- Email: [email protected]
Dose–response relationship in locoregional control for patients with stage II-III esophageal cancer treated with concurrent chemotherapy and radiotherapy
Int. J. Radiation Oncology Biol. Phys., Vol. 61, No. 3, pp. 656 – 664, 2005 Copyright © 2005 Elsevier Inc. Printed in the USA. All rights reserved 0360-3016/05/$–see front matter
doi:10.1016/j.ijrobp.2004.06.022
CLINICAL INVESTIGATION
Esophagus
DOSE–RESPONSE RELATIONSHIP IN LOCOREGIONAL CONTROL FOR PATIENTS WITH STAGE II-III ESOPHAGEAL CANCER TREATED WITH CONCURRENT CHEMOTHERAPY AND RADIOTHERAPY ZHEN ZHANG, M.D.,* ZHONGXING LIAO, M.D.,† JING JIN, M.D.,‡ JAFFER AJANI, M.D.,§ JOE Y. CHANG, M.D., PH.D.,† MELENDA JETER, M.D.,† THOMAS GUERRERO, M.D., PH.D.,† CRAIG W. STEVENS, M.D., PH.D.,† STEPHEN SWISHER, M.D.,¶ LINUS HO, M.D.,§ JAMES YAO, M.D.,§ PAMELA ALLEN, M.P.H.,† JAMES D. COX, M.D.,† AND RITSUKO KOMAKI, M.D.† *Department of Radiation Oncology, Shanghai Fudan University, Shanghai Medical University, Shanghai Cancer Hospital, Shanghai, People’s Republic of China; Departments of †Radiation Oncology, §Gastrointestinal Oncology, and ¶Thoracic and Cardiovascular Surgery, The University of Texas M. D. Anderson Cancer Center, Houston, TX; ‡Department of Radiation Oncology, Cancer Hospital, Peking Union Medical College and Chinese Academy of Medical Sciences, Beijing, People’s Republic of China Purpose: To evaluate the correlation between radiation dose and locoregional control (LRC) for patients with Stage II-III unresectable esophageal cancer treated with concurrent chemotherapy and radiotherapy. Methods and Materials: The medical records of 69 consecutive patients with clinical Stage II or III esophageal cancer treated with definitive chemoradiotherapy at the University of Texas M. D. Anderson Cancer Center between 1990 and 1998 were retrospectively reviewed. Of the 69 patients, 43 had received <51 Gy (lower dose group) and 26 >51 Gy (higher dose group). The median dose in the lower and higher dose groups was 30 Gy (range, 30 –51 Gy) and 59.4 Gy (range, 54 – 64.8 Gy), respectively. Two fractionation schedules were used: rapid fractionation, delivering 30 Gy at 3 Gy/fraction within 2 weeks, and standard fractionation, delivering >45 Gy at 1.8 –2 Gy/fraction daily. Total doses of <50 Gy were usually given with rapid fractionation. Cisplatin and 5-fluorouracil were administrated to 93% of the patients. Results: The patient characteristic that differed between the two groups was that patients in the lower dose group were more likely to have had weight loss >5% (46.2% vs. 23.3%). The lower dose group had more N1 tumors, but the tumor classification and stage grouping were similar in the two groups. The median follow-up time for all patients was 22 months (range, 2–56 months). Patients in the higher dose group had a statistically significant better 3-year local control rate (36% vs. 19%, p ⴝ 0.011), disease-free survival rate (25% vs. 10%, p ⴝ 0.004), and overall survival rate (13% vs. 3%, p ⴝ 0.054). A trend toward a better distant-metastasis-free survival rate was noted in the higher dose group (72% vs. 59%, p ⴝ 0.12). The complete clinical response rate was significantly greater in the higher dose group (46% vs. 23%, p ⴝ 0.048). In both groups, the most common type of first failure was persistence of the primary tumor. Significantly fewer patients in the higher dose group had tumor persistence after treatment (p ⴝ 0.02). No statistically significant difference was found between the two groups in the pattern of locoregional or distant failure. The long-term side effects of chemoradiotherapy were similar in the two groups, although it was difficult to assess the side effects accurately in a retrospective fashion. On multivariate analysis, Stage II (vs. III) disease and radiation dose >51 Gy were independent predictors of improved LRC, and locoregional failure was an independent predictor of worse overall survival. Conclusion: Our data suggested a positive correlation between radiation dose and LRC in the population studied. A higher radiation dose was associated with increased LRC and survival in the dose range studied. The data also suggested that better LRC was associated with a lower rate of distant metastasis. A threshold of tumor response to radiation dose might be present, as suggested by the flattened slope in the high-dose area on the dose–response curve. A carefully designed dose-escalation study is required to confirm this assumption. © 2005 Elsevier Inc. Esophageal cancer, Chemoradiotherapy, Radiation dose–response relation.
INTRODUCTION The treatment of esophageal cancer is difficult. Esophagectomy with or without neoadjuvant chemotherapy and
radiotherapy (RT) is the preferred treatment for patients whose tumors are surgically resectable. For patients whose tumors cannot be surgically removed and for those
Reprint requests to: Zhongxing Liao, M.D., Department of Radiation Oncology, Unit 97, The University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Blvd., Houston, TX 77030. Tel: (713) 792-8653; Fax: (713) 745-6694; E-mail: [email protected] Presented at the 88th Annual Meeting of the Radiological So-
ciety of North America, Chicago, IL, 2002. Supported by a Radiological Society of North America International Radiology Education Program Grant to “Teach the Teachers” from Emerging Nations. Received Jul 9, 2003, and in revised form Jun 14, 2004. Accepted for publication Jun 25, 2004. 656
Radiation dose in esophageal cancer
who have comorbid conditions preventing them from undergoing esophagectomy, the Radiation Therapy Oncology Group (RTOG) trial 85-01 established concurrent chemoradiotherapy as the standard treatment (1). However, the optimal radiation dose in the setting of concurrent chemotherapy has not been adequately defined. RTOG trial 94-05 investigated the effect of the radiation dose on locoregional control (LRC) and overall survival (OS) in patients with locally advanced esophageal cancer who underwent concurrent chemotherapy as the only treatment modality (2). However, 7 of the 11 patients in the high-dose arm died before they had received 50.4 Gy, and no statistically significant difference was found in patient outcome measured by LRC or survival between the high-dose and standard-dose arms. The purpose of the current study was to evaluate the dose–response relationship for LRC, distant metastasis-free survival (DMFS), disease-free survival (DFS), and OS in patients with esophageal cancer treated with chemoradiotherapy.
METHODS AND MATERIALS We searched the patient database of the Department of Radiation Oncology at the University of Texas M. D. Anderson Cancer Center to identify all patients with locally advanced esophageal cancer (Stage II or III according to the American Joint Committee on Cancer 1992 classification) treated with definitive chemoradiotherapy between 1990 and 1998. The search was limited to patients who had undergone biopsy and had histologically proven malignant esophageal cancer confirmed by a pathologist at our institution. A total of 69 consecutive patients met the search criteria and formed the basis of the current study. The RT records and hospital charts were reviewed for each patient for information about pretreatment evaluations, chemotherapy and RT, and outcomes.
Pretreatment evaluation The pretreatment evaluation included a medical history and physical examination, focusing on performance status and a history of smoking, alcohol intake, weight loss, and dysphagia. Laboratory studies included a complete blood cell count and biochemical survey, which included carcinoembryonic antigen measurement. The standard radiographic studies included barium swallow, chest radiography, and CT of the thorax, abdomen, and pelvis. Bone scans were performed when indicated. Esophagogastroduodenoscopy with biopsy was a standard part of the pretreatment evaluation, and endoscopic ultrasonography was performed as indicated.
Treatment and follow-up examinations Radiotherapy was delivered using 60Co or higher energy photons starting on Day 1 of chemotherapy. Two fractionation schedules were used: rapid-fractionation (30 Gy given in 10 fractions within 2 weeks) (3– 6) or standard fractionation (ⱖ45 Gy given at 1.8 –2 Gy/fraction daily). The initial target volume encompassed
● Z. ZHANG et al.
657
the primary tumor with a margin of at least 5 cm above and below the tumor and 2 cm radially. Subclinical disease was treated to at least 30 Gy using rapid fractionation (3– 6) or 45 Gy using standard fractionation before the field size was reduced. Two-dimensional or three-dimensional plans using dedicated CT for RT planning were obtained when possible. The individual and total doses were not corrected for inhomogeneity of the irradiated tissue. Most patients received cisplatin and 5-fluorouracil (5-FU) concurrently with RT. Cisplatin was administered at 20 mg/m2 on Days 1–5, and 5-FU was administered at 300 mg/m2/d as a continuous infusion, using a portable electronic pump Monday through Friday during RT. Patients did not receive chemotherapy on the weekends. Patients usually underwent follow-up examinations every 3– 4 months after treatment completion. Tumor response and nodal disease were evaluated with repeated CT scans, barium swallow studies, and endoscopy.
