The purpose of the current study was to test the hypothesis that a lower clinical TNM stage is associated with a higher rate of pathologic complete response (pathCR) in patients with esophageal carcinoma receiving preoperative chemoradiotherapy and to determine whether outcome after pathCR is related to clinical stage or treatment.
Clinical parameters and surgical specimens of patients with esophageal carcinoma undergoing preoperative chemoradiotherapy were analyzed to identify predictors of pathCR. In patients with pathCR, predictors of overall survival (OS), disease-free survival (DFS), and distant recurrence were studied.
Sixty-nine (29%) of 235 patients achieved pathCR. In patients with American Joint Committee on Cancer (AJCC) Stage II carcinoma, the proportion achieving pathCR was significantly larger than that achieving <pathCR (65% vs. 35%; P = 0.03). The proportion of patients who received induction chemotherapy was higher in the pathCR group than in the <pathCR group (54% vs. 46%; P = 0.05). However, neither TNM classification, primary tumor location, histologic type, gender, therapy sequence, or radiation dose (45 grays [Gy] vs. 50.4 Gy) were found to have any influence on OS or DFS. The median OS from pathCR was significantly longer than that from <pathCR (133 mos vs. 34 mos; P = 0.002). Similarly, DFS was longer in the pathCR group than in the <pathCR (P = 0.001).
Carcinoma of the esophagus, although rare, is an aggressive disease associated with an extremely poor prognosis, even when it is localized. The 5-year survival rate after either surgery alone or definitive chemoradiotherapy has consistently been 15% or less over the past 25 years.1 Although in clinical practice preoperative chemoradiotherapy is commonly recommended,2 the published results from randomized trials investigating preoperative therapy, including preoperative chemoradiotherapy, have not shown this strategy to have an unequivocal statistical benefit.3–7 However, the interest in exploring multimodality preoperative therapies with newer agents remains high. Pathologic complete response (pathCR), defined as the absence of disease in both the esophagus and lymph nodes (T0N0) in the resected specimen, has been correlated with a better outcome than <pathCR.8–15 Unfortunately, the rate of pathCR has remained low and varies substantially from study to study. We reviewed 18 representative studies involving 30 or more patients published since 1993.8, 16–32 In the case of multiple publications from the same institution, either a cumulative report or the report with the largest number of patients was included.21, 22, 31 A total of 828 patients in these studies received preoperative chemoradiotherapy; the median pathCR rate was 27% (range, 6–51%) and the median number of patients with pathCR was 13 (range, 2–27). The cumulative 3-year and 5-year overall survival (OS) rates of patients with pathCR in studies that reported this information were 67% and 64%, respectively,18–20, 22, 24, 25, 28, 29 whereas the cumulative 3-year and 5-year OS rates for all patients, when reported, were 37% and 34%, respectively.18–20, 22–25, 27–31
To our knowledge, the clinical characteristics of patients with pathCR and or its predictors have not been reported to date. Similarly, the predictors of outcome for patients achieving pathCR have not been fully explored.
We analyzed patients with carcinoma of the esophagus at our institution who were receiving preoperative chemoradiotherapy. Our cohort was sufficiently large to identify the potential clinical predictors of pathCR. We also wanted to identify in patients with pathCR determinants of their OS, disease-free survival (DFS), local recurrence, and distant metastases.
Our primary hypothesis was that certain clinical parameters, including TNM stage, histopathologic type, and primary tumor location, would predict the likelihood of a pathCR. We anticipated that a lower clinical TNM stage would be correlated with a higher pathCR probability than would a higher clinical TNM stage. Our secondary hypothesis was that clinical characteristics would also predict the outcome of patients achieving pathCR.
