To our knowledge, this is the first study to investigate the pharmacokinetics and pharmacodynamics of therapeutically used doses of IN DAM in patients with opioid use disorder. We found that self-administered IN DAM doses elicited moderate to strong peak drug effects, with marked reductions in heroin craving and withdrawal symptoms. No clinically relevant respiratory depression and decrease in oxygenation saturation was observed. Subjective effects, including a sense of well-being, rush, and heroin-typical sensations, occurred rapidly within the first 2 min after the start of administration. These effects gradually increased and peaked at 30–35 min, which paralleled the rising plasma concentrations of DAM and 6-MAM. Subsequently, subjective effects remained stable until 60 min, followed by a gradual decline until 180 min, which mirrored the slow decrease of morphine plasma concentrations. Interestingly, M6G levels continued to rise until 180 min, suggesting that it may play a lesser role in the drug’s overall effects. Although M6G is twice as potent as morphine at the mu-opioid receptor in vitro [20], its greater polarity may limit its ability to cross the blood–brain barrier efficiently [21]. Taken together, the initial subjective effects were likely driven by the rapid increases in DAM and 6-MAM, while the sustained subjective effects were likely maintained by morphine. In line with this, the observed counterclockwise hysteresis for DAM and 6-MAM, in contrast to the clockwise hysteresis for morphine and M6G, suggests that the active metabolites morphine and possibly M6G may prolong the drug effects initiated by DAM and 6-MAM [22]. Additionally, the rapid onset of subjective effects within the first 2 min, when 6-MAM plasma levels were still very low, indicates that DAM, rather than 6-MAM, was responsible for this rapid onset. This assumption is consistent with a recent review, which proposed that DAM should be regarded not only as a prodrug but rather as an active compound whose effects, alongside those of 6-MAM, morphine, and M6G, contribute in a sequential and partially overlapping manner to the complex temporal pattern of heroin effects [23].

The present study also described the pharmacokinetics of IN DAM at therapeutically relevant doses. The plasma levels of DAM, 6-MAM, morphine, and M6G peaked at 9, 17, 66, and 129 min, respectively, after the start of administration. DAM and 6-MAM concentrations declined rapidly thereafter, while morphine and M6G reached a plateau and declined much slower. We observed a mean plasma half-life of DAM, 6-MAM, and morphine of 17, 47, and 154 min, respectively. For IN DAM, the elimination half-life was longer (17 versus 5 min), and the tmax occurred later (9 versus 4–5 min) than previously reported [8, 9]. There are several major differences between the studies that may explain this discrepancy. Most importantly, our participants administered up to 30-fold higher doses than in previous studies, which are needed in clinical use for IN opioid-agonist treatment. Owing to the limited capacity of the nasal cavity, these higher doses entail that application can range from a one-off administration in both nostrils to repeated administration in both nostrils over the course of several min. A longer duration of administration is expected to result in a later tmax. Additionally, the nasal absorption of DAM may have been saturated owing to the high doses and volumes administered. Thus, the absorption of the entire DAM dose may have been delayed, which might explain the later tmax and the longer plasma half-life of DAM observed in this study compared with previous studies [8, 9, 12]. Moreover, the overall plasma concentration of DAM and 6-MAM were less clearly dose-dependent compared with those of morphine, M6G, and M3G.

The weaker dose-dependence suggests that part of the dose may have been swallowed and metabolized directly to morphine in the gastrointestinal tract prior to entering the blood stream. Such a first-pass effect of DAM results in negligible levels of DAM and 6-MAM in plasma after oral ingestion [24]. Although we did not detect a distinct second morphine peak as previously described in intranasal heroin users [8, 9, 25], direct metabolization of DAM and 6-MAM to morphine likely contributed to the slow decline in morphine plasma concentrations. Notably, the observed mean plasma half-life of morphine of 154 min represents an underestimation. This is because the half-life could not be determined in 6 out of 14 patients owing to the slow decline, which could not be captured within the 3-h sampling scheme. Moreover, the observed plasma half-lives of the metabolites 6-MAM and morphine do not represent their true elimination half-lives. Their prolonged plasma half-lives are likely attributable to continuous absorption and metabolization of the parent drug into 6-MAM and then subsequently into morphine, which offsets their elimination from plasma.