Definition of outcome A clinically complete response was defined as no clinical, radiographic, endoscopic, or histologic evidence of cancer on follow-up visits. When a discrepancy was found among the findings from the diagnostic procedures, the histologic, endoscopic, radiographic, and clinical evidence was used in decreasing order as the determining factors for defining the response. Survival duration was calculated from the date of diagnosis to that of the first occurrence of the considered event (locoregional recurrence, distant metastasis, or death). The duration of LRC was defined as the period from the date of diagnosis to the date of the first evidence of locoregional disease progression or recurrence. Survival durations were defined as the period from the date of diagnosis to the date of the first evidence of distant metastasis (DMFS); the first evidence of any treatment failure, either locoregional or distant (DFS); or death from any cause (OS).
Statistical analysis Patients were grouped by total radiation dose (ⱕ51 Gy and ⬎51 Gy). The data were analyzed using the Stata, version 7, statistical software program (7). Pearson’s chi-square test was used to assess measures of association in frequency tables. In addition, the survival function was performed using KaplanMeier estimates (8), and the log–rank test was used to assess the equality of the survival function across the groups. The equality of means for continuous variables was assessed using the t test. p ⱕ 0.05 was considered to be statistically significant. Statistical tests were based on a two-sided significance level. The Cox proportional hazards model (9) was used for multivariate analysis to assess the effect of patient characteristics and other prognostic factors on the endpoints. All variables with p ⬍ 0.25 on univariate analysis were entered into the model, and backward elimination was performed. The final model consisted of variables having a statistical significance value of ⱕ0.05 or biologic significance to the model. The estimated hazard was reported. The Wald test was used to assess the role of covariates in the model.
RESULTS Of the 69 patients in our study, 43 received ⱕ51 Gy (lower dose group) and 26 ⬎51 Gy (higher dose group). The pretreat-
658
I. J. Radiation Oncology
● Biology ● Physics
Volume 61, Number 3, 2005
Table 1. Pretreatment patient and tumor characteristics Characteristic Patients (n) Age (y) Median Range Gender (n) Male Female Karnofsky performance status score Median Range Weight loss (%) Median Range Histologic type (n) Adenocarcinoma Squamous carcinoma Other Tumor location (n) Cervical Upper Middle Lower Gastroesophageal junction Unknown T stage (n) T2 T3 T4 N stage (n) 0 1 Stage group (n) II III
Lower dose group (ⱕ51 Gy)
Higher dose group (⬎51 Gy)
43
26
66 24–87 y
67.5 45–82 y
0.42
30 (69.8) 13 (30.2)
19 (73.1) 7 (26.9)
0.77
90 50–100
90 50–100
0.88
4 0–33
8.5 0–23
0.03
16 (37.2) 26 (60.5) 1 (2.3)
4 (15.4) 21 (80.8) 1 (3.8)
0.12
6 (14.0) 7 (16.3) 13 (30.2) 8 (18.6) 7 (16.3) 2 (4.7)
4 (15.4) 6 (23.1) 10 (38.5) 4 (15.4) 2 (7.7) 0
0.83
8 (18.6) 26 (60.5) 9 (20.9)
9 (34.6) 15 (57.7) 2 (7.7)
0.18
16 (37.2) 27 (62.8)
17 (65.4) 9 (34.6)
0.023
16 (37.2) 27 (62.8)
15 (57.7) 11 (42.3)
0.097
p
Data in parentheses are percentages.
ment patient and tumor characteristics for the two groups are listed in Table 1. No statistically significant differences were found between the groups in age, gender, Karnofsky performance status (10), histologic subtype, tumor location, T stage, or clinical stage distribution. Statistically significant differences were present between the groups in the number of N1 tumors and percentage of weight loss, with more patients in the lower dose group having nodal involvement and a greater median percentage of weight loss.
Details of treatment and follow-up The chemotherapy regimens and the numbers of chemotherapy cycles patients received are shown in Table 2. The median radiation dose was 30 Gy (range, 30 –51 Gy) in the lower dose group and 59.4 Gy (range, 54 – 64.8 Gy) in the higher dose group, respectively. The median fraction size was 3 Gy (range, 1.8 –3 Gy) in the lower dose group and 1.8 Gy (range, 1.8 –2 Gy) in the higher dose group. The accelerated course of RT to 30 Gy at 3 Gy/fraction was our institutional policy during the early 1990s, even for patients
Table 2. Chemotherapy regimens
Characteristic Regimen 5-FU 5-FU ⫹ cisplatin 5-FU ⫹ cisplatin ⫹ other Other (not 5-FU) Total Cycles (n) 1 2 3 ⱖ4 Unknown Total
Lower dose group (ⱕ51 Gy) (n ⫽ 43)
Higher dose group (⬎51 Gy) (n ⫽ 26)
22 (51.2) 13 (30.2) 3 (7.0) 5 (11.6) 43 (100)
14 (53.8) 10 (38.5) 0 (0) 2 (7.7) 26 (100)
27 (62.8) 12 (27.9) 2 (4.7) 1 (2.3) 1 (2.3) 43 (100)
16 (61.5) 4 (15.4) 0 (0) 2 (7.7) 4 (15.4) 26 (100)
Abbreviation: 5-FU ⫽ 5-fluorouracil. Data presented as number of patients, with percentage in parentheses.
Radiation dose in esophageal cancer
● Z. ZHANG et al.
659
Fig. 1. Comparison of local-regional control (LRC), distant-metastasis-free survival (DMFS), disease-free survival (DFS), and overall survival (OS) between patients treated with ⱕ51 Gy and ⬎51 Gy. Statistically significant differences were found in LRC, DFS, and OS, favoring the higher dose group. A trend toward better DMFS was also found for the higher dose group.
receiving definitive chemoradiotherapy. The biologically effective dose of this schedule for the tumor is equivalent to 40 Gy at 2 Gy/fraction 5 d/wk (11). The median follow-up time for all patients was 22 months (range, 2–56 months).
Disease control and survival The duration of LRC, DMFS, DFS, and OS for the two patients groups is shown in Fig. 1. Statistically significant differences were found, favoring the higher dose group for all endpoints measured, except for DMFS. The median survival time was 9 months for the lower dose group and 14.5 months for higher dose group. Although the difference in DMFS was not statistically significant, a persistent trend was noted toward better DMFS in the higher dose group. The clinical primary tumor responses in the two groups are summarized in Table 3. The complete response rate was significantly greater in the higher dose group. Of the 69 patients, 45 (32 in the lower dose group and 13 in the higher dose group) developed locoregional recurrence. Fifteen patients had locoregional failure at two sites, and one had locoregional failure at three sites. The sites of first failure
are summarized in Table 4. The most common type of local failure in both groups was primary tumor persistence. However, significantly fewer patients in the higher dose group had tumor persistence after treatment. The most common sites of distant metastasis were lung, bone, and liver. No statistically significant differences were noted between the groups for the sites of distant metastasis. Table 3. Clinical primary tumor response Primary tumor response*
Lower dose group (ⱕ51 Gy) (n ⫽ 43)
Higher dose group (⬎51 Gy) (n ⫽ 26)
Complete response Partial response No response Disease progression Not evaluable Unknown Total
10 (23.3) 13 (30.2) 13 (30.2) 1 (2.3) 5 (16.6) 1 (2.3) 43 (100)
12 (46.2) 6 (23.1) 0 (0) 6 (23.1) 2 (7.7) 0 26 (100)
Data presented as number of patients, with percentage in parentheses. * p ⫽ 0.004.
660
I. J. Radiation Oncology
● Biology ● Physics
Table 4. Sites of first failure
Failure type Locoregional Persistent tumor In-field failure Out-of-field failure Distant metastasis Lung Bone Brain Liver Adrenals Not otherwise specified
Volume 61, Number 3, 2005
radiation dose. A positive correlation was found between radiation dose and local tumor control, although the slope of the dose–response curve became less steep in the high-dose region.
