Preoperative chemoradiation therapy does not improve early survival after esophagectomy for patients with clinical stage III adenocarcinoma of the esophagus

Jessica S. Donington, MD, Daniel L. Miller, MD, Mark S. Allen, MD, Claude Deschamps, MD, Francis C. Nichols III, MD, and Peter C. Pairolero, MD Division of General Thoracic Surgery, Mayo Clinic and Mayo Foundation, Rochester, Minnesota

Background. The optimal treatment for locally advanced esophageal cancer remains controversial. The objective of this study was to determine if preoperative chemoradiation therapy (P-CRT) followed by esophagectomy for patients with clinical stage III adenocarcinoma of the esophagus offered any survival advantage as compared with esophagectomy alone. Methods. Between January 1998 and June 2001, 75 nonrandomized patients with clinical stage III adenocarcinoma of the esophagus underwent either P-CRT and esophagectomy or esophagectomy alone. All patients were staged before initiation of treatment with computed tomography and endoscopic ultrasound. Results. P-CRT followed by esophagectomy was performed in 47 patients (63%) and esophagectomy alone in 28 patients (37%). Although the P-CRT group was younger (median age, 61 years versus 67 years), the two groups were otherwise similar for gender, comorbidities, and symptoms. Overall operative mortality was 4%. Follow-up was complete in all patients and ranged from 5 to

40 months (median, 20 months). Overall, one-, two-, and three-year survivals were 72%, 44%, and 42%, respectively. Three-year survival was identical (42%) for both the P-CRT and surgery alone patients (p ⴝ 0.70). Threeyear disease-free survival for the P-CRT group was 29% as compared with 33% for the surgery only group (p ⴝ 0.51). Conclusions. Patients with clinical stage III adenocarcinoma of the esophagus do not appear to gain an early overall or disease-free survival advantage when treated with P-CRT followed by surgery as compared with surgery alone. However, long-term follow-up is needed. A large, prospective, randomized trial is warranted to address the question of whether P-CRT offers any survival benefit or impact on pattern of recurrence in patients undergoing esophagectomy for locally advanced disease.

T

overall survival, disease free-survival, or pattern of recurrence.

he incidence of adenocarcinoma of the esophagus has increased significantly over the past decade [1] and adenocarcinoma of the distal esophagus now represents 60%–90% of all esophageal cancers [2]. In the United States, the majority of patients have stage III disease at the time of diagnosis indicating that the disease has spread to the regional lymph nodes (N1) and/or invaded adjacent structures (T4) [3]. Stage III patients have a limited three-year survival when treated with surgery alone. Many institutions have adopted preoperative chemoradiation therapy (P-CRT) protocols in an attempt to improve long-term survival after esophagectomy. The purpose of this report was to analyze our experience with P-CRT in the management of locally advanced stage III adenocarcinoma of the esophagus and to determine if P-CRT provides any benefit in early Presented at the Forty-ninth Annual Meeting of the Southern Thoracic Surgical Association, Miami Beach, FL, Nov 7–9, 2002. Address reprint requests to Dr Miller, Section of General Thoracic Surgery, Emory University Clinic, 1365 Clifton Rd NE, Atlanta, GA 30322; e-mail: [email protected].

© 2004 by The Society of Thoracic Surgeons Published by Elsevier Inc

(Ann Thorac Surg 2004;77:1193–9) © 2004 by The Society of Thoracic Surgeons

Material and Methods Between January 1, 1998 and June 30, 2001, 335 patients underwent esophagectomy at the Mayo Clinic in Rochester, Minnesota for adenocarcinoma of the esophagus. Of these patients, 75 (22%) had documented clinical stage III disease (T3N1M0, T4N0M0 or T4N1M0) by computed tomography (CT) and endoscopic ultrasound (EUS). In order to maximize the accuracy of our pretreatment clinical staging and provide the closest matched groups for comparison, patients without a pretreatment CT scan and EUS were excluded from this review. This study was not a randomized trial; the decision on the use of P-CRT was made on an individual basis with input from the patient, the oncologist, and the surgeon. Our standard P-CRT protocol consists of two cycles of chemotherapy with concomitant radiation therapy. Radiation was administered in 28 daily fractions of 180 cGy, 5 0003-4975/04/$30.00 doi:10.1016/j.athoracsur.2003.09.029

