.5 Gy. Finally, we have decided to stop using ICBT, especially for patients who have received chemoradiotherapy. If the data of patients treated with ICBT are excluded, the rate of late adverse effects was 6.1% without treatment-related deaths being included. As shown in Table 5, the addition of ICBT led to a higher rate of esophageal morbidity in several Japanese series. Thus, we think special caution should be taken for the application of ICBT for superficial esophageal cancer, especially in the setting of chemoradiotherapy. However, we could not show the additional analysis of relation of occurrence of esophageal ulcer with the surface dose of ICBT, thus several limitations would exist in our analysis. In our cohort we did not use a prophylactic large field for patients with submucosal disease. One of the reasons for applying a small radiation field is that many of our patients had medical limitations such as cardiopulmonary disease, old age, and existence of second primary. Ishikura et al. [18] have reported fatal late radia-

Table 4 Characteristics and treatment contents of patients developed severe late toxicity. Case

Age (years)

Site

Depth

Event

On set of event

Failure

EBRT dose (Gy)

ICBT dose (Gy)

ICBT fraction

Total dose (Gy)

EQD2

HT

CT

Status

1 2 3 4 5 6 7 8 9 10 11 12 13 14

69 55 62 70 67 58 53 57 67 75 75 64 71 64

Mt Mt Mt Mt Mt Mt Mt Mt Mt Ut Mt Lt Mt Ut

Sm Sm M Sm Sm M Sm Sm Sm M M M M M

Stenosis Stenosis Stenosis Stenosis Stenosis Ulcer Ulcer Ulcer Ulcer PE PE PE RP RP

13M 7M 3M 16M 17M 14M 6M 12M 6M Unknown 18M 15M 7M 3M

Yes Yes No Yes No No No No No No Yes No Yes No

60 60 60 60 60 40 56 55.4 50 40 60 66 66 50

20

5

20 10.5 10 10.5 14

5 3.5 3.3 3.5 3.5

10

3.3

80 60 60 60 60 60 66.5 65.4 60.5 54 60 66 66 60

92 60 60 60 60 72 69.7 67.9 63.7 58.2 60 66 66 62.5

Yes No No No No No No No No No No No No No

No Yes Yes Yes Yes No Yes Yes Yes Yes No No Yes No

LWD LWD NED DOO LWD TRD TRD NED TRD TRD DOO NED DOD NED

EBRT = external beam radiotherapy, ICBT = intra-cavitary brachytherapy, EQD2 = biological dose equivalent as 2 Gy, HT = hyperthermia, CT = chemotherapy, Ut = upper thoracic, Mt = middle thoracic, Lt = lower thoracic, PE = pericardial effusion, RP = radiation pneumonitis, LWD = live with disease, NED = no evidence of disease, DOD = death of current disease, DOO = death of other cause, TRD = treatment-related death, EQD2: alpha/beta = 3 for late reacting tissue.

239

T. Kodaira et al. / Radiotherapy and Oncology 95 (2010) 234–239 Table 5 Reports of radiotherapy for patients with superficial esophageal cancer. Authors

Submucosal lesions (n)

CRT (n)

ICBT (n)

EBRT (Gy/fx)

ICBT (Gy/fx)

5Y OAS (%)

5Y DFS (%)

Local control (%)

Esophageal AE = grade 3 (%)

94 21

NS 4

None None

43 13

60–70/NS 60–69/NS

10/2 8–12/2–3

32.9 (2Y) 50–85

NS NS

NS 45–85 (3Y)

8.3–8.9 14.3 (ICBT)

147 53

95 NS

5 10

92 NS

54–84/NS 40–61/NS

3–36 8–24/2–4

NS NS

NS 74–80

Akagi [16]

35

NS

None

35

50–61/NS

NS

74

0 (EBRT), 4–13 (ICBT) 30 (CRT + ICBT), 3 (RT + ICBT) 12

Yamada [5] Present

63 97

40 71

63 61

52 27

55–66/50–60 38–70.2/25– 30

13.3–27.3/2– 5 10–12/2–3 4–20/2–5

42–62 70–71 (2Y) 38 66.4 63

63.7 48.1

69.9 75.3

4.8 18.5 (ICBT), 5.7 (EBRT)

Okawa [4] Nishimura [3] Nemoto [7] Yorozu [10]

Number

CRT = chemoradiotherapy, ICBT = intra-cavitary brachytherapy, EBRT = external beam radiotherapy, OAS = overall survival, DFS = disease-free survival, AE = adverse event, NS = not stated.