Lower dose group (ⱕ51 Gy) (n ⫽ 43)
Higher dose group (⬎51 Gy) (n ⫽ 26)
29 17 1
9 5 0
0.02 0.08
6 6 2 7 2 3
6 1 0 1 2 2
0.33 0.19 0.26 0.12 0.60 0.91
p
DISCUSSION
The side effects of treatment were similar in the two groups. One patient in the lower dose group had Grade 3 hematologic toxicity. One patient in the higher dose group had Grade 3 pneumonitis. One patient in each group had a Grade 3 fistula, and one in the higher dose group had a Grade 4 fistula. Four patients in the lower dose group and one in the higher dose group had Grade 3 stricture of the esophagus. The results of univariate analysis for LRC and OS are summarized in Tables 5 and 6, respectively. Clinical Stage II (vs. III) disease, complete response to chemoradiotherapy, and radiation dose ⬎51 Gy were statistically significant predictors of improved LRC. A lack of a complete response to chemoradiotherapy and locoregional failure were statistically significant predictors of worse OS. A trend was observed toward improved OS with a higher radiation dose, but this was not statistically significant (p ⫽ 0.063). The results of multivariate analyses for LRC and OS are summarized in Table 7. Stage II (vs. III) disease and a radiation dose ⱖ51 Gy were independent predictors of improved LRC, and locoregional failure was an independent predictor of worse OS. Figure 2 shows the percentage of patients with locoregional failure as a function of
The results of the current retrospective study are consistent with the observation that 50 Gy at 1.8 –2 Gy/ fraction within 5 weeks is adequate to control ⬎90% of subclinical disease in patients with squamous cell carcinoma of the upper aerodigestive tract. At least 60 –70 Gy given at the same fractionation is needed to treat gross tumors (12). Patients in our study who received a total dose ⬎51 Gy had significantly better LRC, DFS, and OS, with a trend toward better DMFS noted in the higher dose group. This finding indicates that eliminating local disease at the primary tumor site with high-dose radiation is important in reducing the risk of cancer dissemination distantly. The combination of increased LRC and decreased distant metastasis was associated with superior OS in patients treated with high-dose RT. However, the optimal radiation dose for treating esophageal cancer is controversial, especially when chemotherapy is delivered concurrently. Sun (13) reported a clear association between a higher radiation dose and improved 5-year survival when RT was the sole therapeutic modality in patients with Stage II or III esophageal cancer. The 5-year survival rate was 10.6% for patients who received 60 – 69 Gy and about 2% for those who received 50 –59 Gy (13). In the series of Coia and associates (14), patients received 5-FU, mitomycin C, and 60 Gy of radiation. Of importance, their trial was the only combined-modality trial in which patients with clinically early-stage esophageal cancer (Stage I and II) were treated and analyzed separately (14). The results of that trial demonstrated a very low local failure rate of 25%, a 5-year actuarial survival rate of 30%, and a 5-year actuarial local relapse-free survival rate of 70% for patients with Stage I disease. However, in a study of 30 patients with clinical Stage I–III disease, John et al. (15) reported
Table 5. Univariate Cox regression analysis for locoregional control Factor
Hazard ratio
p ⬎ |Z|
95% CI
Age ⬍60 y Nonwhite race (vs. white) Male gender KPS 70–80 (vs. 90–100) Weight loss ⬎5% Tumor in lower esophagus or GEJ (vs. upper or middle esophagus) Nonsquamous subtype (vs. squamous) Stage II (vs.III) No clinical CR Distant metastases Radiation dose ⬎51 Gy (vs. ⱕ51 Gy)
1.00 1.39 1.59 0.98 1.02 1.19 0.97 0.44 9.26 0.87 0.45
0.85 0.29 0.19 0.07 0.13 0.58 0.92 0.008 0.000 0.69 0.017
0.97–1.03 0.76–2.55 0.79–3.20 0.95–1.03 0.99–1.05 0.65–2.16 0.52–1.80 0.24–0.81 3.45–24.88 0.44–1.72 0.24–0.87
Abbreviations: CI ⫽ confidence interval; KPS ⫽ Karnofsky performance status; GEJ ⫽ gastroesophageal junction; CR ⫽ complete response.
Radiation dose in esophageal cancer
● Z. ZHANG et al.
661
Table 6. Univariate Cox regression analysis for overall survival Factor
Hazard ratio
p ⬎ |Z|
95% CI
Age ⬍60 y Nonwhite race (vs. white) Gender KPS 70–80 (vs. 90–100) Weight loss ⬎5% Tumor in lower esophagus or GEJ (vs. upper or middle esophagus) Nonsquamous subtype (vs. squamous) Stage II (vs. III) No clinical CR Distant metastases Locoregional failure Radiation dose ⬎51 Gy (vs. ⱕ51 Gy)
0.59 1.14 1.68 0.98 0.99 0.99 0.89 0.76 2.32 1.22 2.40 0.60
0.12 0.64 0.07 0.20 0.97 0.99 0.66 0.29 0.02 0.48 0.04 0.063
3.11–1.13 0.66–1.96 0.95–2.98 0.96–1.00 0.60–1.64 0.58–1.68 0.52–1.51 0.46–1.26 1.17–4.58 0.70–2.13 1.33–4.34 0.35–1.03
Abbreviations as in Table 5.
a similar local failure rate 27% with a lower radiation dose: 40 –50 Gy. The 2-year actuarial survival rate was 29%. Radiation doses of as much as 66 Gy after three cycles of cisplatin and bleomycin have also been used (16). RTOG trial 85-01 (1) established that a radiation dose of 50 Gy is appropriate for patients with esophageal cancer treated with simultaneous cisplatin-based chemotherapy. In that trial, patients received four cycles of 5-FU (1000 mg/m2 on Days 1– 4) and cisplatin (75 mg/m2 on Day 1). It should be emphasized that Cycles 1 and 2 of chemotherapy were given every 4 weeks and Cycles 3 and 4 were given every 3 weeks. This dose intensification after the combined-modality segment may explain, in part, why patients had difficulty completing all four cycles of chemotherapy. RT (50 Gy) was given concurrently with chemotherapy beginning on Day 1. A higher dose of radiation (64 Gy) was used in the RT control arm. At a minimum of 5 years of follow-up, the OS rate was 26% for patients who received combinedmodality therapy and 0% for those who received RT alone. Disease persistence was the most common mode of treatment failure; however, it was less common in the group receiving combined therapy (26%) than in the RT-alone group (37%). Distant metastasis accounted for the first site of treatment failure in 30% of the RT-alone group compared with 16% of the combined-modality group (1). The positive results of RTOG trial 85-01 established combined-modality therapy rather than RT alone as the conventional nonoperative treatment for esophageal cancer. However, despite the positive results in the RTOG 85-01 combined-modality arm, the local failure rate was high (50% at 5 years) and the survival rate was modest (27% at 5 years). RTOG trial 94-05 (2), a follow-up to RTOG 85-01, investigated the possibility of intensification of the radiation dose. In the 94-05 trial, 236 patients with cT1– T4NxM0 squamous cell carcinoma (85%) or adenocarcinoma (15%) of the esophagus without tumor extension to within 2 cm of the stomach were randomized to standarddose combined-modality therapy using a slight modifi-
cation of the combined-modality therapy arm of RTOG 85-01 (50.4 Gy plus concurrent 5-FU and cisplatin on Weeks 1 and 4, repeated 4 weeks after RT completion) or high-dose RT (64.8 Gy) and the same chemotherapy regimen. That trial failed to show any benefit in terms of survival in the high-dose arm. No statistically significant differences were found between the high-dose and standard-dose arms in the median survival time (13.0 vs. 18.1 months), 2-year survival rate (31% vs. 40%), or rate of locoregional failure or locoregional disease persistence (52% vs. 56%). On the basis of these two well-designed prospective randomized trials, 50.4 Gy at 1.8 Gy/fraction for 5 d/wk is currently considered standard when cisplatin-based chemotherapy is administered concomitantly to patients with esophageal cancer treated with a nonoperative approach (1, 2). However, 7 of the 11 treatment-related deaths in the high-dose arm of the RTOG 94-05 trial occurred in patients who had received ⱕ50.4 Gy. In addition, a statistically significant prolongation of treatment time occurred because of breaks required for recovery from side effects after correction for the number of RT sessions and a statistically significant lower dose of 5-FU was given to patients in the high-dose arm. The authors believed that these factors might have contributed, at least in part, to the lack of benefit for patients who received high-dose vs. standard-dose RT (2). Therefore, the findings from the RTOG 94-05 trial were inconclusive regarding whether a radiation dose effect exists in the treatment of cancer of the esophagus. The findings of RTOG 94-05 warrant extensive research on methods to reduce treatment toxicity to allow radiation dose intensification for patients who are not considered candidates for surgery and for whom chemoradiotherapy is the only treatment alternative. Our findings suggest the existence of a correlation between a higher radiation dose and improved LRC when RT is given concurrently with chemotherapy in treating cancer of the esophagus, although the slope of the curve
662
I. J. Radiation Oncology
● Biology ● Physics
Volume 61, Number 3, 2005
Fig. 2. Percentage of patients with locoregional failure 2 years after treatment as a function of radiation dose.
