Drs. Chirieac and Hamilton were supported by the Frederick F. Becker Distinguished University Chair in Cancer Research from The University of Texas.
Posttherapy pathologic stage predicts survival in patients with esophageal carcinoma receiving preoperative chemoradiation
Article first published online: 17 FEB 2005
Copyright © 2005 American Cancer Society
Volume 103, Issue 7, pages 1347–1355, 1 April 2005
How to Cite
Chirieac, L. R., Swisher, S. G., Ajani, J. A., Komaki, R. R., Correa, A. M., Morris, J. S., Roth, J. A., Rashid, A., Hamilton, S. R. and Wu, T.-T. (2005), Posttherapy pathologic stage predicts survival in patients with esophageal carcinoma receiving preoperative chemoradiation. Cancer, 103: 1347–1355. doi: 10.1002/cncr.20916
- Issue published online: 18 MAR 2005
- Article first published online: 17 FEB 2005
- Manuscript Accepted: 3 DEC 2004
- Manuscript Revised: 17 NOV 2004
- Manuscript Received: 20 SEP 2004
- Frederick F. Becker Distinguished University Chair in Cancer Research from The University of Texas
- residual carcinoma;
In patients with locoregional carcinoma of the esophagus or esophagogastric junction who underwent preoperative chemoradiation, it is unclear whether survival was better predicted by pretherapy clinical stage or by posttherapy pathologic stage.
The authors studied 235 consecutive patients with pretherapy clinical Stage II, III, or IVA (according to American Joint Committee on Cancer criteria) carcinoma of the esophagus or esophagogastric junction who were treated with chemoradiation followed by esophagectomy. Posttherapy cancer status was classified using pathologic stage and semiquantitative assessment of residual carcinoma. Clinicopathologic features, residual carcinoma status, and pretherapy and posttherapy stage were compared with disease-free and overall survival.
Posttherapy pathologic stage was Stage 0 in 29% of patients, Stage I in 11% of patients, Stage II in 34% of patients, Stage III in 20% of patients, and Stage IV in 6% of patients. Cancer downstaging occurred in 56% of patients. In univariate analysis, disease-free and overall survival were predicted by posttherapy pathologic stage (both with P < 0.001), margin status (P = 0.002 and P = 0.01, respectively), extent of residual carcinoma (both with P < 0.001), and downstaging (both with P = 0.001), but not by age, gender, type of cancer, pretherapy clinical stage, or preoperative regimen. However, in multivariate analysis, disease-free and overall survival were independently predicted by posttherapy pathologic stage (both with P = 0.02). Extent of residual carcinoma was a marginally significant predictor of overall survival (P = 0.04).
Posttherapy pathologic stage was the best available predictor of outcome for patients with locoregional carcinoma of the esophagus or esophagogastric junction who underwent chemoradiation therapy followed by esophagectomy. The findings in the current study supported the concept of downstaging by preoperative therapy. Cancer 2005. © 2005 American Cancer Society.
Carcinoma of the esophagus and esophagogastric junction is aggressive with a poor prognosis.1–6 Esophagectomy has been the mainstay of therapy for localized esophageal carcinoma. Although pretherapy clinical staging continues to improve, it often differs from the pathologic stage in the resected specimen.7–9 When surgical resection is the primary therapy, the best predictor of disease-free survival and overall survival is pathologic stage.10–14
Multimodality strategies for patients with locoregional esophageal carcinoma, including preoperative chemoradiation followed by esophagectomy, remain popular according to practice surveys,15 despite equivocal data on benefit in randomized studies.2, 16–20 In patients in whom pathologic stage is potentially altered by preoperative therapy, it remains unclear whether pretherapy clinical stage or postsurgical pathologic stage best predicts disease-free and overall survival. Positron emission tomography (PET) scans are a promising technique for assessing response to preoperative therapy, but this technique is evolving and cannot precisely determine the altered stage21–24 or predict pathologic response in the resected specimen.25 Until new technologies have evolved to accurately determine stage and predict survival, established staging techniques must be further evaluated for their value.
The outcome of patients after preoperative chemoradiation is much better if no residual carcinoma (Stage 0) is found in the resected specimen, representing a pathologic complete response (CR).26–34 However, a pathologic CR occurs in < 30% of patients who undergo surgery after preoperative chemoradiation,17, 20, 26, 35, 36 and no reliable method currently exists for predicting disease-free and overall survival for patients who have residual carcinoma in the specimen after preoperative therapy.
