- © 2003 by American Society of Clinical Oncology
Phase III Double-Blind Study of Glutamine Versus Placebo for the Prevention of Acute Diarrhea in Patients Receiving Pelvic Radiation Therapy
- Timothy F. Kozelsky,
- Gregory E. Meyers,
- Jeff A. Sloan,
- Thomas G. Shanahan,
- Stephen J. Dick,
- Randy L. Moore,
- George P. Engeler,
- Albert R. Frank,
- Timothy K. McKone,
- Rodolfo E. Urias,
- Miljenko V. Pilepich,
- Paul J. Novotny and
- James A. Martenson
- From the Mayo Clinic, Rochester; Duluth CCOP, Duluth, MN; Carle Cancer Center Community Clinical Oncology Program [CCOP], Urbana, IL; Siouxland Hematology-Oncology Associates, Sioux City; Iowa Oncology Research Association CCOP, Des Moines, IA; Meritcare Hospital CCOP, Fargo, ND; Toledo Community Hospital Oncology Program CCOP, Toledo, OH; Missouri Valley Cancer Consortium, Omaha, NE; and Ann Arbor Regional CCOP, Ann Arbor, MI.
- Address reprint requests to James A. Martenson, MD, Mayo Clinic, 200 First St, SW, Rochester, MN 55905; email: jmartenson{at}mayo.edu.
Abstract
Purpose: A phase III, randomized, double-blind study was conducted by the North Central Cancer Treatment Group to determine the efficacy and toxicity of oral glutamine for the prevention of acute diarrhea in patients receiving pelvic radiation therapy (RT).
Patients and Methods: All 129 patients enrolled from 14 institutions between February 1998 and October 1999 were eligible. Patients received 4 g of glutamine or placebo orally, twice a day, beginning with the first or second day of RT and continuing for 2 weeks after RT. During treatment, patients were assessed weekly for toxicity, and a bowel function questionnaire was administered. The primary measures of treatment efficacy were diarrhea levels measured by maximum grade of diarrhea, incidence of diarrhea, and average diarrhea score. After completion of RT, the bowel function questionnaire was administered weekly for 4 weeks and at 12 and 24 months. Toxicity was measured by National Cancer Institute common toxicity criteria.
Results: The median age of patients was 69 years (range, 34 to 86 years). The two treatment arms were balanced with respect to all baseline factors. There were no significant differences in toxicity by treatment. Quality-of-life scores and the mean number of problems reported on the bowel function questionnaire were virtually identical for both treatment groups. The incidence of grade 3 or higher diarrhea was 20% for the glutamine arm and 19% for the placebo arm (P = .99). The maximum number of stools per day was 5.1 for the glutamine arm and 5.2 for the placebo arm (P = .99).
Conclusion: There is no evidence of a beneficial effect of glutamine during pelvic RT.
PELVIC RADIATION therapy (RT) is used frequently as adjuvant or primary treatment for patients with gastrointestinal, gynecologic, genitourinary, and other cancers. RT-induced diarrhea is the most frequent acute toxic response for these patients. Primarily manifested by increased stool frequency and volume, RT-induced diarrhea can have a significant negative impact on a patient’s quality of life (QOL) and potentially lead to dehydration that requires intravenous hydration or hospitalization. Severe symptoms may require temporary or permanent cessation of RT before completion of the planned RT program.
In a North Central Cancer Treatment Group (NCCTG) trial involving adjuvant RT for rectal cancer, 35% of the patients treated with concurrent pelvic RT and fluorouracil (FU) experienced grade 2 or greater diarrhea.1 The incidence of grade 3 or greater diarrhea was 20% and 4% for patients treated with RT plus FU and RT alone, respectively.1 In a clinical trial comparing placebo with olsalazine in patients receiving pelvic RT, the incidence of grade 0, 1, 2, and 3 diarrhea in the placebo group was 25%, 36%, 25%, and 14%, respectively.2
Although antidiarrheal agents often are used with some efficacy in the treatment of this problem, some patients experience severe diarrhea despite these medications. Moreover, these agents cannot be used prophylactically to prevent RT-induced diarrhea because of their tendency to cause constipation.