flattened in the high-dose region. The shallower slope of the curve in the high-dose area also suggests the presence of a threshold in the tumor response, such that additional increases in the radiation dose would not improve local tumor control further. A carefully designed dose-escalation study using LRC as the primary endpoint would be required to confirm this assumption. It is worth pointing out that patients received 30 Gy of total radiation given with rapid fractionation, which was, at one time, the treatment chosen in our institution. This program was not meant to be palliative; rather, it was based on the principle that the total radiation dose required to obtain a given biologic effect decreases as the dose per fraction increases. A 30-Gy total dose of radiation given in 10 fractions was considered radiobiologically equivalent to
Table 7. Multivariate Cox regression analysis for locoregional control and overall survival Factor LRC Radiation dose ⬎ 51 Gy (vs. ⱕ51 Gy) Clinical stage II (vs. III) OS Radiation dose ⬎51 Gy (vs. ⱕ51 Gy) Locoregional failure
Hazard ratio
p ⬎ |Z|
95% CI
0.48
0.029
0.25–0.93
0.42
0.007
0.22–0.79
0.69
0.193
0.39–1.21
2.13
0.012
1.19–4.04
Abbreviations: CI ⫽ confidence interval; LRC ⫽ locoregional control; OS ⫽ overall survival.
a standard 5.5-week (50.4 Gy in 28 fractions) program with a shortened overall treatment time. It was designed as a definitive treatment, combined with concurrent chemotherapy for esophageal, pancreatic, and periampullary carcinomas and was used for prospective clinical trials during the early 1990s at the M. D. Anderson Cancer Center (3– 6). However, our data suggested that 30 Gy in 10 fractions was not equivalent to 50.4 Gy in 28 fractions for LRC, even though the lower dose was given in a larger fraction within a shorter overall treatment time. The institutional treatment policy has been changed to use at least 50 Gy for patients receiving definitive chemoradiotherapy since the mid-1990s. It is also important to realize the limitations and drawbacks of a retrospective study, such as the current one, and to be cautious in drawing definitive conclusions from such a study. Although measures were taken to minimize the difference between the two populations, this study has the inherent limitations of a retrospective study in terms of patient selection bias and treatment heterogeneity. A meta-analysis by Ancona et al. (17) provided evidence of a relationship between a higher radiation dose and increased likelihood of a pathologic complete response. In studies of preoperative chemoradiotherapy in patients with esophageal cancer, the 5-year survival rate ranged from 46% to 60% for complete responders, and a statistically significant survival advantage appeared to be limited to patients who showed a complete response to preoperative treatment and subsequently underwent curative resection (17–20). It remains unclear whether patients who respond to chemoradiotherapy and do not undergo surgery have the potential for long-term sur-
Radiation dose in esophageal cancer
vival. In the present study, the clinical complete primary tumor response rate was 47% in the higher dose group (p ⫽ 0.006). The relatively high complete response rate for patients treated with ⬎51 Gy, along with the superior median survival time and 5-year survival observed in these patients, supports the use of high-dose RT with concurrent chemotherapy. That ⬎50% of patients in the RTOG studies had locoregional failure indicated the need for better locoregional treatment. At our institution, a clinical trial of a selective cyclooxygease-2 inhibitor as a radiosensitization agent is currently underway. The use of intensity-modulated RT to reduce the volume of normal tissue irradiated is also being tested in our clinic. The use of cytoprotective agents such as amifostine in esophageal cancer patients is also an active area of research. The hope is to find ways to increase tumor response and reduce radiation-induced damage to normal tissue, making escalation of the radiation dose either clinically feasible or unnecessary. CONCLUSION In the present study, statistically significant better LRC, DFS, and OS were seen in patients who received ⬎51 Gy. In a multivariate analysis, clinical Stage II disease and a radiation dose ⬎51 Gy were independent predictors of improved LRC, and locoregional failure was an independent predictor of worse OS. The data also
● Z. ZHANG et al.
663
suggested that better LRC was associated with a lower risk of distant metastasis. A threshold of tumor response to radiation dose might exist, as suggested by a flattened slope in the high-dose area on the dose–response curve. A carefully designed dose-escalation study is required to confirm this assumption. The side effects were minimal in the studied population and the differences were not statistically significant between the lower dose and higher dose groups. However, the accuracy of the data on toxicity was limited by the retrospective nature of this study. Methods to reduce treatment toxicity and increase the tumor response to RT, thereby increasing the therapeutic ratio, are needed. Modern radiation techniques, such as conformal therapy or intensity-modulated RT, may help to allow dose escalation by minimizing the normal tissue dose around the esophagus. The introduction of a mucosal protective agent such as amifostine may also help to allow further dose escalation. The combination of molecularly targeted therapy, such as cyclooxygenase-2 inhibition to increase tumor radioresponse, with the methods mentioned above may facilitate an increase in the therapeutic ratio. Our data indicate a need for additional study of the optimal radiation dose in a prospective randomized trial, with emphasis on decreasing the treatment-related toxicity of concurrent chemoradiotherapy as the only treatment modality in patients with locally advanced esophageal cancer.
REFERENCES 1. Cooper J, Guo M, Herskovic A, et al., for the Radiation Therapy Oncology Group. Chemoradiotherapy of locally advanced esophageal cancer: Long-term follow-up of a prospective randomized trial (RTOG 85-01). JAMA 1999;281:1623–1627. 2. Minsky B, Pajak T, Ginsberg R, et al. INT 0123 (Radiation Therapy Oncology Group 94-05) phase III trial of combinedmodality therapy for esophageal cancer: High-dose versus standard-dose radiation therapy. J Clin Oncol 2002;20:1167–1174. 3. Spitz F, Abbruzzese J, Lee J, et al. Preoperative and postoperative chemoradiation strategies in patients treated with pancreaticoduodenectomy for adenocarcinoma of the pancreas. J Clin Oncol 1997;15:928 –937. 4. Pisters P, Abbruzzese J, Janjan N, et al. Rapid-fractionation preoperative chemoradiation, pancreaticoduodenectomy, and intraoperative radiation therapy for resectable pancreatic adenocarcinoma. J Clin Oncol 1998;16:3843–3850. 5. Evans D, Abbruzzese J, Cleary K, et al. Rapid-fractionation pre-operative chemoradiation for malignant periampullary neoplasms. J Roy Coll Surg Edinb 1995;40:319 –323. 6. Ajani J, Ryan B, Rich T, et al. Prolonged chemotherapy for localised squamous carcinoma of the oesophagus. Eur J Cancer 1992;28A:880 – 884. 7. Stata Statistical Software. Release 7.0. College Station, TX: Stata Corporation; 2001. 8. Kaplan E, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;53:457– 481. 9. Cox D. Regression models and life tables. J R Stat Soc [Ser B] 1972;34:187–229. 10. Karnofsky D, Burchenal J. The clinical evaluation of chemo-
11. 12.
13. 14. 15. 16.
17.
therapeutic agents in cancer. In: MaCleod C, editor. Evaluation of chemotherapeutic agents. Symposium, microbiology section. New York Academy of Medicine. Vol 1. New York: Columbia University Press; 1949. p. 191–205. Hall E. Radiobiology for the radiologist. Philadelphia: JB Lippincott; 2000. p. 223–227. Withers HR, Peters LJ. Basic principles of radiotherapy: Basic clinical parameters. In: Fletcher GA, editor. Textbook of radiotherapy. 3rd ed. Philadelphia: Lea & Febiger; 1980. p. 180. Sun D. Ten-year follow-up of esophageal cancer treated by radical radiation therapy: Analysis of 869 patients. Int J Radiat Oncol Biol Phys 1989;16:329 –334. Coia L, Engstrom P, Paul A. Long term results of infusional 5-FU, mitomycin-C and radiation as primary management of esophageal cancer. Int J Radiat Oncol Biol Phys 1991;20:29 –36. John M, Flam M, Ager Mowry P, et al. Radiation alone and chemoradiation for nonmetastatic esophageal carcinoma. Cancer 1989;63:2397–2403. Izquierdo MA, Marcuello E, Gomez de Segura G, et al. Unresectable nonmetastatic squamous cell carcinoma of the esophagus managed by sequential chemotherapy (cisplatin and bleomycin) and radiation therapy. Cancer 1993;71:287–292. Ancona E, Ruol A, Santi S, et al. Only pathologic complete response to neoadjuvant chemotherapy improves significantly the long term survival of patients with resectable esophageal squamous cell carcinoma: Final report of a randomized, controlled trial of preoperative chemotherapy versus surgery alone. Cancer 2001;91:2165–2174.
