Optimizing Response to Methadone Maintenance Treatment: Higher Dose Methadone
Sarz Maxwell, MD, Marc Shinderman, MD, Journal of Psychoactive Drugs: Vol-31 #2, April June 1999
ABSTRACT
Using signs, symptoms and serum methadone levels to guide evaluation, we treated 164 patients
in a methadone maintenance program with doses of methadone exceeding 100 mg/d. The mean
dose of these higher dose (HD) patients was 211 mg/d (range 110-780 mg/d). A comparison
group (C) of 101 patients was randomly selected from the general clinic population (mean dose
65 mg/d). At intake the HD group reported $153/day of heroin use vs. $87/day in the C group.
The HD group had more patients whose opiate of choice was an oral pharmaceutical (30%
versus 2% of the C group). Sixty-three percent of the HD group had comorbid Axis I psychiatric
diagnoses compared to 32% of the C group. Response to psychopharmacologic treatment was
enhanced by increased methadone dose in HD patients with "refractory" psychiatric disorders.
Urine toxicologies described as "before" were collected prior to increase over 100 mg/d in the
HD group or at the first routine urine toxicology collection of the calendar year for the C group.
These results were compared to the most recent urine toxicologies for both groups ("after"). The
percentage of toxicologies positive for illicit drugs in the HD group dropped from 87% "before"
to 3% "after". The C group were 54% positive "before" and 37% positive "after". We conclude
that doses of methadone in excess of 100 mg/d (range 110-780 mg/d in our sample of 164
patients) are not only safe but necessary to prevent illicit opiate use, stabilize psychiatric
symptoms, and diminish abuse of alcohol and benzodiazepines in many patients.
INTRODUCTION Methadone has been used since the middle 1960's for the treatment of opiate
addiction. Methadone maintenance treatment (MMT) decreases illicit opiate use (Dole,
Nyswander & Kreek 1966), improves quality of life (Reno & Aiken, 1993), and reduces the
incidence of many associated factors, including crime (Senay 1985), unemployment (Lowenstein
et al, 1992), and HIV infection (Barthwell & Senay 1989).
In 1966 Dole , Nyswander & Kreek recommended a dose of 80 - 120 mg/day of methadone. In
the 1980's, many MMT programs attempted to minimize methadone dose (Craig 1980). Some
used punitive interventions such as reducing methadone dose in response to illicit opiate use
(Iguchi et al 1988). Many MMT clinics prescribed increased psychosocial interventions instead
of increased medication dose for patients who continued to abuse opiates. Some practitioners,
however, continued to view MMT as medical treatment. Clinical utilization of serum methadone
levels began in the 1980's, and became more widespread in the 1990's (Bell et al 1988; Kell
1995; Loimer & Schmid 1992; Tennant 1987; Tennant et al 1984). Though several researchers
have attempted to maintain patients at serum methadone levels as low as 100 ng/ml (Bell et al
1988), it is now generally agreed that 400 ng/ml is the lowest serum methadone level that can
reliably support effective methadone maintenance therapy (CSAT-TIP 1993).
METHODS The Center for Addictive Problems (CAP) is a fee-for-service MMT program
(MMTP) in continuous operation since 1977. The mean daily dose of methadone for the clinic
population of 1,100 at the time of data collection was 78 mg/day. Patients who were identified as
being clinically unstable despite methadone treatment at 100 mg/d were referred by nursing or
counseling staff for medical evaluation. Serum methadone levels were measured while the
patient was taking 100 -120 mg/d and, in some patients, repeated at intervals during dose
titration. Blood was drawn onsite 22-26 hours after an observed dose of methadone. Methadone
serum concentration determinations were performed by Quest and Smith Kline Beecham
Laboratories in San Diego, CA. There was no standard protocol for dose titration: patients'
methadone doses were individually titrated against signs and symptoms of opiate withdrawal.
The charts of these 164 patients were reviewed and the data compared with data from the charts
of 101 patients randomly selected from the general clinic population.
Urine toxicologies from 1-4 weeks prior to increase of dose over 100 mg/d for the 164 patients
requiring higher doses of methadone (HD) were compared to the first urine toxicology of the
calendar year 1996 for the comparison group (C). We identify these results as the "before"
sample. The time between the "before" and "after" urine toxicologies for the C group ranged
from 3 to 6 months. We used this 3-to-6-month time span to compare urine toxicologies because
it was similar to the time span necessary for titration from 100 mg to final dose in the HD group.
