| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Twenty-three patients treated with intracerebroventricular (ICV) morphine in this study not only obtained excellent pain relief without rapid increases in dose, but also experienced a reduction in morphine-related side effects. By 24 h after initiation of ICV morphine, the mean trough cerebrospinal fluid (CSF) morphine concentration (approximately 20 µM) was 50-fold higher than the baseline concentration (approximately 0.4 µM), and the CSF concentration of morphine-6-glucuronide (M6G) was undetectable (<0.01 µM). The mean CSF concentration of morphine-3-glucuronide (M3G) decreased 90%, from a baseline concentration of 1 µM to 0.1 µM by Day 7 postventriculostomy. Thereafter, the mean trough CSF M3G concentration remained relatively constant while ICV morphine was continued, although the concomitant M3G plasma concentrations were undetectable (<0.01 µM). The large increase in the CSF morphine concentration in patients receiving ICV morphine strongly suggests that increased CSF morphine levels are unlikely to be the primary cause of analgesic tolerance or undesirable excitatory side effects (hyperalgesia, myoclonus, seizures) experienced by some patients receiving chronic large-dose systemic morphine. Implications: After initiation of intracerebroventricular morphine, cancer patients experienced excellent pain relief. Although the mean morphine concentration in cerebrospinal fluid increased 50-fold relative to preventriculostomy levels, rapid dose increases did not occur, which suggests that increased cerebrospinal fluid morphine levels are unlikely to be the main cause of analgesic tolerance.
In the last 15 yr, several studies (15) have shown that intracerebroventricular (ICV) morphine can relieve chronic cancer pain when this is incompletely responsive to systemic morphine or when pain relief cannot be achieved without unacceptable morphine-related side effects. Any explanation of the success of ICV morphine administration must take into account the two major pharmacologically active metabolites of morphine: morphine-3-glucuronide (M3G) and morphine-6-glucuronide (M6G) (6). M6G is a more potent analgesic than morphine, with the apparent potency (2- to 800-fold) depending on the species, the route of administration, and the method of assessing pain relief (6). By contrast, M3G is not analgesically active (6); however, when it is administered directly into the central nervous system (CNS) of rodents, it produces a range of dose-dependent excitatory behaviors, including allodynia, hyperalgesia, excessive grooming, "wet-dog shakes," and myoclonus (79), reminiscent of excitatory side effects observed in patients receiving large-dose morphine for chronic cancer pain (10). Furthermore, M3G administered by the ICV, (8,11,12) but not the intrathecal, route (13,14) attenuates the antinociceptive effects of ICV morphine and M6G in rats. These findings suggest that the pain relief and side effects experienced by patients who chronically receive systemic morphine reflect, at least in part, the analgesic properties of morphine and M6G and the excitatory and anti-analgesic effects of M3G. We have used ICV morphine successfully since 1986 in >200 patients for the relief of pain associated with head and neck cancer, midline pain, diffuse pain, and neuropathic pain (3,15). Long-term survivors experience excellent pain relief without rapid dosage escalation. Administration of ICV morphine bypasses liver and/or gut metabolism, and one would expect relatively little in vivo brain metabolism of morphine to its glucuronide metabolites (6). It is likely that the marked improvement in analgesia after initiation of ICV morphine is due to a large increase in the morphine concentration in the cerebrospinal fluid (CSF). Whether the improved side effect profile is due to a reduction in the CSF concentration of one or more of the active metabolites of morphine (M3G and M6G) requires investigation. The aims of this study were 1) to quantify plasma and CSF morphine, M3G, and M6G in patients undergoing ventriculostomy, both immediately before initiation of ICV morphine (baseline concentrations) and subsequently during the period of maintenance of cancer pain control by ICV morphine; and 2) to investigate the possibility that high CSF concentrations of M3G (the excitatory and antianalgesic metabolite) relative to morphine and M6G may contribute significantly to the poor clinical efficacy of systemic morphine in patients in whom subsequent small-dose ICV morphine not only produced effective pain relief, but also reduced side effects.
Approval was obtained from the Human Experimentation Ethical Review Committees of the Royal Brisbane Hospital and The University of Queensland. Patients were eligible for participation if they were scheduled to undergo ventriculostomy to provide access for the administration of ICV morphine. Although 23 patients were recruited for this study, postventriculostomy samples of CSF and plasma were available for only 14 (Table 1). Samples of CSF were available long-term for 5 of these 14 patients who received ICV morphine for 741 wk.
