1
`
`Quality aspects
`
`SCIENTIFIC DISCUSSION
`
`Introduction
`1.1
`Sodium oxybate is a simple molecule, the sodium salt of gamma-hydroxybutyric acid (GHB) and is
`presented in the form of a stabilised oral solution, 500mg/ml
`
`1.2 Active Substance
`Sodium oxybate is the Common Name of the substance butanoic acid 4-hydroxy-monosodium salt. At
`the time of writing this report there is no INN for this substance.
`
`1.2.1 Manufacture
`The manufacturing process is very simple and is basically a one-step hydrolysis of gamma
`butyrolactone under alkaline conditions with sodium hydroxide.
`Starting materials and critical steps are well defined; there are no intermediates.
`The active substance is obtained as a white solid which is dried, ‘de-lumped’ and packed.
`The characterisation of the substance arising from the documented method of synthesis confirms that it
`is indeed sodium oxybate and this has been done by the usual range of spectroscopic methods
`including UV, IR & NMR together with elemental analysis and pKa determination.
`Since sodium oxybate is to be given in solution, polymorphism has not been investigated
`
`1.2.2 Specification
`The specification includes test for identification (IR, HPLC), assay (HPLC), related impurities
`(HPLC) together with tests for water content (KF) residual solvents (GC) and heavy metals, all
`performed by validated methods. It is not necessary to control solid state properties.
`The impurities include a number of named impurities and one un-named impurity, the levels of which
`have all been qualified on a toxicological basis and are considered to present no unnecessary risk.
`Batch analyses (n = 45) from the site of manufacture defined in the dossier, demonstrate satisfactory
`compliance with the agreed specification and indicate good uniformity.
`
`1.2.3 Stability
`In addition to forced degradation studies in the solid state and in solution, stability studies have been
`performed on 6 batches of sodium oxybate under ICH conditions, accelerated and long term. No
`significant negative trends or out of specification results were observed in the formal ICH stability
`investigation, and on the basis of the accumulated results a satisfactory re-test period has been defined.
`
`1.3 Medicinal Product
`
`1.3.1 Pharmaceutical Development
`The product is a simple aqueous solution which is stabilised with malic acid (hydroxysuccinic acid)
`and adjusted to the pH of maximum stability. From a microbiological point of view, investigations on
`this formulation showed that the product also had intrinsic antimicrobial activity and passed the test
`for efficacy of antimicrobial preservatives; therefore it was considered unnecessary to include a
`preservative in the formulation.
`The product is presented as a plastic PET bottle with a child-resistant screw cap closure for the liquid,
`and in order to facilitate accurate dosage a separate dispensing or dosing system is attached at time of
`first use. This consists of a plastic adaptor to fit into the bottle, leading to a syringe dispenser allowing
`the patient to withdraw the accurate dose. In addition, two plastic dosing cups are provided with child-
`resistant closures.
`
`PAR1021
`IPR of U.S. Patent No. 8,772,306
`Page 1 of 30
`
`

`
`1.3.2 Manufacture of the Product
`The active substance is dissolved in purified water and the pH adjusted with malic acid before dilution
`with purified water, filtration, and bottling. The validation of this simple scheme was not
`problematical.
`
`1.3.3 Product Specification
`The product release specification includes relevant tests and limits for physical examination, identity
`(HPLC & IR), assay (HPLC), impurities (degradation products, HPLC), volume in container,
`reproducibility of dosage, pH, microbiological attributes (PhEur), etc. Control of rheological
`properties is not considered necessary for a mobile liquid dosage form.
`Batch analyses (n = 18) indicate satisfactory compliance with the agreed specification and satisfactory
`product uniformity
`
`1.3.4 Stability of the Product
`In all, ten batches of product have been investigated for stability under ICH conditions, accelerated
`and long term, and in all cases the results support the shelflife and storage conditions as defined in the
`SPC.
`In addition, since this is a multidose product with a special adapter and dosing system, additional
`studies were performed with these plastic components in place, in order to mimic the in-use situation,
`and a suitable in-use shelflife has been defined.
`In general the studies show that the plastic adapter and dosing system is compatible with the solution
`and does not encourage degradation.
`Apart from the physical chemical analytical studies carried out during the stability investigations,
`microbiological studies were also performed with satisfactory results.
