Methadone initiation and rotation in the outpatient setting for patients with cancer pain

Abstract

BACKGROUND:

Methadone is an effective and inexpensive opioid for cancer pain treatment. It has been reported as difficult to use in the outpatient setting because of its variable relative potency and long half-life. The purpose of this study was to determine the outcome of methadone initiation or rotation for cancer pain treatment in outpatient settings.

METHODS:

Chart review was done of 189 consecutive patients who underwent methadone initiation or rotation at the authors' palliative care outpatient center. Data were collected regarding demographic and clinical characteristics, symptoms, and opioid side effects at baseline and for 2 follow-up visits (F1, F2). Failure was defined as methadone discontinuation by the palliative care physician or patient's hospitalization for uncontrolled pain or methadone-related side effects at F1.

RESULTS:

One hundred (53%) initiations and 89 (47%) rotations were conducted. Success rates for methadone initiation and rotation were 82 of 89 (92%) and 85 of 100 (84%), respectively. Mean (standard deviation) age was 60 (11) years. One hundred (53%) patients were women, 138 (73%) were white, and 182 (96%) had solid cancers. The main reason for rotation was pain (65 of 89 patients, 47%). Median (interquartile range, IQR) pain scores (Edmonton Symptom Assessment Scale/0-10) were 6 (5-8), 4 (3-6), and 3 (2-5) at baseline, F1, and F2, respectively (P < .0001). Median (IQR) daily methadone dose for initiation and rotation was 10 (5-15) mg and 15 (10-30) mg at F1 (P < .0001) and 10 (8-15) mg and 18 (10-30) mg at F2 (P < .0001), respectively. Constipation and nausea improved (P < .005) after initiation/rotation to methadone. Frequency of sedation, hallucinations, myoclonus, and delirium did not increase after initiation/rotation to methadone.

CONCLUSIONS:

Outpatient methadone initiation and rotation for cancer pain treatment were safe, with high success rates and low side effect profiles. Cancer 2010. © 2010 American Cancer Society.

Pain is extremely common in patients with advanced cancer, with prevalence as high as 80%. The vast majority of patients with cancer pain require opioids for treatment.1, 2 As a result of a major effort by various international groups, the use of opioids has improved significantly, and patients now have earlier exposure to opioids, generally receive higher dosages, and are better managed than in the past.3 This highly desirable increase in the use of opioids, combined with increased vigilance, has resulted in increased detection of several side effects. Side effects can range from constipation and nausea to the more serious opioid-induced neurotoxicity, which includes delirium, seizures, and myoclonus. Opioid toxicity is caused by both the effect of the parent drug as well as by accumulation of opioid metabolites (eg, morphine-3-glucuronide and morphine-6-glucucoronide for morphine).4 Opioid rotation is required when there are intolerable side effects despite good analgesia or inadequate pain control despite dose escalation.

Methadone is a synthetic drug that was developed after the Second World War as an alternative to morphine, having a broad spectrum of suggested actions—μ receptor agonist, N-methyl-D-aspartate receptor antagonist, monoamine reuptake inhibitor, and α receptor agonist—and has been increasingly used as a second-line opioid in recent years in the management of cancer-related pain.5 It has complex pharmacodynamic and pharmacokinetic properties that make it difficult to use. It is an ideal drug for opioid rotation because of its efficacy, higher oral bioavailability, lack of known active metabolites, incomplete cross tolerance with other μ receptor agonists, and lower cost. Several prospective and retrospective trials have demonstrated successful rotation to methadone from other strong opioids in the setting of intolerable side effects and inadequate analgesia despite dose escalation.6-8 However, it has the disadvantages of having a slow elimination phase (15-60 hours) and high interindividual variation in the equianalgesic ratio with other opioids, which have largely restricted methadone use to experienced physicians in pain medicine.9

It is also a desirable alternative first-line strong opioid particularly in developing countries, where cost is an important consideration. There is growing preliminary evidence supporting the use of methadone as a first-line strong opioid.10, 11 Largely, the use of methadone is limited by its complex pharmacokinetic and pharmacodynamic properties, such as the reported elimination metabolism through cytochromes 3A4 and 2B6,12 the participation of P-glycoprotein in the defense against central methadone intoxication,13 and concerns related to its potential cardiotoxic effects (with prolongation of the QT interval),14, 15 which could impede its concomitant use with specific drugs inhibiting the cited cytochromes or P-glycoprotein, or cause QT interval prolongation.

