IJGC_069_076_Roha.qxd

2/15/06

8:04 AM

Page 69

International Journal of Gastrointestinal Cancer, vol. 36, no. 2, 69-76, 2005 © Copyright 2005 by Humana Press Inc. All rights of any nature whatsoever reserved. 0169-4197/05/36:69–76/$30.00

Research Article

Comparison of Clinical Stage, Therapy Response, and Patient Outcome Between Squamous Cell Carcinoma and Adenocarcinoma of the Esophagus Pooja R. Rohatgi, Stephen G. Swisher, Arlene M. Correa, Tsung-T. Wu, Zhongxing Liao, Garrett L. Walsh, Ara A.Vaporciyan, David C. Rice, Norio Fukami, Jack A. Roth, and Jaffer A. Ajani* Departments of Gastrointestinal Medical Oncology, Thoracic and Cardiovascular Surgery, Radiation Oncology, and Pathology, and Gastrointestinal Medicine and Nutrition, The University of Texas MD Anderson Cancer Center, Houston, TX

Abstract Purpose: To analyze the differences in clinical stage, pathologic response to chemoradiotherapy, patterns of failure, and overall survival (OS) between patients with squamous cell carcinoma (SCC) and adenocarcinoma (ACA) of the esophagus. Patients and Methods: We stratified patients by two histologies, ACA and SCC, and statistically compared their clinical stage, post-therapy pathologic response, patterns of failure, and OS. Results: Of the 235 patients who underwent preoperative chemoradiotherapy, 42 (18%) had SCC and 193 (82%) had ACA. Among the ACA patients, a significantly larger proportion was male (93% vs 7%; p <0.001), whereas sex was distributed similarly among SCC patients (55% male vs 45% female; p = 0.5). A significantly larger percentage of SCC patients were classified as lower TN and overall stage than ACA patients (T2 = 41% vs 28%, p <0.0001; N0 = 69% vs 48%, p = 0.01; stage II = 76% vs 55%, p <0.001). A significantly greater portion of SCCs was categorized as pathologic N0 after treatment (71% vs 65%; p = 0.02). Among the pathCR patients in clinical stage II, there were significantly greater proportion of SCC patients (77% vs 63%; p <0.001) than ACA patients. Among the pathCR patients in clinical stage III patients, a significantly greater proportion were ACA patients (38% vs 23%; p <0.001) than SCC patients. The median and 5-yr OS was 53 ± 11 mo and 39% for ACA patients and 35 ± 14 mo and 37% for SCC (median OS, p = 0.3). Among pathCR patients, median OS of ACA patients (133 mo) was longer than that of SCC patients but nonsignificant (29 mo; p = 0.07); results were similar for non-pathCR patients. DFS results were similar in all subgroups. Among the whole cohort, incidence of local-regional recurrence and distant metastases did not vary significantly. The median time to distant metastases did not vary significantly for pathCR and non-pathCR patients. Conclusions: We believe this is the first study that compares failure outcome of ACA and SCC patients with similar clinical stage after trimodality therapy. Our data suggest that significant differences in clinical stage and post-therapy pathologic stage exist between ACA and SCC. Frequent presence of malignant nodes in the resected specimens of ACA patients resulted in a shorter time-to-metastases suggesting that ACA patients need better systemic control. Key Words: Adenocarcinoma of esophagus; squamous cell carcinoma of esophagus; *Author to whom all correspondence and reprint requests should be addressed: Jaffer A. Ajani, MD, Department of Gaschemoradiotherapy; pathologic response. trointestinal Medical Oncology, Unit 426, 1515 Holcombe Blvd, Houston, Texas 77030. Tel: 713-792-2828; Fax: 713-745-1163; E-mail: [email protected].

69

IJGC_069_076_Roha.qxd

2/15/06

8:04 AM

Page 70

70

Rohatgi et al.