We hypothesized that posttherapy pathologic stage rather than pretherapy clinical stage best predicts disease-free and overall survival in patients undergoing preoperative chemoradiation for carcinoma of the esophagus and esophagogastric junction. We therefore studied the clinicopathologic features of 235 patients and related the findings to survival.
MATERIALS AND METHODS
The current study included 235 consecutive patients with histologically confirmed invasive squamous cell carcinoma or adenocarcinoma of the esophagus or esophagogastric junction. Between 1985 and 2003, these patients were treated with chemoradiation followed by esophagectomy at The University of Texas M. D. Anderson Cancer Center (MDACC; Houston, TX). Resected specimen pathology slides were available for review. The study was approved by the MDACC institutional review board.
The mean age of the patients was 60.1 years (95% confidence interval [95% CI], 58.9–61.4 years). There were 203 men and 32 women. Forty-two tumor specimens (18%) were squamous cell carcinoma and 193 (82%) were adenocarcinoma. The majority of tumors (87%) occurred in the distal esophagus or esophagogastric junction with only 13% occurring in the mid or upper esophagus. The pretherapy clinical stage was Stage II in 58% of patients, Stage III in 36% of patients, and Stage IVA with celiac lymph node involvement in 6% of patients (Table 1). Staging was performed according to American Joint Committee on Cancer criteria.
|Characteristics||No. (%) of patients (n = 235) (%)a||Disease-free survival||Overall survival|
|HR||(95% CI)||P value||HR||(95% CI)||P value|
|Female (reference)||32 (14)||1||1|
|Adenocarcinoma (reference)||193 (82)||1||1|
|Squamous cell carcinoma||42 (18)||1.2||(0.76–1.9)||0.43||1.27||(0.78–2.06)||0.34|
|Cervical/upper/middle (reference)||30 (13)||1||1|
|Lower/esophagogastric junction||205 (87)||1.11||(0.65–1.89)||0.71||1.05||(0.6–1.83)||0.87|
|Chemoradiation (reference)||132 (56)||1||1|
|Chemotherapy followed by chemoradiation||103 (44)||0.89||(0.6–1.31)||0.55||0.89||(0.58–1.37)||0.58|
|II (reference)||137 (58)||1||1|
|Pathologic ypTNM stage||< 0.001||< 0.001|
|0 (reference)||69 (29)||1||1|
|III||46 (20)||3.43||(1.98–5.95)||< 0.001||3.35||(1.80–6.24)||< 0.001|
|IV||15 (6)||5.38||(2.59–11.15)||< 0.001||5.39||(2.46–11.78)||< 0.001|
|Residual carcinoma||< 0.001||< 0.001|
|0% (reference)||77 (33)||1||1|
|> 50%||57 (24)||3.45||(2.07–5.73)||< 0.001||4.01||(2.28–7.05)||< 0.001|
|No (reference)||103 (44)||1||1|
|Negative (reference)||211 (90)||1||1|
Patients with clinical Stage I or Stage IVB disease (systemic metastasis) were not eligible for preoperative chemoradiation. Pretherapy clinical stage was determined by barium-swallow esophagogram in 162 patients (69%), computed tomography (CT) scan in 213 patients (91%), endoscopic ultrasonography in 123 patients (52%), and a PET scan in 68 patients (29%). An endoscopic ultrasonograph was used as a pretreatment staging tool in the majority of the patients treated at MDACC after 2001.
Induction chemotherapy followed by chemoradiation was given to 103 (44%) patients, and the remaining patients received only chemoradiation preoperatively (Table 1). Three agents were used for preoperative chemotherapy: 5-fluorouracil in 223 patients (95%), cisplatin in 163 patients (69%), and a taxane in 117 patients (50%). All patients underwent CT scan simulation for radiotherapy. The clinical target volume was defined as the macroscopic tumor volume plus a 5-cm margin superior to the highest extension and inferior to the lowest extension of the carcinoma with a 2-cm radial margin. The planning target volume was defined as the clinical target volume plus a 5-mm margin. The total dose of radiotherapy was 45 gray (Gy) in 25 fractions or 50.4 Gy in 28 fractions prescribed to cover ≥ 95% of the planning target volume. Radiotherapy was given with megavoltage equipment providing > 6 MV of energy with anterior and posterior fields for 20–22 fractions followed by oblique or lateral fields for the remaining fractions to spare the spinal cord.
