The American Journal of Surgery (2014) -, -–-

Assessment of the value of carcinoembryonic antigen reduction ratio as a prognosis factor in rectal cancer Chih-Sheng Huang, M.D.a,b, Jen-Kou Lin, M.D., Ph.D.a,b, Ling-Wei Wang, M.D.b,c, Wen-Yih Liang, M.D.b,d, Chun-Chi Lin, M.D.a,b, Yuan-Tzu Lan, M.D., Ph.D.a,b, Huann-Sheng Wang, M.D., Ph.D.a,b, Shung-Haur Yang, M.D., Ph.D.a,b, Jeng-Kai Jiang, M.D., Ph.D.a,b, Wei-Shone Chen, M.D., Ph.D.a,b, Tzu-Chen Lin, M.D.a,b, Shih-Ching Chang, M.D., Ph.D.a,b,* a

Division of Colon and Rectal Surgery, Department of Surgery, Taipei Veterans General Hospital, No. 201, Section 2. Shih-Pai Road, Taipei 11217, Taiwan; bSchool of Medicine, National Yang-Ming University, Taiwan; cCancer Center, Taipei Veterans General Hospital, Taiwan; dDepartment of Pathology, Taipei Veterans General Hospital, Taiwan

KEYWORDS: Rectal cancer; Preoperative chemoradiotherapy; Carcinoembryonic antigen; CEA reduction ratio

Abstract BACKGROUND: Carcinoembryonic antigen (CEA) is the most widely used tumor marker for colorectal cancer. This study aimed to investigate the role of CEA reduction ratio after preoperative chemoradiotherapy (CRT). METHODS: We enrolled 284 patients who underwent preoperative CRT followed by radical surgical resection. Patients were divided into 3 groups: serum CEA levels before CRT (pre-CRT CEA) less than 5 ng/mL (group 1); pre-CRT CEA of 5 ng/mL or more with CEA reduction ratio of 50% or more (group 2); and pre-CRT CEA of 5 ng/mL or more with CEA reduction ratio less than 50% (group 3). RESULTS: The 5-year disease-free survival (DFS) rate was not different between groups 1 (71.8%) and 2 (69.4%) but was significantly lower in group 3 (49.5%). CEA group, lymph node status after CRT (ypN) stage, and histologic type were independent prognostic factors for DFS on multivariate analysis. CONCLUSIONS: CEA reduction ratio might be an independent prognostic factor for DFS in rectal cancer patients treated with preoperative CRT and radical surgery. Ó 2014 Elsevier Inc. All rights reserved.

Serum carcinoembryonic antigen (CEA) is the most widely used tumor marker for patients with colorectal cancer. Although serum CEA is a poor marker for the C.-S.H. and J.-K.L. contributed equally to this article. There were no relevant financial relationships or any sources of support in the form of grants, equipment, or drugs. * Corresponding author. Tel.: 1886-2-2875-7544; fax: 1886-2-28757639. E-mail address: [email protected] Manuscript received June 28, 2013; revised manuscript August 28, 2013 0002-9610/$ - see front matter Ó 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.amjsurg.2013.08.054

diagnosis of primary colorectal cancer, preoperative CEA values are of prognostic significance. Preoperative CEA levels of 5.0 ng/mL or more have an adverse impact on survival that is independent of tumor stage.1 Besides, elevated preoperative CEA levels that do not normalize after surgical resection imply the presence of persistent disease and the need for further evaluation.2 Most published guidelines, including those from National Comprehensive Cancer Network and American Society of Clinical Oncology, recommend that postoperative CEA testing be performed every 3 to 6 months.2,3