Unfortunately, there are few reports16, 17 on the initiation and/or rotation of methadone in the outpatient setting for cancer pain. The objective of this retrospective study was to determine the efficacy and safety of methadone initiation (in strong opioid-naive patients) or rotation from another strong opioid in treating cancer-related pain in an outpatient palliative care clinic at a comprehensive cancer center.

MATERIALS AND METHODS

This study was reviewed and approved by the institutional review board and was granted a waiver of informed consent. We reviewed the electronic charts of 189 consecutive patients who were prescribed methadone for the first time at the Palliative Care Outpatient Clinic at The University of Texas M. D. Anderson Cancer Center between April 2003 and March 2007.

Patients were included in the study if 1) they were receiving methadone for the first time in the outpatient setting and 2) the previous opioid was completely stopped at the day of methadone initiation. Patients were excluded of the study if they were 1) receiving opioids via intrathecal pump, 2) receiving opioids for reasons other than pain (eg, dyspnea), or 3) admitted to the hospital after the initiation of methadone for reasons other than pain or methadone-related toxicities.

Strong opioid-naive patients were started on methadone as the first-line opioid when concerns regarding the utilization of other opioids were present (ie, renal failure) or when logistic issues were identified, such as the cost of medications for patients with limited resources. Rotations for pain were made by the palliative care specialist after patients underwent multiple unsuccessful titrations with the previous opioid agonist. Rotation usually was decided on when the patient presented early signs or was considered at risk for dose-limiting side effects. Rotations because of side effects took place when patients developed severe opioid toxicity, such as delirium, myoclonus, and/or sedation.

In our Palliative Care Outpatient Clinic, patients were initiated on regular methadone at 5 mg twice a day, with an additional prescription of rescue methadone, according to minor modifications to procedures previously described by our group.10 In the context of opioid rotation, we applied the usual strategy of variable morphine equivalent daily dose:methadone conversion ratios according to the dose of previous opioid (5:1 when previous morphine equivalent daily dose ≤90 mg/d, 8:1 when it was between 91 and 300 mg/d, and 12:1 when it was ≥301mg/d).18 These rotation ratios are modified by the attending palliative care physician based on specific clinical characteristics. The most common modifier is the reason for rotation: patients being rotated because of severe intractable pain were rotated to a lower morphine equivalent daily dose:methadone ratio as compared with patients rotated because of opioid neurotoxicity.

Data were collected regarding patients' demographics, diagnosis, current use of opioids (drug names and total daily dose), date of methadone initiation, and reason for rotation (if previously on opioids). Predictors of poor response to opioid therapy such as presence of delirium, psychological distress, and neuropathic and incidental pain19 were also noted.

Symptoms (namely pain, nausea, and drowsiness) were recorded for the first visit to the Palliative Care Outpatient Clinic and 2 consecutive follow ups using the Edmonton Symptom Assessment Scale (ESAS), a widely used and validated tool to assess 9 symptoms (pain, nausea, drowsiness, dyspnea, anxiety, depression, anorexia, sleep, and fatigue) and general feeling of well-being in a 0-10 scale.20, 21 Other opioid side effects, such as sedation, hallucinations, myoclonus, delirium, and constipation, are routinely assessed in our Palliative Care Outpatient Clinic and were considered present when reported by the attending palliative care physician in the patient's chart.

The total daily opioid dosage was calculated for the first visit to the Palliative Care Outpatient Clinic and 2 consecutive follow-up visits by converting the total opioid dosage during 24 hours to an equivalent dose of oral morphine (morphine equivalent daily dose), following standard equianalgesic conversion tables.22 The ratio between morphine equivalent daily dose and methadone was calculated by dividing the morphine equivalent daily dose previous to rotation by the methadone daily dose at stable pain control (defined as ESAS pain change ≤1 point or ≤30%).