Glutamine plays an important role in the support of mucosal growth and function.3–,8 Many of the major adverse effects of chemotherapy or RT (or both) are related to a loss of mucosal integrity of oral and gut epithelium. Animal studies strongly demonstrate that glutamine protects both upper and lower gastrointestinal tract mucosa from the effects of chemotherapy, RT, or other causes of injury.9–,14 Although it is often assumed that diarrhea caused by pelvic RT is a result of rectal injury, damage to the small bowel also is probably an important factor, as demonstrated by studies that show that the volume of irradiated small bowel is associated with the risk of diarrhea.15–,17
A medical regimen capable of preventing or reducing symptoms of acute diarrhea would be useful in patients with a broad range of pelvic malignancies in which pelvic RT has a role. This phase III clinical trial was undertaken to assess the effectiveness of glutamine in preventing diarrhea during pelvic RT.
PATIENTS AND METHODS
Eligibility Requirements
Patients had to be at least 18 years of age or older and have histologically confirmed adenocarcinoma or squamous cell carcinoma. Patients considered for entry onto the study had to receive pelvic RT at an NCCTG-approved radiation oncology facility. The entire pelvis had to be encompassed by the planned RT fields. The superior border could not be superior to the L4–5 interspace or inferior to the most inferior aspect of the sacroiliac joints.
The total planned dose to the whole pelvic field had to be between 45 and 53.5 Gy (inclusive), with a daily dose of 1.7 to 2.1 Gy. A boost was allowed to the primary tumor or tumor bed. Patients had to be entered onto study before the second RT fraction. Patients were ineligible if they were pregnant, had a known allergy to glutamine, had a history of pelvic RT, had any history of inflammatory bowel disease, were incontinent of stool, had a prior abdominal-perineal resection, or had planned use of leucovorin or cytotoxic chemotherapeutic agents concurrent with RT, other than FU. Written informed consent and institutional review board approval were required before entry of any patient onto this study.
Study Design
This was a two-arm, placebo-controlled, randomized clinical trial. Before random assignment to treatment arms, patients were stratified by the following: history of anterior resection of the rectum versus no prior rectal operation; total planned cumulative dose (including boost fields of external-beam RT) of RT (45 to 53.5 Gy v more than 53.5 to 60 Gy v more than 60 Gy); use of FU (no FU v bolus FU v continuous-infusion FU); and primary tumor site (rectal cancer v prostate cancer v gynecologic cancer v other cancer).
After stratification, patients were randomly assigned, in double-blind manner, to receive one of the following two study medications: glutamine 4 g (8 mL) bid (morning and evening) 7 days per week during RT and for 2 weeks thereafter, or an identical-appearing placebo (glycine), which was administered according to the same schedule. Each study medication was made up of 40 mL of Ora-Sweet (Paddock Laboratories, Inc, Minneapolis, MN) plus 40 mL of Ora-Plus (Paddock Laboratories) and 80 mL of water serving as the vehicle. When the patient was placed onto the study, 112 g of glutamine (or glycine powder) was mixed with the vehicle. This would provide a drug supply for 2 weeks. l-Glutamine is highly unstable in aqueous solution at room temperature and must be refrigerated. Patients, pharmacists, and any persons handling the study medication were informed that the study medication required refrigeration.
At baseline and at weekly intervals while receiving RT, patients were evaluated by a radiation oncologist or a radiation oncology nurse, at which time toxicity was assessed according to the National Cancer Institute common toxicity criteria (Table 1⇓). The use of antidiarrheal medication and laxatives was documented.