Introduction

Patients and Methods

Esophageal carcinoma is rare in the United States, yet has one of the lowest estimated 5-yr survival rates of 10% or less (1–4). Thus combined modality therapies, in a variety of ways, have been tried to improve survival of these patients (3–7). Despite refinements in available therapy, the statistics remain unchanged, suggesting that other aspects of the cancer be analyzed. Surprisingly, the current literature is scarce on the role of histology on patient outcome after trimodality therapy. We reviewed nine studies published since 1996, reporting ≥30 patients who underwent chemoradiotherapy (CTRT) followed by esophagectomy (median number of patients = 44). Of the total 419 patients, 219 (52%) had squamous cell carcinoma (SCC) and 198 (47%) had adenocarcinoma (ACA) (2,4–11). Of the five studies that reported pathCR (pathologic complete response) by histology, the median percentage of SCC and ACA patients who achieved a pathCR were 43% and 33%, respectively. Of all the pathCR patients, a median of 60% were SCC patients and 40% were ACA patients (2,4,5, 8,11). From the two studies that reported 3-yr survival analysis, the 3-yr OS rate for SCC was 38% and for ACA was 41% 7,9. In three studies, median OS did not differ significantly by histology (6,8,9). Only one study found that SCC was associated with better survival (4). Therefore, it would appear that patients with SCC have a slightly higher rate of pathCR; however, survival of these populations is not significantly different. None of these studies have reported on the patterns of failure by histology after preoperative CTRT. Neither did the studies examine the distribution of clinical stage or pathologic stage post-therapy. In the current study, we compared specific pretreatment and post-treatment parameters between SCC and ACA patients. We then compared OS and failure patterns of SCC and ACA patients with similar pathologic response to determine if histology correlates with outcome within different pathologic response groups. We expected that there would be little difference in the type of pathologic response or patient outcome by histology, as has been observed by others. A total of 235 consecutive patients who were seen and had undergone preoperative CTRT at UT MD Anderson Cancer Center were analyzed.

Patient Eligibility

International Journal of Gastrointestinal Cancer

Operable patients with localized, potentially resectable carcinoma of the esophagus evaluated at The University of Texas MD 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 TNM stage II or III esophageal carcinoma were included (included also were patients involving the gastroesophageal junction with M1a nodes), whereas patients with T1N0, T4, or M1b cancer 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 an approved written informed consent.

Pretreatment Evaluation All clinical staging was performed according to AJCC/UICC criteria for staging esophageal carcinoma (12). Clinical staging was performed before therapy and included a complete history and physical examination, chest radiography, computer tomography of the chest and abdomen, upper gastrointestinal double-contrast barium radiography, positron emission tomography (when available), an esophagogastroduodenoscopy with endoscopic ultrasonography (EUS), electrocardiography, a chemistry screen test, complete blood count including platelet count, and measurement of serum, electrolyte, and baseline carcinoembryonic antigen levels. Additional preoperative screening studies were done as needed. All 235 patients were staged with EUS. Criteria for staging by EUS was according to previously published guidelines (13).

Evaluation During Therapy Patients’ symptoms and the results of blood tests were closely monitored during therapy. Patients had repeat endoscopy or other imaging studies as warranted by their clinical symptoms. Volume 36, 2005

IJGC_069_076_Roha.qxd

2/15/06

8:04 AM

Page 71

Comparison of Squamous Cell and Adenocarcinoma of Esophagus

Chemoradiation Therapy Some patients received induction chemotherapy before chemoradiotherapy and others received preoperative chemoradiotherapy only. Details of chemotherapy on this cohort have been published (14). Radiation therapy was delivered using cobalt-60 or higher-energy photons with conventional fractionation, starting at d 1 with concurrent chemotherapy. The initial target volume encompassed the primary cancer with a minimal 5-cm margin cephaladly and caudally and 2-cm margin radially. The radiation dose ranged from 45 Gy in 25 fractions to 50.4 Gy in 28 fractions. Two- or three-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 anteroposterior portals until the spinal cord tolerance dose was reached, followed by oblique fields to avoid the spinal cord. Radiation dose 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 Prior to Surgery Four to six weeks after completion of CRT, patients underwent another complete staging work up including a positron emission tomography (when available) but not EUS (13).

