Fax: (847) 570-8033
Evaluation of pain associated with oral mucositis during the acute period after administration of high-dose chemotherapy
Version of Record online: 4 JUN 2003
Copyright © 2003 American Cancer Society
Volume 98, Issue 2, pages 406–412, 15 July 2003
How to Cite
Cella, D., Pulliam, J., Fuchs, H., Miller, C., Hurd, D., Wingard, J. R., Sonis, S. T., Martin, P. J. and Giles, F. (2003), Evaluation of pain associated with oral mucositis during the acute period after administration of high-dose chemotherapy. Cancer, 98: 406–412. doi: 10.1002/cncr.11505
- Issue online: 2 JUL 2003
- Version of Record online: 4 JUN 2003
- IntraBiotics Pharmaceuticals, Inc. (Mountain View, CA)
- oral mucositis;
- oral pain;
- iseganan HCl
No single oral mucositis (OM) assessment scale is universally accepted; the most commonly used scales are deficient because they combine subjective and objective measures and do not capture the patient's perspective. Because pain is the hallmark symptom of OM, the authors sought to determine whether a simple measure of patient-reported pain was correlated with objective, physician-assessed measures of OM. The findings of the current study may provide a clinical context for understanding the relation between objective indicators and patients' perceptions of OM.
Three hundred twenty-three patients receiving stomatotoxic chemotherapy and randomized to receive either iseganan or placebo for treatment of OM underwent periodic objective and subjective evaluations of OM. Objective measures included clinician scoring of stomatitis and dysphagia using the National Cancer Institute Common Toxicity Criteria scales. A subjective measure was obtained by having patients complete a questionnaire (with questions based on an 11-point numeric scale) regarding oral pain.
More than 90% of scheduled oral assessments were obtained. Mouth pain scores were closely related to stomatitis and dysphagia; peak mouth pain coincided with peak stomatitis and dysphagia. Analgesic use increased by 0.7 days for each unit rise on the pain scale. Patients receiving iseganan had a significantly lower level of peak mouth pain than did patients receiving placebo (P = 0.041).
A separate measurement of patient-reported pain was useful for capturing the patient's perspective on OM and was correlated with the physician's objective assessment. These findings support the use of a simple, patient-reported rating of mouth pain as a clinically relevant and responsive endpoint in clinical trials. This rating system also may provide a straightforward method of following OM in clinical practice. Cancer 2003;98:406–12. © 2003 American Cancer Society.
Patients who have received myeloablative doses of chemotherapy often cite oral mucositis (OM) as the most debilitating side effect.1, 2 The impact of OM is underscored by the finding that over 75% of patients who receive myeloablative chemotherapy experience OM.3–5 The painful sequelae of OM interfere with oral function and frequently are associated with negative outcomes that require parenteral narcotic therapy, supplemental intravenous fluids or nourishment, and extended hospitalization. Patients with OM also are at increased risk of systemic infection and mortality.6 Of all the symptoms associated with OM, oral pain is the one that patients report as the most troublesome.2, 7 In addition, both physicians and patients have identified oral pain as the most important symptom to measure in clinical trials of OM interventions.8
There are a number of interventions for OM that are advancing in development; these interventions include topical approaches and the use of various systemic agents.1, 9 Studies of an intervention should capture measurements that are tangible and important to patients. Current approaches differ with regard to the extent to which they measure objective oral changes, and most approaches do not incorporate information from the patient's perspective. Consequently, it is difficult to compare the extent and effect of oral mucosal changes among studies and to appreciate the impact of these changes as felt by the patient. Pain is a symptom experienced by nearly all patients who develop OM and therefore represents an important and relevant endpoint to measure in clinical trials. A useful scale of pain measurement would be reported directly by patients (as opposed to their health care providers), be sensitive to the evolution of OM, and be correlated with other objective or subjective measures of the disease.
Studies of OM interventions typically use an assessment scale to evaluate changes in the oral mucosa and subjective measures such as pain. Some of the more commonly used OM assessment scales are part of broader instruments like the World Health Organization toxicity grading scale and the National Cancer Institute Common Toxicity Criteria (NCI CTC). One novel approach, the Oral Mucositis Assessment Scale (OMAS), also includes anatomic evaluation of the extent and severity of lesions.6, 10 Due to lack of specificity, lack of patient input, difficulty of use with patients experiencing severe pain, and technical challenges in administration,1 no single scale has emerged as a universal choice for measuring OM. An additional problem is that subjective evaluations are entangled with objective assessments, prohibiting the analysis of the subjective contribution and the relation between the two types of measurements. Furthermore, subjective complications are measured from the clinician's point of view rather than from the patient's. The challenges associated with OM assessment scales highlight the need for a single, simple measure for following the clinical course of OM and evaluating the efficacy of OM interventions.