MATERIALS AND METHODS
Evaluable patients with localized, potentially resectable carcinoma of the esophagus evaluated at The University of Texas M. D. Anderson Cancer Center from 1985 through 2003 were eligible for our analysis. Prior to therapy, patients were evaluated by a medical oncologist, gastroenterologist, radiation oncologist, and thoracic oncology surgeon. Patients with T2, with T3, N0 or N1, or clinical American Joint Committee on Cancer (AJCC) TNM Stage II or III esophageal carcinoma were included (also included were patients with carcinoma involving the gastroesophageal junction with M1a lymph nodes), whereas patients with T1N0, T4, or M1b disease were excluded, as were patients who had uncontrolled medical conditions (e.g., diabetes mellitus, hypertension, heart conditions classified as New York Heart Association Class III or IV, or psychiatric illnesses), were unable to comprehend the purpose of this clinical investigation, or were unable to comply with its requirements. Nutritional counseling was available to a select group of patients based on need and all protocol participating patients signed a written informed consent form.
All clinical staging was performed according to AJCC criteria for staging esophageal carcinoma.32 Clinical staging was performed before therapy and included a complete history and physical examination; chest radiography; computed tomography (CT) scan of the chest and abdomen; upper gastrointestinal double-contrast barium radiography; positron emission tomography (PET, when available); an esophagogastroduodenoscopy with endoscopic ultrasonography (EUS); electrocardiography; SMA-12; a complete blood count including platelet count; and measurement of serum, electrolyte, and baseline carcinoembryonic antigen levels. Additional preoperative screening studies were performed as needed.
Evaluation during Therapy
Patients' symptoms and the results of blood tests were closely monitored during therapy. Patients underwent repeat endoscopy or other imaging studies as warranted by their clinical symptoms.
Some patients received induction chemotherapy before chemoradiotherapy and others received preoperative chemoradiotherapy only. The radiation dose was either 45 grays (Gy) in 25 fractions or 50.4 Gy in 28 fractions. Two- or 3-dimensional plans using dedicated fluoroscopic or CT simulator were routinely used. In patients treated with 2-D radiotherapy, the radiation fields encompassed the primary tumor with a minimal 5-cm margin in the cephalad and caudal directions and a 2-cm margin radially. The radiation treatment usually started with antero-posterior portals until the spinal cord tolerance dose was reached, followed by oblique fields to avoid the spinal cord. Radiation does was prescribed to isocenter.
Chemotherapy agents used either during induction or concurrently with radiotherapy predominantly included cisplatin, 5-fluorouracil, camptothecin-11, and paclitaxel or docetaxel.
Evaluation before Surgery
Four to six weeks after completion of chemoradiotherapy, patients underwent another complete staging work-up including a PET (when available) but not EUS.
A curative radical en bloc esophagectomy was attempted in all patients, and the resected specimens analyzed to determine the posttherapy pathologic AJCC stage. One of three approaches were used in patients: transthoracic surgery, transhiatal surgery, or three-field lymphadenectomy.
Postsurgical staging was based on pathologic findings in the resected specimen, in particular, the residual carcinoma status. Areas showing ulcer or scar formation, indicating the therapy field, were submitted intact for histologic examination to confirm that the specimen was from a macroscopically identified absence of residual carcinoma. A pathologist (T.T.W.) who was blinded to the type and outcome of therapy reviewed all hematoxylin and eosin-stained sections, including margins and lymph nodes. Each specimen was categorized into one of two groups: as showing a pathCR or <pathCR.
Univariate Cox proportional hazard models were fitted, yielding estimates of hazard ratios for death and locoregional recurrence for select parameters (e.g., gender, age, tumor location, histologic characteristics, pretreatment clinical TNM stage, treatment sequence, radiation dose, and type of esophagogastrectomy) within the pathCR group. These parameters were also correlated with instances of distant recurrence in the pathCR group.
Kaplan–Meier survival curves were plotted for the overall data, patients by tumor stage (Stage II vs. Stage III), tumor category (T2 vs. T3), lymph node category (N0 vs. N1), and type of carcinoma (adenocarcinoma vs. squamous cell carcinoma). The OS time was defined as the interval from time of surgery to time of death or date of last follow-up, at which point the data were censored.
P-values were obtained for comparisons of these clinical parameters within the pathCR group and the distributions of these parameters between the pathCR and non-pathCR groups.
Statistical analysis was performed using SPSS software (version 18.104.22.168 for Windows; SPSS Inc., Chicago, IL). All statistical analyses were two-sided; P-values of 0.05 were considered to be statistically significant.