The "after" sample for both groups refers to the most recent urine toxicology available at the
time of data collection. Urine toxicologies were reported as positive if they showed any illicit
drug. Urine toxicologies do not reflect use of illicitly obtained methadone, but patients who
reported regular use of illicit methadone were reported as having a positive toxicology. Patients
in both groups were reported as having concomitant psychiatric illness if they were taking
psychotropic medication, including antidepressants, anxiolytics, mood stabilizers, and
antipsychotics. In most cases these medications were being prescribed by our clinics'
psychiatrists. Statistical analyses included the Cochran test for repeated measures and chi-square
measures of significance, as appropriate. RESULTS The HD population of 164 patients included
106 males and 58 females. Mean methadone dose for the HD patients was 211 mg/d (range 120 780), patients had been in MMT an average of 117 weeks (range 1-844) before taking a dose
exceeding 100 mg/d, and patients had been taking a dose exceeding 100 mg for an average of 31
weeks (range 3-346) at the time of data collection. Mean methadone dose in the C group was 69
mg/d, (range 10-100), and average time in treatment was 252 weeks (range 1-856).
Sixty-three percent of the HD group (n=103) were taking psychotropic medications. These
included selective serotonin reuptake inhibitor antidepressants (55%, n=57), benzodiazepines
(56%, n=58), tricyclic antidepressants (32%, n=33), divalproex (18%, n=19), methylphenidate
(3%, n=4), neuroleptics (2%, n=2) and lithium (2%, n=2). About half of the patients taking
psychotropics (48%, n=49) were taking more than one medication. An exact breakdown of
diagnoses was not possible, but the majority of patients suffered from affective disorders
(predominantly major depression) and anxiety disorders (predominantly post-traumatic stress
and panic disorders). The proportion of C patients taking psychotropic medications (28%, n=29)
was significantly less than that of the HD group (chi-square = 24.22, df = 1, p < .001), but the
breakdown of diagnoses and medications was similar: 62% (n=18) taking selective serotonin
reuptake inhibitor antidepressants; 38% (n=11) tricyclic antidepressants; 45% (n=13) prescribed
benzodiazepines; 21% (n=6) taking divalproex; 7% (n=2) antipsychotic medications and 3%
(3=1) prescribed methylphenidate.
The HD group reported greater amounts of heroin use at intake: average $153/day (range $40 800) compared to $87/d (range $30-200) in the C group (t = 5.15, df = 117, p < .001). These
calculations were made using only patients whose opiate of choice was heroin. Route of heroin
administration was similar for the two groups. Among the HD patients 51% (n=84) were IVDU
and 32% (n=53) used heroin by nasal insufflation. The comparative figures for the C group were
64% (n=65) and 33% (n=34), respectively. In both groups there were a number (5% of C, 12%
of HD) who had been in MMT continuously for more than 15 years. These patients were not
included in computing heroin habit. There was a significant difference between groups in
number of patients who preferentially used prescription opiates: 2% (n=2) of the C group
compared to 34% (n=50) of the HD group (chi-square = 32.21, df = 1, p < .001).
Urine toxicology results for the C group were 55% (n=56) positive/before and 36% (n=37)
positive/after. For the HD group, urine toxicologies were positive for 87% (n=144) before high
dose, and 3% (n=5) after (Q = 338.00, df = 1, p < .001).
Termination rates and patterns for the entire population at the clinic during the period studied
were obtained. Therapeutic termination was defined as one which followed planned, medicallymonitored withdrawal procedures. Non-therapeutic termination describes abrupt discontinuation
of treatment without notice or administrative withdrawal of medication because of
noncompliance with clinic regulations. The overall 12-month termination rate for the entire
clinic was 65%. The majority of these (70%) were non-therapeutic. Transfer to other MMTPs
accounted for 10%, therapeutic withdrawal 8%, death 2% and incarceration 1% of the
terminations in the general Clinic I population. Among the 164 HD patients, there have been
only 23 terminations, for a 12-month termination rate of 14%. Again, the majority of
terminations were non-therapeutic (78%), with two therapeutic terminations and one each
incarceration, death, and transfer.