2

The American Journal of Surgery, Vol -, No -, - 2014

Preoperative chemoradiotherapy (CRT) followed by total mesorectal excision (TME) is the gold standard of treatment for locally advanced rectal cancer. This procedure can produce tumor down staging, resulting in a reduced rate of postoperative local recurrence and a higher preservation rate of the anal sphincter.4–7 The advantage of administering radiotherapy before as opposed to after surgery is that the tissues are better oxygenated; this is proposed to enhance the efficacy of radiotherapy.8 Other advantages of preoperative radiotherapy include the treatment of smaller volumes, less small bowel in the field (which can fall into the pelvis after surgery), avoidance of directly irradiating the healing anastomosis (which could cause an anastomotic leak), and better anorectal function postoperatively.4 Patient compliance with treatment is also greater when radiotherapy is given before surgery.9 The 5-year survival of patients with recurrent disease is less than 7% with a mean life expectancy of 7 months. Therefore, determination of factors predicting pathologic tumor response is of considerable importance, in that it may suggest tailored treatment options and also indicate how individual prognosis should be assessed. Previous studies have suggested that clinical factors, such as the tumor volume,10,11 pretreatment CEA level,12,13 distance from the anal verge,13 and treatment interval between radiation and surgical resection,14 correlate significantly with clinical response. In addition, reduction of CEA levels after CRT may be an independent prognostic factor for disease-free survival (DFS) after preoperative CRT and surgery in rectal cancer patients.15–17 In this setting, the aim of this study was to identify pretreatment clinical factors that may predict DFS after preoperative CRT.

Evaluation

Methods Patients From January 1, 2000, to December 31, 2010, 474 patients diagnosed with rectal adenocarcinoma received preoperative CRT and surgical treatment at the Taipei Veterans General Hospital. Of these, 190 patients were excluded because of the presence of stage IV disease (n 5 114), tumor located within the upper rectum (n 5 5), transanal excision (n 5 7), or a lack of complete data regarding CEA (n 5 64). Thus, 284 patients remained eligible for the study. The computerized database at Taipei Veterans General Hospital was constructed prospectively and updated constantly. The recording variables included the demographic data of the patients; major comorbidities; family history of cancers; location, number, gross, and microscopic pathological characteristics and staging of the tumor; and status of the patient at their last follow-up visit. Tumor staging was classified using the tumor node metastasis system published by the International Union Against Cancer/American Joint Committee on Cancer, Seventh Edition.18

All patients were evaluated with staging workups, including digital rectal examination, complete blood count, liver function test, serum CEA level, colonoscopy, chest radiography, computed tomography (CT) scan of the abdomen, and pelvic MRI. In the context of an abdominal CT scan or MRI, lymph node involvement was regarded as positive when the lymph node was 5 mm or more in size in the short axis. Serum CEA levels were measured in a single laboratory using an immunoassay, with a recommended upper normal limit of 5 ng/mL. Serum samples were obtained from all patients up to 2 weeks before the initiation of preoperative CRT. Postoperative CEA was checked 1 week after surgery. Response to preoperative CRT was evaluated using a tumor regression grade (TRG) system proposed by Dworak et al.19 TRG definitions were as follows: TRG 0, no regression; TRG 1, dominant tumor mass with obvious fibrosis and/or vasculopathy; TRG 2, dominant fibrotic changes with few tumor cells or groups (easy to find); TRG 3, very few (difficult to find microscopically) tumor cells in fibrotic tissue with or without mucous substance; and TRG 4, no tumor cells, only a fibrotic mass (total regression or response). TRG 4 was defined as ‘‘complete response,’’ TRG 3 was defined as ‘‘good response,’’ and TRG 1 or 2 were defined as ‘‘poor response.’’ There was no TRG 0 in this study.

Treatment The details of CRT in the protocol were described in our previous publication.19 The prescription dose to the whole pelvis was 45 Gy in 20 fractions for a period of 4 weeks. For primary T4 disease only, a boost of 5.4 Gy in 3 fractions to the gross rectal tumors with a margin of 1.5 cm was administrated after pelvic irradiation. The median radiotherapy duration was 26 days. Oral chemotherapy agents, tegafur– uracil (UFUR; TTY Biopharm, Taipei, Taiwan), 200 mg/ m2/d, and leucovorin (Wyeth Lederle Laboratories, Taipei, Taiwan), 45 mg/d, were concurrently administered with RT. The total daily doses of both drugs were divided into 3 doses/d. The oral chemotherapy was continued after RT with a dose of 250 mg/m2/d in another 28-day cycle on days 36 to 63. The patients were monitored with an interview, physical examination, and complete blood count every week. Radical surgical resection by experienced colorectal surgeons was performed at 6 to 8 weeks after completion of RT. Pathological staging was available in these patients and was compared with the initial clinical stages. Postoperative adjuvant chemotherapy was considered for those patients with pathologic stage III disease. Of these 71 patients, 11 did not receive adjuvant chemotherapy owing to patient refusal or poor performance status. 5-Fluorouracil (5-FU)/leucovorin was administrated to 34 patients, FOLFOX (5-FU/leucovorin/oxaliplatin) to 18 patients, oral

C.-S. Huang et al.

