• enteropathy;
  • diarrhea;
  • chemotherapy;
  • colon cancer;
  • adjuvant therapy;
  • oxaliplatin;
  • fluorouracil;
  • leucovorin


  1. Top of page
  2. Abstract


Cases of severe gastrointestinal toxicity were monitored prospectively during NSABP C-07, a randomized clinical trial of adjuvant therapy for patients with stage II/III colon cancer.


Patients were treated with weekly bolus 5-fluorouracil (5-FU) and leucovorin (FL; “Roswell Park Regimen”) or the same regimen plus oxaliplatin (FLOX).


Of 1857 patients, 79 (4.3%) developed a syndrome of bowel wall injury (BWI, small or large) characterized by hospitalization for the management of severe diarrhea or dehydration and radiographic or endoscopic evidence of bowel wall thickening or ulceration. Fifty-one (64.6%) of these adverse events occurred in patients treated with FLOX and 28 (35.4%) in those treated with FL (P < .01). Enteric sepsis (ES), characterized by grade 3 or greater diarrhea and grade 4 neutropenia with or without proven bacteremia occurred in 22 patients treated with FLOX, versus 8 in those treated with FL (P = .01). Patients >60 years were at higher risk for BWI after treatment with FLOX (6.7%) versus treatment with FL (2.9%, P < .01). Female patients had a higher incidence of BWI with FLOX (9.1%) than with FL (3.9%, P < .01). Severe gastrointestinal toxicity usually occurred during the third or fourth week on the first cycle of therapy, required hospitalization, and was managed with fluids, antidiarrheals, and antibiotics. There were 5 deaths (0.3%) due to enteropathy, 2 related to ES and 3 related to both BWI and ES. Seventy-one percent of patients resumed treatment with FL after recovery.


Patients treated with adjuvant FL should be closely monitored for diarrhea and aggressively managed, especially if oxaliplatin has been added to the regimen. Cancer 2007. © 2007 American Cancer Society.

Postoperative adjuvant chemotherapy that decreases the risk of cancer recurrence and improves survival has become standard treatment in patients with resected high-risk colon cancer.1 Severe diarrhea associated with 5-fluorouracil (5-FU)-containing chemotherapy regimens can occasionally result in prolonged hospitalization, limiting both the duration and intensity of effective cancer-directed therapy.2, 3

During the National Surgical Adjuvant Breast and Bowel Project (NSABP) C-07 randomized clinical trial comparing the “Roswell Park Regimen” of 5-FU and leucovorin (FL) to the same regimen plus oxaliplatin (FLOX) in patients with resected stage II and III colon cancer, we became aware of a hitherto infrequently recognized diarrheal syndrome occurring in both arms of the study.4 This toxicity was characterized by severe diarrhea or dehydration associated with bowel wall injury (BWI, small or large) on radiographic imaging studies or endoscopy. In response to these adverse events, special reporting of grade 3 or 4 gastrointestinal (GI) toxicity was instituted for the duration of the study. In this article we report our evaluation of BWI or enteric sepsis (ES) in the 2 arms of the study.


  1. Top of page
  2. Abstract

NSABP C-07 was approved by local human investigations committees or institutional review boards in accordance with assurances filed with and approved by the Department of Health and Human Services. Written informed consent was required for participation in the trial. Between February 2000 and November 2002, 2492 patients with resected stage II or III colon cancer entered NSABP clinical trial C-07, which compared 3 8-week cycles of FL (500 mg/m2 5-FU given by bolus injection plus 500 mg/m2 leucovorin given by 2-hour infusion intravenously weekly for 6 weeks followed by a 2-week rest period) or FLOX (FL plus oxaliplatin 85 mg/m2 intravenously over 2 hours on Weeks 1, 3, and 5 of each cycle). Patients were eligible if they had undergone a curative resection within the previous 42 days with negative surgical margins for stage II (T3,4, N0, M0) or stage III (any T, N1,2, M0) colon cancer that was ≥12 cm from the anal verge. Patients had to have adequate hepatic and renal function, performance status 0-2, postoperative granulocyte count ≥1500/mm3, and platelet count ≥100,000/mm3. Those with more than 1 synchronous primary or intestinal obstruction but not free perforation were eligible.