Patients also completed a bowel function questionnaire (Appendix 2, available online at www.jco.org) at baseline, on a weekly basis throughout the course of RT, and weekly for 4 weeks after treatment. The questionnaire was derived from prior studies of the effect of RT on bowel function.18,19 Bowel function questionnaires were also administered at 12 and 24 months after RT, which are the times of routine post-RT assessment. For patients who did not return for the 12- and 24-month assessment, a bowel function questionnaire could be mailed in by the patient or information could be obtained in a telephone interview. Included with the bowel function questionnaire was a UNISCALE QOL measure (Appendix 3, available online at www.jco.org). The UNISCALE is a validated single-item visual analog tool intended to measure overall QOL.20 It has been used successfully in NCCTG trials, and its reliability, validity, and superiority over more complex instruments have been demonstrated.21–,24 Discriminant validity was examined by correlating incidence of toxicity with individual items on the bowel function questionnaire. Baseline, weekly, and posttreatment measures were included in the evaluation.
Evidence of grade 3 or worse diarrhea, rectal bleeding, or abdominal cramping required discontinuation of the study medication. Assessment of toxicity and administration of the bowel function questionnaire continued in patients who discontinued study medication.
The study design was based on an accrual goal of 120 eligible patients, 60 on each of the study arms, so that a two-sided 0.05-level Wilcoxon rank sum statistic would have a power of 97.5% to detect an improvement in which the diarrhea severity of each glutamine-treated patient decreased by one grade compared with the placebo group. The study also had a power of 80.5% to detect an improvement in which the diarrhea severity of 75% of the glutamine-treated patients decreased by one grade compared with the patients treated with placebo. NCCTG policy allows entry of slightly more patients than the minimum required by the study to ensure a minimum of 120 eligible patients. By design, an interim analysis was performed when accrual was half complete (approximately 30 patients entered onto each treatment arm), primarily to stop accrual into the glutamine arm if it was more toxic than the placebo. Nonparametric Wilcoxon and Fisher’s exact tests were used to compare the various measures of efficacy and toxicity between the arms. Type I error rates were at the 5% level for all tests.
RESULTS
Between February 1998 and October 1999, 129 patients were randomly allocated to receive either glutamine (64 patients) or a placebo (65 patients). All 129 patients were eligible and assessable; there were no cancellations. Distribution of study factors is listed in Table 2⇓. There were no differences between the two patient groups. An audit of individual patient research records revealed no reported problems with the manner in which patients or healthcare providers handled the drug, such as failure to refrigerate the preparation.
Primary efficacy measures are listed in Table 3⇓ and the mean diarrhea grades per week in the two groups are shown in Fig 1⇓. There is no indication of any difference between the two treatment arms. No significant differences between the two study arms were found with regard to incidence or grade of diarrhea, number of stools per day (reported by physician and patient), abdominal cramping, tenesmus, or use of antidiarrheal agents. There was also no difference in patient-reported symptoms between the two study arms (Table 4⇓). Information regarding bowel function was provided by 42 and 44 patients at 12 months and 39 and 35 patients at 24 months for patients treated with glutamine and placebo, respectively. No significant differences were observed (Table 5⇓).
QOL data measured by the UNISCALE for the 8 weeks after initiation of RT are shown in Fig 2⇓. There were no significant differences between the two groups with regard to baseline level, weekly assessment, or posttreatment evaluation. Both study arms showed a slight decrease in QOL during RT; however, on completion of therapy, there was a return to baseline QOL in each study arm. The proportion of patients reporting a decrease of 10 points from baseline to the end of treatment was 22% and 23% for the glutamine and placebo arms, respectively (P = .99). QOL information was provided by 71 patients (35 glutamine patients and 36 placebo patients) at 12 months and 61 patients (31 glutamine patients and 30 placebo patients) at 24 months. Median QOL scores for patients treated with glutamine and placebo were 92 and 91 at 12 months (P = .94) and 89 and 84 at 24 months (P = .13), respectively.
An ancillary finding from this study was that the bowel function and QOL data could be collected easily in the 8 weeks after initiation of RT and provided useful information in such studies. Completion rates for the bowel function questionnaire (which included the UNISCALE QOL instrument) exceeded 91% in both arms during this time.