Successful initiation/rotation to methadone was evaluated at the first follow-up visit. Complete success of methadone initiation or rotation was defined as 1) improvement ≥30% or ≥2 points in the ESAS pain score for strong opioid-naive patients who initiated methadone because of pain or for patients who were rotated to methadone from other strong opioids for uncontrolled pain in the absence of side effects; 2) reported evidence of disappearance of side effects at the first follow-up visit in the cases in which the reason for methadone initiation was the development of side effects with the previous strong opioid; 3) improvement ≥30% or ≥2 points in the ESAS pain score, with reported disappearance of side effects for those patients rotated to methadone in the setting of uncontrolled pain and dose-limiting side effects with the previous opioid; and 4) reported evidence of resolution of the logistical issue with the previous opioid (ie, resolution of nonadherence related to financial burden as detected and reported by the specialist palliative care physician as the result of routinely performed adherence assessments), with stable pain control. Methadone initiation or rotation in the outpatient setting was considered a failure if the drug was discontinued, if the patient was admitted to the hospital because of pain or methadone-related side effects on or before the first follow-up visit, or if the patient was lost to follow-up. All cases that did not fall into the complete success or failure categories were considered partial successes.

Data obtained from the second follow-up visit were analyzed with regard to pain control and opioid side effects.

Descriptive statistics were used to summarize the data. Chi-square tests were used to determine associations between categorical variables. Differences between continuous variables were analyzed using t tests for normally distributed data and Wilcoxon rank sum tests for non-normally distributed data. Significance <.05 were considered statistically significant. Spearman correlation coefficient was calculated to evaluate the association between morphine equivalent daily dose of the previous opioid and the morphine equivalent daily dose:methadone dose ratio. To further identify factors associated with the morphine equivalent daily dose:methadone ratios at stable pain control, we performed multivariate analyses of variance based on the ranks of the ratios, because these values were not normally distributed. We performed these analyses first only including the variables found to be univariately associated with the ratios at significance levels <.20, then deleted variables that were not statistically significant (P < .05) 1 at a time from the multivariate model.

RESULTS

One hundred eighty-nine patients underwent methadone initiation or rotation in the Palliative Care Outpatient Clinic. Data collection is described in the flowchart in Figure 1. Patient characteristics are summarized in Table 1. There were no significant differences between groups.

Figure 1. A data collection flowchart is shown. F/U indicates follow-up visit.

Table 1. Demographic Characteristics

Characteristic	Initiation, n=89	Rotation, n=100	Total, N=189	P
SD indicates standard deviation; CAGE, cut-down, annoyed, guilty, eye-opener test.
Mean (SD) age, y	60 (11)	58 (11)	60 (11)	.2
Women	45 (51%)	55 (55%)	100 (53%)	.6
Race
African American	11 (12%)	4 (4%)	15 (8%)	.06
Hispanic	7 (8%)	10 (10%)	17 (9%)	.6
Caucasian	62 (70%)	77 (77%)	139 (73%)	.4
Other	9 (10%)	9 (9%)	18 (10%)	.8
Cancer site
Head/neck/lung	36 (40%)	36 (36%)	72 (38%)	.6
Urologic	15 (17%)	16 (16%)	31 (16%)	1.0
Breast	7 (8%)	15 (15%)	22 (12%)	.2
Gastrointestinal	10 (11%)	9 (9%)	19 (10%)	.6
Hematological	2 (2%)	5 (5%)	7 (3%)	.5
Other	11 (12%)	10 (10%)	21 (11%)	.7
Cancer status
Metastatic	58 (65%)	72 (72%)	130 (69%)	.4
Locally advanced/recurrent	28 (32%)	25 (25%)	53 (28%)	.4
No evidence of disease	3 (3%)	3 (3%)	6 (3%)	1.0
Poor prognostic factors
Neuropathic pain	52 (58%)	57 (57%)	109 (58%)	.9
Incidental pain	23 (26%)	23 (23%)	46 (24%)	.7
Psychological distress	32 (36%)	39 (39%)	71 (38%)	.8
CAGE positive	11 (12%)	13 (13%)	24 (13%)	1.0

By using a conservative intention-to-treat approach, all 7 patients who did not have any follow-up visits were considered failures. Reasons for rotation among these cases are described in Table 2. Because of documentation limitations, it was not possible to determine the reason for rotation to methadone in 4 of 100 (4%) cases.