The ICV catheter was passed into the lateral ventricle and connected to the reservoir implanted subcutaneously, usually under local anesthesia, as described previously (3). Immediately after implantation but before administration of the first ICV morphine dose, samples of CSF (2 mL) and blood (5 mL) were collected for quantification of morphine, M3G, and M6G, thus obtaining baseline values. At the completion of the surgical procedure, patients received the first ICV dose of morphine, the magnitude (0.150.5 mg) of which was determined empirically by an experienced pain management clinician. Respiratory depression, sedation, and constipation were monitored after initiation of ICV morphine. Respiratory depression was assessed by pulse oximetry and clinical observation (no patient required naloxone). Sedation was assessed by discussion with patients and families and by the use of sleep charts. Any concerns about sedation (e.g., possible effects of sepsis, anemia, hypercalcemia, adjuvant drugs) were investigated routinely. Constipation was assessed by reports from patients and by the need for laxatives. The procedure for ICV morphine dosing and CSF collection was developed and validated using in vitro studies (data not shown). Just before ICV morphine administration, CSF (1 mL) was collected from the reservoir using a 1-mL syringe and a 25-gauge butterfly needle infusion set (dead space 0.51 mL) and discarded. Samples of CSF (1 mL) and blood (5 mL) were then collected for quantification of morphine, M3G, and M6G. ICV morphine (diluted with aspirated CSF to 1 mL) was administered into the reservoir, followed by 2 mL of sterile isotonic sodium chloride solution to ensure that the full dose of morphine reached the lateral ventricle and to prevent contamination of any CSF collected subsequently. Blood samples were refrigerated at 4°C before centrifugation. Plasma was transferred to clean polypropylene tubes using polypropylene pipettes to prevent possible adsorptive losses of morphine. Plasma and CSF samples were stored at -20°C before analysis. The CSF and plasma concentrations of morphine, M3G, and M6G were measured using our solid-phase extraction method and high-performance liquid chromatography with electrochemical detection (16). The limits of quantification for morphine, M3G, and M6G were 0.01 µM. The assay was validated over a wide range of plasma and CSF concentrations for morphine, M3G, and M6G (0.023.2, 0.022.7, and 0.022.7 µM, respectively). CSF samples containing morphine or M3G in concentrations that exceed the highest concentration in the respective standard curve were diluted with artificial CSF and reassayed. The intraday and interday coefficients of variation for morphine and its glucuronide metabolites were <23% and <17%, respectively.
Quantification of morphine, M3G, and M6G revealed considerable interpatient variability in the baseline concentrations of all three compounds in both CSF and plasma (Table 2). Examination of the mean (± SEM) CSF to plasma molar concentration ratios of morphine, M3G, and M6G in all baseline samples revealed that this ratio was relatively constant for M3G (0.12 ± 0.004) but was quite variable for morphine and M6G (0.0515.2 and 01.4, respectively) (Table 2). Goucke et al. (17) reported similar findings. In the five patients from whom CSF samples were collected during long-term ICV morphine administration (741 wk), we found that once steady-state CSF concentrations of morphine and M3G were achieved, they changed only relatively slowly with time.
Surprisingly, the trough CSF M6G concentrations were undetectable (<0.01 µM) in all patients by 24 h after initiation of ICV morphine (Fig. 1A), although M6G was still quantifiable in many of the plasma samples (Fig. 1B). By contrast, although the mean CSF M3G concentration decreased by 90% (from 1.0 to 0.1 µM) over a 7-day period postventriculostomy, the trough CSF M3G concentration remained quantifiable ( 0.01 µM) after cessation of systemic morphine in 11 of the 14 patients studied. As expected, by 24 h postventriculostomy, the mean trough CSF morphine concentration (approximately 20 µM) was approximately 50-fold higher than the baseline CSF morphine concentration (approximately 0.4 µM).
In most patients, the plasma concentrations of morphine, M3G, and M6G were undetectable (<0.01 µM) after cessation of systemic morphine, despite the quantifiable concentrations of morphine and M3G in the concomitant CSF samples (Fig. 1A). Although there was considerable interindividual variability in the CSF and plasma concentrations of morphine, M3G, and M6G in patients participating in this study, the data presented for Patients 13 are generally representative (Fig. 2).