GENERAL THORACIC

Preoperative Chemoradiation Therapy Does Not Improve Early Survival After Esophagectomy for Patients With Clinical Stage III Adenocarcinoma of the Esophagus

GENERAL THORACIC

1194

DONINGTON ET AL ESOPHAGECTOMY AFTER NEOADJUVANT THERAPY

Ann Thorac Surg 2004;77:1193–9

Table 1. Preoperative Demographic Findings P-CRT Age, median (range) Male (%) Tobacco use (%) COPD (%) Steroids (%) Diabetes (%) GERD (%) BMI, median (range) Presenting symptoms Dysphagia (%) Pain (%) Bleeding (%) ⬎10 pound weight loss (%) Symptoms to treatment, median (range) T4 classification (%) Barrett’s disease (%) BMI ⫽ body mass index; chemoradiation therapy.

Surgery Alone

p

61 (36 –74) 45 (96) 33 (70) 5 (11) 2 (4) 6 (13) 22 (47) 28 (18 –35)

67 (44 – 81) 24 (86) 22 (79) 2 (7) 2 (7) 4 (14) 18 (65) 28 (20 –37)

0.011 0.188 0.590 0.706 0.626 1.000 0.159 0.801

41 (87) 21 (45) 4 (9) 28 (60) 3.5 (1–28.5) 9 (19) 18 (38)

26 (93) 8 (29) 1 (4) 12 (43) 3.3 (1–14.9) 0 (0) 14 (50)

0.702 0.222 0.645 0.232 0.905 0.014 0.345

COPD ⫽ chronic obstructive pulmonary disease;

days a week, for a total dose of 5040 cGy. Chemotherapy was administered on the first and last 4 days of the radiation therapy. Five-fluorouracil (5-FU) was administered as a continuous 96-hour infusion (1000 mg/m2/d) and cisplatin (75 mg/m2/d) as a daily intravenous bolus over 1 hour. This was followed by a 4 – 6 week recovery period before esophageal resection. Most esophagectomies were performed using the Ivor Lewis approach. After resection all cancers were postsurgically staged using the American Joint Committee for Cancer Staging (AJCC) TNM classification [4]. The medical records of these patients were analyzed for preoperative factors [age, gender, comorbid conditions, Barrett’s metaplasia, body mass index (BMI), symptoms, and time from symptoms to treatment], tumor factors (pretreatment clinical stage, pathologic stage, and response to therapy), hospital course [length of stay, intensive care unit (ICU) admission, need for ventilatory support, transfusion requirement, morbidity, and operative mortality], tumor recurrence, and survival. Operative mortality included those patients who died within the first 30 days after operation and those who died later but during the same hospitalization. Patients were followed every 3 months after hospital dismissal for the first year and every 6 months thereafter. Extensive evaluation for tumor recurrence was initiated only if the patient had symptoms suggestive of recurrence. The pattern of first recurrence was analyzed after excluding those patients with residual disease at the time of resection and those patients in whom the site of first recurrence was unknown. Deaths were verified from hospital or clinic records and/or social security death index. All deaths were presumed to be secondary to esophageal cancer unless the cause of death was known to be unrelated to the malignancy. Survival was estimated by the Kaplan–Meier method using the date of the esophageal resection as the starting

GERD ⫽ gastroesophageal reflux disease;

P-CRT ⫽ preoperative

point and the date of death or last follow-up as the end point [5]. The influence of variables on survival was analyzed using the log-rank test for discrete variables [6] and the Cox proportional hazards model for continuous variables and multivariable models [7–9]. All statistical tests were two-sided with the threshold of significance set at p ⬍ 0.05. All analyses were conducted using SAS (SAS Institute Inc., Cary, NC). The Institutional Review Board of the Mayo Clinic approved this study.

