Preclinical biochemical and behavioural studies^{1, 2} have suggested the endogenous opioid peptides, endorphins, and exogenous agonists may have antipsychotic, antimanic, antipanic, and antidepressant effects^3-5 At least four strategies have been used in studying the possible role of endorphins in humans First is the use of narcotic antagonists (e.g., naloxone) that displace morphine and endogenous opioid-like compounds from opiate receptors. Narcotic antagonists have been used in efforts to treat psychiatric illnesses such as schizophrenia and mania. Second, studies have been conducted and are underway to administer b -endorphin to humans and then to study the behavioural and biochemical changes that are produced. Third, assays have been developed for b -endorphin in urine, plasma, and cerebrospinal fluid to study basal levels in psychiatric patients and the effects of pain, stress, diurnal variation, and activity on b -endorphin activity. Finally, investigators have stimulated the release of b -endorphin by pain, stress, and electrodes implanted in those areas of the brain rich in opiate receptors.

A great deal of effort in the past 5 years relates to the synthesis, degradation, and mapping of opiate receptors, and the relationship of endorphins to other neurotransmitter systems. However, while endorphins have anatomical distribution suggesting possible involvement in psychiatric illnesses, they have not been systematically studied or administered to patients in a double-blind placebo-controlled design to test the hypothesis that opiate agonists have antipsychotic and antidepressant effects. Therefore the physiologic roles of the various endorphins as neurotransmitters have yet to be clearly defined. While we have much to learn about endorphins in psychiatry, the possible implications for clinical medicine seem great because of the extensive effects of narcotic analgesics on behaviour and physiology. Could changes in endorphins be associated with psychiatric illnesses? Could specific alterations in endorphin function produce any of the symptoms of psychiatric illness? Could opiate addicts have preexisting opioid system deficits of deficiencies secondary to opiate abuse? Is methadone or other exogenous opiates sufficiently endorphomimetic to be used in clinical research trials in psychiatry?

In the past, morphine was extensively used in treating manic and other psychoses.³ Comfort⁷ has stated that morphine "can claim to be regarded as the first specific antipsychotic to be subjected to experimentation." Although others have suggested antipsychotic effects of opiate analgesics,⁴ There have been no clinical trials of opiates in the era of modern pharmacology. This is due to the fear of addiction and presence of effective nonaddicting antipsychotics and lithium. Opiates have anxiolytic effects which relate to abuse by musicians and performers. We have presented phenomenological and behavioural data which has demonstrated marked similarities between human opiate withdrawal and naturally occurring panic states.^{9- 11} We have suggested that this may reflect a common central noradrenergic (NE) hyperactivity.^12,13 This locus coeruleus (LC) hyperactivity hypothesis is also supported by our primate studies which demonstrated striking behavioural similarities between opiate withdrawal, electrical stimulation of the LC, pharmacological activation of the central noradrenergic system by piperoxane or yohimbine, and reversal of these anxiety-panic states with clonidine, morphine, or endorphins.^14-17 These data suggested that naturally occurring panic states might result from a functional deficit in endorphin-mediated inhibition of ascending NE activity.^18-21 Opiates (e.g., heroin and methadone) would be expected to provide a self-regulated treatment for this state, which would only manifest itself upon discontinuation of chronic opiate administration.⁵ This would be a possible fifth strategy for investigating endorphin function in man. Finally, addicts might have a pre-existing or drug-induced endorphin abnormality whereas panic-anxiety patients might have a naturally occurring endorphin-LC dysfunction, which could be identified by diagnostic studies of methadone programs and post- detoxification follow-up studies.^{5, 9, 22-25}

If endorphins have antidepressant effects then it might be expected that methadone programs would have an over representation of patients with major depression^{5, 9} and methadone detoxification might be associated with the re-emergence of major depressive syndromes.⁹ We have also studied the hypothesis that endorphin abnormalities might be present in mania and depression using neuroendocrine challenge paradigms. This hypothesis has been evaluated by a number of research groups.^26-29

