Journal of Surgical Oncology 2014;109:580–585
Celecoxib Plus Chemoradiotherapy for Locally Advanced Rectal Cancer: A Phase II TCOG study LING-WEI WANG, MD,1* CHIN-FU HSIAO, PhD,2 WILLIAM TZU-LIANG CHEN, MD,3 HAO-HSIEN LEE, MD,4 TZU-CHEN LIN, MD,5 HUNG-CHANG CHEN, MD,6 HONG-HWA CHEN, MD,7 CHUN-RU CHIEN, MD,8 TZE-YI LIN, MD, DR der MED,9 AND TSANG-WU LIU, MD10 1
Cancer Center, Taipei Veterans General Hospital & School of Medicine, National Yang Ming University, Taipei, Taiwan 2 Division of Biostatistics and Bioinformatics, National Health Research Institute, Miaoli, Taiwan 3 Division of Colorectal Surgery, China Medical University Hospital, Taichung, Taiwan 4 Division of Colorectal Surgery, Chi Mei Hospital-Liuying, Tainan, Taiwan 5 Division of Colorectal Surgery, Taipei Veterans General Hospital, Taipei, Taiwan 6 Division of Colorectal Surgery, Changhua Christian Hospital, Changhua, Taiwan 7 Division of Colorectal Surgery, Chang Gung Memorial Hospital, Kaoshiung, Taiwan 8 Division of Radiation Oncology, China Medical University Hospital, Taichung, Taiwan 9 Department of Pathology, China Medical University Hospital, Taichung, Taiwan 10 Department of Research Planning and Development, National Health Research Institute, Miaoli, Taiwan
Background: To report the results of a phase II trial combining celecoxib and preoperative chemoradiotherapy (CRT) for locally advanced rectal cancer. Patients and Methods: Patients with clinical stage II or III rectal cancer were treated with radiotherapy of 44 Gy in 22 fractions. Concurrent chemotherapy consisted of oral tegafur‐uracil and folinate on days 1–30 and 38–65. Celecoxib (400 mg/day) given from days 1 to 65. Surgery was done on day 70. The expression of cyclooxygenase 2 (COX‐2) in tumor tissues was evaluated microscopically as a prognostic factor. Results: From 2008 to 2011, 53 patients completed CRTþ celecoxib therapy and 47 received radical surgery. Grade 3 diarrhea developed in 5 (9%). Grade 4 anemia was seen in 2 (4%). Pathological complete response (pCR) was seen in 6 (13%). T or N downstaging found in 38 (81%). Sphincter preservation was achieved in 77% of low‐positioned tumors. Patients with tumors expressing high‐level COX‐2 after CRT þ celecoxib treatment had inferior pelvic control (P ¼ 0.01), disease‐free survival (P ¼ 0.04), and overall survival (P ¼ 0.03) than those with low‐level expression. Conclusions: Celecoxib can be safely combined with preoperative CRT for rectal cancer. More intensified adjuvant therapy may be considered for tumors expressing high‐level COX‐2 after CRT and surgery.
J. Surg. Oncol. 2014;109:580–585. ß 2013 Wiley Periodicals, Inc.
KEY WORDS: rectal cancer; COX‐2; celecoxib; chemoradiotherapy
INTRODUCTION Rectal cancer is a prevalent disease, both in Taiwan and worldwide. In treating rectal cancer, neoadjuvant radiotherapy (RT) with or without chemotherapy is often effective in cases of locally advanced disease for improving resectability [1–3], sphincter preservation [4,5], and improved survival [6]. A previous meta‐analysis showed that pre‐ operative radiation combined with chemotherapy may be more effective in the downstaging of locally advanced rectal cancer than radiation therapy alone [7]. Intravenous 5‐fluorouracil (5‐FU) plus leucovorin is the most commonly used chemotherapy regimen. Oral tegafur is a prodrug of 5‐FU. Uracil competes with 5‐FU for the enzyme dihydropyrimidine dehydrogenase, which converts 5‐FU to an inactive metabolite. Studies have demonstrated the feasibility and convenience of oral tegafur‐uracil and radiation combination therapy in the pre‐operative treatment of rectal cancer [8–10]. Cyclooxygenase‐2 (COX‐2) contributes to tumor growth through the expression of prostaglandin E2 and vascular endothelial growth factor [11]. The expression of COX‐2 has also been shown to be a negative prognostic factor for colorectal cancer, and the overexpression of COX‐2 was shown to be predictor of poor response to pre‐operative chemoradiation therapy [12]. The long‐term use of a COX‐2 inhibitor has been found to reduce rectal polyp formation in patients with familial adenosis polyposis [13]. Celecoxib inhibits COX‐2 activity, and
ß 2013 Wiley Periodicals, Inc.