It recently was reported that iseganan HCl increased the proportion of patients who did not develop ulcerative oral mucositis.11 At Day 21, 43% of patients receiving iseganan had not developed ulcerative oral mucositis, compared with 33% of patients receiving placebo (P = 0.067). Iseganan also significantly reduced stomatitis severity, patient-reported peak and average mouth pain, peak throat pain, and average difficulty swallowing. From the same trial that yielded these findings, we now report the results of a patient-reported pain assessment and the correlation between the results of this assessment and objective measures of OM severity. Our goal is to provide a clinical context for understanding the relation between objective indicators of OM and the patient's perception of his or her condition. Better understanding of the impact of OM from the patient's perspective can provide support for the clinical relevance of objective ratings of OM and for the validity of subjective assessment, provided that objective and subjective ratings are closely related to one another in terms of timing and intensity.
MATERIALS AND METHODS
Patients were enrolled in 1 of 24 centers across the United States, the United Kingdom, and Germany; the study was conducted with institutional review board approval at all sites. Patient eligibility criteria included age ≥ 7 years, no evidence of OM at randomization, and planned treatment consisting of chemotherapy associated with NCI CTC Grade 2 or worse stomatitis in at least 50% of patients. Patients were required to be free of oral mucositis and to give written informed consent before trial entry. Patients were randomized to rinse with an oral solution of either iseganan HCl or placebo 6 times daily during the first 21 to 28 days of chemotherapy or radiotherapy; patients, study personnel, and the sponsor were blinded with regard to who received iseganan and who received placebo. Patients were allowed to use any concomitant medications, including narcotics, as prescribed by their clinicians. Parenteral narcotic analgesic use was recorded during the study.
Evaluations of OM began on the first day of cytotoxic therapy, and for 3 weeks, patients were evaluated 3 times weekly, using both subjective and objective measures, for the presence of OM and its sequelae. Objective measures of OM included clinician scoring of stomatitis and dysphagia using the NCI CTC Stomatitis for Myeloablative Chemotherapy Scale and the NCI CTC Dysphagia Scale (Table 1). These scales range from 0 (no evidence of abnormality) to 4 (severe or life-threatening impairment). A subjective measure of OM was obtained by having patients complete a brief questionnaire describing their oral pain on an 11-point numeric scale ranging from 0 (no pain) to 10 (worst pain imaginable). Specifically, at each of 3 weekly visits over a 22-day period (10 scheduled visits, on Days 1, 3, 5, 8, 10, 12, 15, 17, 19, and 22), patients were asked to rate the worst mouth pain they had experienced during the previous 24 hours. The 0–10 numeric rating scale and anchors are similar to those used in the Brief Pain Inventory.12
|0||No stomatitis||No dysphagia|
|1||Painless ulcers, erythema, or mild soreness in the absence of lesions||Mild dysphagia, but patient can eat regular diet|
|2||Painful erythema, edema, or ulcers, but patient can swallow||Dysphagia that requires a predominantly pureed, soft, or liquid diet|
|3||Painful erythema, edema, or ulcers that prevent swallowing or necessitate hydration or parenteral (or enteral) nutritional support||Dysphagia that requires intravenous hydration|
|4||Severe ulceration that requires prophylactic intubation or results in documented aspiration pneumonia||Complete obstruction (patient cannot swallow saliva); ulceration with bleeding not induced by minor trauma or abrasion or perforation|
For each measure (stomatitis, dysphagia, and mouth pain), each patient's highest (peak) score was determined, and the number of days from Study Day 1 to the first occurrence of the peak score was calculated. Summary statistics (mean scores by visit, timing and intensity of peak scores) were calculated using all available data; no scores were imputed or carried forward. Reductions in peak mouth pain for the iseganan and placebo groups were calculated using the Wilcoxon rank sum test.
