21 December 2010

Say no to "just say no"! Give in, with therapeutic strings.

Article printed in "OF SUBSTANCE".

"View from the coal face" … in Redfern, inner Sydney.

Commentary on Addiction editorial on benzodiazepine maintenance.

After many years of wrestling with the problem of benzodiazepine use in opioid dependency patients it was reassuring to read the prominent paper by Liebrenz and colleagues. Their hypothesis is an approach using what appear to be harm reduction principles, parallel to methadone maintenance. Our original practice policy was to ‘just say no’ but despite our entreaties, about one third of our patients continued to use benzodiazepines on urine testing. A number did succeed at abstinence, only to relapse with significant harms occurring due to disinhibited behaviour, often involving amnesia for the events.

Some patients were able to function almost normally while taking illicit benzodiazepines. Others became disorganised regarding their finances, housing and interpersonal relationships, some even coming to the attention of the police or emergency departments.

Although there appeared to be a number of patters of tranquillizer use, from binge and recreational use to quasi-therapeutic, we treated all such patients the same way initially, using diazepam 5mg tablets supervised at the clinic. Those currently abusing alcohol were excluded. Each patient needed to return at least once, about 3 hours after a witnessed dose for a brief examination to confirm their tolerance and exclude intoxication. All patients also had to agree to random urine testing and regular medical consultations to assess progress.

Our impression has been that when given access to diazepam under close supervision, stability returned to most such patients. A recent audit of our referral dependency practice showed that of 167 pharmacotherapy patients, (80% methadone, 20% buprenorphine) 30% were being prescribed benzodiazepines, mostly under supervision at the practice. The mean dose was 14mg daily (range 2mg - 25mg). One third were gainfully employed.

Thus we can confirm that some of the protocols alluded to in the forward thinking item in Addiction are feasible and are ripe for research. Enquiries showed that many of our colleagues had one or two pharmacotherapy patients taking long-term benzodiazepines and nearly all had organised supervised dosing at least once.

Benzodiazepine use has been the ‘elephant in the room’ in addiction treatment. While many centres still use an abstinence approach, many patients continue to use these drugs. Since benzodiazepines, along with alcohol, constitute a major source of drug-related harm it may be timely to reassess our approach. Severe restrictions on supply alone have historically never solved drug problems. Such restrictions also necessarily reduce access to those who need the drugs therapeutically. As with many other areas of public health, we believe that it is possible to translate the principles of ‘harm reduction’ to benzodiazepine use by utilizing the protocols of ‘universal precautions’ espoused by Dr Gourlay in Canada.

The use of benzodiazepine maintenance is probably at the same stage of ‘evidence’ as methadone treatment was in about 1980. It appears to be acceptable to the patient population; it appears to be safe in practice, yet definitive research is awaited to prove its effects … and to identify optimal dosing, supports and necessary supervision. Likewise oral morphine, injectable methadone and heroin assisted treatments are being trialled in several countries currently. Thus in our own patient group we found that supervised, low dose diazepam was worth offering to those who were unable or unwilling to give up benzodiazepines.


Liebrenz M, Boesch L, Stohler R, Caflisch C. Agonist substitution-a treatment alternative for high-dose benzodiazepine-dependent patients? Addiction 2010 105;11:1870–1874

Gourlay DL, Heit HA, Almahrezi A. Universal precautions in pain medicine: a rational approach to the treatment of chronic pain. Pain Med (2005) 6;2:107-112

Darke S, Ross J, Mills K, Teesson M, Williamson A, Harvard A. Benzodiazepine use among heroin users: Baseline use, current use and clinical outcome. D&A Review 2010 29:3:250-255

Byrne A. Benzodiazepines: the end of a dream. Aust Fam Physician 1994 23;8:1584-1585

See article summary by Libby Topp http://www.ofsubstance.org.au/images/archive/pdf/ofsubstance_2010_3.pdf

17 December 2010

More drop-outs with buprenorphine in pregnancy.

Reduced retention makes buprenorphine a poor second to methadone in pregnancy.