Reports of adverse events were received by the NSABP Biostatistical Center. Through weekly reviews of reported toxicity, it became apparent that a cohort of patients was developing similar GI adverse events. In the fall of 2000, NSABP investigators were instructed to report the following adverse events from both arms of the trial in an expedited manner: Bowel Wall Injury (BWI): Hospitalization for the management of diarrhea or dehydration, with radiographic or endoscopic evidence of bowel wall thickening or ulceration. An example of bowel wall thickening on a computed tomography (CT) scan in a patient receiving FL is shown in Figure 1. Enteric Sepsis (ES): Grade 3 or greater diarrhea and grade 4 neutropenia with or without proven bacteremia.

thumbnail image

Figure 1. Computed tomography scan of the abdomen demonstrating thickening of the bowel wall (arrow) in a patient with bowel wall injury.

Download figure to PowerPoint

Of the total population of patients participating in NSABP C-07, 1857 patients (929 FL, 928 FLOX) were included in this analysis. These were eligible patients who were randomized on or after November 1 2000 and started treatment, coinciding with the implementation of forms collecting data on BWI and ES on both treatment arms prospectively.

NCI Common Toxicity Criteria (v. 2) were used to grade diarrhea, neutropenia, and other toxicities. Rates between the 2 treatment arms were compared using the chi-square test. All reported P-values are 2-sided.


  1. Top of page
  2. Abstract

Patient characteristics of age, sex, and race were well balanced between the 2 treatment arms (Table 1). Women made up 43.9% of the study population. Slightly more women received FLOX (46.2%) than FL (41.7%). Of the patients treated with FL, 16.3% were older than 70 years, compared with 15.8% of those treated with FLOX. The majority of patients on both treatment arms were white (85.1%).

Table 1. Incidence of Severe Gastrointestinal Toxicity Among Various Patient Groups Treated on NSABP Study C-07 and Randomized on or After November 1, 2000
 FL (N = 929)FLOX (N = 928)
 Total %BWI no. (%)ES no.BWI or ES no. (%)Total %BWI no. (%)ES no.BWI or ES no. (%)
  1. BWI indicates bowel wall injury; ES, enteric sepsis.

Age, y        
 <6051.73.1 (15)43.8 (18)53.24.5 (22)55.3 (26)
 60-6932.13.0 (9)23.4 (10)30.96.6 (19)148.7 (25)
 ≥7016.32.7 (4)23.3 (5)15.86.8 (10)38.8 (13)
 Men58.32.4 (13)33.0 (16)53.82.4 (12)43.0 (15)
 Women41.73.9 (15)54.4 (17)46.29.1 (39)1811.4 (49)

Bowel wall injury was reported in 51 (5.5%) patients treated with the FLOX regimen and in 28 (3.0%) patients who received FL without oxaliplatin (P < .01; Table 2). In both arms the small bowel was involved more frequently with BWI than was the colon. Involvement of the ileum was more common than involvement of other sites within the small bowel. BWI involving both small bowel and colon occurred in 12 (2.4%) patients treated with FLOX versus only 1 (0.1%) who received FL. ES was also more common in patients treated with FLOX, 2.4% versus 0.9% (P = .01; Table 2). Few patients had both BWI and ES, 3 patients on FL and 9 on FLOX.

Table 2. Incidence of Bowel Wall Injury (BWI) and Enteric Sepsis (ES) Among 1857 Patients Randomized on or After November 1, 2000 on NSABP C-07*
 FL % N = 929FLOX % N = 928
  • FL indicates 5-fluorouracil and leucovorin; FLOX, FL and oxaliplatin.

  • *

    Patients may be included in more than 1 category.

  • Data collected prospectively.

Patients with BWI3.0 (28)5.5 (51)
Colon0.7 (6)2.2 (20)
Small bowel2.5 (23)4.5 (42)
Unspecified0.2 (2)0.0 (0)
Patients with ES0.9 (8)2.4 (22)
Either BWI or ES3.6 (33)6.9 (64)
Both BWI and ES0.3 (3)1.0 (9)

The incidence of both BWI and ES are shown by patient characteristics in Table 1. The incidence of BWI was similar in patients younger than 60 years of age treated with either FL (3.1%) or FLOX (4.5%, P = .28). Those ≥60 years of age, however, were at higher risk for BWI after treatment with FLOX (6.7%) than with FL (2.9%, P < .01).