demonstrates effective radiosensitization in various types of tumor cells expressing COX‐2 [14]. However, increased cardiovascular risk of patients taking COX‐2 inhibitors or celecoxib for colorectal adenoma prevention was reported [15–17]. Phase II clinical studies have also shown that celecoxib and chemoradiation (CRT) combination therapy for rectal cancer is tolerable in most patients [18,19]. However, one trial reported occurrence of skin rash in 49% of the participants [20]. In these clinical trials, the most common celecoxib dosage was 400 mg twice daily [18,20,21], which is higher than that required for mild to moderate pain, such as that induced by moderate doses (50 Gy) of radiation. Lower doses of celecoxib (e.g., 400 mg/day) during and after radiotherapy may be sufficient for radiosensitization and increase tolerability. Realizing the potential
Grant sponsor: National Health Research Institute of Taiwan, R.O.C.. *Correspondence to: Ling‐Wei Wang, MD, Cancer Center, Taipei Veterans General Hospital, No. 201, Sec. 2, Shih‐Pai Road, Taipei 112, Taiwan. Fax: þ886‐2‐28749425. E‐mail: [email protected] Received 30 August 2013; Accepted 30 November 2013 DOI 10.1002/jso.23538 Published online 24 December 2013 in Wiley Online Library (wileyonlinelibrary.com).
A Celecoxib Trial for Rectal Cancer benefits of such a regimen, we proposed a clinical trial of low‐dose celecoxib treatment combined with chemoradiotherapy for rectal cancer. In 2008, we initiated a multi‐center phase II trial sponsored by the Taiwan Cooperative Oncology Group (TCOG) of the National Health Research Institute, Taiwan for the study of celecoxib and CRT for the treatment of locally advanced rectal cancer. The primary objectives of the trial were to evaluate the efficacy of combining preoperative tegafur‐ uracil, folinate, radiation, and low‐dose celecoxib treatments for locally advanced rectal cancer, and to determine whether any toxicity is associated with this regimen. The prognostic significance of COX‐2 expression in pathological specimens before and after combination therapy was also examined.
PATIENTS AND METHODS Eligibility Our study was approved by the institutional review boards of all the seven participating hospitals in the TCOG, and was registered on www. clinicaltrials.gov as trial NCT00931203. Patients meeting the following criteria were included in our study: (1) a diagnosis of adenocarcinoma based on the pathology results of biopsy specimens; (2) clinical stage T2N1‐2M0 or T3‐4N0‐2M0, according to the American Joint Committee on Cancer (AJCC) 2002 staging system; (3) resectable or potentially resectable adenocarcinoma of the rectum; (4) underwent magnetic resonance imaging (MRI) of the pelvis, ultrasonography of liver, and chest X ray (CXR); (5) bi‐dimensionally measurable tumor based on MRI using a pelvic array coil and an intrarectal tube; (6) superior margin of the tumor below the L5‐S1 spine junction; (7) between the ages of 18 and 80 years; (8) a functional status 3.5 109/L, a neutrophil count > 1.5 109/L, a platelet count > 100 109/L, a serum bilirubin level < 125% of the reference‐range maximum (ULN), an AST/ ALT < 300% ULN, and a serum creatinine level 50%. Based on the product of the intensity and extent of staining scores, the
582
Wang et al.
sections were divided into low (total score ¼ 0–4) and high (total score ¼ 6 or 9) staining groups.