71

ual 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 grossly identified absence of residual carcinoma. Apathologist (TTW), 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 in one of two groups: as showing a pathCR or
Statistical Methods Chi-square test or Fisher’s exact test, when necessary, were performed to compare proportion of pathologic and clinical parameters, and failure pattern rates between SCC and ACA patients. Kaplan– Meier curves were plotted for OS, DFS, localregional recurrence-free survival and distant metastases-free survival of patients from the whole cohort, of only non-pathCR patients, and of only pathCR patients, stratified by histology. Log-rank test was used to compare survival curves of SCC and ACA patients. Multivariable Cox regression analysis was also performed to compare time to failure between SCC and ACA patients. Statistical analysis was performed using SPSS software (version 11.5.2.1 for Windows; SPSS, Chicago, IL). All statistical analyses were performed with a two-sided significance value of 0.05.

Results Patient Characteristics

A curative radical en-bloc esophagectomy was attempted in all patients, and the resected specimens analyzed to determine the post-therapy pathologic AJCC/ UICC stage. The surgical approaches used were Ivor–Lewis esophagectomy (abdominal-right thoracic approach) in 130 patients (55%), three-field McKeown esophagectomy (right thoracic abdominal-cervical approach) in 44 patients (19%), and transhiatal esophagectomy (abdominal-cervical approach) in 59 patients (25%). Two patients (1%) underwent two-stage esophagectomy (15,16).

Two hundred and thirty five consecutive patients who were evaluated at the University of Texas MD Anderson Cancer Center from 1985 to 2003, were histologically confirmed to have either ACA or SCC of the esophagus, and underwent pretreatment clinical staging followed by preoperative chemoradiotherapy were analyzed. Of these, 193 (82%) patients were diagnosed with ACA and 42 (18%) with SCC. Distribution of sex varied significantly between the two histologies (p <0.001; Table 1). Of the ACA patients, a significantly larger portion was male (93% vs 7%; p <0.001), whereas sex was distributed nearly equal (55% male vs 45% female; p = 0.5) among SCC patients.

Post-Surgical Evaluation

Clinical and Pathologic Parameters

Postsurgical staging was based on pathologic findings in the resected specimen, in particular the resid-

A significantly larger portion of SCC than ACA was clinically classified as lower T2-stage(T2= 41%

Surgery

International Journal of Gastrointestinal Cancer

Volume 36, 2005

IJGC_069_076_Roha.qxd

2/15/06

8:04 AM

Page 72

72

Rohatgi et al. Table 1 Comparison of Clinical and Pathologic Parameters by Histology ACA (n = 193)

SCC (n = 42)

n

(%)a

n

(%)a

Male Female

180 13

(93) (7)

23 19

(55) (45)

T2 T3 T4

53 137 3

(28) (71) (2)

17 25 0

(41) (60) (0)

92 101

(48) (52)

29 13

(69) (31)

105 76 12

(55) (39) (6)

32 9 1

(76) (21) (2)

0 125 68

(0) (65) (35)

2 30 10

(5) (71) (24)

Sex

p value <0.001

T

N N0 N1 Clinical Stage Stage II Stage III Stage IVA Pathologic Nb NX N0 N1

0.2 <0.001 <0.001 0.01 <0.001 <0.001 0.06 <0.001 <0.001 0.02 <0.001 <0.001

First p value for each parameter obtained from chi-square test or Fisher’s exact test (where necessary) comparing overall results between histologies to determine correlation. Subsequent numbers under each parameter obtained from chi-square tests to determine whether counts for subparameter are significantly different between two histologies. a Represents percent within histology. Percentages are rounded so total may not equal 100. b Pathologic N was determined post-treatment.