`
`1.4 Discussion on chemical, pharmaceutical and biological aspects
`The simple synthesis and manufacture of the product are described and controlled in a relevant
`manner, and the specifications of the active substance and medicinal product are considered to be
`relevant for a product of this type. The stability of the product has been well-investigated, both in the
`unopened form, and with the adaptor and dosing system in place during use.
`Satisfactory uniformity of dose has been demonstrated, and there are no unresolved quality issues that
`could have an impact on the benefit/risk balance for the patient.
`
`2
`
`Non-clinical aspects
`
`2.1
`
`Introduction
`
`2.2 Pharmacology
`
`2.2.1 Primary pharmacodynamics (in vitro/in vivo)
` Oxybate (GHB) is a metabolite of γ-aminobutyric acid (GABA) which is synthesised and
`accumulated by neurones in the brain. It is present at µM concentrations in all brain regions
`investigated as well as in several peripheral organs, particularly in the gastro-intestinal system.
`Neuronal depolarisation releases GHB into the extracellular space in a Ca2+-dependent manner. A
`family of GHB receptors in rat brain have been identified and cloned and most probably belong to the
`G-protein-coupled receptors. High-affinity receptors for GHB are present only in neurones, with a
`restricted specific distribution in the hippocampus, cortex and dopaminergic structures of rat brain.
`In general, stimulation of these receptors with low (physiological) amounts of GHB induces
`hyperpolarisation in dopaminergic structures with a reduction of dopamine release. However, in the
`hippocampus and frontal cortex, GHB seems to induce depolarisation with an accumulation of cGMP
`and an increase in inositol phosphate turnover. However, at higher (therapeutic) exposures, GHB
`
`Page 2/30
`
`PAR1021
`IPR of U.S. Patent No. 8,772,306
`Page 2 of 30
`
`

`
`receptors are saturated and probably de-sensitised and down-regulated. Such GHBergic potentiations
`induce dopaminergic hyperactivity, strong sedation with anaesthesia and EEG changes that are
`consistent with normal sleep and/or epileptic spikes.
`The pathogenesis of narcolepsy is still unknown, but an imbalance between monoamines and
`acetylcholine is generally accepted. Recent research has found a marked reduction of the neuropeptide
`hypocretin type 1 in the cerebrospinal fluid of a majority of patients and a global loss of hypocretins in
`post-mortem brain tissue of narcoleptic subjects. The hypocretins are synthesised by a small group of
`neurones predominantly located in the lateral hypothalamic and perifornical regions of the
`hypothalamus. The hypothalamic system directly and strongly innervates and potently excites
`noradrenergic, dopaminergic, serotoninergic, histaminergic and cholinergic neurones. The effect of
`GHB on this system has not been investigated. However, the available data indicate that its mode of
`action is likely to relate to non-specific dopaminergic stimulation rather than the hypocretin system.
`Formal nonclinical pharmacology studies to investigate the primary pharmacodynamics have not been
`conducted by the applicant, rather a comprehensive review of the scientific literature has been
`conducted. The publications included have been selected based on their relevance to the proposed
`indications, based on evidence of efficacy from early clinical studies. In addition, animal models of
`cataplexy and narcolepsy are continuing to be developed, but have not yet been fully validated. Little
`nonclinical information is available on its effects on narcolepsy in general, and cataplexy in particular.
`Available, directly relevant data, from the published literature, has been reviewed but the current
`understanding of the role of GHB in the CNS does not provide a mechanistic explanation of the
`positive clinical effects reported in the dossier. GHB had no effect on cataplexy in dogs with
`hereditary narcolepsy when administered as a single dose of 500 mg/kg i.v. or 50 mg/kg/day p.o. for 3
`consecutive days. However, although such dogs have a mutation of the type 2 hypocretin receptor, the
`clinical relevance of this model remains to be established. Moreover, a dose of 75 mg/kg/day for at
`least 14 days is required for efficacy in humans.
`Though the precise mode of action is unknown, the sedative properties of GHB and its effects on sleep
`may play a role in the efficacy observed in humans.
