Effect of topical morphine for mucositis-associated pain following concomitant chemoradiotherapy for head and neck carcinoma
Oral mucositis is the dose-limiting toxicity for patients receiving concurrent chemoradiotherapy regimens for tumors of the head and neck area. Currently, the management of established mucositis includes the use of topical anesthetics and systemic analgesics. Based on the clinical evidence of pain alleviation by topical morphine in patients with some inflammatory and painful conditions, a clinical study was undertaken to determine this effect on mucositis-associated pain.
Twenty-six patients with head and neck malignancies treated with concomitant chemoradiotherapy for head and neck carcinoma who had severe painful mucositis (World Health Organization Grade 2 or higher) were enrolled. Patients were randomly assigned to morphine mouthwash (MO; 14 patients) or magic mouthwash (MG), a mixture of equal parts of lidocaine, diphenhydramine, and magnesium aluminum hydroxide (12 patients).
The duration of severe pain was 3.5 days less in the MO group compared with the MG group (P = 0.032). The intensity of oral pain was also significantly lower in the MO group compared with the MG group (P = 0.038). No patient in the MO group required third-step opiates for alleviation of the mouth pain. There was a significant difference in duration of severe functional impairment (P = 0.017). Five patients in the MG group complained of local side effects and only one in the MO group (P = 0.007).
For patients with head and neck carcinomas receiving concomitant chemoradiotherapy, MO is a simple and effective treatment to decrease the severity and duration of pain and the duration of functional impairment. Cancer 2002;95:2230–6. © 2002 American Cancer Society.
Oral mucositis is a frequent and potentially severe complication of cancer therapy. It is painful, impairs adequate nutritional and liquid intake, affects the quality of life, and may result in serious clinical complications. Oral mucositis is the dose-limiting toxicity for patients receiving concurrent chemoradiotherapy regimens for tumors of the head and neck area and for patients receiving 5-fluorouracil (5-FU)-based therapies, especially when administered by continuous infusion.1, 2
Cisplatin and 5-FU remain the most used and active regimen in previously untreated patients with head and neck carcinoma.3, 4 Administering multiple cytotoxic drugs during radiation substantially increases toxicity and leads to interruptions of radiotherapy. The advantages of multiagent chemotherapy plus radiotherapy are improved locoregional control, fewer distant metastases, and longer periods of survival.5–11 The combination of cisplatin and 5-FU with concurrent radiation is currently the most used regimen for patients with locally advanced, unresectable disease.12–15 The majority of patients who receive this type of concurrent chemoradiotherapy develop painful mucositis that may result in dose reductions in subsequent cycles or radiotherapy interruptions, which may ultimately affect patients' response to therapy.16, 17
Currently, no intervention exists that is completely successful at preventing oral mucositis and the management of established mucositis can be difficult for both the patient and health provider. General approaches include effective oral care, dietary modifications, and topical mucosal protectants. The appropriate use of topical anesthetics and systemic analgesics remains the cornerstone of therapy.18 Despite the lack of definitive clinical data to support or refute the routine use of mucosal coating agents with analgesic properties, they are widely used.2, 19 Magic mouthwash (MG), a mixture of equal parts of magnesium aluminum hydroxide, diphenhydramine, and 2% viscous lidocaine, is used to relieve oral pain.19, 20
When oral mucositis is severe and interferes with nutritional intake and quality of life, it is appropriate to use oral or, if necessary, parenteral narcotics. Effective use of systemic opioid drugs requires balancing pain relief and the undesirable side effects of nausea, vomiting, mental clouding, constipation, sedation, tolerance, and physical dependence. The opioid analgesics, of which morphine is the prototype, produce their analgesic effects by binding to opiate receptors in the central nervous system and the peripheral terminals of afferent nerves.21–23 Basic and clinical research suggest that the analgesic effects of exogenous opioids applied locally are particularly prominent in patients with painful inflammatory conditions.24–27
We designed this study to compare the effect of locally applied morphine (morphine mouthwash [MO]) with MG on mucositis-related oral pain and on the maintenance of oral intake in patients with tumors of the head and neck area treated with a chemoradiotherapy regimen. Additional objectives were to evaluate the safety of MO by determining the frequency of treatment-emergent drug-related adverse events (local and systemic) or hematologic and biochemical abnormalities, the intensity of chemoradiotherapy administered, tumor response, weight loss, need of a nasogastric tube (NGT), and mucositis-related hospitalizations.