Women had a higher incidence of BWI in the FLOX treatment arm (9.1%) than in the FL arm (3.9%, P < .01). Treatment with FLOX did not increase toxicity in men (2.4% in each arm). ES was also more common in women (2.8%) than in men (0.7%, P < .001). Similar gender differences were apparent for other GI toxicities such as nausea, vomiting, and dehydration, all of which were seen more frequently in women than in men treated with either FLOX or FL (data not shown). The distribution of toxicity by race could not be adequately studied because of small numbers of minorities recruited into the trial.

BWI and ES usually occurred during the first cycle of therapy, peaking in the third week for those receiving FLOX and in the fourth week for those receiving FL treatment (Fig. 2). Only 11.4% of all BWI events occurred after the first cycle of therapy. Duration of BWI was similar between the FL and FLOX groups (median 13 days from diagnosis to discharge from hospital). Clostridium difficile infection in patients with BWI or ES was rare, reported in only 1 of 33 patients receiving FL and in 4 of 64 patients treated with FLOX. Surgical intervention was required in 9 patients with BWI (11.4%), 5 FL-treated, and 4 FLOX-treated. Two patients who developed ES with FLOX (9.0%) were taken to surgery; both of these patients also had BWI. None of those who encountered ES after treatment with FL required subsequent surgery.

thumbnail image

Figure 2. Number of bowel wall injury and/or enteric sepsis events among patients by cycle and treatment. One case of enteric sepsis on the FL arm is omitted from the figure because the cycle and treatment of onset could not be determined. FL, 5-FU and leucovorin; FLOX, 5-FU and leucovorin plus oxaliplatin.

Download figure to PowerPoint

Seventy-one percent of those with BWI were retreated with FL after recovery from toxicity as allowed per protocol. One of these patients developed BWI after retreatment. After the development of ES, 15 of 30 patients resumed treatment with FL and 1 patient had further evidence of severe diarrhea and sepsis.

Five patients (0.3%) in this cohort of 1857 patients randomized after November 1 2000 died with severe bowel toxicity. Death in 2 FLOX-treated patients was attributable to ES. Two other FLOX-treated patients and 1 FL-treated patient died from complications related to both BWI and ES.


  1. Top of page
  2. Abstract

Injury of the bowel wall by a variety of chemotherapeutic agents has been previously reported.5–12 Neutropenic enterocolitis has been associated with high-dose chemotherapy and bone marrow transplant and with standard-dose chemotherapy in patients with solid tumors. BWI in the absence of neutropenia has also been reported in patients treated with FL.9 Grade 4 neutropenia was not a component of the BWI syndrome in the present report.

Diarrhea and dehydration with or without neutropenia is a common toxicity reported in patients treated with 5-FU combination chemotherapy regimens.1, 2 This can lead to causing further morbidity and delay of therapy. Patients with GI toxicity from therapy, especially in the adjuvant setting, may not complete the prescribed treatment, thereby potentially compromising treatment efficacy. In addition, the estimated cost of treatment for this toxicity is substantial.13

The schedule of FL given weekly for 6 weeks developed at Roswell Park Memorial Institute14 as given in the FL regimen on NSABP C-07 was associated with diarrhea in 40% of previously untreated patients with advanced colorectal carcinoma, half of whom required hospitalization for intravenous hydration.14 However, the present report is the first large, prospective evaluation of BWI and ES in patients treated with this regimen with or without the addition of oxaliplatin in the adjuvant setting.

Bowel wall thickening on CT scan has been described in association with ischemia because of a wide variety of causes,15 including cancer chemotherapy. This radiographic finding is therefore nonspecific, reflecting edema of the bowel wall. In our study, bowel wall thickening was associated with severe diarrhea and dehydration and required surgical intervention in 11.4% of patients who developed this abnormality.

BWI and ES in our study occurred in patients treated with either FL or FLOX but were more common and were seen somewhat earlier in those treated with FLOX, suggesting that oxaliplatin independently contributed to GI toxicity. Regardless of which treatment patients received, BWI more frequently involved the small bowel rather than the large bowel. Preferential involvement of small bowel with 5-FU-induced GI toxicity has been noted by others.12 In the vast majority of patients in this report, the sites of BWI involvement were detected by CT scan. Barium radiographs of the colon or small bowel were used in some cases, and a few cases were confirmed by endoscopy or surgery.