CEA reduction ratio and rectal cancer

UFUR to 7 patients, and oral capecitabine to 1 patient. Postoperative adjuvant chemotherapy was also administrated to patients with pathologic stage II disease accompanied with other risk factors (including infiltration extent after CRT [ypT stage] 3 to 4, lymphovascular invasion, perineural invasion, and anastomosis leakage). Of these patients, oral UFUR was administrated to 41 patients, 5-FU/leucovorin to 3 patients, and FOLFOX (5-FU/leucovorin/oxaliplatin) to 6 patients.

Carcinoembryonic antigen group Serum CEA levels before CRT (pre-CRT CEA) were measured around 1 week before CRT, and serum CEA levels after CRT (post-CRT CEA) were measured within 1 week before surgery. In this study, the normal limit of serum CEA measured by RIA was set as less than 5 ng/mL and the extent of CEA reduction was evaluated by calculating the CEA ratio (defined as post-CRT CEA divided by pre-CRT CEA). In the early 1980s, the World Health Organization (WHO) developed recommendations in an attempt to standardize criteria for response assessment, and these criteria were adopted as the standard method for evaluating tumor response in image studies.20 Specifically, a 50% reduction or more in tumor size was considered as a partial response. As a result, we chose 50% reduction as an optimal cutoff value of reduction ratio in the high preCRT CEA group. According to this cutoff value, all patients were classified into 3 CEA change groups: pre-CRT CEA less than 5 ng/mL (group 1), pre-CRT CEA of 5 ng/mL or more with CEA reduction ratio of 50% or more (group 2), and pre-CRT CEA of 5 ng/mL or more with CEA reduction ratio less than 50%.

Follow-up All patients were followed up in the outpatient department every 3 months in the first 2 years, every 6 months in the third and fourth years, and annually thereafter. The follow-up examinations included chest radiography, serum CEA levels, abdominal sonography, abdominal/pelvis CT, and colonoscopy. Chest CT was arranged when a suspicious metastatic lesion was evident on a regular chest radiograph. It is our policy to perform the first follow-up colonoscopy 3 to 6 months after surgery for those patients in whom a complete colonoscopic study had not been or could not be performed before surgery. If the patient had received complete colonoscopy before surgery, the first colonoscopic surveillance was arranged 1 year after the surgery. Our follow-up strategy is then adjusted according to the presence of polyps. If the patient had a negative colonoscopy or 1 single polyp smaller than 5 mm, further colonoscopy was arranged 2 or 3 years later. If the patient had more than 2 polyps or 1 single polyp more than 5 mm, then further colonoscopic examination was arranged 1 year after polypectomy. The interval of surveillance was increased to 5 years if there were 2 consecutive negative colonoscopic surveillances.

3

Statistics The data were analyzed using the Statistical Package for Social Science (SPSS V 16.0; SPSS, Inc., Chicago, IL) statistical software. Ages of the patients were compared using independent t test. We used chi-square or Fisher exact tests to reveal associations between categorical variables. The survival curve was plotted using the Kaplan–Meier method and compared using the log-rank test. Statistical significance was defined as P less than .05.

Results Clinicopathological features of the patients Of the 284 patients, 186 (65.5%) were men. Median age was 64 years (range 27 to 93 years), and median pre- and post-CRT CEA were 3.76 ng/mL (range ,1 to 400 ng/mL) and 2.20 ng/mL (,1 to 50.1 ng/mL), respectively. Low anterior resection was performed in 230 patients (80.1%), and free resection margin (,1 mm) were found in all surgical specimens. Fifty-three patients (18.6%) achieved pathologic complete response after CRT. Local recurrence occurred in 23 patients (8.0%) and distant recurrence occurred in 78 patients (27.4%). Sites of distant metastases were the lung (n 5 51), liver (n 5 27), bone (n 5 6), brain (n 5 6), peritoneal carcinomatosis (n 5 5), and para-aortic lymph nodes (n 5 3) in descending order of frequency. The median follow-up interval was 50 months (range 4 to 150 months).