Results There were 69 men and 6 women. Forty-seven patients (63%) underwent P-CRT followed by esophagectomy and 28 patients (37%) underwent esophagectomy alone. Preoperative demographics for both groups are shown in Table 1. Patients who had P-CRT had significantly more clinical evidence of tumor invasion into adjacent organs (T4) (p ⫽ 0.014) and were significantly younger than those who had surgery alone (p ⫽ 0.01). The two groups were otherwise similar for gender, comorbidities, presenting symptoms, time-interval from symptoms to treatment,

Table 2. Operative Findings P-CRT Surgical approach Ivor Lewis Extended Ivor Lewis (three incisions) Transhiatal Positive margins (%) Intraoperative XRT (%) EBL, median (range)

Surgery Alone

p 0.272

42 (90) 5 (10) 0 (0) 2 (4) 9 (19) 300 (100 –900)

26 (92) 1 (4) 1 (4) 5 (18) 0 (0) 300 (150 –1800)

0.095 0.022 0.912

EBL ⫽ estimated blood loss in cubic centimeters; P-CRT ⫽ preoperative chemoradiation therapy; XRT ⫽ radiation therapy.

DONINGTON ET AL ESOPHAGECTOMY AFTER NEOADJUVANT THERAPY

1195

Table 3. Hospital Course

ICU admission (%) ICU days, median (range) LOS in days, median (range) Any complication (%) Arrhythmia (%) Respiratory complication (%) Transfusion (%) Leak (%) Death (%) ICU ⫽ intensive care unit;

P-CRT

Surgery Alone

p

16 (34) 2.5 (1–39) 10 (0 –52) 19 (40) 10 (21) 8 (17) 23 (49) 3 (6) 3 (6)

6 (21) 7 (1–16) 9 (7– 68) 10 (36) 5 (18) 5 (18) 6 (21) 1 (4) 0 (0)

0.301 0.570 0.929 0.808 0.775 1.000 0.027 1.000 0.289

LOS ⫽ length of stay.

and presence of Barrett’s metaplasia. Thirty-four patients (72%) completed two cycles of chemotherapy and 43 patients (91%) received at least 4500 cGy of radiation. The median time interval between the completion of P-CRT and esophagectomy was 40 days (range, 17–94 days). Details of the operations are shown in Table 2. There was no difference between the groups with regard to surgical approach (p ⫽ 0.272). Ten centimeter proximal and distal resection margins were obtained whenever possible. Seven patients (9%) had resection margins microscopically positive for carcinoma; 2 patients (4%) in the P-CRT group and 5 patients (18%) in the surgery alone group (p ⫽ 0.095). All patients had a two field lymphadenectomy (perigastric and periesophageal) and a pyloroplasty or pyloromyotomy. All anastomosis in the chest were placed above the level of the azygos vein and 52 (70%) were reinforced with an omental or pleura flap. All anastomoses were hand sewn. Suture selection and anastomotic technique varied from surgeon to surgeon. Nine patients underwent intraoperative radiation therapy (IORT), all in the P-CRT group (p ⫽ 0.02). Patients who received IORT in general had larger bulkier tumors on pretreatment staging and this was the main indication for the use of IORT. Thirty-three percent of the patients who received IORT had evidence of T4 tumor classification on pretreatment staging, compared with only 9% of the other patients (p ⫽ 0.071). We observed no significant difference in the rate of morbidity (33% versus 39%, p ⫽ 0.726) or mortality (11% versus 3%, p ⫽ 0.322) after esophagectomy with the use of IORT. Overall, there were three operative deaths (4.0%), all in

Fig 1. Probability of overall survival (death from any cause) in 75 patients with clinical stage III adenocarcinoma of the esophagus who underwent preoperative chemoradiation therapy followed by esophagectomy or esophagectomy alone. Zero time on the abscissa represents the date of esophagectomy.