Statistical Analysis The Simon minimax two‐stage design (a ¼ 0.05, power ¼ 80%, two‐ sided test) was used for the statistical analysis of our study results. We considered the addition of celecoxib to chemoradiation as non‐effective for a pCR rate of 15% or lower, or as in need of further study for a pCR rate of 30% or higher. In the first stage of the trial, 23 patients were enrolled. If more than 3 patients had pCR, another 25 patients were enrolled. The test regimen was considered effective if 11 patients had pCR in the second stage of the trial. Overall survival was calculated from the date of enrollment in the study to the date of death. Disease‐free survival was calculated from the date of surgery to the date of treatment failure. The disease‐free survival and overall survival rates were calculated using the Kaplan–Meier method. The log‐rank test was used to compare the survival curves, and the Wilcoxon signed‐rank test was used to compare the total scores of COX‐2 staining of specimens before and after radiotherapy. A two‐sided exact test was used to evaluate the association between variables. The P‐value was considered significant when it was 6
Journal of Surgical Oncology
53 57 [26–76] 40 (75) 13 (25) 48 (91) 5 (9) 48 (91) 5 (9) 27 (51) 26 (49) 2 (4) 7 (13) 37 (70) 1 (2) 6 (12) 7 (13) 38 (72) 3 (6) 5 (9) 5 [0–12] 37 (70) 16 (30)
Most of the patients tolerated neoadjuvant treatment well. Two patients had RT interruptions, and 4 had concurrent chemotherapy interruptions because of toxicity. Four patients required modification of the celecoxib dosage, and celecoxib was withheld temporarily from 2 other patients because of adverse effects. Cases of acute toxicity are shown in Table II, most of which were grades 1–2. These toxicities included fatigue, radiation‐related rash, diarrhea, and anemia. The most common grade‐3 toxicity was diarrhea (five patients). Three participants (6%) had grade‐4 toxicities, including two patients with anemia and 1 with reversible hypercreatinemia. No grade‐5 toxicities occurred. Five patients refused surgery. One patient experienced disease progression after chemoradiation, and withdrew from the study. A total of 47 patients received TME. The median interval between RT and surgery was 44 days (range, 30–58 days). Thirty‐nine patients received low anterior resection (LAR), whereas 8 underwent abdominoperineal resection (APR). Sphincter preservation occurred in 20 of 26 patients (77%) with low‐positioned tumors (6 cm from anal verge before treatment. However, only one patient had both positive circumferential and proximal margins. The median distal margin was 2 cm, and all the distal margins were negative. The tumor response and down‐staging results are shown in Table III. Thirty‐five patients (75%) exhibited good tumor regression (TRG 0 þ 1), including six patients (13%) experienced pCR. After surgery, 42 patients received adjuvant chemotherapy, and 40 of them completed all six cycles. The median follow‐up period was 33 TABLE II. Acute Toxicities Before Surgery (N ¼ 53) Item Allergy Gastrointestinal Anorexia Diarrhea Oral mucositis Nausea Ulcer, GI Vomiting Hematology Hemoglobin Leukocytes Platelets Neutrophils Cardiovascular Hypertension Hypotension Constitutional Fatigue Weight loss Dermatology Rash (radiation‐related) Rash (hand‐foot reaction) Laboratory (biochemical) Bilirubin‐T Creatinine SGOT SGPT Hemorrhage, GI Pain Anal pain Dysuria Other pain Genitourinary Urinary frequency
Grade 1 (%)
Grade 2 (%)
Grade 3 (%)
Grade 4 (%)
2 (4)
2 (4)
1 (2)
0 (0)
17 23 3 8 0 1
(32) (43) (6) (15) (0) (2)
3 9 0 0 1 0
(6) (17) (0) (0) (2) (0)
0 5 0 0 0 0
(0) (9) (0) (0) (0) (0)
0 0 0 0 0 0
(0) (0) (0) (0) (0) (0)
28 15 17 4
(53) (28) (32) (8)
3 4 2 1
(6) (8) (4) (2)
0 0 0 0
(0) (0) (0) (0)
2 0 0 0
(4) (0) (0) (0)
4 (8) 1 (2)
0 (0) 0 (0)
0 (0) 0 (0)
0 (0) 0 (0)
30 (57) 3 (6)
5 (9) 0 (0)
0 (0) 0 (0)
0 (0) 0 (0)
22 (42) 0 (0)
13 (25) 0 (0)
1 (2) 0 (0)
0 (0) 0 (0)
0 0 0 0 1
0 1 0 0 0
12 1 6 7 8
(23) (2) (11) (13) (15)
0 0 0 1 0
(0) (0) (0) (2) (0)
(0) (0) (0) (0) (2)
(0) (2) (0) (0) (0)
6 (11) 2 (4) 5 (9)
8 (15) 0 (0) 3 (6)
0 (0) 0 (0) 1 (2)
0 (0) 0 (0) 0 (0)
6 (11)
0 (0)
0 (0)
0 (0)
A Celecoxib Trial for Rectal Cancer
583
TABLE III. Tumor Regression and Downstaging in 47 Patients
TRG 0 (¼pCR) 1 2 3 Pathological staging (yp) T0N0 T1–2N0 T3–4N0 T2–4N1 T1–3N2 T4N0M1a T downstaging N downstaging
N
%
6 29 9 3
13 62 19 6
6 13 16 9 2 1 22 31
13 28 34 19 4 2 47 66
TRG, tumor regression grading; pCR, pathological complete response. Peritoneal seeding found during operation.
