22 September 2005

Naltrexone implants: Claims and counter claims: None so blind

Addiction Biology 2005 10:201-204


Letters to the Editor


Naltrexone implants as treatment for heroin dependence:



Contents of this post





Part I of letter


There are a number of important methodological problems with this recent study of naltrexone implants (Hulse et al., 2004a). The selection of patients was opportunistic. No inclusion and exclusion patient selection criteria are provided. The authors do not state exactly how many specimens were obtained from each of the patients with either the 1.7g or the 3.4g implant, and the testing was not done according to any fixed protocol.
The most serious problem is the presentation of the data. Only blood concentrations �standardised� to 70kg body weight are reported, apparently to compare the two sizes of implant. It is, however, misleading to use plasma concentrations standardised to 70kg bodyweight in discussing the minimum effective concentration (2ng/ml). In the examination of kinetic data, relationships between plasma concentrations and body weight can be assessed, but actual plasma concentrations should have been reported, as has been done in the other reported studies on sustained release naltrexone preparations (Comer et al., 2002; Olsen et al., 2004). Concentrations �standardised� for bodyweight are not a useful clinical measure, and this study provides no helpful information into how to adjust the implantation protocols for bodyweight.
A high proportion of bodyweight-standardised blood concentrations reported in this study were below 2 ng/ml. Further, in a second study of sequential naltrexone implants (Hulse et al., 2004b), two out of five patients are reported as having plasma concentrations, standardised to 70kg body-weight, below 1ng ml for at least some of the time. Based upon the data presented in these studies it appears likely that blood concentrations were inadequate in several patients. Without actual concentration data, however, these studies allow no conclusions to be drawn about the proportion of subjects who had naltrexone �above therapeutic levels�.
These flaws, especially in the data analysis, make it very difficult to draw conclusions about the clinical performance of naltrexone implants in these studies.

Dr Andrew Byrne* MB BS FAChAM, Professor Garry Graham�, Dr Richard Hallinan* B, Dr Bridin Murnion BSc, MBChB


* Dependency physicians, 75 Redfern St, Redfern, New South Wales, Australia

# Department of Clinical Pharmacology and Toxicology, St Vincent�s Hospital, Sydney, Australia


References


Hulse GK, Arnold-Reed DE, O�Neil G, Chan C,T, Hansson R, O�Neil P. (2004a) Blood naltrexone and 6-�-naltrexol levels following naltrexone implant: comparing two naltrexone implants. Addiction Biology 9:59-65
Hulse GK, Arnold-Reed DE, O�NeiI G, Chan C-T, Hansson R. (2004b) Achieving long-term continuous blood naltrexone and 6-�-naltrexol coverage following sequential naltrexone implants. Addiction Biology 9:67 72
Comer SD, Collins ED, Kleber HD, Nuwayser ES, Kerrigan JH, Fischman MW. (2002) Depot naltrexone: long-lasting antagonism of the effects of heroin in humans. Psychopharmacology (Berl) 159:351 60
Olsen L, Christophersen AS, Frogopsahl G, Waal H, Morland J. (2004) Plasma concentrations during naltrexone implant treatment of opiate-dependent patients. BrJ Clin Pharmacol 58:219



Reply to Part I:


I write in response to Dr Byrne et al�s letter commenting on our recent publication (Hulse et al., 2004) in Addiction Biology. First, Dr Byrne comments on the vagueness of patient selection and study design. This is despite the methods section clearly stating that the study was a retrospective review of blood sample results from clinical records with subject selection, data inclusion criteria and number of samples collected per patient also delineated in the manuscript.
Secondly, with regard to Dr Byrne�s comments on the "serious" flaws regarding the validity of using body-weight standardised blood concentrations of naltrexone, we feel that this is open to interpretation. Our original manuscript submitted to Addiction Biology reported non-standard blood levels (as advocated by Dr Byrne). However, the standardisation was made in response to the journal reviewer�s comments on our paper. The rationale of the reviewer and journal was that as all subjects received a fixed quantity of naltrexone in their implant, if non standardised values were used then we would have to reconcile larger body mass (male) subjects with a lower circulating blood naltrexone concentration with smaller body mass (female) subjects with a higher blood naltrexone concentration when estimating longevity of implants. To overcome this we standardised blood concentrations to a body weight of 70kg. Exponential curves were fitted to the standardised values based on a least squares best fit method. These calculations take into account the variation in blood naltrexone levels between subjects.
Lastly, this paper is a research article and was not written to provide a clinical framework for the use of implants. It is therefore erroneous to attempt its use as such.