the P-CRT group (Table 3). The observed difference in mortality between the P-CRT group (6%) and surgery alone (0%) was not statistically significant (p ⫽ 0.289). All deaths were associated with prolonged ICU stays. Cause of death was Pseudomonas aeruginosa sepsis in 2 patients and myocardial infarction in one. Our mortality rate of 4% compares favorably with rates from our own institution [10] and rates for other national cancer institutions and hospitals performing a large number of esophagectomies [11]. Complication rates were similar between the two groups (p ⫽ 0.808). ICU admissions and length of hospitalization were similar but significantly more patients undergoing P-CRT required a blood transfusion in the perioperative period (p ⫽ 0.027). Atrial arrhythmias and respiratory complications including pneumonia, respiratory failure, aspiration, and adult respiratory distress syndrome (ARDS) were the most frequent complications in both groups. Anastomotic leaks occurred in 5%

Table 4. Postsurgical Pathological Stagea Stage

P-CRT

Surgery Alone

T0N0 (%) I (%) IIA (%) IIB (%) III (%) IV (%)

12 (26) 3 (6) 8 (17) 7 (15) 16 (34) 1 (2)

0 (0) 0 (0) 6 (21) 0 (0) 21 (75) 1 (4)

a

Downstaging P-CRT versus surgery alone (p ⫽ 0.001).

P-CRT ⫽ preoperative chemoradiation therapy.

Fig 2. Probability of survival (death from any cause) in 47 patients with clinical stage III adenocarcinoma of the esophagus who underwent preoperative chemoradiation therapy and esophagectomy compared with 28 patients who underwent esophagectomy alone. Zero time on the abscissa represents the date of esophagectomy.

GENERAL THORACIC

Ann Thorac Surg 2004;77:1193–9

GENERAL THORACIC

1196

DONINGTON ET AL ESOPHAGECTOMY AFTER NEOADJUVANT THERAPY

Fig 3. Probability of overall disease-free survival (death from any cause) in 75 patients with clinical stage III adenocarcinoma of the esophagus who underwent preoperative chemoradiation therapy followed by esophagectomy or esophagectomy alone. Zero time on the abscissa represents the date of esophagectomy.

of patients. Complication rates were similar between the groups. Postoperative data for both groups are summarized in Table 3. Postsurgical pathologic stage is outlined in Table 4. Significant downstaging was observed in those patients who received P-CRT (p ⫽ 0.001). Twelve patients (26%) in the P-CRT group had a complete pathologic response (CR) with no viable tumor detected in the resected specimen (T0N0); 3 patients (6%) were stage I and 15 patients (32%) were stage II. Stage III classification did not change in 16 of the 17 remaining P-CRT patients (34%). The remaining patient (4%) was found to have metastatic pulmonary disease at the time of resection and was reclassified as stage IV. In the surgery alone group, 21 patients (75%) remained in stage III after resection, 6 patients (21%) were reclassified as stage II because of false positive lymph node staging by CT scan and EUS, and 1 patient had metastatic disease to the lung (stage IV). Follow-up was complete in all patients and ranged from 5 to 40 months (median, 20 months). In the P-CRT

Fig 4. Probability of disease-free survival (death from any cause) in 47 patients with clinical stage III adenocarcinoma of the esophagus who underwent preoperative chemoradiation therapy and esophagectomy compared with 28 patients who underwent esophagectomy alone. Zero time on the abscissa represents the date of esophagectomy.

Ann Thorac Surg 2004;77:1193–9

Fig 5. Probability of survival (death from any cause) in 12 patients with clinical stage III adenocarcinoma of the esophagus who, after preoperative chemoradiation (CR) therapy, had a complete pathologic response at esophagectomy as compared with 35 patients who had residual disease. Zero time on the abscissa represents the date of esophagectomy.