MATERIALS AND METHODS
We conducted a unicenter, randomized, controlled, parallel comparison between MO and MG. Patients were assigned randomly to either the MO or MG study group according to a computer-generated randomization scheme (in a 1:1 ratio in blocks of 10).
Patients were required to be 18 years of age or older. Other criteria included histologically documented diagnosis of squamous carcinoma of the head and neck region, clinically unresectable locoregional advanced head and neck carcinoma, treatment with chemoradiotherapy affecting part of the oral mucosa, a Karnofsky performance score of at least 60%, a World Health Organization (WHO) rating of global mucositis of Grade 2 or higher,28 and a Mini-Mental Status Exam score of at least 24/30. Patients provided written informed consent.
Patients were excluded if they had undergone previous surgery for head and neck carcinoma and if there was a risk of patients swallowing the mouthwash solution, determined by a recovery of less than 90% of the 15 mL of normal saline used as a tester. Severe renal (creatinine clearance < 30 mL/min) and hepatic insufficiency (bilirubin level > 2 mg/dL, ascitis, albumin level < 3.5 g/dL, and/or prothrombin time < 50%) were also exclusion criteria. In addition, pregnant women were ineligible for the study.
Twenty-six patients were enrolled in the study between October 2000 and July 2001. All patients had a dental and oral examination before treatment. After initiation of chemotherapy, they were checked weekly until the initiation of radiotherapy when they were checked daily for painful mucositis. Patients were followed up until there was total alleviation of mouth pain (numeric rating scale [NRS] ≤ 1 and a rating of low or none on the verbal scale [VS]) and recovery of oral function (functional impairment scale [FIS] = 1 or 2).
The MO group used an oral rinse of 15 mL 2% morphine solution (2000 mg morphine clorhydrate diluted in 1000 mL of water) every 3 hours six times a day. The morphine solution was prepared by the hospital pharmacy. The MG group used an oral rinse (a mixture of 5 mL magnesium aluminum hydroxide [Mylanta, Parke-Davis SA, Buenos Aires, Argentina], 5 mL 2% viscous lidocaine [Xylocaina Viscosa, Astra SA, Buenos Aires, Argentina], and 5 mL [12.5 mg] diphenhydramine [Benadryl, Parke-Davis SA]), every 3 hours six times a day.
Patients in both groups were instructed not to swallow during the rinses and to hold the mouthwash for 2 minutes. Both groups were subjected to the same general procedures (removal of dentures, debridement of necrotic tissues by a dental professional, and oral rinsing with normal saline as needed) and were ordered to follow the same dietary guidelines. No patients received steroids and/or antimicrobials before inclusion.
Chemotherapy consisted of two inpatient cycles of cisplatin 100 mg/m2 (3-hour infusions on Days 1 and 21) plus a continuous infusion of 5-FU 1000 mg/m2 (Days 1–5 and Days 21–25), followed by a concurrent outpatient regimen of chemoradiotherapy. Radiotherapy started on Day 28 of the treatment schedule. All patients were irradiated with 60Co beams using two parallel opposed fields to treat the primary tumor, involving lymph nodes, and the relevant areas of lymphatic drainage. A total dose of 70 Gy over 5 weeks was administered, with an individual dose of 1.5 Gy two times a day from Monday to Friday with a 6-hour interval between treatments. The uninvolved areas received a total dose of 50 Gy (one daily fraction of 2 Gy) with one anterior appositional field. The spinal cord dose was limited to 45 Gy including the scattered radiation contribution after exclusion. Concomitant chemotherapy consisted of a weekly administration (Days 35, 42, 49, and 56) of a 1-hour infusion of cisplatin 30 mg/m2 followed by a 1-hour infusion of 5-FU 300 mg/m2.