The daily doses of oral morphine (150220 mg) administered to Patient 1 in the week before ventriculostomy, together with the daily doses of ICV morphine administered over a 7.5-wk interval (until death), are shown in Figure 2A. The ICV dose remained relatively constant as the disease state progressed. The CSF morphine concentration increased 100-fold, from 0.23 µM (baseline) to a steady-state concentration of approximately 23 µM, by 5 days postventriculostomy (Fig. 2B). At the same time, the CSF M3G concentration decreased eightfold, from 1.2 to 0.15 µM, and the CSF M6G concentration became undetectable by 24 h postventriculostomy, although the plasma M6G concentration at this time was relatively high (1.1 µM) and still detectable for the next 4 days until cessation of systemic morphine. The CSF and plasma samples collected on Days 27, 51, and 52 postventriculostomy show that the trough CSF concentrations of morphine (approximately 20 µM) and M3G (approximately 0.15 µM) remained relatively constant. The plasma concentrations were undetectable (<0.01 µM and <0.01 µM, respectively) on Day 27. Patient 2 was referred for admission receiving 4600 mg of morphine daily by subcutaneous (SC) infusion for neuropathic cancer pain. He was confused and exhibited excitatory opioid side effects (allodynia, restlessness, and myoclonic jerks). The dose of morphine was gradually reduced to 1100 mg (Fig. 2C) over the next 10 days, by which time the patient's mental state had cleared and he was competent to consent to ventriculostomy. After initiation of ICV morphine, the daily dose of systemic morphine was reduced until it was stopped on Day 9 (Fig. 2C). The ICV morphine dosing requirements remained constant (3 mg) from Day 8 until death (Day 14). The trough CSF M6G concentration was undetectable (<0.01 µM) by 24 h after initiation of ICV morphine (Fig. 2D). However, the preventriculostomy CSF M6G concentration was also undetectable (<0.01 µM), despite the corresponding plasma M6G concentration being relatively high (1.8 µM) and remaining quantifiable until Day 6 (0.28 µM) postventriculostomy. The CSF morphine concentration increased from approximately 1.4 µM (baseline) to a steady-state concentration of approximately 20 µM by Day 3 postventriculostomy (Fig. 2D). The concomitant CSF M3G concentration decreased from approximately 1.8 µM (baseline) to approximately 0.05 µM by Day 6 and was undetectable on Day 10. The corresponding plasma M3G concentrations decreased from approximately 7 µM (baseline) to undetectable levels (<0.01 µM) once systemic morphine administration had ceased (Day 9). The absence of M6G (potent analgesic metabolite of morphine) in preventriculostomy CSF was not unique to Patient 2. This was also observed in 9 of 20 patients for whom samples were available. For the 5 days before ventriculostomy, the daily dose of SC morphine administered to Patient 3 was 300 mg, which was gradually withdrawn over 3 days after the initiation of ICV morphine. The ICV morphine dose remained constant (3 mg/d) from Day 3 until the patient was discharged from the hospital on Day 14 (Fig. 2E). After initiation of ICV morphine, there was a 100-fold increase in the CSF morphine concentration, from approximately 0.5 µM (baseline) to approximately 50 µM at steady state, on Day 3 (Fig. 2F). The CSF M3G concentration decreased threefold throughout the postventriculostomy period, from 0.18 to 0.06 µM (Day 10), and the concomitant plasma M3G concentrations decreased from a baseline concentration of approximately 2.0 µM to undetectable levels by Day 10 (Fig. 2F). M6G was undetectable (<0.01 µM) in all samples of CSF collected after initiation of ICV morphine, a finding common to all patients.
As previously reported (3,15), ICV morphine produced effective pain relief with a reduced incidence of unacceptable side effects in patients whose pain had been only partially responsive to oral/parenteral morphine, or in whom pain relief could not be achieved without unacceptable morphine-related side effects. A significant feature was the mental alertness of these patients. Because the administration of morphine into the lateral ventricle bypasses metabolism in the liver/gut wall and delivers morphine to within close proximity of areas of the brain that are rich in opioid receptors (periaqueductal gray, locus coereleus), it was not unexpected that patients experienced excellent initial pain relief. However, experience with approximately 200 patients has shown that patients continue to experience excellent pain relief and that their ICV dosing requirements remain relatively stable, even the long-term survivors (3,15). In this study, by 24 h after initiation of ICV morphine, the mean trough CSF morphine concentration (approximately 20 µM) was approximately 50-fold higher than the respective baseline CSF concentration (approximately 0.4 µM) that was maintained while patients received ICV morphine. Although the magnitude of the increase in the morphine concentration at CNS opioid receptors need not necessarily be the same as that found in the CSF, it is clear that CNS neurons and their associated opioid receptors would have been exposed to markedly increased morphine concentrations relative to baseline. One would have expected that such exposure would have greatly accelerated the development of analgesic tolerance, manifested by rapid dosage escalation. That this did not happen strongly suggests that morphine itself is not primarily responsible for the development of analgesic tolerance or the production of the undesirable excitatory side effects (hyperalgesia, myoclonus, seizures) experienced by some patients receiving chronic, large-dose, systemic morphine. It is more likely that one or more of morphine's active glucuronide metabolites are involved.