Four to six weeks after completing preoperative therapy, patients underwent a radical en bloc esophagectomy. The surgical approaches used were Ivor–Lewis esophagectomy (abdominal-right thoracic approach) in 130 patients (55%), 3-field McKeown esophagectomy (right thoracic abdominal-cervical approach) in 44 patients (19%), and transhiatal esophagectomy (abdominal-cervical approach) in 59 patients (25%).37, 38 Two patients (1%) underwent 2-stage esophagectomy.
Assessment of Residual Carcinoma and Pathologic Stage
Residual carcinoma status was determined in each esophagectomy specimen. If residual carcinoma was identified by macroscopic evaluation, the tumor specimen was sampled for histologic evaluation in an average of 15 slides (95% CI, 12.1–18.0). If no residual carcinoma was identified by macroscopic evaluation, areas with ulcer or scar indicating the therapy fields were submitted completely for histologic examination and generated an average of 16 slides (95% CI, 14.9–17.7).
A gastrointestinal pathologist (T.-T.W.) who had no knowledge of the therapy or outcome reviewed all hematoxylin-eosin–stained sections, including margins and lymph nodes. Each specimen was evaluated for depth of invasion and lymph node metastasis and staged according to American Joint Committee on Cancer criteria for esophageal carcinoma.11
Extent of residual carcinoma was assessed semiquantitatively irrespective of lymph node status based on the estimated percentage of residual carcinoma in relation to total carcinoma area, including amount of radiotherapy-induced tissue injury, in mural histologic sections.39 Extent of residual carcinoma in the esophagectomy specimen was assigned to 1 of 4 categories: no residual carcinoma, 1–10% residual carcinoma, 11–50% residual carcinoma, and > 50% residual carcinoma (Fig. 1), as modified from selected published grading systems for esophageal and gastric carcinomas.39, 40 The extent of residual carcinoma in regional lymph nodes was not assessed.
Chi-square or Fisher exact tests were used to compare categoric data. Disease-free survival was computed from the time of surgery to the time of clinical diagnosis of recurrent carcinoma or death without evidence of disease recurrence, at which point data were censored. Overall survival was calculated from the time of surgery to the time of death of any cause or to the time of last follow-up, at which point the data were censored. Disease-free and overall survival curves were constructed using the Kaplan–Meier method, and the log-rank test was used to evaluate the statistical significance of differences.
The prognostic significance of clinicopathologic characteristics was determined using univariate Cox regression analysis. Cox proportional hazards models were fitted for multivariate analysis. After interactions between variables were examined, a backward stepwise procedure was used to derive the best-fitting model.
Statistical analysis was performed using SPSS software (Version 220.127.116.11 for Windows; SPSS, Chicago, IL). Kaplan–Meier survival curves were drawn with GraphPad Prism (Version 4 for Windows; GraphPad Software, San Diego, CA). We used a 2-sided significance level of 0.05 and a power of 0.90 for all statistical analyses.
Residual Carcinoma Status and Pathologic Stage
In the esophagus and esophagogastric junction, 77 patients (33%) had no residual carcinoma, 58 patients (25%) had 1–10% residual carcinoma, 43 patients (18%) had 11–50% residual carcinoma, and 57 patients (24%) had 51–100% residual carcinoma (Table 1). Of the 77 cancers with no residual carcinoma in the esophagus or esophagogastric junction, 69 were classified as Stage 0 (ypT0N0M0), and 8 with lymph node metastases (ypT0N1M0) were classified as Stage II. Overall, 25 patients (11%) had pathologic Stage I disease, 80 (34%) had Stage II disease, 46 (20%) had Stage III disease, and 15 (6%) had Stage IV disease. Two patients were excluded from the pathologic stage analysis because of inadequate lymph node sampling. Residual carcinoma was found in the proximal, distal, or radial resection margins in 24 patients (10%).
Downstaging, as defined by posttherapy pathologic stage lower than pretherapy clinical stage, was detected in 132 patients (56%; Table 1). Among the patients with pathologic Stage 0, 65% (45 of 69) have clinical pretreatment clinical Stage II, as compared with 55% (92 of 166) of the patients with residual carcinoma (P = 0.16).
The median potential follow-up time using censored data was 30.8 months. Univariate Cox regression analysis (Table 1) showed that pathologic stage,11 extent of residual carcinoma, downstaging, and presence of disease in a resection margin were prognostic indicators for both disease-free and overall survival (Fig. 2).