The 235 consecutive patients evaluated at the M. D. Anderson Cancer Center between 1985–2003 who had histologically confirmed adenocarcinoma or squamous cell carcinoma of the esophagus and who underwent pretreatment clinical staging followed by preoperative chemoradiotherapy were analyzed. Sixty-nine of these patients (29%) achieved a pathCR. Patient characteristics are summarized in Table 1.
Table 1. Clinical Characteristics and Predictors of Survival of 69 Patients with pathCR
pathCR: pathologic complete response; GEJ: gastroesophageal junction; EUST: endoscopic ultrasonography-diagnosed T classification; EUSN: endoscopic ultrasonography-diagnosed N classification; AJCC: American Joint Committee on Cancer; Chemo: chemotherapy; RT: radiation therapy; Gy: grays.
Determined by the chi-square test.
Median age in years (range)
Squamous cell carcinoma
Clinical AJCC stage
Chemo then Chemo/RT
Type of surgery
Pretreatment Staging of Patients with pathCR
Pretreatment clinical stages were as follows: 23 patients (33%) had EUS-diagnosed T2 carcinoma, 46 patients (67%) had EUS-diagnosed T3 carcinoma, 40 patients (58%) had EUS-diagnosed N0 lymph nodes, and 29 patients (42%) had EUS-diagnosed N1 lymph nodes. Forty-five patients (65%) had Stage II carcinoma and 24 (35%) had Stage III carcinoma.
Using chi-square analysis, we determined that the distributions of gender, age, primary location, histologic type, clinical staging, preoperative radiation dose, and type of surgery were statistically significantly different (Table 1) between the 69 patients with a pathCR compared with the 166 patients with <pathCR.
Treatment Characteristics of the Patients with a pathCR
Thirty-seven patients (54%) received induction chemotherapy followed by preoperative chemoradiotherapy, whereas 32 patients (46%) received preoperative chemoradiotherapy only. The radiation dose was 45 Gy in 25 fractions for 49 patients (71%) and 50.4 Gy in 28 fractions for 20 patients (29%).
Cisplatin was administered to 54 patients (78%), 5-fluorouracil was administered to 68 patients (99%), paclitaxel or docetaxel was administered to 36 patients (52%), and 17 patients (25%) received other agents.
A transthoracic approach along with regional lymphadenectomy was performed in 37 patients (54%), and a transhiatal approach was used in 18 patients (26%). The remaining 14 patients (20%) underwent a 3-field lymphadenectomy.
Survival Analysis of Patients with pathCR
The median follow-up time was 37 months (range, 1–191 mos) and the median survival time was 133 months (range, 1–133 mos) for all patients with a pathCR (Fig. 1).
The median survival time for patients with EUS-diagnosed T3, EUS-diagnosed N0, and clinical Stage II carcinomas was 133 months (range, 1–133 mos) (Fig. 2). At the time of last follow-up, no patients with EUS-diagnosed T2, EUS-diagnosed N1, or clinical Stage III carcinomas had reached the median survival time.
The median survival time for patients with adenocarcinoma was 133 months (range, 1–133 mos) and that for patients with squamous cell carcinoma was 29 months (range, 1–59 mos), but this difference was not statistically significant (P = 0.08).
Parameters Predictive of pathCR
We compared clinical variables such as gender, age, primary tumor location, histologic type, EUS-diagnosed T and EUS-diagnosed N classifications, clinical stage, chemotherapy type and its sequence with chemoradiotherapy, and preoperative radiation dose in the 69 patients who achieved pathCR with those in the 166 patients with < a pathCR. The proportion of Stage II patients who achieved a pathCR was significantly greater than those who did not achieve a pathCR (65% vs. 35%; P = 0.03). In patients who received induction chemotherapy, the proportion who achieved a pathCR was also statistically (borderline) greater than that of patients with < a pathCR (54% vs. 46%; P = 0.05). None of the other variables examined was found to be significantly correlated with pathCR (Table 2).