On average, peak DAM plasma concentrations were elevated in a dose-dependent manner. However, there was substantial variability in plasma concentrations across patients. This variability can be attributed to factors such as differences in administration duration, techniques of intranasal self-administration, variability in metabolizing enzyme activity, and the short elimination half-life of DAM, among others. Importantly, intrasubject plasma levels showed less variation, suggesting that the blood sampling procedures and analytical method used were adequately measuring plasma levels of DAM and its metabolites.

Compared with existing literature, the peak plasma concentrations and overall exposure of DAM and 6-MAM were substantially lower in our study. For example, an IN DAM dose of 12 mg resulted in peak DAM plasma levels of 10–40 ng/ml [8, 9]. Assuming a linear dose–concentration relationship, a peak plasma concentration of 260–1040 ng/ml would be expected for a mean IN dose of 346 mg DAM. However, we observed 3- to 13-fold lower peak DAM plasma concentrations, ranging from 20 to 247 ng/ml in this study. In contrast, the peak plasma levels of morphine, M6G, and M3G were comparable when adjusted for dose differences. This discrepancy further supports the hypotheses that at higher IN doses and volumes, the absorption of DAM may be saturated, and that a significant amount of DAM may be swallowed and metabolized directly into morphine. A more concentrated DAM solution would reduce the administered volume, potentially improving absorption.

Notably, participants rated the rush, the good, and heroin-typical effects almost similarly, which suggests that they perceived them as overlapping or even synonymous. While these constructs were presented as distinct in the questionnaire (“rush” referring to the immediate and intense euphoria typically associated with IV DAM administration; “good effects” referring more broadly to any positive, pleasurable sensation; and “heroin-typical effects” referring to the subjective experience that users associate with heroin use), the similar ratings suggest that participants may have been unable to distinguish these items. Additionally, participants may have interpreted “rush” not in a pharmacological sense (i.e., the rapid onset after IV administration) but rather as a broader, less route-specific positive effect. This interpretation is supported by the fact that all participants were already treated with IN DAM and would therefore be familiar with its pharmacodynamic profile. Moreover, several participants had never injected DAM and may therefore not have had a reference point for the characteristic IV induced “rush.” Therefore, the near-identical ratings do not necessarily imply a lack of awareness of differences in onset but may rather point to semantic overlap in how these items on subjective effects were understood and reported by participants. This interpretation was also confirmed by feedback from individuals with lived experience of IN DAM treatment, who indicated that these terms were indeed perceived as overlapping or difficult to distinguish in practice.

The study has notable strengths. First, it investigated therapeutically used doses of IN DAM, thus enabling the evaluation of potential clinical challenges related to administration technique and saturation of absorption at high doses and volumes. Second, the study provided a comprehensive pharmacokinetic profile of DAM and its pharmacologically active metabolites over a 3-h period. Furthermore, participants received their usual oral opioid maintenance dose only after the 3-h observation period, ensuring that the subjective effects of IN DAM were accurately assessed. From a pharmacological standpoint, this study demonstrates that IN DAM is a viable treatment option for patients who sniff illicit heroin. Moreover, IN DAM should be considered as a harm-reduction strategy for individuals injecting prescribed DAM. Compared with previously published harm reduction approaches, e.g., improving needle exchange programs by considering acidity-related vein damage, IN DAM is associated with a lower risk of overdose and no injection-related complications [26]. However, the reduced subjective opioid effects compared with IV DAM might lead to a lower acceptance. A comprehensive discussion of IN DAM as a harm-reduction strategy is beyond the scope of this paper and available in literature [3]. Notwithstanding these strengths, the present study has several limitations. The sample size was relatively small, with only 14 participants, and the findings would need to be replicated in a larger population. Additionally, many sources of variability were present, including a wide range of doses, differing baseline morphine plasma levels, unknown baseline levels of other opioids such as methadone or levomethadone, and variations in administration techniques, all of which could have influenced dose–response effects and pharmacokinetic parameter estimates. Moreover, participants may not have been able to make the distinction between the “rush” and other typical heroin effects, which could have affected the accuracy of their responses on the self-report questionnaire.