`Evidence from a human clinical study (Study OMC-SXB-20) where GHB was administered to
`narcoleptic patients and overnight polysomnograms (PSG) were recorded, suggests that GHB modifies
`sleep architecture, specifically a dose-related increase in Stage 3 & 4 slow wave sleep (SWS, delta
`sleep). The cause of human narcolepsy and cataplexy is, as yet, unknown. Recent evidence points to
`the loss of hypocretin-containing neurones, possibly due to autoimmune attack, as a likely cause
`(Scammell 2003). Hypocretin is a neurotransmitter that has roles amongst others, in sleep-wake
`regulation. Alterations in hypocretin neurotransmission have also been observed in mouse and dog
`models of narcolepsy, although no studies have been undertaken with GHB in these models. Animal
`models of cataplexy and narcolepsy are continuing to be developed (Gerashchenko et al, 2003), but
`the effects of GHB in these models, have yet to be investigated.
`
`2.2.2 Secondary pharmacodynamics
`Published literature reports are presented that discuss the potential for effects on the respiratory,
`cardiovascular, gastrointestinal, renal and endocrine function, together with relevant findings from the
`toxicology studies.
`GHB may increase growth hormone secretion, but this effect is inconsistent across species and dose
`levels. GHB has no other relevant secondary pharmacodynamic effects in animals.
`GHB consistently decreases respiration by effects on minute volume and respiratory rate, with
`younger animals being more susceptible to these effects. In halothane-anaesthetized rats, GHB (187.5-
`750 mg/kg i.p.) dose-dependently decreased basal minute volume and respiratory rate compared to
`pre-injection control, with a maximum decrease to about 60% of pre-injection values for each
`parameter at the highest dose of GHB (Hedner et al, 1980).
`Effects on cardiovascular parameters were also studied as part of the repeat dose toxicology studies in
`dogs, including heart rhythm and P-QRS-T complexes determined from ECGs, and there was no
`evidence of any dramatic changes in these parameters during the studies at doses up to 600 mg/kg/day
`(corresponding to male and female AUC0-24 of 3363.05 and 3631.35 µg·hr/mL and Cmax of 583.0
`
`Page 3/30
`
`PAR1021
`IPR of U.S. Patent No. 8,772,306
`Page 3 of 30
`
`

`
`and 726.7 µg/mL). Additionally, in several human clinical studies, there were no significant effects of
`GHB administration on ECGs. The applicant claims that given the relatively long established clinical
`use of GHB, as an anaesthetic and sedative and other uses such as for the treatment of alcohol
`withdrawal, the undesirable effects and risk potential of GHB are well known, and additional
`nonclinical safety pharmacology studies are not justified. Results of some studies indicate weak
`rewarding effects and possible development of tolerance in rats and mice, however there is no
`compelling evidence that GHB represents a significant drug dependence hazard. Interaction of GHB
`with ethanol and other central nervous system depressants generally result in greater central depressant
`effects than seen with either drug alone. Numerous case reports of GHB poisoning demonstrate that
`overdosing in humans is associated with many of the same signs and symptoms as in animals: a rapid
`onset of drowsiness, nausea, vomiting, myoclonic seizures, respiratory depression progressing to
`apnoea, and coma.
`
`2.2.3 Safety pharmacology
`The applicant has not conducted animal safety pharmacology studies. However, there is ample
`evidence in the published literature that GHB is a potent CNS depressant, may cause convulsions and
`potentially fatal respiratory depression and cardiac failure.
`On the basis of available information, the lack of conventional safety pharmacology studies is
`considered acceptable. However, since co-medication is probable, the effects of GHB on respiratory
`pattern in the presence of other CNS depressing agents like ethanol, and inhibitors of GHB
`metabolism
`like
`valproic
`acid
`and
`ethosuximide
`are mentioned
`in
`the
`SPC.
`A weak tolerance to GHB administration has been demonstrated in a number of specific animal
`behavioural studies and also the development of cross-tolerance between GHB and ethanol. Therefore,
`potentially, an acute toxic effect (e.g. acute respiratory depression) could be experienced after drug
`intake following a period of drug withdrawal, sufficient for the disappearance of tolerance. Caution is
`advised if treatment is re-started after discontinuation. (SPC, section 4.2). Clinical data (open label
`study OMC-GHB-3) have failed to show any major development of tolerance on efficacy and the
`AUC after 8-weeks compared to the first dose was not significantly increased (study OMC-SXB-10,
`see clinical section). However, as these clinical data are too limited to draw firm conclusions, the
`potential for development of tolerance, especially with concomitant intake of ethanol, cannot be
`excluded and is mentioned in the SPC.
`
`2.2.4 Pharmacodynamic drug interactions
`Formal studies of pharmacodynamic drug interactions have not been conducted. According to the
`published literature, concomitant administration of GHB and other CNS depressants (benzodiazepines,
`barbiturates, alcohol) results in an additive increase in sedation.