a
Fig. 2. COX‐2 expressions scores before and after treatment (N ¼ 29). (range, 2.7–54) months. Thirteen cases (27%) developed distant metastases. The most common metastatic sites after surgery were the lung (N ¼ 9) and the liver (N ¼ 5). The 3‐year pelvic control, disease‐ free and overall survival rates were 89%, 63%, and 88%, respectively. The disease‐free survival curve is shown in Figure 1. Immunohistochemistry for COX‐2 expression was performed in 39 available biopsy specimens and 38 surgical specimens (excluding those from the six patients that had pCR and three with very few residual tumor cells). For the pretreatment specimens, 77% (30/39) expressed high‐ level COX‐2 expression. The paired comparison of the biopsy and surgical specimens revealed a trend toward a decreased total score for COX‐2 staining with celecoxib treatment (N ¼ 29, P < 0.001, Fig. 2). There was no association between the pre‐RT COX‐2 staining scores and the TRG of tumors after chemoradiotherapy (P ¼ 1.0). The post‐RT COX‐2 staining scores were significantly associated with pelvic recurrence (P ¼ 0.01), disease‐free survival (P ¼ 0.04, see Fig. 3), and
Fig. 1.
Disease‐free survival of all patients receiving surgery (N ¼ 47).
Journal of Surgical Oncology
overall survival (P ¼ 0.03), but were not associated with the T stage (P ¼ 0.34) or the TRG (P ¼ 1.0).
DISCUSSION The expression of COX‐2 plays important role in the carcinogenesis and the invasiveness of rectal cancer. In our clinical trial, we observed the feasibility and efficacy of combining the COX‐2 inhibitor celecoxib with pre‐operative chemoradiotherapy for the treatment of locally advanced rectal cancer, and demonstrated the role of COX‐2 expression in surgical specimens as a prognostic factor in rectal cancer. In our trial, the CRT and celecoxib treatment were well tolerated. Most (98%) patients’ pain was grade 1 or 2, which is consistent with the
Fig. 3. Disease‐free survival by post‐CRT COX‐2 scoring high (N ¼ 12) and low (N ¼ 26) groups.
584
Wang et al.
anti‐inflammatory and analgesic effects of celecoxib. A smaller fraction (10%) of patients experienced allergic reactions manifested as skin rashes compared with the study by Jakobsen et al. [20]. The severity of cardiovascular adverse effects were all grade 1 and its incidence was low (Table II). These results may be related to the low dosage and short duration of celecoxib treatment. All the grade‐4 toxicities were managed conservatively and recovered. Overall, a safe toxicity profile was observed for celecoxib. With regard to efficacy, though our pCR rate is just 13%, a high rate of the down‐staging of disease and good tumor regression (TRG 0 þ 1) were still observed. The latter accounted for approximately 75% (35/47) of the patients that received surgery (Table III). Only one case of a positive circumferential margin and a high pelvic control rate occurred, which may reflect the combined effects of chemoradiotherapy, celecoxib treatment, and effective surgical intervention. A high sphincter preservation rate (77%) was also achieved among low‐positioned tumors. Pathological complete response (pCR) rate is a common measure for efficacy of radiosensitizers in the clinic and the primary end point of this study. One randomized phase II trial by Debucquoy’s et al. [18] revealed that rectal cancer patients treated with CRT (45 Gy) and celecoxib (800 mg/day for 10–12 weeks) tended to have a better pCR rate (39%) compared with that of patients treated with CRT and placebo (29%). However, the difference in pCR rates for the two groups was not statistically significant. A phase I trial by Malik et al. [21] using celecoxib (800 mg/day for 5.6 weeks) and CRT (50.4 Gy) obtained a pCR rate of 15%. A recent phase II study by Unger et al. [19] combining celecoxib (400 mg/day for 5.6–6 weeks) with CRT (50.4 or 54 Gy) achieved a pCR rate of 25%. All the above phase I/II studies were done with small patient number (
