Hydromorphone is a semi-synthetic opioid analgesic with potent mu-agonist activity. It was first marketed in the U.S. in the 1920s.

Hydromorphone is structurally very similar to morphine (Figure 1); differing by the presence of a 6-keto group and the hydrogenation of the double bond at the 7–8 position of the molecule (15).

It is marketed in various formulations, including injection solution, (extended-) tablet, oral solution and suppository.

The PK of hydromorphone is well described in adults. In this population, after oral administration hydromorphone is rapidly absorbed and is subject to a significant first-pass effect, leading to a mean systemic oral bioavailability of 32% with wide interindividual variation (17–62%) (16–19); the maximum serum concentration (Cmax) is reached in less than an hour for immediate-release forms. After intranasal administration of the injection solution, a bioavailability of 50–60% is described (20, 21). Rectal administration has also been evaluated in small studies (n < 10) and has been found to have a bioavailability of around 30% (10–65%) (17, 18). Hydromorphone is a lipophilic molecule with a limited protein binding capacity of 7–19%, and its apparent Vd is relatively small, estimated to be approximately 1.22-4 L/kg. It is extensively metabolized (>95%) in the liver by uridine 5'-diphospho-glucuronosyltransferase 2B7 (UGT2B7) to hydromorphone-3-glucuronide (H3G), which has no intrinsic pain-relieving effects but is thought to have neuroexcitatory adverse effects (22–25). Other metabolites are dihydromorphine (<1%), dihydroisomorphine (1%) and their glucuronides. The involvement of P-glycoprotein (P-gp) in hydromorphone transport is not clearly established to date and rapid membrane crossing, including the BBB, is observed due to the liposolubility of hydromorphone (26).

Hydromorphone is therefore not only structurally but also pharmacokinetically very similar to morphine. Both molecules have interindividual variation in oral bioavailability, undergo glucuronidation primarily by UGT2B7, are metabolized to a 3-glucuronide metabolite and are eliminated by renal route. Response to both hydromorphone and morphine treatments may be influenced by polymorphisms in the μ-receptor gene (OPRM1), as well as drug interactions involving the UGT2B7 (27, 28). In adults, their key PK parameters such as bioavailability, Vd and half-life, are comparable (Table 1). Their differences are mainly for hydromorphone (i) a less well-defined role of the P-gp efflux transporter in the BBB penetration and brain disposition (26, 31, 32) and (ii) the lack of active 6-glucuronide metabolite formation. For this last reason, hydromorphone use in patients with severe renal impairment is often viewed as a safer alternative to morphine. However, evidence of a larger safety margin in renal failure is limited and both molecules should be used with caution due to the accumulation of their 3-glucunonide metabolites.

All mentioned data and observations are from studies in adults. In young children, ontogenic changes and other age-related differences can significantly alter the PK of drugs, for both morphine and hydromorphone, making simple extrapolation of adult data inappropriate.

Regarding morphine, the effect of ontogeny is well described. It is thus known that the estimated oral bioavailability is higher in very young infants than in healthy adults (33). Data on the ontogeny of UGTs are scarce, but UGT2B7 isoenzyme activity is reduced at birth and seems to reach adult activity levels between 1 and 12 months of age (34). Consistent with the immaturity of hepatic glucuronidation by UGT2B7, the limited ability of neonates to glucuroconjugate morphine is well documented (35, 36). Renal function which is represented by glomerular filtration rate (GFR) changes quickly with the maturation of young children, reaching adults' capacity between 6–12 months (37). Morphine clearance is typically slower in infants and approaches adult values by 6 months of age (38); therefore, the half-life is longer in the earliest stages of life and decreases as metabolic pathways develop (39). The neonatal BBB shows a lower barrier capacity than in adults, due to lower expression of barrier-related proteins and lower function of the P-gp, which reaches adult activity between 3 and 6 months of age (32, 40). This increase in permeability contributes, amongst others factors, to the increased sensitivity of neonates and young infants to the central depressant effects of morphine (32, 41).

As hydromorphone is pharmacokinetically very similar to morphine, the same changes, as described above, could be expected. However, data on the PK of hydromorphone in children are much sparser. We found only two studies that evaluated hydromorphone PK in the pediatric population (42, 43). The first study by Collins et al. included 10 children randomly assigned to receive either morphine or hydromorphone by patient-controlled analgesia (PCA) (mean ages 13.7 and 15.3 years respectively) to manage mucositis pain. Blood samples were drawn 2, 4, and 6 h after the start of a continuous infusion and only clearance was determined (51.7 mL/min/kg; range, 28.6–98.2). In the second, more recent prospective study by Balyan, 34 children [mean age 13.5 (4–18 years), bodyweight 56.7 (23–89.6 kg)] undergoing elective surgery (spine, neurological, or abdominal surgery) were treated with IV hydromorphone boluses followed by PCA. The PK profile was determined by measuring hydromorphone concentrations before and 3, 10, 30, and 90 min after the first dose and by using nonlinear mixed-effects modeling. The study demonstrated that body weight was a significant covariate for clearance while gender, race and type of surgery were not. Vd was comparable to the one described in prior adult studies (33 L/70kg vs. 3.35–42.7 L/70kg) and clearance value was smaller (0.738 L/min/70kg vs. 1.02-1.81 L/min/70kg) (17, 44, 45). Therefore, these two studies give us no information regarding other relevant PK properties, such as bioavailability or time to reach maximum concentration (Tmax), and above all, they provide no PK data for young children, particularly for infants younger than 6–12 months in whom the effect of ontogeny is the most expected.

Hydromorphone is a non-selective opioid receptor agonist with predominant affinity for μ-receptors and lower affinity for k- and d-receptors.

