- © 1999 by American Society of Clinical Oncology
Insufficient Effectiveness of 5-Hydroxytryptamine-3 Receptor Antagonists Due to Oral Morphine Administration in Patients With Cisplatin-Induced Emesis
- Akira Shoji,
- Mariko Toda,
- Kaneo Suzuki,
- Hiroshi Takahashi,
- Kenichi Takahashi,
- Yasuhiro Yoshiike,
- Takashi Ogura,
- Yuji Watanuki,
- Harumi Nishiyama and
- Shigeki Odagiri
- From the Division of Respiratory Disease, Kanagawa Prefectural Cardiovascular and Respiratory Center, Kanagawa, Japan.
- Address reprint requests to Akira Shoji, MD, Division of Respiratory Disease, Kanagawa Prefectural Cardiovascular and Respiratory Center, 6-16-1 Tomioka-higashi, Kanazawa-ku, Yokohama, Kanagawa 236-0051, Japan
Abstract
PURPOSE: To compare the effect of 5-hydroxytryptamine-3 (5HT3) receptor antagonists in cancer patients receiving chemotherapy including cisplatin (CDDP), with or without sustained-release oral morphine (MS Contin; Shionogi Co, Osaka, Japan).
PATIENTS AND METHODS: We retrospectively studied 58 lung cancer patients given chemotherapy including at least 50 mg/m2 CDDP with 5-HT3 receptor antagonists between January 1996 and December 1997. Number of vomiting episodes, average porportions of hospital-supplied meals consumed (0 to 100%, as an index of appetite), and nausea severity scores (0 to 2 points, subjective patient judgment) were compared between oral morphine–administered (+) and morphine-free (−) groups.
RESULTS: Sixteen morphine(+) and 42 morphine(−) cases were used. In cases of acute emesis (within 24 hours after CDDP injection), morphine(+) and morphine(−) groups were significantly different in number of vomiting episodes (1.25 and 0.14, respectively; P < .0001), appetite (58.13% and 90.24%; P < .0001), and nausea severity scores (1.63 and 0.62; P < .0001). In delayed-emesis cases (24 to 120 hours after CDDP), these groups differed significantly in number of vomiting episodes (1.94 and 0.43, respectively; P = .0001), appetite (23.13% and 52.08%; P < .0001), and nausea severity (1.38 and 0.91; P = .009). There were no significant differences in sex, age, anticancer drugs concurrent with CDDP, CDDP dose, corticosteroid administration, clinical stage, or type of 5-HT3 antagonist.
CONCLUSIONS: These data suggest that morphine can markedly reduce the effectiveness of 5-HT3 receptor antagonists in patients receiving chemotherapy that includes CDDP. These results require confirmation by reinvestigation of clinical data on the efficacy of 5-HT3 receptor antagonists and by extensive prospective analyses.
PATIENTS WITH ADVANCED lung cancer have been administered radiation therapy, chemotherapy, or both. Of the many drugs available, mainly the platinum agents have been used, but these drugs have many side effects, such as neutropenia, nausea, and kidney dysfunction, to name a few.
The American Society of Clinical Oncology issued guidelines for the treatment of advanced non–small-cell lung cancer.1 In this document, it was pointed out that (1) chemotherapy for advanced non–small-cell lung cancer should include the platinum formulations and (2) patients with a poor performance status (PS) should not be given chemotherapy. It is generally known that PS is an independent prognostic factor for lung cancer.2 Thus, a course of chemotherapy must avoid aggravating PS, because patients lose their tolerance to chemotherapy and lose their desire to be cured of the cancer. Nausea and vomiting are disagreeable side effects of anticancer drugs that act counterproductively in regard to their purpose. Many agents, especially cisplatin (CDDP), cause strong emesis.3,4
Recently developed 5-hydroxytryptamine-3 (5-HT3) receptor antagonists have been administered to many cancer patients receiving chemotherapy and have enabled the majority of these patients to escape severe nausea and vomiting. There are now two major problems among the gastrointestinal side effects induced by chemotherapy. The first problem is delayed emesis occurring in almost all patients about 2 to 7 days after CDDP injection and generally reduced appetite. The other problem is lack of effectiveness of 5-HT3 antagonists against acute emesis caused by chemotherapy; it occurs within about 24 hours after CDDP injection, with some patients developing severe nausea and vomiting despite administration of 5-HT3 receptor antagonists. It has been reported that the frequency of ineffective cases is about 20%.5-8
On the other hand, morphine is necessary for the control of pain in cancer patients. It has generally been used in patients with pain induced by, for example, bone metastasis and direct invasion of bones or the chest wall. The World Health Organization has suggested that there is no fixed maximum dose of morphine for cancer patients and that morphine should be administered in sufficient amounts to remove the pain of the disease.9
We found out incidentally that the effects of 5-HT3 antagonists may be reduced in patients given oral morphine for emesis induced by CDDP-based chemotherapy. It is, of course, generally known that morphine sulfate alone induces nausea and vomiting in the early phase of its administration and that these side effects disappear within about 2 weeks. However, we observed that 5-HT3 receptor antagonists are ineffective even in patients who do not have nausea and vomiting in the early phase or who stop vomiting after 2 weeks of oral morphine administration. In this study, we examined retrospectively the effect of morphine sulfate on the action of 5-HT3 receptor antagonists in patients treated with chemotherapy including CDDP.