Celecoxib Plus Chemoradiotherapy for Locally Advanced Rectal Cancer: A Phase II TCOG study LING-WEI WANG, MD,1* CHIN-FU HSIAO, PhD,2 WILLIAM TZU-LIANG CHEN, MD,3 HAO-HSIEN LEE, MD,4 TZU-CHEN LIN, MD,5 HUNG-CHANG CHEN, MD,6 HONG-HWA CHEN, MD,7 CHUN-RU CHIEN, MD,8 TZE-YI LIN, MD, DR der MED,9 AND TSANG-WU LIU, MD10 1
Cancer Center, Taipei Veterans General Hospital & School of Medicine, National Yang Ming University, Taipei, Taiwan 2 Division of Biostatistics and Bioinformatics, National Health Research Institute, Miaoli, Taiwan 3 Division of Colorectal Surgery, China Medical University Hospital, Taichung, Taiwan 4 Division of Colorectal Surgery, Chi Mei Hospital-Liuying, Tainan, Taiwan 5 Division of Colorectal Surgery, Taipei Veterans General Hospital, Taipei, Taiwan 6 Division of Colorectal Surgery, Changhua Christian Hospital, Changhua, Taiwan 7 Division of Colorectal Surgery, Chang Gung Memorial Hospital, Kaoshiung, Taiwan 8 Division of Radiation Oncology, China Medical University Hospital, Taichung, Taiwan 9 Department of Pathology, China Medical University Hospital, Taichung, Taiwan 10 Department of Research Planning and Development, National Health Research Institute, Miaoli, Taiwan
Background: To report the results of a phase II trial combining celecoxib and preoperative chemoradiotherapy (CRT) for locally advanced rectal cancer. Patients and Methods: Patients with clinical stage II or III rectal cancer were treated with radiotherapy of 44 Gy in 22 fractions. Concurrent chemotherapy consisted of oral tegafur‐uracil and folinate on days 1–30 and 38–65. Celecoxib (400 mg/day) given from days 1 to 65. Surgery was done on day 70. The expression of cyclooxygenase 2 (COX‐2) in tumor tissues was evaluated microscopically as a prognostic factor. Results: From 2008 to 2011, 53 patients completed CRTþ celecoxib therapy and 47 received radical surgery. Grade 3 diarrhea developed in 5 (9%). Grade 4 anemia was seen in 2 (4%). Pathological complete response (pCR) was seen in 6 (13%). T or N downstaging found in 38 (81%). Sphincter preservation was achieved in 77% of low‐positioned tumors. Patients with tumors expressing high‐level COX‐2 after CRT þ celecoxib treatment had inferior pelvic control (P ¼ 0.01), disease‐free survival (P ¼ 0.04), and overall survival (P ¼ 0.03) than those with low‐level expression. Conclusions: Celecoxib can be safely combined with preoperative CRT for rectal cancer. More intensified adjuvant therapy may be considered for tumors expressing high‐level COX‐2 after CRT and surgery.
J. Surg. Oncol. 2014;109:580–585. ß 2013 Wiley Periodicals, Inc.
KEY WORDS: rectal cancer; COX‐2; celecoxib; chemoradiotherapy
INTRODUCTION Rectal cancer is a prevalent disease, both in Taiwan and worldwide. In treating rectal cancer, neoadjuvant radiotherapy (RT) with or without chemotherapy is often effective in cases of locally advanced disease for improving resectability [1–3], sphincter preservation [4,5], and improved survival [6]. A previous meta‐analysis showed that pre‐ operative radiation combined with chemotherapy may be more effective in the downstaging of locally advanced rectal cancer than radiation therapy alone [7]. Intravenous 5‐fluorouracil (5‐FU) plus leucovorin is the most commonly used chemotherapy regimen. Oral tegafur is a prodrug of 5‐FU. Uracil competes with 5‐FU for the enzyme dihydropyrimidine dehydrogenase, which converts 5‐FU to an inactive metabolite. Studies have demonstrated the feasibility and convenience of oral tegafur‐uracil and radiation combination therapy in the pre‐operative treatment of rectal cancer [8–10]. Cyclooxygenase‐2 (COX‐2) contributes to tumor growth through the expression of prostaglandin E2 and vascular endothelial growth factor [11]. The expression of COX‐2 has also been shown to be a negative prognostic factor for colorectal cancer, and the overexpression of COX‐2 was shown to be predictor of poor response to pre‐operative chemoradiation therapy [12]. The long‐term use of a COX‐2 inhibitor has been found to reduce rectal polyp formation in patients with familial adenosis polyposis [13]. Celecoxib inhibits COX‐2 activity, and
ß 2013 Wiley Periodicals, Inc.