DISCUSSION
Diarrhea and other changes in bowel habits are commonly observed during and after pelvic RT.1,2,25,26 Chemotherapy seems to result in increased toxicity. For example, a study by Miller et al1 reported that the incidence of grade 2 or greater diarrhea in patients receiving RT was 20%, compared with 35% for patients receiving RT and FU. Studies have used QOL measures to assess treatment-related toxicity.18,19 Agents that could reduce such treatment-related toxicity would have a significant impact in the clinic. However, such agents need to be without significant toxicity.
Limited evidence indicates that glutamine may be effective in preventing RT- or chemotherapy-induced mucosal injury.8–,12 Many of the major adverse effects of cancer treatment are related to lack of mucosal integrity from chemotherapy- or RT-induced damage of oral and gut epithelium. Any agent that could reduce such adverse effects would have far-reaching implications in oncology.
Two studies evaluated the benefits of glutamine in patients undergoing RT alone. In a small randomized trial by Huang et al,27 the authors found a statistically significant reduction in duration and severity of objective oral mucositis with the use of glutamine but not with placebo in patients undergoing RT for head and neck cancers.
A study by Richards et al28 evaluated the benefits of glutamine in the prevention of RT-induced enterocolitis in patients undergoing RT for prostate cancer. Detailed patient accounts of the incidence and frequency of clinical symptoms associated with RT-induced enterocolitis and histologic and morphometric analysis of endoscopically directed rectosigmoid biopsies were evaluated in a group of 26 patients. Patients were randomly assigned to receive 21 g (7 g three times daily with meals) of either glutamine or placebo. Although the average number of stools, the stool volume, and the stool consistency did not differ between the groups, biopsies (performed before RT and 11 days after RT) revealed significant improvements in the group receiving glutamine.
Glutamine supplementation in patients undergoing chemotherapy alone has had mixed results. A randomized study by Okuno et al29 evaluated the use of glutamine (4 g swish and swallow twice a day) or placebo in 134 patients receiving FU-based chemotherapy. No statistically significant differences were seen in the incidence or severity of oral mucositis. Another randomized study by Jebb et al30 failed to show a benefit of oral glutamine (16 g per day) compared with placebo in patients receiving FU chemotherapy.
Anderson et al31 evaluated the efficacy of glutamine (2 g/m2 bid or placebo in 16 children and eight adults) to reduce stomatitis after cytotoxic chemotherapy. This dose was similar to that used in our trial. Low-dose glutamine supplementation during and for 2 weeks after chemotherapy significantly reduced both the duration and severity of chemotherapy-associated stomatitis. A small study recently reported the effectiveness of glutamine (10 g twice a day) for the prevention of symptomatic myalgia and arthralgias in patients receiving paclitaxel chemotherapy.32
Bozzetti et al33 evaluated the efficacy of glutamine in preventing doxifluridine-induced diarrhea in a double-blind randomized trial in 65 patients. Patients received glutamine 30 g per day (divided into three doses of 10 g each) or a placebo during 8 consecutive days between chemotherapy. Diarrhea was assessed, and severity was scored by the National Cancer Institute classification. A median of 10 cycles of chemotherapy were administered to each group; there were 34 and 32 episodes of diarrhea in the glutamine and placebo groups, respectively. No significant difference was seen between the two groups.
A study by Daniele et al34 evaluated the effects of glutamine in the prevention of FU-induced intestinal tract toxicity in a double-blind, placebo-controlled randomized trial. Patients with colorectal cancer were randomly assigned to either 18 g of oral glutamine per day or placebo. Importantly, the glutamine (or placebo) was started 5 days before chemotherapy; both intestinal absorption and intestinal permeability were measured. The researchers found a statistically significant decrease in intestinal absorption and a larger reduction of intestinal permeability in the placebo arm. The investigators concluded that intestinal absorption and permeability caused by FU can be reduced by glutamine, which may also offer a protective effect on chemotherapy-induced diarrhea.