Characteristics of patients according to pre-CRT CEA Patients with high (R5.0 ng/mL) pre-CRT CEA concentrations were likely to have more advanced clinical T stage than those with normal (,5.0 ng/mL) pre-CRT CEA. Sex distribution, tumor location, histologic differentiation, pathologic stage, downstage rate, lymphovascular and perineural invasion status, tumor recurrence rate, and 5year DFS rate did not differ significantly between the normal and high pre-CRT CEA groups (Table 1).

Clinicopathologic features, local control, and DFS of patients with reference to CEA reduction ratio Using 50% reduction as the cutoff point, all patients were categorized into 3 groups according to pre- and postCRT CEA: pre-CRT CEA less than 5 ng/mL (group 1), preCRT CEA of 5 ng/mL or more with CEA reduction ratio of 50% or more (group 2), and pre-CRT CEA of 5 ng/mL or more with CEA reduction ratio less than 50% (group 3). When we compared clinicopathologic characteristics among the 3 groups, we found that the mean pre-CRT CEA was highest in group 2 and mean post-CRT CEA was

The American Journal of Surgery, Vol -, No -, - 2014

4 Table 1

Clinical characteristics according to pre-CRT CEA

Patient number Sex Male Female Age, mean 6 SD Tumor location Mid rectum Low rectum Surgery type LAR APR Pre-CRT CEA Post-CRT CEA cT cT2 cT3–4 cN cN0 cN1–2 ypT ypT0–2 ypT3–4 ypN ypN0 ypN1–2 Down stage Differentiation Well/moderate Poor LVi PNi Complete response Recurrence Local Distant

Normal pre-CRT CEA

High pre-CRT CEA

176

108

110 (62.5%) 66 (37.5%) 62.3 6 13.0

76 (70.4%) 32 (29.6%) 64.4 6 12.1

.110

84 (47.7%) 92 (52.3%)

60 (55.6%) 48 (44.4%)

.123

142 (80.7%) 34 (19.3%) 2.8 6 .9 2.4 6 1.6

88 (81.5%) 20 (18.5%) 25.3 6 47.6 5.4 6 7.1

.499

25 (14.2%) 151 (85.8%)

7 (6.5%) 101 (93.5%)

.033

44 (25.0%) 132 (75.0%)

23 (21.3%) 85 (78.7%)

.286

85 (48.3%) 91 (51.7%)

41 (38.0%) 67 (62.0%)

.057

133 (75.6%) 43 (24.4%) 132 (75.0%)

80 (74.1%) 28 (25.9%) 83 (76.9%)

.442

168 (95.5%) 8 (4.5%) 12 (6.8%) 5 (2.8%) 32 (18.2%) 47 (26.7%) 12 (6.8%) 45(25.6%)

106 (98.1%) 2 (1.9%) 4 (3.7%) 3 (2.8%) 21 (19.4%) 38 (35.2%) 9 (8.3%) 35 (32.4%)

P

.768

,.001 ,.001

.419 .197 .203 .641 .454 .084 .400 .134

APR 5 abdominal perineal resection; CEA 5 carcinoembryonic antigen; CRT 5 chemoradiotherapy; LAR 5 low anterior resection; LVi 5 lymphovascular invasion; PNi 5 perineural invasion.

greatest in group 3. Additionally, group 3 was composed predominantly of male patients. Although pathological features did not differ significantly between group 2 and group 3 patients, the group 3 patients had a higher incidence of tumor recurrence (Table 2). The 5-year DFS rate was similar in group 1 (71.8%) and group 2 (69.4%) but significantly lower in group 3 (49.5%) (P 5 .025; Fig. 1). Univariate analysis showed that 7 clinicopathologic parameters (post-CRT CEA, CEA group, infiltration extent after CRT [ypT stage], lymph node status after CRT [ypN stage], TRG, tumor differentiation, and perineural invasion status) were predictive of DFS (Table 3). Multivariate analysis showed that ypN stage, tumor differentiation, and CEA groups were independent, statistically significant prognostic factors for DFS (Table 4).