Table 2. Methadone Initiation/Rotation Outcomes: Number of Successes (Complete and Partial) and Failures of Methadone Rotation and Initiation

Rotation Characteristic	Complete Success	Partial Success	Overall Success	Failure
a P=.17 for rotation versus initiation.
Reasons for rotation
Pain, n=65	28 (48%)	30 (52%)	58/65 (89%)	7/65 (11%)
Pain and side effects, n=23	10 (59%)	7 (41%)	17/23 (74%)	6/23 (26%)
Side effects only, n=5	5 (100%)	0 (0)	5/5 (100%)	0/23 (0)
Logistic, n=3	2 (67%)	1 (33%)	3/3 (100%)	0/23 (0)
Unclear, n=4	2 (100%)	0 (0)	2/4 (50%)	2/4 (50%)
Rotation total, n=100	47 (55%)	38 (45%)	85/100 (85%a)	15/100 (15%)
Initiation, n=89	53 (65%)	29 (35%)	82/89 (92%a)	7/89 (8%)

Median (interquartile range, IQR) time between consultation and first follow-up visit was 13 (6-21) days, and did not differ between initiation and rotation patients (P = .22). Median (IQR) time between first and second follow-up visits was 8 (1-5) days for initiation patients and 15 (10-25) days for rotation patients (P < .0001). Median (IQR) time between baseline and second follow-up visit was 37 (23-60) days (P = .59).

Median (IQR) morphine equivalent daily dose before methadone initiation was 100 (60-185) mg/d. The median (IQR) methadone dose at first follow-up visit was 10 (5-15) mg/d for initiation patients and 15 (10-30) for rotation patients (P < .0001), and at second follow-up visit it was 10 (8-15) mg/d for initiation patients and 18 (10-30) mg/d for rotation patients (P < .0001). Median dose increase between first and second follow-up visits was 1.75 (0-5) mg (P < .0001) for initiation and 2.5 (0-10) mg for rotation patients (P = .12). Considering previous reports of variable dose-dependent ratio for conversions from other opioids to methadone,23-26 the correlation between the morphine equivalent daily dose before rotation and the morphine equivalent daily dose:methadone dose ratio at stable pain control was studied, and its results are shown in Figure 2. The findings confirm the significant correlation between previous opioid dose and opioid:methadone ratio (r = 0.55, P < .0001). The morphine equivalent daily dose:methadone ratio was univariately analyzed according to age, sex, reason for rotation, and morphine equivalent daily dose of the previous opioid. Multivariate analyses were then performed with the 3 variables (age, reason for rotation, and morphine equivalent daily dose before rotation) found to be significant in the univariate analyses. Age was not significant and was removed from the model. Results for both univariate and multivariate analyses are summarized in Table 3.

Figure 2. Association between morphine equivalent daily dose of the previous opioid and morphine equivalent daily dose (MEDD):methadone ratio is shown.

Table 3. Median MEDD:Methadone Ratio at Stable Pain Control

Variable	No.	Median MEDD: Methadone Ratio (Interquartile Range)	Univariate Analysis P	Multivariate AnalysisaP
MEDD indicates morphine equivalent daily dose. a Multivariate model R^2: 0.269, P=.0002. b Statistically significant.
Age, y
≤65	47	6:1 (3:1-8:1)
>65	16	9:1 (6:1-10:1)	.0302b	—
Sex
Women	34	6:1 (4:1-10:1)
Men	29	6:1 (3:1-10:1)	.8302	—
Reason for rotation
Pain alone	43	6:1 (3:1-8:1)
Side effects (±pain)	16	9:1 (6:1-11:1)	.0266b	.0032b
Previous opioid MEDD
≤90 mg/d	29	5:1 (3:1-6:1)
>90 mg/d	34	8:1 (5:1-10:1)	.0022b	.0005b

Reasons for opioid rotation are described in Table 2. Success (complete and partial) of methadone initiation after 1 follow-up visit occurred in 82 of 89 (92%) initiation patients and in 85 of 100 (85%) rotation patients (P = .2). Reasons for failure of the rotation to methadone were nonimprovement/increase in pain in 12 of 15 (80%) cases and appearance/persistence of side effects in 3 of 15 (20%) cases. For opioid naive patients initiating methadone, reasons for failure were nonimprovement of pain in 3 of 7 (43%) cases, nonadherence in 1 of 7 (14%) cases, and appearance/persistency of side effects in 3 of 7 (43%) cases.