demonstrates effective radiosensitization in various types of tumor cells expressing COX‐2 [14]. However, increased cardiovascular risk of patients taking COX‐2 inhibitors or celecoxib for colorectal adenoma prevention was reported [15–17]. Phase II clinical studies have also shown that celecoxib and chemoradiation (CRT) combination therapy for rectal cancer is tolerable in most patients [18,19]. However, one trial reported occurrence of skin rash in 49% of the participants [20]. In these clinical trials, the most common celecoxib dosage was 400 mg twice daily [18,20,21], which is higher than that required for mild to moderate pain, such as that induced by moderate doses (50 Gy) of radiation. Lower doses of celecoxib (e.g., 400 mg/day) during and after radiotherapy may be sufficient for radiosensitization and increase tolerability. Realizing the potential
Grant sponsor: National Health Research Institute of Taiwan, R.O.C.. *Correspondence to: Ling‐Wei Wang, MD, Cancer Center, Taipei Veterans General Hospital, No. 201, Sec. 2, Shih‐Pai Road, Taipei 112, Taiwan. Fax: þ886‐2‐28749425. E‐mail: [email protected] Received 30 August 2013; Accepted 30 November 2013 DOI 10.1002/jso.23538 Published online 24 December 2013 in Wiley Online Library (wileyonlinelibrary.com).
A Celecoxib Trial for Rectal Cancer benefits of such a regimen, we proposed a clinical trial of low‐dose celecoxib treatment combined with chemoradiotherapy for rectal cancer. In 2008, we initiated a multi‐center phase II trial sponsored by the Taiwan Cooperative Oncology Group (TCOG) of the National Health Research Institute, Taiwan for the study of celecoxib and CRT for the treatment of locally advanced rectal cancer. The primary objectives of the trial were to evaluate the efficacy of combining preoperative tegafur‐ uracil, folinate, radiation, and low‐dose celecoxib treatments for locally advanced rectal cancer, and to determine whether any toxicity is associated with this regimen. The prognostic significance of COX‐2 expression in pathological specimens before and after combination therapy was also examined.
PATIENTS AND METHODS Eligibility Our study was approved by the institutional review boards of all the seven participating hospitals in the TCOG, and was registered on www. clinicaltrials.gov as trial NCT00931203. Patients meeting the following criteria were included in our study: (1) a diagnosis of adenocarcinoma based on the pathology results of biopsy specimens; (2) clinical stage T2N1‐2M0 or T3‐4N0‐2M0, according to the American Joint Committee on Cancer (AJCC) 2002 staging system; (3) resectable or potentially resectable adenocarcinoma of the rectum; (4) underwent magnetic resonance imaging (MRI) of the pelvis, ultrasonography of liver, and chest X ray (CXR); (5) bi‐dimensionally measurable tumor based on MRI using a pelvic array coil and an intrarectal tube; (6) superior margin of the tumor below the L5‐S1 spine junction; (7) between the ages of 18 and 80 years; (8) a functional status 3.5 109/L, a neutrophil count > 1.5 109/L, a platelet count > 100 109/L, a serum bilirubin level < 125% of the reference‐range maximum (ULN), an AST/ ALT < 300% ULN, and a serum creatinine level 50%. Based on the product of the intensity and extent of staining scores, the
582
Wang et al.
sections were divided into low (total score ¼ 0–4) and high (total score ¼ 6 or 9) staining groups.