Older patients were more susceptible to the GI toxicity associated with FLOX. It is unclear if this was due to associated medical conditions, delay in treatment of symptoms, differences in drug metabolism, or greater intrinsic susceptibility to bowel wall toxicity from oxaliplatin given in combination with 5-FU in this cohort of patients. The enhanced susceptibility to GI toxicity among older patients cannot be explained by poorer pretreatment performance status, however, because there was no significant association between performance status and age. Women also had a higher incidence of BWI and ES than did men, especially with FLOX. Enhanced susceptibility of women to various toxicities from fluorinated pyrimidines has been previously reported,16, 17 and therefore the differences in toxicity according to gender observed in this study are not new. The reason for these gender differences is unclear, but is not related to cumulative toxicity resulting from multiple doses of chemotherapy because the vast majority of cases of enteropathy among both men and women occurred after only a few doses of chemotherapy in the first cycle of therapy.

In this study the management of established BWI or ES required hospitalization, which was often prolonged before diarrhea resolved. Recovery was only rarely delayed by C. difficile superinfection. Despite the severity of symptoms, two-thirds of patients subsequently received further adjuvant therapy, as allowed per protocol. These patients were to be treated with a reduced dose of 5-FU in combination with leucovorin, which may explain the low incidence (1 case) of reported episodes of recurrent bowel injury. The outcome in 8 patients with BWI or ES retreated with an oxaliplatin-containing regimen after initially receiving FLOX is unknown, except for 1 patient who had recurrent ES after only 1 additional dose of FLOX, and a second patient who experienced recurrent BWI.

The therapeutic results of C-0718 confirm the earlier conclusions from the MOSAIC trial19 that a combination of 5-FU, leucovorin, and oxaliplatin improves disease-free survival after resection for stage II and III colon cancer. Analysis of disease-free survival in each of these clinical trials demonstrated a relative risk reduction of approximately 20% when oxaliplatin was added to 5-FU and leucovorin. GI toxicity associated with the bolus and continuous infusion 5-FU-containing FOLFOX4 regimen used in the MOSAIC trial appears to be much less than in the “Roswell Park Regimen” of weekly bolus 5-FU and 2-hour leucovorin infusion FLOX regimen. The incidence of grade 3 and 4 diarrhea was reported to be 10.8% for FOLFOX4 and 38.0% for FLOX. Neither the BWI nor the ES syndrome was reported after treatment with FOLFOX4. There was more grade 3/4 neutropenia associated with FOLFOX4 than with FLOX, 41.1% versus <5%, but neutropenic fever or infection documented by culture was reported in only 1.8% of patients treated with FOLFOX4. The higher incidence of neurotoxicity noted in the MOSAIC trial for FOLFOX4-treated patients (12.4%) over that reported in C-07 for FLOX-treated patients (8.2%) is likely related to the 25% decreased cumulative dose of oxaliplatin used in the latter trial. Overall treatment-related mortality was reported in 0.5% of patients treated with FOLFOX4 versus 1.2% of patients treated with FLOX.

The higher incidence of GI toxicity from FLOX should be considered when choosing an adjuvant therapy for patients with resected colon cancer. The incidence of BWI or ES from FLOX is 6.9%. Although toxicity can be severe and sometimes prolonged, upon recovery many patients can continue therapy safely with a reduced dose of the FL regimen. Frequent monitoring for the onset of diarrhea is recommended for patients on FLOX because mild diarrhea can progress rapidly to more severe episodes of toxicity.1 Patients should be warned about severe diarrhea and instructed to initiate oral therapy with an antidiarrheal agent at the first sign of toxicity. A low threshold should be maintained to withhold chemotherapy and begin intravenous hydration in patients who suffer more than 4 diarrheal episodes per day. Hospitalization and aggressive management for possible infection and fluid and electrolyte disorders is mandated in patients who have evidence of BWI or ES. Clostridium difficile should be ruled out in such cases and patients should receive antibiotics. Consideration should be given for the use of a long-acting formulation of octreotide, which has documented efficacy in the treatment of diarrhea.20 With the resolution of symptoms, patients may be retreated, but we do not recommend the use of additional oxaliplatin.

The FLOX and FOLFOX regimens appear to be similarly effective as postoperative adjuvant therapy for stage II and III colon cancer.18, 19 The lower incidence of neurotoxicity along with the simplicity of bolus delivery versus an implanted catheter or port and infusion pump may make adjuvant therapy with FLOX a better choice in selected patients, provided care is taken to minimize and to aggressively manage the enteropathy described in this report.