There were 29 of 89 (35%) partial successes among initiation patients between baseline and first follow-up visit. Reasons for partial successes were pain in 21 of 29 (72%) patients, nonadherence in 3 of 29 (10%) patients, and side effects in 5 of 29 (17%) patients. Among rotation patients, there were 38 of 85 (44%) partial successes between baseline and first follow-up visit. In this group, reasons for partial successes were pain in 28 of 38 (74%), nonadherence in 3 of 38 (9%), and side effects in 7 of 38 (18%) patients.

Data were available for 63 of 89 (70%) rotation and 63 of 93 (68%) initiation patients at the time of the second follow-up visit. At this time point, 53 of 63 (84%) rotation and 59 of 61 (96%) initiation patients continued receiving methadone (P = .03).

ESAS pain scores improved in all groups, regardless of their reason for initiation. Nausea and drowsiness were also reduced in all groups, as summarized in Figure 3. Sedation, hallucinations, myoclonus, delirium, and constipation, symptoms frequently related to opioid toxicity, did not increase in frequency after methadone initiation/rotation, as described in Figure 4.

Figure 3. Median and interquartile range Edmonton Symptom Assessment Scale (ESAS) symptom scores are shown. F/U indicates follow-up visit.

Figure 4. Frequencies of opioid toxicity-related symptoms in patients who underwent opioid initiation/rotation to methadone are shown.

DISCUSSION

Caution in the use of methadone in the outpatient setting was previously recommended in several articles by our group and others.27-30 We present the results of a retrospective chart review of 189 advanced cancer patients who underwent initiation or rotation to methadone in an outpatient palliative care clinic. In the majority (88%) of the patients, this strategy was successful. The use of methadone as a first-line opioid and especially in the outpatient setting is still scarce, probably because of concerns about the complex management of this drug and its side effects.9 In our study, 89 patients were started on methadone as first strong opioid, and the overall success rate was of 92%. In a randomized study by Mercadante et al with 108 patients receiving slow-release morphine, fentanyl, or methadone as initial strong opioid, no difference was found between the 3 drugs regarding effectiveness, need for coanalgesia, and safety. Methadone had the advantage of lower cost; however, it required more dose titration, prompting the authors to suggest that its use requires more expertise in pain management compared with other opioids.11 Our group has also reported in a randomized, double-blind study of methadone versus morphine as first-line opioid for the treatment of cancer pain that no difference was found regarding tolerability and analgesia between the drugs.10

The use of methadone in the process of opioid rotation has been more extensively studied. In general, the switch to methadone has been considered safe and effective. However, the majority of the studies were performed in the inpatient setting.6, 8, 23, 31-34 Our study shows that in the context of opioid rotation, the use of methadone for cancer-related pain in the outpatient setting is successful (85% of the cases). Specifically, in the outpatient setting, Hagen and Wasylenko reviewed 29 charts of patients with cancer-related pain who underwent rotation to methadone in the outpatient setting and found that rotation to methadone was successful in 17 (62%) patients within a 32-day average. The authors concluded that patients can be safely and effectively rotated to methadone in the outpatient setting despite a long period to reach pain stability.16 Hernansanz et al reported success in 10 of 14 (71%) cases in which opioid rotation was performed in the home setting in advanced cancer patients.17 Our study builds on this previous evidence, shedding light on the possibility of safely using this drug in the outpatient setting.

After 2 follow-up visits, methadone was continued in the majority of patients in our study, more frequently among opioid-naive patients than in patients undergoing opioid rotation. This is consistent with previous data showing that the titration of methadone was easier in patients who had not previously received opioids.35

Methadone used as first-line strong opioid in general resulted in improvement in pain and nausea, with stable drowsiness. The improvement of pain is consistent throughout studies of methadone rotation/initiation. In a retrospective study of 196 patients who received methadone either as a first-line opioid (54%) or for opioid rotation from morphine (18%), only pain and insomnia improved, and other symptoms such as drowsiness, constipation, and confusion worsened.36 The improvement in nausea shown in the current study might be related to advances in the interdisciplinary management of this symptom. Delayed onset sedation was previously described by our group in a study with 103 patients randomized to either morphine or methadone as first-line opioid, in which those who received methadone reported more delayed onset sedation, with no difference in nausea, constipation, or confusion.10