Table 2

Clinical characteristics between CEA groups 2 and 3

Patient number Sex Male Female Age, mean 6 SD Tumor location Mid rectum Low rectum Surgery type LAR APR Pre-CRT CEA Post-CRT CEA cT cT2 cT3–4 cN cN0 cN1–2 ypT ypT0–2 ypT3–4 ypN ypN0 ypN1–2 Down stage Differentiation Well/moderate Poor LVi PNi Complete response Recurrence Local Distant

P

Group 2

Group 3

77

31

49 (63.6%) 28 (36.4%) 64.0 6 13.0

27 (87.1%) 4 (12.9%) 65.3 6 9.8

.012

41 (53.2%) 36 (46.8%)

19 (61.3%) 12 (38.7%)

.293

63 (81.8%) 14 (18.2%) 30.4 6 55.0 3.8 6 3.4

25 (80.6%) 6 (19.4%) 12.76 14.1 9.7 6 11.2

.542

4 (5.2%) 73 (94.8%)

3 (9.7%) 28 (90.3%)

.321

16 (20.8%) 61 (79.2%)

7 (22.6%) 24 (77.4%)

.512

31 (40.3%) 46 (59.7%)

10 (32.3%) 21 (67.7%)

.291

58 (75.3%) 19 (24.7%) 59 (76.6%)

22 (71.0%) 9 (29.0%) 24 (77.4%)

.405

76 (95.5%) 1 (4.5%) 1 (1.3%) 1 (1.3%) 18 (23.4%) 22 (28.6%) 9 (11.7%) 19 (24.7%)

30 (96.8%) 1 (3.2%) 3 (9.7%) 2 (6.5%) 3 (9.7%) 16 (51.6%) 2 (6.5%) 14 (45.2%)

.494

.109

.041 ,.001

.572

.070 .197 .083 .021 .335 .033

APR 5 abdominal perineal resection; CEA 5 carcinoembryonic antigen; CRT 5 chemoradiotherapy; LAR 5 low anterior resection; LVi 5 lymphovascular invasion; PNi 5 perineural invasion.

Comments Serum CEA is a widely accepted tumor marker, especially for colorectal cancer, and its measurement is standardized, inexpensive, and available. Elevated preoperative CEA level has been considered as an independent prognostic factor for DFS in colorectal carcinoma.12,13 Furthermore, postoperative CEA monitoring has been valuable for early detection of recurrence after curative surgery and for assessing responses to chemotherapy in patients with metastatic colorectal cancer. Preoperative CRT followed by TME had become the gold standard of treatment for locally advanced rectal cancers.7 Recent studies have investigated the utility of pre- and post-CRT CEA concentrations as predictors of response or as prognostic factors in rectal cancer patients who receive

C.-S. Huang et al. Table 3

CEA reduction ratio and rectal cancer

5

Univariate analysis of prognostic factors for DFS

Variable Sex Male Female Age ,70 R70 Tumor location Mid rectum Low rectum Surgery type LAR APR Pre-CRT CEA R5 ,5 Post-CRT CEA R5 ,5 CEA group Group 1 Group 2 Group 3 ypT ypT0–2 ypT3–4 ypN ypN0 ypN1–2 Regression grade Complete Good/moderate Poor Differentiation Well/moderate Poor LVi Yes No PNi Yes No

No. of patients

5-y DFS rate (%)

186 98

70.5 64.9

.329

183 101

66.6 70.3

.208

144 140

67.6 69.7

.862

230 54

69.2 66.6

.482

108 176

63.5 71.8

.149

36 248

52.0 71.8

.029

176 77 31

71.8 69.4 49.5

.025

126 158

80.9 57.9

,.001

213 71

75.7 47.1

,.001

53 196 35

87.1 68.4 30.3

.004

207 11

71.0 18.2

,.001

8 268

37.5 69.7

.177

8 276

37.5 69.5

,.001

P

APR 5 abdominal perineal resection; CEA 5 carcinoembryonic antigen; CRT 5 chemoradiotherapy; DFS 5 disease-free survival; LAR 5 low anterior resection; LVi 5 lymphovascular invasion; PNi 5 perineural invasion.