tion toxicities following chemoradiotherapy for esophageal cancer. They have reported late adverse effects in patients treated with extended-field chemoradiotherapy. These patients received radiotherapy including the cervical, mediastinal, and upper abdominal regions. Multiportal three-dimensional conformal radiotherapy was not performed in this report. As another reason for limiting the CTV, we considered the pattern of recurrence in superficial esophageal cancer. Regional lymph node recurrence with control of the primary lesion was observed in a small number of our patients (six of 97; 6.1%). We could re-treat three of the six recurrences with additional radiotherapy. One patient had both local and nodal recurrences, and then underwent a salvage operation. The most frequent site of recurrent disease, including new esophageal tumors, was local site (25 of 97; 25.8%). Thus, we believe careful periodical screening is essential. For successful salvage after local recurrence, lesion limited to mucosal layer is identical because salvage EMR could easily undergo for such conditions. Because prognoses have improved for patients with superficial esophageal disease, we now apply three-dimensional conformal radiotherapy with multiport EBRT to reduce late pulmonary and cardiac toxicities. We believe this technique will help to achieve safer clinical outcome for these patients. In our cohort we found that the addition of chemotherapy did not improve survival rates; however, the selection bias of chemoradiotherapy for more invasive lesions could have affected the results. On multivariate analysis, the addition of chemotherapy was a significantly favorable factor for the LFF rate. We believe that chemoradiotherapy should be a standard treatment choice for patients with stage I disease if allowed by their medical condition. Conclusion We have reported outcomes at a single institute of the treatment with radiotherapy of patients with superficial esophageal cancer including mostly submucosal lesions. Our results are promising and suggest that chemoradiotherapy should be considered a standard treatment for stage I esophageal cancer. The ICBT would be carefully used not to increase severe esophageal morbidity, especially in case with chemoradiotherapy. Conflicts of interest statement The authors declare that there is no conflict of interest.

References [1] Center for Cancer Control and Information Services, National Cancer Center, Japan. [2] Registration Committee for Esophageal Cancer (ed.), Comprehensive registry of esophageal cancer in Japan, The Japan Society for Esophageal Diseases. [3] Nishimura Y, Okuno Y, Ono K, Mitsumori M, Nagata Y, Hiraoka M. External beam radiation therapy with or without high-dose-rate intraluminal brachytherapy for patients with superficial esophageal carcinoma. Cancer 1999;86:220–8. [4] Okawa T, Dokiya T, Nishio M, Hishikawa Y, Morita K. Multi-institutional randomized trial of external radiotherapy with and without intraluminal brachytherapy for esophageal cancer in Japan. Japanese Society of Therapeutic Radiology and Oncology (JASTRO) Study Group. Int J Radiat Oncol Biol Phys 1999;45:623–8. [5] Yamada K, Murakami M, Okamoto Y, et al. Treatment results of chemoradiotherapy for clinical stage I (T1N0M0) esophageal carcinoma. Int J Radiat Oncol Biol Phys 2006;64:1106–11. [6] Nemoto K, Matsumoto Y, Yamakawa M, et al. Treatment of superficial esophageal cancer by external radiation therapy alone: results of a multiinstitutional experience. Int J Radiat Oncol Biol Phys 2000;46:921–5. [7] Nemoto K, Yamada S, Hareyama M, Nagakura H, Hirokawa Y. Radiation therapy for superficial esophageal cancer: a comparison of radiotherapy methods. Int J Radiat Oncol Biol Phys 2001;50:639–44. [8] Kodaira T, Fuwa N, Kamata M, Furutani K, Tachibana H, Yamazaki T. Singleinstitute phase I/II trial of alternating chemoradiotherapy with 5-FU and nedaplatin for esophageal carcinoma. Anticancer Res 2006;26:471–8. [9] Kodaira T, Fuwa N, Itoh Y, et al. Multivariate analysis of treatment outcome in patients with esophageal carcinoma treated with definitive radiotherapy. Am J Clin Oncol 2003;26:392–7. [10] Yorozu A, Dokiya T, Oki Y. High-dose-rate brachytherapy boost following concurrent chemoradiotherapy for esophageal carcinoma. Int J Radiat Oncol Biol Phys 1999;45:271–5. [11] Fuwa N, Nomoto Y, Shouji K, Kodaira T, Kamata M, Ito Y. Therapeutic effects of simultaneous intraluminal irradiation and intraluminal hyperthermia on oesophageal carcinoma. Br J Radiol 2001;74:709–14. [12] Kaplan E, Meier P. Non-parametric estimation from incomplete observation. J Am Stat Assoc 1958;53:457–81. [13] Mantle N. Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Rep 1966;50:163–70. [14] Cox D. Regression models and life table. J R Stat Soc 1972;34:197–220. [15] Shinoda M, Hatooka S, Mori S, Mitsudomi T. Clinical aspects of multimodality therapy for resectable locoregional esophageal cancer. Ann Thorac Cardiovasc Surg 2006;12:234–41. [16] Akagi Y, Hirokawa Y, Kagemoto M, et al. Optimum fractionation for high-doserate endoesophageal brachytherapy following external irradiation of early stage esophageal cancer. Int J Radiat Oncol Biol Phys 1999;43:525–30. [17] Gaspar LE, Winter K, Kocha WI, Coia LR, Herskovic A, Graham M. A phase I/II study of external beam radiation, brachytherapy, and concurrent chemotherapy for patients with localized carcinoma of the esophagus (Radiation Therapy Oncology Group Study 9207): final report. Cancer 2000;88:988–95. [18] Ishikura S, Nihei K, Ohtsu A, et al. Long-term toxicity after definitive chemoradiotherapy for squamous cell carcinoma of the thoracic esophagus. J Clin Oncol 2003;21:2697–702.