Unplanned admissions to two Sydney public hospitals after naltrexone implants

Naltrexone is an opioid antagonist licensed internationally as an oral preparation for relapse prevention in treating heroin and alcohol dependence. However, its effectiveness for heroin dependence is compromised by poor treatment adherence and high relapse rates, with outcomes not significantly better than placebo for most patients.1 This has led to the development of long-acting naltrexone products that seek to overcome problems of the patients’ poor motivation and adherence to medication. A depot intramuscular injection product is licensed in the United States for treating alcohol dependence,2-4 and a pilot study has reported favourable outcomes in opioid users.5

Several long-acting naltrexone implants have also been developed. The implant is inserted subcutaneously, typically in the abdominal wall, and releases therapeutic plasma levels of naltrexone for 3 to 6 months.6 The treatment process usually commences with “rapid” detoxification from opiates, often using naloxone and sedation. Studies of several hundred individuals have been reported.7-10 To date, no naltrexone implant has been approved for human use in Australia or internationally by regulatory bodies such as the Therapeutics Goods Administration (TGA).

Despite naltrexone implants being unlicensed, private clinics are delivering this treatment, and thousands of Australian cases have been reported.11,12 One mechanism enabling the use of unlicensed products is the TGA Special Access Scheme for “persons who are seriously ill with a condition from which death is reasonably likely to occur within a matter of months, or from which premature death is reasonably likely to occur in the absence of early treatment”.13 However, critics argue that naltrexone implants have not undergone the rigorous safety and efficacy licensing assessments required of medicinal products, and that this is of particular concern in view of their use in this vulnerable patient population.14,15 Public facilities may be reluctant to deliver this intervention in the absence of a licensed therapeutic product and because of concerns about minimal safety and efficacy data and medicolegal indemnity.

A key issue in the debate is the extent to which naltrexone implants are associated with significant adverse events. The previously cited naltrexone implant studies reported no severe adverse events. In contrast, another study reported five deaths in patients treated with naltrexone implants: three after the proposed duration of the implant had expired, one non-opiate overdose, and one opiate overdose despite an “active” implant.14 Another reported six severe adverse events: one case each of dehydration, seizure, pulmonary oedema, baclofen toxicity, benzodiazepine withdrawal, and death.16 It has been argued that these complications were probably due to the detoxification procedure rather than the implant, and that such complications are not seen in Australia where the implant is usually preceded by safer rapid detoxification.11,12 Nevertheless, it has been estimated that about 10% of the patients undergoing rapid detoxification and naltrexone implants at one clinic are referred to a local emergency department for assessment and management involving a “variable period of inpatient supportive care”.11

Recently, a series of adverse events associated with naltrexone implants resulting in hospital admission has been encountered in two Sydney public hospitals that warrants reporting, given the paucity of published severe adverse events associated with the controversial and unregulated use of naltrexone implants.

We report a case series of presentations to two teaching hospitals in Sydney associated with recent treatment with naltrexone implants. The series includes only patients referred to Drug and Alcohol Consultation–Liaison services at Liverpool Hospital (a 650-bed teaching hospital) and Royal Prince Alfred Hospital (a 700-bed tertiary care teaching hospital) over a 12-month period (August 2006 to July 2007), in which patients were noted to have recently had naltrexone implants inserted (within the preceding 6 months). This series does not necessarily reflect all implant-related presentations to the hospitals, as not all such patients may have been referred to the liaison services.

Hospital records were reviewed by an addiction medicine specialist, and key features of the presentation were deidentified and extracted with a structured data instrument agreed to by the authors. Clinicians involved in the management of each patient reviewed and confirmed the details of each case summary. The project was approved as a quality assurance project of the Sydney South West Area Health Service Ethics Committee. One further case, of a patient with a panic attack immediately before the implant procedure, was found (not included).

Twelve patients were identified (Box).

This case series identifies severe adverse events associated with the use of naltrexone implants, of particular relevance to staff in emergency departments, drug and alcohol services, pain management services, service providers engaged in delivering naltrexone implants, and regulatory agencies. These events challenge the notion that a naltrexone implant is a safe procedure, and indicate the need for careful case selection, careful clinical management and closer regulatory monitoring to protect this marginalised and vulnerable population.

Most of these cases (8/12) can be attributed to the naltrexone implant or implantation procedure. The case of Patient 8 was possibly implant-related, given the temporal sequence of events, and relapse of mental health conditions is a recognised complication of opiate withdrawal. The cardiac arrhythmia in Patient 9 was probably related to daily intravenous cocaine use.17 Three further cases were unrelated to the naltrexone implant procedure, but reflect difficulties in providing these patients with pain relief. As previously reported,18 non-opioid analgesia was effective where high opiate doses were not. However, a challenge for hospital staff is that patients reported that their implants had variable durations of action ranging from 3 to 12 months! Given that self-report is not always reliable, and that many patients presented after hours without referral and without reliable product information, it can be difficult to establish whether a naltrexone implant is still providing opioid blockade. Under such circumstances a trial of opioid analgesia (with close monitoring) may be warranted.

Clinical, regulatory and ethical implications arise from this case series. Patients should be warned of the associated risks, and appropriate procedures planned to respond to any complications. The emergence of severe psychiatric symptoms in one case highlights the importance of screening patients for underlying medical or psychiatric conditions, and importantly, coordinating with relevant service providers. Similarly, a close relationship between naltrexone implant providers and local emergency departments is important.11 This was not evident in most of these cases, and it is concerning that the management of expected complications after this private treatment was “unexpectedly” transferred to the public health system.

Further, many of the adverse events may relate to the method of induction of naltrexone using rapid opiate detoxification rather than problems of the implant itself. In future, this may be avoided by delaying implant insertion until the patient can tolerate a single oral naltrexone challenge following standard (or rapid) opiate detoxification. Ultimately, uncoupling rapid detoxification from naltrexone implant treatment is not only required for regulatory and clinical assessment of safety, but would make any future treatment with naltrexone implants less expensive and more accessible to patients.

We are unable to estimate the incidence of severe adverse events associated with naltrexone implants. It is clear from discharge records that no consistent diagnosis-related group or International classification of diseases, 10th revision (ICD-10) code used for these admissions would enable such patients to be identified through an electronic search of hospital records. Furthermore, there was no documentation in the clinical notes that these cases had been reported to the TGA Adverse Drug Reaction Reporting System, suggesting an under-reporting of implant-related adverse events. The presentation of at least 10 cases to one hospital in a 12-month period suggests that adverse events with naltrexone implants are not rare or uncommon.

Controlled trials are required to show the safety and efficacy of naltrexone implants. It may be appropriate to consider unlicensed treatments with a limited evidence base under special access schemes if a patient has a life-threatening condition and has not benefited from or cannot tolerate conventional treatment. It is not apparent how all of these cases met these criteria, as several patients reported no previous alcohol and drug specialist treatment, and others were actively enrolled in effective methadone programs. The widespread and unregulated use of naltrexone implants without appropriate safeguards for patients, their families and service providers should be restricted until this therapeutic product has been assessed for safety and effectiveness.