We assessed the severity of oral pain daily by using the NRS (0–10) and the VS (worst possible, severe, moderate, low, and none) and the severity of dysphagia was assessed daily using the FIS (1 = able to eat solid foods, 2 = able to eat soft foods, 3 = able to eat liquids, and 4 = oral alimentation not possible). Supplemental analgesic medications were prescribed according to the WHO analgesic ladder. The first step was 500 mg acetaminophen four times a day up to 3000 mg per day. The second step was the first step plus 50 mg tramadol three times a day up to 600 mg per day. The third step was the first step plus oral or parenteral morphine clorhydrate as needed. Local infections were also monitored. Patients with documented or high clinical suspicion of mycotic and/or viral infection were treated with nystatin rinses (changing to oral or parenteral fluconazole if nystatin failed) and/or acyclovir (oral or parenteral), respectively. Bacterial infections were treated according to the culture and the sensitivity test results with oral or parenteral antibiotics. We evaluated the frequency and severity (assessed by several NRS) of treatment-emergent drug-related (probably, or certainly related) local (loss of taste, viscous saliva, dry mouth, excessive anesthesia, itching, burning, nausea and others) and systemic side effects (constipation, nausea, vomiting, somnolence, miosis, respiratory depression, and others). We assessed patients' need for an NGT or similar hydration-nutrition device. Hospitalizations due to mucositis-related complications were recorded. Weight changes were monitored by comparing weights at the beginning and at the end of the oncologic treatment. Intensity of the chemoradiotherapy was recorded as the percentage of the theoretical-proposed intensity regimen. Tumor response was determined by the investigator according to the Eastern Cooperative Oncology Group (ECOG) criteria.29 Performance status (ECOG scale) was compared at the beginning and at the end of oncologic treatment. Clinical laboratory parameters were monitored carefully (at the beginning and at the end of each treatment course), including white blood count and levels of granulocytes, lymphocytes, platelets, albumin, urea, creatinine clearance, potassium, sodium, magnesium, ionic calcium, alanine and aspartate aminotransaminase, alkaline phosphatase, and bilirubin.
The principal end points were the duration (number of days) of severe pain (NRS > 5 and/or VS worst possible or severe), the intensity of pain (the median of the three highest NRS scores for each patient [M3H-NRS], and the duration (number of days) of severe functional impairment (FIS = 3 or 4). Addition end points were time elapsed (in days) before the first supplemental analgesic and total amount taken (mg per day). Patients who required the second and third-steps of the analgesic ladder were excluded for the analysis of systemic side effects because they were premedicated with laxatives (during all the treatment) and metoclopramide (first five days of the treatment).
Data are expressed as the mean with a 95% confidence interval, or as the median with either median absolute deviation (MAD), or interquartile range. Demographic data were analyzed using the chi-square test. The Wilcoxon's rank sum test was used to compare the duration of severe pain, duration of functional impairment, and intensity of pain. Other parametric and nonparametric tests were used to compare the other variables.30
All P values cited were two sided and P values less than 0.05 were judged as statistically significant. All calculations were done with the statistics program Statistix 7.0 (Analytical Software2000, Tallahasse, FL).
Table 1 shows the characteristics of the patients included. The median (MAD) radiation dose delivered at the time of enrollment was 39 (7) Gy for the MO group and 38 (6) Gy for the MG group. Other characteristics were also similar between both branches. None of the patients were neutropenic at randomization or during the study.