vs 28%; p<0.001). A significantly larger proportion of SCC than ACA also had no clinical nodal involvement (N0 = 69% vs 48%; p = 0.01). Overall, SCC was associated with clinical stage II, while ACA was associated with clinical stage III, although not significantly (p = 0.06); however, when clinical stages were analyzed separately, a significantly larger percentage of SCC than ACA was classified as lower clinical stage (stage II = 76% vs 55%; p <0.001). Post-treatment pathologic T or pathologic stage distribution did not vary significantly between the two types of histology (p = 0.3 and 0.09, respectively). A significantly greater portion of SCC than ACA had no pathologic nodal involvement after treatment (N0=71% vs 65%; p = 0.02). Comparison data are summarized in Table 1. International Journal of Gastrointestinal Cancer

Pathologic Response Rates There was no significant association between either of the two histologies and overall pathCR rate (p = 0.8) or pathCR rate stratified by clinical stage (p = 0.5). However, when analyzed separately, pathCR in clinical stage II had a significantly greater proportion of SCC patients than ACA patients (77% vs 63%; p <0.001) while pathCR in clinical stage III had a significantly greater proportion of ACApatients than SCC patients (38% vs 23%; p <0.001). These results are summarized in Table 2.

Survival and Disease-Free Survival Analyses Median follow-up time was 37 mo. Median OS for ACApatients was 53 ± 11 mo and for SCC patients was 35 ± 14 mo (p = 0.3). Five-year OS rate was Volume 36, 2005

IJGC_069_076_Roha.qxd

2/15/06

8:04 AM

Page 73

Comparison of Squamous Cell and Adenocarcinoma of Esophagus

73

Table 2 Comparison of Pathologic Response Patterns by Histology ACA (n = 193)

Pathologic Response Complete (pathCR) Not Complete PathCR Rate by Clin.b Stage Stage II Stage III

SCC (n = 42)

n

(%)a

n

(%)a

56 137

(29) (71)

13 29

(31) (69)

35 21

(63) (38)

10 3

(77) (23)

p value 0.8

0.5 <0.001 <0.001

a Represents percent within histology. Percentages are rounded so total may not equal 100. b Clin.=clinical.

Table 3 Comparison of Failure Patterns by Histology ACA (n = 193)

Local-regional Recurrence Distant metastases Death Death or Relapse

SCC (n = 42)

n

(%)a

n

(%)a

p value

20 53 67 83

(10) (27) (35) (43)

5 8 23 25

(12) (19) (55) (60)

0.8 0.3 0.02 0.05

p values obtained from chi-square or Fisher’s exact test (where necessary) comparing overall results between histologies to determine correlation. a Represents percentage within histology. Percentages are rounded so total may not equal 100.

39% for ACA patients and 37% for SCC patients. A significantly larger proportion of ACA patients (n = 126; 65%) than SCC patients (n = 19; 45%) remain alive (p = 0.02). Among pathCR patients, median OS for ACA patients (133 mo) was close to significantly greater than median OS for SCC patients (29 mo; p = 0.07). Kaplan–Meier curves suggest a trend between ACApatients and longer OS among pathCR patients (Fig. 1). Among non-pathCR patients, median OS for SCC patients was 35 ± 16 mo and for ACA patients was 32 ± 5 mo (p = 0.9). Among stage II patients, OS did not vary significantly between histology (p = 0.3) and among stage III patients, OS did not vary significantly between histology (p = 0.9). DFS followed the same patterns as OS. A significantly larger proportion of ACA patients (n = 110; 57%) than SCC patients (n = 17; 40%) survived withInternational Journal of Gastrointestinal Cancer

out disease (p = 0.05). Five-year DFS rate was 38% for ACA patients and 36% for SCC patients. Median DFS for SCC patients was 29 ± 13 mo and for ACA patients was 20 ± 9 mo (p = 0.4). Among pathCR patients, median DFS for ACA patients was 133 mo and for SCC patients was 29 ± 11 mo (p = 0.1). Kaplan–Meier curves suggest a trend between ACA patients and longer DFS time among pathCR patients. Among non-pathCR patients, median DFS for SCC patients was 29 ± 19 mo and for ACApatients was 17 ± 4 mo (p = 1.0).