As previously reported,41 we found no significant differences in either disease-free or overall survival between patients with adenocarcinoma and those with squamous cell carcinoma (P = 0.43 and P = 0.34, respectively), between men and women (P = 0.88 and P = 0.96, respectively), or between different sequences of chemoradiation (P = 0.55 and P = 0.58, respectively) (Table 1). Moreover, pretherapy clinical stage was not a prognostic indicator of disease-free or overall survival (P = 0.25 and P = 0.20, respectively).
The disease-free survival rate was significantly better for patients who had pathologic Stage 0 carcinoma (69% at 2 years and 60% at 5 years) than it was for patients with more advanced stages. The median disease-free survival time was significantly better among patients with pathologic Stage 0 carcinoma (133.2 months) than it was among patients with Stage I (35.7 months), Stage II (33.8 months), Stage III (9.4 months), or Stage IV (6.3 months) (P < 0.001) (Fig. 2). Disease-free survival did not differ significantly between patients with 1–10% residual carcinoma and patients with 11–50% residual carcinoma (P = 0.67).
The overall survival rate for patients who had pathologic Stage 0 carcinoma was 78% at 2 years and 65% at 5 years, compared with 55% at 2 years and 29% at 5 years in patients with residual carcinoma (P = 0.003). The median overall survival time was significantly better among patients with pathologic Stage 0 carcinoma (133.2 months) than it was among patients with Stage II (38.1 months), Stage III (20.7 months), or Stage IV (10.5 months) (P < 0.001) (Fig. 2). Overall survival did not differ significantly between patients with 1–10% and 11–50% residual carcinoma (P = 0.90).
After adjusting for significant variables, we found that both posttherapy pathologic stage (P = 0.02) and the extent of residual carcinoma (P = 0.04) were independent predictors of overall survival. Disease-free survival was predicted by posttherapy pathologic stage (P = 0.02), but not by the extent of residual carcinoma (P = 0.08) (Table 2).
|Variables||No. of patients (n = 235) (%)||Disease-free survival||Overall survival|
|HR||(95% CI)||P value||HR||(95% CI)||P value|
|Extent of residual carcinoma||0.08||0.04|
|0% (reference)||77 (33)||1.00||1.00|
|> 50%||57 (24)||2.29||(0.67–7.87)||0.19||4.15||(0.95–18.25)||0.06|
|Pathologic (ypTNM) stage||0.02||0.02|
|0 (reference)||69 (29)||1.00||1.00|
|No (reference)||103 (44)||1.00||1.00|
|Any margin positive|
|No (reference)||211 (90)||1.00||1.00|
Because most patients with carcinoma of the esophagus or esophagogastric junction present with advanced-stage disease, the overall prognosis after surgical resection alone is poor. The tumor is resectable in approximately 60% of patients, and the median survival is approximately 18 months for patients in whom a curative resection (R0) is achieved.17 Therefore, the strategy of preoperative therapy has received increasing attention, based on the theory that reducing the size and stage of the cancer would result in a higher rate of complete resection, reduced local disease recurrence, and prolonged disease-free and overall survival. In general, the use of preoperative chemoradiation, although popular,41–44 has not been consistently beneficial in randomized trials.2, 16, 17, 19, 20, 45, 46
Although the benefits of preoperative chemoradiation remain unsettled, it is well known that patients who have a pathologic CR fare better than those who have residual carcinoma in their resection specimen.26–33 It has been argued that patients with no residual carcinoma have a better outcome because of selection bias rather than the effects of preoperative therapy. Why cancers respond differently to the same preoperative therapy is unclear. Until highly effective individualized therapy can be developed, a reliable method for predicting prognosis after empirical therapies is needed. Pretherapy clinical stage and posttherapy pathologic stage are possible predictors.
We studied patients treated on one of several chemoradiation protocols at a single institution. Although patients included in the current study were not treated uniformly with one regimen or one esophagectomy technique, therapeutic modality did not play a significant role in disease-free and overall survival (Table 1). However, the proportion of patients with pathologic Stage 0 is slightly higher in patients treated with induction chemotherapy followed by chemoradiation (36% [37 of 103 patients]) compared with patients with residual carcinoma (24% [32 of 132 patients]).47 The detail of clinical response after induction chemotherapy is not available in this retrospective study. However, approximately 50% of patients have a clinical response after induction chemotherapy in our previously published study.48 These findings indicate that the outcome of patients treated with preoperative therapy is not determined by the regimens used.