preoperative CRT. Some studies reported that pre-CRT CEA concentration more than 5 ng/mL is associated with poor tumor response to preoperative CRT,12,13,21 suggesting that this cutoff may serve as a valuable clinical predictor of pathologic tumor response. Perez et al22 reported that post-CRT CEA less than 5 ng/mL was associated with increased overall survival and DFS. In our study, the 5-year DFS did not differ significantly between the normal (,5 ng/mL) and high (R5 ng/mL) pre-CRT CEA groups (71.8% vs 63.5%, P 5 .149), but the 5-year DFS rate was significantly lower

Figure 1 The 5-year DFS rate relative to CEA groupings categorized with respect to pre- and post-CRT CEA. Group 1: pre-CRT CEA less than 5 ng/mL; group 2: pre-CRT CEA of 5 ng/mL or more with CEA reduction ratio of 50% or more; group 3: pre-CRT CEA of 5 ng/mL or more with CEA reduction ratio less than 50%.

in patients with high (R5 ng/mL) than with normal (,5 ng/mL) post-CRT CEA (52.0% vs 71.8%, P 5 .029). According to our results, we suggest that the post-CRT CEA is more valuable than the pre-CRT CEA as a prognostic factor for DFS in patients with rectal cancers. We also found that patients with high levels of post-CRT CEA had a higher risk of liver metastasis, indicating that the prognostic value of post-CRT CEA is not restricted to primary rectal cancer alone. Of the 36 patients with high post-CRT CEA, 9 developed liver metastases (25.0%). In contrast, of the 248 patients with normal post-CRT CEA, 18 had liver metastases (7.3%). Therefore, in patients with high levels of post-CRT CEA, aggressive screening for liver metastasis should be mandatory. Other risks of liver metastasis included advanced ypT and ypN stage. We also found that histologically poor differentiation, perineural invasion, and advanced ypT and ypN stage are risk factors of lung metastasis (Table 5).

Table 4 Multivariate analysis of prognostic factors for disease-free survival Variable CEA group Group 1 Group 2 Group 3 ypN ypN0 ypN1–2 Differentiation Well/moderate Poor

Hazard ratio

95% CI

P

1 2.190 9.951

.999–4.768 2.927–33.830

.050 ,.001

1 2.823

1.230–6.481

1 8.572

2.181–33.689

.014

.002

CEA 5 carcinoembryonic antigen; CI 5 confidence interval.

The American Journal of Surgery, Vol -, No -, - 2014

6 Table 5 Variable

Risk factors for tumor recurrence No. of patients

Pre-CRT CEA ,5 176 R5 108 P Post-CRT CEA ,5 248 R5 36 P ypT ypT0–2 126 ypT3–4 158 P ypN ypN0 213 ypN1–2 71 P Differentiation Well/ 274 moderate Poor 10 P LVi No 268 Yes 16 P PNi No 276 Yes 8 P

Lung metastasis Liver metastasis (n 5 51) (n 5 27) 31 (17.6%) 20 (18.5%) .483

13 (7.4%) 14 (13.0%) .090

44 (17.7%) 7 (19.4%) .478

18 (7.3%) 9 (25.0%) .003

15 (11.9%) 36 (22.8%) .012

5 (4.0%) 22 (19.3%) .003

32 (15.0%) 19 (26.8%) .022

13 (6.1%) 14 (19.7%) .001

45 (16.4%)

24 (8.8%)

6 (60.0%) .003

3 (30.0%) .059

48 (17.9%) 3 (18.8 %) .573

24 (9.0%) 3 (18.8%) .186

46 (16.7%) 5 (62.5%) .006

25 (9.1%) 2 (25.0%) .171

CEA 5 carcinoembryonic antigen; CRT 5 chemoradiotherapy; LVi 5 lymphovascular invasion; PNi 5 perineural invasion.