`
`2.3 Pharmacokinetics
`The applicant has not conducted animal PK studies, with the justification that the more relevant
`pharmacokinetic data are derived from human exposure. Some data have been compiled from a review
`of the published literature. Data on non-clinical absorption, distribution, metabolism, and excretion
`have been compiled from a review of the published literature
`
`2.3.1 Absorption- Bioavailability
`In the rat, oral bioavailability was about 50-80%. Kinetics was non-linear, with oral dose increments
`resulting in an under-proportional increase in Cmax and an over-proportional increase in the AUC. By
`contrast, i.v. administration resulted in an over-proportional increase in Cmax. Thus, both absorption
`from the gut and elimination may depend on saturable mechanisms. Saturable absorption from the gut
`was confirmed in an everted rat intestine model.
`
`2.3.2 Distribution
`Whole-body autoradiography following i.v. injection of 14C-GHB in mice showed a fairly uniform
`distribution pattern of radioactivity due to GHB and/or its metabolites. Shortly after injection, lower
`radioactivity was found in fatty tissues such as thymus, brown fat, and the white and grey brain matter
`
`Page 4/30
`
`PAR1021
`IPR of U.S. Patent No. 8,772,306
`Page 4 of 30
`
`

`
`than in other tissues, including plasma. However, by 30 minutes after injection, radioactivity was
`distributed throughout the body, including brain, skeletal muscle, myocardium, kidney, spleen, liver,
`lung, thymus, urinary bladder, stomach, intestines, and, in pregnant mice, the foetus. GHB distributed
`rapidly to the brain of rats, dogs and monkeys, producing brain concentrations several orders of
`magnitude above the physiological level. In the dog, the highest concentration was found in the white
`matter of the temporal lobe. There are no data on plasma protein binding in animals, but this is likely
`to be negligible.
`
`2.3.3 Metabolism (in vitro/in vivo)
`Metabolism of GHB is rapid and complete and proceeds via succinic semialdehyde, succinate and the
`Krebs cycle or through γ-hydroxybutyrate and β-oxidation.
`The potential for inhibition of CYP isozymes was tested in pooled human liver microsome fractions
`using standard markers for CYP1A2, 2C9, 2C19, 2D6, 2E1 and 3A. In all cases, the IC50 was > 3000
`µM (> 378 µg/ml). Since the average maximum human exposure is 142 µg/ml, GHB is not expected to
`show pharmacokinetic interactions with drugs metabolised by these isozymes. Non P450 mediated
`effects on GHB metabolism are unclear.
`A rat study found that co-administration of compounds stimulating or inhibiting GHB dehydrogenase
`were able to decrease or increase plasma levels of GHB by up to 1/3. The interactions resulting from
`the stimulation or inhibition of GHB dehydrogenase, namely with anticonvulsivant drugs and L-dopa
`are considered to be clinically relevant and are mentioned in the SPC.
`
`2.3.4 Excretion
`Clearance is predominantly by biotransformation, with limited amounts of unchanged drug recovered
`from the urine or faeces. Radiospirometric studies in rats showed that 14C-GHB was rapidly
`converted to exhaled CO2 and about 2/3 of the dose was excreted by respiration within 6 hours and an
`additional 10- 20% over the next 18 hours. After oral administration of 14C-labelled GHB
`(200 mg/kg) to rats, the urinary recovery over 48 hours was 5.5% of the radioactive dose, and only
`1.5% was recovered in the faeces. There are no data on the excretion of GHB in the milk of lactating
`animals. The proposed SPC contains an appropriate statement to this effect.
`T½ in rats following oral administration of a single dose of 200 mg/kg was 0.75 h for the α- and 2.68
`h for the β-phase. Similar T½ values were observed in dogs and monkeys. In rats, Cmax and AUC
`values tended to be higher in females than in males, whereas the opposite applied to dogs.
`The applicant has been asked to discuss the comparative pharmacokinetics in humans and
`experimental animals and the implications for a critical appraisal of the relevance of the main species
`used in the toxicity testing for human safety assessment. In summary, the rat and dog showed similar
`pharmacokinetic characteristics, although exposure measured as AUC was higher in human than in
`either rat or dog at the NOAEL. The exposures measured in the maximum tolerated dose toxicokinetic
`studies (conducted in support of mouse and rat carcinogenicity studies) were, however, greater than in
`human subjects (Cmax 2.60- and 2.76-fold; AUC 1.21- and 1.64-fold for mouse and rat, respectively).