We have additional clinical evidence for the antipsychotic, antidepressant and antipanic effects of opiate agonists as a result of our experience wit the previously described procedure for rapid opiate detoxification with clonidine.³⁰ We have treated 200 individuals addicted to methadone, heroin, or synthetic opiates by discontinuing opiates and administering clonidine to suppress the signs and symptoms of opiate withdrawal.^{30, 34} Clonidine markedly reduces opiate withdrawal symptoms and enables addicts who have been maintained for years on 100mg/day of methadone or less to discontinue use of opiates and clonidine in fewer than 14 days without significant withdrawal symptoms.^32-35 We have followed these patients for at least 4 weeks, in some cases for 24 months. Within 120 days of abrupt methadone discontinuation, 8 of these opiate addicts had significant depressive episodes, 5 had episodes of panic anxiety, and 3 had acute mania that required treatment. We will describe 3 of these patients, one who developed a significant depressive episode, another who developed a panic-anxiety disorder, and another who developed an acute mania to illustrate the natural history of the phenomenon and the similarities of these episodes to naturally occurring major depressive and panic-anxiety episodes as defined by the Research Diagnostic Criteria.³⁶

Mr. A, a 25-year-old married man, had a 2-year history of unsuccessful attempts to detoxify from methadone and a 9-year history of episodic opiate use. He was admitted to the clonidine study and detoxified from 50mg/day of methadone. Mr. A was able to return to work within a week of his hospital discharge, but approximately 2 weeks later he complained of difficulty falling asleep, a 2.7-kg weight loss, decreased libido, energy, concentration, no appetite, and a depressed mood. He failed to keep a scheduled appointment for follow-up and evaluation of these complaints. Five weeks after Mr A’s discharge from the hospital, his wife found him with a large kitchen knife, tearful, and expressing the desire to die. When he was admitted to a psychiatric hospital, he stated that he was unable to work for the previous week and that he was "hearing voices in my head saying I'm no good and should die." He had a quinine-positive urine and said he had transient symptom relief after heroin use. Mr. A'’ family gave a history of two previous depressive episodes, one of which resulted in psychiatric hospitalization (diagnosis of psychotic depression) and treatment with ECT. He was not addicted to opiates at this time. He expressed the desire to return to the methadone program. Methadone was administered and he was discharged from the hospital after 30 days. Mr. A returned to work on a maintenance dose of 25mg/day of methadone.

Mr. B, a 25-year-old married professional man, was successfully detoxified with clonidine from addiction to hydromorphone after an intermediate stabilization of 35 mg/day of methadone. He said he had used cocaine and amphetamine in the past, but they made him "too nervous to function." H had had now previous psychiatric hospitalizations or treatment but gave a history of episodic nervousness and feelings of impending doom when he was not taking opiates. Mr. B did well after hospital discharge, he returned to work immediately, and his urine specimens were negative for opiates. He had no complaints of any physical or psychic discomfort for weeks, but he then noted the abrupt onset of one to three "attacks" of nervousness, sweating, irritability, and fear: he said he sometimes felt like he should "run away somewhere, or hide, or do something, but I didn’t know really what to do." These symptoms were not clearly relatable to environmental events or changes. He began to "lose confidence in myself" but continued working. there were no neurovegetative signs of symptoms of depression, although he said he was depressed. Medical consultation and evaluation failed to find evidence of a medical syndrome that would explain his symptoms. He was given low doses of imipramine (25 mg p.o. t.i.d.) and reported that his attacks disappeared within weeks. He also said he had obtained a prescription for clonidine and was taking 0.2mg of clonidine with his imipramine.