Kim et al16 reported that reduction ratio of pre- to postCRT CEA may be an independent prognostic factor for DFS after preoperative CRT and surgery. They classified rectal cancer patients into 3 groups according to pre- and post-CRT CEA, though with different cutoff values (.6 ng/mL or %6 ng/mL). Patients with pre-CRT CEA more than 6 ng/mL were further divided based on whether post-CRT CEA was at least more than or equal to 70%, which is lower than pre-CRT CEA (CEA reduction ratio R 70%). The 5-year DFS rate in the above study was significantly lower in patients with pre-CRT CEA more than 6 ng/mL and CEA reduction ratio of 70% or more in comparison with the other groups. Although cutoff points for the groups based on CEA concentration differed, our previous study reported similar results.17 Using ROC curves, Kim et al determined the cutoff point of this predictive ratio (70%). However, the predictive ratio will vary in different cohorts according to the maximum Youden index (sensitivity 1 specificity 2 1) from the ROC curves. For example, the maximum Youden index corresponded to a 59% reduction in our study. To determine a more objective CEA reduction ratio model, we modified the WHO response criteria, which is a simple, globally accepted,

and validated evaluation tool. In this setting, we found that there was no significant difference in the 5-year DFS rate between groups 1 (71.8%) and 2 (69.4%) but that 5year DFS was significantly lower in group 3 (49.5%). In addition, our study demonstrated that CEA groups were independent prognostic factors for DFS on multivariate analysis. We suggested that significant CEA reduction plays a role in supporting significant tumor regression and a good clinical response after CRT for patients with high preCRT CEA. Further large multicenter studies are required to confirm the feasibility of a modified WHO response criteria based on the CEA reduction ratio model. In addition to CEA groups, we found that ypN stage and histologic type were independent prognostic factors for DFS on multivariate analysis. Regional lymph node involvement is one of the strongest predictors of outcome after surgical resection of rectal cancers, second only to distant metastasis. Nodal spread is an indication for adjuvant therapy for rectal cancer in most guidelines.3 Poorly differentiated tumors are high-grade carcinomas that generally behave in a biologically aggressive fashion. Most studies suggest that ypN is an independent prognostic factor for DFS in patients with rectal cancers,21,23 but the significance of histological differentiation remained controversial.15,16 The major limitation when performing comparative studies is the variability of adjuvant chemotherapy regimens. Considering the potential effects of different chemotherapy regimens, we analyzed several groups (5-FU/ leucovorin, FOLFOX, oral UFUR and oral capecitabine) and found that there was no DFS difference between patients treated with these 4 regimens. Other limitations of this study included its retrospective design, relatively small cohorts, and short follow-up period in some patients. In the future, large prospective studies in which CEA kinetics are analyzed through routine measurement of CEA levels at follow-up are required.

Conclusions In conclusion, CEA grouping according to the reduction ratio of pre- to post-CRT CEA might be an independent prognostic factor for DFS in rectal cancer patients with high (.5 ng/mL) pre-CRT CEA treated with preoperative CRT and radical surgery. A modified WHO response criteria incorporating the CEA reduction ratio model may be a feasible tool for use in prediction of the pathologic tumor response after preoperative CRT. In addition, an aggressive surveillance protocol for liver metastasis should be used for those patients with high post-CRT CEA.

References 1. Thirunavukarasu P, Sukumar S, Sathaiah M, et al. C-stage in colon cancer: implications of carcinoembryonic antigen biomarker in staging, prognosis, and management. J Natl Cancer Inst 2011;103:689–97.

C.-S. Huang et al.