Prof G K Hulse, Dr D E Arnold-Reed


School of Psychiatry and Clinical Neurosciences, University of Western Australia, 35 Stirling Highway, Crawley, WA 6009, Australia


Reference


Hulse GK, Arnold-Reed DE, O�Neil G, Chan C,T, Hansson R, O�Neil P. (2004) Blood naltrexone and 6-�-naltrexol levels following naltrexone implant: comparing two naltrexone implants. Addiction Biology 9:59-65



Part II of letter


We have serious ethical concerns about a recent paper (Hulse et al., 2004) reporting naltrexone and 6-�-naltrexol levels following naltrexone implant.
In this paper, the authors have not followed the generally accepted sequence of studies in drug development. They have commenced a Phase II study before they (or others) have reported Phase I studies. Researchers in Phase I clinical trials test a new drug in a small group of people (generally 20�80) for the first time to evaluate the safety and likely safe dosage range and to identify side effects. In contrast, researchers in Phase II clinical trials give the study drug to a larger group of people (100�300) to see if it is effective and to further evaluate the safety of the drug.
Phase II studies of disulfiram implants were also conducted before Phase I studies had been undertaken (Johnson and Morland, 1991; Konoplitskaya et al., 1991). The Phase II trials were unsuccessful because extensive fibrosis sealed off the implants. Had Phase I studies been conducted first, this would have been detected and ineffective Phase II studies would have been thus avoided. Ideally, Phase I trials should always be prospective and this is a retrospective study.
There were other methodological problems including opportunistic selection of patients, no published inclusion and exclusion patient selection criteria and lack of clarity about the number of specimens obtained from each of the patients with either the 1.7g or the 3.4g implant.
The authors state that they have ethics committee permission to review patient records and to perform the study. However, they do not state explicitly that they have ethics committee approval for the novel naltrexone implant itself. The authors do not refer to any protocol to detect, record and consider adverse effects. The lack of such a protocol may well have led to an underestimation of adverse effects. Serious adverse events with implanted naltrexone have been reported (Hamilton et al., 2002). The authors do not provide or refer to any plan of action in the case of disease or trauma that would normally require standard opioid agonist therapy.
The treatment in this study has been approved by the Therapeutic Goods Administration (TGA), Australian Government Department of Health and Ageing, under the Special Access Scheme (SAS) (http: www.tga.gov.au does html sasinfo.htm). This scheme enables medical practitioners to supply unapproved medications to some very seriously ill patients under clearly specified conditions on a case-by-case basis. However for trial purposes, use of this therapy should properly be approved under the Clinical Trial Exemption Scheme (CTX) or the Clinical Trial Notification Scheme (CTN) under the Australian NHMRC guidelines (http: www.tga.gov.au does html clintrials.htm). In Australia, naltrexone-implants are required to be stamped with a notice in large print warning against use in human subjects. It is not clear if patients enrolled in this study have been informed about this notice.
The declaration of interest states that �one of the authors (G.O�N.) is a director of GoMedical, the medical device company which manufactures the implants�. Yet this declaration states that this author only �would be expected to receive pecuniary benefit (amount unspecified) from GoMedical in his role as surgeon performing the implants�. The declaration does not state whether or not any of the authors stand to gain from marketing the devices.
Studies of naltrexone implants should meet the same high and scientific standards as other studies evaluating novel medical treatments for the same condition. We believe that this paper does not meet the same standards now accepted for the evaluation of substitution treatment for heroin dependence.

Dr. Alex Wodak, Dr Robert Graham


Alcohol and Drug Service, St. Vincent�s Hospital, Darlinghurst, NSW, 2010, Australia


References


Hulse GK, Arnold-Reed DE, O�Neil G, Chan C-T, Hansson R, O�Neil P. (2004) Blood naltrexone and 6-�-naltrexol levels following naltrexone implant: comparing two naltrexone implants. Addiction Biology 9:59-65
Konoplitskaya KL. Pkhakadze GA. Narazayko LF. (1991) Biocompatibility of a prolonged-action anti-alcohol preparation. Biomaterials 12:701 704
Johnsen J. Morland J. (1991) Disulfiram implant: a double-blind placebo controlled follow-up on treatment outcome. Alcoholism: Clinical & Experimental Research 15:532 536
Hamilton RJ, Olmedo RE, Shah S, Hung OL, Howland MA, Perrone J, Nelson LS, Lewin NL, Hoffman RS. (2002) Complications of Ultrarapid Opioid Detoxification with Subcutaneous Naltrexone Pellets. Academic Emergency Medicine 9:63 68


Reply to Part II:


Doctors Wodak and Graham are correct in their assertion that the development and registration of pharmaceutical drugs commonly involves the successive testing of the drugs in phase I prior to phase II and subsequently phase III trials. There is however, within many countries, the ability for physicians under clearly specified conditions on a case-by-case basis to prescribe or use non-approved pharmaceuticals for some seriously ill patients. Legislation that allows this commonly holds that prescribing of non-registered pharmaceuticals can occur where there is a high risk of mortality associated with the morbidity under treatment. Perhaps the most recognised instances of this supplement use is oncology, where use of cancer treatment drugs, which have been successfully used overseas, but not registered in the practitioner�s country, might be used by the practitioner to treat a cancer. Clearly where this takes place, it is important to evaluate outcomes associated with these treatments.
As noted by Drs Wodak and Graham, this legislation in Australia is known as the Special Access System (SAS) and administered by the Commonwealth Therapeutic Goods Administration (TGA). Given increased risk of mortality amongst heroin users, this SAS has been used by a number of Australian community and hospital based physicians in at least 4 of the 6 Australian States to administer sustained release naltrexone preparations. These preparations have included a number of those developed in the US with a likely clinical pharmacokinetic life of 6-8 weeks and the locally Australian produced GoMedical implant. For the last 5 years the TGA in Australia has continued to allow naltrexone implants to be administered under SAS. It is estimated that in the past few years, in excess of 3,000 heroin dependent persons have been treated by this method in the States of Victoria, New South Wales, Queensland and Western Australia.
Clearly, as with the cancer scenario, it would be foolhardy and I suggest irresponsible not to superimpose some type of monitoring and evaluation on these patients. The result of analysing naltrexone blood samples from a cohort of SAS treated patients was in fact the basis for the two papers published in Addiction Biology. As such, this was not a phase II drug development study. This is made clear in the paper, in that Drs Wodak and Graham should seek to suggest otherwise is misleading. Lack of methodological clarity raised by the two doctors is also misleading, given that this again this is clearly delineated in the paper. Clearly an international journal such as Addiction Biology has in place checks and balances to ensure that this type of information is made available to the readers and they and the paper�s separate and independent assessors found it adequate.
Drs Wodak and Graham refer to the lack of protocol to detect, record and consider adverse events. This is the type of protocol that would be required where a randomised clinical trial was in place, yet these were SAS not clinical trial treated patients. In the context of SAS patients, monitoring of adverse outcomes is the responsibility of the treating physician, and I would hasten to note that the TGA has mechanisms for collecting this adverse information from SAS treated patients. In making their case the doctors refer to "serious adverse events associated with implant naltrexone." (Hamilton et al., 2002). Of course those who are familiar with this literature will recognise that this article refers to morbidity and mortality in the US resulting from the use of implants to precipitate rapid opiate detoxification and not sustained release naltrexone for the purposes of naltrexone maintenance. This typifies Drs Wodak and Graham�s letter with subtle non-truths to create illusion to support a fictitious case; words " twisted by knaves to make a trap for fools". There is no place for this type of misrepresentation and Drs Wodak and Graham stand reprimanded for this deceptive behaviour. This includes both the misrepresentation of the aforementioned article and the attempt to present SAS patient data collection as a clinical trial.
Questions are also raised about the statement of declaration and pecuniary interests of the authors in marketing the implant. I would hasten to note that the wording of this disclaimer was superimposed by the Journal and perhaps any issues of concern should be raised with them. However, for the record, we state categorically that neither I nor any member of my research team has, or will receive pecuniary interest in marketing of this implant product. We are a University clinical research facility and our sole objective is the unbiased reporting and evaluation on clinical activities. For Drs Wodak and Graham to suggest otherwise is both disturbing and a personal affront.
Of course, the reader could be forgiven for believing that all this information will be new to Drs Wodak and Graham, but alas, I understand they are already conversant with all the facts. That these two doctors suggest that evaluation of SAS patients treated with naltrexone implants should not take place is to me not only abhorrent, but lacking in vision. In fact the opportunity to monitor and access valuable information from SAS treated patients with implant naltrexone in Australia has been recognised by the Commonwealth Department of Health and Human Services, who in consultation with the TGA (a subsection of this Branch) identified additional funding and identified areas of information to be garnished from these SAS patients. This includes histological information on tissue reactivity around the site of implant, ultrasound to assess biodegradability and I might add additional pharmacokinetic blood collected similarly to that published in Addiction Biology. Similarly, the National Health and Research Medical Council (NH&MRC) has recently funded for hospital morbidity, mortality and mental health data to be collated from a large cohort of these SAS treated patients. Ultimately, this information, coupled with more data collected from randomised double blind clinical trials will be used to assess the viability of the Australian implant for transTasman registration. I also understand from the Australian Commonwealth Department of Health that as part of this assessment, the National Drug and Alcohol Centre in Sydney, of which Dr Wodak is an adjunct senior lecturer, has been or will be funded to garnish morbidity and mortality data from similarly treated SAS patients. This would put Dr Wodak�s facility in exactly the same position as my research team, yet Dr Wodak makes no reference to this.
It would appear that Drs Wodak and Graham are not only at odds with myself, but also the Commonwealth Department of Health, the Therapeutic Goods Administration and the NH & MRC in their objection of data being sought from SAS treated patients. For clinical researchers to not garnish valuable information from patients treated by new methods under the SAS legislation would be a tragedy, yet for some, this simple concept seems incomprehensible. "There are none so blind as those who will not see."

Prof G K Hulse


School of Psychiatry and Clinical Neurosciences, University of Western Australia, 35 Stirling Highway, Crawley, WA 6009, Australia


Reference


Hamilton RJ, Olmedo RE, Shah S, Hung OL, Howland MA, Perrone J, Nelson LS, Lewin NL, Hoffman RS. (2002) Complications of Ultrarapid Opioid detoxification with Subcutaneous Naltrexone Pellets. Academic Emergency Medicine 9:63 68