group 27 patients (57%) are currently alive and 19 patients exhibit no evidence of disease. Twenty patients (43%) have died; 16 patients (34%) died from cancer, 3 patients (6%) died from postoperative complications, and 1 patient (2%) died from respiratory failure. In the surgery group 13 patients (46%) are alive; 9 patients with no evidence of disease. Fifteen patients (54%) died, all from cancer. The overall one-, two- and three-year survivals were 72%, 44%, and 42%, respectively (Fig 1). The one-, two-, and three-year survival after P-CRT therapy was 71%, 42%, and 42% and did not differ significantly from the one-, two-, and three-year survivals observed after surgery alone (p ⫽ 0.70) (Fig 2). Median survival was 20 months in the P-CRT group and 23 months in the surgery alone group. Five patients in the surgery alone group and 2 patients in the P-CRT group had positive surgical margins. One patient in each group was found to have pulmonary metastases at the time of resection. All of these patients were assigned a disease free survival of zero months. Overall one-, two- and three-year disease-free survival was 55%, 30%, and 30%, respectively (Fig 3). Disease-free survival was similar between the treatment groups. The one-, two-, and three-year disease-free survival after P-CRT was 53%, 29%, and 29%, respectively, as compared with 60%, 33%, and 33%, respectively, for the surgery alone group (p ⫽ 0.51) (Fig 4). The disease-free survival was improved from a median of 6 to 16 months in those patients who underwent IORT compared with those who did not (p ⫽ 0.023) but the use of IORT had no effect on overall survival (p ⫽ 0.275). The pattern of first recurrence was evaluated excluding all patients with residual disease at the time of resection and those with an unknown site recurrence during follow-up. More patients in the surgery alone group relapsed locally (6 patients, 30%) than those who received P-CRT (3 patients, 8%). This observation was not significant (p ⫽ 0.10), but then again more patients in the P-CRT group had T4 disease. The rate of distance relapse

was similar between the groups; 9 patients (24%) in the P-CRT group and 3 patients (16%) in the surgery alone group (p ⫽ 0.14). Patients in the P-CRT group were assessed for pathologic response to treatment at the time of surgery. No residual disease (T0N0) was found in resected specimens in 12 patients (26%) and all were considered to have a complete pathologic response (CR). However, no correlation was observed between CR and survival (p ⫽ 0.97). Patients with CR had one-, two- and three-year survival of 69%, 46%, and 46%, respectively, as compared with 71%, 42%, and 42%, respectively, for the 35 patients with residual disease (Fig 5).

Comment Resection of the esophagus for cancer has become a much safer procedure because of improved anesthetic and surgical techniques. However despite these advances the long-term prognosis for this disease remains poor [12]. Based on the favorable results with P-CRT for patients with colorectal cancer [13–15], similar treatment protocols have been developed for patients with esophageal cancer [16]. Potential advantages of P-CRT include cancer downstaging, increased respectability, and reduction in micrometastases. In addition chemotherapeutic agents such as cisplatin, 5-FU, and mitomycin all possess radiosensitizing properties and have proven effective in the management of esophageal cancer [2]. Combining these agents with radiation therapy has become widely accepted as the standard of care despite the lack of evidence of clinical benefit [17]. Only two randomized trials evaluating the effectiveness of P-CRT have been reported [16, 18]. The first trial was reported by Walsh and associates, who randomized 113 patients to 5-FU, cisplatin, and 4000 cGy of radiation plus esophagectomy versus esophagectomy alone [16]. Although the threeyear follow-up results demonstrated a highly significant survival advantage (p ⫽ 0.01) in the P-CRT group, the study has been widely criticized for the disproportionally poor outcome in the surgery alone group. This study also lacked staging techniques available today such as EUS. In addition CT was obtained only selectively. The second prospective randomized trial from the University of Michigan by Urba and associates [18] randomized 100 patients with carcinoma of the esophagus to 5-FU, cisplatin, vinblastine, and 4500 cGy of radiation plus surgery versus surgery alone. Seventy-five percent of patients had adenocarcinoma and 25% had squamous cell carcinoma. Whereas a prior pilot trial [19, 20] from the same institution demonstrated a three-year survival advantage with P-CRT, the larger randomized trial failed to demonstrate a significant survival advantage with P-CRT, a finding similar to our retrospective study of stage III patients. Failure to clearly demonstrate the effectiveness of PCRT was the impetus for the development of the multicenter phase III Intergroup trial C9781. The Intergroup trial was to enroll 500 patients, but, unfortunately, accrual was poor and the study was closed prematurely without