DISCUSSION In the 1980's, practitioners were under a great deal of pressure to apply restrictive
and punitive measures to addiction treatment. In MMT programs, contingency dosing and very
low ceiling doses were attempted and shown to be of suboptimal efficacy. Numerous studies
concluded that urine toxicology results positive for illicit opiates show robust inverse correlation
to methadone dose (Hartel et al 1995; Siassi, Angle & Alston 1977; Strain et al 1993). In our
sample 37% of patients in the control group (C), representing the general clinic population, had
positive urine toxicologies in the "after" sampling. This is consistent with the rate of illicit drug
use in MMTPs nationwide (US-GAO 1990). We identified a group of 164 patients who, despite
doses up to 100 mg/day, had a very high rate of illicit opiate use (87%). When these patients
were individually titrated to doses of methadone that resulted in no objective signs or subjective
symptoms of withdrawal, the percentage of patients with positive urine toxicologies dropped to
3% (see Figure 1), a highly significant change (p<.001). A 3% rate of illicit drug use is several
orders of magnitude lower than most clinics achieve (US-GAO 1990).
Another common outcome measure for MMT is retention in treatment. Nationally, the six-month
retention rate averages 54% (US-GAO 1990). There is a clear correlation between increased
methadone dose and increased retention in treatment (Capelhorn & Bell 1991). This association
of methadone dose with retention in treatment is strongly illustrated in our sample of HD
patients, 86% of whom were retained in treatment for a year or more. In this age of HIVspectrum disease, retention in treatment is a primary treatment goal. Discharging patients
because of illicit opiate use can not be justified.
All patients at CAP are educated about the need for adequate doses of methadone to eliminate
illicit drug use and craving. Patients with consistently positive urine toxicologies are referred for
medical evaluation by nursing and/or counseling staff. The patient is evaluated for signs and
symptoms of opiate abstinence syndrome (OAS), and dose adjustments are made accordingly. As
these dosing practices developed and became known throughout the clinic population, more
patients became willing to spontaneously report illicit drug use, including use of illicit
methadone. Unsurprisingly, patients were more likely to report illicit use when they received
appropriate treatment rather than punishment for their candor.
When we first started treating patients with doses exceeding 100 mg/d we were aware that we
were prescribing differently than most programs in the US. Consequently we performed serum
methadone levels (SMLs) often for the first year, and our clinical experience during this time
was consistent with the empirical findings of Hiltunen et al (1995). Our conclusions are more
clinical than empirical, as we did not do serial SMLs for the majority of HD patients, and we did
no SMLs for comparison purposes in the C group. A schematic representation of ur clinical
findings is presented in Figure 2. When the serum methadone level (SML) is very low (<200
ng/ml), most patients exhibit the familiar objective signs of OAS: mydriasis, rhinorrhea,
piloerection, arthralgia, and nausea/ vomiting/ diarrhea. As the dose of methadone is increased
and the SML rises above 200 ng/ml, these objective criteria for OAS disappear. This may be one
reason for reports that 100-150 ng/ml was an adequate SML (Bell et al 1988). However, patients
may continue to experience dysphoria, insomnia, anxiety, and craving. Further increases in
methadone dose alleviate these subjective symptoms. We found SML helpful in confirming
inadequate dose, but less helpful in determining adequate dose. Although 400 ng/ml is the lowest
SML that is usually effective in alleviating objective signs and subjective symptoms of OAS, ,
some HD patients had SMLs of 800 -1200 ng/ml when titrated to doses resulting in no signs of
intoxication, no symptoms of withdrawal, and no illicit substance use.
These higher-than-anticipated SML results may be due to differential effects and metabolism of
R and S enantiomers (stereoisomers) of methadone (Olsen et al, 1977; Eap et al, 1996). Recent
European work examining inter-individual response to dose in methadone-dependent patients
describes the R enantiomer as 50 times more biologically active than the S enantiomer (Eap et al
1998). Methadone products in the US are racemic (1:1 ratio of R:S enantiomers) but standard
tests for serum methadone levels report the sum of both enantiomers (R+S). Eap's group
measured serum levels of the R and S forms separately. They found seven-fold inter-individual
variability in the ratio of the enantiomers in the serum of different patients on equivalent doses.
Eap attributes 47% of the variability to heritable cytochrome P450 mechanisms.
Thus an adequate, or even toxic, SML may be reported for a patient whose serum level of active
R enantiomer is inadequate for maintenance. Clinically, this occurs in patients whose R:S
methadone ratios are 0.65 or less (Ortelli et al 1997). Other patients may have R:S ratios much
greater than 1, and will be clinically stable (no signs or symptoms of OAS, no illicit drug use)
with SMLs well below 400 ng/ml. Therefore, relying on a standard SML alone to determine
adequate methadone dosage will result in dosing errors in both directions for a significant
percentage of patients.