`
`2.4 Toxicology
`All toxicology studies were conducted by the applicant with the exception of data from literature for
`single dose toxicity and carcinogenicity in mice.
`2.4.1 Single dose toxicity
`Formal single dose toxicity studies were not conducted. A review of published literature data
`identified a number of references providing LD50 values in several species. In the mouse, LD50
`values of 2960 – 3700 mg/kg following i.p. injection were identified; in the rat, LD50 values were
`9990 mg/kg following p.o. administration, and 1700 mg/kg following i.p. injection. In the rabbit and
`dog, LD50 values in excess of 1000 mg/kg were reported following i.v. administration, which could
`be increased to over 7000 mg/kg with artificial respiration without lethality.
`
`Page 5/30
`
`PAR1021
`IPR of U.S. Patent No. 8,772,306
`Page 5 of 30
`
`

`
`2.4.2 Repeat dose toxicity (with toxicokinetics)
`Repeat-dose toxicity studies comprised 3- and 6-month toxicity studies in rats and 3- and 12-month
`studies in dogs. Treatment-related clinical signs were mainly related to sedation, reduced food
`consumption and secondary changes in body weight, body weight gain and organ weights.
`In rats, the only treatment-related clinical chemistry changes were a slight reduction of serum albumin
`and WBC in rats that may have been related to changes in nutritional status. The lowest NOAEL value
`was 350 mg/kg/day in rats (AUC ≈200 µg.h/ml) based on bodyweight changes.
`In dogs, three repeat dose studies have been conducted, an initial rising dose study, followed by 90
`day and 52 week exposure studies. In the rising dose study single doses ranging from 150 – 1800
`mg/kg/day were investigated, followed by a 5 day continuous dosing phase at 600 mg/kg/day.
`Treatment related clinical signs included emesis following dosing at 600 mg/kg and 1200 mg/kg and
`emesis, hypersalivation, ataxia and hypoactivity following the 1800 mg/kg dose. Emesis and ataxia
`were also observed during the 600 mg/kg/day daily dosing phase. Plasma concentrations were
`observed to increase in approximate proportion to increasing dose, with emesis leading to decreased
`plasma concentrations. These signs subsided after the first few weeks of the study, and led to the
`decision to increase the high dose to 900 mg/kg/day at Week 32. Following this dose increase, similar
`clinical signs were noted in the 900 mg/kg/day dose group.
`There were no treatment-related clinical chemistry, ophthalmology or ECG findings. Necropsy
`findings included a dark area on the ileal mucosa of one animal, as a result of emesis during the study.
`A dose-dependent atrophy of the salivary and submucosal oesophageal glands was observed. Such
`changes are not uncommon with drugs acting on the parasympathetic nervous system. The NOAEL
`value was 150 mg/kg/day (AUC ≈300 µg.h/ml) based on bodyweight changes and salivary gland
`atrophy.
`The toxicokinetic parameters (mean values calculated from male and female data, since no apparent
`gender difference has been reported) at the NOAEL and/or NOELs in the repeat dose toxicology
`studies in rats and dogs are presented in the next table, together with pharmacokinetic parameters for
`humans from a study where the highest proposed dose of GHB was administered (2 x 4.5 g, 4 h apart,
`Study OMC-SXB-9).
`
`
`
`At the NOAEL observed in both rat repeat dose toxicology studies, mean AUC and Cmax values were
`approx 0.5 fold the values in humans. After the second 4.5 g dose in humans, the Cmax values for rats
`and humans were approximately the same. At the lowest NOEL for dogs, observed in the dog 52 week
`toxicology study (150 mg/kg/day), the exposure margin for AUC was approximately 0.5 fold the
`values observed in humans. The exposure margin for Cmax was approx 0.5 fold the value following
`the first dose in humans and approximately the same following the second dose in humans. Thus, with
`the exception of Cmax values following the second GHB dose in humans, exposure to GHB in both
`rats and dogs was less than observed for humans at the highest proposed therapeutic dose.