Jones HE, Kaltenbach K, Heil SH, Stine SM, Coyle MG, Arria AM, O'Grady KE, Selby P, Martin PR, Fischer G. Neonatal Abstinence Syndrome after Methadone or Buprenorphine Exposure. NEJM 2010; 363:2320-2331

Dear Colleagues,

These authors assessed 1074 pregnant women from clinics in Austria, Canada and the USA for this comparison of methadone and buprenorphine regarding neonatal abstinence syndrome (NAS). There were 250 who declined to participate and another 650 did not fit inclusion criteria (alcohol, benzo use, medical conditions, for example) leaving 175 in the randomised comparison groups, blinded by a morphine wash-out period in hospital and use of ‘double dummy’ placebo. Pure buprenorphine (Subutex) was used to avoid prenatal exposure to naloxone (Suboxone is contraindicated in pregnancy). Over $1000 was available to subjects contingent on clear urine toxicology during the trial.

There were two peri-natal deaths reported from those who remained in the study, one from each group. There were two miscarriages (both in the MMT group). “The percentage of neonates requiring NAS treatment did not differ significantly between groups (P = 0.26), nor did the groups differ significantly with respect to the peak NAS score (P = 0.04) or head circumference (P = 0.04).” However, two other primary outcomes were significantly better in the buprenorphine group: (1) quantities of morphine used for neonatal abstinence syndrome (NAS) and (2): ‘total hospital stay’. The prognostic significance of these outcomes is unknown to my best knowledge.

The numbers of enrolled subjects were not large enough to show anything but very major outcome effects as being statistically significant. Previous experience has consistently shown only modest differences between methadone and buprenorphine regarding neonatal outcomes. Hence this trial was under-powered on the numbers to find minor differences to statistical significance.

Aside from the unsurprising neonatal outcomes, there were some dramatic and remarkable maternal findings which are mentioned but not brought to the prominence they deserve in my view. These differences are so great that they may even invalidate the neonatal findings (for example if those with high tolerance or rapid metabolism were more likely to drop out).

To my mind the two most important findings of the study are as follows:

The methadone group (n=89) had 16 drop-outs (18%) while the buprenorphine group (n=86) had 28 drop-outs (33%) [p=0.02]. Even more dramatic, of the 16 methadone drop-outs, 2 were due to ‘dissatisfaction with the medication’ while the corresponding number for the buprenorphine candidates was 20, a massive difference. These findings are discussed in the text but should probably also be enshrined in the title of the article. There is no point in having favourable neonatal outcomes if a third of the mothers have left treatment before term.

Even with its failings, I believe that this study provides the most persuasive evidence to date showing the safety and effectiveness of methadone in pregnancy. Indeed, it clearly demonstrates that, in the absence of contraindications, methadone should be the first line drug for opiate dependence in pregnancy (as in the non-pregnant). Yet despite their finding that almost twice as many women dropped out in the buprenorphine group, these authors state, somewhat clumsily: “Although there were no significant differences in overall rates of NAS among infants exposed to buprenorphine and those exposed to methadone, the benefits of buprenorphine in reducing the severity of NAS among neonates with this complication suggest that it should be considered a first-line treatment option in pregnancy.”

Rather than an option, “first line” status implies an obligation in my book. Since buprenorphine is the only alternative medication licensed for opioid dependence then it is obviously an ‘option’, if a somewhat less effective one. This RCT confirms that during pregnancy, consistent with the existing body of research evidence, (pure) buprenorphine should be the second line drug and only used when methadone is found to be clinically inappropriate. I believe that it is unethical to prescribe Suboxone (buprenorphine/naloxone).

Comments by Andrew Byrne ..


15 December 2010

French study shows torsade rare to vanishing.

Perrin-Terrin A, Pathak A, Lapeyre-Mestre M. QT interval prolongation: prevalence, risk factors and pharmacovigilance data among methadone-treated patients in France. Fundam Clin Pharmacol. 2010 Sep 6;

Dear Colleagues,

These French authors have done our field a great service by replicating Anchersen’s national study from Norway which was so reassuring concerning the use of methadone in addiction treatment.

This study examined QT intervals in a small sample of methadone patients and then compared the results with national reports on adverse events involving methadone and/or cardiac issues including sudden death. These authors found a mean QT interval of 414 ± 29ms with no readings over 500ms, the level at which the risk of arrhythmia is believed to become significant. We are told that analysis of the 42 cases showed that longer QT intervals were associated with recent increases in methadone dose, tobacco smoking, the use of other medications and pre-existing cardiac disease. These findings are consistent with most other work published on this matter, with the notable exception of Wedam. The authors quote QT ‘dispersion’ which is the difference between the shortest and longest QT interval among the twelve leads readings on the standard cardiograph. Despite some initial promise, the significance of ‘dispersion’ has been questioned more recently and its relevance to methadone treatment is uncertain. Even normal values are not agreed upon by cardiologists.