664
I. J. Radiation Oncology
● Biology ● Physics
18. Kitamura K, Kuwano H, Araki K, et al. Clinicopathologic features of patients with oesophageal cancer obtaining a histological complete response for preoperative hyperthermo-chemo-radiotherapy. Int J Hyperthermia 1998;14: 233–243. 19. Vogel SB, Mendenhall WM, Sombeck MD, et al. Downstag-
Volume 61, Number 3, 2005
ing of esophageal cancer after preoperative radiation and chemotherapy. Ann Surg 1995;221:685– 693. 20. Forastiere A, Orringer M, Perez-Tamayo C, et al. Preoperative chemoradiation followed by transhiatal esophagectomy for carcinoma of the esophagus: Final report. J Clin Oncol 1993; 11:1118 –1123.
doi:10.1016/j.ijrobp.2004.06.022
CLINICAL INVESTIGATION
Esophagus
DOSE–RESPONSE RELATIONSHIP IN LOCOREGIONAL CONTROL FOR PATIENTS WITH STAGE II-III ESOPHAGEAL CANCER TREATED WITH CONCURRENT CHEMOTHERAPY AND RADIOTHERAPY ZHEN ZHANG, M.D.,* ZHONGXING LIAO, M.D.,† JING JIN, M.D.,‡ JAFFER AJANI, M.D.,§ JOE Y. CHANG, M.D., PH.D.,† MELENDA JETER, M.D.,† THOMAS GUERRERO, M.D., PH.D.,† CRAIG W. STEVENS, M.D., PH.D.,† STEPHEN SWISHER, M.D.,¶ LINUS HO, M.D.,§ JAMES YAO, M.D.,§ PAMELA ALLEN, M.P.H.,† JAMES D. COX, M.D.,† AND RITSUKO KOMAKI, M.D.† *Department of Radiation Oncology, Shanghai Fudan University, Shanghai Medical University, Shanghai Cancer Hospital, Shanghai, People’s Republic of China; Departments of †Radiation Oncology, §Gastrointestinal Oncology, and ¶Thoracic and Cardiovascular Surgery, The University of Texas M. D. Anderson Cancer Center, Houston, TX; ‡Department of Radiation Oncology, Cancer Hospital, Peking Union Medical College and Chinese Academy of Medical Sciences, Beijing, People’s Republic of China Purpose: To evaluate the correlation between radiation dose and locoregional control (LRC) for patients with Stage II-III unresectable esophageal cancer treated with concurrent chemotherapy and radiotherapy. Methods and Materials: The medical records of 69 consecutive patients with clinical Stage II or III esophageal cancer treated with definitive chemoradiotherapy at the University of Texas M. D. Anderson Cancer Center between 1990 and 1998 were retrospectively reviewed. Of the 69 patients, 43 had received <51 Gy (lower dose group) and 26 >51 Gy (higher dose group). The median dose in the lower and higher dose groups was 30 Gy (range, 30 –51 Gy) and 59.4 Gy (range, 54 – 64.8 Gy), respectively. Two fractionation schedules were used: rapid fractionation, delivering 30 Gy at 3 Gy/fraction within 2 weeks, and standard fractionation, delivering >45 Gy at 1.8 –2 Gy/fraction daily. Total doses of <50 Gy were usually given with rapid fractionation. Cisplatin and 5-fluorouracil were administrated to 93% of the patients. Results: The patient characteristic that differed between the two groups was that patients in the lower dose group were more likely to have had weight loss >5% (46.2% vs. 23.3%). The lower dose group had more N1 tumors, but the tumor classification and stage grouping were similar in the two groups. The median follow-up time for all patients was 22 months (range, 2–56 months). Patients in the higher dose group had a statistically significant better 3-year local control rate (36% vs. 19%, p ⴝ 0.011), disease-free survival rate (25% vs. 10%, p ⴝ 0.004), and overall survival rate (13% vs. 3%, p ⴝ 0.054). A trend toward a better distant-metastasis-free survival rate was noted in the higher dose group (72% vs. 59%, p ⴝ 0.12). The complete clinical response rate was significantly greater in the higher dose group (46% vs. 23%, p ⴝ 0.048). In both groups, the most common type of first failure was persistence of the primary tumor. Significantly fewer patients in the higher dose group had tumor persistence after treatment (p ⴝ 0.02). No statistically significant difference was found between the two groups in the pattern of locoregional or distant failure. The long-term side effects of chemoradiotherapy were similar in the two groups, although it was difficult to assess the side effects accurately in a retrospective fashion. On multivariate analysis, Stage II (vs. III) disease and radiation dose >51 Gy were independent predictors of improved LRC, and locoregional failure was an independent predictor of worse overall survival. Conclusion: Our data suggested a positive correlation between radiation dose and LRC in the population studied. A higher radiation dose was associated with increased LRC and survival in the dose range studied. The data also suggested that better LRC was associated with a lower rate of distant metastasis. A threshold of tumor response to radiation dose might be present, as suggested by the flattened slope in the high-dose area on the dose–response curve. A carefully designed dose-escalation study is required to confirm this assumption. © 2005 Elsevier Inc. Esophageal cancer, Chemoradiotherapy, Radiation dose–response relation.
INTRODUCTION The treatment of esophageal cancer is difficult. Esophagectomy with or without neoadjuvant chemotherapy and
radiotherapy (RT) is the preferred treatment for patients whose tumors are surgically resectable. For patients whose tumors cannot be surgically removed and for those
Reprint requests to: Zhongxing Liao, M.D., Department of Radiation Oncology, Unit 97, The University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Blvd., Houston, TX 77030. Tel: (713) 792-8653; Fax: (713) 745-6694; E-mail: [email protected] Presented at the 88th Annual Meeting of the Radiological So-
ciety of North America, Chicago, IL, 2002. Supported by a Radiological Society of North America International Radiology Education Program Grant to “Teach the Teachers” from Emerging Nations. Received Jul 9, 2003, and in revised form Jun 14, 2004. Accepted for publication Jun 25, 2004. 656
Radiation dose in esophageal cancer
who have comorbid conditions preventing them from undergoing esophagectomy, the Radiation Therapy Oncology Group (RTOG) trial 85-01 established concurrent chemoradiotherapy as the standard treatment (1). However, the optimal radiation dose in the setting of concurrent chemotherapy has not been adequately defined. RTOG trial 94-05 investigated the effect of the radiation dose on locoregional control (LRC) and overall survival (OS) in patients with locally advanced esophageal cancer who underwent concurrent chemotherapy as the only treatment modality (2). However, 7 of the 11 patients in the high-dose arm died before they had received 50.4 Gy, and no statistically significant difference was found in patient outcome measured by LRC or survival between the high-dose and standard-dose arms. The purpose of the current study was to evaluate the dose–response relationship for LRC, distant metastasis-free survival (DMFS), disease-free survival (DFS), and OS in patients with esophageal cancer treated with chemoradiotherapy.
METHODS AND MATERIALS We searched the patient database of the Department of Radiation Oncology at the University of Texas M. D. Anderson Cancer Center to identify all patients with locally advanced esophageal cancer (Stage II or III according to the American Joint Committee on Cancer 1992 classification) treated with definitive chemoradiotherapy between 1990 and 1998. The search was limited to patients who had undergone biopsy and had histologically proven malignant esophageal cancer confirmed by a pathologist at our institution. A total of 69 consecutive patients met the search criteria and formed the basis of the current study. The RT records and hospital charts were reviewed for each patient for information about pretreatment evaluations, chemotherapy and RT, and outcomes.
Pretreatment evaluation The pretreatment evaluation included a medical history and physical examination, focusing on performance status and a history of smoking, alcohol intake, weight loss, and dysphagia. Laboratory studies included a complete blood cell count and biochemical survey, which included carcinoembryonic antigen measurement. The standard radiographic studies included barium swallow, chest radiography, and CT of the thorax, abdomen, and pelvis. Bone scans were performed when indicated. Esophagogastroduodenoscopy with biopsy was a standard part of the pretreatment evaluation, and endoscopic ultrasonography was performed as indicated.
Treatment and follow-up examinations Radiotherapy was delivered using 60Co or higher energy photons starting on Day 1 of chemotherapy. Two fractionation schedules were used: rapid-fractionation (30 Gy given in 10 fractions within 2 weeks) (3– 6) or standard fractionation (ⱖ45 Gy given at 1.8 –2 Gy/fraction daily). The initial target volume encompassed
● Z. ZHANG et al.
657
the primary tumor with a margin of at least 5 cm above and below the tumor and 2 cm radially. Subclinical disease was treated to at least 30 Gy using rapid fractionation (3– 6) or 45 Gy using standard fractionation before the field size was reduced. Two-dimensional or three-dimensional plans using dedicated CT for RT planning were obtained when possible. The individual and total doses were not corrected for inhomogeneity of the irradiated tissue. Most patients received cisplatin and 5-fluorouracil (5-FU) concurrently with RT. Cisplatin was administered at 20 mg/m2 on Days 1–5, and 5-FU was administered at 300 mg/m2/d as a continuous infusion, using a portable electronic pump Monday through Friday during RT. Patients did not receive chemotherapy on the weekends. Patients usually underwent follow-up examinations every 3– 4 months after treatment completion. Tumor response and nodal disease were evaluated with repeated CT scans, barium swallow studies, and endoscopy.
