The theory is the kappa-opioid system that should inhibition emotions and give a state of anhedonia and depersonalisation . The affinity for naltrexone for the kappa system is low so you need a high dose above 150.mg to have a partiel effect.
In a 6.weeks trail with naltrexone the mean reduction was 30%. In the trail with rTMS at the right VLPFC for 15.min once the reduction was 25% the day after.
https://www.ncbi.nlm.nih.gov/pubmed/15876908
There is another problem. There has been several brain imaging studies of people with depersonalisation with the model that areas in the prefrontal cortex makes the inhibition of emotions and the insula, the in the structure in the brain that gives the sense of self.
3.structures are very active in this inhibition of emotions in DP; The right ventrolateral prefrontal cortex, the right dorsolateral prefrontal cortex and the medial prefrontal cortex. The ventrolateral prefrontal cortex is regarded as the most potent of the 3 in DP
if one tries to locate kappa opioid receptor in these structures that can inhibit emotions and anxiety only the medial prefrontal cortex can do that. So, if the inhibition of emotion in the other structures is likely with the current knowledge not done by the opioid system and antagonist will not have any effect on their inhibition. There might be people with depersonalisation where the medial prefrontal cortex makes a large inhibition and have some larger effect but in general the right ventrolateral prefrontal is the most potent in DP. It might explain that some have a high response, some partial and some none response to a opiate antagonist. It only affect one of the structures.
You can in theory try to inhibit them with rTMS. I don't think all can be tried in one session. The right VLPFC and the right DLPFC could be tried in a trail.
The medial prefrontal cortex is a part of the "default mode network" that is the network the brain is in when is not exposed to stimuli and at rest. The brain is self-reflective. That is likely to be disrupted in depersonalisation and other strutures related to the network is also very hyperactive active in DP. So, to work medial prefrontal cortex with rTMS one should properly look at depersonalisation as a abnormal and hyperactive "default mode network" that makes the brain highly self-reflective, introspective, defensive to emotions with difficulties to make shifts to other networks that is extroverted towards the world. There is no publications indicating the the medial prefrontal cortex has been tried in depersonalisation like there is for the right ventrolateral prefrontal cortex and the right dorsolateral cortex so I think it is extremely difficult to get someone to work with as it is untried unless it is tried with by researcher with knowledge of DP. There is either no publications where depersonalisation is viewed in from the view of the "default mode network". But, i know that someone has expressed an interest in that perspective. So, it might come in the future. There is a rTMS trail with depersonalisation in France with inhibition of angular gyrus that is a part of the "default mode network". The Temporo-parietal junction( TPJ) is also a part of the "default mode network" and rTMS trails have some significant effect in 50% when neuronavigation for location is used. To include the medial prefrontal cortex with either TPJ or angular gyrus might normalise the depersonalisation more. But, in rTMS trails it is one locations only and not two. If normalisation of a network two location might be more potent.
https://en.wikipedia.org/wiki/Default_mode_network