The 550 reports to the official French ‘pharmcovigilence’ centre regarding methadone are of great interest and relevance to the current debate over supposed cardiac effects of methadone. In the ten years to 2007 only 5 reports (0.9%) involved QT problems, three of these with torsade de pointes including one death, a 19 year old who died after unsuccessful resuscitation. This case is not tabulated as having torsade but if he did, as implied in the text, he would represent the first confirmed death from torsade de pointes in the literature to my knowledge.

Over the ten years there were also 7 sudden deaths (1.3%) which the authors tabulate in detail. For some reason they postulate that some or all might be ascribed to arrhythmia. They state: “it is conceivable that serious toxicity might be mediated in part through cardiac effects rather than solely via respiratory depression.” Almost anything is ‘conceivable’ yet the real quesion is whether it is likely or even ‘credible’. In fact the data presented make it highly unlikely that torsade is involved to any significant degree, if at all. Were torsade the cause of just half of these deaths, and considering a reported mortality well under 10% (some say zero) then one would expect hundreds of non-fatal cases of torsade across France - yet only two were ever reported in the whole country over a ten year period. The expected under-reporting in such a voluntary scheme would apply to some extent to both mortality and torsade, even though the latter has become almost ‘notorious’ and was originally described in France and with a French name.

Further, these authors state of several deceased cases: ‘no overdose of methadone’. This is largely based on drug levels found at post-mortem. However, these are now known to be most unreliable alone in determining cause of death. Indeed, serious toxicity and death from narcotism have frequently been found with post-mortem blood methadone in the ‘therapeutic’ range while some patients in normal treatment have levels in or near the toxic range. Also, there are major changes in the levels of measurable methadone following death.

It seemed extraordinary that five of the deceased patients were being treated for psychosis so I wrote to the authors. I have now learned that certain antipsychotics are used commonly in France as alternatives to benzodiazepines, hence not all of these patients were schizophrenic. Each of these 5 patients who were prescribed anti-psychotics died in the first 5 days of methadone maintenance treatment for addiction and two were taking doses which were in excess of current guidelines. Nearly all were receiving four or more prescribed medications apart from the methadone. HIV drugs were involved in two cases. Hence these well-documented deaths also make it clear that polypharmacy is a major factor and the patients would have been self-evident as extremely high risk candidates for treatment.

This study makes it hard to understand the findings of Krantz who found more cases in his own small district than in the whole of France in a decade. Suffice it to say that the findings are even more reassuring than the national figures from Norway published by Anchersen and colleagues.

Comments by Andrew Byrne ..

Related article (includes one of the same authors): Molokhia M, Pathak A, Lapeyre-Mestre M, et al. Case ascertainment and estimated incidence of drug-induced long-QT syndrome: study in Southwest France. Br. J. Clin. Pharmacol. 2008 66:386–395

In this exhaustive paper using a rational methodology an estimate is made of the incidence of drug induced long QT syndrome (LQTS) at 11 per million per year. The literature shows that over 50% are due to class III antiarrhythmics (presumably in those with existing cardiac rhythm disturbances), closely followed by anti-infective drugs and antihistamines (with presumably some antipsychotics). The drug methadone is not even mentioned in the entire paper so we presume that in 2008 it was not even on the radar for these experienced and thorough researchers. This casts yet further doubt of the contention of Dr Mori Krantz that cardiac problems constitute a public health priority in methadone treatment in America.

14 December 2010

QT changes in babies aged one day uncertain relevance.

Parikh R, Hussain T, Holder G, Bhoyar A, Ewer AK. Maternal methadone therapy increases QTc interval in newborn infants. Arch Dis Child Fetal Neonatal Ed 2010 doi:10.1136/adc.2009.181701

Dear Colleagues,

These authors set out to measure the QT interval in newborns for the first week of life from both ‘uncomplicated’ methadone mothers and ‘healthy’ or drug-free controls. The finding was that there was a small but significant increase on days 1 and 2 (~30ms) which was absent by days 4 and 7. On the first day of life there were 4 of 26 methadone exposed babies with QTc of greater than 500 yet these were all outliers, the 5th longest interval being less than 460ms. By day two only one was near 500ms while later readings were all below these levels. There were no cases of torsade de pointes nor any other rhythm disturbance.