Table 1. Characteristics of the Patients at Randomization
|Mean age (range)||58.85 (44–80)||52.83 (27–77)||NS|
|Gender|| || || |
|Tumor localization (%)|| || || |
| Oral cavity||5 (35.7)||5 (41.6)||NS|
| Nasopharynx||2 (14.3)||1 (8.4)||NS|
| Oropharynx||6 (42.8)||6 (50)||NS|
| Paranasal sinuses||1 (7.2)||0||NS|
|Stage distribution (%)31|| || || |
| E III||6 (42.8)||6 (50)|| |
| E IV||8 (57.2)||6 (50)||NS|
|Mucositis grade (%)|| || || |
| 2||4 (28.6)||6 (50)|| |
| 3||10 (71.4)||6 (50)||NS|
|ECOG performance status (%)|| || || |
| 1||8 (57)||8 (67)|| |
| 2||6 (43)||4 (33)||NS|
|Median radiation dose delivered at the time of enrolled (MAD)||39 (7) Gy||38 (6) Gy||NS|
The duration of severe pain (NRS ≥ 5 and/or VS severe or worst possible) was 3.5 days less in the MO group compared with the MG group (P = 0.032). The intensity of oral pain (M3H-NRS) was also significantly lower in the MO group compared with the MG group (6 vs. 7.5; Table 2). More patients in the MG group needed supplementary (oral or parenteral) analgesia compared with the MO group (P = 0.019). Nevertheless, the time elapsed before the first supplemental analgesic and the total amount of analgesics taken was similar for both groups. Of 12 patients in the MG group, 3 (25%) and none in the MO group required third-step opiates for alleviation of mouth pain. However, the differences in the maximum WHO step needed for control of pain were not statistically significant (Table 3).
Table 2. Primary End Points after the Treatment
|Duration of severe functionalimpairment (days mean ± SD)||1.85 ± 1.53||7.67 ± 4.3||0.017|
|Duration of severe pain (days mean ± SD)||5.07 ± 1.79||8.58 ± 2.65||0.032|
|M3H-NRS (1st–3rd quartile)||6 (5–6.75)||7.5 (6.16–8.83)||0.038|
Table 3. Secondary End Points after Treatment
|Need supplementary analgesia (no. of patients)||3 (21%)||8 (67%)||0.019|
|Time elapsed before first supp. analgesia (mean days)||2.34||2.28||NS|
|Maximum gWHO step in analgesic ladder (%)|| || || |
| Step 1||1 (7)||0|| |
| Step 2||2 (14)||4 (34)|| |
| Step 3||0||3 (25)||0.088|
|Document Infections (%)|| || || |
| HSV||2||2|| |
| Candida sp.||5||6|| |
| Total||7 (50)||8 (67)||NS|
|Body weight change (kg)||+0.2||−0.3||NS|
|Intensity of treatment (patients with intensity > 90%)||14 (100%)||12 (100%)||NS|
|Local side effects (probably, or certainly related)||1 (7%)||5 (41.6%)||0.007|
There was a significant difference in duration of severe functional impairment (Table 2). Nevertheless, the body weight change was similar for both groups. There were no significant differences in documented or highly suspected infections, change in performance status, tumor response rate, and intensity of the chemoradiotherapy delivered between the two treatment groups (Table 3). No patients required hospitalization due to mucositis during the study.
Patients in the MG group reported more local side effects (Table 3). Of the 12 patients in this group, 5 reported one or more symptoms (nausea, 1 patient; loss of taste, 3 patient; viscous saliva, 2 patients; dry mouth, 2 patients; excessive anesthesia, 1 patient). One patient in the MO group experienced dry mouth and a burning sensation that required topical diphenhydramine for 1 day.
None of the patients inadvertently swallowed the mouthwash. The adverse effects of the treatments and the taste of the treatment solutions were not deemed sufficiently unpleasant to cause inadvertent drug expectoration. No systemic clinically relevant adverse effects (signs, symptoms, and/or changes in clinical laboratory parameters) were noted in either of the two arms of the study.