Three hundred twenty-three patients (191 male and 132 female; age range, 11–76 years) were enrolled between April 2000 and January 2001. Most patients were receiving stomatotoxic therapy for lymphoma, acute leukemia, chronic leukemia, or multiple myeloma. Of the 323 enrolled patients, 308 received ablative chemotherapy followed by a stem cell transplant, 13 patients received nonablative stomatotoxic chemotherapy, and 2 patients did not receive any chemotherapy. The most commonly used chemotherapy and radiotherapy regimens are listed in Table 2.
|Regimen||No. of patients|
|Total body irradiation||109|
|Methotrexate (for GVHD prophylaxis)||77|
|Melphalan, 200 mg/m2||46|
|Busulfan, ≥ 16 mg/kg||38|
|Busulfan + cyclophosphamide||28|
|Etoposide, > 800 mg/m2 cumulative dose||26|
|BCNU + etoposide + cytarabine + melphalan||21|
|Cyclophosphamide + BCNU + etoposide||15|
Patients completed 91% of the scheduled subjective assessments (i.e., the mouth pain questionnaires) through Day 22, which covered the period during which peak stomatitis and peak pain were most likely to occur (Fig. 1). Of the 98 patients who were missing mouth pain questionnaires at the end of the study (Day 22), most reported no mouth pain as their last recorded value. This finding suggests that the primary reason for missing data in the last week of the study was recovery from mucositis. Similar results were observed for the scheduled objective oral assessments of stomatitis and dysphagia; 92% of these assessments were completed through Day 22.
Oral pain was negligible at the start of the study, which coincided with the first administration of chemotherapy; peaked on Days 14–16; and subsequently subsided. Over the course of the study, average mouth pain scores increased and decreased as the average stomatitis and dysphagia scores increased and decreased (Fig. 2). Most patients (n = 185; 57%) reported their peak mouth pain within 2 days of recording their peak stomatitis. Most patients (n = 248; 77%) experienced peak mouth pain within 1 week of peak stomatitis. Similarly, most patients (n = 192; 59%) reported peak mouth pain within 2 days of recorded peak dysphagia, and 248 patients (77%) reported peak mouth pain within 1 week of peak dysphagia. Most patients experienced peaks in one measure within a day or two of their peaks in the other measures. The occurrence of peak mouth pain also followed a time course similar to the courses followed by peak stomatitis and peak dysphagia (Fig. 3). Several patients did not experience any increase in mouth pain (n = 95; 29.4%), stomatitis (n = 62; 19.2%), and/or dysphagia (n = 77; 23.8%) after the Day 1 visit; data from these patients are not included in Figure 3.
Correlations between mouth pain score and toxicity grade (for stomatitis and dysphagia) are detailed in Figure 4; the coefficients of correlation between peak mouth pain and peak stomatitis and between peak mouth pain and peak dysphagia were 0.692 and 0.667, respectively. For both toxicities, there was a monotonic rise in average pain score from Grade 0 to Grade 3 and no increase from Grade 3 to Grade 4. Variability in pain scores was minimal at the Grade 0–1 toxicity level and considerably greater at the Grade 2–4 level. Although mouth pain and NCI CTC toxicity clearly are related, the variability in pain in the higher toxicity range supports the use of pain assessment, in addition to toxicity grade evaluation alone, to gain a better understanding of the impact of stomatitis and dysphagia.
As peak mouth pain increased, the number of days of parenteral (transdermal, injected, or intravenous) narcotic analgesic use rose incrementally, from approximately 3 days when no pain was reported to almost 10 days at the worst pain level (Fig. 5). Each unit change on the pain scale reflected an approximate increase of 0.7 days of parenteral narcotic use; this finding indicated a clear relation between patient ratings of peak pain and the duration of analgesic therapy.
At all pain levels, iseganan HCl reduced peak mouth pain compared with placebo (P = 0.041) (Fig. 6). This was particularly true for patients experiencing the worst peak mouth pain. Fifteen percent of patients in the placebo group reported peak mouth pain scores of 9–10 (the worst pain imaginable), compared with only 6% of patients in the iseganan HCl group. In contrast, 13.5% of patients in the iseganan group reported peak mouth pain scores of 7–8, compared with 8.7% of patients in the placebo group. In pain categories 7–10 combined, there were 31 patients (19.1%) from the iseganan group and 38 patients (23.7%) from the placebo group. This represented in a 67% reduction in the number of patients who received the longest courses of parenteral narcotics. Iseganan may have prevented some patients from reaching the most severe pain levels (9–10) and decreased the percentage of patients with pain scores in the range of 7–10.
In the current trial, we were able to obtain both objective and subjective measures of OM in a high proportion of patients throughout the peritransplant period. Clinician assessments were completed during both the rise and peak periods of stomatitis and dysphagia, and the vast majority of patients (91%) were able to complete self-assessments of mouth pain throughout the trial. It is noteworthy that missing questionnaires at the end of the study were attributed primarily to patients whose last recorded score was 0 (no mouth pain); this finding suggests that patients who continued to experience mouth pain continued to complete self-assessments. The completeness of data capture, for both objective and subjective measures, is encouraging in light of the discomfort that patients with severe oral pain can experience when undergoing oral examinations. Assessment of patients with severe pain due to OM using scales that require an extensive oral examination can be difficult to use and may result in underreporting of OM pain levels.1 A simple, patient-reported measure may eliminate the need for an extended oral examination to obtain a measure of the course of OM. Others also have reported the feasibility of obtaining self-reported pain assessments: McGuire et al.13 used more complex questionnaires than those in the current study to assess oral pain during the peritransplant period, and de Wit et al.14 and Maunsell et al.15 used patient diaries to assess pain in outpatients with cancer.