DONINGTON ET AL ESOPHAGECTOMY AFTER NEOADJUVANT THERAPY

1197

properly addressing the issue. Many believe that the reason for the poor accrual was the widespread acceptance of P-CRT for esophageal cancer despite the lack of evidence for clinical benefit. At our institution a treatment bias clearly exists where patients with clinical stage I and II disease are more likely to undergo resection alone whereas patients with stage III are more frequently treated with P-CRT. Because the majority of patients in the United States present with locally advanced disease we limited our review to patients with stage III disease and only included patients with adenocarcinoma of the esophagus who had undergone extensive pretreatment staging with CT and EUS. Patients without both CT and EUS pretreatment staging were excluded from this evaluation in order to achieve the best clinical staging currently available. This was not a randomized trial and the decision to undergo P-CRT or surgery alone was made on an individual basis. This may explain why younger patients and those with bulkier tumors were in the P-CRT arm. In our patients no survival advantage was demonstrated with P-CRT. Median survival was 20 months in the P-CRT group and 23 months in the surgery alone group. Whereas our study is retrospective and nonrandomized the survival results for both our P-CRT and surgery alone patients (41% threeyear survival) are consistent with the survival results for the P-CRT patients reported by both Walsh and associates [16] and Urba and associates [18]. Despite the failure to demonstrate a survival advantage with P-CRT one promising aspect with this regime is the potential for tumor downstaging. All major series report a consistent patient group that varies from 17%– 40% who do not have evidence of residual tumor in their resected specimens [16, 17–25]. Five-year survival in this subgroup of patients has been reported as high as 60% [22]. Unfortunately, no pretreatment characteristics have been identified that indicate which patients are more likely to respond to therapy [10, 26]. Although 26% of our patients had a CR at the time of resection, no early survival advantage was demonstrated. One possible explanation for our observation may be the relatively short follow-up period in the CR group. Another explanation may be that we excluded patients with early stage disease from this analysis and these may be the patients that see the greatest survival benefit after P-CRT [10]. In summary this retrospective study suggests that preoperative chemoradiation therapy followed by esophageal resection for adenocarcinoma of the esophagus provides better local control than surgery alone but has no effect on early overall or disease-free survival. Complete response after P-CRT did not improve early survival. A large multicenter randomized prospective trial is needed to answer the question of whether preoperative chemoradiation therapy is of any benefit. Until then surgery should remain the standard of care for patients with resectable adenocarcinoma of the esophagus, whereas preoperative chemoradiation therapy should be limited to clinical trials.

GENERAL THORACIC

Ann Thorac Surg 2004;77:1193–9

GENERAL THORACIC

1198

DONINGTON ET AL ESOPHAGECTOMY AFTER NEOADJUVANT THERAPY

Ann Thorac Surg 2004;77:1193–9

References 1. Kirby TJ, Rice TW. The epidemiology of esophageal carcinoma: the changing face of the disease. Chest Surg Clin N Am 1994;4:217–25. 2. Forastiere AA, Heitmiller RF, Kleinberg L. Multimodality therapy for esophageal cancer. Chest 1997;112:195S–200S. 3. DeCamp MM, Swanson SJ, Jaklitsch MT. Esophagectomy after induction chemoradiation. Chest 1999;116:466S–9S. 4. American Joint Committee for Cancer Staging and EndResult Reporting. Cancer of the esophagus. In: Beahr OH, Henderson DE, Hutter RVP, Kennedy BJ, eds. Manual for staging of cancer, 4th ed. JB Lippincott: Philadelphia, PA, 1993:75–9. 5. Kaplan E, Meier P. Non-parametric estimation from incomplete observations. J Am Stat Assoc 1958;53:457–81. 6. Peto R, Peto J. Asymptotically efficient rank invariant procedures (with discussion). J R Stat Soc Ser A 1972;135:185–207. 7. Cox DR. Regression models and life-tasks. J R Stat Soc Ser B 1972;34:187–220. 8. Snedecor GW, Cochran WG. Statistical methods. Iowa State University Press: Ames, IA, 1967:100 – 6. 9. Mehta CR, Patel NR. A network algorithm for performing Fisher’s Exact Test in rXc Contingency Tables. J Am Stat Assoc 1983;78:427–34. 10. Donington JS, Miller DL, Allen MS, et al. Tumor response to induction chemoradiation: influence on survival after esophagectomy. Eur J Cardiothor Surg (in press). 11. Swisher SG, Deford L, Merriman KW, et al. Effects of operative volume on morbidity and mortality, and hospital use after esophageal cancer. J Thorac Cardiovasc Surg 2000; 119:1126 –34. 12. Orringer MB, Marshall B, Iannettoni MD. Transhiatal esophagectomy: clinical experience and refinements. Ann Surg 1999;230:392–403. 13. Buroker TR, Nigro N, Bradley G, et al. Combined therapy for cancer of the anal canal: a follow-up report. Dis Colon Rectum 1977;20:677–8. 14. Newman HK, Quan SH. Multimodality therapy for epidermoid carcinoma of the anus. Cancer 1976;37:12–9. 15. Sischy B, Remington JH, Sobel SH, Savlov ED. Treatment of carcinoma of the rectum and squamous carcinoma of the