Definition of outcome A clinically complete response was defined as no clinical, radiographic, endoscopic, or histologic evidence of cancer on follow-up visits. When a discrepancy was found among the findings from the diagnostic procedures, the histologic, endoscopic, radiographic, and clinical evidence was used in decreasing order as the determining factors for defining the response. Survival duration was calculated from the date of diagnosis to that of the first occurrence of the considered event (locoregional recurrence, distant metastasis, or death). The duration of LRC was defined as the period from the date of diagnosis to the date of the first evidence of locoregional disease progression or recurrence. Survival durations were defined as the period from the date of diagnosis to the date of the first evidence of distant metastasis (DMFS); the first evidence of any treatment failure, either locoregional or distant (DFS); or death from any cause (OS).
Statistical analysis Patients were grouped by total radiation dose (ⱕ51 Gy and ⬎51 Gy). The data were analyzed using the Stata, version 7, statistical software program (7). Pearson’s chi-square test was used to assess measures of association in frequency tables. In addition, the survival function was performed using KaplanMeier estimates (8), and the log–rank test was used to assess the equality of the survival function across the groups. The equality of means for continuous variables was assessed using the t test. p ⱕ 0.05 was considered to be statistically significant. Statistical tests were based on a two-sided significance level. The Cox proportional hazards model (9) was used for multivariate analysis to assess the effect of patient characteristics and other prognostic factors on the endpoints. All variables with p ⬍ 0.25 on univariate analysis were entered into the model, and backward elimination was performed. The final model consisted of variables having a statistical significance value of ⱕ0.05 or biologic significance to the model. The estimated hazard was reported. The Wald test was used to assess the role of covariates in the model.
RESULTS Of the 69 patients in our study, 43 received ⱕ51 Gy (lower dose group) and 26 ⬎51 Gy (higher dose group). The pretreat-
658
I. J. Radiation Oncology
● Biology ● Physics
Volume 61, Number 3, 2005
Table 1. Pretreatment patient and tumor characteristics Characteristic Patients (n) Age (y) Median Range Gender (n) Male Female Karnofsky performance status score Median Range Weight loss (%) Median Range Histologic type (n) Adenocarcinoma Squamous carcinoma Other Tumor location (n) Cervical Upper Middle Lower Gastroesophageal junction Unknown T stage (n) T2 T3 T4 N stage (n) 0 1 Stage group (n) II III
Lower dose group (ⱕ51 Gy)
Higher dose group (⬎51 Gy)
43
26
66 24–87 y
67.5 45–82 y
0.42
30 (69.8) 13 (30.2)
19 (73.1) 7 (26.9)
0.77
90 50–100
90 50–100
0.88
4 0–33
8.5 0–23
0.03
16 (37.2) 26 (60.5) 1 (2.3)
4 (15.4) 21 (80.8) 1 (3.8)
0.12
6 (14.0) 7 (16.3) 13 (30.2) 8 (18.6) 7 (16.3) 2 (4.7)
4 (15.4) 6 (23.1) 10 (38.5) 4 (15.4) 2 (7.7) 0
0.83
8 (18.6) 26 (60.5) 9 (20.9)
9 (34.6) 15 (57.7) 2 (7.7)
0.18
16 (37.2) 27 (62.8)
17 (65.4) 9 (34.6)
0.023
16 (37.2) 27 (62.8)
15 (57.7) 11 (42.3)
0.097
p
Data in parentheses are percentages.
ment patient and tumor characteristics for the two groups are listed in Table 1. No statistically significant differences were found between the groups in age, gender, Karnofsky performance status (10), histologic subtype, tumor location, T stage, or clinical stage distribution. Statistically significant differences were present between the groups in the number of N1 tumors and percentage of weight loss, with more patients in the lower dose group having nodal involvement and a greater median percentage of weight loss.
Details of treatment and follow-up The chemotherapy regimens and the numbers of chemotherapy cycles patients received are shown in Table 2. The median radiation dose was 30 Gy (range, 30 –51 Gy) in the lower dose group and 59.4 Gy (range, 54 – 64.8 Gy) in the higher dose group, respectively. The median fraction size was 3 Gy (range, 1.8 –3 Gy) in the lower dose group and 1.8 Gy (range, 1.8 –2 Gy) in the higher dose group. The accelerated course of RT to 30 Gy at 3 Gy/fraction was our institutional policy during the early 1990s, even for patients
Table 2. Chemotherapy regimens
Characteristic Regimen 5-FU 5-FU ⫹ cisplatin 5-FU ⫹ cisplatin ⫹ other Other (not 5-FU) Total Cycles (n) 1 2 3 ⱖ4 Unknown Total
Lower dose group (ⱕ51 Gy) (n ⫽ 43)
Higher dose group (⬎51 Gy) (n ⫽ 26)
22 (51.2) 13 (30.2) 3 (7.0) 5 (11.6) 43 (100)
14 (53.8) 10 (38.5) 0 (0) 2 (7.7) 26 (100)
27 (62.8) 12 (27.9) 2 (4.7) 1 (2.3) 1 (2.3) 43 (100)
16 (61.5) 4 (15.4) 0 (0) 2 (7.7) 4 (15.4) 26 (100)
Abbreviation: 5-FU ⫽ 5-fluorouracil. Data presented as number of patients, with percentage in parentheses.
Radiation dose in esophageal cancer
● Z. ZHANG et al.
659
Fig. 1. Comparison of local-regional control (LRC), distant-metastasis-free survival (DMFS), disease-free survival (DFS), and overall survival (OS) between patients treated with ⱕ51 Gy and ⬎51 Gy. Statistically significant differences were found in LRC, DFS, and OS, favoring the higher dose group. A trend toward better DMFS was also found for the higher dose group.
receiving definitive chemoradiotherapy. The biologically effective dose of this schedule for the tumor is equivalent to 40 Gy at 2 Gy/fraction 5 d/wk (11). The median follow-up time for all patients was 22 months (range, 2–56 months).
Disease control and survival The duration of LRC, DMFS, DFS, and OS for the two patients groups is shown in Fig. 1. Statistically significant differences were found, favoring the higher dose group for all endpoints measured, except for DMFS. The median survival time was 9 months for the lower dose group and 14.5 months for higher dose group. Although the difference in DMFS was not statistically significant, a persistent trend was noted toward better DMFS in the higher dose group. The clinical primary tumor responses in the two groups are summarized in Table 3. The complete response rate was significantly greater in the higher dose group. Of the 69 patients, 45 (32 in the lower dose group and 13 in the higher dose group) developed locoregional recurrence. Fifteen patients had locoregional failure at two sites, and one had locoregional failure at three sites. The sites of first failure
are summarized in Table 4. The most common type of local failure in both groups was primary tumor persistence. However, significantly fewer patients in the higher dose group had tumor persistence after treatment. The most common sites of distant metastasis were lung, bone, and liver. No statistically significant differences were noted between the groups for the sites of distant metastasis. Table 3. Clinical primary tumor response Primary tumor response*
Lower dose group (ⱕ51 Gy) (n ⫽ 43)
Higher dose group (⬎51 Gy) (n ⫽ 26)
Complete response Partial response No response Disease progression Not evaluable Unknown Total
10 (23.3) 13 (30.2) 13 (30.2) 1 (2.3) 5 (16.6) 1 (2.3) 43 (100)
12 (46.2) 6 (23.1) 0 (0) 6 (23.1) 2 (7.7) 0 26 (100)
Data presented as number of patients, with percentage in parentheses. * p ⫽ 0.004.
660
I. J. Radiation Oncology
● Biology ● Physics
Table 4. Sites of first failure
Failure type Locoregional Persistent tumor In-field failure Out-of-field failure Distant metastasis Lung Bone Brain Liver Adrenals Not otherwise specified
Volume 61, Number 3, 2005
radiation dose. A positive correlation was found between radiation dose and local tumor control, although the slope of the dose–response curve became less steep in the high-dose region.