We found that patient-reported mouth pain was a responsive measure of the patient's clinical course of OM, as evidenced by the parallel rise, peak, and fall of average mouth pain scores compared with stomatitis scores. In addition, the subjective measure of patient-reported mouth pain was correlated with objective, clinician-reported measures of OM, such as stomatitis and dysphagia. Peak mouth pain, peak stomatitis, and peak dysphagia occurred at similar times for each patient. Stomatitis and dysphagia also peaked at the same time for most patients. The observed correlations indicate that the pain scores and stomatitis and dysphagia ratings share about 50% of the variance of the data; this finding clearly suggests a strong correlation and also illustrates that the data are complementary. An earlier trial, which validated the OMAS scale, also demonstrated a close correlation between objective and subjective findings related to OM.10 Objective scoring of erythema using OMAS, as well as the presence or absence of ulcers and pseudomembranes, was correlated with subjective symptoms of pain, swallowing, and the ability to eat. Although the OMAS scale was not formally applied in the current trial, the results are consistent with those found using OMAS.6, 10 As in the current trial, patient-reported visual analog scales were used in the OMAS validation to assess subjective measures of pain and swallowing, and were found to be correlated with several measures of OM, including the mean mucositis score, the weighted mean mucositis score, the worst site score, and the extent of mucositis score. Although the specificity of a pain scale is important, it must be recognized that patients experiencing pain in multiple body sites may have difficulty discerning the exact source of pain.
The NCI CTC stomatitis scale for patients receiving chemotherapy incorporates a subjective measure of pain as it relates to the presence of objective findings of erythema, edema, and ulcers. It was not possible to extract the pain component from the stomatitis scale and compare purely subjective findings with purely objective findings. In addition, the physician assessment can include a statement of whether pain is present or absent, but it cannot evaluate the severity of pain as a subjective patient experience. These limitations do not diminish our finding that a patient-reported measure was correlated with physical changes observed by clinicians throughout the trial.
Another noteworthy observation was the number of patients reporting a pain score of 10 (worst pain imaginable) at least once during the study. There are several possible explanations: providers may fail to take patient complaints of pain seriously; providers may fail to provide an adequate level of analgesic medication due to concerns regarding adverse effects; patients may underreport pain in an effort to please the provider; patients may dislike the adverse effects associated with analgesics so much that they would rather experience pain; or the pain associated with oral mucositis may be intractable and not satisfactorily manageable with narcotic analgesics. It is likely that a combination of these factors is responsible for the severe oral pain that patients report.
As a subjective measure of OM, patient-reported mouth pain was sensitive to the effects of an intervention, namely iseganan HCl. The duration of treatment with parenteral narcotics, which is an objective indicator of pain, was correlated with changes in peak oral pain. Specifically, a 1-unit change in peak mouth pain was correlated with an additional 0.7 days of parenteral narcotic use. This finding suggests that reductions in pain should lead to a decrease in parenteral narcotic use. Parenteral narcotic use was significantly associated with peak OM score; a 1-point increase in OM score resulted in almost 3 additional days of narcotic use.6 We note that although patients receiving iseganan HCl were less likely to report peak mouth pain scores of 9–10 than were patients receiving placebo, they were more likely to report peak mouth pain scores of 7–8. This finding may reflect the ability of iseganan HCl to reduce the severity of peak mouth pain from the highest level to a lower level, rather than to eliminate pain completely. It remains to be seen whether this observation will be repeated in future studies.
In conclusion, the results of the current study suggest that an independent assessment of pain by the patient is not only useful for capturing the patient's perspective on his or her condition but also complementary to the physician's objective assessment of the disease. It is noteworthy that objective oral assessments could be obtained for most patients during the rise and peak periods of OM-related pain. The current study showed that mouth pain rose and fell predictably during the course of OM and that mouth pain was closely related to objective measures of oral disease. Patients receiving iseganan had significantly less mucositis and associated pain. A confirmatory trial evaluating iseganan in patients receiving stomatotoxic chemotherapy is underway. The results of the current study support the use of a simple, patient-reported rating of mouth pain as a highly relevant and responsive endpoint in clinical trials. This rating system also may provide a straightforward method of following OM in clinical practice.