Statistical Analysis The Simon minimax two‐stage design (a ¼ 0.05, power ¼ 80%, two‐ sided test) was used for the statistical analysis of our study results. We considered the addition of celecoxib to chemoradiation as non‐effective for a pCR rate of 15% or lower, or as in need of further study for a pCR rate of 30% or higher. In the first stage of the trial, 23 patients were enrolled. If more than 3 patients had pCR, another 25 patients were enrolled. The test regimen was considered effective if 11 patients had pCR in the second stage of the trial. Overall survival was calculated from the date of enrollment in the study to the date of death. Disease‐free survival was calculated from the date of surgery to the date of treatment failure. The disease‐free survival and overall survival rates were calculated using the Kaplan–Meier method. The log‐rank test was used to compare the survival curves, and the Wilcoxon signed‐rank test was used to compare the total scores of COX‐2 staining of specimens before and after radiotherapy. A two‐sided exact test was used to evaluate the association between variables. The P‐value was considered significant when it was 6
Journal of Surgical Oncology
53 57 [26–76] 40 (75) 13 (25) 48 (91) 5 (9) 48 (91) 5 (9) 27 (51) 26 (49) 2 (4) 7 (13) 37 (70) 1 (2) 6 (12) 7 (13) 38 (72) 3 (6) 5 (9) 5 [0–12] 37 (70) 16 (30)
Most of the patients tolerated neoadjuvant treatment well. Two patients had RT interruptions, and 4 had concurrent chemotherapy interruptions because of toxicity. Four patients required modification of the celecoxib dosage, and celecoxib was withheld temporarily from 2 other patients because of adverse effects. Cases of acute toxicity are shown in Table II, most of which were grades 1–2. These toxicities included fatigue, radiation‐related rash, diarrhea, and anemia. The most common grade‐3 toxicity was diarrhea (five patients). Three participants (6%) had grade‐4 toxicities, including two patients with anemia and 1 with reversible hypercreatinemia. No grade‐5 toxicities occurred. Five patients refused surgery. One patient experienced disease progression after chemoradiation, and withdrew from the study. A total of 47 patients received TME. The median interval between RT and surgery was 44 days (range, 30–58 days). Thirty‐nine patients received low anterior resection (LAR), whereas 8 underwent abdominoperineal resection (APR). Sphincter preservation occurred in 20 of 26 patients (77%) with low‐positioned tumors (6 cm from anal verge before treatment. However, only one patient had both positive circumferential and proximal margins. The median distal margin was 2 cm, and all the distal margins were negative. The tumor response and down‐staging results are shown in Table III. Thirty‐five patients (75%) exhibited good tumor regression (TRG 0 þ 1), including six patients (13%) experienced pCR. After surgery, 42 patients received adjuvant chemotherapy, and 40 of them completed all six cycles. The median follow‐up period was 33 TABLE II. Acute Toxicities Before Surgery (N ¼ 53) Item Allergy Gastrointestinal Anorexia Diarrhea Oral mucositis Nausea Ulcer, GI Vomiting Hematology Hemoglobin Leukocytes Platelets Neutrophils Cardiovascular Hypertension Hypotension Constitutional Fatigue Weight loss Dermatology Rash (radiation‐related) Rash (hand‐foot reaction) Laboratory (biochemical) Bilirubin‐T Creatinine SGOT SGPT Hemorrhage, GI Pain Anal pain Dysuria Other pain Genitourinary Urinary frequency
Grade 1 (%)
Grade 2 (%)
Grade 3 (%)
Grade 4 (%)
2 (4)
2 (4)
1 (2)
0 (0)
17 23 3 8 0 1
(32) (43) (6) (15) (0) (2)
3 9 0 0 1 0
(6) (17) (0) (0) (2) (0)
0 5 0 0 0 0
(0) (9) (0) (0) (0) (0)
0 0 0 0 0 0
(0) (0) (0) (0) (0) (0)
28 15 17 4
(53) (28) (32) (8)
3 4 2 1
(6) (8) (4) (2)
0 0 0 0
(0) (0) (0) (0)
2 0 0 0
(4) (0) (0) (0)
4 (8) 1 (2)
0 (0) 0 (0)
0 (0) 0 (0)
0 (0) 0 (0)
30 (57) 3 (6)
5 (9) 0 (0)
0 (0) 0 (0)
0 (0) 0 (0)
22 (42) 0 (0)
13 (25) 0 (0)
1 (2) 0 (0)
0 (0) 0 (0)
0 0 0 0 1
0 1 0 0 0
12 1 6 7 8
(23) (2) (11) (13) (15)
0 0 0 1 0
(0) (0) (0) (2) (0)
(0) (0) (0) (0) (2)
(0) (2) (0) (0) (0)
6 (11) 2 (4) 5 (9)
8 (15) 0 (0) 3 (6)
0 (0) 0 (0) 1 (2)
0 (0) 0 (0) 0 (0)
6 (11)
0 (0)
0 (0)
0 (0)
A Celecoxib Trial for Rectal Cancer
583
TABLE III. Tumor Regression and Downstaging in 47 Patients
TRG 0 (¼pCR) 1 2 3 Pathological staging (yp) T0N0 T1–2N0 T3–4N0 T2–4N1 T1–3N2 T4N0M1a T downstaging N downstaging
N
%
6 29 9 3
13 62 19 6
6 13 16 9 2 1 22 31
13 28 34 19 4 2 47 66
TRG, tumor regression grading; pCR, pathological complete response. Peritoneal seeding found during operation.