Mr. C was a 32-year-old physician who had a 14-month history of morphine abuse and addiction. He had a history of marijuana and methaqualone (Quaalude) abuse. He had no previous psychiatric treatment but gave a history of a short-lived depression during his freshman year of college. This depression was described as being "paralytic," but after "sleeping for 5 days" the episode remitted. The patient also gave a history consistent with 2 days of grandiosity and complete insomnia after taking amphetamines during his first year of medical school. Recently the patient was successfully detoxified as an inpatient after a comprehensive evaluation. Approximately three weeks after discharge the patient’s wife called to say her husband had begun to act in an "atypical" manner. He became unable to sleep, hypersexual, and lost his appetite. He stayed at his office working on a manuscript "exposing" his colleagues. He refused to be interrupted in his work by patient appointments or his nurses and office staff. He told his wife he was uncomfortable with his house and bought a new house. He was hospitalized and treated with lithium carbonate.

The association between opiate addiction and psychopatholoty has a long history, which is new partially supported by empirical data. For example, opiate addicts have been shown to have high rates of depression,^37-43 antisocial personality characteristics,⁴⁴ schizophrenia or schizotypical fearures,^45-47 manic symptomatology,^{8, 44, 48} and alcoholism.⁴⁹ The major problem to date in the studies of psychopathology in opiate addicts, is that the assessment of psychopathology has usually been with symptom or personality scales. Diagnostic techniques, particularly the more recently improved measures and neurobiological testing,^50-53 have rarely been applied to the opiate addict. The result is that there has been a gap between general psychiatric practice and the treatment of opiate abusers. This gap is reflected in the fact that opiate addicts like other drug abusers are usually treated in separate unidimensional specialty clinics, like patients with tuberculosis had been treated in the past. The isolation of the addict in separate "programs" and from doctors and hospitals transmits a profound message to these people and isolates them from recent developments in psychiatric diagnostic practice, which leads to misdiagnosis and mistreatment.⁵⁴ For example, the opiate addict who has agorophobia might benefit from monoamine oxidase (MAO) inhibitors, as a bipolar patient might benefit from lithium; or the addict who has a major depression by ratings and biological testing would benefit from treatment with tricyclic antidepressants. We can only conclude after frequently observing the post-detoxification emergence of major psychopathology that many opiate addicts have been self-medicating and could be treated with alternative treatments.

As part of this study, we have also seen the patients described above and 7 other patients who had signs and symptoms of primary major depressive disorder (one with psychotic features), 2 with acute mania, and another 4 who had complaints consistent with the diagnosis of panic state. These patients had symptoms and courses similar to those of the patients described above. Six depressed patients were successfully treated with tricyclics and the other 2 were readmitted to the methadone program. Three panic patients were successfully treated with TCAs, one with an MAO inhibitor, and one was readmitted to the methadone program. The 2 manic patients were treated acutely with haloperidol and ultimately maintained on lithium alone. One of these manic patients was also readmitted to a methadone program. The clinical observations presented here support the notion that for some patients methadone maintenance does not merely help control opiate abuse, but also serves as a psychotropic maintenance program⁵ for depressive, manic, psychotic, and panic states.⁹

The pattern of signs and symptoms exhibited by patients withdrawing from opiates and by patients experiencing spontaneous attacks of panic-anxiety is similar, and this pattern is also observed in humans after the administration of drugs such as piperoxane or yohimbine, which markedly activate the noradrenergic nucleus locus coeruleus, and after abrupt discontinuation of chronic clonidine administration. These similarities suggest a possible common endorphin-noradrenergic mechanism mediating the syndromes of opiate withdrawal and panic anxiety. This had suggested to us that clonidine could be used in methadone withdrawal, and methadone in clonidine withdrawal. However, there are no reports of direct and quantitative comparison of these two syndromes.

Ten opiate addicts (8 males, 2 females) and 10 patients with histories of frequent attacks of panic anxiety (4 males, 6 females) were studied. The test battery included vital signs, the Addiction Research Center Inventory for Weak Opiate Withdrawal, the Speilberger State Anxiety Inventory, and analog self-rating scales for anxiety, fear irritability, unpleasantness, anger, and euphoria. This battery was administered to opiate addicts twice daily during a baseline period of methadone maintenance, then every 6 hours for 36 hours after the abrupt discontinuation of methadone. For the panic-anxiety patients, the battery was administered twice daily for 7 medication-free days and immediately upon the onset of a panic attack. All patients had at least 2 spontaneous panic attacks during the study period.