`
`2.4.3 Genotoxicity in vitro and in vivo
`Genotoxicity studies in vitro (bacterial mutation assays in salmonella and E.coli and chromosomal
`aberrations in CHO cells in absence and presence of metabolic activation) and in vivo (rat
`
`Page 6/30
`
`PAR1021
`IPR of U.S. Patent No. 8,772,306
`Page 6 of 30
`
`

`
`micronucleus test) did not identify a cause for concern. Sodium oxybate may be considered as non-
`genotoxic.
`
`2.4.4 Carcinogenicity (with toxicokinetics)
`There is available data on rat and mouse with γ-butyrolactone (GBL which converts in GHB rapidly in
`the body) in two NTP (National Toxicology Program) studies and with sodium oxybate in rats
`(applicant-sponsored 2-year study). In the rat carcinogenicity study the active substance was
`administered as sodium oxybate (maximum dose 1000 mg/kg). In the National Toxicology Program
`mouse carcinogenicity study (CAS No. 96-48-0) the active substance was administered as GBL
`(maximum dose 525 mg/kg). Provided that no toxicokinetic data was included in the NTP studies,
`single dose and 14 day dose bridging studies were conducted by the applicant in both rat and mouse to
`estimate the exposure in these species following administration of sodium oxybate or γ-butyrolactone
`(GBL) in support of these carcinogenicity studies. The exposures in rat and mouse at the doses
`corresponding to the maximum doses in the respective carcinogenicity studies was compared to the
`exposures in humans at the maximum recommended dose of 9g, administered as two 4.5g doses 4
`hours apart (OMC-SXB-9). The animal/human exposure ratios for Cmax were 2.60 and 2.76 for mouse
`and rat, respectively. Similarly the animal/human exposure ratios for AUC were 1.21 and 1.64 for
`mouse and rat, respectively. It is not clear whether data from short-term administration (up to 14 days)
`is informative regarding the exposure of animals at the end of the study, provided that accumulation
`seemed to occur at least at high doses, as suggested by comparative toxicokinetic analysis of the
`values obtained in repeated dose studies in rats and dogs. Underestimation of exposure may therefore
`be given in the bridging studies.
`GBL has been classified by NTP as non-carcinogenic in rats and equivocal carcinogen in mice, due to
`slight increase of pheochromocytomas which was difficult to interpret due to high mortality in the
`high dose group. With GBL a non-significant increase of hyperplasia of adrenal medulla was observed
`also in rats, which poses the possibility of a drug-related effect in both species. Decreased incidence of
`several neoplasm types in mice (hepatocellular tumors) and rats (mammary fibroadenomas)
`administered with GBL were also observed in the NTP study.
`In rats, both sodium oxybate (applicant-sponsored study) and GBL were classified as non-
`carcinogenic. In the oxybate study 2/50 pituitary carcinomas were observed in high-dose female rats
`compared to 0/50 in all other groups including controls. However, this finding was of doubtful
`statistical significance and the incidence was at the upper bound of historical controls. Since GHB was
`non-genotoxic and the pituitary carcinomas in the rat were of marginal statistical significance, there is
`sufficient information to assume that Xyrem is unlikely to be a potential carcinogen in humans. The
`safety ratios calculated against predicted human exposure were still low but new studies do not seem
`necessary as no concern has been raised from the available toxicological data.
`
`2.4.5 Reproductive and developmental studies
`GHB had no effect on mating, general fertility or sperm parameters and did not produce embryo-foetal
`toxicity in rats exposed to up 1000 mg/kg/day GHB. As there are no PK data in pregnant animals, the
`corresponding exposure margin, calculated from non-pregnant animals is 1.64 times the human one.
`In rabbits, foetotoxicity was slight and did not reach statistical significance. The only notable
`abnormality was hydrocephalus in two foetuses from the same litter in a mid-dose female. Based on
`historical data provided and taking into consideration that there were no similar findings in any other
`litter or dose group, this finding was considered to be unrelated to treatment. The rabbit study included
`toxicokinetics. The highest Cmax recorded was 454 µg/ml, which is 3-fold higher than the predicted
`human value.
`GHB had no adverse effects on the F0 animals in a conventional Segment III study in rats, except for
`an increase in the incidence of post-dose sedation, low bodyweight gain and reduced food
`consumption in the high-dose group. Perinatal mortality was increased and mean pup weight was
`decreased during the lactation period in high-dose F1 animals. Though not dose related, a relationship
`of these mortalities with the treatment cannot be ruled out. GHB is not to be recommended during
`pregnancy or breast-feeding and this is reflected in the SPC.
