Double trouble—pregnancy and antagonist treatment in opioid dependence; two contentious issues needing further consideration and research

The stimulating paper by Jones et al. 1 highlights two areas of contention in the field of addiction medicine, namely the management of substance use in pregnancy and the potential role of opioid antagonists in the treatment of opioid dependence, particularly in comparison to opioid agonist maintenance treatment (OAMT; sometimes referred to as as opioid substitution treatment (OST)). The result of combining these two contentious issues into one research proposal may meet unpredictable responses from any given human research ethics committee, and the way forward does not seem at all clear. Separating out these two issues might help to clarify the next steps. While somewhat counter-intuitive and a little challenging to those outside the field of addiction, it is now fairly uncontroversial within the profession that one of the main tasks of treating people with opioid dependence (pregnant or otherwise) is to help them understand the risks of seeking abstinence in the short term, and to understand the likely benefits of a significant period of OAMT 2. Given that in RCTs of adequate sample size, OAMT outperforms all other treatments for opioid dependence, it has come to be regarded as the ‘standard of care’ for the medical treatment of opioid dependence, both in and out of pregnancy, and that to which other treatments should ordinarily be compared. There is less agreement, particularly when it comes to pregnancy, on what to do when OAMT is not available (as is the case in many countries and many rural areas) or when women make informed decisions not to take OAMT, as is their right. While some commentators would suggest that opioid detoxification in pregnancy is ‘contra-indicated’ 3, most recent guidelines make limited exceptions for its use 4-6. Certainly there are increased risks of detoxification in pregnancy. There is clear evidence that opioid withdrawal can put the fetus under stress, and this has resulted in cases of first-trimester miscarriage and third-trimester premature labour, a small number of which are often quoted in the literature. The risks of medication-assisted detoxification are perhaps not as great as was thought earlier, however, with larger case-series demonstrating similar rates to the general population of both miscarriage and premature labour 7. Most guidelines make no mention of the advisability or otherwise of accelerated detoxification regimens (which are sometimes used when commencing antagonist treatment), whether buprenorphine can or should be used for detoxification or how to monitor for fetal distress. There is also conflicting advice on whether buprenorphine should be used for maintenance treatment in pregnancy. What we do find are recommendations to refer to specialized obstetric substance abuse services, although these services are few and far between. In defence of those guideline groups, there is very little evidence upon which to base more detailed guidelines, and this is essentially the crux of the matter. Opioid-dependent women frequently have children, and by and large we are not publishing how we are treating them during their pregnancy and the associated outcomes. The inconsistency and lack of details of clinical guidelines on this issue make it difficult for practitioners and researchers to move forwards, and perhaps also make some people reluctant to publish what they do. Further consensus on ‘acceptable’ approaches to the management of opioid dependence in pregnancy, including detoxification approaches and if and how to monitor the safety of the fetus, would be useful and would go a long way to prepare for antagonist treatment trials. The World Health Organization is currently in the process of developing clinical guidelines on the management of substance use in pregnancy, and is accepting submissions on the topic 8. The second issue raised by Jones et al. is how to research antagonist treatment in opioid dependence, moving beyond the placebo-controlled RCT normally requested by regulatory authorities. Oral naltrexone may have demonstrated minimal advantages over placebo in western countries, where initial trials have been conducted 2, but higher success rates have been demonstrated in lower-income settings 9, 10, and it could be argued in some settings that the oral opioid antagonists such as naltrexone may have a role to play in the management of women who refuse or who do not have access to OAMT. The long-acting formulations are more promising, and it could easily be argued that they are worth comparing to standard detoxification in women who refuse or do not have access to OAMT. More controversial would be the use of long-acting opioid antagonists in women who are prepared to take OAMT, but would prefer an abstinence approach if the effectiveness was the same. At this stage it would be difficult to support such an approach, given the lack of data comparing long-acting opioid antagonists to OAMT in any population and, as pointed out by Jones, some of the missing steps along the way: how to predictably detoxify safely and effectively and commence opioid antagonists with minimal withdrawal severity; and is randomization feasible between such different treatment approaches as agonists and antagonists? To our knowledge, there are four studies in which there was random allocation (or pseudo-random in one case) to either opioid agonists or antagonists in the treatment of opioid dependence 11-14. While one study had problems with people accepting the randomization (mainly refusing to take the implants when OAMT was desired), on the whole participants appeared to accept the random allocation to maintenance versus detoxification and antagonist approaches; so while Jones raises the possibility that the RCT may not be the best study design to compare opioid antagonists to OAMT in pregnancy, the existing attempts would not preclude it. An ideal design for antagonist versus OAMT trials in any population would be the conduct of two simultaneous trials, one for those who would prefer OAMT as the control condition and one for those who would prefer detoxification, thus reducing the risk of participants refusing the allocated treatment post-randomization. It would also seem prudent to ensure that participation in such research is not the only way to access OAMT. However, as Jones suggests, we are not yet at that point. Nevertheless, if long-acting opioid antagonists do turn out to have similar or better treatment outcomes to OAMT in the non-pregnant opioid-dependent population, we suspect that there will be plenty of interest in them from pregnant opioid-dependent women. NC and AV are employed by WHO, which is developing clinical guidelines for the identification and management of substance use in pregnancy. They have no financial competing interests.