Currently, testing for serum R:S enantiomer ratios is relatively costly. Medical evaluation of
objective signs and subjective symptoms is a sensitive, reliable, and cost-effective method of
dose titration. Using these clinical indicators to determine adequate methadone dose results in
dramatic inter-individual variability in methadone dose. We have learned that continued use of
illicit opiates is the best indication for increasing methadone dose, and supersedes standard SML
results. For 53 (32%) of the HD group no SML determinations were performed. For many
patients (n=95; 58%) SML was not repeated after the initial determination at 100-120 mg/d. We
have become confident using the same criteria for higher-dose titration as we have always used
for routine dose titration: signs and symptoms of OAS.
Our original expectation was that some patients would seek intoxicating doses of medication.
This concern was rapidly dispelled. Not one patient continued to ask for increases in methadone
dose after reaching clinical stabilization. In fact, many patients were reluctant to accept the
increases that we recommended, and required support and reassurance to achieve an adequate
dose.
The United States GAO's findings about continued illicit drug use in MMT patients (US-GAO)
support our finding that, when allowed to set their own dose of methadone, patients tend to
maintain themselves at subtherapeutic doses. Significant numbers of patients continue to use
illicit drugs while saying "I don't want to go too high on the methadone". MMTPs may
discourage adequate methadone dose by imposing arbitrary dose ceilings, restricting privileges
based on dose, communicating prejudices regarding dosage, and supporting negative myths
about methadone.
Additionally, the disinclination to take any medication as prescribed is at least as prevalent in
methadone-maintained heroin addicts as in patients with other chronic diseases, such as diabetes,
schizophrenia, depression, and hypertension. Methadone is different from the medications used
for those illnesses; it is a medication encumbered with overwhelming stigma and significant
practical barriers to compliance. MMT patients need even more support and education for
adequate pharmacological maintenance than do patients with other chronic disorders.
Measurement of SML is very useful as an educational tool to encourage a fearful patient to
accept an adequate dose or to assure a less experienced colleague that a dose increase is justified.
We also measure SML when clinical signs and symptoms are unclear or conflicting. We rely
upon objective signs such as continued substance abuse and signs of OAS as indications of the
need for increased dosage. We have also learned to take patients' subjective complaints of
distress very seriously. We have confirmed that patients' reports of subjective symptoms are
usually reliable, and are a more appropriate criterion for dose titration than costly, invasive, timeconsuming, and (as described above) sometimes unreliable SML. Using these guidelines we have
been able to bring levels of illicit drug use in this high-risk population to unprecedented lows.
We found a significantly higher incidence of psychiatric diagnoses in the HD group (63%) than
in the C group (32%) (p < .001) (See Figure 3). This difference may be an artifact, as patients in
the HD group had more physician contact than those in the C group. Physician contact was
required for a dose increase over 100 mg/d and for subsequent increases. Probability of
recognizing a psychiatric disorder was therefore increased. The 63% incidence in our HD sample
is consistent with epidemiological studies of psychopathology in addicted patients (Abbott,
Weller & Walker 1994; Jacobs, Doft & Koger 1981; Milby et al 1996; Nunes et al 1994).
It is possible that patients treated with psychotropic medications could require methadone dose
adjustments due to drug interactions. Methadone is metabolized by the Cytochrome P450
enzymes CYP2D6 and CYP3A4, as are many serotonin-reuptake inhibitor antidepressants
(SSRIs), tricyclic antidepressants (TCAs), benzodiazepines, antipsychotics, and anticonvulsants
(Iribarne et al 1997). It is well known that certain anticonvulsants, particularly carbemazepine
and phenytoin, induce both the CYP2D and CYP3A enzyme families, (Panesar, Bandiera &
Abbott, 1996) resulting in clinically significant decreases in methadone level. However,
antidepressants (both SSRIs and TCAs) tend to be inhibitors of CYP2D6 and CYP3A4 (Sproule
et al, 1997). Benzodiazepines also tend to inhibit cytochrome P450 activity in vivo (Nims et al,
1997). This inhibition of enzyme activity would tend to result in an increase in methadone levels.
This is not consistent with our finding that psychotropic-treated patients tend to require higher
doses of methadone. Metabolic drug interactions do not seem to explain this finding. It is
possible that there is a correlation between Axis I psychopathology and high methadone dose
requirement. Such a correlation has been reported by Maremmani (1997). This study found that a
significantly higher mean dose of methadone was required to stabilize patients with psychotic,
bipolar, depressive disorders and alcohol abuse disorders than that required by patients with
heroin dependence alone.