`
`Page 7/30
`
`PAR1021
`IPR of U.S. Patent No. 8,772,306
`Page 7 of 30
`
`

`
`2.4.6 Other toxicity studies
`Immunotoxicity studies were not carried out. However, there were no signals of direct immunotoxicity
`in any of the repeat-dose toxicity studies.
`Dependance studies
`GHB is a known substance of abuse and in the United States the use of Xyrem is subject to a rigid risk
`management program. The applicant did not conduct specific animal tests for dependence. However, a
`review of the published literature was carried out to collect information on the effects of GHB in
`models of drug discrimination, self-administration and tolerance. Drug discrimination studies show
`that GHB produces a unique discriminative stimulus that in some respects is similar to that of alcohol,
`morphine and certain GABA-mimetic drugs. However, the characteristics of these effects differ with
`dose, suggesting the involvement of multiple receptor systems with varied affinities for GHB. Self-
`administration studies in rats, mice and monkeys have produced conflicting results, whereas tolerance
`to GHB as well as cross-tolerance to alcohol has been clearly demonstrated in rodents.
`The data generated in the file does not allow evaluation of the potential of sodium oxybate to induce
`withdrawal phenomena. The literature review suggest that mild effects could occur in mice and rats
`after frequent daily administration and in primates after continuous long term administration (more
`than 30 days) of 750mg/kg/day dose. The potential for withdrawal seems therefore to exist, though
`limited. This is reflected in the SPC and will be further evaluated clinically as a post-marketing
`commitment.
`2.5 Environmental risk assessment
`A formal environmental risk assessment has not been carried out. It is nevertheless agreed that sodium
`oxybate is unlikely to pose any perceivable risk to the environment.
`
`2.6 Discussion on the non-clinical aspects
`The precise mechanism by which sodium oxybate produces an effect on cataplexy is unknown,
`however sodium oxybate is thought to act by promoting slow (delta) wave sleep and consolidating
`night-time sleep. The cause of human narcolepsy and cataplexy is, as yet, unknown. Recent evidence
`points to the loss of hypocretin-containing neurons, possibly due to autoimmune attack, as a likely
`cause. Animal models of cataplexy and narcolepsy are continuing to be developed, but the effects of
`GHB in these models, have yet to be investigated.
`In addition to its sedative properties, GHB is a potent CNS depressant and may increase growth
`hormone secretion. The potential for acute respiratory depression after reintroduction of treatment in
`patients who might have developed tolerance is mentioned in the SPC (4.2) as well as a warning in
`case of concomitant administration with other CNS depressant, and especially alcohol. (SPC, 4.4 and
`4.5)
`The interactions resulting from the stimulation or inhibition of GHB dehydrogenase, namely with
`anticonvulsivant drugs and L-dopa are considered to be potentially relevant in the clinic and are
`mentioned in the SPC.(section 4.5 and 5.3).
`The bioavailability in rats is about 50-80% and the clearance is rapid (T½ ≈1h) with extensive
`biotransformation which is not P450 dependant. The kinetic is non-linear, due to saturable
`mechanisms of both absorption and elimination resulting in reduced Cmax and increased AUC from
`linearity with increasing doses. As GHB is metabolised by γ hydroxybutyrate (GHB) dehydrogenase,
`there is a potential interaction with drugs that inhibit this enzyme. This is reflected in the SPC
`(sections 4.5 and 5.3)
`In repeat-dose toxicity, treatment-related clinical signs were mainly related to sedation. Safety
`margins, based on body weight changes and salivary gland atrophy, are low or non-existent. This is in
`accordance with the frequent occurrence in humans of adverse effects such as nausea, anorexia and
`parasympathetic disorders (blurred vision, enuresis and sweating).
`GHB is non-genotoxic and not considered to present a carcinogenic risk to humans.
`In reproductive toxicity studies, foetoxicity was observed in rats and rabbits. The use of GHB is not
`recommended during pregnancy or breast-feeding and this is reflected in the SPC.
`
`Page 8/30
`
`PAR1021
`IPR of U.S. Patent No. 8,772,306
`Page 8 of 30
`
`

`
`Tolerance to GHB as well as cross-tolerance to alcohol has been clearly demonstrated in rodents. The
`potential of GHB to induce withdrawal phenomena, demonstrated in animal models, is low, and the
`relevance for humans will be monitored i