The efficacy and safety of hydromorphone are documented in adults, regardless of route of administration (46–48). As with all opioids, there is a large interindividual variability in the dose-efficacy-toxicity relationship. The “appropriate” dose for a given patient varies depending on many factors, including individual factors (gender, weight, comorbidities, organ function, previous exposure to opioids, ontogeny...) as well as genetic and environmental factors (comedications, diet...). The recommended initial dose often needs to be adjusted according to individual pain intensity, efficacy and occurrence of adverse drug reactions (ADRs). The most commonly described ADRs of hydromorphone are related to its binding to the μ-opioid receptors and are therefore, at equianalgesic doses, similar to the ADRs of other opioids. They consist mainly of dizziness, nausea, confusion, drowsiness, vomiting, constipation, pruritus and dry mouth; more rarely, respiratory depression and impaired consciousness. In adults, no study has demonstrated a different ADR profile, including nausea and pruritus, between hydromorphone and morphine at equianalgesic doses (48–50). The higher affinity for μ-receptors makes hydromorphone a more potent analgesic than morphine. The equianalgesic dose ratio between parenteral hydromorphone and morphine, calculated from adult studies, is approximately 1:5–7 (48, 51, 52). The same is true for the oral equianalgesic dose (52, 53).

In children, the efficacy of hydromorphone to treat perioperative pain has been demonstrated in a small number of studies when administered IV, either in bolus, continuous or PCA (14, 43, 54–60). The efficacy of epidural administration has also been established (61–67) and a recent study showed the efficacy of intranasal administration (68). Hydromorphone appears to be as effective as morphine, fentanyl and sufentanyl. These studies, whatever the route of administration, primarily included children and adolescents. Only two of them included infants (54, 59). These studies showed good tolerance of hydromorphone in infants, children and adolescents. Adverse effects were comparable to those described in adults, mainly nausea, vomiting and pruritus (14, 43, 57, 58, 60–68).

Spénard et al. recently published an excellent systematical review that sought to compare the efficacy and safety of hydromorphone and morphine in children (69). Among 754 abstracts reviewed, they found only four randomized controlled trials that compared the PD of hydromorphone and morphine in children (43, 56, 57, 61). In three of them, treatment was administered IV (43, 56, 57), in bolus or PCA doses, with equianalgesic dose ratio ranging from 5:1–7:1. The last of the four studies involved epidural administration and none involved oral administration. More than 150 children and teenagers were included, but none were younger than 3 years of age. Two of the studies involving IV administration showed no statistically significant difference in pain scores with morphine compared with hydromorphone. Only the study by Chen et al. showed that significantly more patients in the morphine group required extra fentanyl for pain relief, however with no significant difference in analgesia satisfaction score between the two groups (56). The three studies reporting the use of the IV route showed no significant difference in adverse effects, including nausea, sedation and pruritus (43, 56, 57). Only the study in which hydromorphone and morphine were administered epidurally found a higher incidence of pruritus related to the use of morphine (8% for hydromorphone vs 35% for morphine) (61). These findings should be taken with caution, as the relatively low (8%) incidence of pruritus on hydromorphone described in this study does not corroborate with the 30% to almost 70% incidence of pruritus reported in other studies (64, 65, 67).

Regarding the hydromorphone to morphine equianalgesic dose ratio, only one pediatric study has assessed the equipotence of hydromorphone vs. morphine (43). In this double-blind three-period crossover study, 10 children (mean ages 13.7 and 15.3 years for group 1 and 2, respectively) with mucositis pain received morphine or hydromorphone by PCA in a 7:1 ratio. Analysis of variance of total opioid doses indicated that patients used 27% more hydromorphone than expected, suggesting a mean equipotence of 5:1, comparable to that derived from adult's studies. No study has determined the equianalgesic dose ratio between oral hydromorphone and morphine in children and the same ratio is used in children of all ages, including infants, without taking into account the ontogenic considerations described above.

Marketing authorization for hydromorphone administration in children is restricted and varies from country to country (Table 2). Due to the few studies available on its epidural or intranasal administration, the only routes of administration approved by the majority of national regulatory authorities are oral, SC and IV injection (bolus, continuous or via PCA). In the United States (US), there is no labeled indication in children, regardless of the route of administration.

Various international expert opinions and formularies (70–79) have issued dosing recommendations for IV and oral hydromorphone in children. These recommendations vary widely and their scientific evidence is not described (Table 3).

As summarized in Table 3, the majority of IV recommendations tend to agree on a dosage of 0.01–0.02 mg/kg/dose every 3–4 h or 0.003–0.006 mg/kg/h for continuous infusion regardless of patient's age, but most often specifying an age older than 6 months or a weight higher than 10 kg. The Dutch Kinderformularium, a database developed by the Dutch Knowledge Centre for Pediatric Pharmacotherapy (Nederlands Kenniscentrum Farmacotherapie bij Kinderen: NKFK), available online at www.kinderformularium.nl, provides specific dosing for young infants, distinguishing between infants under or over 10 kg. They recommend a much lower dosage in infants under 10 kg: 0.003–0.005 mg/kg/dose every 3–4 hours (Kinderformularium.nl). Berde et al., in a small comment note under their guidelines table, specify that “in infants under 6 months, initial per-kilogram doses should begin at roughly 25 percent of the per-kilogram doses recommended” in older children (76).

In line with the oral bioavailability of hydromorphone described in adults, the most commonly recommended oral pediatric dose is 0.03–0.06 mg/kg/dose every 3–4 h. For young infants under 10 kg, the Dutch Kinderformularium recommends a dosage of 0.01–0.02 mg/kg/dose every 3–4 h. As with IV administration, Berde et al. recommend that “in infants under 6 months, initial per-kilogram doses should begin at roughly 25 percent of the per-kilogram doses recommended” in older children (76).