Lower dose group (ⱕ51 Gy) (n ⫽ 43)
Higher dose group (⬎51 Gy) (n ⫽ 26)
29 17 1
9 5 0
0.02 0.08
6 6 2 7 2 3
6 1 0 1 2 2
0.33 0.19 0.26 0.12 0.60 0.91
p
DISCUSSION
The side effects of treatment were similar in the two groups. One patient in the lower dose group had Grade 3 hematologic toxicity. One patient in the higher dose group had Grade 3 pneumonitis. One patient in each group had a Grade 3 fistula, and one in the higher dose group had a Grade 4 fistula. Four patients in the lower dose group and one in the higher dose group had Grade 3 stricture of the esophagus. The results of univariate analysis for LRC and OS are summarized in Tables 5 and 6, respectively. Clinical Stage II (vs. III) disease, complete response to chemoradiotherapy, and radiation dose ⬎51 Gy were statistically significant predictors of improved LRC. A lack of a complete response to chemoradiotherapy and locoregional failure were statistically significant predictors of worse OS. A trend was observed toward improved OS with a higher radiation dose, but this was not statistically significant (p ⫽ 0.063). The results of multivariate analyses for LRC and OS are summarized in Table 7. Stage II (vs. III) disease and a radiation dose ⱖ51 Gy were independent predictors of improved LRC, and locoregional failure was an independent predictor of worse OS. Figure 2 shows the percentage of patients with locoregional failure as a function of
The results of the current retrospective study are consistent with the observation that 50 Gy at 1.8 –2 Gy/ fraction within 5 weeks is adequate to control ⬎90% of subclinical disease in patients with squamous cell carcinoma of the upper aerodigestive tract. At least 60 –70 Gy given at the same fractionation is needed to treat gross tumors (12). Patients in our study who received a total dose ⬎51 Gy had significantly better LRC, DFS, and OS, with a trend toward better DMFS noted in the higher dose group. This finding indicates that eliminating local disease at the primary tumor site with high-dose radiation is important in reducing the risk of cancer dissemination distantly. The combination of increased LRC and decreased distant metastasis was associated with superior OS in patients treated with high-dose RT. However, the optimal radiation dose for treating esophageal cancer is controversial, especially when chemotherapy is delivered concurrently. Sun (13) reported a clear association between a higher radiation dose and improved 5-year survival when RT was the sole therapeutic modality in patients with Stage II or III esophageal cancer. The 5-year survival rate was 10.6% for patients who received 60 – 69 Gy and about 2% for those who received 50 –59 Gy (13). In the series of Coia and associates (14), patients received 5-FU, mitomycin C, and 60 Gy of radiation. Of importance, their trial was the only combined-modality trial in which patients with clinically early-stage esophageal cancer (Stage I and II) were treated and analyzed separately (14). The results of that trial demonstrated a very low local failure rate of 25%, a 5-year actuarial survival rate of 30%, and a 5-year actuarial local relapse-free survival rate of 70% for patients with Stage I disease. However, in a study of 30 patients with clinical Stage I–III disease, John et al. (15) reported
Table 5. Univariate Cox regression analysis for locoregional control Factor
Hazard ratio
p ⬎ |Z|
95% CI
Age ⬍60 y Nonwhite race (vs. white) Male gender KPS 70–80 (vs. 90–100) Weight loss ⬎5% Tumor in lower esophagus or GEJ (vs. upper or middle esophagus) Nonsquamous subtype (vs. squamous) Stage II (vs.III) No clinical CR Distant metastases Radiation dose ⬎51 Gy (vs. ⱕ51 Gy)
1.00 1.39 1.59 0.98 1.02 1.19 0.97 0.44 9.26 0.87 0.45
0.85 0.29 0.19 0.07 0.13 0.58 0.92 0.008 0.000 0.69 0.017
0.97–1.03 0.76–2.55 0.79–3.20 0.95–1.03 0.99–1.05 0.65–2.16 0.52–1.80 0.24–0.81 3.45–24.88 0.44–1.72 0.24–0.87
Abbreviations: CI ⫽ confidence interval; KPS ⫽ Karnofsky performance status; GEJ ⫽ gastroesophageal junction; CR ⫽ complete response.
Radiation dose in esophageal cancer
● Z. ZHANG et al.
661
Table 6. Univariate Cox regression analysis for overall survival Factor
Hazard ratio
p ⬎ |Z|
95% CI
Age ⬍60 y Nonwhite race (vs. white) Gender KPS 70–80 (vs. 90–100) Weight loss ⬎5% Tumor in lower esophagus or GEJ (vs. upper or middle esophagus) Nonsquamous subtype (vs. squamous) Stage II (vs. III) No clinical CR Distant metastases Locoregional failure Radiation dose ⬎51 Gy (vs. ⱕ51 Gy)
0.59 1.14 1.68 0.98 0.99 0.99 0.89 0.76 2.32 1.22 2.40 0.60
0.12 0.64 0.07 0.20 0.97 0.99 0.66 0.29 0.02 0.48 0.04 0.063
3.11–1.13 0.66–1.96 0.95–2.98 0.96–1.00 0.60–1.64 0.58–1.68 0.52–1.51 0.46–1.26 1.17–4.58 0.70–2.13 1.33–4.34 0.35–1.03
Abbreviations as in Table 5.
a similar local failure rate 27% with a lower radiation dose: 40 –50 Gy. The 2-year actuarial survival rate was 29%. Radiation doses of as much as 66 Gy after three cycles of cisplatin and bleomycin have also been used (16). RTOG trial 85-01 (1) established that a radiation dose of 50 Gy is appropriate for patients with esophageal cancer treated with simultaneous cisplatin-based chemotherapy. In that trial, patients received four cycles of 5-FU (1000 mg/m2 on Days 1– 4) and cisplatin (75 mg/m2 on Day 1). It should be emphasized that Cycles 1 and 2 of chemotherapy were given every 4 weeks and Cycles 3 and 4 were given every 3 weeks. This dose intensification after the combined-modality segment may explain, in part, why patients had difficulty completing all four cycles of chemotherapy. RT (50 Gy) was given concurrently with chemotherapy beginning on Day 1. A higher dose of radiation (64 Gy) was used in the RT control arm. At a minimum of 5 years of follow-up, the OS rate was 26% for patients who received combinedmodality therapy and 0% for those who received RT alone. Disease persistence was the most common mode of treatment failure; however, it was less common in the group receiving combined therapy (26%) than in the RT-alone group (37%). Distant metastasis accounted for the first site of treatment failure in 30% of the RT-alone group compared with 16% of the combined-modality group (1). The positive results of RTOG trial 85-01 established combined-modality therapy rather than RT alone as the conventional nonoperative treatment for esophageal cancer. However, despite the positive results in the RTOG 85-01 combined-modality arm, the local failure rate was high (50% at 5 years) and the survival rate was modest (27% at 5 years). RTOG trial 94-05 (2), a follow-up to RTOG 85-01, investigated the possibility of intensification of the radiation dose. In the 94-05 trial, 236 patients with cT1– T4NxM0 squamous cell carcinoma (85%) or adenocarcinoma (15%) of the esophagus without tumor extension to within 2 cm of the stomach were randomized to standarddose combined-modality therapy using a slight modifi-
cation of the combined-modality therapy arm of RTOG 85-01 (50.4 Gy plus concurrent 5-FU and cisplatin on Weeks 1 and 4, repeated 4 weeks after RT completion) or high-dose RT (64.8 Gy) and the same chemotherapy regimen. That trial failed to show any benefit in terms of survival in the high-dose arm. No statistically significant differences were found between the high-dose and standard-dose arms in the median survival time (13.0 vs. 18.1 months), 2-year survival rate (31% vs. 40%), or rate of locoregional failure or locoregional disease persistence (52% vs. 56%). On the basis of these two well-designed prospective randomized trials, 50.4 Gy at 1.8 Gy/fraction for 5 d/wk is currently considered standard when cisplatin-based chemotherapy is administered concomitantly to patients with esophageal cancer treated with a nonoperative approach (1, 2). However, 7 of the 11 treatment-related deaths in the high-dose arm of the RTOG 94-05 trial occurred in patients who had received ⱕ50.4 Gy. In addition, a statistically significant prolongation of treatment time occurred because of breaks required for recovery from side effects after correction for the number of RT sessions and a statistically significant lower dose of 5-FU was given to patients in the high-dose arm. The authors believed that these factors might have contributed, at least in part, to the lack of benefit for patients who received high-dose vs. standard-dose RT (2). Therefore, the findings from the RTOG 94-05 trial were inconclusive regarding whether a radiation dose effect exists in the treatment of cancer of the esophagus. The findings of RTOG 94-05 warrant extensive research on methods to reduce treatment toxicity to allow radiation dose intensification for patients who are not considered candidates for surgery and for whom chemoradiotherapy is the only treatment alternative. Our findings suggest the existence of a correlation between a higher radiation dose and improved LRC when RT is given concurrently with chemotherapy in treating cancer of the esophagus, although the slope of the curve
662
I. J. Radiation Oncology
● Biology ● Physics
Volume 61, Number 3, 2005
Fig. 2. Percentage of patients with locoregional failure 2 years after treatment as a function of radiation dose.