16. 17. 18.

19.

20.

21. 22. 23.

24.

25.

26.

anus by combination chemotherapy, radiotherapy and operation. Surg Gynecol Obst 1980;151:369 –71. Walsh TN, Noonan N, Hollywood D, et al. A comparison of multimodality therapy and surgery for esophageal adenocarcinoma. N Engl J Med 1999;335:462–7. Forastiere AA, Heitmiller RF, Kleinberg L. Multimodality therapy for esophageal cancer. Chest 1997;112(suppl):195S– 200S. Urba SG, Orringer MB, Turrisi A, et al. Randomized trial of preoperative chemoradiation versus surgery alone in patients with locoregional esophageal carcinoma. J Clin Oncol 2001;19:305–13. Forastiere AA, Orringer MB, Perez-Tamayo C, et al. Concurrent chemotherapy and radiation therapy followed by transhiatal esophagectomy for local-regional cancer of the esophagus. J Clin Oncol 1990;8:119 –27. Forastiere AA, Orringer MB, Perez-Tamayo C, et al. Preoperative chemoradiation followed by transhiatal esophagectomy for carcinoma of the esophagus: final report. J Clin Oncol 1993;11:1118 –23. Popli E, Flemming T, Leichman L, et al. Combined therapies for squamous cell carcinomas of the esophagus, a Southwest Oncology Group (SWOG 8037). J Clin Oncol 1987;5:622–8. Naunheim KS, Petrosku PJ, Roy TS, et al. Multimodality therapy for adenocarcinoma of the esophagus. Ann Thorac Surg 1995;59:1085–91. Adelstein DJ, Rice TW, Becker M, et al. Use of concurrent chemotherapy, accelerated fractional radiation, and surgery for patients with esophageal carcinoma. Cancer 1997;80: 1011–20. Forastiere AA, Heitmiller RF, Lee DJ, et al. Intensive chemoradiation followed by esophagectomy for squamous cell and adenocarcinoma of the esophagus. Cancer J Sci Am 1997;3: 144 –52. Jones DR, Detterbeck FC, Egan TM, et al. Induction chemoradiotherapy followed by esophagectomy in patients with carcinoma of the esophagus. Ann Thorac Surg 1997;64:185– 91. Zuccaro G, Rice TW, Goldblum J, et al. Endoscopic ultrasound cannot determine suitability for esophagectomy after aggressive chemotherapy for esophageal cancer. Am J Gastroenterology 1999;94:906 –12.

DISCUSSION DR BRIAN F. MEYERS (St. Louis, MO): I congratulate you on results of taking care of a tough set of patients; the morbidity and the mortality in both arms of that trial are admirable. I think all of us who do esophagectomies have had our hand forced by the results of the Walsh trial in 1996 that really ignited a lot of enthusiasm for induction chemoradiation. In addition to the criticisms that many have had with that trial, there has subsequently been a lot of information, including your own paper, that raises serious questions about the benefit to patients from induction chemoradiation. In addition to your work, Dr. Rice at the Cleveland Clinic showed us that patients with the lowest stages likely experienced the negative effects of the induction therapy without receiving much benefit, and the group at the Brigham, with Scott Swanson presenting, showed us that older patients in particular may be at greater risk for harm by trimodality therapy. So I congratulate you, and I don’t have a criticism or a question, but I point out that some bias present in this study would actually tend to favor the induction therapy patients and thereby reinforce your conclusions. You had younger patients getting the induction therapy, so if they did better, one might criticize the study by saying that the induction patients did better only because they were younger, but that wasn’t the case.