The 2 groups did not differ during the baseline period. During withdrawal or panic, both groups showed significant increases in heart rate, blood pressure, temperature, tremulousness, anorexia, insomnia, restlessness, and gastrointestinal discomfort (p<0.01). The opiate withdrawal patients demonstrated a significantly greater elevation of anger and irritability than did the panic-anxiety patients, who in turn, demonstrated a significantly greater elevation of fear ratings (p<0.01). All opiate withdrawal patients responded to treatment with clonidine with a rapid decrease in symptoms. Three of the panic anxiety patients with previous unsuccessful responses to imipramine were given clonidine during a panic-anxiety attack. Two of these three patients showed nearly complete alleviation of panic symptoms, while the third patient showed a significant, but partial, response to clonidine.

These findings again suggest a common neurobiological mediation of opiate withdrawal and panic anxiety. Theoretical implications of these results have been discussed,¹⁰ but include the use of opiates by physicians to treat intractable panic and anxiety states.

Clonidine is an efficacious nonopiate treatment for opiate withdrawal; however, it does not prevent readdiction or modify the numerous factors that are related to opiate use and addiction.^30,31 One of these factors may be an underlying psychopathology or a vulnerability to major psychoses or panic-anxiety states in patients maintained on opiates. If opiate agonists are effective or somewhat effective in the treatment or self-medication of major psychoses or panic states, then discontinuation of maintenance treatment might be expected to result in an exacerbation of the previously controlled syndrome. This may explain the clinical data reported here on the post-detoxification emergence of panic states, major depression, mania and psychosis. Although psychological and social factors or drug-induced changes might be invoked to explain the emergence of severe psychiatric symptomatology after clonidine detoxification, neurobiologic explanations invoking a pre-existing illness may be more parsimonious.

The discovery of opioid receptors and endogenous opioid peptides (endorphins) in the brain has kindled interest in the possible role of endorphin systems in psychiatric disorders.^{34, 55, 56} The analgesic, anxiolytic, euphoric, and calming effects of exogenous opiates suggest that decreased functional activity in endogenous opioid peptide systems could be linked to the pathophysiology of depression. Opiate receptors and endorphins are widely distributed in limbic and hypothalamic areas thought to be involved in depression. Endogenous opioid peptides may modulate noradrenergic and dopaminergic neuronal systems that are implicated in the etiology and pharmacotherapy of depression.

There are anecdotal reports from the prepsychotropic era of the efficacy of opiates in treating depression. We have observed and reported here significant depression in opiate addicts after detoxification. Some of these patients have been successfully treated with opiates and tricyclic antidepressants, suggesting that opiate discontinuation resembles the discontinuation of effective psychopharmacololgic maintenance treatment. Administration of b -endorphins has ameliorated depressive symptoms in some patients.⁵⁶ These data support the hypothesis that endogenous opioids influence the maintenance of pathological mood states.

We used a neuroendocrine challenge paradigm to study in vivo endogenous opioid systems in depressed patients. The normally pronounced increase in serum prolactin produced by morphine⁵⁷ was markedly blunted in patients with major depressive disorder.

Both exogenous opioids and endogenous opioid peptides are potent stimulators of secretion of the pituitary hormone prolactin in animals⁵⁸ and man.^{3, 27, 59} Morphine in the dosage range we used has been reported to produce large and reliable increases in serum prolactin in normal subjects.⁵⁷ Prolactin secretion is controlled in part by the dopaminergic tuberoinfundibular tract, which exerts an inhibitory effect over the secretion of prolactin.⁶⁰ Serotonergic neurons have a stimulatory effect on prolactin secretion; and other neurotransmitters and neuromodulators, including norepinephrine and epinephrine, have been reported to modulate prolactin secretion as well.^58-60 When opioid receptors located on dopaminergic neurons⁵⁵ are activated, they inhibit the dopaminergic tonic inhibition of prolactin secretion. Such opioid receptor activation would therefore allow increased secretion of prolactin after administration of opioids.⁵⁸ The mean maximal prolactin response of 7.2 ± 2.7 ng/ml in the depressed group was significantly lower than that of 31.9 ± 9.5 ng/ml in the control group (p<0.01).