Patterns of Failure Twenty (10%) ACA patients and 5 (12%) SCC patients had local-regional recurrence (p = 0.8). Fiftythree (27%) ACA patients and 8 (19%) SCC patients had distant-metastases (p = 0.3; Table 3). Median time-to-distant metastases for ACA patients was 28 Volume 36, 2005

IJGC_069_076_Roha.qxd

2/15/06

8:04 AM

Page 74

74

Rohatgi et al.

Fig. 1. OS stratified by cancer histology of (A) all patients, (B) patients with less than pathCR, and (C) patients with pathCR, using Kaplan–Meier method.

± 13 mo, while it was not reached for SCC patients (p = 0.1). Multivariable Cox regression analysis suggests a close to significant correlation between SCC and longer time to distant metastases in the whole cohort (p = 0.07). Among pathCR patients, median time to distant metastases for ACA and SCC patients was not reached (p = 0.5). Among non-pathCR patients, median time to distant metastases for ACA was 17 ± 2 mo, while it was not reached for SCC patients (p = 0.1). Kaplan–Meier curves suggest a trend between SCC patients and longer time to distant metastases for all three groups (Fig. 2). Time to local-regional metastases or time to event, defined as death or relapse, did not vary significantly between the two types of histology in any of the three International Journal of Gastrointestinal Cancer

groups, based on multivariable Cox regression or log-rank analyses.

Discussion Despite advancements in the treatment of esophageal carcinoma, survivorship remains unacceptably low, perhaps reflecting our failure to select appropriate therapy for each case. Therefore, it may be important to identify parameters that could further characterize each cancer. We felt that there must be difference in clinical biology of patients presenting with two distinct types of histology. Although, as stated previously, the limited information in the literature would suggest that one is less likely to find Volume 36, 2005

IJGC_069_076_Roha.qxd

2/15/06

8:04 AM

Page 75

Comparison of Squamous Cell and Adenocarcinoma of Esophagus

75

Fig. 2. Distant-metastases-free survival by cancer histology of (A) all patients, (B) patients with less than pathCR, and (C) patients with pathCR, using Kaplan–Meier method.

differences in overall survival but no detailed analyses have been published. Our study, with one of the largest patient populations with localized carcinoma of the esophagus treated similarly, reveals many interesting differences. As expected, the proportion of men with ACA is much higher than it is with SCC; this is most likely due to the fact that SCC is closely associated with lifestyle than it is with gender. Surprisingly, a significantly higher proportion of patients presented with T2 in the SCC group and significantly higher number presented with T3 in the ACA group. ACA patients were more likely to have N+ cancer than SCC patients. A larger proportion of SCC patients had clinical stage II cancer and a larger proportion International Journal of Gastrointestinal Cancer

of ACA patients had either stage III or IVa. We hypothesize that because SCC occurs predominantly in the upper esophagus, where other structures crowd the thoracic cavity leading to less esophageal pliability, physical obstruction by a mass will be detected earlier. We are thus able to diagnose SCC at an earlier stage. These differences obviously have a significant impact on the clinical biology and treatment outcome of the cancers. Like other studies, the OS for both groups was similar and the pathCR rates for the two groups was similar; however, when analyzing the rate of pathCR by stage, we found that in stage II patients, SCC was in greater proportion and in stage III patients, it was ACA in greater proportion. When ACA patients Volume 36, 2005