  1. Top of page
  2. Abstract
  • 1
    NCCN Practice Guidelines in Oncology-v.2.2006. Available at URL: Accessed on January 18, 2007.
  • 2
    Saltz LB. Understanding and managing chemotherapy-induced diarrhea. J Supp Oncol. 2003; 1: 3546.
  • 3
    O'Brien BE,Kaklamani VG,Benson AB3rd. The assessment and management of cancer treatment-related diarrhea. Clin Colorectal Cancer. 2005; 4: 375381.
  • 4
    Smith RE,Colangelo L,Wieand S, et al. The occurrence of severe enteropathy among patients with stage II/III resected colon cancer (CC) treated with 5-fluorouracil/leucovorin (FL) plus oxaliplatin (FLOX) [abstract]. Proc Am Soc Clin Oncol. 2003: 22. Abstract 1181.
  • 5
    Takaoka E,Kawal K,Ando S, et al. Neutropenic colitis during standard dose combination chemotherapy with nedaplatin and irinotecan for testicular cancer. Jpn J Clin Oncol. 2006; 36: 6063.
  • 6
    Wenzel C,Urbauer E,Schwarz C, et al. Severe enteropathy associated with ralitrexed and oxaliplatin chemotherapy: report of two patients experiencing this rare, potentially lethal gastrointestinal adverse event. Anticancer Drugs. 2006; 17: 865868.
  • 7
    Avigan D,Richardson P,Elias A, et al. Neutropenic enterocolitis as a complication of high dose chemotherapy with stem cell rescue in patients with solid tumors: a case series of a review of the literature. Cancer. 1998; 83: 409414.
  • 8
    Tham RT,Vlasveld LT,Willemze R. Gastrointestinal complications of cytosine-arabinoside chemotherapy: findings on plain abdominal radiographs. AJR Am J Roentgenol. 1990; 154: 9598.
  • 9
    Zamani R,Heldmann M. Enteritis as a complication of adjuvant combination chemotherapy using 5-fluorouracil and leukovorin: clinical and helical computed tomographic features. J La State Med Soc. 2004; 156: 143144.
  • 10
    Kelvin FM,Gramm HF,Gluck WL, et al. Radiologic manifestations of small bowel toxicity due to floxuridine therapy. AJR Am J Roentgenol. 1986; 146: 3943.
  • 11
    Hiehle JFJr,Levine MS. Gastrointestinal toxicity of 5-FU and 5-FUDR: radiographic findings. Can Assoc Radiol J. 1991; 42: 109112.
  • 12
    Fata F,Ron IG,Kemeny N, et al. 5-fluorouracil-induced small bowel toxicity in patients with colorectal carcinoma. Cancer. 1999; 86: 11291134.
  • 13
    Dranitsaris G,Maroun J,Shah A. Estimating the cost of illness in colorectal cancer patients who were hospitalized for severe chemotherapy-induced diarrhea. Can J Gastroenterol. 2005; 19: 8387.
  • 14
    Petrelli N,Herrera L,Rustum Y, et al. A prospective randomized trial of 5-fluorouracil versus 5-fluorouracil and high-dose leucovorin versus 5-fluorouracil and methotrexate in previously untreated patients with advanced colorectal carcinoma. J Clin Oncol. 1987; 5: 15591565.
  • 15
    Chou CK. CT manifestations of bowel ischemia. AJR Am J Roentgenol. 2002; 178: 8791.
  • 16
    Sloan JA,Goldberg RM,Sargent DJ, et al. Women experience greater toxicity with fluorouracil-based chemotherapy for colorectal cancer. J Clin Oncol. 2002; 20: 14911498.
  • 17
    Chansky K,Benedetti J,Macdonald JS. Differences in toxicity between men and women treated with 5-fluorouracil therapy for colorectal carcinoma. Cancer. 2005; 103: 11651171.
  • 18
    Kuebler P,Wieand HS,O'Connell MJ, et al. Oxaliplatin combined with weekly bolus 5-fluorouracil and leucovorin as surgical adjuvant chemotherapy for stage II and III colon cancer. Results from NSABP protocol C-07. J Clin Oncol. 2007; 25: 21982204.
  • 19
    Andre T,Boni C,Mounedji-Boudiaf L, et al. Oxaliplatin, fluorouracil and leucovorin as adjuvant treatment for colon cancer. N Engl J Med. 2004; 350: 23432351.
  • 20
    Rosenoff S. Resolution of refractory chemotherapy-induced diarrhea (CID) with octreotide long-acting formulation in cancer patients: 11 case studies. Support Care Cancer. 2004; 12: 561570.