Thus the absent or blunted increase in serum prolactin occurring after morphine infusion in our patients with major depressive disorder may reflect abnormalities in central endorphin, dopamine, serotonin, or other neuroregulatory systems. Possible abnormalities in endorphin systems that could account for a blunted prolactin response to morphine in major depression include an opioid receptor deficit, an excess of endogenous opioid antagonist, or elevated endorphin levels with compensatory down regulation of opioid receptors. This might suggest a role for endorphin testing in depression and a possible use of endorphins in the treatment of intractable depression.

The mood effects produced by the acute and chronic use of narcotic analgesics offer some support for these neurendocrine data and suggest that endorphins may play a role in affective disorders. In one outpatient study, cyclazocine, a mixed agonist-antagonist, was found to be an antidepressant.⁶¹ When patients who are depressed were given acute infusions of naloxone, no effect was seen on depressed mood.⁶² These clinical data reported here are also in agreement with our previously described hypothesis for bipolar affective illness.⁸

We have previously described the self-medication of schizophrenic patients with opiates in a methadone maintenance program⁵ and the theoretic and clinical data suggesting the antipsychotic efficacy of opiate antagonists.^{3, 5} These schizophrenic patients were well known to us and were not detoxified from methadone maintenance. We did not observe the emergence of a schizophrenic of schizophrenic-like psychosis after rapid detoxification in the sample of patients reported here. What we did observe was the appearance of significant depression and mania, with and without psychotic features, and panic episodes that were virtually identical to naturally occurring depressive and panic episodes.³⁶ In addition, these patients were successfully treated with opiates, tricyclic antidepressants, or a combination of tricyclic antidepressants and clonidine, which suggests that opiate discontinuation in these patients was similar to the discontinuation of an effective psychopharmacologic maintenance treatment.

These clinical data may offer some support for the hypothesis that opiates have antidepressant, antimanic, and antipanic effects. This hypothesis should be studied directly by double-blind studies of the effects of exogenous and endogenous opioid peptides in patients with major depressive illness, panic and anxiety states, schizophrenia, and schizo-affective illness. These clinical data support our studies in nonhuman primates and man with suggest a common LC or NE hyperactivity may underlie both drug withdrawal and spontaneous panic states.^{63, 64} Whether endorphin deficiency or derangements account for the postulated NE hyperactivity needs additional study and we will discuss our preliminary work later. Failure of endorphins to terminate bursts in LC firing rate and NE release may be responsible for both of these types of panic states.^{6, 65-67} In addicts, this mechanism could exist prior to opiate use, or abuse of potent exogenous endorphinomentic compound may cause an endorphin-abnormality.⁶⁸ Both of these possibilities would be compensated by continuous opiate maintenance.

Methadone maintenance is a complicated psychiatric, psychological, and social phenomenon. Further studies are necessary to evaluate the role of opiate maintenance in treating or suppressing the emergence of underlying psychopathology. Previous psychiatric hospitalization or treatment for a schizophrenic or affective illness may contraindicate absolutely the use of clonidine or other rapid detoxification methods. These data suggest the possibility of substituting a nonaddicting psychotropic medication for opiates in some patients who are self-medicators. The clinical data support other data suggesting the potential antipsychotic, antidepressant, and antianxiety/antipanic effects of the endogenous opioids, endorphins, and exogenous opioids, endorphins, and exogenous opiates.^{56, 69-74} These and other data suggest potential utility for opioid agonists and endorphin testing in psychiatric treatment and diagnosis.