In most patients (18 of 20), the baseline CSF morphine concentrations were less than the threshold (1 µM) required in vitro to inhibit (shorten the action potential duration) sensory neurons in the CNS, a process generally accepted as being integral to the establishment and maintenance of analgesia (18). However, after initiation of ICV morphine, the mean trough CSF morphine concentration was approximately 20 µM, a concentration that would have activated not only µ-opioid, but also Our findings also support the proposal by Hanks et al. (21) that after conventional oral doses of morphine, the resulting CSF morphine concentrations may be insufficient to elicit analgesia and that M6G, the potent analgesic metabolite, is important for the production and maintenance of analgesia. The lack of M6G in the baseline CSF obtained from 9 of 20 patients whose pain was poorly responsive to systemic morphine in this study lends further support to the view that M6G contributes to the analgesic effects of morphine given by systemic routes. M6G was not detected in baseline CSF from nine patients, despite measurable concentrations of M6G (0.071.79 µM) in many of the concomitant plasma samples (Table 2). This lack of M6G may have contributed significantly to the relatively poor analgesia provided by oral or parenteral morphine in the preventriculostomy period. These unexpected findings clearly indicate that even relatively high plasma concentrations of M6G do not guarantee detectable concentrations of M6G in the CSF. Furthermore, these findings suggest that M6G may be actively transported, rather than passively diffused, across the blood-brain barrier and that this mechanism may be defective or inhibited in some patients. The proportion of patients lacking M6G in their baseline CSF samples (45%) is probably overrepresented relative to the general population receiving chronic morphine for cancer pain, as systemic morphine had provided unsatisfactory analgesia for the patients in this study. Further research is required to investigate this issue. Because animal studies have shown that ICV M6G is a more potent respiratory depressant than morphine (22,23), the absence of M6G in patient CSF may also explain why patients receiving ICV morphine did not experience respiratory depression. Additionally, the lack of respiratory depression may have been due to the continued presence of M3G in the CSF of most patients (albeit at a 10-fold lower concentration than preventriculostomy), as ICV M3G has been shown to attenuate the respiratory depressant effects of ICV morphine and M6G in both rats and dogs (22,23). The decreased constipation reported by patients after initiation of ICV morphine indicates that this side effect is not primarily a central effect of morphine. The fact that ICV M6G more potently inhibits gastrointestinal transit than morphine in mice (24) suggests that the absence of M6G in patient CSF contributes to the relief of constipation. Additionally, cessation of systemic morphine removes the local effects of morphine and its metabolites on opioid receptors in the gut wall. The continued presence of M3G in the CSF of 11 of the 14 patients studied in the postventriculostomy period, together with the undetectable levels of both morphine and M3G in patient plasma after cessation of systemic morphine, indicates that the brain was metabolizing morphine to M3G. By implication, for patients receiving morphine by systemic routes, both brain and liver/gut metabolism of morphine to M3G contribute to the CSF concentrations of M3G, whereas only liver/gut metabolism contributes significantly to the CSF M6G concentrations in the same patients. Another factor that may have contributed to the efficacy and potency of ICV morphine and the lack of discernible development of analgesic tolerance in these patients is that the mean CSF molar concentration ratio of putative antianalgesic to analgesic substances, M3G/(morphine + M6G), decreased by 23 orders of magnitude, coinciding with markedly improved pain relief. This observation is supported by studies in rats, in which M3G was produced by in vivo metabolism (25,26), but not by studies in which M3G was given by systemic routes (27,28). Previous studies in our laboratory have shown that when M3G is formed endogenously after the administration of three different chronic IV dosing regimens, the degree of pain relief produced in rats is highly inversely correlated with the mean plasma molar concentration ratio, M3G/morphine (antianalgesic to analgesic ratio) (25). Similarly, Barjavel et al. (26) showed a highly significant inverse correlation between levels of antinociception in rats and the mean M3G/morphine molar concentration ratio in the cortical extracellular fluid after single SC doses of morphine (10 mg/kg). In summary, the administration of ICV morphine to cancer patients with pain only partially responsive to systemic morphine provided excellent pain relief with a reduction in morphine-related side effects. Quantification of the CSF morphine, M3G, and M6G in these patients revealed that, by 24 h after the initiation of ICV morphine, M6G was undetectable in the CSF, which may explain, at least in part, the absence of respiratory depression and the improvement in constipation. The absence of a rapid escalation in the ICV morphine dosing requirements in these patients, despite the approximately 50-fold increase in the mean trough CSF morphine concentrations, provides strong evidence to support our view that morphine itself is unlikely to be the primary causative agent of either analgesic tolerance or excitatory opioid side effects in patients receiving morphine by systemic routes. It is more likely that M3G, the excitatory and putative antianalgesic metabolite of morphine, has a major role in producing these undesirable effects of chronic morphine administration in humans.
This research was supported by The University of Queensland Foundation, The Queensland Cancer Fund, and The University of Queensland Research Grants Scheme. The authors sincerely thank the nursing staff of the Neurosurgical Unit and Ward 2C at Royal Brisbane Hospital for assistance with blood and CSF collection from patients participating in this study.
This research was presented in preliminary form at the 7th World Congress on Pain.
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|