IJGC_069_076_Roha.qxd

2/15/06

8:04 AM

Page 76

76 achieved a pathCR, they tended to have a better outcome than when SCC patients achieved a pathCR. The possible explanation for this type of outcome lies in differences in cancer biology. The patterns of failure were consistent with stage of the cancer. Among those who did not achieve a pathCR, ACApatients fared worse than SCC patients because of the frequent Npositive cancer in the ACA group. Kaplan–Meier curves suggest that SCC patients have longer time-to-distant metastases probably due to lower clinical stage than those with ACA. To the best of our knowledge, this is the only study that compares outcome after trimodality therapy of ACA and SCC patients within similar pathologic groups. Further studies are needed to pursue these observations. Nonetheless, it is clear that to improve the outcome of ACA patients, a more frequent histology in the West, we need improved therapeutic approaches to control the occult metastases, in addition to localized one.

Acknowledgements Supported in part from by grants from the Rivercreek Foundation, and the Smith, Cantu, and Dallas families.

References 1. Posner MC, Gooding WE, Landreneau RJ, et al. Preoperative chemoradiotherapy for carcinoma of the esophagus and gastroesophageal junction. Cancer J Sci Am 1998; 4(4):237–246. 2. Jones DR, Detterbeck FC, Egan TM, et al. Induction chemoradiotherapy followed by esophagectomy in patients with carcinoma of the esophagus. Soc Thoracic Surg 1997; 64:185–192. 3. 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(1):119–127.

International Journal of Gastrointestinal Cancer

Rohatgi et al. 4. De Vita F, Martino ND, Orditura M, et al. Preoperative chemoradiotherapy for squamous cell carcinoma and adenocarcinoma of the esophagus. Chest 2002;122:1302–1308. 5. Makary MA, Kiernan PD, Sheridan MJ, et al. Multimodality treatment for esophageal cancer: the role of surgery and neoadjuvant therapy. Am Surg 2003;69:693–702. 6. Heath EI, Burtness BA, Heitmiller RF, et al. Phase II evaluation of preoperative chemoradiation and postoperative adjuvant chemotherapy for squamous cell and adenocarcinoma of the esophagus. J Clin Oncol 2000;18(4):868– 876. 7. Lew JI, Gooding WE, Ribeiro U Jr, et al. Long-term survival following induction chemoradiotherapy and esophagectomy for esophageal carcinoma. Arch Surg 2001;136(7):737–743. 8. 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(3):144–152. 9. Stahl M, Wilke H, Fink U, et al. Combined preoperative chemotherapy and radiotherapy in patients with locally advanced esophageal cancer. Interim analysis of a phase II trial. J Clin Oncol 1996;14(3):829–837. 10. Meneu-Diaz JC, Blazquez LA, Vicente E, et al. The role of multimodality therapy for resectable esophageal cancer. Am J Surg 2000;179:508–513. 11. Slater MS, Holland J, Faigel DO, et al. Does neoadjuvant chemoradiation downstage esophageal carcinoma? Am J Surg 2001;181:440–444. 12. Greene FL, Page DL, Fleming ID, et al. AJCC Cancer Staging Manual, 6th Ed. Springer, New York, NY; 2002; pp. 91–98. 13. Kalha I, Kaw M, Fukami, N, et al. The accuracy of endoscopic ultrasound for restaging esophagel carcinoma after chemoradiation therapy. Cancer 2004;101(5):940–947. 14. Chirieac LR, Swisher SG, Ajani JA, et al. Posttherapy pathologic stage predicts survival in patients with esophageal carcinoma receiving preoperative chemoradiation. Cancer 2005;103(7):1347–1355. 15. Kitajima M, Kitagawa Y. Surgical treatment of esophageal cancer—the advent of the era of individualizaiton. N Engl J Med 2002;347:1705–1709. 16. Hulscher JB, van Sandick JW, de Boer AG, et al. Extended transthoracic resection compared with limited transhiatal resection for adenocarcinoma of the esophagus. N Engl J Med 2002;347:1662–1669.

Volume 36, 2005