Our data on 235 patients demonstrated that preoperative characteristics, including clinical stage, had no impact on disease-free and overall survival. Rather, the extent of residual carcinoma predicted overall survival (Fig. 2) and confirmed the findings in previous reports that patients who had no residual carcinoma in the esophagus or esophagogastric junction had a considerable survival advantage compared with those who had residual carcinoma remaining in their resection specimen.19, 26, 30, 31 The 5-year overall survival rate in the current study (i.e., 65% in patients with pathologic Stage 0 and 29% in patients with residual carcinoma) is similar to the overall survival rate published by Kleinberg et al.49 We found no significant differences in disease-free and overall survival between patients with 1–10% residual carcinoma and those with 11–50% residual carcinoma. Therefore, evaluating the extent of residual carcinoma in the primary site in the esophagus or esophagogastric junction without considering lymph node status provided 3 predictive categories: no residual carcinoma, a partial response (1–50% residual carcinoma), and no response (> 50% residual carcinoma).19, 26, 30, 31
In our study, similar to previous reports, only 29% of patients achieved absence of residual carcinoma in the esophagus and lymph nodes.19, 26, 30, 31, 39 In the remaining 71% of patients who had residual carcinoma, our data demonstrated that posttherapy pathologic stage reliably predicted disease-free and overall survival (Fig. 2), supporting that posttherapy pathology stage can be used as a reliable intermediate end point for controlled Phase II and Phase III trials. No contemporaneous control group of patients with carcinoma of the esophagus or esophagogastric junction treated with surgery alone was available for our study. However, a comparison of our results to earlier data from our institution and from published reports2, 4, 10, 12–14, 50 showed that the altered posttherapy pathologic stage in our patients treated with preoperative therapy yielded similar survival rates to those stages in patients treated with surgery alone (data not shown).
Current imaging techniques for assessing clinical stage have limited accuracy.7–9, 51 Thus, any assessment of downstaging by comparing posttherapy pathologic stage to pretreatment clinical stage may be biased. Unfortunately, there are no better tools currently available for preoperative staging. As better methods become available, it will be possible to more accurately identify downstaging. However, the predictive ability of postsurgery pathologic stage combined with the finding that a large proportion of patients were downstaged (Table 1) supports the idea that preoperative therapy can be used to obtain the lowest possible stage after surgery.
Preoperative therapy has not consistently demonstrated benefit in patients with localized carcinoma of the esophagus or esophagogastric junction, possibly due to lack of effective multimodality therapy. With the development of more effective therapy, a larger fraction of patients could be shifted to a lower posttherapy stage and thus have a better outcome. Our data suggest that posttherapy pathologic staging is an effective tool for communicating prognosis to patients and health care professionals. Posttherapy staging could also serve as an intermediate end point for well designed Phase II and Phase III trials.
- 11GreeneFL, PageDL, FlemingID, et al., editors. AJCC cancer staging manual. 6th edition. New York: Springer-Verlag, 2002.
- 14Tumors of the esophagogastric junction. Long-term survival in relation to the pattern of lymph node metastasis and a critical analysis of the accuracy or inaccuracy of pTNM classification. J Thorac Cardiovasc Surg. 1996; 111: 85–94; discussion 94–95., , , , , .
- 25Failure in presumption of residual disease by quantification of FDG uptake in esophageal squamous cell carcinoma immediately after radiotherapy. Radiat Med. 2002; 20: 181–186., , , et al.
- 28Irinotecan/cisplatin followed by 5-FU/paclitaxel/radiotherapy and surgery in esophageal cancer. Oncology (Huntingt). 2003; 17: 20–22., , , et al.
- 34Esophagogastrectomy for carcinoma of the esophagus and cardia: a comparison of findings and results after standard resection in three consecutive eight-year intervals with improved staging criteria. J Thorac Cardiovasc Surg. 1997; 113: 836–846; discussion 846–848., , , , .
- 46Pre-operative radiotherapy prolongs survival in operable esophageal carcinoma: a randomized, multicenter study of pre-operative radiotherapy and chemotherapy. The second Scandinavian trial in esophageal cancer. World J Surg. 1992; 16: 1104–1109; discussion 1110., , , et al.
- 47Prediction of pathologic complete response after preoperative chemoradiotherapy in carcinoma of the esophagus and outcome after pathologic complete response. Cancer. In press., , , et al.