CEA reduction ratio and rectal cancer

2. Locker GY, Hamilton S, Harris J, et al. ASCO 2006 update of recommendations for the use of tumor markers in gastrointestinal cancer. J Clin Oncol 2006;24:5313–27. 3. National Comprehensive Cancer Network (NCCN). Rectal Cancer Clinical Practice Guidelines in Oncology. Available at: http://www. nccn.org/index.asp. Accessed November 26, 2012. 4. Sauer R, Becker H, Hohenberger W, et al. Preoperative versus postoperative chemoradiotherapy for rectal cancer. N Engl J Med 2004;351: 1731–40. 5. Glimelius B. Radiotherapy in rectal cancer. Br Med Bull 2002;64: 141–57. 6. Folkesson J, Birgisson H, Pahlman L, et al. Swedish Rectal Cancer Trial: long lasting benefits from radiotherapy on survival and local recurrence rate. J Clin Oncol 2005;23:5644–50. 7. McCarthy K, Pearson K, Fulton R, et al. Pre-operative chemoradiation for non-metastatic locally advanced rectal cancer. Cochr Database Syst Rev 2012;12:CD008368. 8. Gray R, Hills R, Stowe R, et al. Adjuvant radiotherapy for rectal cancer: a systematic overview of 8,507 patients from 22 randomised trials. Lancet 2001;358:1291–304. 9. Bosset JF, Collette L, Calais G, et al. Chemotherapy with preoperative radiotherapy in rectal cancer. N Engl J Med 2006;355:1114–23. 10. Yeo SG, Kim DY, Park JW, et al. Tumor volume reduction rate after preoperative chemoradiotherapy as a prognostic factor in locally advanced rectal cancer. Int J Radiat Oncol Biol Phys 2012;82:e193–9. 11. Yeo SG, Kim DY, Kim TH, et al. Tumor volume reduction rate measured by magnetic resonance volumetry correlated with pathologic tumor response of preoperative chemoradiotherapy for rectal cancer. Int J Radiat Oncol Biol Phys 2010;78:164–71. 12. Yoon SM, Kim DY, Kim TH, et al. Clinical parameters predicting pathologic tumor response after preoperative chemoradiotherapy for rectal cancer. Int J Radiat Oncol Biol Phys 2007;69:1167–72. 13. Das P, Skibber JM, Rodriguez-Bigas MA, et al. Predictors of tumor response and downstaging in patients who receive preoperative chemoradiation for rectal cancer. Cancer 2007;109:1750–5.

7 14. Kalady MF, de Campos-Lobato LF, Stocchi L, et al. Predictive factors of pathologic complete response after neoadjuvant chemoradiation for rectal cancer. Ann Surg 2009;250:582–9. 15. Jang NY, Kang SB, Kim DW, et al. The role of carcinoembryonic antigen after neoadjuvant chemoradiotherapy in patients with rectal cancer. Dis Colon Rectum 2011;54:245–52. 16. Kim CW, Yu CS, Yang SS, et al. Clinical significance of pre- to postchemoradiotherapy s-CEA reduction ratio in rectal cancer patients treated with preoperative chemoradiotherapy and curative resection. Ann Surg Oncol 2011;18:3271–7. 17. Yang KL, Yang SH, Liang WY, et al. Carcinoembryonic antigen (CEA) level, CEA ratio, and treatment outcome of rectal cancer patients receiving pre-operative chemoradiation and surgery. Radiat Oncol 2013;8:43. 18. Edge SB, Byrd DR, Compton CC, et al. AJCC Cancer Staging Manual. 7th ed. Springer; 2010. 19. Dworak O, Keilholz L, Hoffmann A. Pathological features of rectal cancer after preoperative radiochemotherapy. Int J Colorectal Dis 1997;12:19–23. 20. Miller AB, Hoogstraten B, Staquet M, et al. Reporting results of cancer treatment. Cancer 1981;47:207–14. 21. Kim TH, Chang HJ, Kim DY, et al. Pathologic nodal classification is the most discriminating prognostic factor for disease-free survival in rectal cancer patients treated with preoperative chemoradiotherapy and curative resection. Int J Radiat Oncol Biol Phys 2010;77: 1158–65. 22. Perez RO, Sao Juliao GP, Habr-Gama A, et al. The role of carcinoembriogenic antigen in predicting response and survival to neoadjuvant chemoradiotherapy for distal rectal cancer. Dis Colon Rectum 2009; 52:1137–43. 23. Kim NK, Baik SH, Seong JS, et al. Oncologic outcomes after neoadjuvant chemoradiation followed by curative resection with tumorspecific mesorectal excision for fixed locally advanced rectal cancer: impact of postirradiated pathologic downstaging on local recurrence and survival. Ann Surg 2006;244:1024–30.