A study by Yoshida et al35 evaluated the effects of glutamine in patients with locally advanced esophageal cancer receiving combined-modality therapy (chemotherapy and RT). This was a randomized trial of 30 g per day of oral glutamine given at the start of chemoradiotherapy and for the subsequent 28 days. Patients received FU and cisplatin chemotherapy concurrently with hyperfractionated irradiation, 1.2 Gy bid for a total dose of 36.2 Gy. Oral glutamine preserved the lymphocyte count and attenuated the gut permeability compared with the control group.
In the present study, glutamine at a dose of 4 g orally twice daily was not effective in preventing diarrhea or other treatment-related toxicity seen during pelvic RT. The schedule and dose used for glutamine administration were similar to those used in the studies by Anderson et al31 and Okuno et al.29 QOL and toxicity were similar in both treatment arms.
One possible explanation for the variable results with glutamine may be the variation in the doses, schedules, or drug formulations used in different studies. No study has addressed dose-response relationships, so the optimal dose is not currently known. It is possible that under severe stress, such as that encountered with pelvic RT, much higher levels of glutamine may be needed to maintain epithelial integrity in the gut.32 Hall et al3 indicated that under stress 20 to 40 g per day of glutamine may be necessary to maintain homeostasis. Another issue is the timing of glutamine administration. In the study with positive results by Daniele et al,34 glutamine administration was started 5 days before chemotherapy rather than with the start of treatment, as in the current study. A final concern is the formulation of the drug used in our trial, which required refrigeration. An audit failed to reveal any problems. The possibility cannot be excluded, however, that a more stable formulation of the drug might have yielded different study results.
In conclusion, in the dose, schedule, and formulation used in this trial, glutamine was well tolerated but was not effective in preventing diarrhea during pelvic RT. In addition, our data showed no advantage or benefit of glutamine to prevent other treatment-related toxic responses of the gastrointestinal tract during pelvic RT. QOL was similar in patients treated with glutamine or placebo. Our results do not rule out the possibility that future trials might demonstrate a benefit for glutamine using a different dose, schedule, or formulation. Clinical trials using higher doses of glutamine would be especially interesting. At present, however, the use of glutamine during pelvic RT to reduce treatment-related diarrhea cannot be recommended, except in the context of a controlled clinical trial.
APPENDIX 1
Additional participating investigators and institutions included: Ferdinand Addo, MD, Medcenter One Health Systems, Mid Dakota Clinic, Bismarck; Daniel J. Walsh, MD, Altru Health Systems, Grand Forks, ND; Martin Wiesenfeld, MD, Cedar Rapids Oncology Project Community Clinical Oncology Program (CCOP), Cedar Rapids, IA; Harold E. Windschitl, MD, CentraCare Clinic, St Cloud, MN; Tom R. Fitch, MD, Scottsdale CCOP, Scottsdale, AZ; and Maria Tria Tirona, MD, Saskatoon Cancer Centre, Saskatoon, and Allan Blair Cancer Centre, Regina, Saskatchewan, Canada).
APPENDIX 2
Patient Questionnaire on Bowel Function
APPENDIX 3
UNISCALE–Patient
National Cancer Institute Common Toxicity Criteria
Distribution of Baseline Factors in Assessable Patients
Measures of Efficacy
Patient-Reported Symptoms, by Treatment
Patient-Reported Symptoms at 12 and 24 Months, by Treatment
Mean diarrhea grade by week of treatment with radiation therapy. There was no significant difference between the patients who received placebo and those who received glutamine.
Mean quality of life (QOL) score by week of treatment with radiation therapy. A score of 100 represents the highest QOL and a score of 0 represents the lowest QOL. There was no significant difference between the patients who received placebo and those who received glutamine.
Footnotes
-
This study was conducted as a collaborative trial of the North Central Cancer Treatment Group and Mayo Clinic and was supported in part by Public Health Service grant nos. CA-25224, CA-37404, CA-15083, CA-63826, CA-35195, CA-35103, CA-37417, CA-35415, CA-63849, CA-35101, CA-35269, and CA-63848.
-
Presented, in part, at the Thirty-Seventh Annual Meeting of the American Society of Clinical Oncology, San Francisco, CA, May 13, 2001.
- Received May 7, 2002.
- Accepted February 3, 2003.