As with other reports, the ability of experts to read QT intervals was limited. The inter-observer and intra-observer variations were given as minus 14 to plus 21ms. Hence it appears that a 30-40 ms difference in QT interval is a rather imprecise figure, since these 95% confidence limits are so wide.

A single case study by these authors in 2007 indicated movement of methadone across the placenta and changes in QT in the newborn, something which is hardly surprising but of unknown clinical significance.

I was not sure whether or not I should bring this to the attention of a wider readership, so slanted is the emphasis of the research and so lacking is it in practical clinical relevance. Ever since the commencement of the campaign to talk up the relevance of QT changes in methadone patients, new and supposedly ever more worrying facets of the problem have been ‘exposed’ - most recently questing whether testosterone levels are the cause! Or that racemic methadone was the problem and levo-methadone the solution (see refs below).

The premise here again seems to be that there are unexplained sudden deaths in methadone patients and their offspring and that a proportion of these may be due to torsade de pointes arrhythmia. Yet such a death has never been reported to my best knowledge. While some still-births or SIDS cases may possibly be due to torsades, the proportion must be extremely low owing to the paucity of reports of non-fatal cases (I could only find one confirmed case from the 1970s and it is reported that torsade mortality is very low or even zero).

Regarding adult cases, despite an aging population with a high rate of other serious illnesses and drug taking, reports of torsade de pointes arrhythmia in methadone patients continue to be sparse indeed. Furthermore, nearly all can be linked to significant risk factors other than (or as well as) the methadone. It is still possible that methadone actually lowers the rate of torsade de pointes - but only large prospective studies could prove that point … and such work would be impractical and expensive (and very probably unethical, considering the proven benefits of methadone treatment both for pregnancy and other outcomes).

Mori Krantz wrote that cardiac safety (in adults) was now a ‘national priority’. The references he used to support his thesis of increasing unexplained deaths in methadone patients (Balesteros; Sorg; Gagajewski; Shah) describe precisely the opposite on my own careful reading (none in Ballesteros; one from Sorg; none in Gagajewski; reducing, not increasing death rates in Shah’s report from New Mexico).

The main issue with the present item is balancing the small risks of methadone in pregnant women with the enormous risks of street heroin to the mother and baby. While a good alternative, buprenorphine is still less effective and technically not approved in pregnancy.

So in this case, as for adult opioid treatment, research energies have been devoted to a problem which is largely theoretical. Equally, we have seen new ‘guidelines’ and recommendations promulgated by health authorities, Colleges, hospitals, etc, each addressing an almost non-existent ‘problem’. One can only speculate at the reasons behind such moves concerning an established, effective drug. The same thing is happening for propoxyphene (Darvon, Digesic, Doloxene) which has just been withdrawn in America based on limited evidence of potential harm. This is against the actual evidence of 50 years of safe and effective use across the world as a second or third line opiate analgesic. According to some authors denigrating old drugs is a time-honoured tactic of drug companies to promote profits derived from more modern, patented and often very expensive drugs. I hasten to state that there is no evidence of this occurring with the current case.

The final line of this abstract says it all: “Bradycardia, tachycardia or an irregular heart rate in an infant born to a mother on methadone treatment should prompt investigation with a 12-lead ECG.” I would be concerned about any baby with an atypical pulse, NOT JUST the babies of mothers taking methadone.

Comments by Andrew Byrne ..


Daniell HW. Does Methadone Prolong QTc Intervals by Depleting Testosterone Levels? Arch Int Med 2010 170;15:1407-8

Ansermot N, Albayrak O, Schlapfer J, et al. Substitution of (R,S)-methadone by (R)-methadone: impact on QTc interval. Arch Intern Med. 2010;170(6):529-536

Dr Andrew Byrne MB BS (Syd) FAChAM (RACP)
Dependency Medicine,
75 Redfern Street, Redfern,
New South Wales, 2016, Australia
Email - ajbyrne@ozemail.com.au
Tel (61 - 2) 9319 5524 Fax 9318 0631 NO MOBILE

I am a deeply religious non-believer - this is a somewhat new kind of religion.
Albert Einstein d.1954. Me too! Andrew Byrne b.1954.