Table 2. Predictors of PathCR and Comparison of Clinical Parameters in Patients with pathCR and Less than a pathCR
(n = 69) (%)
(n = 166) (%)
(n = 235) (%)
pathCR: pathologic complete response; SD: standard deviation; GEJ: gastroesophageal junction; EUST: endoscopic ultrasonography-diagnosed T classification; EUSN: endoscopic ultrasonography-diagnosed N classification; AJCC: American Joint Committee on Cancer; Chemo: chemotherapy; RT: radiation therapy; Gy: grays.
Age in yrs
Squamous cell carcinoma
Clinical AJCC stage
Chemo then Chemo/RT
Preoperative radiation dose
≤ 45 Gy
> 45 Gy
Survival of Patients with a pathCR and < a pathCR
Patients with pathCR had a statistically significantly longer OS time than did those with <pathCR (Fig. 1). The median OS time for patients with pathCR was 133 months (range, 1–133 mos), whereas that for patients with <pathCR was 34 months (range, 1 to >186 mos; P = 0.002).
The DFS time was significantly longer for patients with a pathCR than for patients with < a pathCR (Fig. 3). The median DFS time for patients with < a pathCR was 16 months (range, 0 to >100 mos) ant that for patients with a pathCR had not been reached at the time of last follow-up, a median of 37 months (P = 0.001).
Predictors of Survival in Patients with a pathCR
Because the 5-year survival rate of patients with pathCR is approximately 65%, we analyzed our data for any association between OS time and gender, primary tumor location, histology type, TNM staging, chemotherapy and chemoradiation sequence, radiation dose, type of chemotherapy, or type of surgery, but none of these parameters was found to be correlated (Table 1), with the exception of age, which was associated only as a continuous variable (P = 0.05).
Achievement of a pathCR was the only significant variable (P = 0.008) in a multivariate analysis of all 235 patients with regard to OS.
Patterns of Disease Recurrence among Patients with a pathCR and < a pathCR
Locoregional recurrence occurred in only 5 of the 69 patients with a pathCR (7%). None of the examined parameters (gender, age, primary location, histology, pretreatment TNM staging, chemotherapy chemoradiation sequence, radiation dose, type of chemotherapy, histology type, and type of surgery) was found to be correlated with time to locoregional recurrence.
Distant disease recurrence occurred in 10 of the 69 patients with pathCR (14%) and none of the clinical parameters could be correlated. However, the distant recurrence-free survival time was significantly longer for patients who achieved a pathCR compared with those with < a pathCR (P = 0.0009) (Fig. 4).
To our knowledge to date, no clinical or molecular parameter has been associated with the prediction of pathCR in patients with carcinoma of the esophagus receiving preoperative chemoradiotherapy. We hypothesized that a lower clinical stage might be predictive of pathCR. Our data in 235 patients suggest that patients with clinical Stage II disease are more likely to achieve a pathCR than those with clinical Stage III disease (P = 0.03). We believe that the association between lower stage disease and pathCR may be the most important observation of the current study. Suntharalingam et al.31 reported on 24 pathCR patients who, after laparoscopic thoracic and abdominal staging, received preoperative chemoradiotherapy. They noted that patients with N1 disease were less likely to achieve pathCR compared with those with N0 disease. However, a full analysis of patients with and without pathCR was not reported and pretreatment stage was not found to be correlated with pathCR. Invasive (laparoscopic) staging is not routinely performed in patients with carcinoma of the esophagus, and the 235 patients in the current study were staged by a conventional noninvasive method. Therefore, a comparison of our analysis and that of Suntharalingam et al.31 may not be appropriate.
We also observed that patients who received induction chemotherapy prior to chemoradiotherapy had a statistically greater chance of achieving a pathCR (P = 0.05). This is intriguing and hypothesis generating. Although we have reported on induction chemotherapy previously,10, 11, 33 to our knowledge its contribution to patient outcome is not yet established.