a
Fig. 2. COX‐2 expressions scores before and after treatment (N ¼ 29). (range, 2.7–54) months. Thirteen cases (27%) developed distant metastases. The most common metastatic sites after surgery were the lung (N ¼ 9) and the liver (N ¼ 5). The 3‐year pelvic control, disease‐ free and overall survival rates were 89%, 63%, and 88%, respectively. The disease‐free survival curve is shown in Figure 1. Immunohistochemistry for COX‐2 expression was performed in 39 available biopsy specimens and 38 surgical specimens (excluding those from the six patients that had pCR and three with very few residual tumor cells). For the pretreatment specimens, 77% (30/39) expressed high‐ level COX‐2 expression. The paired comparison of the biopsy and surgical specimens revealed a trend toward a decreased total score for COX‐2 staining with celecoxib treatment (N ¼ 29, P < 0.001, Fig. 2). There was no association between the pre‐RT COX‐2 staining scores and the TRG of tumors after chemoradiotherapy (P ¼ 1.0). The post‐RT COX‐2 staining scores were significantly associated with pelvic recurrence (P ¼ 0.01), disease‐free survival (P ¼ 0.04, see Fig. 3), and
Fig. 1.
Disease‐free survival of all patients receiving surgery (N ¼ 47).
Journal of Surgical Oncology
overall survival (P ¼ 0.03), but were not associated with the T stage (P ¼ 0.34) or the TRG (P ¼ 1.0).
DISCUSSION The expression of COX‐2 plays important role in the carcinogenesis and the invasiveness of rectal cancer. In our clinical trial, we observed the feasibility and efficacy of combining the COX‐2 inhibitor celecoxib with pre‐operative chemoradiotherapy for the treatment of locally advanced rectal cancer, and demonstrated the role of COX‐2 expression in surgical specimens as a prognostic factor in rectal cancer. In our trial, the CRT and celecoxib treatment were well tolerated. Most (98%) patients’ pain was grade 1 or 2, which is consistent with the
Fig. 3. Disease‐free survival by post‐CRT COX‐2 scoring high (N ¼ 12) and low (N ¼ 26) groups.
584
Wang et al.
anti‐inflammatory and analgesic effects of celecoxib. A smaller fraction (10%) of patients experienced allergic reactions manifested as skin rashes compared with the study by Jakobsen et al. [20]. The severity of cardiovascular adverse effects were all grade 1 and its incidence was low (Table II). These results may be related to the low dosage and short duration of celecoxib treatment. All the grade‐4 toxicities were managed conservatively and recovered. Overall, a safe toxicity profile was observed for celecoxib. With regard to efficacy, though our pCR rate is just 13%, a high rate of the down‐staging of disease and good tumor regression (TRG 0 þ 1) were still observed. The latter accounted for approximately 75% (35/47) of the patients that received surgery (Table III). Only one case of a positive circumferential margin and a high pelvic control rate occurred, which may reflect the combined effects of chemoradiotherapy, celecoxib treatment, and effective surgical intervention. A high sphincter preservation rate (77%) was also achieved among low‐positioned tumors. Pathological complete response (pCR) rate is a common measure for efficacy of radiosensitizers in the clinic and the primary end point of this study. One randomized phase II trial by Debucquoy’s et al. [18] revealed that rectal cancer patients treated with CRT (45 Gy) and celecoxib (800 mg/day for 10–12 weeks) tended to have a better pCR rate (39%) compared with that of patients treated with CRT and placebo (29%). However, the difference in pCR rates for the two groups was not statistically significant. A phase I trial by Malik et al. [21] using celecoxib (800 mg/day for 5.6 weeks) and CRT (50.4 Gy) obtained a pCR rate of 15%. A recent phase II study by Unger et al. [19] combining celecoxib (400 mg/day for 5.6–6 weeks) with CRT (50.4 or 54 Gy) achieved a pCR rate of 25%. All the above phase I/II studies were done with small patient number (