flattened in the high-dose region. The shallower slope of the curve in the high-dose area also suggests the presence of a threshold in the tumor response, such that additional increases in the radiation dose would not improve local tumor control further. A carefully designed dose-escalation study using LRC as the primary endpoint would be required to confirm this assumption. It is worth pointing out that patients received 30 Gy of total radiation given with rapid fractionation, which was, at one time, the treatment chosen in our institution. This program was not meant to be palliative; rather, it was based on the principle that the total radiation dose required to obtain a given biologic effect decreases as the dose per fraction increases. A 30-Gy total dose of radiation given in 10 fractions was considered radiobiologically equivalent to
Table 7. Multivariate Cox regression analysis for locoregional control and overall survival Factor LRC Radiation dose ⬎ 51 Gy (vs. ⱕ51 Gy) Clinical stage II (vs. III) OS Radiation dose ⬎51 Gy (vs. ⱕ51 Gy) Locoregional failure
Hazard ratio
p ⬎ |Z|
95% CI
0.48
0.029
0.25–0.93
0.42
0.007
0.22–0.79
0.69
0.193
0.39–1.21
2.13
0.012
1.19–4.04
Abbreviations: CI ⫽ confidence interval; LRC ⫽ locoregional control; OS ⫽ overall survival.
a standard 5.5-week (50.4 Gy in 28 fractions) program with a shortened overall treatment time. It was designed as a definitive treatment, combined with concurrent chemotherapy for esophageal, pancreatic, and periampullary carcinomas and was used for prospective clinical trials during the early 1990s at the M. D. Anderson Cancer Center (3– 6). However, our data suggested that 30 Gy in 10 fractions was not equivalent to 50.4 Gy in 28 fractions for LRC, even though the lower dose was given in a larger fraction within a shorter overall treatment time. The institutional treatment policy has been changed to use at least 50 Gy for patients receiving definitive chemoradiotherapy since the mid-1990s. It is also important to realize the limitations and drawbacks of a retrospective study, such as the current one, and to be cautious in drawing definitive conclusions from such a study. Although measures were taken to minimize the difference between the two populations, this study has the inherent limitations of a retrospective study in terms of patient selection bias and treatment heterogeneity. A meta-analysis by Ancona et al. (17) provided evidence of a relationship between a higher radiation dose and increased likelihood of a pathologic complete response. In studies of preoperative chemoradiotherapy in patients with esophageal cancer, the 5-year survival rate ranged from 46% to 60% for complete responders, and a statistically significant survival advantage appeared to be limited to patients who showed a complete response to preoperative treatment and subsequently underwent curative resection (17–20). It remains unclear whether patients who respond to chemoradiotherapy and do not undergo surgery have the potential for long-term sur-
Radiation dose in esophageal cancer
vival. In the present study, the clinical complete primary tumor response rate was 47% in the higher dose group (p ⫽ 0.006). The relatively high complete response rate for patients treated with ⬎51 Gy, along with the superior median survival time and 5-year survival observed in these patients, supports the use of high-dose RT with concurrent chemotherapy. That ⬎50% of patients in the RTOG studies had locoregional failure indicated the need for better locoregional treatment. At our institution, a clinical trial of a selective cyclooxygease-2 inhibitor as a radiosensitization agent is currently underway. The use of intensity-modulated RT to reduce the volume of normal tissue irradiated is also being tested in our clinic. The use of cytoprotective agents such as amifostine in esophageal cancer patients is also an active area of research. The hope is to find ways to increase tumor response and reduce radiation-induced damage to normal tissue, making escalation of the radiation dose either clinically feasible or unnecessary. CONCLUSION In the present study, statistically significant better LRC, DFS, and OS were seen in patients who received ⬎51 Gy. In a multivariate analysis, clinical Stage II disease and a radiation dose ⬎51 Gy were independent predictors of improved LRC, and locoregional failure was an independent predictor of worse OS. The data also
● Z. ZHANG et al.
663
suggested that better LRC was associated with a lower risk of distant metastasis. A threshold of tumor response to radiation dose might exist, as suggested by a flattened slope in the high-dose area on the dose–response curve. A carefully designed dose-escalation study is required to confirm this assumption. The side effects were minimal in the studied population and the differences were not statistically significant between the lower dose and higher dose groups. However, the accuracy of the data on toxicity was limited by the retrospective nature of this study. Methods to reduce treatment toxicity and increase the tumor response to RT, thereby increasing the therapeutic ratio, are needed. Modern radiation techniques, such as conformal therapy or intensity-modulated RT, may help to allow dose escalation by minimizing the normal tissue dose around the esophagus. The introduction of a mucosal protective agent such as amifostine may also help to allow further dose escalation. The combination of molecularly targeted therapy, such as cyclooxygenase-2 inhibition to increase tumor radioresponse, with the methods mentioned above may facilitate an increase in the therapeutic ratio. Our data indicate a need for additional study of the optimal radiation dose in a prospective randomized trial, with emphasis on decreasing the treatment-related toxicity of concurrent chemoradiotherapy as the only treatment modality in patients with locally advanced esophageal cancer.
REFERENCES 1. Cooper J, Guo M, Herskovic A, et al., for the Radiation Therapy Oncology Group. Chemoradiotherapy of locally advanced esophageal cancer: Long-term follow-up of a prospective randomized trial (RTOG 85-01). JAMA 1999;281:1623–1627. 2. Minsky B, Pajak T, Ginsberg R, et al. INT 0123 (Radiation Therapy Oncology Group 94-05) phase III trial of combinedmodality therapy for esophageal cancer: High-dose versus standard-dose radiation therapy. J Clin Oncol 2002;20:1167–1174. 3. Spitz F, Abbruzzese J, Lee J, et al. Preoperative and postoperative chemoradiation strategies in patients treated with pancreaticoduodenectomy for adenocarcinoma of the pancreas. J Clin Oncol 1997;15:928 –937. 4. Pisters P, Abbruzzese J, Janjan N, et al. Rapid-fractionation preoperative chemoradiation, pancreaticoduodenectomy, and intraoperative radiation therapy for resectable pancreatic adenocarcinoma. J Clin Oncol 1998;16:3843–3850. 5. Evans D, Abbruzzese J, Cleary K, et al. Rapid-fractionation pre-operative chemoradiation for malignant periampullary neoplasms. J Roy Coll Surg Edinb 1995;40:319 –323. 6. Ajani J, Ryan B, Rich T, et al. Prolonged chemotherapy for localised squamous carcinoma of the oesophagus. Eur J Cancer 1992;28A:880 – 884. 7. Stata Statistical Software. Release 7.0. College Station, TX: Stata Corporation; 2001. 8. Kaplan E, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;53:457– 481. 9. Cox D. Regression models and life tables. J R Stat Soc [Ser B] 1972;34:187–229. 10. Karnofsky D, Burchenal J. The clinical evaluation of chemo-
11. 12.
13. 14. 15. 16.
17.
therapeutic agents in cancer. In: MaCleod C, editor. Evaluation of chemotherapeutic agents. Symposium, microbiology section. New York Academy of Medicine. Vol 1. New York: Columbia University Press; 1949. p. 191–205. Hall E. Radiobiology for the radiologist. Philadelphia: JB Lippincott; 2000. p. 223–227. Withers HR, Peters LJ. Basic principles of radiotherapy: Basic clinical parameters. In: Fletcher GA, editor. Textbook of radiotherapy. 3rd ed. Philadelphia: Lea & Febiger; 1980. p. 180. Sun D. Ten-year follow-up of esophageal cancer treated by radical radiation therapy: Analysis of 869 patients. Int J Radiat Oncol Biol Phys 1989;16:329 –334. Coia L, Engstrom P, Paul A. Long term results of infusional 5-FU, mitomycin-C and radiation as primary management of esophageal cancer. Int J Radiat Oncol Biol Phys 1991;20:29 –36. John M, Flam M, Ager Mowry P, et al. Radiation alone and chemoradiation for nonmetastatic esophageal carcinoma. Cancer 1989;63:2397–2403. Izquierdo MA, Marcuello E, Gomez de Segura G, et al. Unresectable nonmetastatic squamous cell carcinoma of the esophagus managed by sequential chemotherapy (cisplatin and bleomycin) and radiation therapy. Cancer 1993;71:287–292. Ancona E, Ruol A, Santi S, et al. Only pathologic complete response to neoadjuvant chemotherapy improves significantly the long term survival of patients with resectable esophageal squamous cell carcinoma: Final report of a randomized, controlled trial of preoperative chemotherapy versus surgery alone. Cancer 2001;91:2165–2174.
664
I. J. Radiation Oncology
● Biology ● Physics
18. Kitamura K, Kuwano H, Araki K, et al. Clinicopathologic features of patients with oesophageal cancer obtaining a histological complete response for preoperative hyperthermo-chemo-radiotherapy. Int J Hyperthermia 1998;14: 233–243. 19. Vogel SB, Mendenhall WM, Sombeck MD, et al. Downstag-
Volume 61, Number 3, 2005
ing of esophageal cancer after preoperative radiation and chemotherapy. Ann Surg 1995;221:685– 693. 20. Forastiere A, Orringer M, Perez-Tamayo C, et al. Preoperative chemoradiation followed by transhiatal esophagectomy for carcinoma of the esophagus: Final report. J Clin Oncol 1993; 11:1118 –1123.