PATIENTS AND METHODS
Patients
We selected patients with advanced lung cancer who were given chemotherapy in our hospital from January 1996 to December 1997. Of these, we selected those who were treated with combination chemotherapy including CDDP (≥ 50 mg/m2/d) together with 5-HT3 receptor antagonists. From these patients, we excluded those with symptomatic brain metastasis or abdominal symptoms such as nausea, vomiting, and ileus for any reason before chemotherapy. Because morphine induces nausea and vomiting early during its administration, we excluded those patients who had experienced such side effects after receiving oral morphine before chemotherapy. The degree of nausea and vomiting induced by chemotherapy in the selected cases was investigated.
Efficacy Assessment
In our hospital, pulse rate, blood pressure, and body temperature are routinely recorded as part of the patient's record. In addition to these parameters, in lung cancer patients receiving chemotherapy, urine volume, frequency and character of stool, food consumption, number of vomiting episodes, and nausea scale according to the patient's subjective judgment (none, slight, severe) are recorded daily. Using these records, we assessed the effectiveness of 5-HT3 receptor antagonists in the following three categories: the number of vomiting episodes, the mean proportion of hospital meals consumed (0 to 100%) (appetite), and the nausea severity score classified by stage (severe, 2 points; slight, 1 point; none, 0 points) according to the patient's subjective judgment. We used the worst of the daily nausea severity scores in the periods defined as follows: the early phase was within 24 hours after CDDP injection (acute emesis) and the late phase was from 24 to 120 hours after CDDP injection (delayed emesis).
Variables
We examined the severity scores according to whether oral administration of morphine (MS Contin; Shionogi Co, Osaka, Japan) was given and by other categories that may have an effect on the emesis induced by chemotherapy, namely, sex, age, types of anticancer drugs combined with CDDP, dosage of CDDP, administration of corticosteroids, clinical stage (using the International Tumor-Node-Metastasis Staging System), and type of 5-HT3 receptor antagonist (granisetron [Kytril; SmithKline Beecham Pharmaceuticals, Tokyo, Japan], ondansetron [Zofran; Glaxo Co, Tokyo, Japan], azasetron [Serotone; Yoshitomi Pharmaceutical Co, Fukuoka, Japan], or DAU 6215 [Boehringer Ingelheim Co, Tokyo, Japan].
Statistical Methods
All analyses were conducted using StatView for Macintosh Version 4.5 (Abacus Concepts, Inc, Berkeley, CA). Data are expressed as mean ± SD and were analyzed using the independent t test, one-way analysis of variance, and Scheffe's multiple comparison test. A P value of less than .05 was considered statistically significant.
RESULTS
Acute Emesis
In our hospital, from January 1996 to December 1997, 63 patients received chemotherapy that included 50 mg/m2 or more of CDDP. Five of these patients were excluded from the study because they had abdominal symptoms such as nausea, vomiting, and ileus of various causes at the start of chemotherapy. The remaining 58 patients were studied. Table 1 shows the results of acute emesis induced by CDDP injection classified according to whether morphine was administered (morphine(+), 16 cases) or not (morphine(−), 42 cases) as well as by age, 5-HT3 receptor antagonist, use or nonuse of corticosteroids, CDDP dosage, types of antitumor drug given concurrently with CDDP, and other variables. In acute emesis, the numbers of vomiting episodes were 1.25 ± 1.06 and 0.14 ± 0.42, the estimates of appetite were 58.13% ± 29.71% and 90.24% ± 16.89%, and the nausea severity scores were 1.63 ± 0.62 and 0.62 ± 0.62 in the morphine(+) and morphine(−) groups, respectively. These three comparisons revealed statistically significant differences (P < .0001), but no such differences were found with the other variables.
CDDP-Induced Acute Emesis (within 0 to 24 hours after injection) by Clinical Characteristics and Treatment History
We examined the effects of dosage and duration of administration of oral morphine on the same three variables. There was no significant difference in these variables between the high-dose (≥ 60 mg/d; n = 5) and low-dose (< 60 mg/d; n = 11) administration groups (data not shown). The patients given oral morphine were placed in two groups: one consisted of the patients given oral morphine for 14 days or more at the time of CDDP injection (n = 8) and the other of patients who received morphine for less than 14 days (n = 8). This examination of administration periods revealed no significant effect on the variables (data not shown).
Delayed Emesis
Table 2 shows the results for delayed emesis and is classified in the same way as Table 1. We found significant differences in the numbers of episodes of vomiting (1.94 ± 0.43 and 0.43 ± 1.04; P = .0001), the estimates of appetite (23.13% ± 18.20% and 52.08% ± 24.89%; P < .0001), and the nausea severity scores (1.38 ± 0.50 and 0.91 ± 0.62; P = .009) between the morphine(+) and morphine(−) groups (respectively). In the present study, we detected no effect of corticosteroids on delayed emesis.