Whether causally-related or not, the juxtaposition of psychiatric disorder and higher dosage
requirement allowed us to make some interesting observations regarding clinical management of
these dually-diagnosed patients. The use of opiate agonists as psychotropic agents is not new.
Methadone has established psychotropic efficacy (Musselman & Kell 1995). We observed a
subset the present group of patients whose chronic depressive disorders had been refractory for
years to many varied and heroic antidepressant regimens. When these patients' methadone doses
were titrated against their depressive symptoms, antidepressant efficacy was achieved. In some
cases antidepressant medications were able to be decreased or discontinued and in all cases the
efficacy of the antidepressants was greatly enhanced. It reminded us that methadone is a
powerful psychotropic tool, and we now use methadone more aggressively in the treatment of
psychiatric symptoms. In several patients, dysphoria and anxiety were the sole subjective
symptoms against which we titrated the last 10-15% of methadone dose. These phenomena
certainly deserve further study. Our clinical experiences with benzodiazepine abuse also bear
mention. Others have reported that methadone-maintained patients use benzodiazepines to
enhance methadone effects (Preston et al 1984). We now treat abuse of benzodiazepines in MMT
patients as a symptom of inadequate methadone dose. Titration of methadone dose against use of
benzodiazepines has been very useful in minimizing or eliminating abuse of benzodiazepines in
motivated patients. The physician initiates the cooperative process by prescribing
benzodiazepines ad libitum as long as the patient agrees to upward titration of the methadone
dose. The patient's benzodiazepine requirement (licit and illicit) is converted to a single long-
acting benzodiazepine, administered and monitored by medical staff. We have been able to
eliminate inappropriate use of benzodiazepines in many patients.
Our experience with alcohol abuse is similar to that with benzodiazepines. Most alcohol-abusing
patients find that increased methadone dose helps to curb alcohol cravings. In these patients we
titrate methadone dose against the patient's report of alcohol cravings, though dose titration may
be complicated by acute alcohol intoxication. Our clinic policy is to withhold methadone if the
breathalyzer reading exceeds 0.05. The efficacy of opiate substitution in relieving alcohol
cravings is not surprising, as the opioidergic mechanism of alcohol's action is well known
(Sinclair, Adkins & Walker 1973), and naltrexone, an opiate antagonist, is effective in the
treatment of alcohol use disorders. (O'Malley et al 1996; O'Brien, Volpicelli & Volpicelli 1996).
We have encountered very few patients whose alcohol misuse did not respond to methadone
dose titration.
Cocaine is more problematic. The literature provides conflicting reports regarding opiate agonist
efficacy in the treatment of cocaine abuse (Preston et al 1996; Rawson et al 1994; Tennant &
Shannon 1995). Although we do not have the data to conclude that methadone is an efficacious
treatment for cocaine abuse, we can report our clinical experience. In patients with primary
opiate dependence and secondary cocaine use, aggressive methadone titration combined with
antidepressant therapy can be helpful for motivated patients. Zubieta et al (1996) reported
upregulation of mu receptor binding in response to cocaine bingeing. This has two implications
for the methadone-maintained patient. First, it is evidence for opioidergic involvement in cocaine
dependence and craving. Secondly, it helps us understand the increased methadone requirement
seen in patients who are using cocaine. This elevated methadone dose requirement must be met
by appropriate increases in methadone dose. Retention in treatment is crucial to achieve any long
term improvement and retention in treatment is dependent upon adequate methadone dose.
The concept of opioid addiction as a metabolic disease involving deficiency of endogenous
ligand is not new. It was first suggested in 1966 in Dole, Nyswander, & Kreek's original article,
and has subsequently been discussed by many authors (Martin, Payte & Zweben 1991; Goldstein
1991; Kreek 1992). Most patients on an adequate dose of methadone can achieve a
pharmacologic approximation of the "normal" or "healthy" neurophysiologic state. If the dose of
methadone is inadequate, illicit supplementation will be necessary to approximate healthy
neurophysiology (See Figure 4).
In conclusion, we have found that the dose of methadone necessary for effective pharmacologic
treatment of opiate addiction has wider inter-individual variability than has been generally
acknowledged in most treatment settings. Patients treated with inadequate doses of methadone
commonly supplement their dose with illicit opiates, benzodiazepines, and alcohol. Treatment
response can be maximized by increasing the dose of methadone in response to patient reports of
illicit drug use and subjective distress. Patients do not request inappropriate increases of
methadone. We have yet to encounter patients whose illicit opiate use cannot be eliminated by
titrating methadone against symptoms of OAS and signs of drug abuse.