A method with a high probability of identifying patients likely to achieve a pathCR before surgery may lead to the implementation of more effective strategies. For example, the esophagus could be preserved in such patients. Our finding that patients with a pathCR had only a 7% rate of locoregional disease recurrence compared with 19% for those with < a pathCR (P = 0.05) supports the strategy to delay surgery if patients can be identified accurately. The concept of esophageal preservation can be realized only if a substantial number of patients treated with preoperative chemoradiotherapy can achieve a pathCR.
Could one extrapolate that patients with clinical Stage I esophageal carcinoma might achieve an even higher rate of pathCR? The current therapy for patients with Stage I esophageal carcinoma is surgery. However, Ura et al.34 reported a 77% 5-year survival rate for 73 patients with squamous cell carcinoma of the esophagus who were treated with definitive chemoradiotherapy. Based on the current study data, we believe this strategy may be appropriate for patients with Stage I adenocarcinoma. A clinical study using preoperative chemoradiotherapy to determine the pathCR rate would first be necessary before avoiding surgery in this group of patients.
The University of Texas M. D. Anderson Cancer Center's studies in the past have demonstrated a higher pathCR rate and survival compared with other studies in the literature. However, the proportion of patients with clinical Stage II was 58%, and 36% had Stage III disease, but from our literature search we found 442 patients who were clinically staged prior to treatment, 36% of whom were classified as having Stage II disease, 60% with Stage III disease, and 4% with Stage IV disease. However, some of these studies employed other guidelines to assess clinical staging.8, 17–19, 24, 27 These data potentially suggest that we may have reached an obstacle to further improvement in the pathCR rate. Our nonspecific cytotoxic approach is no longer altering the biology of this cancer, and we may be better off with more specific approaches, such as targeted therapies to overcome chemoradiotherapy resistance.
Our data regarding OS for 235 patients are consistent with those in representative studies (Table 3). Specifically, the 3- and 5-year OS for patients with a pathCR and the entire cohort are similar, but it is also clear that 35% of patients with a pathCR do not live 5 years. Our data further indicate that pretreatment parameters (e.g., EUS-diagnosed T classification, EUS-diagnosed N classification, or disease stage) do not correlate with survival once a patient achieves a pathCR. In addition, none of the treatment variables (e.g., radiation dose, type of chemoradiotherapy treatment, or treatment sequence) was found to have a bearing on OS or DFS time. Therefore, identification of patients at high risk for disease recurrence is also important, but to our knowledge there currently is no method to accomplish it.
Table 3. Comparison of 3-Year and 5-Year Survival Rates between Published Data and the Current Study
These studies reported survival of all patients.
These studies reported survival of those patients who achieved a pathologic complete response.
Published data: Stahl M et al., 199618; Stockeld et al., 200119; Vogel et al., 199520; Meneu-Diaz et al., 200023; Donington et al., 200324; De Vita et al.200225; Kim et al.,200127;Makary et al., 200329; and Sunthraralingam et al., 200131a Stahl M et al., 199618; Stockeld et al., 200119; Vogel et al., 199520; Donington et al., 200324; and De Vita et al.200225b
37%% (n = 561)
67% (n = 90)
53% (n = 235)
70% (n = 69)
Published data: Stockeld et al., 200119; Vogel et al., 199520; Forastiere et al., 199322; Kim et al.,200127; Imdahl et al., 200028; Makary et al., 200329; and Tsujinaka et al., 199930a Vogel et al., 199520; Forastiere et al., 199322; Imdahl et al., 200028; and Makary et al., 200329b
35% (n = 435)
64% (n = 50)
42% (n = 235)
65% (n = 69)
Clinical stage was not found to correlate with OS but pathCR did appear to correlate with OS and DFS. There is no clear explanation for this observation. Clinical stage remains suboptimal, whereas the pathologic stage is definite. Another possibility is that there might be intrinsic differences in the biology of this tumor. Patients who are destined to live longer may be achieving a pathCR. However, not all patients with Stage II carcinoma are likely to have a favorable biology and do not have a homogeneous outcome.
Data from the current study confirm that the outcome is best when a pathCR is achieved but, more important, that patients with a lower clinical stage of disease are more likely to achieve a pathCR than are patients with a higher stage. This observation could be exploited to preserve the esophagus in clinical Stage I patients.