CDDP-Induced Delayed Emesis (from 24 to 120 hours after injection) by Clinical Characteristics and Treatment History
We examined the effects of dosage and duration of administration of oral morphine on the same three variables as studied in acute emesis. There was no significant difference in the these variables between high- and low-dose or between long and short administration period groups (data not shown).
DISCUSSION
This study revealed that the effects of 5-HT3 receptor antagonists on patients given oral morphine were reduced. We encountered four patients who received two or more courses of the same regimen of chemotherapy with and without oral morphine because the pain disappeared on treatment or appeared despite the treatment. They had severe emesis only after chemotherapy with oral morphine. One of the major risk factors for severe acute emesis, despite the administration of 5-HT3 receptor antagonists, may be the administration of this drug. Of course, a few patients with severe emesis were not given oral morphine. Therefore, administration of this drug is not the only reason for insufficient 5-HT3 receptor antagonist effectiveness.
Before CDDP injection, the patients in this study were given the Japanese addiction doses of 5-HT3 receptor antagonists, ie, granisetron 3 mg, ondansetron 4 mg, azasetron 10 mg, or DAU-6215 5 or 10 mg. If the patients failed to have adequate control of nausea and vomiting, the same kind of 5-HT3 receptor antagonist, with or without metocloplamide, domperidone, and chlorpromazine, was administered. Therefore, there were no differences between the morphine(−) and morphine(+) groups in the dosage of 5-HT3 receptor antagonist administered before chemotherapy. The total dosage of 5-HT3 receptor antagonists and other antiemetics administered during the chemotherapy course were greater in the morphine(+) group than in the morphine(−) group.
Although it has been reported that some characteristics of patients, such as age and female sex, may be risk factors for insufficient 5-HT3 receptor antagonist effectiveness, and although the administration of corticosteroid reduces delayed emesis,10-12 in the present study, there were no significant differences in acute or delayed emesis in relation to other factors, such as age, sex, corticosteroid administration, type of 5-HT3 receptor antagonist, and so on.
Because of the high risk of emesis in the patients given oral morphine, we have hypothesized two possible mechanisms. The first is that 5-HT3 receptor antagonists simply become unable to inhibit the sum total of stimuli to the vomiting center (VC) that are due to the side effects of oral morphine and CDDP. We realized, on the basis of our observations of patients who suffered severe emesis despite the administration of 5-HT3 receptor antagonists, that such episodes occurred mainly in circumstances such as the following: when the patients made head or body movements, for example, while eating, brushing teeth, or using the toilet, even though not the slightest feeling of nausea had been present before those actions were made. This suggests that the mechanisms of these emetic episodes were activated not only by stimuli via the chemoreceptor trigger zone but also via direct stimulation of the VC from the vestibular nuclei and by other direct stimuli to the VC caused by elevation of the intragastric pressure by the gastric contents and carried through the vagus and greater splanchnic nerve fibers. Generally speaking, direct VC stimulation from the vestibular nuclei is more frequently seen as a side effect of morphine sulfate, whereas the direct VC stimulation that is due to increased intragastric pressure and travels through the vagus and greater splanchnic nerve fibers is more often observed as a result of the side effects of both CDDP and morphine that cause reduction of peristalsis in the digestive tract.13-16 On the basis of these facts, we propose the following hypothesis regarding this vomiting: we suggest that, even if the patient experiences no nausea, because of the oral administration of morphine, the stimuli to the VC approach the vomiting threshold latently, and then both the administration of CDDP during treatment with 5-HT3 receptor antagonists and the additional slight stimulation afforded by the processes just described result inthe vomiting threshold being reached, so that the patient ultimately vomits.
The second possible mechanism is that changes of pharmacokinetics caused by drug-drug interactions probably induce severe nausea and vomiting. The serum concentration of CDDP or morphine may be increased or that of 5-HT3 receptor antagonists may be decreased.
Administration of oral morphine is a major latent risk factor in insufficient 5-HT3 receptor antagonist effectiveness in patients treated with chemotherapy that includes CDDP. It is necessary to clarify this mechanism. It was also reported that 5-HT3 receptor antagonists are ineffective in about 20% of patients treated with CDDP. However, until the present study, no published reports indicated the need for restriction of the use of oral morphine or mentioned any differences in the effects of these drugs between patients given and not given oral morphine. It is necessary, therefore, to reinvestigate the clinically evaluated data on the efficacy of 5-HT3 receptor antagonists.
A large prospective analysis is needed to confirm the results. It is, moreover, necessary to clarify this mechanism both by a pharmacokinetic examination of these drugs and by clinical trials of antihistamine drugs or gastrointestinal peristalsis promoters, or both, combined with 5-HT3 receptor antagonists.
- Received September 4, 1998.
- Accepted February 18, 1999.