Also, one finds that with close scrutiny of retrospective reports of induction therapy that some patients become M1 with disease progression and get excluded from surgery after induction therapy. If you only look at the patients who have had an operation, rather than intention to treat, you focus on a better prognostic group. Neither of those biases seemed to help your induction therapy arm, and I think your work reinforces that there remain a lot of doubt about the blind acceptance of induction therapy. That was a very nice paper. DR CAROLYN E. REED (Charleston, SC): I also rise to congratulate you on this very well presented paper, and I think it points out that we still have an incredible amount of work to do in esophageal cancer. It is a shame that one attempt at a multiinstitutional trial to answer your question was not successful, and I think that it is actually going to be very hard to mount a multi-institutional trial based on some previously published papers. I am actually interested in asking a question about a subset of your patients. I think I saw this correctly that 19 of your patients had intraoperative radiation therapy. If that is true, that is about 40% of your patients that had neoadjuvant therapy, and I just

wish you would tell me a little bit more about that group and why they got intraoperative therapy, if that is a routine now at your institution, and if so, what are the criteria to use intraoperative treatment? I also wondered if you would tell me how many of your stage III patients were T4, not T3N1 but actually T4 patients. Do you operate on all of your T4 patients or did you exclude some of the T4 patients? DR DONINGTON: Thank you for your comments. To start with, I may have misspoke, but the intraoperative radiation therapy was actually given to nine patients in the neo-adjuvant group, which represents 19%. I could have easily said that backwards. That number makes sense and it is something I remember because there are also nine patients who are T4. And I would like to say that all of them received all the IORT, but the majority of the intraoperative radiation therapy was given to patients who had a pretreatment T4 and there was only one patient who had a T3 who received intraoperative IORT. So therefore one of the main criteria for our use would be invasion into adjacent organs.

DONINGTON ET AL ESOPHAGECTOMY AFTER NEOADJUVANT THERAPY

1199

DR KAMAL A. MANSOUR (Atlanta, GA): I enjoyed the paper very much. I think I need to compliment the group at Mayo Clinic for an excellent mortality rate. You have a 4% mortality, and I think that is the acceptable standard now, 4 to 5%; we are far ahead the 25% in the past reported by Mr. Andrew Logan in the early sixties. The question I have, actually, you answered it right in your last sentence of your conclusion, is a need for a randomized multicenter trial for long-term survival. I noticed that the longest follow-up in your series is 36 months? DR DONINGTON: Right, the follow-up was just over three years in the longest patient.

DR REED: Most of us would reject T4 patients as not being operative candidates. Is that your feeling now?

DR MANSOUR: And that worries me, because if you leave any patient with cancer of the esophagus alone without chemotherapy or surgery, it will take him about two to three years to die. So I am in agreement with you as I never use any preoperative chemotherapy, and that is the main problem. I have been watching this for many years; one paper is for it and another against it, and I don’t know whether we will ever have the final answer in our lifetime. So my question is, will these results in the final analysis stand the test of time being so short a followup? Thank you.

DR DONINGTON: That we would reject them? Actually not, because actually that group of patients had a very mixed result. Of the nine patients who were T4, one patient actually had a complete pathologic response, three of them are still alive, and only one patient had positive margins at the time of resection. So, therefore, now at the Mayo Clinic having a T4 tumor is not something that would definitely make you not a surgical candidate, especially in the face of neoadjuvant treatment.

DR DONINGTON: Thank you for your comments. I do think that the follow-up is short here and that this data may change as it matures. One group that experienced that was the group out of the University of Michigan where their pilot trials looking at neoadjuvant therapy originally looked like they had a benefit, but the longer that they watched it grow, that benefit seemed to disappear. It would be nice to look and see what this data looks like three to five years from now.

GENERAL THORACIC

Ann Thorac Surg 2004;77:1193–9