As observed in previous studies, patients who started or rotated to methadone showed considerable reduction in the frequency of constipation as reported by the attending palliative care physician. This has been previously reported and is postulated as being secondary to differences in methadone's affinity for intestinal opioid receptors and also to bowel withdrawal syndrome, because of the significantly smaller doses of methadone needed to reach analgesia, binding to fewer intestinal opioid receptors as compared with other opioids.8, 37 Frequency of other symptoms related to opioid toxicity such as myoclonus, delirium, hallucinations, and sedation did not increase with methadone rotation; however, it is not possible to positively conclude trends of reduction because of the number of drop-offs between follow-up visits. Response to methadone when used for opioid rotation has been attributed to an absence of circulating metabolites as compared with other μ agonists.18

It has been reported that methadone elimination mostly occurs through cytochrome 3A4.12 Although very recently published evidence began to question this metabolic route in animal studies,38, 39 it is still advisable to take into consideration that the use of potent cytochrome 3A4 inhibitors (such as fluconazole, ciprofloxacin, venlafaxine, fluoxetine, and macrolides)40, 41 has the potential to interfere in the elimination of methadone, increasing the risk of elevation of the plasma levels of the drug toward toxic levels. The same effect could be expected when inhibitors of cytochrome 2B6 (such as haloperidol, levopromazine, paroxetine, sertraline, fluoxetine, and chlorimipramine)40 are used concomitantly with methadone, because it is also related with methadone metabolism. Doxorubicin, vimblastin, and actinomycin, frequently used in cancer patients, are inhibitors of P-glycoprotein, and by impairing the ability of the blood-brain barrier to block excessive transference of methadone to the central nervous system, might predispose patients to have increasing plasmatic levels of methadone despite stable dosing.

Prolongation of the QT interval is associated with increased risk of severe arrhythmias42 and has been reported to be associated with the use of methadone.14, 15, 43 We did not find any clinical evidence of cardiac rhythm alterations after the introduction of methadone. However, because neither cardiac effects nor the concomitant use of other potentially arrhythmogenic drugs were formally assessed in this retrospective study, we are unable to draw definitive conclusions regarding the cardiac safety of methadone. Therefore, it is still advisable to obtain information regarding the presence of electrolyte imbalances or the concomitant use of other cardiotoxic drugs such as tricyclic antidepressants, diuretics, and antipsychotics.44

The significant association between morphine equivalent daily dose of the previous opioid (Fig. 2) and the variable morphine equivalent daily dose:methadone ratio confirms previous reports from other authors.23-26, 45 We found in univariate analysis that greater morphine equivalent daily dose:methadone ratios are associated with older age, rotation because of side effects, and higher previous opioid doses. In the multivariate model, reason for rotation and previous opioid doses (grouped according to previously published evidence45) remained independently associated with greater morphine equivalent daily dose:methadone ratio. Our findings in the outpatient setting are similar to those recently reported in a study with 54 patients who underwent rotation in the inpatient setting, which showed that the median morphine:methadone ratio ranged from 2:1 to 15:1, being independently associated with previous opioid doses and reason for rotation.45 These findings suggest that methadone behaves in a different manner as compared with other opioids with regard to opioid sensitivity, and caution should be exercised in patients who had escalation of the previous opioid.

We were not able to find a significant number of patients who were rotated to methadone because of financial issues in our limited sample. This might be because of the finding that when financial cost is identified by the palliative care team as a concern, patients are already started on methadone as a first-line strong opioid. However, it was impossible to obtain this detailed information for methadone initiations in this retrospective study. Our group and others45-48 have already reported that methadone is a very inexpensive drug, being an ideal alternative to other opioids, particularly in developing countries where cost of treatment is often a barrier in providing excellent care to patients with cancer pain.

Conclusions

Methadone was effective for use as both first- and second-line opioid. Patients exhibited tolerable side effects to the drug. Our study adds to the growing evidence in the literature to recommend such use. Further prospective studies on the use of methadone as a first-line strong opioid with longer follow-up periods are required to confirm our clinical findings.

Acknowledgements

We thank Laura Navarrete, Timotheos Paraskevopoulos, Elva Ramirez, and Lynn Roberts for their invaluable help with data collection.

Conflict of Interest Disclosures

Eduardo Bruera is supported in part by National Cancer Institute R01 grants CA122292-01 and CA124481-01, and National Institute of Nursing Research grant NR010162-01A1.