The biologic complexities associated with chemotherapy32 and radiation-induced33 mucosal damage have only recently been delineated. The cancer therapy-mediated injury to the mucosa is attributed to the apoptotic effects on basal epithelial34 and endothelial cells of the microvasculature.33 Oral mucositis is a multiple-phase model32 in which inflammation plays a key role. Shortly after the administration of radiation or chemotherapy, cytokines (interleukin [IL]-1, IL-6, tumor necrosis factor-α, and transforming growth factor β3) are released from the mucosal tissue, causing local tissue damage and leading to ulcer formation and pain. For several reasons, the inflamed and damaged tissue may constitute a unique clinical model for the action of locally applied exogenous opioids. First, although peripheral opioid receptors are not readily detectable in normal tissue, they appear minutes to hours after the initiation of an inflammatory reaction21, 26, 35 and the analgesic efficacy of peripheral opioids increases linearly with the duration of inflammation.36 Second, opioid agonists (endogenous and exogenous locally applied) have easier access to neuronal opioid receptors during inflammation because inflammation disrupts the perineurium35 and the number of peripheral sensory-nerve terminals is increased in inflamed tissues (sprouting).37 Finally, qualitative and quantitative changes in saliva and oral flora following radiation therapy may lead to a decrease in oral pH from a normal level of 7.0 to 5.5 or less.19 A low pH level increases the efficacy of opioid agonists in vitro by augmenting the interaction of opioid receptors with G-proteins in neuronal membranes.38 In addition, mucositis-associated inflammation (a long-lasting stimulus) results in a typical dull, aching pain, which is more susceptible to amelioration by local opioids than sharp pain.27 For these reasons we applied local opioids to treat mucositis-associated pain.
We used morphine because the opioid ligands with a preference for μ-receptors are generally most potent when applied locally21 and because morphine has a predictable low trans-mucose absorption. The trans-mucose and sublingual absorptions of drugs are dependent on both pH level and lipid solubility.39, 40 Therefore, the less lipophilic opioids (morphine partition coefficient = 0.00001) and opioids ionized at the low pH level of the oral cavity (morphine pKa at 37 °C = 7.9) are poorly absorbed.41–43 Absorption also depends on the concentration and contact time. Weinberg et al.42 estimated that sublingual morphine at high concentrations (5 mg/mL) and a relatively long contact time (10 minutes) has a bioavailability of 9%. The same authors demonstrated a tissue reservoir formation for opioids in the oral cavity. For at least one opioid (methadone), extraction is difficult and back-diffusion insignificant once the drug has traversed the mucosa. In a previous workstudy, we measured the serum levels of morphine in a similar group of patients with head and neck carcinoma receiving morphine mouthwashes (2 mg/mL) under equivalent conditions (2-minute rinses for patients with damaged oral mucosa). No detectable concentrations of systemically active morphine were found in circulation (< 10 ng/mL). We can assume that the dose of morphine administered with the mouthwash in the current study was systemically inactive.
The availability of hematopoietic growth factors to reduce the duration and severity of myelosuppression associated with cancer therapy has dramatically increased the importance of oral mucositis as a significant side effect. In turn, cancer disease remission and cure rates might be enhanced if oncologists used more intensive therapies without the limitations of oral mucosal injury and associated impaired patient compliance with the regimens. Although a variety of new approaches to oral mucositis have been taken, a single efficacious intervention or agent for the prophylaxis or management of radiotherapy or chemotherapy-induced mucositis has not yet been identified.44 Adequate oral hygiene, treatment of established infectious complications, and appropriate use of topical anesthetics and systemic analgesics remain the cornerstone of therapy.18, 44 Despite the lack of definitive clinical data to support or refute the routine use of mucosal-coating agents with analgesic properties, they are being used widely.2, 19 In the current study, we showed that one of the anesthetic cocktails (MG) relieved oral pain, but the relief was short-lived, started or worsened oral discomfort (viscous saliva, dry mouth), and prevented taste perception, which can interfere with food intake.
Our results indicate that for patients with head and neck carcinomas receiving concomitant chemoradiotherapy, MO deserves further study. Two of the major advantages of local morphine administration (MO) are the simplicity and low incidence of side effects compared with MG. We are currently performing additional studies to determine the most effective schedule, dose, and possible associations for MO.
The authors thank Mrs. Elsa Gomez for her generous assistance, the staff of the Department of Pharmacy, and Dr. Maria E. Cabalar for her valuable help and continuous support.